SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY NUMBER 3 EDITOR j. Thomas Dutro, jr. Paleozoic Perspectives A Paleontological Tribute to G. Arthur Cooper ISSUED FEB 211971 SMITHSONIAN INSTITUTION PRESS CITY OF WASHINGTON I971 SERIAL PUBLICATIONS OF THE SMITHSONIAN INSTITUTION The emphasis upon publications as a means of diffusing knowledge was expressed by the first Secretaiy of the Smithsonian Institution. In his formal plan for the Institution, Joseph Henry articulated a program that included the following statement: "It is proposed to publish a series of reports, giving an account of the new discoveries in science, and of the changes made from year to year in all branches of knowledge not strictly professional." This keynote of basic research has been adhered to over the years in the issuance of thousands of titles in serial publications under the Smithsonian imprint, commencing with Smithsonian Con- tributions to Knowledge in 1848 and continuing with the following active series: Smithsonian Annals of Flight Smithsonian Contributions to Anthropology Smithsonian Contributions to Astrophysics Smithsonian Contributions to Botany Smithsonian Contributions to the Earth Sciences Smithsonian Contributions to Paleobiology Smithsonian Contributions to Zoology Smithsonian Studies in History and Technology In these series, the Institution publishes original articles and monographs deal- ing with the research and collections of its several museums and offices and of professional colleagues at other institutions of learning. These papers report newly acquired facts, synoptic interpretations of data, or original theory in specialized fields. Each publication is distributed by mailing lists to libraries, laboratories, institutes, and interested specialists throughout the world. Individual copies may be obtained from the Smithsonian Institution Press as long as stocks are available. S. DILLON RIPLEY Secretary Smithsonian Institution Official publication date is handstamped in a limited number of initial copies and is recorded in the Institution's annual report, Smithsonian Year. U.S. GOVERNMENT PRINTING OFFICE WASHINGTON • 1971 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C., 20402 - Prke $4 Foreword This collection of papers was solicited from colleagues and students of G. Arthur Cooper as a Festchrift in recognition of his profound influence on the study of brachiopods and their biostratigraphic application to geologic problems, especially in the Paleozoic Era. Dr. Cooper initiated a period of growth in both the research staff and the National Collections of Fossils that guided the Smithsonian Institution to its present position of leadership in paleontological research. His own superb studies of fossil and living brachiopods are unsurpassed in breadth and paleontological significance. Mainly through his efforts, the Smithsonian has acquired an outstanding reference collection of invertebrate fossils that is the envy of the scientific community. It is a pleasure to add my own heartfelt appreciation for Cooper's contributions to the science of paleontology and to the Smithsonian Institution. Porter M. Kier, Chairman Department of Paleobiology, United States National Museum of Natural History, Smithsonian Institution m An Appreciation of G. Arthur Cooper G. Arthur Cooper, geologist and curator, is an old-fashioned scientist. I mean "old-fashioned" in the very best sense of the word, but perhaps "traditional" is the more appropriate word. One dictionary defines a traditionalist as "one who passes on an accumulated store of knowledge to his successors." This definition sums up much of Cooper's career. In one way or another, he has devoted his life to the accumulation of geological and paleontological knowledge and to passing this remarkable store of information on to his colleagues. Some of this has been accomplished through a magnificent series of publications. But much of it has been on a less formal basis— through joint field experiences, discussions, talks at meetings, and the brachiopod seminars at the United States National Museum. Although he was never formally a teacher, there is scarcely a researcher in brachiopods in the western world who has not acquired much of his lore from Cooper. On at least three occasions during the 1950s Coop conducted a series of seminars on brachiopodology, each of which was attended by six to eight persons. At these informal sessions he unselfishly passed on to these young people much of his accumu- lated knowledge on the collection, preparation, and interpretation of brachiopods and their stratigraphic significance. He has taught more paleontologists more about fossils than many academicians at die universities. While a certain amount of his knowledge was acquired from his predecessors, a large part of Cooper's contribution to geology is of his own making. His taxonomic studies represent a quantum jump beyond those of earlier workers. Further, his biostratigraphic methods and facies studies were pioneering efforts that only now are being understood and applied by a younger generation. G. A. Cooper was born 9 February 1902, at College Point, New York, and he has retained a flavor of that part of New York ever since. Who hasn't detected the Brookly- nese creeping into his speech now and then? And he is an avid baseball fan to this day. An early urge to collect and identify things was satisfied with mineral collecting. His love for minerals remains, and his support of the development of the Smithsonian Institution's mineral and gem collection reflects this continuing interest. Fossils became a passion with Coop during his undergraduate days at Colgate. The great wealth of geologic material in the Hamilton beds near the college attracted bis attention early, and he spent much of his spare time poking about the quarry on the Colgate campus and in the creeks and glens nearby. This familiarity with Hamilton rocks and fossils led inevitably to the classic treatment of Hamilton stratigraphy that was Cooper's thesis at Yale. His meticulous detail regarding the physical stratigraphy and biostratigraphy of these Middle Devonian beds has not been approached in any later work. The recent New York State Correlation Chart accepts in toto die concepts of that epoch as developed by Cooper nearly 40 years ago. Field work and fossil collecting remain a large part of Coop's life even today. Nearly every year has found him in die field, measuring, observing, collecting, inter- SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY preting—always adding to his own subjective "feel" for the geology and augmenting the collections of the United States National Museum. During the first ten years of his work at the Smithsonian, Cooper wrote short summaries of his field experiences yearly in the Smithsonian's "Explorations and Field Work" series, an interesting publication that expired after 1940, doubtless a victim of the war years. His first paper was aptly titled "Dry-Dredging in East-Central New York." During these ten years Cooper was collecting in New York, the Gaspe, the Appalachians, the Midwest, Michigan and Ontario, Nevada, New Mexico, Utah and, inevitably, western Texas. During the past 20 years he has revisited these areas time and again with many associates. The first results of etching for silicified fossils were reported in 1940. Of this trip, Cooper wrote: "Two weeks were spent in the Glass Mountains searching for blocks suitable for etching. Many were obtained, and a fine collection of hitherto rare species is confidently anticipated when the limestone lumps are finally decalcified." His wildest expectations have been exceeded a hundred-fold. After many revisits to the Glass Mountains, the Permian brachiopod collection has been augmented by tens of thou- sands of specimens, and I understand that Coop confidently looks forward to one last trip to pick up a few rare species in the spring of 1970. The residents of Marathon are so accustomed to these "last trips" to western Texas that diey greet him each year with "Well, Dr. Cooper, it's good to see you back again for die last time!" Cooper went to Yale where he received his doctor's degree in 1929. Here he came under the influence of two great brachiopod workers, Charles Schuchert and Carl Dunbar. Thus, he is a direct intellectual descendant of the first great American brachiopodologist, James Hall. This experience turned him from fossil molluscs, his first love at Colgate, to brachiopods, and we are all the beneficiaries of this good fortune. The breadth of his brachiopod research is unequaled by any other paleon- ologist, past or present. He knows, intimately, brachiopods through their entire geologic range, and he has published definitive monographs on many taxonomic groups—from orthoids to productoids and rhynchonelloids to terebratuloids. He is a remarkable collector and an unexcelled preparer of these beasts. I will never forget the magnificent demonstration of needlework diat he gave at one of his seminars. Taking a terebratuloid with a chalky filling, he skillfully and quickly removed a part of the shell, excavated the interior, and laid bare the loop that was widiin. It looked remarkably easy until the rest of us tried our hand at it. This care and precision crops out in all his work. He is the curator par excellence. Always on the lookout for ways to increase the collections of the museum, he takes advantage of every oppor- tunity to add new material. Each field trip is an experience in the fine art of fossil collecting. No chance is lost, both going to and returning from a major project area, to enrich the museum's collections. I remember, especially, a trip to New Mexico, when our main goal was to collect Upper Devonian brachiopods; but there were one or two places where we just had to stop to get a few rare shells. And before we were finished we had made fine collections of Upper Ordovician from central New Mexico, of Pennsylvanian and Permian from various parts of western Texas, of Mississippian from Lake Valley, New Mexico, of Upper Pennsylvanian and Cretaceous from north-central Texas, and of Middle Ordovician from the Arbuckles. The broader implications of Cooper's research, however, are never obscured by the immediate collecting, sorting, labeling, and curating aspects of his daily work. The Devonian correlation chart, published by the GSA in 1942, is largely a result of NUMBER 3 vJj his efforts. His concepts of facies, zonation, and stages went into the chart, and it is only a matter of time before all geologists realize the true significance of this contribu- tion. Characteristically, Cooper is busily revising that chart—in his spare time at home so as not to interfere with his "regular" work. The Chazyan brachiopod monograph of 1956 also incorporates large amounts of stratigraphic information, much of which was developed by Cooper, and his non- relative Byron Cooper, in the central Appalachians. It will be another decade before all these data are assimilated into the working geologic backgrounds of even the more knowledgeable stratigraphers and generalists. Probably no one else in North America knows as much as G. Arthur Cooper about the geology of three great Paleozoic Systems—the Ordovician, the Devonian, and the Permian. His paleoecological interpretations of the Middle Devonian of the eastern and central United States, in the Geological Society of America's Paleoecology Memoir are models of cautious extrapolations into the distant past, and they have not been surpassed in clarity and insight into ancient geography and fossil environments. Comprehensive studies of large groups of brachiopods regularly have received a large part of Cooper's attention. The first of these was the orthoid and pentameroid book that he wrote with Professor Schuchert. There followed studies of Cambrian and Ordovician shells (with Ulrich), the Tully fauna (with Stewart Williams) the productoids (with Miss Muir-Wood) and the magnificent Chazyan brachiopod biostratigraphic study. And, as is well known, Cooper is now completing his West Texas Permian monograph (with R. E. Grant), a publication that will be unparalleled in the history of brachiopod research. Thousands of photographs will illustrate this magnificent taxonomic work. Its publication will be the crowning event of an illustrious paleontologic career. A major classification of brachiopods, first put forth in the brachiopod chapter of the Shimer and Shrock book Index Fossils, is reflected in all of Cooper's taxonomic work and is used today by practically all brachiopod students. It is characteristic that Cooper never intended this to be a formal classification—merely a working hypothesis by which to test continuing systematic research. Cooper is not just a collector, classifier, monographer, and stratigrapher. He is a curator of unmatched skill at the Smithsonian. He has worked assiduously for more than 35 years to build what is undoubtedly the greatest reference collection of brachio- pods in the world. But his interests do not stop there. He is keenly appreciative of all invertebrate fossils. His own research has led him to studies of pelecypods, echino- derms, and trilobites; and he is nearly as thrilled at finding new and beautiful corals or bryozoans as he is at describing new brachiopod genera. To have shared the experience of a field collecting expedition with Cooper is to have been very fortunate indeed. His enthusiasm and quiet good humor are infectious. He is at ease with all sorts of people: scientists, ranchers, laborers, teachers, oilmen, and spinster amateur collectors. No geologist who has been closely associated with him doubts for one moment that here, indeed, is a great and good man. This recitation of what Cooper has been doing in the last 40 years barely scratches the surface of the man himself. As head curator of geology and chairman of paleo- biology at the United States National Museum, he quiety distinguished himself as an administrator of scientific research. I am certain that he would be the last person to agree with this evaluation, but, after observing for a number of years the developing careers and scientific contributions of the people he gathered around him in the Museum, I can attribute much of the success of this paleontological venture only to SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Cooper's administrative skill. He is a kind, generous, thoughtful man who, as he waged the paper-work campaigns over the years, kept uppermost in his mind the best interests of his staff, paleontology in general, and the Smithsonian in particular. Nevertheless, his personal interests do not center solely on paleontology and geology. He is an avid student of history, having read widely about early American history, the Civil War, the opening of the West, and explorations in all parts of the globe. His interests range through the whole field of natural history. Many are aware of his long-standing love affair with the National Zoological Park. He and his wife are regular visitors there, and they follow the fortunes of the various animals almost as carefully as do the keepers themselves. He is also an afficionado of flowering plants and can rapturize over the stunning displays of blooming cacti in the southwestern United States. His photographic skill has caught much of this fascination with natural wonders wherever he has traveled. Of course, the most constant objects of his lenses are his brachiopods. Where else would one see a man, at this advanced stage in a distinguished research career, patiently taking hundreds of pictures a day to illustrate a taxonomic paper? Many honors have come Cooper's way during the course of his career at the Smithsonian. He received, in 1953, an honorary doctor of science degree from Colgate, his alma mater; he is a fellow of the Geological Society of America; and is a past- president of the Paleontological Society, the Geological Society of Washington, and the Paleontological Society of Washington. In 1958 he received one of the National Academy's highest honors, the Mary Clark Thompson medal, and in 1964 he became the second recipient of the Paleontological Society medal in recognition for his outstanding contributions to paleontology. All the contributors to this book have been brushed by the genius of this man. Each of us has gone to Cooper to learn of brachiopods and stratigraphy and has come away a much richer and wiser scientist. My own life has been the fuller for having known G. Arthur Cooper—geologist, paleontologist, curator, natural philosopher, and friend. J. THOMAS DUTRO, Jr., United States Geological Survey, Washington, D.C. The Publications of G. Arthur Cooper 1930. Fossil Fauna of the Marl Deposits in the Vicinity of New Milford (Connecticut), in R. F. Flint, The Glacial Geology of Connecticut. Connecticut State Geological and Natural History Survey Bulletin 47:238-259, plates 2-4A. 1930. Stratigraphy of the Hamilton Group of New York. Parts 1 and 2. American Journal Science, series 5, 19:116-134, 214-236, figures 1-6. 1930. [with Charles Schuchert] Upper Ordovician and Lower Devonian Stratigraphy and Paleontology of Perce, Quebec. American Journal Science, series 5, 20:161-174, 4 figures. 1930. New species from the Upper Ordovician of Perce. American Journal Science, series 5, 20:265-288, 365- 392, plates 2-5. 1930. The Brachiopod Genus Pionodema and Its Home- omorphs. Journal of Paleontology 4(4) :369-382, plates 35-37, 1 figure. 1931. Lepidechinoides Olsson, a Genus of Devonian Echinoids. Journal of Paleontology 5(2) : 127-142, plates 18, 19, figures 1, 2. 1931. A New Species of the Echinoid Lepidesthes. American Journal of Science, series 5, 22:531-538, plate 2, 3 figures. 1931. [with Charles Schuchert] Synopsis of the Brachiopod Genera of the Suborders Orthoidea and Pentamero- idea, with Notes on the Telotremata. American Journal of Science, series 5, 22:241-251. 1931. Concerning the Authorship of the "Preliminary Notice of the Lamellibranch Shells of the Upper Helderberg, Hamilton and Chemung Groups, etc., Part 2." Wash- ington Academy of Sciences Journal, 21(18) : 459— 467, figure 1. 1932. Dry-dredging in Eastern Central New York. Smith- sonian Institution Explorations and Field-Work in 1931, pages 19-22, figures 16-18. 1932. A New Accent in Paleontology. [Abstract.]1 Washing- ton Academy of Sciences Journal, 22(15) :457. 1932. [with Charles Schuchert] Brachiopod Genera of the Suborders Ordioidea and Pentameroidea. Yale Uni- versity Peabody Museum Natural History Memoirs, 4(1) :i-xii, 1-270, plates A, 1-29, figures 1-36. 1933. Collecting Fossils in Gaspe. Smithsonian Institution Explorations and Field-Work in 1932, pages 9-12, figures 8—12. 1933. Stratigraphic Studies in Eastern New York. Smith- sonian Institution Explorations and Field-Work in 1932, pages 13-16, figures 13-17. 1933. Stratigraphy of the Hamilton Group of New York. [Abstract.] Washington Academy of Sciences Journal, 23(8):402. 1933. [and Lawrence Whitcomb]. Salonia, a New Ordovi- cian Brachiopod Genus. Washington Academy Sciences Journal, 23(11) :496-503, figures 1-23. 1933. A Method for the Preparation of Fossils. Science, new series, 77:394. 1933. Evaluation of Internal Characters in the Classification of the Brachiopoda. [Abstract.] Geological Society of America Bulletin, 44(12) : 193-194. 1934. Stratigraphy of the Hamilton Group of Eastern New York. American Journal Science, series 5, 26: 537— 551, figures 1-3; 27:1-12. 1934. Reports on the Collections Obtained by the First Johnson-Smithsonian Deep-Sea Expedition to the Puerto Rican Deep. New Brachiopods. Smithsonian Miscellaneous Collections 91(10): 1-5, plates 1, 2. 1934. [with E. O. Ulrich] Syntrophinella Ulrich and Cooper, New Genus, and Syntrophinella typica Ulrich and Cooper, New Species. Japanese Journal of Geology and Geography Transactions and Abstracts 11 (3,4): 164-165, plate 1. 1935. Oligorhynchia, a New Ordovician (Chazy) Brachi- opod. American Journal of Science, series 5, 29:48— 53, plate 1, figures 1-4. 1935. Young Stages of the Devonian Trilobite Dipleura dekayi Green. Journal of Paleontology, 9(1): 3-5, plate 1. 1935. [and J. Stewart Williams] Tully Formation of New York. Geological Society of America Bulletin, 46(5) : 781-868, plates 54-60, figures 1-7. 1935. [with A. S. Warthin, Jr.] Devonian Studies in Southwestern Ontario and Michigan. Smithsonian Institution Explorations and Field-Work in 1934, pages 13-16, figures 10-11. 1935. [with A. S. Warthin, Jr.] New Formation Names in the Michigan Devonian. Washington Academy of Sciences Journal, 25 (12) : 525—526. 1936. Facies Relationships in the Hamilton Group of New York. [Abstract.] 16th International Geological Con- gress (1933) Report, 2:1106. 1936. Studies of Middle Devonian Rocks in the Mid-West. Smithsonian Institution Explorations and Field-Work in 1935, pages 9-12, figures 8-9. 1936. New Cambrian Brachiopods from Alaska. Journal of Paleontology, 10(3) :210-214, plate 26. 1936. [and A. S. Wardiin, Jr.] New Correlations of Hamil- ton Rocks. [Abstract.] Geological Society of America Proceedings, 1935, pages 376-377. ix SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY 1936. [with E. O. Ulrich] New Silurian Brachiopods of the Family Triplesiidae. Journal of Paleontology, 10(5) :331-347, plates 48-50, figure 1. 1936. [and C. H. Kindle] New Brachiopods and Trilobites from the Upper Ordovician of Perce, Quebec. Jour- nal of Paleontology, 10(5): 348-372, plates 51-53, figures 1-10. 1936. [with E. O. Ulrich] New Genera and Species of Ozarkian and Canadian Brachiopods. Journal of Paleontology, 10 (7) : 616-631. 1937. Collecting Devonian Fossils in the Mid-West. Smithsonian Institution Explorations and Field-Wo-rk in 1936, pages 15-18, figures 11-12. 1937. [with E. O. Ulrich] Cambrotrophia, New Name for Eostrophia Ulrich and Cooper, Not Dall. Journal of Paleontology, 11(1): 78. 1938. Collecting Fossils in Michigan, Pennsylvania, New York and Canada. Smithsonian Institution Explora- tions and Field-Work in 1937, pages 9-12, figures 9-10. 1938. [and P. E. Cloud] New Devonian Fossils from Cal- houn County, Illinois. Journal of Paleontology, 12 (5) : 444-460, plates 54-55. 1938. [with E. O. Ulrich] Ozarkian and Canadian Brachi- opoda. Geological Society of America Special Paper, 13: i—viii-f-1-323, plates 1-58, figures 1-14. 1938. Brachiopod Ecology and Paleoecology. National Re- search Council, Division Geology and Geography Re- port Committee on Paleoecology, 1936-1937', pages 26-53. 1939. Collecting Fossils in the Catskills of New York. Smith- sonian Institution Explorations and Field-Work in 1938, pages 29-32, figures 26-27. 1940. [with Josiah Bridge] Collecting Fossils in Utah, Nevada, Texas and the Mid-West. Smithsonian In- stitution Explorations and Field-Work in 1939, pages 9-16, figures 8-18. 1940. Collecting Ordovician Fossils in the Southern Ap- palachians. Smithsonian Institution Explorations and Field-Work in 1939, pages 17-20, figures 19-22. 1941. New Devonian Stratigraphic Units. Washington Academy of Sciences Journal, 31(5) : 179-181. 1941. [and A. S. Warthin, Jr.] New Middle Devonian Stratigraphic Names. Washington Academy of Sciences Journal, 31(6) : 259-260. 1941. Geologizing in Texas and Tennessee. Smithsonian Institution Explorations and Field-Work in 1940, pages 9-12, figures 11, 12. 1941. Facies Relations of the Middle Devonian (Hamilton Group) Along the Catskill Front. [Abstract.] Geo- logical Society of America Bulletin, 52(12) : 1893. 1941. [with E. O. Ulrich] Chazyan and Related Brachi- opods. [Abstract.] Geological Society of America Bulletin, 52(12): 1976. 1942. [with E. O. Ulrich] New Genera of Ordovician Brachiopods. Journal of Paleontology, 16(5) :620- 626, plate 90. 1942. [and A. S. Warthin, Jr.] New Devonian (Hamilton) Correlations. Geological Society of America Bulletin, 52(6):873-888, illustrated. 1942. [with committee members] Correlation of the Devo- nian Sedimentary Formations of North America (Chart No. 4). Geological Society of America Bulle- tin, 53(12): 1729-1794, chart. 1942. Ecology of Some Permian Brachiopods. National Re- search Council Division Geology and Geography Re- port, Committee on Marine Ecology as Related to Paleontology, Annual Report, appendix N, pages 36-37. 1942. New Genera of North American Brachiopods. Wash- ington Academy of Sciences Journal, 32(8) : 228-235. 1943. [with A. S. Warthin, Jr.] Traverse Rocks of Thun- der Bay Region, Michigan. American Association of Petroleum Geologists Bulletin, 27(5) :571-595, fig- ures 1—8. 1943. Obituary of Charles Schuchert (1858-1942). Wash- ington Academy of Sciences Journal, 33(11) : 352. 1944. Phylum Brachiopoda, in H. W. Shimer and R. R. Shrock, Index Fossils of North America, pages 277— 365, plates 105-143. New York: John Wiley and Sons, Inc. 1944. Remarks on Correlation of Devonian Formations in Illinois and Adjacent States, in Symposium of Devo- nian Stratigraphy. Illinois Geological Survey Bulletin, 68:217-222, chart. 1944. [with A. S. Warthin, Jr.] Middle Devonian Subsur- face Formations in Illinois. American Association of Petroleum Geologists Bulletin, 28(10) : 1519-1527, figures 1—5. 1944. New Species of Brachiopods from the Devonian of Illinois and Missouri. Journal of Paleontology, 19(5) : 479-489, plates 63-64. 1946. [with B. N. Cooper] Lower Middle Ordovician Stratigraphy of the Shenandoah Valley, Virginia. Geological Society of America Bulletin, 57(1) :35- 114, plates 1-3. 1946. [with J. B. Knight] Permian Studies at the Smith- sonian Institution, Washington. Science, 104(2688) : 15-16. 1946. [and A. R. V. Arellano] Stratigraphy near Caborca, Northwest Sonora, Mexico. American Association of Petroleum Geologists Bulletin, 30(4) :606-611, figure 1. 1946. Silicified Fossils (Permian, Texas, Ordovician, Virginia. [Abstract.] Virginia Academy of Sciences Proceedings, 1945-46, page 87. 1947. [and Cecil Haldane Kindle] Remopleurides sinclairi, New Name. Journal of Paleontology, 21(1) : 76. 1948. Annotated Bibliography of Brachiopod Ecology. National Research Council Report, Committee on u Treatise of Marine Ecology and Paleoecology, 1946- 1947,7:38-44. 1948. Annotated Bibliography of Brachiopod Paleoecology. National Research Council Report, Committee on a Treatise of Marine Ecology and Paleoecology, 1946- 1947, 7:45-53. NUMBER 3 XI 1948. A New Genus of Brachiopoda from the Longview Limestone of Virginia. Harvard College Museum of Comparative Zoology Bulletin, 100(6) :463-474, illustrated. 1950. Permian faunas of the Glass Mountains, Texas, and Their Environment. New York Academy of Sciences Transactions, series 2, 12(3) : 80—81. 1951. New Brachiopods from the Lower Cambrian of Vir- ginia. Washington Academy of Sciences Journal, 41(1) :4-8, illustrated. 1951. [and Helen M. Muir-Wood] Brachiopod Homonyms Washington Academy of Sciences Journal, 41(6):195-196. 1951. [with H. M. Muir-Wood] A New Species of the Jurassic Brachiopod Genus Septirhynchia. Smith- sonian Miscellaneous Collections, 116(6): 1-6, plates 1,2. 1951. Unusual Specimens of the Brachiopod Family Iso- grammidae. Journal of Paleontology, 26(1) : 113— 119, plates 21-23. 1951. Cambrian Stratigraphy and Paleontology near Caborca, Northwestern Sonora, Mexico. Introduc- tion and Stratigraphy [with A. R. V. Arrellano]: Brachiopoda. Smithsonian Miscellaneous Collections, 119(1) : 1-18,36-48, plates 1-5, 11-13. 1951. New and Unusual Species of Brachiopods from the Arbuckle Group in Oklahoma. Smithsonian Miscel- laneous Collections, 117 (14) : 1-35, figure 1, plates 1-4. 1951. [and Alwyn Williams] Significance of the Strati- graphic Distribution of Brachiopods, in L. G. Henbest (editor), Distribution of Evolutionary Explosions in Geologic Time—a Symposium. Journal of Paleontol- ogy, 26(3) :326-337, figures 1-12. 1953. Permian Fauna at El Antimonio, Western Sonora, Mexico. Smithsonian Miscellaneous Collections, 119(2): 1-13,21-82,92-106, plates 1, 4-24. 1953. Permian Faunal Studies in the Glass Mountains, Texas. West Texas Geological Society Spring Field Trip, May 1953, pages 70-76. 1954. Unusual Devonian Brachiopods. Journal of Paleon- tology, 28(3) : 325-332, plates 36, 37, figures 1-5. 1954. Recent Brachiopods, Bikini and Nearby Atolls, Marshall Islands. United States Geological Survey Professional Paper, 260-G: 315-318, plates 80, 81. 1954. Brachiopoda Occurring in the Gulf of Mexico, in P. S. Galtsoff (coordinator), Gulf of Mexico, Its Origin, Waters, and Marine Life. United States Fish and Wildlife Service Fishery Bulletin, 89:363-365. 1955. New Genera of Middle Paleozoic Brachiopods. Journal of Paleontology, 29(1) :45-63, plates 11-14, figure 1. 1955. New Brachiopods from Cuba. Journal of Paleon- tology, 29(l):64-70, plate 15. 1955. [and F. G. Stehli] New Genera of Permian Brachio- pods from West Texas. Journal of Paleontology, 29(3) : 469-474, plates 52-54. 1955. New Cretaceous Brachiopoda from Arizona. Smith- sonian Miscellaneous Collections, 131 (4): 1—18, plates 1-4. 1955. Faunal Suites of the Appalachian Middle Ordovician. [Abstract.] Geological Society of America Bulletin, 66(12):1686-1687. 1956. New Pennsylvanian Brachiopods. Journal of Paleon- tology, 30(3) :521-530, plate 61, figure 1. 1956. Chazyan and Related Brachiopods. Smithsonian Mis- cellaneous Collections, volume 127:part 1 (text), pages i-xvi+1-1024; part 2 (plates), pages 1025- 1245, plates 1-269. 1956. A New Upper Canadian Fauna from a Deep Well in Tennessee. Journal of Paleontology, 30(1) : 29— 34, plate 5. 1957. Loop Development of the Pennsylvanian Brachiopod Cryptacanthia. Smithsonian Miscellaneous Collec- tions, 134(3) : 1-18, plates 1, 2. 1957. Tertiary and Pleistocene Brachiopods of Okinawa, Ryukyu Islands. United States Geological Survey Professional Paper, 314-A:i-iv+l-20, plates 1-5. 1957. Study of the Wolfcamp and Related Faunas of the Glass Mountains, Texas. Society of Economic Paleon- tologists and Mineralogists, Permian Basin Section, April 1957, Guidebook, pages 8-12. 1957. Paleoecology of Middle Devonian of Eastern and Central United States, in H. S. Ladd (editor), Paleoecology. Geological Society of American Memoir, 67:249-277, plate 1, figures 1, 2. 1957. Brachiopods—Annotated Bibliography, in H. S. Ladd (editor), Paleoecology. Geological Society of America Memoir, 67: 801-804. 1957. Asterozoa of the Paleozoic—Annotated Bibliography, in H. S. Ladd (editor), Paleoecology. Geological Society of America Memoir, 67:973—974. 1957. Echinoids of the Paleozoic—Annotated Bibliography, in H. S. Ladd (editor), Paleoecology. Geological Society of America Memoir, 67:979-980. 1957. Memorial to Edwin Kirk (1884-1955). Geological Society of America Proceedings 1956, pages 141-146. 1957. Permian Brachiopods from Central Oregon. Smith- sonian Miscellaneous Collections, 134(12) : i—iv-f- 1-79, plates 1-12. 1957. [and P. B. King] Guidebook, 1957 Fall Field Trip, Glass Mountains (Texas), October 25—26, 1957, pages 1-39, illustrated. 1958. Presidential Address: The Science of Paleontology. Journal of Paleontology, 32(5) : 1010-1018, figure 1. 1959. Genera of Tertiary and Recent Rhynchonelloid Brachiopods. Smithsonian Miscellaneous Collections, 139(5) :i-iv+l-90, plates 1-22. 1960. Correction of Brachiopod Names. Journal of Paleon- tology, 34(3):601. 1960. [with H. M. Muir-Wood] Morphology, Classifica- tion, and Life Habits of Productoidea (Brachiopoda). Geological Society of America Memoir, 81 :i—xi4- 1-447, plates 1-135. 1962. Pseudopunctate Brachiopods. [Abstract] Geological Society America Special Paper, 68:155-156. 1962. [and R. E. Grant] Torynechus: New Name for Permian Brachiopod Uncinuloides King. Journal of Paleontology, 36(5) : 1128-1129. SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY 1964. [and R. E. Grant] Permian Brachiopods of the Glass Mountains, West Texas. [Abstract.] Geological So- ciety of America Special Paper, 76:36-37. 1964. [and R. E. Grant] New Permian Stratigraphic Units in Glass Mountains, West Texas. American Associa- ciation of Petroleum Geologists Bulletin, 48(9) : 1581-1588, figures 1, 2. 1965. [and H. B. Whittington] Use of Acids in Preparation of Fossils. Handbook of Paleontological Techniques, pages 294-300. San Francisco: W. H. Freeman and Company. 1966. [and Thomas Phelan] Stringocephalus in the De- vonian of Indiana. Smithsonian Miscellaneous Collec- tions, 151(1): 1-20, plates 1-5. 1966. [and R. E. Grant] Permian Rock Units in the Glass Mountains, West Texas. United States Geological Survey Bulletin, 1244-E:E1-E9, illustrated. 1967. [and H. M. Muir-Wood] New Names for Brachiopod Homonyms. Journal of Paleontology, 41 (3) :808. 1968. The Brachiopod Family, Richthofeniacea. [Abstract.] Geological Society America Program and Abstracts, Mexico City Annual Meetings, page 62. 1968. Age and Correlation of the Tully and Cedar Valley Formations in the United States. International Sym- posium on the Devonian System, Calgary, Alberta, 1967 (Proceedings), 2:701-709, illustrated. Calgary, Alberta: Alberta Society of Petroleum Geologists. 1969. [and R. E. Grant] New Permian Brachiopods from West Texas. Smithsonian Contributions to Paleo- biology, 1:1-20, plates 1-5. Contents GENERAL TOPICS Page BYRON N. COOPER : Roles of Fossils in Appalachian Stratigraphy 3 MICHAEL R. HOUSE : The Goniatite Wrinkle-Layer 23 VALDAR JAANUSSON : Evolution of the Brachiopod Hinge 33 ALWYN WILLIAMS: Comments on the Growth of the Shell of Articulate Brachiopods ... 47 PRECAMBRIAN-CAMBRIAN A. J. ROWELL: Supposed Pre-Cambrian Brachiopods 71 ORDOVICIAN GERTRUDA BIERNAT: On Branched Surface Spines in Some Inarticulate Brachiopods ... 83 SHOU-HWA CHUANG: The Morphology and Paleobiology of Trematis elliptopora Cooper (Inarticulata, Brachiopoda) 93 ROUSSEAU H. FLOWER: Cephalopods of the Whiterock Stage 101 ROBERT B. NEUMAN: An Early Middle Ordovician Brachiopod Assemblage from Maine, New Brunswick, and Northern Newfoundland 113 REUBEN JAMES Ross, JR. : A New Middle Ordovician Syntrophopsid Genus 125 H. B. WHITTINGTON: A New Calymenid Trilobite from the Maquoketa Shale, Iowa . . . . 129 SILURIAN ARTURO J. AMOS AND S. NOIRAT: A New Species of Ancillotoechia from the Zapla Formation, NorUiern Argentina 139 THOMAS W. AMSDEN: Triplesia alata Ulrich and Cooper 143 ARTHUR J. BOUCOT: Aenigmastrophia, New Genus, a Difficult Silurian Brachiopod 155 DEVONIAN JEAN M. BERDAN: Some Ostracodes from die Schoharie Formation (Lower Devonian) of New York 161 PRESTON E. CLOUD, JR., AND ARTHUR J. BOUCOT: Dzieduszyckia in Nevada 175 J. THOMAS DUTRO, JR.: The Brachiopod Pentagonia in the Devonian of Eastern United States 181 WILLIAM A. OLIVER, JR. : The Coral Fauna and Age of the Famine Limestone in Quebec . . 193 PAUL SARTENAER: Redescription of the Brachiopod Genus Yunnanella Grabau, 1923 (Rhynchonellida) 203 J. STEWART WILLIAMS: The Beirdneau and Hyrum Formations of North-Central Utah ... 219 ELLIS L. YOCHELSON: A New Late Devonian Gastropod and Its Bearing on Problems of Open Coiling and Septation 231 CARBONIFEROUS JOHN L. CARTER: New Early Mississippian Silicified Brachiopods from Central Iowa . . . 245 MACKENZIE GORDON, JR.: Carlinia, a Late Mississippian Genus of Producddae from the Western United States 257 M. J. S. RUDWICK: The Functional Morphology of the Pennsylvanian Oldhaminoid Brachiopod Poikilosakos 267 Xlil XIV SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY PERMIAN ROBERT M. FINKS: Sponge Zonation in the West Texas Permian Type Section 285 WILLIAM M. FURNISH AND BRIAN F. GLENISTER: Permian Gonioloboceratidae (Am- monoidea) 301 RICHARD E. GRANT: Taxonomy and Autecology of Two Arctic Permian Rhynchonellid Brachiopods 313 FRANCIS G. STEHLI: Tethyan and Boreal Permian Faunas and Their Significance 337 J. B. WATERHOUSE: The Permian Brachiopod Genus Terrakea Booker, 1930 347 GARNER L. WILDE: Phylogeny of Pseudofusulinella and Its Bearing on Early Permian Stratigraphy 363 SUBJECT INDEX 381 INDEX TO GENERA AND SPECIES 385 GENERAL TOPICS Byron JV*. Cooper Roles of Fossils in Appalachi an Stratigrap hy ABSTRACT Fossils play a dominant role in unravelling the Paleozoic stratigraphy of the Appalachians. The stratigrapher and mapping geologist should acquire a good working knowledge of the general sequence of fossil assemblages and should work in association with competent paleon- tologists. Teamwork of this kind can lead to the solution of the many difficult problems related to migrating lithofacies and biofacies, recurrent common rock types and faunas, environmental influences on both lithol- ogies and faunas, correlations of all kinds, and struc- tural interpretations. The doubtful panaceas of guide fossils, key rock types, and faith in any single kind of fossil should be avoided if meaningful geologic results are to be ob- tained. Dominant themes of this analysis include algae and echinoderms as rock-builders, recurring faunas, the failure of the guide fossil concept, and the difficulties in relying solely on lithologic criteria for mapping pur- poses. Comprehensive systematic monographs of all groups of invertebrate fossils are considered indispensa- ble tools in furthering the knowledge of the geology of the Appalachians. Fossil faunas in Appalachian sedimentary formations serve a variety of scientific purposes, some of which are fully appreciated solely by the systematic paleobiolo- gist, others of which are best appreciated by the sea- soned field geologist. It is the purpose of this paper to record some personal observations about fossils that are useful aids in recognizing and mapping Appala- chian sedimentary formations. Generic and specific identities are not of critical importance in these re- marks and will be mentioned only where necessary. Byron N. Cooper, Department of Geological Sciences, Vir- ginia Polytechnic Institute, Blacksburg, Virginia 24061. Some things noted are cited as mere curiosities or enigmas. Many of the details enumerated were passed along to me by two great field teachers—Charles Butts and G. Arthur Cooper. Others have been gathered during the course of 35 field seasons in the Appalachians. All are offered in the hope that they may be useful to other stratigraphers working in the Appalachians. One of the truly inspiring personal experiences I have had was an extended field trip in 1945 down the Appalachians with Carl O. Dunbar, Percy Morris, G. A. Cooper, Dr. Wang of the Geological Survey of China, and Raymond S. Edmundson with whom I was associated for many enjoyable field seasons. I gained enormous respect for the value of fossils by watching Dr. Dunbar examine the fossiliferous rocks from Vir- ginia to Alabama and, largely from his knowledge of the evaluation of shell structures in brachiopods and other fossil groups, methodically and calmly zero-in on relative ages of the beds, often without even mention- ing names of the fossils. I witnessed a master stratig- rapher and paleontologist at work. I recall particu- larly his observations atop Draper Mountain in Pulaski County, Virginia. For years I had taken geologists up Draper Moun- tain on old U.S. Route 11 and tested their awareness of a curiosity—all the rocks on Draper Mountain along the highway are overturned. To an Appalachian ge- ologist unfamiliar with Draper Mountain, the ride up the mountain could fail to convey the message of the rocks. The southeast-dipping Millboro could be mis- taken for Martinsburg, the "overlying" red beds could pass for Juniata, and the overturned Clinch atop the mountain could be taken as part of the normal succes- sion. I used to keep a confidential box score of how visitors responded to Draper Mountain; many of them who could not or did not read the rocks judged the 3 372-386 O—71- section as normal, but not Dr. Dunbar. On getting out of the car on top of Draper Mountain he nosed around the beautiful exposure and remarked, "What a museum piece these exposures are! There's inverted forest bed- ding in that layer, overturned fracture cleavage in that little shale bed, and, look, did you ever see so many Ar- throphycus peering at you wrong side up!" Those brief observations of a master stratigrapher and paleontolo- gist in unfamiliar country summed up the evidence of deformation structures, sedimentary structures, and of the fossils, which pointed to inversion of the section. He had not missed a thing. I realized a little better the resources a stratigrapher has at his disposal after seeing an "old pro" react so quickly and competently to what he saw. Scolithus in Lower Paleozoic Rocks Some fossils are so abundant as to determine the lith- ology of many Appalachian stratigraphic zones. The oldest organic remains in the Appalachian Paleozoics that literally give the rock its peculiar appearance are the so-called Scolithus tubes in the Erwin or Antietam Formations. Perhaps the best display of Scolithus in the Appalachians is in a slightly feldspathic orthoquartzite zone about 80 feet thick, exposed on the broad crest of a southwest-plunging anticline along historic Jor- dan (pronounced "Jerdan") Road between Buena Vista, Rockbridge County, and the Blue Ridge Park- way. The Scolithus are straight tubular structures in which the sand filling is commonly very micaceous and, characteristically, pale green (see Plate 1). The align- ment of the Scolithus in the Erwin beds along Jordan Road is so perfect, the tubes so abundant, and the spacing so close that the rock has fracture cleavage (see Plate 2). The cleavage mullions obviously are the result of the Scolithus but, alas, despite their abundance we do not really know what organism made the tubular markings and casts. These markings are elsewhere commonly segregated in definite zones, as they are along Jordan Road. When "Scolithus-bearing sand- stone" is mentioned in the geologic literature, the spe- cial rock type evident along Jordan Road is generally what the stratigrapher is trying to describe (see Plate 3). Abundance is the contributor to the peculiar ap- pearance of the rocks. Scolithus is a very common fossil found in beds as old as Unicoi but not tremendously abundant at that level except in a fairly pure-sand litho- facies. The occurrences in the Erwin and its equiva- SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY lents to the north (Antietam and Potsdam) and south (Weisner) are especially noteworthy, and Scolithus in true abundance is a valid indication of medial to upper Erwin beds. Algae as Rock Builders The role of algae as a prime lithologic determinant perhaps has been the least appreciated of all lithologic phenomena in Appalachian rocks. The characteristic "ribbon-banded" algal-matte limestones are widely dis- tributed in the lov/er part of the Paleozoic section begin- ning with the Patterson Limestone Member of the Shady Formation and extending up to the early Tren- tonian Witten Limestone. The algal masses in the Pat- terson limestones show the greatest range in size and configuration of any in the entire Appalachian succes- sion. The commonest forms of algal growths are "tufted pancake" or "pizza"-types with strange little digitate, inosculate projections. The "pancakes" are blue-gray limestone, especially bluish in tint only because they are interleaved with equally irregular interstitial fillings of buff-yellow to reddish brown dolomite or dolomitic siltstone. The smaller algae in the Patterson carbon- ates range downward from about ten centimeters in diameter to virtually "eyeball" size, and they are al- most—but never quite—spherical in shape. The smaller the algae, the better structure lamination they show. Some of the Patterson algal beds are chevron-flexed or tufted and provide a very distinctive lithology. Not uncommonly, this variety is associated with coarse re- crystalline dolomites, some of which have acicular and geopetal structures. Nests of coarse white to pink dolo- mite are especially common. The lower and middle Patterson algal limestones are especially noteworthy for two features which prob- ably are better illustrated in Wythe County, Virginia (in the Patterson-Foster Falls-Austinville-Ivanhoe- Cripple Creek-Speedwell sector), than anywhere else. Coarse "eyeball" algal masses with crystalline quartz centers are especialy noteworthy. Probably the most fascinating feature of layered, petaloid algal growths is the structures that resulted from partial desiccation of algal plates or matte during periods of low tide (see Plates 4, 5c). Exposure probably caused the algal plates to contract and to cup upward into saucer-like shapes. Waters evaporating in these natural petri dishes pro- duced radiating or culstered nests of anhydrite crystals which were later converted to dolomite-—probably NUMBER 3 - PLATE 1.—Close-up of Scolithus structures in quartzite, showing influence of the tubes on the lithology; in west foothills of Blue Ridge Mountains, three miles east of Grottoes and 13.5 miles southwest of Elkton, Rockingham County, Virginia. (Photo by W. D. Lowry.) during an early return of water cover. These curious dolomite pseudomorphs after anhydrite are com- monly prismatic in habit, but many of the crystals show flared, fibrous extensions at one or both ends (see Plate 5A) . The crystal clusters, not at all rare, range through at least 100 feet of beds, most of which are partially dolomitized algal-matte limestones (see Plate 5B). In some of the Patterson algal-matte limestones there are, also, curiously zoned oolites with conspicu- ous zonation of radiating fibers of calcite and external rinds of iron-rich dolomite or poorly crystalline silica. Some of the oolites show coalescence into linear chains, and there is a suggestion that these structures may be of algal origin. Some of the Patterson algal-matte limestones are completely dolomitized without loss of original struc- tures. The dolomite identity of the altered limestone is easy to recognize on weathered surfaces, but one needs the acid bottle to distinguish limestone from dolomite on fresh surfaces. The algal-matte limestones with tufted pancake to tubiferous algae are common in the Rome Formation and in lower, equivalent parts of the Conasauga For- mation. At this level the algal-matte limestones have irregular copper-red to burnt-sienna platelets of dolo- mite—evidently an especially iron-rich type of dolo- mite. In the southern Appalachians the Rutledge Lime- stone, Rogersville Shale, and Maryville Limestone con- tain probably the thickest development of algal-matte limestones. These beds are so similar in lithology that an algal limestone in the Rutledge cannot be distin- guished from one in the Maryville. Unlike dolomitized algal limestones in the Patterson member of the Shady, the algal limestones in the Rutledge and Maryville commonly are dolomitized to blue-gray dolomites of even grain and structureless appearance that show ab- solutely no residual algal masses of any kind. Such dolomites, composing the Honaker Dolomite, are about as homogeneous as any in the Paleozoic column. SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY PLATE 2.—Fracture cleavage in Erwin (or Antietam) quartzites, controlled by abundance of Scolithus tubes; from same locality as strata shown in Plate 1. (Photo by W. D. Lowry.) Good algal-matte limestones of Maryville type com- monly contain considerable magnesia or dolomite (up to 20 percent magnesia calculated as MgC03). The blue-gray algal mattes and buff-yellow dolomitic in- terstitial fillings form a color contrast that is very distinctive. There are hundreds of feet of algal-matte limestone in the Maryville Limestone just south of Sunbright or Hortons Summit on U.S. Route 23 in Scott County, Virginia. The nodular polyp-like tu- berosities on the algal plates are distinctive. The lower Rutledge Limestone directly overlying the Rome along U.S. Route 25E between Thorn Hill and Evans Ferry in Tennessee shows somewhat nodular or conspicuously dimpled algal mattes, and the inter- stitial dolomite is a little more yellowish. Such rock is quarried for marble just northeast of U.S. Route 25E. When cut into slabs on the bias to bedding, the blocks accentuate the thickness of the alternate golden yellow and blue-black algal mattes to yield a kind of marble closely resembling the famous Belgian Gold marbles of commerce. Locally, the Rogersville Shale between the Rutledge and Maryville Limestones contains thin zones of algal matte-limestone, as does the Nolichucky Shale that overlies the Maryville in Scott, Russell, and Tazewell Counties, Virginia, and in Sullivan and Hawkins Counties, Tennessee. Such algal-matte limestones generally have dolomite fillings with the copper- red characteristics of the Rome algal-matte lime- stones. Throughout much of eastern Tennessee and the ad- jacent counties of Virginia, and locally as far north- east as New River, the upper part of the Nolichucky Formation is a distinctive magnesian, ribbon-banded, algal-matte limestone possessing somewhat thicker al- gal mattes and thicker interstitial zones of dolomite— each ranging from five-tenths centimeter to two centi- meters in thickness. The chevron structure or "pleat- NUMBER 3 PLATE 3.—"Hob-nail" bedding surface of Scolithus-bearing quartzite showing effect of tubes on the lithology; Erwin Formation, west foothills of Blue Ridge Mountains on Hellgate Creek, four miles east-southeast of Natural Bridge, Rockbridge County, Virginia. (Photo by W. D. Lowry.) ing" of the algal limestone mattes at this stratigraphic level is a common characteristic. These beds are now widely identified as the Maynardville Limestone or Maynardville Limestone Member of the Nolichucky Formation. Generally, but especially in the Copper Ridge outcrop belt, the basal beds of the overlying Copper Ridge Dolomite are laminated carbonates of similar type but they are (almost, if not completely) dolomitized algal-matte limestones. In southeastern belts of the central and southern folded Appalachians, the Elbrook Formation (equiva- lent to the Rutledge-Rogersville-Maryville-Honaker- Nolichucky succession) possesses three especially in- teresting kinds of algae. One type is characteristically displayed in nodular tufts about the size of a man's thumb, and these give a distinctive, nodiform surface to limestone layers. The best exposure of this kind of algal biostrome occurs east of Wytheville, Virginia, in a deep rail cut just southwest of the bridge of the eastbound lane of U.S. Route 11. Similar nodular algal beds occur along Route 81 about three miles northeast of Lexington, Virginia. Far down in the Elbrook is a widespread persistent zone of large-diameter (up to 45 centimeters), plate- shaped algal colonies that grow upward convexly, are finely laminated, and are commonly 7.5 to 15.0 centi- meters thick. These invariably are dolomitized algae that weather a fawn-buff to yellowish gray. Prob- ably the finest exhibit of these algae is in the southwest wall of the Blue Ridge Stone Corpora- tion quarry near the boundary line of Botetourt and Bedford Counties, just north of U.S. Route 460 in Virginia. To my knowledge, these broad plates are found only in the Elbrook. Algal-matte limestones of the distinctive Maryville type also abound in the Elbrook, and doubtless the close similarity of these formations comes from the fact that much of the Elbrook is of Maryville age. SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY PLATE 4.—Algal-matte limestone in Patterson Member of Shady Formation (Lower Cambrian), 1.7 miles east of Foster Falls Station, on Foster Falls Farm of Neuhoff Packing Company, Wythe County, Virginia. Light areas are dolomite. Among the formations hardest to distinguish are the Elbrook and Conococheague of the Upper Cam- brian. I doubt whether the two actually have a dis- crete boundary. The Conococheague algal limestones are commonly algal-mattes with flexed or chevron- shaped plates, generally from three to eight centimeters thick, and with the intervening layers not as yellow or as uniformly dolomitic as in the older Cambrian formations. The beds or ledges generally are much thicker than those in the Elbrook. Some of the algal matte limestones of the Copper Ridge are composed of large discoid, reniform to sub- spherical Cryptozoon masses. They also are prominent in the Conococheague between Verona and Harrison- burg, Virginia, particularly in the vicinity of the Berg- town Klippe of the Pulaski-Staunton fault block. Cauliflower-shaped to spherical algal colonies com- monly called Crytozoon occur as limestone masses and as siliceous or completely silicified masses which weather out of the Conococheague or equivalent Cop- per Ridge Dolomite of the southern Appalachians. These masses occur also in the Beekmantown lime- stones and dolomites, and in cherts in the equivalent Longview, Kingsport, Mascot successions of eastern Tennessee and adjacent areas in Virginia. The diffi- culty with the algal chert masses is that they creep away from their parent ledges and move downhill where they mingle with other cherts; therefore, they have little value as stratigraphic markers in mapping. Copper Ridge algae commonly form nodiform to large subspherical colonies up to 60 centimeters in diameter with the structure preserved in alternate black and gray chert laminae. The larger colonies, when broken open, show incomplete silicification with curved zonal vacuities lined with drusy euhedral quartz crystals. The banded gray-black structures are fairly diagnostic of the Copper Ridge dolomites. Silic- ification of the algal masses may be due more to NUMBER 3 weathering than to a penecontemporaneous action, but the time of silicification is open to question. The Longview Limestone—a persistent zone of par- tially dolomitized micrite with fossil gastropods but lit- tle else in way of invertebrate fossils—contains charac- teristic small algal microlaminated bodies that grew around abandoned gastropod shells. These irregular, billowy, tufted masses of algal limestone are scattered through the limestone layers, and in some places they determine the lithology because they are so numerous. A similar algal occurrence in the Appalachian Paleo- zoic beds is in the much younger Hillsdale (St. Louis) Limestone. The Middle Ordovician limestones contain some algae of Solenopora type; these are especially promi- nent in the Wassum Limestone of the Wilderness Stage in Wassum Valley, four miles northwest of Marion, Smyth County, Virginia, and also are exposed along State Route 16 on the northwest slope of Walker Mountain south of Chatham Hill in the same county. These pinkish encrinal limestones or calcarenites pre- serve the algae as large distinct intraclasts in a matrix of bioclastic spar calcite. The margins of the algal masses suggest they were of a doughy consistency when broken up and transported. The old familiar algal-matte limestones of Maryville type make their highest stratigraphic appearance in the slabby dove-gray limestones of the upper Witten; and along with the fucoid markings of Camarocladia and Buthotrephis they form a distinctive facies that has been identified from the northern Shenandoah Valley southwestward far into eastern Tennessee. In south- west Virginia these Maryville-type algal-matte lime- stones contain dolomitic partings and also fillings of Camarocladia tubes that weather buff-yellow. In my opinion, this is the most definitive lithozone in the Paleozoic succession of the southern Appalachians. The next higher stratigraphic zone in which algae appear in any abundance is in the Hillsdale Limestone, the approximate correlative of the well-known St. Louis Limestone. Little is known about the nature of these algal masses. They are invariably very irregular in shape, possess sharp outlines, and appear to be intra- clasts derived from more localized indigenous growths of algae. Despite the experience of examining many exposures of these beds over a strike length of 100 miles or more, I do not recall ever having seen even a suggestion of a fossil algal colony in place. The peculiar appearance of bluish gray algal micrite masses in the characteristically black Hillsdale Limestone is a lithologic quality worth noting because it is one of the more useful criteria for recognizing St. Louis age beds in the field. The familiar Lithostrotionella colo- nial coralla—both the "mammilare," free-standing corallites and the "castlenaui" basaltiform corallites of the compact variety—are not as common in the Ap- palachian Hillsdale Limestone as they are in the era- tonic, sheet limestones with extensive Lithostrotionella biostromes found in the Middle Mississippi Valley. Some of the limestones in the lower portions of the Chester and upper part of the Meramec series in west- ern outcrop belts of Alabama, Tennessee, Virginia, and West Virginia contain similar intraclasts of algal lime- stone ; this suggests a persistence of a widespread paleo- ecological condition wherein masses of biogenic micrite induced by algal growths existed in the same general environment in which black, pyritic, petrolifer- ous lime mud was accumulating. Possibly, the algae grew on little mounds rising above the redox threshold in shallower waters. The highest occurrence of algal masses I have seen are clasts in the locally conglomeratic Princeton For- mation of late Chesterian age. These masses are always the single kind of limestone clast to survive dissolution and, therefore, present the familiar blue-gray color which shines through the dead, iron-stained appear- ance of weathered "wormeaten" surfaces of Princeton conglomerates. Worm Burrows, Markings, and Trails Fossil worms, preserved as pseudomorphic casts of bur- rows or as trails of ordered setae impressions, do not make much of an impression on the Appalachian field geologist, although trails of Pteridichnites biseriatus in the monotonous, thick sand-shale successions of the Brallier Formation (Portage age) are so abundant and widely distributed as to be a sort of identification label for beds of Portage type. But the Portage rock hardly needs fossil labels because its general lithology is so distinctive. Worm markings—chiefly pseudomorphic fillings of segmented impressions or burrows—occur on the un- dersides of Tuscarora or Clinch sandstone ledges at so many places in the Appalachians that one can be assured of finding them if only he looks for them. These markings, widely known as Arthrophycus—both A. harlani and A. allegheniensis—are so abundant in places as to be a lithologic determinant. To find the undersides of projecting ledges plastered over with 10 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Arthrophycus raises the question of how the things were formed. The worms evidently frequented muds, not sands. When the tides ebbed, the mud began to shrink and dry out, and this induced the worm to really flex its segmented body and to thresh around so that its free part flapped crazily here and there around the vise of drying mud that held the rest of the worm captive. Thus, after the mud had dried thoroughly the mud surface around a worm hole had a mass of markings. Each worm that eventually died for lack of water had made segmented, linear impressions that converge toward the hole where the trapped part of the worm was evidently held. Hence the "turkey track" type of impression. Arthrophycus has a special value that I have found useful in mapping. The normal mode of occurrence is on the underside of sand beds. When the sand layer was deposited, loose sand was washed into the bur- rows and the segmented, groove-like impressions in the sand became firmly cemented and welded into orthoquartzite. The Clinch and Tuscarora, especially south of the James River in Virginia, in many places are involved in faulting and overturning of successions of beds. Arthrophycus provides the readiest cue as to whether the beds are right-side up or upside down. Without doubt the best display of Arthrophycus in in- verted successions of ledgemaking sandstone is on Draper Mountain in Pulaski County, Virginia, in the cut made by old U.S. Route 11. As previously indi- cated, the casts of worm burrows are a part of a three-fold display of lithologic features that testify to the inverted position of the beds. In the same cuts, one ledge of the Clinch shows upside-down foreset bedding and an adjacent shale interbed shows fracture cleav- age that likewise is indicative of inverted beds (B. N. Cooper, 1961, pi. 28A) . Thus, the familiar though enig- matic Arthrophycus doubles as a primary sedimentary feature that helps the stratigrapher determine when the beds are in normal succession or overturned. Echinoderms in Carbonate Strata Perhaps the most significant contribution by in- vertebrates to the lithology of Appalachian sedimentary formations is made by echinoderms—particularly cystids, crinoids, blastoids, paleoasteroids, and edrio- asteroids—in producing the texture of so many Ap- palachian limestones. Cystids, crinoids, and blastoid calyx and stem plates are secreted as crystallographic units which, to some extent, are knit together. When the animal dies, the skeleton may, with the decay of soft parts, yield an aggregate of loose plates or a fused monolithic structure which shows plate outlines but which will break either along plate boundaries or across them. Stem and pinnule plates are fragile and brittle and seldom survive much transportation and abrasion. This is especially true of the stem plates of cystids, which have an inordinately large lumen compared to crinoids. Except for extremely rare biocoenoses of crinoids, cystids, and blastoids, such as the rare Crawfordsville, Indiana, crinoid beds, echinoderms occur overwhelm- ingly as cleavage fragments of biogenic spar calcite. Rocks composed of cleavage fragments of encrinal or- igin have a peculiar, coarse granularity characteristic of calcarenites or biosparites. Echinoderms in the Ap- palachian section make their first appearance as cystid- like tests in early Acadian rocks. In the southern Appalachians, the Maryville Limestone is probably the oldest Cambrian unit known to contain echinodermal remains. Unlike most of the younger biosparites or calcarenites, those in the Maryville are invariably dark-colored and are almost invariably oolitic. They are well exposed in the classic Thorn Hill, Tennessee, section but are best displayed in natural exposures about three miles southwest of Speers Ferry bridge over Clinch River on U.S. Routes 23 and 58 and about 600 feet north of the riverbank. Considerably comminuted trilobite "hash" occurs in the same beds. These oolitic biosparites are forerunners of the great masses of en- crinal and cystoidal debris that were to be deposited in the Appalachian geosyncline beginning in Cham- plainian time (Middle Ordovician). The grain of an Appalachian Middle Ordovician calcarenite results in a distinctive and unmistakable roughness of the sur- face, produced by closely spaced, projecting corners of cleavage rhombohedra. Such rocks, however, are not indicative of any specific stratigraphic level but occur at many different levels, as Keith (1895, text) so clearly described in Tennessee. A personal experience illustrates how important it is to appreciate the feel of a calcarenite. In 1944, I was asked to give a United States Geo- logical Survey ground-water geologist a tour of western Virginia to help him become familiar with the strati- graphic section. At that time, the Pulaski fault was in- terpreted to cross U.S. Route 11 about 1.5 miles north- east of Buffalo Creek where Mid-Ordovician lime- stones supposedly were resting on Martinsburg shales (Butts, 1933). At that time the exposures were fresh NUMBER 3 11 ^v*,/* . '^ * k 4 E £ IV- BBK~ s ??%&&?£ - ,11 A , f; r PLATE 5.—Features of algal-matte limestones, A, Nests of dolomite crystals pseudomorphic after anhydrite, from Patterson Member of Shady Formation, Ward Creek section, northwest of Cripple Creek, Wythe County, Virginia, B, Algal-matte limestone cup or "petri dish" containing coarsely crystalline dolomite probably pseudomorphic after anhydrite, with center core of quartz crystals; from same locality as above, c, Hand-specimen of typical algal-matte limestone (natural scale), from Patterson Member of Shady Formation, 1.25 miles east of Huddle, Wythe County, Virginia. 12 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY and worthy of photographing. While my colleague set up his camera for a photograph of the supposed fault contact, I casually rested the palm of my hand against the rocks on the northwest side of the cut. The familiar feel of a calcarenite led me to look at it closely. The presence of orthid brachiopods and shreds of bryozoans confirmed my "Braille" impression. The rock was a calcarenite commonly found beneath the Liberty Hall black limestone-black shale lithofacies of the Edinburg Formation, so I began to suspect that the supposed Martinsburg actually might be Liberty Hall shaly lime- stone. This was quickly confirmed by finding graptolites in the black shale near the boundary with the calcar- enite on the opposite side of the road. These discoveries led to curiosity about the dolomites under the Middle Ordovician limestone that also are fully exposed along the highway. Eventually we found the Ceratopea and Lecanospira faunules of the upper and middle parts of the Beekmantown beds in normal sequence under the Champlainian rocks, and the fault about to be photo- graphed "disappeared." Downgrade to the southwest, just northeast of Buffalo Creek near old Buffalo Mills, I found the real Pulaski fault where the lower Beek- mantown had been overridden by the Elbrook Forma- tion and where a good development of Max Meadows crush conglomerate characteristic of the Pulaski fault zone separated the Beekmantown from the Elbrook. Indeed, the whole train of discoveries had been trig- gered by the recognition of the significance of the cal- carenite, merely by feeling the rock. Middle Ordovician calcarenites are very commonly pinkish-tinged because of the very characteristic color of the crinoid and cystoid debris of which the rocks are composed. Although these echinodermal groups in the form of abraded fragments are overwhelmingly predominant as rockmakers and are distinctive litho- logic determinants, other fossil fragments also abound, including intraclasts of Solenopora algae, bryozoans of many types, brachiopods (especially tumid triplesids such as Oxoplecia), and triangular, strongly plicate Oligorhynchia. One especially significant biosparite occurs at the very base of the Middle Ordovician Tumbez Lime- stone. It is often red or pink and is loaded with detri- tal chert clasts like its dolomite counterpart, the Black- ford Formation. Almost invariably the biosparite is loaded with myriads of Rostricellula, a strongly plicate, tumid rhynchonellid that withstood rolling very well without dismemberment of the valves. The elegant little shells occur by the millions in pink encrinal lime- stones in the Clinch Mountain belt of outcrop from western Russell County, Virginia, southwestward all the way to the southwest end of the outcrop belt near Chesney, Union County, Tennessee. Rocks so full of fossils as are these beds owe their color, coarse texture, characteristic cross-bedding, and stylolite markings to the abundance of organic remains. Such rocks, which as Ulrich (1911) indicated years ago aggregate nearly 700 feet of section in the Luttrell- Chesney district of East Tennessee, constitute a rock type commonly called Tennessee pink or red marble (Holston marble) which is very abundant in Ten- nessee and in southern Virginia. How does one explain the many hundreds of feet of encrinal calcarenite whose pink to red colors come from the mass effect of pink or red calcite-spar fragments of crinoids and cystoids? Where were the crinoid and cystoid beds from which billions of tons of encrinal fragments could be derived by wave and current ac- tion and be transported as spar calite sands to form this dominant lithofacies of Ordovician limestone? What curious combination of marine sedimentary processes could continue to yield vast tonnages of spar calcite by recurrent disruption of enormous, growing areas for crinoids and cystoids sufficient to furnish hundreds of feet of nearly pure calcitic sediment? The enigma is further heightened by the fact that no one has yet re- ported a token occurrence of lush-growing crinoids or cystoids that could supply the sediment. That waves and currents swept the abraded echinodermal debris across the sea floor to its final resting place is indicated by the abundant development of bold foreset bedding in many of these calcarenites, particularly in those utilized as polished marble slabs. These echinodermal calcarenites are obviously a lithofacies that is repeated in the Ordovician Appalachian section of Virginia and Tennessee from the base of the Champlainian (as typi- fied by the calcarenites at the base of the Lenoir Lime- stone at Friendsville, Tennessee) nearly all the way to the top of the Wilderness Stage. A facies variant of the typictal Holston marble litho- facies is represented by the Effna Limestone on the Saltville fault block in Virginia and Tennessee. This unusual limestone is developed sporadically in the lower beds of the Porterfield Stage. The Effna is a series of isolated biohermal mounds of nearly white limestone that are laced together by gray to pinkish cal- carenites. Some of the mounds and associated calcare- nite sandbanks interfinger with black graptolitic shales as in Porterfield Quarry of Olin Mathieson Chemical NUMBER 3 13 Corporation in Rich Valley, Smyth County, Virginia. The main reef body consists of limestones of algal ori- gin but which are rich in bryozoans, octacorals, brachi- opods, and trilobites and are mixed with pockets and stringers of calcarenite. The bedded parts of the Effna form bold ledges, and the unbedded reefy masses, literally mounds of fossils of many kinds, form huge clints. A curious thing about the Effna reefy bodies is the inordinately low magnesia content—in many places less than 0.60 per- cent calculated as MgCOa. Most of the fossils prob- ably are bryozoans, and these animals always secrete shells with from 3.5 to 8 percent (or even more) of mag- nesia calculated as MgC03. How is magnesium se- lectively removed in such rocks? Coarse calcarenites make a strong appearance in the Helderbergian of the Appalachians. In the northern Appalachians the encrinal calcarenite is the Becraft Limestone, which is loaded with round crinoid colum- nals. In the Central Appalachians much of the coarse limestone that occurs in the Helderberg is in the so- called Coeymans Limestone, which looks nothing like the Coeymans on Becraft Mountain but resembles the New York Becraft. The Central Appalachian Coey- mans is locally a pink marble made up of a distinctive thick type of columnal with interlocking milled or ridged faces. This kind of rock is so like the Mid- Ordovician Tennessee marbles tiiat it is uncanny. Yet close examination readily shows a difference, the Mid- Ordovician pink calcarenites are loaded with cystoid columnals whereas the Devonian pink marbles contain essentially no cystoid debris. In the Clifton Forge district of western Virginia there is an aberrant pink calcarenite that occurs in the Keyser Limestone, which otherwise is mainly bio- stromal coralline limestone. The occurrence of two "Coeymans Marble" zones is not uncommon. What makes the crinoid stem plates pink? This question has never been answered, but the durability of the pink color is attested by slabs of Craigsville marble that were hewn out of a small quarry near old Craigsville on the C & O Railway. These slabs were hauled out by wagon through Buffalo Gap west of Staunton, Augusta County, to the Tinkling Springs Church cemetery west of Waynesboro, where some of them have reposed for 225 years without losing the delicate pinkish color of the Craigsville marble (a local variant of the Coeymans Limestone). Lush crinoid beds are unknown in the Central and Southern Appalachians, but cystoids are locally plenti- ful in the late Silurian Keyser Limestone in Maryland. The origin of the Devonian encrinal calcarenites is as puzzling as that of the Middle Ordovician calcarenites of Holston-marble type. Strata composed largely of crinoidal remains abound also in the Mississippian limestone lithofacies from Pennsylvania to Alabama. The Mississippian cal- carenites are readily distinguished by the abundance of unbroken discrete stem and calyx plates, which are distributed through a finer grained matrix. Such cal- carenites grade rather abruptly into oolitic limestone in which spar calcite is plentifully mixed. Mississippian calcarenites in the Appalachian region seldom are as coarse-grained as their Ordovician counterparts, ex- cept for the unbroken stem segments which are the hallmark of the younger calcarenites. The abundance of crinoidal calcarenite in the Appalachian Missis- sippian raises mystifying questions about the paleo- ecology. The various calcarenites in the Appalachian section can generally be distinguished by using a combination of lithologic and fossil characteristics. The Mississippian calcarenites and the oldest ones in the Acadian lime- stones are variably oolitic, but oolites are essentially absent in the Ordovician, Devonian, and Late Silurian calcarenites. Ordovician calcarenites can be distin- guished from Devonian or Late Silurian ones because the first-named are predominantly cystoidal, whereas cystoidal calcarenites are rare in the Silurian and Devonian. The latter are coarser and not uncommonly contain large spar-calcite intraclasts of fused columnal plates up to 2.5 centimeters in diameter. The Missis- sippian calcarenites are generally impure, clayey, and more erratic in thickness. The Tennessee marbles of commerce, so lavishly used in the National Gallery of Art in Washington, are composed entirely of fossil fragments, all somewhat recrystallized without loss of structural detail, fused into a coarsely crystalline mass devoid of porosity. Brachiopods as Lithic Components Brachiopods are surprisingly abundant rock makers. Waves and currents have segregated brachiopods in many places to form limestones of peculiar and highly distinctive lithologies. In many of these beds the par- ticular fossils are sufficiently abundant to be the over- whelmingly predominant lithologic determinant. 14 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY The oldest Appalachian limestones to be literally composed of brachiopods are the Rostricellula bio- sparites of basal Lenoir or equivalent Tumbez Lime- stone or basal New Market limestone of the Shenan- doah Valley of Virginia, and the lowermost Row Park limestones of Maryland and Pennsylvania. These rather erratically distributed beds were deposited as discontinuous concentrations of easily rolled shells which ultimately came to rest on deeper recesses of the post-Canadian surface of unconformity. The irregu- larities on this surface, including undercut projecting ledges, sink holes, and deep cut channels, account for some amazing occurrences of this Marmorian faunule sometimes found far down in sequestered dissolution cavities or clefts down in the Knox dolomites. At Eagle Rock, Botetourt County, Virginia, the Rostricellula brachiopods flushed into a narrow, current-cut, V- shaped cleft in the surface of unconformity developed in the Knox. The channel filling of Rostricellula brachiopods crops out as an inverted wedge of brachi- opod pudding stone that projects upward into over- turned, older Knox dolomites. The Elway limestones of the basal Ashby Stage con- tain one of the more profuse displays of Dinorthis known, with D. holdeni the oldest known species. One large slab in the collections of Virginia Polytechnic In- stitute contains about 6,000 shells exposed on one sur- face. Some are silicified in blue-gray limestone, but many occur in black chert. All the shells are of about the same size and shape and degree of resupinacy. But on U.S. Route 25W south of Clinton, Tennessee, the cherty Elway has yielded thousands of free silicified shells of Dinorthis that can be sorted into various sizes and shapes and degrees of resupinacy presaging many of the species of Dinorthis that range as high as the Richmondian. Dinorthis holdeni in company with other Elway fos- sils occurs in a wide range of limestone lithologies: pink Holston-type calcarenites at Chesney, Union County, Tennessee; cherty micrites in Elk Garden, Russell County, Virginia; and in black fine-grained limestones containing abundant lenses of black chal- cedonic chert. The typical occurrence is in chert blocks that have been bleached and somewhat devitrified. Although brachiopods are abundant throughout the Middle Ordovician, the Sowerbyella-Yimestones at the base of the Martinsburg Formation are the only other abundantly brachiopodiferous limestones in the Middle Ordovician formations. Any experienced Appalachian field geologist can be handed a piece of this Sower- byella-bea.ring limestone and will identify it immedi- ately because of the special look that the fossils give to the limestone. The slabs are packed with brachio- pods—particularly Sowerbyella, Zygospira, and stro- phomenids—and profuse bryozoans that include Hallopora, Rhinidictya, and Prasopora. The fossil shells are embedded in a matrix of spar calcite, finely ground shell debris, and quartz silt. The limestones are slabby and irregularly bedded. Individual layers of shelly limestone are interbedded with black pyritic shales which weather greenish gray to brownish buff. The brachiopodiferous slabs of lower Martinsburg, or "Trenton" limestone, are commonly strewn over foot- hills at the base of obsequent slopes of typical Appala- chian hogback ridges upheld by the basal Silurian Tuscarora or Clinch Sandstone. One can recognize them without even picking up the slabs. The curved, crisp, thin valves of Sowerbyella can be recognized either plastered over bedding surfaces or exposed on surfaces normal to bedding where the shells show as thin profiles. Finally, the luster of the shells imparts a silky appearance to the beds that is very distinctive. The high Martinsburg from northern Virginia to Alabama and Georgia is represented by or contains limestone beds packed with tumid Orthorhynchula shells, many of which are filled with white calcite. The concentrations of these shells and their coarse ribbing is so distinctive as to impart a definite appearance to the rock as a whole. The rock in which the shells occur is very impure, and is composed largely of a silt matrix. The weathered limestones, full of holes where shells have been dissolved out, yield a pulp that is readily identifiable. Not uncommonly the leached rock will show whitish thread-like coatings of hydroapatite, evidently derived from myriads of Lingula brachiopods which have been partially dissolved by percolating waters that later evaporated to leave the whitish residues. Silurian and Devonian Limestones Of all the calcarenitic or highly fossiliferous limestones, the ones most difficult to distinguish in the field are the Cayugan and Helderbergian limestones. The Coeymans calcarenites unfortunately are mimicked by local coarse calcarenites in the Late Silurian Keyser Lime- stone. The true Coeymans has characteristic brachio- pods, but in the calcite-spar matrix of the calcarenites they are impossible to extract without decorticating the shells so that they are useless for identification. The NUMBER 3 15 pinkish calcarenites in the Keyser tend to be adult- erated with quartz sand, but to some extent this is also true of the Coeymans itself. The true Coeymans of the Clifton Forge, Virginia, district is easily the coarsest grained limestone in the Paleozoic succession. Keyser limestones are blue-gray limestones that con- tain black chert and thus resemble, superficially, many of the Middle Ordovician limestones such as the Lin- colnshire Limestone. But if one puts the lens on the Keyser limestones some clue of Cayugan age among the fossils can be detected very quickly. For example, Cladopora corals, or favositids, stropheodontids, or lirate to finely plicate spiriferids generally can be identi- fied. The Keyser contains irregular biohermal bodies of very fossiliferous limestone permeated by coarse cal- carenite. Such masses have a distinctively massive ap- pearance but show ropy projections of chert and silicified shells. The experienced eye can recognize these biohermal beds, which are not duplicated in the Licking Creek Limestone or in the Coeymans Limestone. The main difficulty with the Devonian limestones is their lack of good exposure or evident continuity of ex- posures, and their soil cover is more often thick than thin. Fossils in Devonian and Younger Clastic Strata One might think that the fossiliferous Oriskany, so full of molds of large fossil shells, would be easy to recognize, but that is not so unless it is the only sand- stone in the Devonian part of the section of Cayugan and Helderberg limestones. Generally, there are other sandstones—for example, those in the base of the Keyser Formation (generally but not everywhere indi- cated by a predominantly foreset-bedding structure), the Healing Springs Sandstone of New Scotland age, and the arenaceous zones in the Becraft which weather much like the Oriskany and pass southward into a sandstone known as the Rocky Gap Formation. The stratigrapher working in these beds will do well to depend on sequence to distinguish these sandstones. Of course, this is another way of saying that the lithologies themselves are not distinct. In the great mass of Devonian elastics, fossils con- tribute notably to the recognition of only a few zones. The Needmore Formation (Butts, 1933, 1942), olive- drab shales of "Onondaga" affinities just above the Oriskany, is a unique lithology commonly characterized by almost unbelievably abundant molds of all the diminutive members of the Onondaga-Schoharie fauna—Tentaculites, Leptostrophia, Ambocoelia. Gen- erally, the eye-catcher of the Needmore is a com- bination of two things: the lumpiness or lack of fissility and the curious surfaces formed by myriads of Amphi- genia brachiopods which are so strongly curved that molds of the shells look like pebbles or granules. The Millboro black shales contrast in color with the gray-green Needmore, but the former lacks any obvious implantation of fossils on lithologic character- istics of the rock. The great thickness of rocks of Portage lithofacies, which in the Southern Appalachians extends upward from the upper biofacies zone of the Millboro Forma- tion—Butt's (1942) Naples division—all the way up to the Maccrady red beds of Warsaw (Mississippian) age, at first glance or even after considerable study, appear to be essentially unfossiliferous. They are char- acterized by a monotonous repetition of rusty-weath- ering sandstones and drab shales, and their only note- worthy attribute is their great thickness. A very close examination will indicate some distinctive lithofacies and biofacies, but none of these are reliable strati- graphic markers. Much of the tremendously thick Brallier Formation is virtually devoid of fossils, but in the lower 300 to 500 feet of the formation there are crumbly, limy beds very commonly packed with brachiopods of the so-called Ithaca biofacies. The as- semblage is dominated by brachiopods and clams, gen- erally including Rhipidomella, Spirifer mesastrialis, athyrids showing interiors, Chonetes, Leptodesma, Grammysia, and Paleoneilo. One can learn to spot these beds after examining them and collecting fossils from them, but just what is distinctive about them is elusive. A stratigrapher learns how to spot beds of different ages but, like the untrained musician who plays by an instinctive ear, he may find it very difficult to con- vey to others how he recognizes certain parts of the section. The inadequacy of language in descrip- tion is illustrated by the following incident. Thirty-four years ago, I was driving with Dr. Charles Butts, dean of Appalachian geologists in his time and (luckily for me) my mentor, during a heavy rainstorm over Brushy Mountain along U.S. Routes 52 and 21 in Bland County, Virginia. The car win- dows were no more than translucent from con- densate moisture. Suddenly, Dr. Butts exclaimed: "Stop on the shoulder! We will go back and collect some Productella shells and goniatities out of the Braillier; I just saw them back a little ways." It was 16 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY hard to believe, but my doubts that he could see so much were wiped away when we walked only about 150 feet back up the road and I saw some soft, crumbly, rusty-weathering, lumpy rocks with just the fossils he said he had seen out of the corner of his eye. When I pressed him why or how he knew what he had seen, he said, "You cannot always put into words what you can see." Perhaps someone else, having a stronger emotional tie with these beds than I, can do justice in pinpointing what I am incapable of describing as distinctive about the the "Ithaca beds" and Naples beds in the Brallier Formation. The upper part of the Portage lithofacies from James River southward—even down into Tennessee— contains a zone that is characterized by vast quantities of Spirifer disjunctus, Productella, Camarotoechia, and Chonetes brachiopods and great quantities of Melocrinus stem plates that are distinguished by strongly radial striations on the flat sides of the "but- tons." No one has any trouble spotting this zone from a distance of 10 to 50 feet—so long as the assemblage contains plenty of thick Spirifer shells. If the road cut is fresh, the calcite of the shell glistens in the sun and outlines what could only be Spirifer; if the exposure is weathered, the Spirifers have dissolved away, leav- ing a peculiarly shaped cleft or indentation in the rock. Where these fossils are very abundant, both fresh and weathered surfaces have a look all their own. Not infrequently exposures of the Spirifer dis- junctus beds contain some red-colored beds, probably signifying an admixture of in-washed Catskill red material. The Spirifer disjunctus biofacies in the Por- tage lithofacies is very calcareous—so strongly so that some of the rock could be called an impure limestone without stretching the imagination. Commonly, the fossils in this zone occur in relatively thick beds of sandstone-siltstone which aid in the eye-catching qual- ities of the fossils or their impressions. But a little higher in the sequence, the Spirifers lessen in abundance and essentially drop out of the faunules— leaving Camarotoechia, Productella, Chonetes, and crinoid buttons. This is an equivocal, deceiving, frus- trating assemblage that has but one redeeming fea- ture—abundance of fossils. These beds are easy to spot, but almost always they occur as loose slabs of rock not uncommonly weathered and coated with a light- gray, case-hardened exterior. These fossils, forming the same kind of sandstone, cover quite a stratigraphic interval, and if one follows this zone upward he will in due course see the Mississippian spiriferids Pseudo- syrinx and Syringothyris. Accompanying these fossils, besides Productella, Chonetes, and Camarotoechia, are Burlington-Keokuk brachiopods, notably Tetracam- era. Thus, this confounding biofacies, which ranges down to the level of the Chemung—which Chadwick (1935) evaluated as just a "brachiopodiferous facies of the Portage"—surely ranges as high as the Missis- sippian Price Formation and its Tennessee counter- part, the upper part of the Grainger Formation. What impressed me most about this troublesome group of distinctive fossiliferous beds of the same gen- eral lithology is that the mythical systemic boundary between the Devonian and Mississippian cannot be spotted. But this is not a rare phenomenon. Most sys- temic "boundaries" in the Appalachian section are figments of the imagination (Cooper, 1960). Many a vicious controversy among former geological friends is recorded in geological literature in the polemics of where to place these "boundaries"—many abetted principally by paleontologists. The Price-Fort Payne-New Providence faunules of Burlington-Keokuk age are favored by one lithozone that aids in locating the biozone with Pseudosyrinx and its associates. This faunule is always found very close to a distinctive glauconitic siltstone. The Mississippian Greenbrier limestones above the Hillsdale, which are of Ste. Genevieve and Gasper ages, are distinctive because of the prominent display of crinoid and blastoid remains, the former generally as isolated plates, the latter commonly as whole calyces. The Ste. Genevieve beds, which Wells (1950) has renamed the Denmar Limestone, are either calcilutites or calcarenites. The fine-grained beds are sparsely fos- siliferous, but the calcarenites commonly are loaded with crinoid columnals among which the distinctive stems of Platycrinites "huntsvillae" are most conspic- uous. The stem plates of that crinoid are elliptical in shape and have a spinose periphery. In the stem, each element is offset slightly to produce a flattened, ribbon- like spiral that must have had greater flexibility than most crinoid stems. In good light, these stem plates, which are only about 0.5 centimeter in maximum diameter, can be spotted ten feet away from the out- crop. Higher Mississippian limestones that have been widely called "Gasper" contain wing plates of two or three genera of crinoids of which Talarocrinus and Pterotocrinus are most easily recognized, and they are truly abundant. Mississippian crinoids clearly were great rockmakers. Most of the limestones of Gasper age are also characterized by rather common blastoids, NUMBER 3 17 of moderate size, that were rolled around by waves and currents without appreciable abrasion damage to the monolithic calyx, which was about as resistant to break-up as a hard pebble of rock. The impress of crinoids and blastoids on the Mississippian Greenbrier limestones is best conveyed by the 1,000 feet or more of limestones exposed along the lower slopes of Stony Ridge along State Route 16 southeast of Bishop, Taze- well County, Virginia (B. N. Cooper, 1945). Fossils also contribute in a remarkable way to the peculiar, rusty-weathering calcareous shales of the basal part of the Bluefield Formation. The finest sec- tion to see the features is opposite the junction of West Virginia Route 12 and U.S. Route 460 just north of the rail underpass at the northeastern limit of Blue- field, West Virginia. The crisp, delicate markings made by thin fragile fossil shells contribute a diagnostic feature of these beds. Most of the fossils in this zone were evidently indigenous and little disturbed by deposition of muds around the shells as they accumu- lated. Bold markings on such beds are made by the screw-spiralled stolons of Archimedes. In just the right light the profusion of Archimedes shows up very prom- inently 10 to 25 feet away. But for one or two exceptions, the lower Bluefield brachiopod and bryozoan limestones mark the highest level in the Appalachian Paleozoic sequence where fossils literally make rocks. This zone also is just below the tremendous thickness of predominantly clastic rocks of the red-green-black Mauch Chunk lithofacies or magnafacies which persists up to the base of the Pennsylvanian Lee Formation. In medial and upper portions of the Bluefield Formation in the great Hurri- cane Ridge syncline (B. N. Cooper, 1961, 1964; Thomas, 1966) and also in the calcareous portions of the Hinton Formation and even higher Bluestone Formation, the impress of fossils is less evident. Never- theless, it should be mentioned that the abraded, broken, disoriented character of the invertebrate fos- sils in the Bluefield-Hinton-Bluestone successions— aggregating nearly 5,000 feet in both the Hurricane Ridge and Greendale synclines—does point to one gen- eral conclusion. The fossils clearly are part of a great accumulation of what Weeks (1952) aptly described as "dump deposits," and in their beaten, bruised, broken, and corraded condition they constitute the most clear-cut examples of transported thanatocoe- noses in the entire Appalachian column. Two markedly fossiliferous zones above the basal Bluefield Archimedes beds are valuable as key beds for mapping purposes and stratigraphic reference. One is a rather persistent zone of calcareous mudstone bear- ing whole or fragmented giant pentremites (Pen- tremites mccalleyi, the largest known Appalachian pentremite, rivaling in size P. obesus of the Mississippi Valley region). These fossils make a puddingstone rock zone of very distinctive appearance of up to four feet or so in thickness. They are almost unbelievably abundant in this one zone, but they have a consider- ably greater stratigraphic range through beds in which they occur as more or less isolated mavericks. The upper of the two high Mississippian lithologic zones composed primarily by fossils is the Avis Lime- stone Member of the Hinton Formation. Near its top it has a thin argillaceous limestone in which the com- plete shells of an elliptical Composita abound to such an extent as to outweigh the matrix in many places. This is a widespread thin zone capable of yielding more pounds of whole fossils per cubic foot than any other division of the Paleozoic of the Appalachians, yet the still lowly estate of Appalachian Paleozoic in- vertebrate paleontology is attested by the fact that these Composita shells have never been described specifically. Appalachian Stratigraphy without Fossils For nearly 50 years prior to the early 1940s Appalach- ian stratigraphy was plagued by serious misconceptions which would not have existed had the invertebrate faunas been studied. The confusion of the Middle Ordovician Bays Formation with the Juniata-Clinch as described by Rodgers (1953, pp. 78, 97) is an example. Among the Ordovician limestone forma- tions, one rock type was singled out by Ulrich (1911) as a lithologically distinctive zone which he named the Mosheim Limestone. Two other zones of the same kind of limestone, separated by a few hundred feet of beds and neither the same as the Tennessee Mosheim, were confused with the Mosheim in southwestern Vir- ginia. This duplication of the Mosheim lithology (first described by Cooper and Prouty, 1944) was a source of much confusion. About six different stratigraphic zones of Holston-type limestone, another distinctive lithofacies, occur in Tennessee. The lowest occurrence in the basal Marmorian is directly on top of the Knox Dolomite of Canadian age. The highest occurrences are in the Wilderness Stage in Virginia and Tennessee. All of the different Holston marbles could be easily distinguished if the enclosing rocks maintained some 18 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY semblance of lithologic continuity from belt to belt, but the entire succession containing marble beds ranges greatly in lithology from place to place. The stratigraphic positions of all the marble lentils could be worked out if the fossils in the individual marble beds were studied in detail. Conodont studies may prove helpful in solving problems with the marble zones. It is regrettable to me that the term Chickamauga was revived and used so widely by Rodgers in Ten- nessee. Rodgers's lettered zones of the Chickamauga can be related to a number of Virginia formations that have an established faunal basis for their recognition, whereas the lettered zones do not. The only reliable and consistent way to work stratigraphic successions is to use all the information that the rock supplies. Assumption of unique nature of some limestone lithologies was a serious retardant in Appalachian stratigraphy that could have been avoided had fossils been used in conjunction with the lithologic characteristics. Perhaps the part of the section that best illustrates the deceptive quality of lithology criteria alone is along U.S. Route 25E south of Evans Ferry bridge over an inlet of Norris Reservoir. There are three prominent zones of buff-weathering cobbly limestone—all of which have been confused with the so-called Ottosee Limestone of Ulrich (1911). The lowermost zone of this cobbly buff-weathering limestone, the lower part of the Lincolnshire Limestone of Ashby age, contains the Lincolnshire brachiopod fauna. The middle zone of cobbly limestone along U.S. Route 25E in the Evans Ferry section contains abundant Porterfield fossils that are easily distinguishable from those of the Ashby Stage. The third cobbly zone of "Ottosee" type is the Wardell Limestone of the Wilderness Stage. Fossils in the Wardell are markedly different from those in either the middle or lower zones of cobbly limestone com- monly called Ottosee. Knowledge of the fossils is essential to consistent stratigraphic identification of recurrent zones of the same lithofacies. Discreditation of Guide Fossils The question might be asked: "What is the value of any individual fossil in the recognition of any strati- graphic unit in the Appalachian Valley?" One would have to respond by saying that no single index fossil has stood the test of time. A number of individual fossils were misused in the past for identification of certain ages of beds, with rather disastrous results. Among such fossils were Lecanospira, the depressed spired gastropod once thought to be confined to the Longview Limestone; various species of Tetradium, some of which are but growth stages of others; the enigmatic Cryptophragmus; Cyrtospirifer disjunctus; and many others. All of these were touted to be "dependable guides." The only way to use fossils meaningfully in stratigraphic work is to make full use of the various elements of the fauna or faunule that are found and identified. There is safety in using num- bers of genera and species in stratigraphy. The facts regarding the wide stratigraphic range of the fossil Crypto phragmus antiquatus, long supposed to be a valid guide to beds of Lowville age, exemplify the danger of relying on a single kind of fossil for strati- graphic work (Cooper and Cooper, 1946, pp. 58, 59). The last Appalachian Paleozoic guide fossil to fall from grace was the genus Lithostrotionella, which was long respected as a valid guide to the Appalachian Hillsdale Limestone, or its equivalents, commonly identified as St. Louis Limestone by Butts (1927, 1942). This dis- tinctive coral is now known to range outside the St. Louis Limestone or Hillsdale Limestone; it has been found in the Little Valley Limestone in the Greendale syncline. Value of Maverick Fossils The fact that fossils make up so much bulk of strata and actually determine the physical appearance of so much rock is the best reason for the stratigrapher to seek help from or to join up with a good, experienced paleontologist and thereby learn what a great help fossils can be. One kind of fossil that has special value in stratigraphic work is the maverick organism that became separated from its paleoecological habitat and was buried and preserved in a lithofacies environment, and in the midst of other buried organisms, foreign to its natural habitat. The maverick fossil has special value in contributing information indicating the essen- tial contemporaneity of different sediments containing relatively different fossil assemblages. In the Appalachians, some faunal assemblages are little known outside a given lithofacies. For example, the Normanskill graptolite fauna seldom is observed in rocks other than black shales. When G. A. Cooper and I were working on the Middle Ordovician of the Shenandoah Valley, the problem of the relation of the Normanskill graptolite fauna, in the rocks long called NUMBER 3 19 Athens black shale by Butts and others, to the fauna of the rocks formerly called Chambersburg Lime- stone was finally resolved in two ways. First, detailed collecting showed clearly that the trilobites and at least some of the brachiopods occurred both in the beds identified as Athens and in those to the north called Chambersburg. This did not convince everyone of the equivalency of the two formations. Detailed collecting in the so-called Athens platy black lime- stones and graptolitic shales at Lacey Spring in Rock- ingham County, Virginia, led to the discovery of one of the most characteristic Chambersburg fossils, Chris- tiania lamellosa. Extensive collections of fossils from the Chambersburg farther north in Shenandoah County, Virginia, led to the discovery of Normanskill graptolites in token quantities in Chambersburg lime- stones. Doubtless the graptolites and the characteristic Chambersburg brachiopod, Christiania lamellosa, were strays and were atypical of the rocks in which they were found. But the token occurrences of the two in both the Athens and Chambersburg beds provided additional information needed to clinch the correla- tion. The possibility of finding maverick fossils in for- eign lithofacies is sufficient reason to look for them, but diligent collecting may be necessary to round up even a few such fossils. Stray fossils are most apt to be found in the border fringes of two adjacent but rela- tively distinct lithofacies, and that is where the search for them is most likely to be fruitful. The maverick fossil may also be turned up in the etching of lime- stones. Rare fossils that cannot expectably be broken out of rock may be dissolved out and freed for study. Thus, all the material recovered from etching must be combed over to see if a "one-in-a-million" fossil find can be made. Recurring Faunules Recurring faunules have all too often put the paleon- tologists on the defensive in answering the question: "If fossils are any good for stratigraphic work, how can they be depended on if they recur?" Recurring fau- nules should be expected if faunas evolve gradually and if the depositional environment exercises a strong con- trol over the dispersion of a given faunal assemblage. The key to the understanding of recurrences lies in faunal associations. For example, in the Edinburg Limestone of the Shenandoah Valley the distinctive cystoid Echinosphaerites occurs in two relatively widely separated zones, one at the base of the Edinburg 372-386 0—71 3 and the other several hundred feet higher in the over- lying Oranda Formation, which is the approximate equivalent of the Trentonian Salona Formation of Pennsylvania. The lower occurrence is the main one; the Oranda occurrence is the anomalous one. The re- currence of Echinosphaerites is in an association of brachiopods, trilobites, and bryozoans that are very different from the associated fossils in the lower part of the Edinburg Formation (Cooper and Cooper, 1946, pp. 35-113). In the geographic area of the re- currences of Echinosphaerites, the phasing out of the Edinburg fauna and phasing in of the Oranda fauna expectably would be consummated gradually, with some favored organisms managing to survive longer than other members of the earlier fauna. Echinosphae- rites, which probably had a drifting habit, could have survived a bit longer in some secluded place and then could have broken out of captivity and spread over the area a second time before being eliminated. Lack of similarity of the associated fossils is indication of a different age for the two occurrences, and the occur- rences of Echinosphaerites are subordinate in impor- tance to the preponderant weight of faunal evidence indicated by dissimilar nature of the sets of fossils as- sociated with the two occurrences of Echinosphaerites. In the area immediately north of Strasburg, Vir- ginia, near the junction of U.S. Route 11 and State Highway 81, the Edinburg slabby limestones contain thin intercalations of shaly material bearing anomalous occurrences of lower Martinsburg brachiopods and bryozoans that are normally found higher in the sec- tion in a faunule commonly referred to as the Sinuites Beds. The anomalous occurrences of Martinsburg fos- sils in the subjacent Edinburg is understandable if one appreciates the fact that, on the east side of the Mas- sanutten syncline, the base of the Martinsburg litho- facies drops way down in the section and supplants much of the limestone constituting the Edinburg For- mation. The different faunas of the Edinburg and Martinsburg are largely facies-controlled. The occurrence of members of the lower Martins- burg faunule in the Porterfield Stage in the Edinburg Formation is a clear indication that the characteristic Trentonian assemblage actually came into existence during Porterfield time but thrived only within the environmental limits in which Martinsburg elastics were being deposited. Narrow enclaves of shale in the Edinburg Limestone north of Strasburg are logically interpreted as the result of significant shifts in the boundary between the Edinburg lithofacies and the 20 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Martisburg lithofacies during the time when the Mar- tinsburg fauna was making its debut principally— though not entirely—within an area lying to the east of the environment in which the Edinburg Limestone was accumulating. The overlapping of fossil faunas of essentially different ages should be expected as a normal phenomenon; it should not be taken as an indication that fossils do not mean anything. Multiple occurrences of the same faunule in thin intercalations within beds containing different fossils are no less to be expetced than interfingering of facies, and both phenomena point to shifting boundaries of partially contemporary lithofacies. In making such interpretations a much more sophisticated view of de- positional conditions is necessary to comprehend the paleoecological conditions which so often are sensitive to bathymetric changes, such as could result from tidal fluctuations. If we look at the Appalachian succession, we can see that sharp changes in fossils are essentially everywhere controlled by depositional environments of enclosing sediments. If depositional environments change suddenly, faunal changes will be just as sud- den; if the depositional boundaries shift slowly over a period of time, facies intertonguing and repetition of separate biofacies can be expected. Some Concluding Observations The Paleozoic faunal succession in the Appalachian region is distinctive enough in a general way so that, if one appreciates the evolutionary development of even one or two groups of invertebrates and knows when shell features were introduced and how long they survived, the most elemental paleontological informa- tion can save the stratigrapher from making serious mistakes in mapping and correlation. Paleontology does have its weaknesses insofar as stra- tigraphy is concerned. Fossils just simply do not show up as often as the stratigrapher needs them, which is all the time and everywhere. The hapless stratigrapher who follows on the heels of an avid collector more often than not finds little left but unidentifiable scraps. Furthermore, many important sections are based upon road cuts, which are notoriously poor for fossil col- lecting. As the cuts gradually weather, the fossils may begin to show themselves, but by that time the ex- posures are not as complete and original boundaries of lithologic units may be lost. The transient inadequacies of fossils can be overcome largely by using all the lithologic features as well as all the fossils that are available. In my opinion, neither fossils nor lithology alone can be a basis for sound stratigraphy, but together they can provide a sound answer. The local geologic mapper has a great advan- tage over the stratigraphic intruder sent out by his major professor to make a regional study of the beds by sampling them every 20 to 30 miles. The local mapper sees virtually all the exposures where remnant caches of fossils provided by weathering have escaped the less trained eye. The "road-runner" geologist seldom will see these natural caches. The greatest recent breakthrough in paleontology was undoubtedly the technique of etching fossils, first used with tremendous success in the United States by G. A. Cooper. The probable causes for selective silici- fication of fossils in a given rock are not very well understood—not even the time of silicification. But etching has yielded fossil assemblages surpassing any that might have derived by natural forces of weather- ing. Etching has made it possible to resample the same locality for more fossils as long as there is any rock left to collect. Etching of fossils has produced biologi- cally significant numbers of fossils from a single lo- cality for the first time. With large numbers of speci- mens of a given fossil, the paleontologist can pick and choose the best specimens for illustration and descrip- tion; he can determine the range in size and the char- acters of young, mature, and gerontic tests; and, most important of all, he can obtain both interior and ex- terior morphologic features. With this background of information at his disposal, the stratigrapher can make more use of his scraps of fossils or his good single spec- imens. The paleontology of all megascopic fossils is tremendously useful— that is, what there is of it. But the truth is that, in the Appalachians, except for the brachiopods and trilobites, the groups of abundant fossils in the Paleozoic rocks—ostracodes, bryozoans, algae, graptolites, cystids, crinoids, blastoids—are yet to be monographed in modern systematic detail. Even whole groups of brachiopods and trilobites await study. It is hoped that more work can be stimulated in this classic region, where sparsity of substantive data on fossils has so long handicapped stratigraphy. A new way of collecting and describing fossils has come to the fore during the past five years, but its usefulness to the mapping stratigrapher is extremely limited, if not completely lacking. Systematic grab or channel sampling of well-exposed stratigraphic sec- tions in 5-, 10-, or 20-foot intervals, measuring the sec- NUMBER 3 21 tions, and returning to home base for time-consuming digestion of the rock in chloracetic and/or formic acid will yield conodonts in considerable numbers. Conodont paleontology, looked on by many as the panacea of regional stratigraphy, will, without doubt, prove to be less than that. The pragmatic value of conodont paleon- tology is still to be proven. I would be disinclined to view conodonts as of much use to the field stratig- rapher who must make his stratigraphic decisions in the field as he goes about his business of mapping and of geologic interpretation. The value of fossil remains whose biological relationships are yet so poorly known that a military classification for them still prevails cannot have the scientific value of etched megafossils from the same rocks. Conodont paleontology has two great attractions. First, it is relatively quick and one can cover a great deal of section in little field time by systematic samp- ling. Second, there is no denying that teeth from a mobile, cosmopolitan swimmer can be far more widely distributed than the sessile, semisedentary, or passively moved members among invertebrates could possibly be. But they will not and cannot provide the last word on biostratigraphic correlations. The last word will await all the systematic paleontology that has to be done, and which the stratigrapher continues to need. In the fast tempo of today's scientific work, the ten- dency will be for the paleontologist to work on the fossil groups that will provide him with material for the greatest number of papers. It is hoped that some paleontologists will elect the harder and more time- consuming groups of megafossils for detailed study even though their study and monographing will require more preparatory time, more acid, more photography, more plates, and better descriptions to supplement the illustrations. The future progress of stratigraphy will be controlled by the amount of good paleontology that is available for the stratigrapher to use. The common interests of stratigraphy and paleon- tology are rooted in paleoecology, and if we are to make progress in unravelling details of ancient environments both the rocks and their contained fossils must be stud- ied exhaustively. Many of the problems in stratigraphy, and also in paleontology, are far too profound to be fully comprehended by one person. Continuing as- sociations of stratigraphic mappers and competent paleontologists alone can provide the answers to many current geologic problems. The status of our knowledge of Appalachian in- vertebrate fossils is far from satisfactory. The first and only truly exhaustive and comprehensive monograph dealing with an important group of Paleozoic fossils is G. A. Cooper's (1956) "Chazyan and Related Bra- chiopods," which sets a standard of excellence far above that of any other paleontologic monograph on a group of Paleozoic fossils. If treatises of similar na- ture and quality were available on Appalachian Paleo- zoic gastropods, cephalopods, pelecypods, bryozoans, corals, ostracodes, trilobites, and other fossil groups it would be possible to quicken the painfully slow advance of Appalachian Paleozoic stratigraphy and to accom- plish work of considerably higher quality than that now being produced. Literature Cited Butts, C. 1927. The Paleozoic Rocks. In, Geology of Alabama. Alabama Geological Survey Special Report, 14:41— 230, plates 3-76, figures 2-4. 1933. Geologic Map of the Appalachian Valley of Vir- ginia, with Explanatory Text. Virginia Geological Survey Bulletin, 42:1—56. 1942. Geology of the Appalachian Valley in Virginia. Virginia Geological Survey Bulletin, 52(1) : 1—568; 52(2): 1-271, 135 plates. Chadwick, G. H. 1935. Chemung is Portage. Geological Society of America Bulletin, 46(2) :343-354, 2 figures. Cooper, B. N. 1945. Industrial Limestones and Dolomites in Virginia; Clinch Valley District. Virginia Geological Survey Bulletin, 66:72-78. 1960. Systemic Boundaries in the Appalachians. Mineral Industries Journal, 7(4) :5—8. 1961. Grand Appalachian Excursion. In, Geological So- ciety of America Geological Guidebook, 1. Virginia Polytechnic Institute Engineering and Experimental Station Series, Geology Guidebook, 1:1—170, illus- trations, geologic maps. 1964. Relation of Stratigraphy to Structure in the South- ern Appalachians. Virginia Polytechnic Institute, Department of Geological Science Memoir, 1:81— 114. Cooper, B. N., and C. E. Prouty 1944. Lower Middle Ordovician Stratigraphy of Tazewell County, Virginia. Geological Society of America Bulletin, 54:819-886. 5 plates, 3 figures. Cooper, G. A. 1956. Chazyan and Related Brachiopods. Smithsonian Miscellaneous Collections, volume 127: Part 1 (text), xvi+1024 pages; Part 2 (plates), pages 1025-1245, plates 1-269. Cooper, G. A., and B. N. Cooper 1946. Lower Middle Ordovician Stratigraphy of the Shenandoah Valley, Virginia. Geological Society of America Bulletin, 57(1): 35-114, plates 1-3. 22 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Keith, A. 1895. Geologic Description of the Knoxville Sheet. United States Geological Survey Geologic Atlas, folio 16, unnumbered text, maps. Rodgers, J. 1953. Geologic Map of East Tennessee with Explanatory Text. Tennessee Division of Geology Bulletin, 58 (2) : 1-168, 15 plates, 7 figures. Thomas, W. A. 1966. Late Mississippian Fold of a Syncline in the West- ern Appalachians of West Virginia and Virginia. Geological Society of America Bulletin, 77:473— 494, 11 figures. Ulrich, E. O. 1911. Revision of the Paleozoic Systems. Geological So- ciety of America Bulletin, 22:281-680, plates 25-27. Weeks, L. G. 1952. Factors of Sedimentary Basin Development that Control Oil Occurrence. American Association of Petroleum Geologists Bulletin, 36(11) :2071-2124, 26 figures. Wells, D. 1950. Lower Middle Mississippian of Southeastern West Virginia. American Association of Petroleum Geol- ogists Bulletin, 34(5):882-922, 6 figures. Michael R. House The Goniatite Wrinkle- Layer ABSTRACT The wrinkle-layer, a finely striate or ridged structure on the surface of certain goniatite shells, is probably related to some feature of the extended nacreous layer—the ostracum, as broadly interpreted—in the cephalopod animal. Dorsal and ventral structures are demonstrated for some genera in the Anarcestaceae and Pharcicerataceae; only the dorsal wrinkle-layer is known from the Cheilocerataceae and, probably, the Goniatitaceae; restriction to an irregularly patterned structure is found in the Clymeniina. All groups with wrinkle-layer structure are composed of smooth, non- ribbed forms. description and illustration of these structures. Fur- thermore, a review of the occurrence of wrinkle-layer structures among the early ammonoids reveals differ- ences in pattern and arrangement which appear to be of taxonomic importance; also, some of the analogies which have been drawn with the living Nautilus prove difficult to substantiate while others appear more likely. Hence, with some brevity, these matters are discussed here in the hope that more detailed studies and records of them will be made in the future. Be- cause some of the best known examples are Devonian, such occurrences are discussed more fully than later Keyserling, in his classic report on an expedition to the Pechora region in 1843, drew attention to the oc- currence of finely striate or ridged structures upon the surface of some goniatite shells. He aptly compared them with human fingerprints (Keyserling, 1846, p. 274). Subsequently, structures of this type have been described on many Paleozoic ammonoids, and several terms have been applied to them. The best known of such terms is the wrinkle-layer, or Runzelschicht; others include Ritzstreifen, epidermides, couche ridee, stries creuse, and hypostracum. Various analogies have been drawn between these structures and those found in the living Nautilus. In recent years confusion has grown concerning the use and interpretation of many of the terms. This is due partly to the lack of good illustrations of the struc- tures and partly to the difficulty in their interpretation. Perhaps these difficulties explain the omission of all such terms from a glossary of "Morphological Terms Applied to the Ammonoidea" (Arkell et al., in R. C. Moore, 1957, p. L2). The time is clearly ripe for a Michael R. House, Department of Geology, The University of Hull, Hull, England. Historical Review The significant literature on the subject is about a century old. Guido Sandberger (1851), in his neglected paper "Organisation der Goniatiten" has given the best illustrations of the structures in Devonian gonia- tites. The terms Runzelschicht and Ritzstreifen were proposed in 1850 by Guido Sandberger and his brother Fridolin. They applied the former term to the struc- tures found on the dorsal side of the goniatite body chamber, as described in Beloceras, and others which had been noted as early as 1842 by A. d'Archiac and M. E. de Verneuil. The Sandbergers considered the Runzelschicht to be homologous with the black-layer of Nautilus. The term Ritzstreifen, or Einritzung, was applied to the structures seen on internal molds show- ing the flanks and ventral parts of the whorl, as common in Gyroceratites; they thought this analagous with the fine striations seen on the inside of the ventral portions of the shell of Nautilus (Sandberger and Sandberger, 1850, pp. 11, 12). Later, Barrande (1867, p. 23) ex- pressed the opinion that the two structures were but two styles of preservation of the same feature, but he used the terms couche ridee and stries creuse for them. This anomaly he later rationalized (Barrande, 23 24 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY 1877, p. 1189), and used the term epidermides for both. Suess (1870, p. 10) confirmed Barrande's view on the identity of the Runzelschicht and Ritzstreifen but argued that both represented an incomplete nacreous layer. It is curious, therefore, that Foord and Crick (1897, p. xx), who established the anglicized versions, con- sidered the wrinkle-layer to be the equivalent of "the 'black layer' near the aperture of Nautilus . . the Runzelschicht of G. & F. Sandberger and the couche ridee of Barrande." They used the term epidermids for "the Ritzstreifen of G. & F. Sandberger and the stries creuses or epidermides of Barrande" and re- garded them as "reproducing the surface of the man- tle." Few detailed records of the structures in the older goniatites have been made subsequently; it hap- pens that the structure most commonly found in post- Devonian ammonoids is of the Runzelschicht type, and it is this that has been referred to as the wrinkle-layer. Miller, Furnish, and Schindewolf (in R. C. Moore, 1957, p. L12) refer to the structure as the hypostracum, implying that it is distinct from and below the ostra- cum. Ruzhencev (in Orlov, 1962, p. 247), however, expresses the view that wrinkle-layer represents the wrinkled inner surface of the ostracum and that the hypostracum is distinct from it. Review of the Occurrence of Wrinkle-Layer Structures Barrande's conclusion that the Runzelschicht and Ritz- streifen are but two expressions of the same structure is confirmed in the present work. Not all goniatites show both, or either, however. Particularly well-pre- served specimens showing both are instructive. Speci- mens of Manticoceras sinuosum (Hall) preserved as barytic replacements found in the Cashauqua Shale of New York (Plate 1: figures 1, 2) often show the full development. These show that in the body chamber the wrinkle-layer may extend over both the ventral and dorsal parts of the internal surface and that the struc- ture is continuous over these areas (Figure 1). In the dorsal part, the layer overlies the previous whorl and may be referred to as the dorsal wrinkle-layer. In the ventral parts, the layer forms the innermost layer usually preserved, and this may be referred to as the ventral wrinkle-layer. The terms as used here are, for practical purposes, the same as the Runzelschicht and Ritzstreifen, respectively, of the Sandberger broth- ers. In a large number of genera, however, only the dorsal wrinkle-layer is known, and it is conjectural whether a ventral structure was ever developed. It seems preferable to keep some reference to position of development in the terminology, notwithstanding. The term wrinkle-layer is preferred to the.term hypostracum because there still is uncertainty whether the implied homology is the correct one, but this is discussed below. For convenience, the following systematic review conforms to the classification for Palaeozoic ammo- noids as given in the treatise edited by R. C. Moore (1957). ANARCESTACEAE.—So far, examples of the wrinkle- layer have not been described from those most primi- tive Lower Devonian goniatites which, as Erben has so elegantly shown, form the origin of the Ammonoi- dea. The later Emsian and Eifelian genus Gyrocera- tites, however, shows well the ventral wrinkle-layer on pyritic internal molds from the Wissenbacher Schiefer (Plate 1: figure 4; also Sandberger 1851, pi. 3, figs. 23, 25; Erben, 1953, pp. 190, 191). These layers form incised grooves on the mold. The genus shows no im- pressed area, and it is uncertain whether these struc- tures continue on to the dorsal part of the whorl; they are best developed in the body chamber. The dorsal wrinkle-layer has been noted in one specimen of Agoniatites oxynotus Wedekind; the layer is recrystallized and shows little radial pattern (Plate 2: figure 2), but the site is on the dorsum close to the aperture where the structure is frequently weakly and irregularly developed. The structure which has been referred to as the wrinkle-layer in Mimagoniatites fidelis (Barrande, 1865, pi. 8, fig. 20) does not appear to be so; rather, PLATE 1: figures 1, 2.—Manticoceras sinuosum (Hall). Views of the wrinkle-layer on the inside of the umbilical wall, the flanks, and the dorsum of a specimen (in NYSM) preserved as a barytic shell replacement. From a concretionary nodule from near the top of the Cashaqua Shale, Shurtleff's Gully, Livonia, Livingstone County, New York, collected by Professor J. W. Wells. (Both views about X 6.) Figure 3.—Sudeticeras ordinatum Moore. A topotype showing a structure in the shell of the outer whorl, which is pressed against the usual crenulate ornament of the preceding whorl; aperture to right. Specimen from the Visean Pa at Kinckley, Yorkshire; BMc74994 (X 15.) Figure 4.—Beloceras sagittarium (G. & F. Sandberger). The dorsal wrinkle-layer (X 6) on a specimen (figured in Glenister, 1958, pi. 14, fig. 2) from the Frasnian of the Fitzroy Basin, Western Australia. NUMBER 3 25 PLATE 1 26 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY it is reminiscent of the structure figured here under Sudeticeras (Plate 1: figure 3). A feature that may be the ventral wrinkle-layer is illustrated for Anarcestes praecursor (Freeh) by Bar- rande (1865, pi. 7, fig. 9), showing transverse and ir- regular striae crossing the outer flanks and venter. Illustrations are available (Holzapfel 1895, pi. 6, fig. 6b) for the dorsal wrinkle-layer of Maenioceras terebratum (G. & F. Sandberger). A specimen figured here (Plate 2: figure 5) shows it developed through much of the body chamber, and it has been recorded, with a diameter of about 40 millimeters, as being slightly prosiradiate and continuing to the position of the base of the apertural lappet (House, 1962, p. 269, pi. 43, fig. 8). PROLOBITACEAE.—The only representative known to show the wrinkle-layer is Sobolewia, and the writer has doubts that this is the correct taxonomic position. In S. virginiana House (Plate 2: figure 1) the dorsal wrinkle-layer has been observed for a quarter-whorl on the dorsal side of the apicad part of the body chamber. The course bears no relation to the course of the growth lines (the rule), and the lirae are firm and continuous, branching occasionally. VENTRAL WRINKLE LAYER DORSAL WRINKLE LAYER PHARCICERATACEAE.—Some comment has already been made on the contguity of the dorsal and ventral wrinkle-layer as shown in well-preserved specimens of Manticoceras (Figure 1; Plate 1: figures 1, 2). The ventral wrinkle-layer shows frequently on both the EXTERNAL CALLUS OSTRACUM OF PRECEDING WHORL OSTRACUM phragmocone and body chamber of pyritic molds; for example, on Budesheim specimens. Illustrations have been given for several species (G. Sandberger, 1851, pi. 3, figs. 19, 21, 23, 24; Sandberger and Sandberger, 1850, pi. 7, fig. 9). The pattern ventrally takes the form of almost rectiradiate striae, but with a slight forward sweep in some specimens and with a slight backward sweep in others. At least one species group can be distinguished by the wrinkle-layer. The M. affine group is characterized by a much finer pattern than M. cordatum and its allies. Ponticeras genundewa (Clarke) from the Genundewa Limestone of New York occasionally shows the wrinkle-layer in pyritic forms (e.g., NYSM x 3645) as close-set, rather prorsi- radiate striae on the lower flanks at diameters of less than 5.0 millimeters. Forms from the Cashaqua Shale barytized horizon that have been referred to Probelo- ceras by the writer show the dorsal wrinkle-layer par- ticularly well (Figure 2B; Plate 2: figure 6) but no ventral structure has been observed. These specimens show that the layer terminates near the aperture. Keyserling (1844, pi. B, fig. 1) illustrated a similar structure in Ponticeras (?) uchtense. In Timanites keyserlingi (Miller) the dorsal wrinkle-layer striae have been described as "passing slightly backwards from the umbilicus" and continuing "in an irregular rectilinear course towards the venter, frequently bi- furcating" (House and Pedder, 1963, p. 526, pi. 75, fig. 5). In the Pharciceratidae, G. Sandberger (1851, pi. 3, fig. 36) illustrated the structure of the dorsal wrinkle- layer in Pharciceras lunulicosta. In this specimen the the lirae pass across the flanks and are slightly prorsi- radiate, the lirae numbering some 14 per millimeter, and without the spiral tendency close to the umbilical seam as shown, for example, in Beloceras and Probelo- ceras. The striae are coarser close to the seam, however. Among the Beloceratidae, the structure in Beloceras is best known (d'Archiac and de Verneuil, 1842, pi. 26, figs. 7a, 8; G. Sandberger, 1851, pi. 3, fig. 35; Glenister, 1958, pi. 5, fig. 12, and pi. 14, fig. 2). The writer is indebted to Professor Glenister for supplying the negatives for the enlargements shown here (Plate 1: figure 4; Plate 3: figure 9). Only a dorsal wrinkle- layer has been observed and, again, it terminates at FIGURE 1.—The relationship between the dorsal and ventral wrinkle-layers, based on a specimen of Manticoceras sinuosum (Hall) from the Cashaqua Shale of New York (see Plate 1: figures 1, 2). 1 New York State Museum, Albany. Other abbreviations used in this paper: BM, British Museum (Natural History) ; USNM, United States National Museum. NUMBER 3 27 FIGURE 2.—The dorsal wrinkle-layer pattern in certain genera, based on specimens discussed in the text: A, Sobolewia, Givetian; B, Probeloceras, Frasnian; c, Beloceras Frasnian; D, Tornoceras, Middle and Upper Devonian. the aperture. A diagrammatic representation is illus- trated in Figure 2c. CLYMENIINA.—Giimbel, in refiguring most of Minister's species, illustrated the dorsal wrinkle-layer in a number of genera, including Cymaclymenia, Clymenia and ?Cyrtocylmenia sp. (Giimbel, 1863, pi. 15, fig. 6e) ; and in these the pattern is distinctly irregular. The figures for Kosmoclymenia undulata (Minister) illustrated by Giimbel (1863, pi. 17, fig. Ik) show an irregular transverse arrangement near the dorsal line associated with a raised band along the midline. This arrangement also has been observed by the writer in Devon specimens of Cymaclymenia. The ventral wrinkle-layer has not been described in this group, but it seems that the dorsal layer tends to be best developed close to the dorsal midline. CHEILOCERATACEAE.—So far, only the dorsal wrinkle-layer has been met with in the Tornoceratidae, notwithstanding the large variety in the preserved specimens examined. The pattern is, however, remark- ably distinctive, characterized by lirae which spiral backward from the umbilicus, become radial or slightly backwardly directed across the flanks, and are dis- rupted or vaguely spiral on the venter (of the preced- ing whorl). This pattern is illustrated diagrammati- cally in Figure 2D. The best illustrations are of Torno- ceras uniangulare obesum from the Cashaqua Shale barytized fauna (House, 1965, pi. 8, fig. 72, and pi. 9, figs. 78, 79). Other illustrations have been given by Whidborne (1890, pi. 6, fig. 3a), Keyserling (1844, pi. A, fig. 5d), Sandberger and Sandberger (1852, pi. 10, fig. 14), and G. Sandberger (1853, pi. 3, fig. 37). In this genus also it can be demonstrated that the dorsal wrinkle-layer is not restricted to the body cham- ber. This is shown at least by T. arcuatum from the Squaw Bay Limestone of Michigan (House, 1965, pi. 8, fig. 7; it may be remarked that the writer never intended his recognition of specimens conforming to this species in the Tully Limestone to imply a correla- tion with that formation, the Squaw Bay would seem to be slightly younger, but older than the Genundewa). The same distinctive pattern is shown by Parodiceras. The only remaining genus of this family in which the wrinkle-layer has been described is Pseudoclymenia, which was illustrated by G. Sandberger (1853, pi. 8, fig. 4b) under the name Clymenia pseudogoniatites. In the Cheiloceratidae the pattern seems the same as in Tornoceras. It has been recorded in Cheiloceras (House and Pedder, 1963, p. 530), but only the dorsal wrinkle-layer is known. If Neoaganides belongs in this family, then Miller and Furnish (1957, pi. 131, fig. 4) have illustrated another example. GONIATITACEAE.—Knowledge of the wrinkle-layer in Carboniferous goniatites suffers from frequent men- tion but infrequent description. Some records are as follows: Beyrichoceras (de Koninck, 1880, pi. 49, fig. 10; Bisat, 1924, p. 83; 1934, p. 293; Gordon, 1957, p. 41); Beyrichoceratoides (Currie, 1954, p. 553); Munsteroceras (Gordon, 1957, p. 35) ; "Pericyclus" (Crick, 1899, p. 432, fig. 2; Foord 1901, p. 145) ; Gir- tyoceras (E. W. J. Moore, 1945, p. 414). Certain struc- tures apparently connected with the ostracum are il- lustrated here in Sudeticeras (Plate 1: figure 3), but such structures are not comparable with the wrinkle- layer (and the same may be true of other records). Some of the finest illustrations of the dorsal wrinkle- layer occur in Platygoniatites described by Ruzhencev (1956; 1963; in Orlov, 1962, p. 247, figs. 2a,b) ; they show backwardly spiralling striae close to the umbilicus becoming more radial outwards, a course shown also in Tornoceras. DIMORPHOCERATACEAE.—Ruzhencev (1956) has recorded a wrinkle-layer in Delepinoceras (which he would place in the Goniatitaceae). PROLECANITACEAE.—There appears to be an occur- rence reported by Foord (1903, p. 205, pi. 48, fig. 4a) in Merocanites compressus (Sowerby). 28 MEDLICOTTIACEAE.—A wrinkle-layer is noted by Gordon in ?Pronorites, but the structure on a figure of Pronorites ludfordi (Bisat, 1957, pi. 3, fig. 4) is a fingerprint. Miller and Furnish (1947, pp. 2, 44) have commented on the structure in Artinskia as follows: "fine lines which are not parallel to the growth lines; each forms a deep narrow rounded sinus with parallel sides on the ventral portion of the conch and a promi- nent rounded salient on the flanks." Of the stratigraphically younger occurrences of wrinkle-layer structures, little can be reported. The Jurassic records have not been confirmed, and their general absence from ribbed ammonoids is perhaps significant. There are a number of Triassic records for which Mojsisovics (1893) is a primary source, although some of his attributions may be questioned. Two good examples showing a dorsal wrinkle-layer are illustrated here—Arcestes (Proarcestes) and Owenites (Plate 3: figures 1, 5-7). These suggest that a system- atic survey might be profitable. Generalizations From the foregoing references and illustrations several general comments can be made. Firstly, the only groups in which both dorsal and ventral wrinkle-layers are authenticated are the Anarcestaceae and Pharcicera- taceae; and in both of these groups the dorsal wrinkle- layer striae have a nearly radial or forwardly project- ing course. Secondly, in these groups the structure is well enough known to demonstrate that certain species groups have distinctive styles of wrinkle-layer striae. Thirdly, in the Cheilocerataceae only the dorsal wrin- kle-layer is known, and it contrasts with the preceding in that the striae are backwardly directed close to the umbilical seam, although they become more radial out- ward. Fourthly, the Clymeniina show a restriction of the structure, the dorsal midline is commonly asso- ciated with a raised band and the pattern is irregular. Fifthly, goniatitacean structures, whilst still poorly known, seem to agree with those of the Cheilocera- taceae; that is, with their ancestors as generally in- terpreted. Sixthly, all groups showing the wrinkle-layer are smooth, nonribbed forms. It has been demonstrated that the structure when complete continues to the aperture on the inside sur- face of the body chamber, but that the course of the striae bears no relation to that of the growth lines or form of the aperture. The structure is often found SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY on the early whorls and phragmocone and is not a feature of maturity alone. Interpretation It is appropriate to discuss briefly the structures in the living Nautilus that might be homologous. More point is given to this when it is recalled that fossil nautiloids also show structures which have been de- scribed as the wrinkle-layer (for example, see Plate 2: figure 3). Six structures in Nautilus have been con- sidered to be homologous with the wrinkle-layer: (1) The black layer (G. Sandberger, 1851, pars). (2) Reproduction of the surface of the mantle (Foord and Crick, 1897). (3) Post annular lirae (G. Sandberger, 1851, pars). (4) The ostracum (Ruzhencev, 1963; Teichert, in R. C, Moore, 1964, pp. K13, K15). (5) Nacreous layer (Suess, 1870). (6) Hypostracum (Miller and Furnish, in R. C. Moore, 1957, p. L12). There is, among these recorded opinions, a no- menclatorical confusion, for, as Stenzel (in R. C. Moore, 1964, p. K77) has pointed out, in some usages PLATE 2: figure 1.—Sobolewia virginiana House. Dorsal wrinkle-layer on the posterior part of the body chamber (the last chamber is shown) ; aperture to right, umbilicus below. Specimen from the Millboro Shale (Bed 3 of Butts, 1940, p. 311) 0.4 mile due north of Hayter's Gap, Virginia; USNM 137650 (X 9). Figure 2.—Agoniatites oxynotus Wedekind. Dorsal wrinkle- layer on the posterior part of the body chamber; orad direc- tion to left. Specimen from the Middle Devonian of Ober- scheld, Germany; Berlin Museum c401 (X 4.6). Figure 3.—Vestinautilus sp. The dorsal wrinkle-layer (be- low) and ventral wrinkle-layer (above) joining across the site of the umbilicial seam; orad direction to left. From the Lower Carboniferous of the Askeaton Limeworks, County Limerick, Eire. BM c74996; (X 1). Figure 4.—Gyroceratites gracilis (Bronn). An internal mold of the body chamber showing the ventral wrinkle-layer; orad direction to left. From the Wissenbacher Schiefer, Dillenburg, Germany. BM c55963 (X 8). Figure 5.—Maenioceras terebratum (G. & F. Sandberger). The dorsal wrinkle-layer in the posterior part of the body chamber; outer part of the body chamber, with growth lines, is shown above. From the Brilon Ironstone, Martenburg, Adorf, Germany. BM c74995 (X 4). Figure 6.—Probeloceras aff. lutheri (Clarke). The dorsal wrinkle-layer in the body chamber (last septum is at top right). Specimen is a barytic replacement, probably from the upper part of the Cashaqua Shale near Honeoye Lake, New York; NYSM 4063 (X 9). NUMBER 3 29 PLATE 2 30 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY 10 PLATE 3 NUMBER 3 31 the ostracum can include both the nacreous layer and the hypostracum. The analogy with the black layer has little to com- mend it, for, whilst it is a line of weakness in Nautilus, it has no linear structure, and there is no dominant de- velopment on the ventral surface. Also, Foord and Crick's interpretation is difficult to substantiate. Simi- larly, correspondence with postannular lirae is im- probable in view of the restriction of that structure to the posterior part of the body chamber (and a posterior annular elevation has been described also in a number of ammonoids by Crick, 1898). This, then, leaves the ostracum in its broad interpretation. It is, of course, the dorsal nacreous layer which corresponds to the posi- tion of the dorsal wrinkle-layer. As is well-known, dis- solution of Nautilus ostracum produces a wrinkled etched surface (Miller 1947, p. 20), as shown in Plate 3 (figures 8, 9). The dorsal nacreous layer is in con- tinuation with the ventral nacreous layer—the inner part of the outer ostracum as used by Stenzel (in R. C Moore, 1964, p. K78)—and, hence, continuation be- tween a dorsal and a ventral wrinkle-layer could be expected. But dissolution is not wholly satisfactory as it has been shown that the wrinkle-layer structures can be preserved in the phragmocone (where they might be less subject to dissolution), and the pyritic preserva- tions would demand dissolution before the internal molds were formed. PLATE 3: figures 1, 7.—Arcestes (Proarcestes) hauieli (Welter). The dorsal wrinkle-layer. Specimen from the Triassic of Timor (figured in Welter, 1914, pi. 29, fig. 6) ; in Bonn Geologisch und Palaeontologisch Institut. (Figure 1, X 1; figure 7, X 10.) Figures 2-4.—The aperture of a supposed goniatite ( X 6) from the Frasnian of Biidesheim, Germany, collected by the writer from outcrops 400 meters south of the village church. Figures 5, 6.-—Owenites egrediens Welter. Specimen (fig- ured in Welter, 1914, pi. 14, fig. 24) showing the dorsal wrin- kle-layer; in Bonn Geologisch und Palaeontologisch Institut. (Figure 5, X 1; Figure 6, X10.) Figures 8, 9.—Nautilus pompilius Linne. Views (X4) of two stages in the dissolution of ostracum from the ventral part of the body chamber during etching in dilute hydrochloric acid. Figure 10.—Beloceras sagittarium (G. & F. Sandberger). View (X7.5) showing the dorsal wrinkle-layer close to the aperture on a specimen (figured in Glenister, 1958, pi. 5, fig. 12) from the Frasnian of the Fitzroy Basin, Western Australia. It was considered that there might be a relation between the apertural "pallial line," inferred by the writer on another occasion to be close to the goniatite aperture (House, 1960). But there is little evidence to relate this to the course of the ventral wrinkle-layer striae. A chance find of a curious aperture at Biidesheim (Plate 3: figures 2-4), possibly of Tornoceras (it shows traces of growth lines with ventrolateral lappets), in- dicates how little it known of apertural conditions and emphasizes how uncertain any interpretations must remain. That the wrinkle-layer reflects in some way structures in some part of an extended nacreous layer, however, seems probable. Literature Cited d'Archiac, A., and M. E. de Verneuil 1842. On the Fossils of the Older Deposits of the Rhenish Provinces. Transactions of the Geological Society of London, series 2, 6: 303-410. Barrande, J. 1865- Systeme silurien du centre de la Boheme, Premiere 1877. Partie. Recherches paleontologiques, vol. 2, Classe des Mollusques, Order des Cephalopodes. Parts 1-9 and supplement. Prague. Bisat, W. S. 1924. The Carboniferous Goniatites of the North of England and Their Zones. Proceedings of the York- shire Geological Society, new series, 20: 40-124, plates 1-10. 1934. The Goniatites of the Beyrichoceras Zone in the North of England. Proceedings of the Yorkshire Geological Society, new series, 22: 280-309, plates 17-24. 1957. Upper Visean Goniatites from the Manifold Valley, North Staffordshire. Palaeontology, 1: 16-21, plates 3,4. Butts, C. 1940. Geology of the Appalachian Valley in Virginia. Part 1, Geologic text and illustrations. Virginia Geological Survey Bulletin, 52(1): 1—568, 63 plates, 10 figures. Crick, G. C. 1898. On the Muscular Attachment of the Animal to Its Shell in Some Fossil Cephalopoda (Ammonoidea). Transactions of the Linnaean Society of London (Zool.), 7:71-113. 1899. On Some New or Little Known Goniatites from the Carboniferous Limestone of Ireland. Annals and Magazine of Natural History, series 7, 3: 429- 454. Currie, E. D. 1954. Scottish Carboniferous Goniatites. Transactions of the Royal Society of Edinburgh, 67: 527-602, plates 1—4. 32 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY de Koninck, L.-G. 1880. Fauna du .Calcaire Carbonifere de la Belgique, Parte 2. Museum royale d'Histoire naturelles Bel- gique Annates, 5: 1-133. Erben, H. K. 1953. Goniatitacea (Ceph.) aus dem Unterdevon und den Unteren Mitteldevon. Neues Jahrbuch fur Geologie und Palaontologie Abhandlungen, 98:175-225, plates 17-19. Foord, A. H. 1897- Carboniferous Cephalopoda of Ireland. Palaento- 1903. graphical Society Monographs, 234 pages, 49 plates. Foord, A. H, and G. C. Crick. 1897. Catalogue of the Fossil Cephalopoda in the British Museum (Natural History). Part 3, xxxiii + 303 pages, London. Glenister, B. F. 1958. Upper Devonian Ammonoids from the Manticoceras Zone, Fitzroy Basin. Journal of Paleontology, 32:58-96, plates 5-15. Gordon, M., Jr. 1957. Mississippian Cephalopods of Northern and East- ern Alaska. United States Geological Survey Profes- sional Paper 283, 61 pages, 6 plates. Gumbel, C. W. 1863. Ueber Clymenien in den Uebergangsgebilden des Fichtelgebirges. Palaeontographica, 11:85-165, plates 15-21. House, M. R . 1960. Abnormal Growths in Some Devonian Goniatites. Palaeontology, 3: 129-136, plate 24. 1962. Observations on the Ammonoid Succession of the North American Devonian. Journal of Paleontol- ogy, 36: 247-284, plates 43-48. 1965. A Study in the Tornoceratidae: The Succession of Tornoceras and Related Genera in the North American Devonian. Philosophical Transactions of the Royal Society of London, series B, 250: 79—130, plates 5—11. House, M. R., and A. E. H. Pedder 1963. Devonian Goniatites and Stratigraphical Correla- tions in Western Canada. Palaeontology, 6: 491— 539, plates 70-77. Keyserling, A. 1844. Wissenschaftliche Beobachtungen auf einer Reise in das Petschora-Land im Jahre 1843. 465 pages, 22 plates. St. Petersburg. 1846. Beschreibung einiger Goniatiten aus dcm Domanik Schiefer. Russisch-Kaiserliche Mineralogische Ge- sellschaft, Verhandlungen, pages 217-238, plate A. Miller, A. K. 1947. Tertiary Nautiloids of the Americas. Geological Society of America Memoir, 23: 1—234, plates 1— 100. Miller, A. K, and W. M. Furnish 1957. Permian Ammonoids from Southern Arabia. Journal of Paleontology, 31: 1043-1051, plates 131-132. Mojsisovics, E. 1893. Die cephalopoden der Hallstatter Kalke. Kaiser- lich-konigliche Geologische Reichsansta.lt, Ab- handlungen, 6(2) : 1-835, plates 71-200. Wien. Moore, E. W. J. 1945. The Carboniferous Genera Girtyoceras and Eumorphoceras, Proceedings of the Yorkshire Geological Society, 25: 387-445, plates 22-27. Moore, R. C. (editor) 1957. Treatise on Invertebrate Paleontology. Part L, Mol- lusca 4. 490 pages, 558 figures. New York City and Lawrence, Kansas: Geological Society of America and University of Kansas Press. 1964. Treatise on Invertebrate Paleontology. Part K, Mol- lusca 3. 519 pages, 361 figures. New York City and Lawrence, Kansas: Geological Society of America and University of Kansas Press. Orlov, Y. A. (editor) 1962. Osnovy Paleontologii, 5, Mollyuski: Golovonogie 1. 438 pages, 32 plates. Moscow: Akademii Nauk SSSR. Ruzhencev, V. E. 1956. On Some New Genera of Ammonoids. [In Russian.] Akademii Nauk SSSR, Doklady, new series, 107: 158-161. 1963. On the Position of the Southern Ural Genus Dele- pinoceras. [In Russian.] Akademii Nauk SSSR, Doklady, new series, 122:293-296. Sandberger, G. 1851. Organisation der Goniatiten. Nassauischer Verein fur Naturkinde, Jdhrbucher (1851), pages 1—15, plates 3, 4. 1853. Einige Beobachtungen iiber Clymenien; mit besonderer Rucksicht auf die Westphalischen Arten. Naturhistorischer Verein der Rheinlande Verhand- lungen, 10:1—46, plates 6-8. Sandberger, G., and F. Sandberger 1850- Systematische Beschreibung and Abbildung der 1856. Versteinerungen des Rheinischensystems in Nassau. 136 pages, 18 plates (in atlas). Wiesbaden. Suess, E. 1870. Uber Ammoniten. Kaiserliche Akademie der Wis- senschaften, Mathematische-naturwissenschaftliche Klasse, Sitzungsberichte, 52:7-89. Welter, O. A. 1914. Die Obertriadischen Ammoniten und Nautiliden von Timor. In J. Wanner, Palaontologie von Timor, vol. 1, 258 pages, 36 plates. Whidborne, G. F. 1890. A Monograph of the Devonian Fauna of the South of England. Part 2. Palaeontographical Society Monographs, pages 47—154, plates 5—15. Valdar Jaanusson Evolution of the Brachiopo d Hinge ABSTRACT In articulate brachiopods the hinge-teeth grow either by simple addition of shell material distally (deltidio- dont hinge-teeth) or by means of a complex process of secretion distally and resorption proximally (cyrto- matodont hinge-teeth). It is shown that the mode of formation of the hinge-teeth influences the construc- tion and function of the whole hinge mechanism as well as the growth of the posterior margin of the shell. During evolution of the hinge a minor and a major functional discontinuity had to be overcome. The minor functional threshold or instability was at the change from strophic to nonstrophic shells with delti- diodont dentition. The major threshold had to be passed when resorption in forming the hinge-teeth was introduced. Some evolutionary implications of the process of passing these functional thresholds are discussed. The groups of articulate brachiopods with deltidio- dont and with cyrtomatodont dentition correspond to Beecher's subdivisions Protremata and Telotremata, respectively. It is suggested that the use of these sub- divisions should be reintroduced, for example, as sub- classes of the Class Articulata. In the past, little attention has been paid to the morphology, mode of growth, and mechanism of the articulating structures in articulate brachiopods. This is astonishing, because the general morphology and microstructure of the shell have been studied in con- siderable detail in this group. The purpose of the present paper is to draw atten- tion to the importance of the articulating mechanism in the function and construction of the articulate brachiopod shell. The observations included herein summarize some of the results on the functional mor- phology of the brachiopod shell which the writer has Valdar Jaanusson, Natu-rhistoriska Risksmuseet, Paleozoologiska Sektionen, Stockholm 50, Sweden. carried out during many years. An early version of the work was presented at the Symposium on Early Palaeozoic Invertebrates held in Warsaw in 1961. In this paper, attention is focused on dentition. Other components of the articulating mechanism and asso- ciated structures will be treated elsewhere. For this study it was necessary to have access to a wide variety of articulate brachiopod groups with specimens showing internal morphological details. Al- though the paper is mainly based on material de- posited in the Departments of Palaeozoology and Invertebrate Zoology, Swedish Museum of Natural History, Stockholm, many additional specimens were examined in other collections. The opportunity, in 1959, to examine the world's largest collection of brachiopods—in the United States National Museum, Washington, D. C.—has been of great value, and the writer is greatly indebted to Dr. G. A. Cooper for his hospitality and generous help. Dr. T. W. Amsden, Oklahoma Geological Survey, Dr. H. A. Lowenstam, California Institute of Tech- nology, and Dr. A. Romusoks, Tartu University, kindly placed specimens pertinent to this study at the writer's disposal. Thanks also are due the officials of the De- partment of Invertebrate Zoology, Swedish Museum of Natural History j Zoological Museum, Uppsala Uni- versity; and Laboratoire Arago, Banyuls-sur-Mer (through Dr. H. Mutvei, Stockholm) for material of modern brachiopods. Further acknowledgment is made to Dr. H. Neu- haus, Institute of Physics, Stockholm University, for valuable suggestions; to Mr. G. Andersson, Institute of Palaeontology, Uppsala University, for photo- graphs; to Mr. E. Stahl, of the same institute, for con- structing models of articulating mechanism; to Mr. B. Bliicher, Department of Palaeozoology, Swedish Museum of Natural History, for drawing text figures; and to Dr. D. Bruton, Palaeontological Museum, Oslo, for linguistic revision of the manuscript. 33 34 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY All figured specimens, with the exception of the one shown on Plate 2: figure 10, belong to the Department of Palaeozoology, Swedish Museum of Natural His- tory. For phototechnical reasons the ventral valves are consistently orientated with the ventral or posterior side up, and the dorsal valve with the dorsal or pos- terior side down. In order to facilitate comparison, this orientation of the valves is maintained also in the text figures. The Hinge Structures The main components of the hinge in articulate brachiopods normally are a pair of ventral hinge-teeth which accommodate into dorsal hinge-sockets and a pair of dorsal shell projections which, in certain groups, fit into a groove (crural fossette of Kozlowski, 1927) at the base of the hinge-teeth. The terminology used for the dorsal positive element of the hinge differs in various groups of articulate brachiopods (brachio- phore, socket ridge, inner socket ridge). As this paper deals with articulate brachiopods in general, the use of different terms for the same structure is confusing. This difficulty has provisionally been solved here by the use of the descriptive term "dorsal hinge process." In some groups of articulate brachiopods with a hinge- line, the articulation becomes simplified secondarily by reduction of normal hinge structures; limitation of space prevents a further treatment of these modifications. In addition to the main articulating devices there occur accessory hinge structures, the main function of which seems to be to direct the free valve edges when the shell closes. Such structures are not further treated in this paper. Examination of articulated shells of modern brachio- pods (rhynchonellaceans, terebratulaceans, terebratel- laceans, and thecideaceans) reveals that the articu- lating structures serve four main functions. (1) They make possible a rotating movement of the valves rela- tive to each other. In nonstrophic shells (Rudwick, 1959) they determine the position of the axis of rota- tion; in strophic shells the axis of rotation coincides with the hinge-line, and the construction of the articu- lating devices is such that this position of the axis of rotation is maintained at all degrees of opening or closing of the shell. (2) They prevent the transverse (lateral) and longitudinal movement of the valves rela- tive to each other. (3) When the shell is being closed they direct the valves so that the free valve edges can fit snugly. (4) They limit the extent to which the shell can be opened. In an articulated shell it is not possible to move the valves sideways or backwards and forwards relative to each other without damaging some of the hinge structures. When the valves are forced apart beyond the limit allowed by the articulating devices, some of these structures break (Thomson 1927, p. 77). These functions are integrated in the hinge mecha- nism, and all seem to be important to the animal; thus, an effective hinge mechanism would fulfill all these functions. When dealing with extinct groups, study of the functions of the articulating mechanism imposes problems because, as a rule, it is not possible to examine the hinge in action. In some cases simple models, con- structed on the basis of serial sections, have been useful. In this paper the functions are discussed separately for each type of dentition. Growth of the Hinge-Teeth FIGURE 1.—Posterior part of ventral valves of strophic shells with deltidiodont hinge-teeth, A, Hesperorthis davidsoni (Ver- neuil) ; B, Resserella elegantula (Dalman). There are two main types of growth of the hinge-teeth. In one large group of articulate brachiopods the hinge- teeth grow as simple projections of the shell by addition of shell material distally (Figure 1; Plate 1: figures 1, 3, 5). The entire secreted hinge-tooth substance is fully preserved from the apex of the valve to the functional hinge-teeth. This kind of hinge-teeth is here termed the "deltidiodont type" (alluding to the often triangular, delta-like track of growth on the interarea). In the other group of articulate brachiopods the hinge-teeth either are knoblike (Figure 2A; Plate 2: figures 5, 7, 10), rising directly from the floor of the valve, or they are somewhat hook-shaped with a posteromedially protruding process (Figure 2B; Plate 2: figures 1, 8, 9). NUMBER 3 35 FIGURE 2.—Posterior part of ventral valves with cyrtomato- dont hinge-teeth, A, Hemithyris psittacea (Gmelin), non- strophic, with knoblike hinge-teeth; B, Eospirifer radiatus (Sowerby), strophic, with hook-shaped hinge-teeth. Hinge-teeth of such shapes cannot possibly grow with- out a continuous resorption of the previously secreted tooth substance behind the protruding portion of the hinge-tooth. For this type of hinge-teeth the term "cyrtomatodont" (from the Greek word meaning "knob") is here proposed. The deltidiodont hinge-teeth characterize (in the classification of Williams and Rowell, in Moore, 1965) all members of the orders Orthida, Strophomenida (except those groups in which the hinge-teeth have become secondarily reduced), and Pentamerida. Fur- ther exceptions are Tropidoleptus, Perditocardinia, and Cadomella, which have hinge-teeth of the cyrtomato- dont type. In the writer's opinion, these three genera should be classified, respectively, with the Terebratu- lida, Rhynchonellida, and Spiriferida (see section on classification, below). In a few other genera of some- what doubtful taxonomic position—such as Enantio- sphen and Thecospira—the dentition is poorly known. Among groups with deltidiodont hinge-teeth, only two known examples indicate the use of resorption during the growth of a shell structure. In Dicoelosia the hook-shaped dorsal hinge process obviously requires resorption for its growth (Wright, 1968), and there is some doubt whether or not some other advanced en- teletaceans formed dieir dorsal hinge process in the same manner. In Porambonites the pedicle foramen 372-386 0—71 4 became enlarged during the ontogenetic development, and, here again, resorption evidently was involved. Apart from these exceptions, brachiopods possessing deltidiodont hinge-teeth do not seem to have acquired the ability to use resorption for the construction of their shells. A few specimens exhibit a hook-shaped hinge- tooth (or similar features of other internal structures) which could not possibly grow and preserve this shape if resorption had not been involved. Such examples that are known to the writer, however, obviously belong to gerontic individuals which have ceased to grow, and most specimens of the same species do not show these features. All members of the orders Rhynchonellida, Spiri- ferida, and Terebratulida and of the suborder Theci- deidina possess cyrtomatodont hinge-teeth. In Spiri- ferida and Terebratulida resorption also is widely used for the construction of other shell structures, such as brachidia, loops, and pedicle foramina. The extent of resorption during the growth of the cyrtomatodont hinge-teeth varies within wide limits. In most instances, previous growth stages of a hinge- tooth have been completely removed by resorption, and the hinge-tooth then has the shape of a knob which rises directly from the floor of the valve (Figure 2A; Plate 2: figures 5, 7, 10). In other cases only the shell material behind the posteromedially protruding por- tion of the hinge-teeth is resorbed, and the rest of the once-secreted hinge-tooth is preserved; the hinge- teeth then have a hooklike appearance (Figure 2B; Plate 2: figures 1, 8, 9). The hooklike hinge-tooth is associated with strophic shells such as those of the Spiriferacea (Figure 2B) and Thecideacea (Plate 2: figures 8, 9) but occasionally it also occurs in non- strophic shells. It is interesting that very small shells of certain rhynchonellaceans possess hooklike hinge- teeth (Plate 2: figure 11) which, during the onto- genetic development, gradually change into knob- shaped hinge-teeth (Figure 2A) . Gradations seem to exist between these two shapes of cyrtomatodont hinge- teeth and, thus, gradations in the degree of resorption used for producing hinge-teeth of this type. Articulation in Strophic Shells with Deltidiodont Hinge-Teeth In the Orthida, Strophomenida, and Porambonitacea (excluding Camerellidae, Parastrophinidae, and Bre- vicameridae, which in this paper are tentatively in- cluded in a separate superfamily, the Camerellacea) 36 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY the hinge-teeth accommodate into dorsal hinge-sock- ets. As in articulate brachiopods in general, the hinge- sockets cannot be formed as simple pits in the floor of the valve as the thinness of the test sets a definite limit to the depth of such pits; consequently, the main part of the hinge-sockets has to be constructed as a kind of pouch upon the floor of the valve. In the brach- iopods under consideration, the posterior shell wall is used as the posterior wall of the hinge-socket and the anteromedian wall of the socket is formed by the secreting of a ridge, plate, or boss upon the inner sur- face of the valve (the "brachiophore" of Schuchert and Cooper, 1932; the "socket ridge" of Williams, 1953; and the "dorsal hinge process," as in the present paper). In certain groups the posterolateral surface, corner, or edge of this process accommodates into a depression or groove ("fossette crurale" of Kozlowski, 1927) at the base of the hinge-teeth; other groups, such as clitambonitaceans, strophomenaceans, and triplesi- aceans, lack crural fossette. In some plectambonitac- ean genera the crural fossette is so deep that the hinge-tooth appears to be bifid. Laterally the hinge- socket has no wall, and the reason for this is clear: if such a wall were present in the direction of growth of the hinge-socket it would have to be continuously resorbed and secreted during the growth of the hinge- teeth. All brachiopods under consideration have a strophic shell, which means that they have a hinge-line (Rud- wick, 1959). Continued existence of the hinge-line during the growth of the shell produces an interarea, at least ventrally. Contrary to the opinion of Rudwick (1959, p. 21), Porambonites has a well-defined hinge- line, although it is very narrow. In these shells the position of the axis of rotation was fixed along the hinge-line. During opening or closing of the shell the anterodorsal surface of the hinge-teeth glided upon the floor of the hinge-socket. The interlocking arrange- ment of the hinge-teeth and the dorsal hinge processes prevented lateral and longitudinal movements of the valves relative to each other and, during closing of the shell, directed the valves so that their edges would fit snugly. One of the main functions of this tight inter- locking arrangement was to maintain the position of the axis of rotation exactly along the hinge-line at all degrees of opening or closing of the shell. In this kind of articulation it is difficult to find a mechanism which effectively controlled the maximum degree of opening of the shell; it is probable that, as in inverte- brate brachiopods, this function has been exerted by the muscles. Some few members of the group under considera- tion developed structures which could, and possibly did, control the maximum width of the gap between the valves. In Dicoelosia the shell could be opened until the unique, hook-shaped dorsal hinge process became pressed against the base of the hinge-teeth. In Porambonites the distance between the beaks of the valves is short, and a relatively narrow gap be- tween the valves would bring the beaks into contact and thus prevent a further increase of the width of the gap- Articulation in Nonstrophic Shells with Deltidiodont Hinge-Teeth The pentameraceans and camerellaceans that I ex- amined do not possess a hinge-line. It has been re- peatedly stated, for example, that Stricklandia and Costistricklandia are provided with an interarea, at least on the ventral valve (Schuchert and Cooper, 1932; Rudwick, 1959; Amsden, in Moore, 1965; and others). In these genera, however, the plane areas on either side of the delthyrium or cardinalia could not have been formed by the growth of a hinge-line be- cause the anterior margin of these areas is not straight (Plate 1: figures 1,2). Moreover, in these brachiopods the plane areas on either side of the cardinalia cannot represent interareas if one considers the mechanics of articulation. The topographically anterior margin of these areas is situated in front of the main part of the hinge-notch and, thus, in front of the axis of rotation. The following discussion is based on a study of the pentameraceans; however, the examination of a few well-preserved interiors of the camerellaceans indicates that all the important points mentioned apply also to the Camerellacea. The small size of the hinge-teeth in the pentamera- ceans makes it difficult to study their shape with serial sections. In isolated ventral valves the hinge-teeth usually are partly or completely broken off, thus giving the impression that the posterior edges of the spondy- lium acted as articulating structures. The writer has mainly studied the hinge-teeth of pentameraceans in articulated shells either by removing parts of the dorsal valve around the hinge-teeth or by cleaning the interior of the shell. NUMBER 3 37 Growth track or"/he hinge-notch FIGURE 3.—Costistricklandia lirata (Sowerby). Posterior part of a dorsal valve in ventral view (cf. Plate 1: figure 2). The hinge-teeth of the pentameraceans and came- rellaceans are of the deltidiodont type, the whole onto- genetic development of a hinge-tooth being preserved along the delthyrial margin (Plate 1: figure 1). As a rule the hinge-teeth grow along a longitudinally strongly curved line which is reflected by the strong curvature of the beak of the ventral valve. As a result, the posterior side of a hinge-tooth is somewhat con- cave. The cross-section of a hinge-tooth is commonly more or less rounded in contrast to the often pro- nouncedly triangular shape in the brachiopods that possess deltidiodont hinge-teeth and a strophic shell. The pentameraceans lack a normal hinge-socket (St. Joseph, 1935 p. 404). The posterior side of the hinge- teeth commonly accommodates into a notch at the lateral margin of the cardinalia, here termed the "hinge-notch" (Figure 3; Plate 1: figure 2). In the pentameraceans with hinge-notches the distal part of the hinge-teeth projects freely into the interior of the dorsal valve, and the anterior sides of the teeth are free and unprotected (Figure 4; Plate 2: figure 4). In some pentameraceans (e.g.,Sieberella, Plate 2: figure 3) and many camerellaceans (e.g., Anastrophia) the hinge- teeth fit into a niche-like recess in the anterior wall of the cardinalia. The growth-track of a "hinge-niche" resembles a normal hinge-socket in strophic shells ex- cept that its longest axis is more anteriorly directed. FIGURE 4.—Gypidula galeata (Dalman). The interior of an articulated shell showing the fitting of the hinge-teeth into hinge-notches (sp. = spondylium); after specimen Br. 2738. Anteromedially the hinge-notch or the hinge-niche is bounded by the lateral edge of what, in the penta- meracean terminology, is called the "inner ? plate." This structure does not take part in the immediate construction of the hinge-notch or hinge-niche, and it is situated at a higher level ventrally. Its general posi- tion relative to a hinge-tooth is similar to that of the dorsal hinge process in strophic shells with deltidio- dont hinge-teeth, and the same term can be used provisionally for this structure both in the camerella- ceans and pentameraceans. The dorsal hinge process grew from the apex of the valve obliquely to the axis of rotation, and it assumes the same oblique position anteromedially in front of the hinge-teeth (Figure 3; Plate 1: figure 2). No trace of a depression or groove at the base of the hinge-teeth for the accommodation of the edge of this process has been observed in any camerellacean or pentameracean. There were no essential differences in the process of articulation whether the hinge-teeth accommodate into hinge-notches or hinge-niches. When the shell opened the posterior sides of the hinge-teeth were pressed against the posterior wall of the hinge-notch or the posteroventral edge of the hinge-niche, and the contact surface between these dorsal and ventral hinge structures obviously determined the position of the axis of rotation. The distal part of the hinge-teeth moved freely in anterior direction as there is no dorsal surface upon which it could have glided. The dorsal hinge processes are situated ventral to the apparent position of the axis of rotation and their role in the process of articulation is not quite clear. As these proc- esses lie obliquely in front of the hinge-teeth, however, they prevented a longitudinal movement of the valves relative to each other. In this articulating mechanism, too, there are no structures which effectively control the maximum degree of opening of the shell; however, these shells possess another mechanism which potentially could have served this function. It is difficult to ascertain whether this mechanism actually was used or the pivotal movement of the valves became arrested at an earlier stage by the action of the muscles. In nonstrophic shells with cyrtomatodont hinge- teeth the diductor muscles are attached dorsally to the inner side of the incurved beak of the dorsal valve. The axis of rotation is situated fairly deep within the dorsal valve. When the shell opens, the force of the diductor muscles moves the beak of the dorsal valve into the ventral valve until the interaction between the ventral and dorsal articulating structures makes a further movement impossible. The movement takes place just at the dorsal margin of the delthyrium and deltidial 38 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY hinge process FIGURE 5.—The relationship between some hinge structures and the posterior margin of the dorsal valve in closed, articulated, nonstrophic shells. A, in shell with deltidiodont hinge-teeth (a pentameracean); B, in shell with cyrtomatodont hinge-teeth (a rhynchonellacean). Dashed lines, structures of the ventral valve; dotted lines, structures of the dorsal valve; d.m., area of insertion of the diductor muscles. structures (Figure 5B) or, if these are lacking, at the posterior margin of the ventral valve. In nonstrophic shells with deltidiodont hinge-teeth the diductor muscles also were attached to the beak of die dorsal valve, but on opening the shell only a re- stricted movement of the dorsal beak into the ventral valve was possible. In these brachiopods the delthyrial opening cannot possibly be wider than the distance be- tween the hinge-teeth, provided that no resorption was involved. The hinge-teeth have to fit into the dorsal analogues of the hinge-sockets formed on the inner side of the dorsal valve (Figure 5A). AS a result of these conditions the total width of the dorsal beak in the pen- tameraceans and camerellaceans is always larger than the distance between the hinge-teeth. When the shell opened, the beak of the dorsal valve could move until it became pressed against the delthyrial margins and the movement was arrested. In the pentameracean and camerellacean shells examined the amount of such movement allowed varies, but in several genera only a narrow gap could have been produced between the free edges of the valves. Articulation in Shells with Cyrtomatodont Hinge-Teeth The method of using resorption in the formation of the hinge-teeth opened a new way in the construction of the articulating structures. An effective mechanism could be constructed in the articulating devices for con- trolling the maximum width of the gap between the valves. Resorption of earlier ontogenetic stages of the hinge-teeth allows the delthyrial-deltidial region to be- come wider than in shells with deltidiodont hinge-teeth where the distance between the hinge-teeth sets a definite limit to the width of the delthyrium. Thus, when the shell opens, the beak of the dorsal valve can move into the ventral valve as described in the fore- going section. As a result, the contact between the beak of the dorsal valve and the delthyrial margins lost the potential function of controlling the maximum degree of opening of the shell, and this function was taken over by the articulating structures. In nonstrophic shells with cyrtomatodont hinge- teeth the hinge-socket grows along a curved line follow- ing the curvature of the posterior margin of the dorsal valve. This is the case also with the homologues of the hinge-socket in the camerellaceans and pentamera- ceans. Due to the resorption of the hinge-teedi proxi- mally, however, the resulting knob- or hook-shaped hinge-tooth can fit into a socket and perform a limited rotating movement relative to the hinge-socket. The anteroventral wall of the socket is formed by the dorsal hinge process (inner socket ridge) which has a thick- ened, rounded margin. As a result of the direction of growth, the hinge-sockets, the long axis of the hinge- teeth, and usually also the dorsal hinge process, are directed anteromedially and thus obliquely to the axis of rotation. When the shell opens, die convex posterior surface of the hinge-tooth performs a pivotal move- ment upon the floor of the socket until the truncated proximal end of the hinge-tooth is pressed against the bottom of the hinge-socket, and the inner margin of the hinge process becomes pressed against the base of NUMBER 3 39 the hinge-tooth. In some terebratulaceans the interac- tion between the dorsal and ventral hinge structures is more complicated but the general result is the same. A furmer opening of the shell then is impossible with- out breaking some of the articulating structures. This mechanism effectively controls the maximum degree of opening of the shell. The axis of rotation is not fixed in a certain position but makes a slight translative move- ment during the pivotal movement of the hinge-teeth. Brachiopods widi cyrtomatodont hinge-teeth and a strophic shell have the same principal construction of the hinge as those with a nonstrophic shell. A transla- tion of the axis of rotation does not take place, how- ever, and the position of the axis of rotation is fixed along the hinge-line. The hinge-sockets grow along a straight line (Figure 6A) as do the hinge-teeth. Com- pared with strophic shells which possess deltidiodont dentition, the growtii of the hinge-socket is different. In shells of the former type the hinge-sockets grow by addition of shell within the socket; thus, their floor is continuously raised. In shells with cyrtomatodont hinge-teeth the growth proceeds in an anterolateral direction by continuously widening the hinge-socket. In this case no further shell is secreted upon the floor of the socket except secondarily at a later stage. As a result of this, in shells with a dorsal interarea the growth-track of the hinge-socket often remains open, forming a fur- row from the beak of the valve to the functional hinge- socket (Figure 6A; Plate 2: figure 2). Evolution of the Brachiopod Hinge Dorsal hinge process Hinge-socket FIGURE 6.—Posterior parts of dorsal valves in shells with a dor- sal interarea, showing the position of the hinge-socket relative on the notothyrial margin. A, in telotremate brachiopods (Eo- spirifer) ; B, in protremate brachiopods (Hesperorthis). All Cambrian articulate brachiopods possess deltidi- odont hinge-teeth and all the shells are strophic. The growth of the articulating structures is simple, and the animal obviously had not acquired the ability to use resorption for their construction. It is probable that the other types of dentition have evolved from the kind found in strophic shells with deltidiodont hinge-teeth. The ancestors of the camerellaceans and penta- meraceans probably should be looked for among the porambonitaceans (Williams and Rowell in Moore, 1965). Early in the Ordovician there was a tendency among some lineages of the porambonitaceans to reduce the length of the hinge-line and thus form shells which are narrow posteriorly. One such group is the Lower Ordovician (Ontikan) Angusticardini- idae (placed among Orthida and Enteletacea by Wil- liams and others (in Moore, 1965), but evidently be- longing to the Porambonitacea, cf. Rubel, 1961). Schuchert and Cooper (1932, p. 84) suggested that Angusticardinia is the earliest rhynchonelliform shell which might, with more complete reduction of the interareas, have given rise to rhynchonellids. Unfortu- nately, in this family as well as in early camerellaceans the dentition is still poorly known. The change from strophic to nonstrophic shells in brachiopods with deltidiodont dentition involved the loss of the anterior wall of the hinge-socket. Owing to the growth along a transversely curved line, the hinge- socket grew in anterolateral to anterior, instead of lateral, direction. As explained earlier, no wall could be formed in the direction of growth of the hinge- socket; otherwise this wall would have to be con- tinuously resorbed and secreted during the growth of the shell. The lack of the anterior wall of the hinge- socket implies that no surface existed for die hinge- teeth to glide upon, and thus the distal surface of the hinge-teeth did not actively participate in the process of articulation. During the pivotal movement of the valves, the posterior side of the hinge-teeth became pressed against the margin of a hinge-notch or a hinge- niche by the action of the normal force. The change also involved a decrease in the size of the hinge-teeth. In strophic shells a large contact area was needed between the hinge-teeth and hinge-sockets in connection with the function of keeping the position of the axis of rotation fixed along a long hinge-line. When the hinge-teeth became secondarily reduced, accessory structures such as the denticles in stropheo- dontids usually were developed for fulfilling the same function. Nonstrophic shells with deltidiodont denti- tion are narrow posteriorly and, since the distal surface of the hinge-teeth could not be used in the process of 40 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY articulation, there was no need of forming large hinge- teeth. The small size of the hinge-teeth in the pen- tameraceans has led to the opinion that their denti- tion is of a degenerate nature and that in this group the articulation would have been of no use in main- taining the valves together, a function that must have devolved almost wholly on the muscles (St. Joseph, 1938, p. 247). There is no doubt, however, that in the pentameraceans the dentition was sufficiently strong to allow a pivotal movement of the valves and to pre- vent a lateral movement of them relative to each other. There is, of course, the question whether or not the normal force caused by the action of the diductor muscles was sufficiently strong for holding the hinge- teeth in place. If not, the presence of an additional muscle, with its line of action directed transversely to the long axis of the hinge-teeth and going through the axis of rotation, would have been necessary. The sole function of this additional muscle would have been to hold the hinge-teeth in their proper position by the action of the normal force during the pivotal move- ment of the valves. At the present state of our knowl- edge, however, this question is purely theoretical. Neither the necessity for nor the existence of such a muscle in the articulating mechanism of the pen- tameraceans has been proven. The articulation in nonstrophic shells with deltidio- dont dentition can easily be derived from that in strophic shells. There obviously is a functional in- stability involved at the transition to anteriorly open hinge-sockets, but in order to discuss this a more de- tailed knowledge is needed on the dentition of late porambonitaceans and early camerellaceans. The use of resorption in forming hinge-teeth cer- tainly marks one of the major steps in the evolution of articulate brachiopods. The first shells that made use of this invention appeared early in the Middle Ordovi- cian; gradually, shells with cyrtomatodont dentition began to dominate among articulate brachiopods. All modern articulate brachiopods have cyrtomatodont hinge-teeth. From the standpoint of mechanics, a hinge with cyr- tomatodont hinge-teeth is a more elegant solution to the problem of articulation in the brachiopod shell than is a hinge with deltidiodont hinge-teeth. The axis of rotation could have been placed fairly deep inside the dorsal valve, and at the same time there could have been constructed a hinge which efficiently fulfilled all essential functions of the articulation. The former con- dition made it necessary to find a solution for attach- ing the diductor muscles dorsally so that the force of these muscles would attain the greatest possible mo- ment when the shell opens (Jaanusson and Neuhaus, 1965). There is an obvious functional discontinuity be- tween the cyrtomatodont and deltidiodont dentition. The shells either possessed or lacked the ability to use resorption when forming the hinge-teeth, and no inter- mediary stages could have existed. Cyrtomatodont hinge-teeth agree with the deltidio- dont hinge-teeth as developed in the camerellaceans and pentameraceans in that, in the process of articula- tion, their morphologically posterior side is the func- tionally active surface and not, as in the deltidiodont hinge-teeth of strophic shells, the distal surface. In fact, in cyrtomatodont hinge-teeth, the morphologi- cally distal surface does not participate in the process of articulation at all. This and other considerations strongly suggest that the cyrtomatodont dentition has been derived from the deltidiodont hinge-teeth in non- strophic shells. Most brachiopods with cyrtomatodont dentition have nonstrophic shells, but strophic shells are not un- common and they characterize separate genera in some groups or all members in other groups. The cyrtomato- dont dentition is easily adapted to both kinds of shells. The strophic condition of the shells with cyrtomato- dont dentition obviously is secondary (Williams and Rowell, in Williams et al., 1965) and genera or groups with such shells probably have evolved from ancestors with nonstrophic shells and with die same type of dentition. A derivation of strophic shells with cyrto- matodont dentition from those with deltidiodont PLATE 1: figures 1, 2.—Costistricklandia lirata (Sowerby), Gotland, Visby, Norderstrand. Silurian, Lower Visby Marl. 1, Ventral valve in posterior view (X5) showing well-pre- served deltidiodont hinge-teeth; Br. 102331. 2, Cardinal re- gion of the dorsal valve, in ventral view (X5) ; Br. 102332. Figures 3, 4.—Hesperorthis davidsoni (Verneuil), Got- land. Silurian, Lower Visby Marl. 3, Ventral valve in pos- terodorsal view (X5) showing deltidiodont hinge-teeth; Br. 44011, Vastkinde. 4, Interior of the posterior part of a dorsal valve in posteroventral view (X6) ; Br. 102333, Visby, Nor- derstrand. Figures 5, 6.—Clitambonites squamatus (Pahlen), Esto- nia, Kohtla. Middle Ordovician Kukruse Stage. 5, Ventral valve in posterior view (X4) showing deltidium and well- preserved deltidiodont hinge-teeth; Br. 102334. 6, Interior of the posterior part of a dorsal valve in posteroventral view (X4); Br. 102335. NUMBER 3 41 ^iF> PLATE 1 42 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY dentition would have required a number of coordi- nated changes in the articulating mechanism, and this is considered unlikely. In die evolution of the hinge in articulate brachio- pods there are tiius both minor and major functional discontinuities. Once a functional threshold had been passed and an advantageous construction created, the dentition remained remarkably constant during long periods of geological time. The dentition of the Ordovician rhynchonellaceans is practically identical witii that of the modern representatives of the group, and in no important respect does it differ from the dentition of other groups with cyrtomatodont hinge- teeth. With the exception of reduction phenomena, the morphology of the hinge-teeth varies but slightly within the group that possesses deltidiodont hinge- teeth and strophic shells. This is one of die reasons why details of the hinge-teeth are almost completely ignored in descriptions of articulate brachiopods. Study of the evolution of the brachiopod dentition gives results of great theoretical interest. It furnishes excellent examples of passing functional thresholds and the creation of new advantageous constructions which lead to an adaptive radiation of the groups with the new construction. Moreover, in this case there is the possibility for us to actually understand what happens when a functional threshold is passed during the evolution. This is possible because it is easy to analyze the function in terms of the mechanics involved; and the growth of the components that constitute the mechanism can be followed in often well-preserved and abundant fossil material. Much work remains to be done in following the changes of the dentition and analyzing the mechanism, especially in the groups which are assumed to be near the functional threshold. Some tentative conclusions can be made with the material at hand, however. The functional thresholds under consideration are of such nature that, in order to fix the innovation in a population gene pool, individual brachiopods with and without the new construction had to co-exist in one population. This implies that there probably existed at least one population of brachiopods widi deltidiodont dentition in which some individuals had a very short hinge-line and the others lacked this structure. Likewise, it is probable that a population or populations existed in which some individuals had acquired the ability to use resorption in forming hinge- teeth and that these individuals were able to cross- breed with others that lacked this ability. In other words, the populations under consideration were van- able enough to include the morphological types of die shell and articulation on either side of the functional threshold. The populations probably lived in an environment wherein the selection pressure against articulating mechanism was extremely low, because, in its early stages, the new construction of the dentition was not likely to have served its functions as well as did die old construction. In fact, a selective advantage in some other characters that were useful in the particular environment and were pleiotropically linked to the genes which carried the characters of the new con- struction possibly was needed in order to carry the new construction through the functionally unstable stage. In the process of passing the functional threshold the species probably had very small, thin shells. In small, light shells the forces acting upon the articulating devices are weak, and a certain laxity in the construc- tion of the hinge mechanism is easier for such shells to endure than in large, heavy shells where strong forces are acting upon relatively massive articulating devices. All of the earliest representatives of brachio- pods widi cyrtomatodont dentition have very small shells. Classification of the Articulate Brachiopods The groups of articulates with deltidiodont and cyrtomatodont hinge-teeth correspond closely to Beecher's subdivisions Protremata and Telotremata, respectively. When defining these groups Beecher (1891) overemphasized the importance of the pedicle opening and the role of the pedicle in influencing the growth of the posterior part of the shell. The criteria used for distinguishing these groups in adult shells were based mainly on deltidial structures, and later work showed the general appearance of the deltidial structures to overlap in the two groups. Following criticism by Cooper (1944), the use of Beecher's classi- fication was discontinued (cf. also Williams 1956, p. 259, and Williams and others in Moore 1965, p. H222). The results reported in the present paper, however, strongly support Beecher's classification. In fact, char- acters also exist in the deltidial structures that con- tribute to defining these groups; however, the factors controlling the differences in the delthyrial-deltidial region are not connected with the pedicle but are influenced by the growth and function of parts of the articulating mechanism. NUMBER 3 43 In shells with deltidiodont hinge-teeth the delthy- rium cannot possibly be broader than the distance between the hinge-teeth, and in these shells the growdi track of the hinge-teeth can be said to define the mar- gins of the deltiiyrium (Figure 1; Plate 1: figures 1, 3, 5). The inner margin of the delthyrium and that of the hinge-teeth do not correspond exactly since in many protremates the convexity of the hinge-teeth causes their inner margins to protrude farther medially than the margin of the delthyrium as defined at the level of the interarea. If die hinge-teeth are bridged by a continuous or discontinuous shell cover (deltidial structures), the growth lines on the cover commonly are continuous with those of the shell lateral to the hinge-teeth (Plate 1: figure 5). In shells with deltidio- dont hinge-teeth and a dorsal interarea the hinge- sockets invariably are situated lateral to the noto- thyrial margin (Figure 6B; Plate 1: figures 4, 6). The definition of delthyrium as taken from shells with deltidiodont dentition has been applied to this structure in all articulate brachiopods (Williams, 1956, p. 257; Williams and others in Moore, 1965, p. H93). However, in brachiopods with a cyrtomatodont denti- tion and delthyrial-deltidial structures, the growth- track of die hinge-teeth invariably is situated medially to the delthyrial margin or to the lateral margin of the deltidial structures (Figure 2; Plate 2: figures 1, 5). In nonstrophic shells the reason for this is the closeness of the hinge-teedi to the posterior margin of the shell. During opening or closing of the shell the beak of the dorsal valve moves into the ventral valve along a line just in front of the deltidial structures and behind the hinge-teeth (Figure 5B ). Because of this, when the posterior shell margin is secreted a discontinuity or a sharp flexure develops in the growth lines laterally or posterolaterally to the hinge-teeth (Figure 2A). This discontinuity or flexure in the growth lines sharply deliniates a triangular area on the posterior side of the ventral valve. If die hinge-teeth are situated farther anteriorly, so that they do not interfere with the growth of the posterior margin of the shell, the shell is secreted continuously and no triangular area is formed; this is the case with Bouchardia (Plate 2: figure 10). In strophic shells with cyrtomatodont dentition the reason for the development of delthyrial-deltidial structures is somewhat different. Here the beak of the dorsal valve does not become tucked into the ven- tral valve. In these shells parts of the hinge-teeth nor- mally project posteriorly beyond the hinge-line, caus- ing a sharp flexure in the growth lines of the posterior margin of the ventral valve lateral to the main part of the hinge-teeth (Figure 2B; Plate 2: figure 1). Again, if the hinge-teeth are placed farther anteriorly, as is the case with the thecideaceans (Plate 2: figures 8, 9), no median triangular area is developed (cf. Thecidellina). If an area is developed (cf. Lacazella), the cause of its presence lies in the configuration of dor- sal structures. In shells with a dorsal interarea, the notothyrial margin invariably is lateral to the hinge- sockets, and the lateral margin of the hinge-socket defines the notothryial margin (Figure 6A; Plate 2: figure 2). Thus the delthyrial-deltidial and notothyrial- chilidial structures are not strictly homologous in Pro- tremata and Telotremata. The phylogenetic results presented in this paper closely agree with those outlined for articulate brachi- opods by Williams and Rowell (in Moore, 1965) but they are based on a quite different set of characters. The addition of characters of dentition to those dis- cussed by Williams and Rowell brings certain points more clearly into focus and causes some changes in emphasis. Williams and Rowell (in Moore, 1965, p. HI80) suggest a close affinity of the groups here in- cluded in the Telotremata, but since those authors failed to find exclusive characters by which this group could be defined they preferred not to express this affinity in the classification. As outlined in preceding sections in this paper, the main event in the evolution of the articulate brachio- pods was the change from the delitidiodont to cyrtoma- todont dentition. At present there is no evidence in the Early Paleozoic which suggests that the func- tional threshold associated with this change was passed repeatedly in different phylogenetic lineages of ar- ticulate brachiopods. On the basis of dentition and associated structures, the class Articulata can be divided into two main, well- defined subdivisions. There is no need to replace Beecher's names Protremata and Telotremata for these subdivisions despite the fact that Beecher distinguished these groups by the use of characters that subsequently have been found to be incorrect or poorly defined. In the classification used by Williams and Rowell (in Moore, 1965) the subclass Telotremata would com- prise the orders Rhynchonellida, Spiriferida, and Terebratulida and the suborder Thecideidina. In ad- dition, the characters of the dentition reveal that cer- tain genera—which, in the current classification are included in other orders—ought to be classified with the Telotremata. 44 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY The loop-bearing Devonian genus Tropidoleptus currently is considered to be a member of the Entele- tacea (Williams and Wright, 1961; Wright, in Moore, 1965). The dentition of Tropidoleptus is of the cyrto- matodont type (Hall and Clarke, 1894, pi. 82, fig. 35; Williams and Wright, 1961, fig. lOf). Similarly, the hinge-sockets are like those in the other telotremates, and are bordered posteriorly by the chilidium (Wil- liams and Wright, 1961, fig. 10a,b). In protremate genera that possess a chilidium the hinge-sockets in- variably are situated lateral of the chilidium. In the writer's opinion the presence of a chilidium (or anty- gidium, if a distinction between these two structures is found to be advantageous) and the lack of a func- tional pedicle are weak arguments for placing Tropi- doleptus with the protremates. The dentition and the presence of a loop strongly indicate that Tropidoleptus is a member of Telotremata. The genus Cadomella was included in the separate superfamily Cadomellacea within the suborder Chone- tidina by Muir-Wood (in Williams et al., 1965). Cowen and Rudwick (1966) described a spiral bra- chidium in Cadomella and transferred that genus to the Koninckinacea, which they included in Chone- tidina. The dentition in this Lower Jurassic genus is clearly of the cyrtomatodont type (Muir-Wood, in Moore, 1965, fig. 295:2b), and the writer regards Cadomella as a late member of the order Spiriferida. The Koninckinidae were classified by Boucot and others (in Moore, 1965) with the Spiriferida, and, although the dentition in this family is still poorly known, there seems to be no serious reason to doubt the correctness of this classification. The Mississippian genus Perditocardinia currently is included in the Enteletacea but it has no hinge-line and the hinge-teeth are of the cyrtomatodont type (Schu- chert and Cooper 1932, pi. 19, figs. 14, 21). No crura, brachidium, or loop have ever been described in this genus but the absence of crura should be confirmed by serial sections. Perditocardinia might be a member of Rhynchonellida but only further studies could prove this. The Devonian Enantiosphen and the Triassic The- cospira are two additional loop-bearing genera that currently are included in protremate orders (Penta- merida-Pentameracea and Strophomenida-Davidsoni- acea, respectively). In both these genera the dentition is poorly known and the writer regards their taxonomic position as uncertain. Should it be proved that Enantio- sphen and Thecospira possess hinge-teeth of the cyrto- matodont type, then the Subclass Protremata would not include any genera with a loop or a brachidium. Moreover, with the exception of the Triassic old- haminidinid Bactrynium, the subclass would be con- fined to the Palaeozoic. Another important step in the evolution of the dentition has taken place within the subclass Protre- mata. The writer is inclined to place greater emphasis on the differences between the protremate groups with strophic and nonstrophic shells than do Williams and Rowell (in Moore, 1965). In order, however, to deter- mine the taxonomic rank of the differences between the group comprising Camerellacea and Pentameracea and the rest of the Protremata (the orders Orthida and Strophomenida, and the superfamily-Porambonitacea PLATE 2: figures 1, 2.—Megerlia truncata (L.), Mediter- ranean, exact locality unknown. Recent. 1, Ventral valve in posterodorsal view (X6) showing interarea and hook-shaped cyrtomatodont hinge-teeth; Br.102341. 2, Dorsal valve, in ventral view (X6) showing interarea and the lateral margin of the hinge-socket forming the notothyrial margin; Br. 102342. Figure 3.—Sieberella roemeri (Hall and Clarke). Ten- nessee. Silurian, Brownsport Formation. Articulated shell (Br.102336, X4) with lateral part removed to show the fitting of a hinge-tooth in a hinge-niche (sp., spondylium). Figure 4.—Antirhynchonella linguifera (Sowerby). Got- land, Stora Karlso. Silurian, probably Slite Marl. Interior of an articulated shell (Br.3068, X6) showing how the hinge- teeth fit into hinge-notches and that the distal part of the hinge-teeth project free into the cavity of the dorsal valve. The distal part of the spondylium (sp.) is broken off. Figures 5—7.—Terebratulina retusa (L.). Sweden, Gullmar Fjord, Skar. Recent. 5, Posterior part of a ventral valve in dorsal view (X6); the cyrtomatodont hinge-teeth are com- plete and have been exposed in an articulated shell by care- fully removing the hinge-structures of the dorsal valve with a needle; Br.102337. 6, Posterior part of a dorsal valve in ven- tral view (X6); Br.102338. 7, Posterolateral view (X6) of complete hinge-tooth in ventral valve (same specimen as shown in figure 5). Figures 8, 9.—Lacazella mediter-ranea (Risso). Western Mediterranean, exact locality unknown. Recent. Ventral valve in dorsal and posterodorsal views (Xl5) to show the hook-shaped cyrtomatodont hinge-teeth; Br. 102339. Figure 10.—Bouchardia sp. Locality unknown (stated to be from Singapore). Recent. Eugenia Expedition. Ventral valve in dorsal view (X8) to show the anterior position of the hinge-teeth and the continuous posterior margin of the valve; Department of Invertebrate Zoology, Brachiopoda, Dry Collection No. 75. Figure 11.—Hemithyris psittacea (Gmelin). White Sea, exact locality unknown. Posterior part of a very small (1.3 mm long) ventral valve in dorsal view (X40) to show the hook-shaped cyrtomatodont hinge-teeth; Br. 102340. NUMBER 3 45 PLATE 2 46 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY with the exclusion of members with a nonstrophic shell), more information is needed about the dentition and the mechanism of articulation in early camerel- laceans and late porambonitaceans. Literature Cited Beecher, C. E. 1891. Development of the Brachiopoda. Part I. Introduc- tion. American Journal of Science, 41:343—357, plate 17. Cooper, G. A. 1944. Phylum Brachiopoda, in H. W. Shimer and R. R. Shrock, Index Fossils of North America, pages 277- 365, plates 105-143. New York: John Wiley and Sons, Inc. Cowen, R., and M. J. S. Rudwick 1966. A Spiral Brachidium in the Jurassic Chonetoid Brachiopod Cadomella. Geological Magazine, 103: 403-406, plate 18. Hall, J., and J. M. Clarke 1894. An Introduction to the Study of the Genera of Palaeozoic Brachiopoda. New York Geological Survey, 8(2) : 1-394, plates 21-84. Jaanusson, V., and H. Neuhaus 1965. Mechanism of the Diductor Muscles in Articulate Brachiopods. Stockholm Contributions to Geology, 13:1-8. Kozlowski, R. 1927. Sur certains Orthides ordoviciens des environs de St. Petersburg. Cura et Suptibus Universitatis Liberae Polonae, series A, 17:3-21, 1 plate, 2 figures. Moore, R. C. (editor) 1965. Treatise on Invertebrate Paleontology, Part H. Brachiopoda. 927 pages, 746 figures. New York City and Lawrence, Kansas: Geological Society of America and University of Kansas Press. Rubel, M. 1961. Lower Ordovician Brachiopods of the Superfamilies Orthacea, Dalmanellacea and Syntrophiacea of Eastern Baltic. ENSV Teaduste Akadeemia Geoloo- gia Instituudi Uurimused, 6:141—224, plates 1—27. [In Russian, with Estonian and English summaries.] Rudwick, M. J. S. 1959. The growth and form of brachiopod shells. Geologi- cal Magazine, 96:1—24. Schuchert, C, and G. A. Cooper 1932. Brachiopod Genera of the Suborders Orthoidea and Pentameroidea. Yale University, Peabody Mu- seum of Natural History Memoir, 4(1): 1—270, plates A and 1-29. St. Joseph, J. K. S. 1935. A Critical Examination of Stricklandia (=Strick- landinia) lirata (J. de C. Sowerby) 1839 forma typica. Geological Magazine, 72:401—424, plates 16, 17. 1938. The Pentameracea of the Oslo Region. Norsk Geologisk Tidsskrift, 17:225-336, plates 1-8. Thomson, J. A. 1927. Brachiopod Morphology and Genera (Recent and Tertiary). New Zealand Board of Science and Art Manual, 7 :1-338, 2 plates. Williams, A. 1953. The Classification of the Strophomenoid Brachio- pods. Washington Academy of Sciences Journal, 43:1-13. 1956. The Calcareous Shell of the Brachiopoda and Its Importance to Their Classification. Biological Re- views, 31:243-287. Williams, A., and A. D. Wright 1961. The Origin of the Loop in Articulate Brachiopods. Palaeontology, 4:149-176. Wright, A. D. 1968. The Brachiopod Dicoelosia biloba (Linnaeus) and Related Species. Arkiv for Zoologi, series 2, 20(14) : 261-319, plates 1-7. Alwyn Williams Comment s on the Growth of the Shell of Articulat e Brachiop ods ABSTRACT Growth of the shell of articulate brachiopods, which is controlled at the mantle margin of each valve by a "conveyor belt" of secreting cells proliferated from an intramarginal generative zone, involves two distinct types of surfaces. Isotopic surfaces separate different shell layers and represent regular changes in the secre- tory regime of migrating outer epithelial cells. Syn- chronous boundaries indicate contemporaneous sur- faces of secretion, including superficial surfaces, such as the valve floors, and a fine banding—discernible in the microfabric of the calcareous shell—which prob- ably represents daily deposition. Ultrastructural studies of tiiese surfaces show that disruptive forward movements of the mantle (trans- gressions) are less frequently recorded than retraction from the shell edge (regressions), which may be fa- cilitated by the secretion of a protein membrane between the outer epithelium and the exoskeleton. Fol- lowing regression, all cells resume shell deposition at that phase in die secretory regime where they left off. Partitions reducing the internal volume of pedicle valves in certain fossil species also were formed by mantle retraction, and their deposition must have been preceded by secretion of a proteinous seeding membrane. The brachiopod phylum, dominated as it is by the richness and diversity of its past relics, has furnished many paleontologists with an unrivalled wealth of ma- terial for die exercise of their talents. Indeed, so pro- found and sustained has been the challenge posed by the fossil record tiiat a disproportionate number of those paleontologists, who are freely acknowledged to be among the greatest students of the science, have Alwyn Williams, Department of Geology, Queens University, Belfast, BT7 INN, Northern Ireland, United Kingdom. attained this select rank by life-long researches on the phylum. Thomas Davidson and James Hall are ob- vious examples of such paleontologists of the last cen- tury; in such pedigree of scholarship, G. Arthur Cooper is their natural successor today. The researches for which Dr. Cooper is universally acclaimed have ranged throughout Phanerozoic time, and his output of original work is remarkable for its quantity as well as quality. In fact, it is no small achievement to become familiar with the full measure of his systematic and morphological studies of the Brachiopoda, let alone his contributions to strati- graphic correlation. In both pursuits, the balance of his researches has been concerned with Paleozoic faunas, especially those of Ordovician age. But die impending publication of his account of the Permian brachiopods, the culmination of an intellectual drive that has been sustained for very many years now, will outstrip even his classic "Chazyan and Related Brachiopods" (1956) in the number, novelty, and morphological diversity of species studied. Nor should it be forgotten that, between the issues of his mono- graphs beginning with those written in collaboration with C. Schuchert (1932) and E. O. Ulrich (1938), he has published a host of significant papers varying from a definitive survey of Cainozoic rhynchonellides (1959) to penetrating analyses of specialized Devonian articulates (1954, 1955). Three invariable features of Dr. Cooper's work have always combined to set the seal of authority on his publications: die completeness of his collections, the acuteness of his observations, and the quality of his illustrations. These features reflect his constant con- cern widi the variation of constituent species as well as the entirety of an assemblage, and tiiey stem from his brilliant fieldwork and his unrivalled skill in the mechanical and chemical cleaning of embedded speci- 47 48 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY mens. One need only see his superb illustrations of the fossils etched and dissected by him during die preparation of his monographs on Ordovician brach- iopods (1956) and the Productoidea (with H. M. Muir-Wood, 1960) to have some indication of our indebtedness to him. For above all, he has made us aware of the importance of morphological detail; and by demonstrating the constancy of subtle differences in skeletal characters he has introduced new and wholly acceptable standards of precision into the notoriously opinionated world of systematic procedure. In many respects, one of his earliest papers—a study, published in 1930, of a number of North Ameri- can species of Ordovician brachiopods that had been masquerading under die name of Pionodema—was a prelude to die scale of reform that was about to be set into motion by his researches. He showed that if the microstructure of the shell, the disposition of the ventral muscle field, and the composition and mor- phology of the cardinalia are taken into account, a meaningless jumble of species could be segregated into three well-defined groups typified by Pionodema, Doleroides, and Mimella. At the same time, he drew attention to the remarkable degree of homeomorphy that existed between the punctate Pionodema-Schizo- phoria and the impunctate Doleroides-Hebertella lines, not only in many internal features but even in shell shape and the development of hollow ribs. This neat demonstration of the similarity between plector- thids and early enteletids is now recognized as a classic example of parallel evolution; and, like very many other discoveries of his that could have been quoted, is itself a lasting tribute to his palaeontological insight. Diverse as Dr. Cooper's brachiopod studies are, they are necessarily concerned, as indeed are all palaeonto- logical researches, with the morphological variation of the exoskeleton in time and space. Such variation, whether it be manifest during the development of the individual or the emergence and subsequent evolution of a species, is an expression of differential shell growth. It is, therefore, essential to be at least familiar with the processes of shell growth if one wishes to conduct a viable systematic study of a fossil assemblage, or to trace the origin and development of an exoskeletal feature. These processes are very complex, despite the fact that any increase in shell area is controlled at the mantle margins and is intimately related to the prolif- eration of cells from a pair of intramarginal generative zones. All macroscopic variation from slight modifica- tions in the shape of the shell to complicated internal and external extensions of either valve are brought about by differential changes in rates of exoskeletal secretion that, in turn, are reflected by ultrastructural realinements in the shell fabric. Indeed, certain exo- skeletal outgrowths involve such accelerations in secre- tion that they can be catered for only by the development of secondary generative zones which may be located anywhere on the mantle surface. Moreover, temporary checks or accelerations in secretion are not the only means of fashioning the exoskeleton. Under certain conditions, localized or widespread retreats of the mantle margin can take place, while resorption of the calcareous shell normally occurs simultaneously with secretion during the growth of internally dis- posed projections of the exoskeleton. Even shell repair is simply a distinctive phase of growth which has been promoted by abnormal environmental conditions. Ex- cept for the last, these several processes are briefly considered below. They are discussed mainly in die light of researches carried out on recent articulate brachiopods which afford the means of relating the exoskeletal ultrastructure to the morphology and secre- tory activities of the outer epithelial cell. This may not be palaeontology in the traditional sense, but die con- clusions derived from such studies of living species are directly relevant to the interpretation of fossil fabrics which, in the majority of stocks, are known to have been stable since at least early Cambrian times (Williams, 1968c). Material referred to in this paper was examined under an EM6B electron microscope and a Cambridge "Stereoscan" scanning electron microscope. Prepara- tion of sections for the study of soft tissues involved the fixing of living specimens in glutaraldehyde followed by their decalcification in up to 10 percent EDTA (Ethylenediamenetetraacetic acid), postosmication and embedding in "Epon Araldite"; sections were cut with a Porter-Blum microtome and stained with aque- ous uranyl acetate and lead citrate. Calcareous exo- skeletons were prepared for electron miscroscopy by removing adherent tissue and periostracum with 1 per- cent aqueous solution of potassium hydroxide or deter- gents. Shell sections were polished, briefly etched with 1 percent EDTA, and replicated by cellulose acetate strips which, before being dissolved away, were shad- owed with gold-palladium at 1-to-l and coated with carbon to provide casts for examination. Shell surfaces were coated with gold-palladium before being viewed under the "Stereoscan." NUMBER 3 49 I am indebted to Mr. Christopher Bang of Blom- sterdalen, Norway, Dr. C. A. Fleming, F.R.S., of the Geological Survey of New Zealand, Dr. D. G. Jenkins of the University of Canterbury, and Dr. R. T. Paine of the University of Washington, all of whom so generously provided me with living specimens of spe- cies referred to in the text. I also thank Dr. Jean Graham and Dr. Katharine McClure, research as- sistants in the Department of Geology of the Queen's University, Belfast, for their help in the preparation of illustrations and some of the sections figured here. Growth of the Articulate Exoskeleton The growth of the brachiopod shell can be analyzed in two seemingly different ways. Fundamentally, it is a biochemical process controlled by the mantle lining the shell (Williams 1956, 1968a, 1968b). Alternatively, growth can be thought of as a microscopic modification of the exoskeleton (Rudwick, 1959). The former proc- ess is the "momentary" growth of Rudwick (1959, p. 2) and has to be considered as an expression of cellular activity. The latter is Rudwick's "cumulative" growth (1959, p. 2) and is best understood in terms of growth vectors summarizing three dimensional variations in the rate of growth.1 As Westbroek (1967) has shown, differences between the two concepts are important. Yet, they are also readily reconcilable in that a single cell performs several distinct secretory operations in turn and thereby successively contributes to a con- sistently multilayered skeleton; whereas the mantle, in its entirety, is made up of cells—simultaneously en- gaged in every phase of the secretory regime—which deposit an assortment of materials on a contemporane- ous growth surface. Hence, a brief review of the way in which cells secrete the exoskeleton and become incor- porated into the mantle is a necessary introduction to any commentary on shell growth. In respect of both shell secretion and growth, the processes worked out 1 Shell growth is here understood to include any reduction or expansion of any part of the organic and mineral exoskele- ton resulting from its incremental resorption or secretion, respectively, by mantle epithelium. Pauses in growth consti- tuting a state of equilibrium between these two conditions probably are much rarer than is generally assumed from the frequency of "growth lines" on the brachiopod shell. Such indications of differences in growth rates are as likely to represent condensed mineral deposition, temporary reversions to organic secretion, or even resorption, as they do absolute breaks in biochemical activity of the mantle. for the rhynchonellide Notosaria nigricans (Sowerby) have been used as standards. Apart from the advan- tages of being readily available in all stages of growth, this species is impunctate, so the development of the exoskeleton is not complicated by the presence of caeca, while the strongly lamellose shell affords an opportu- nity to study the effects of periodic retraction of the mantle. The Standard Secretory Regime Marginal increase of the mantle lining a brachiopod valve is controlled by a generative zone located within a circumferential groove separating the outer and inner lobes of the mantle edge (Figure 1). The inner lobe is part of the ciliated inner epithelium and need not be considered further. The outer lobe is a fold com- posed of rotating secretory cells that acts like a "con- veyor belt" to supply periostracum to a growing exo- skeleton and rows of cells to an expanding outer epi- thelium. The core or axis of this conveyor belt, around which the cells released from the generative zone rotate, is composed of connective tissue. As each cell moves around this axis, its secretory surface initially faces inward and is part of the inner side of the outer mantle lobe. But when the secreting surface reaches the tip of the lobe, having been "moved along" by the steady proliferation of younger cells from the genera- tive zone, it is rotated to face outward and becomes an integral part of the outer side of the mantle respon- sible for the secretion of the calcareous shell. In Noto- saria, the secretory regime of every outer epithelial cell originating and migrating in this manner consists of six different operations which occur in the order (Plate 1: figure 1) described below. At first, the secreting surface of a newly released cell, with its short prostrate microvilli, exudes a muco- polysaccharide. But as the cell approaches the tip of the outer lobe, the periostracum is quickly assembled under the protective mucopolysaccharide cover before the cell is rotated externally to the tip of the outer lobe. The first periostracal constituents to be secreted are protein rods, flattened distally and arranged in arrays in a 75-degree rhombic pattern. They are about 20nm thick and stand about 35nm above a triple-unit mem- brane which is deposited almost simultaneously with the rods. The membrane is about 140A thick and, once formed, is pushed away from the secreting surface of the cell by further exudation which gives rise to a var- 50 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY iable organic layer up to 1/x.m thick. The finely divided osmiophilic granules composing the layer suggest tiiat it initially has a gel-like consistency. It is probably a form of mucoprotein because it is seen to be built up from the densely osmiophilic contents of secretion droplets of the cell, which are histochemically identi- fiable as mucoprotein. Apart from ellipsoidal mucin inclusions and commonly occurring, long fibrils lying more or less normal to the bounding surfaces, the com- pleted layer invariably contains sporadically distrib- uted vesicles of various sizes which are always bounded by a triple-unit membrane comparable in thickness with those forming the limiting surfaces of die perio- stracum. The vesicle coats are discards from the secret- ing cells, which have been trapped like bubbles in the solidifying mucoprotein before it is sealed off by an inner membrane. This membrane seems to be identical with the outer bounding one and is also associated with a distal constituent consisting of a mat of erect fibrils which extend into die mucoprotein layer for about 30nm and may be only morphologically different from the rods attached to the distal surfaces of die outer bounding membrane. Secretion of the inner bounding membrane com- pletes the deposition of the periostracum, although exudation of impersistent patches of protein as em- bedding cement for the first calcite nuclei may still take place. At this phase in the secretory regime die cell occupies the tip of the outer lobe, but as the secretory surface of the cell is further rotated to face outward it begins to deposit isolated calcite rhombs. With further secretion, these calcite seeds start amalga- mating with one another and with the continuous front of the primary layer (Plate 1:figure 2). In this way the calcareous shell expands peripherally and the outer epithelium becomes separated from the periostracum which it has already secreted by an increasing thickness of primary shell. As deposition proceeds, the cells, now regularly arranged in alternate rows, leave traces of microvillous trails and periodic banding within the fabric of the thickening primary layer. However, when a row of cells comes to occupy a certain distance behind the tip of the outer mantle lobe through the addition of new cells at the tip, it starts secreting the calcareous- organic secondary shell. Each cell exudes a triple-unit membrane, along an arcuate anterior strip of its secret- mucoprotein inner bounding membrane protein cement ?^^secondary fibre outer bounding membrane mucopolysaccharide generative zone inner epithelium FIGURE 1.—Stylized longitudinal section of a valve edge of young Notosaria showing the relation- ship between the outer mantle lobe and the triple-layered exoskeleton of periostracum, primary and secondary shell. , 0'5/um , PLATE 1.—Electron micrographs of parts of longitudinal sections of the outer mantle lobe of a young Notosaria nigricans (Sowerby) showing various features of the periostracum: 1, exuded mucoprotein and inner bounding membrane (to the left side of figure) and a regression (on the right) affecting these two layers; 2, a transgression within the periostracum (in top left quarter of the outer bounding membrane and mucopolysaccharide (in bottom right quarter of figure) ; 3, rhythmic layering causing a repetition of the outer bounding membrane and mucopolysaccharide (in bottom right quarter of figure). 372-386 0—71 5 52 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY ing cell surface. The membrane is comparable in thick- ness and possibly composition with those in the periostracum, and initially intervenes between the pri- mary shell and the base of a calcite fiber being secreted by the posterior part of the cell surface, which is now more or less free of microvilli. But because the same operation is carried out in all adjacent cells, the mem- branes join up with one another to form a series of sheaths, each containing a long, thin, calcite fiber nor- mally inclined at about 10 degrees to the primary shell (Plate 2:figures 3-5). Each fiber has a distinctive cross section reflecting the disposition of the concave secret- ing cell surface controlling its growth, and occurs within a characteristically stacked series, the regularity of which reflects the orderly arrangement of the cells relative to one another (Plate 2:figure 6). The sec- ondary shell continues to grow so long as the cells remain functional, and in this respect it is quite unlike the periostracum or the primary shell, both of which have a more or less constant thickness for any given period in the growth of the individual. Thus, any cell originating within the intramarginal generative zone of the mantle of Notosaria secretes six successive covers in the course of its functional life; mucopolysaccharide, outer bounding membrane, muco- protein, inner bounding membrane, calcareous primary layer, and calcareous-organic secondary layer. The mucopolysaccharide layer may be a relict of the proto- typic brachiopod exoskeleton (Williams, 1968d, p. 285), but it rarely persists beyond the tip of the outer mantle lobe, in living species at least, and can be ignored. The mucoprotein, with its inner and outer bounding membranes, is variably developed in other species of articulate brachiopods, but together the mem- branes make up the periostracum, which is always present, in some form or other, as a seeding sheet for the mineral shell. Hence, this layer, although lost during fossilization, plays an important part in the growth of the exoskeleton and the molding of its topo- graphic detail. The primary and secondary layers are the most permanent constituents of the exoskeleton. Recognizable traces of their ultrastructure survive even in early Cambrian shells (Williams, 1968c) and show that the fabric of a primary cryptocrystalline layer and orthodoxly stacked secondary fibers, so characteristic of the terebratulides and rhynchonellides, was equally prevalent in extinct groups. Consequently, detailed studies of these layers in relationship to the outer epithelium will afford as valid a guide to the skeletal growth of extinct stocks as to that of living species. Indeed, even for groups like the strophomenides and thecideidines, in which the development of the primary-secondary layers greatly differs from that char- acteristic of other articulates, it is likely that the basic growth processes are the same. In examining the relationships between the outer epithelial secretory surface and each of the three prin- cipal exoskeletal layers—periostracum, primary shell and secondary shell—it is essential to distinguish phys- ically between momentary and cumulative growth. Westbroek (1967, pp. 20-24) has discussed fully the problems involved. He found it convenient to regard the outer epithelium underlying the shell as constitut- ing a number of units, each consisting of a suite of cells performing the same secretory function and con- sequently giving rise to a homogeneous shell unit. These are his isotopic shell units, which are sepa- rated from each other by well-defined isotopic shell unit boundaries and are expressions of cumulative growth (Figures 1,4). He also recognized that isotopic (shell unit) boundaries are quite different from sur- faces of active shell deposition or resorption, i.e., momentary growth, which he termed the superficial shell unit boundaries. Unlike the former, the latter are contemporaneous as well as anisotopic and are more appropriately referred to as synchronous (shell unit) boundaries because they have proved to be as easily identifiable as isotopic boundaries throughout the exoskeletal fabric (Figure 4). The actual surface of deposition at the moment of death can then be referred to as the superficial synchronous boundary. The use of these terms with the meanings just given has helped to sort out the various aspects of growth described below. They also help to correlate processes of shell growth of the Brachiopoda and Bivalvia. The "matrix" of Lison (1949, p. 47) is really the commis- sural zone of the synchronous shell unit minus the outermost rim of calcite representing the first mineral secretion by cells that have only just been rotated into position. The rate of incorporating such cells into the main spread of the mantle lining the shell is con- sidered by Carter (1967, p. 271) to be partly respon- sible for generating the actual curve of the external surface of a bivalve. Carter's conclusions were based on considerations of shell geometry, but they are in line with the idea that consistent biochemical changes in migrating cells generate similar isotopic surfaces in brachiopods. His views on brachiopod growth (Car- ter, 1967, p. 277) are less agreeable because he con- NUMBER 3 53 outer bounding membrane FIGURE 2.—Formation of a transgression (aa') in the periostracum of Notosaria shown in diagram (c), by forward movement of the cell surface from its normal depositional attitude (A) through the formation of a temporary loop-like contraction (B) . 54 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY PLATE 2 NUMBER 3 55 fused primary and secondary shell and erroneously as- serted that shell deposition is normal to the long axes of secondary fibers. Growth Features of the Periostracum In Notosaria, the threefold layering of the periostra- cum greatly facilitates study of growth variation at the mantle edge. The two bounding membranes are physi- cally distinguishable from each other by the nature of their distal constituents, although there are indica- tions that the fibrils assembled during the construc- tion of the inner bounding membrane may initially be coiled into rods, like those characteristic of the outer bounding membrane. Both surfaces are isotopic bound- aries to the mucoproteinous filling which constitutes the bulk of the periostracum. They are simultaneously assembled on a cell plasmalemma(s) which subtends an outwardly facing angle of about 5 degrees with the membranes; and the various secreted products forming die surface immediately underlying the plas- malemma can be regarded as the superficial syn- chronous boundary to the periostracum. Variation in periostracal growth obviously will be related to bio- chemical changes in the cell(s) or positional shifts in the plasmalemma(s). Thus, one would expect to find features reflecting variation in the rate of secretion as well as the forward or backward slip of the outer mantle lobe. Such clues do occur and, in addition, in- dicate that different phases of the secretory regime can be repeated to give a rhythmic layering in parts of the periostracum. PLATE 2: figure 1.—Stereoscan view of a section of the primary-secondary shell of a young Notosaria nigricans (Sowerby) showing well-developed banding in the primary layer. Figure 2.—Electron micrograph of the surface of the pri- mary shell of Notosaria nigricans (Sowerby) immediately below the periostracum. Figures 3, 5.—Stereoscan views of a tangential section of secondary fibers of Terebratulina retusa (Linn.) showing well-developed banding in individual fibers. Figure 4.—Stereoscan view of a longitudinal section of the primary-secondary shell junction in Notosaria nigricans (Sowerby) showing the disposition of banding developed in the secondary fibers. Figure 6.—Stereoscan view of a regressed lower succession of secondary shell, showing primary layer and calcite pads underlying the normal upper succession of secondary and primary shell. The external surface of the periostracum of Noto- saria is thrown into a series of impersistent folds ar- ranged more or less concentrically to the shell edge (Plate 1: figure 2). In section, these are seen as anti- clines and synclines with a wavelength of about 0.4 /mi and an amplitude of 0.25 /mi, which do not affect the inner bounding membrane. The folds therefore cor- respond to recurrent changes in the thickness of the mucoprotein layer. Yet, with rare exceptions, this layer is evenly exuded between the membranes, which are more or less parallel to one another when they lie on the inner surface of the outer mantle lobe. It is likely, therefore, that this folding is not an expres- sion of variation in secretion but a postdepositional shrinkage, arising from polymerization and dehydra- tion, which does not wrinkle the inner bounding mem- brane because it is kept taut across the underlying surface of the primary shell. There are, however, sig- nificant differences in the average thickness of the periostracum from one part of the shell surface to the other that do not arise from either folding of the outer membrane or movement of the secreting surface. More consistent variations may eventually prove to be con- trolled by periodically operating factors like tempera- ture; but other variations must reflect innate differ- ences in the biochemical efficiency of the outer mantle lobes. In one young shell of Notosaria, for example, a strip of periostracum on one valve, with an average thickness of about 0.6 /mi, was three times as thick as contemporaneous periostracum on the opposing valve. Repetition of one or more secretory operations, as indicated by recurrences of part of the periostracum within the same vertical section, is fairly common. Usually it is seen to have been brought about by minor displacements of the secretory surface; and even where short strips of the same membrane recur one above the other as rhytiimic layering within a section, an oscillation of the cell plasmalemma rather man a repe- tition of its biochemical activity cannot be ruled out entirely. Rhythmic layering can be quite striking. One of the examples examined involved the repetition of the mucopolysaccharide and outer bounding mem- brane four times before periostracal deposition was completed with the exudation of the mucoprotein and the inner bounding membrane. This kind of repe- tition is the commonest (Plate 1: figure 3). Opera- tions leading to the secretion of mucoprotein let alone the inner bounding membrane, widiout discernible displacement of die cell surface, have not yet been 56 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY seen; nor has evidence been found so far of any of the basic operations of the secretory regime being omitted during deposition of the periostracum. The most spectacular changes in the periostracal succession are undoubtedly those brought about by removal of the secretory plasmalemma of the cell(s) from the superficial synchronous boundary of die periostracum, thereby terminating exudation on that particular surface. The displacement is die result of an abrupt forward extension or a sudden backward retraction of part or all of the outer mantle lobe, and it seems appropriate to refer to the physical effects on the exoskeleton of these respective movements as transgression and regression (Williams and Rowell, in Moore, 1965, p. H79). They differ from each other in the degree to which tiiey affect the rest of the exo- skeleton. Manifestation of transgressions as structural discontinuities appears to be limited to the periostra- cum; they probably are not represented by anything more spectacular than an acceleration in secretion in the rest of the shell. Regresions, on the other hand, can frequently be traced well into the secondary layer, reflecting a deep withdrawal of the entire mantle edge. Discussion of the latter, therefore, is best postponed until growth of the primary and secondary shell has been reviewed. Transgressions are relatively rare and cannot involve more than ultramicroscopic slip because the forward moving tip of the outer lobe must still be accommo- dated within a comparatively rigid rim made up of completed periostracum and the first-formed primary shell. Indeed, the cell surface may never move much outside the area representing the superficial synchron- ous boundary of the periostracum because all known transgressions are like that shown in Plate 1: figure 2. In these examples, that part of the cell surface con- cerned with the deposition of the outer bounding membrane slips forward along the superficial syn- chronous boundary of the mucoprotein to make contact with the old inner bounding membrane (figure 2B ). Thereafter, it continues to assemble the new outer bounding membrane but in physical continuity with the old inner one. The new inner bounding mem- brane also starts off in the vicinity of this junction. This suggests that transgression is really brought about by a sudden contraction of the cell(s) that carries the rear end of the secreting plasmalemma forward to lie adjacent to its front end. However, deposition soon starts up again and the plasmalemma, initially con- tracted into a loop, must revert very quickly to its normally extended contact with the superficial syn- chronous boundary of the periostracum. One interest- ing aspect of the transgressions so far examined is die way secretion of the outer bounding membrane is continued witiiout any perceptible break or readjust- ment on the edge of the old inner boundary. Repeti- tion in periostracal successions indicates that strips of outer bounding membranes can be secreted in isolation; but it may well be that as the plasmalemma (s) slides forward, the close physical and chemical similarity between the outer and inner bounding membranes promotes a linkage between the free end of the inner bounding membrane and that part of the plasmalemma responsible for the secretion of the outer bounding membrane, thereby putting a brake on further transgression. Growth Features of the Primary Layer Fabric studies of the primary shell show that it is made up of a succession of microscopic layers representing synchronous shell units that were deposited obliquely to two bounding surfaces (Figure 4). These outer and inner limiting surfaces are sharply defined isotopic boundaries and are, respectively, the periostracal- primary and primary-secondary interfaces. They coin- cide with profound biochemical changes that affect all migrating outer epithelial cells released from the intra- marginal generative zone. The continuity of these boundaries throughout the growing shell is itself proof of the regularity with which separate operations of the secretory regime follow one another in strict order. In this regard, the most interesting aspect of the primary layer of Notosaria is the increase in its maximum thick- ness from umbo to anterior margin; in effect die perios- tracal-primary and primary-secondary interfaces be- come more widely separated as individuals grow older. The relationship is illustrated by variation in the thickness of the primary layer of a brachial valve of Notosaria along a medial surface length of 15.5 milli- meters. Periodic retraction of the outer lobe of die mantle, which occurred throughout the growth of the shell, continually reduced the primary layer to a negli- gible thickness and greatly complicated the pattern (Figure 3B) . However, the maximum thickness of the primary layer increases steadily from the umbo by NUMBER 3 57 5 10 length of valve length of valve 11KJ layer mary o. 50 'o thickn ess r =0-461 a( var a) = 3-676 (0-818) FIGURE 3.—Maximum thickness of the primary shell per mm (A) and per 0.25 mm (B) meas- ured along a medial section of a brachial valve of Notosaria. synchronous unit synchronous boundary periostracum superficial synchronous boundary secondary fibre FIGURE 4.—Stylized medial longitudinal section of valve edge of young Notosaria showing the relationship between various isotopic and synchronous surfaces. about 12/A per millimeter of surface length (Figure 3A) , which represents the average increment accompanying full recovery from each retraction. The wider implica- tions of this increase cannot be profitably discussed until the synchronous shell units have been considered. Synchronous shell units are represented in sections of the primary layer by a distinct parallel banding which is inclined at an acute angle to the primary- secondary interface and opens out in the directions of growth, i.e., radially from the umbo (Figure 4 and 58 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Plate 3: figure 1). In one segment of the primary layer of a brachial valve of Notosaria, 19 measure- ments of this angle averaged 13.9 degrees (range 10- 25 degrees) ; and sampling of other sections indicates that such an estimate is typical not only of this species but others as well. The banding, being traces of old depositional surfaces, is parallel to the section of the superficial synchronous boundary on which the outer face of the outer lobe of the mantle rested at the moment of death of die animal. It is likely that the size of cells at the mantle edge does not vary very much throughout the life of the individual because enlargement of the mantle margin always takes place by the addition of new cells proliferated from the intramarginal generative zone. The average maximum widths (with ranges) of 7 and 14 newly formed fibers as an index of cell size at the junction between the primary and secondary layer were 6.7/mi (4.8 to 7.9/mi) and 6.5/mi (4.9 to 8.2/mi) for two brachial valves with median lengths of 3.2 and 13.5 millimeters respectively. The implications now become obvious. If the maxi- mum thickness of the primary layer, measured normal to its isotopic boundaries, increases throughout growth, while the size of the cells occupying the superficial synchronous boundary and the disposition of that boundary remain more or less constant, the outer mantle lobe becomes absolutely longer through an increase in the number of cells secreting the primary layer. The relationship can even be used as a rough guide to the number of cells simultaneously deposit- ing primary shell. The average medial length of 20 terminal faces of fibers seen in sequence immediately behind the primary-secondary junction was 10ju.ni. As- suming this mean not to be significantly different from the average length of cells secreting primary shell, then: N= Y 10 sin e of the periostracum. The slowing down is quite marked. In a young pedicle valve of Notosaria, 2.6 millimeters long, three times as many cells occupied a radial strip of the primary shell at the shell edge than at the umbo. Also, since there is no important difference in the thicknesses of the synchronous shell units in the compared sections, one can assume that each cell took three times as long to pass through the phase of depositing primary shell as did its predecessors near the umbo. The synchronous shell units, which are seen in sec- tion as variably imprinted bands, constitute the deposi- tional layering of the primary shell. The layers are of the right order of thickness to have been secreted diurnally (cf. Williams 1968a, p. 43) and, as is to be expected, vary more along the length of the shell than within any localized section normal to the isotopic boundaries. Thus, measurements taken near the anteromedian edge of two valves of Notosaria (one 2.6 millimeters long and the other between 7 and 9 milli- meters long) gave mean thicknesses (with range) of 0.35ft (0.23-0.41/*) for 86 layers and 0.58/* (0.53- 0.8//,) for 155 layers, respectively. The significance of such variation is still being studied. It may reflect sea- sonal changes, with the thinner layers representing winter deposition. Alternatively or additionally, daily secretion of calcite may be heavier in adult shells than in younger ones. Removal of the periostracum shows that the syn- chronous shell units intersect the isotopic external sur- face of the primary shell as a series of microscopic scarps facing outward and disposed concentrically to the shell margin and to any growth lines ornamenting the surface (Plate 2: figure 2). When die surface is slightly fretted by die pulling away of the periostra- cum, the position of the first-formed calcite seeds is where N is the number of cells along a radial vector of the mantle lobe, Y the thickness of the primary layer between its two isotopic boundaries expressed in /mi; and Z.Q the disposition of the synchronous boundaries relative to the primary-secondary interface. In bio- chemical terms, cells arising at the intramarginal gen- erative zone take an increasingly longer time to com- plete the preliminary stages of the secretory regime as the shell grows older. This is demonstrably so for that phase concerned with the deposition of the primary layer, and is likely also to prove true for the exudation PLATE 3: figure 1.—Stereoscan view of a regression within the secondary shell of Notosaria nigricans (Sowerby) showing the calcite pads developed between two sets of fibers. Figure 2.—Stereoscan view of a regression (indicated by a white crenulated band representing a resin replacement of an organic seal),, affecting the primary and secondary layers and caeca of Magasella sanguinea (Leach). Figures 3, 4.—Stereoscan view of the junction between a partition and the floor of the pedicle valve (extending from right to left in the lower half of figure 3) of Syringospira prima (Kindle) with a detail of the partition (in figure 4) showing the disposition of the secondary fibers. NUMBER 3 59 PLATE 3 60 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY well seen as culminating rhombohedral peaks along the tops of each scarp (Williams 1968a, p. 3). The peaks stand on successively bigger crystal surfaces dis- posed as contour-like steps, which join with one an- other to form the dip slope surfaces behind the scarps. Each dip slope is the top of a simultaneously secreted layer of calcite; it is also the interface between the periostracum and the primary shell. The ultra- structural arrangement of the outer isotopic boundary to the primary layer, then, is a combination of isotopic and synchronous surfaces represented by the scarps and dip slopes respectively (Figure 4). Growth Features of the Secondary Layer The secondary shell accumulates by a more com- plicated process than either the periostracum or primary layer in that each cell involved in its secre- tion simultaneously exudes both organic and mineral compounds and, in so doing, initiates and controls the growth of a calcite fiber and its ensheathing membrane in the manner already described. Consequently, the secondary shell is defined by two sorts of isotopic boundaries as well as by a superficial synchronous boundary (Figure 4). One isotopic boundary separates the primary layer from the fibers making up the sec- ondary shell (Plate 2: figures 1, 4). This boundary is less well defined tiian that between the periostracum and the primary shell. The change from primary to sec- ondary shell deposition is rapid, but it is immediately distinguishable only by the presence of the first-formed ends of the membranes, the alinement of which traces the course of the boundary. The second kind of isotopic surface is that which segregates each fiber from its neighbors and is conveniently equated with the mem- branes pervading the secondary shell. Within its sheath, each fiber is an expression of cumulative growth be- cause it is built up of a succession of synchronous shell units (Plate 2: figures 3-5), the nature of which was first recognized by Krans (1965, p. 84). The shape of a unit, which is determined by the disposition of the cell surface responsible for its secretion, is highly characteristic and comes about in the following way. The anterior arc of the external plasmalemma of a typical outer epithelial cell underlying the secondary shell secretes an organic membrane at a very much slower rate than the rest of the plasmalemma deposits calcite. The thickness of the membrane is about 14nm compared with an average of 354nm for 60 synchro- nous calcite shell units, measured medially immediately behind the membrane where the deposition of calcite is fastest. Away from this zone of maximum deposition, the section of calcite dwindles to nothing along the trail and sides of the exposed fiber, which are being covered by the membranes exuded by adjacent cells. Simultaneous secretion at these differential rates gives rise to a synchronous shell unit which is like a 74- degree sector of a circle in outline, corresponding to the margins of the terminal face of a fiber; and which is like a curved wedge in medial section with the thin end inserted between the stalk and the covering mem- brane. The concave surface of a synchronous shell unit, therefore, faces externally, as does the posterior part of the plasmalemma of the cell responsible for its secretion. The course followed by fibers during their growth and the use of their mosaics (i.e., the arrangement of the exposed parts of fibers on the valve floor) to pre- pare growth maps of articulate shells have been de- scribed (Williams, 1968a, p. 9) and need only a brief review here. For such purposes, the long axis of the exposed part of a fiber can be taken as its growth vector at the moment of death, and average vectors for groups of adjacent fibers can be systematically plotted for the entire internal surface of a valve. The exercise is instructive not only for illustrating how the micro- scopic growth of various features takes place but also for revealing the characteristic growth of a fiber and the location of zones of secondary proliferations of cells and calcite resorption. In Notosaria, newly formed fi- bers bordering the junction of the primary and sec- ondary shells develop normal to the shell margin. Within this narrow zone, however, in the two lateral areas of a valve interior, the terminal faces of fibers quickly become re-orientated more or less parallel to the commissure and grow towards the medial zone where fibers also tend to lie more or less normal to the anteromedian margin, unless they are involved in topographic complications like septa. Since these ma- ture fibers were initially normal to the shell commis- sure, their growth away from the primary layer must trace a spiral arc, rotating clockwise in the right half of a valve and counterclockwise in the left half. Such a spiral course of growth may be described as the characteristic trace of fibers. The characteristic trace is greatly modified for those fibers contributing to the formation of skeletal outgrowths composed of sec- ondary shell. It may also complete a spiral or double back on itself as a consequence of localized swirls or regroupings of cells. NUMBER 3 61 Further specialization of the outer epithelium lin- ing the secondary shell—as a preliminary to the se- cretion of elaborate internal features or the accom- modation of growing tissue like muscle—can lead to gross alteration of the skeletal fabric. In particular, it can promote the formation of a number of distinctive isotopic shell units. Two of these, the units controlled by muscle bases and secondary generative zones, are closely associated (Williams, 1968a, pp. 14-19) but they receive only passing mention here because they are still under investigation. Essentially, muscle bases (including tendons binding the lophophore to the crura) change the biochemistry of the underlying outer epithelium sufficiently to inhibit membrane secretion with a consequential loss in identity of fibers. Calcite secretion also is greatly affected within an area occu- pied by a muscle complex and can vary from exces- sive deposition to form large scalenohedral or rhom- bohedral faces to differential resorption of calcareous surfaces. Resorption is especially important in control- ling the size of features, such as lophophore supports, that have to accommodate growing soft parts of die animal. It also is involved in the growth of interlocking devices arising from both valves, like those defining the inner surfaces of the tootii ridge and the posterior part of the dental plate and the complementary outer surface of the inner socket ridge. Isotopic shell units such as these spread out and migrate across orthodox secondary layers as the shell grows. As they migrate, die disused parts of the isotopic units become buried by comparatively narrow zones of closely packed, small, regular fibers which are taken to indicate proliferation of cells from generative zones located well within the shell margin. So far, these secondary generative zones, are known to exist in Notosaria as confining posterior arcs to the main muscle areas on the floor of the valves, on the outer faces and the anterior part of the inner faces and dorsal edges of the dental plates, and on the anterodorsal surfaces of the crural bases. There is, therefore, a delicately balanced relationship between the differential growth and specialized functions of out- growth composed of secondary shell; tiiat relation- ship is reflected in the microstructure of such features. Retraction of Outer Epithelium Once the outer epithelium becomes detached from any part of the superficial synchronous boundary of the shell, retraction of the mantle or the posterior body wall can take place. The most profound changes in the exoskeleton arise from retraction of the mantle because the outer mantle lobe of punctate as well as impunctate brachiopods is free to retreat well within the primary- secondary junction, and in so doing interrupts every phase of the secretory regime being carried out by the affected outer epithelium. The actual operation is fa- cilitated by the intervention of new forms of secretion which temporarily have precedence over all the ortho- dox secretory activities pursued by cells within the area of retraction. The area affected by retraction can vary from microscopic strips normally occupied by periostra- cum to macroscopic zones showing regression through the entire thickness of the shell. Both movements obvi- ously involve the same sequence of events, and it is only a matter of descriptive convenience to treat them separately as has been done below. More spectacular morphologically but, in fact, less profound biochemi- cally were the repeated retractions of the posterior body wall away from the floor of the pedicle valves of certain spiriferide and strophomenide species. Al- though they resulted in the growth of new elevated partitions within the valves, such movements could only have involved brief lapses from the normal secre- tory operation of mature outer epithelium. Mantle Retraction in Impunctate Species Restricted mantle retraction in Notosaria is immedi- ately preceded by cessation of normal secretory activity, and the outer lobe of the mantle is pulled away from the superficial synchronous boundaries for the primary layer and periostracum (Figure 5 and Plate 3: figure 1). As the tip of the lobe retreats, newly formed outer bounding membrane is applied as a wrapping to the inner edge of the primary layer, as are the inner bound- ing membrane and mucoprotein layer after they have been sealed off by an irregular layer of a rapidly exuded, densely osmiophilic substance which is proba- bly a protein. In fact, this substance is laid down as an undercoat to the primary layer for the entire distance of retreat away from the edge of the periostracum. It probably acts as an organic slide as well as a seal, which facilitates the breaking off of contact with the super- ficial synchronous surface and the smooth witiidrawal of cells to their new positions. In all observed examples of periostracal regression the outer bounding mem- brane remains unbroken, implying tiiat an acceleration in its secretion occurred during retraction of the mantle inner bounding membrane 62 (A) (B) SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY m u c o p r o t e i n o u t e r mucopolysaccharide protein bounding membran e FIGURE 5.—Formation of a regression (aa') in the periostracum of Notosaria shown embedded in primary shell in diagram (D), by retraction of the mantle lobe from its normal depositional attitude (A) and the concomitant secretion of the outer boundary membrane and a sealing pro- teinous layer (B) before forward movement of the mantle and normal depositon began again (c). NUMBER 3 63 lobe. It is even possible that no matter how extensive the mantle's retreat from the shell edge, secretion of the outer bounding membrane keeps pace with retrac- tion; and, even if this were not so, it is likely that exudation of protein always takes place in sufficient quantities to ensure a continuous organic seal. Insofar as one can judge from such studies as have been carried out, protein actually is laid down by tiiose cells which have already passed tiirough that phase of the secretory regime concerned with the exudation of the inner bounding membrane. Cells previously depositing the inner bounding membrane and mucoprotein appear to suspend operations during retreat but start up again—when re-advance of the outer lobe takes place—at that phase in the secretory regime where they left off. Even those cells which secrete protein during mantle retraction revert to the correct phase of the regime once normal growth is restored. Thus, new primary shell is seeded on the protein undercoat of die old, the inner bounding mem- brane and mucoprotein start off anew from junctions with the protein layer, and secretion of the inner bounding membrane continues. In this way a fully op- erating regime is quickly re-established. More powerful retractions, which cause the tip of the outer mantle lobe to retreat within die primary- secondary shell boundary, may or may not involve the widespread secretion of protein as has been observed in Magasella sanguinea (Leach) and Notosaria, re- spectively. Future work probably will confirm that these differences have no systematic significance be- cause deposition of protein accompanies mantle re- traction in Hemithiris psittacea (Gmelin) but occurs only discontinuously beneath a regression in the one specimen of Terebratulina retusa (Linnaeus) examined. In the secondary shell of Notosaria, die first stages in mantle retraction involve an abrupt termination of all secretion, so that retreating outer epithelial cells move across calcific cores separated from one another by proteinous strips which represent die terminal faces and the exposed ends of the organic sheatiis of fibers. As soon as retreat has ended deposition begins again, but in a highly selective way (Figure 6; Plate 2: figure 6; Plate 3: figure 1). The retracted part of the mantle must undergo microscopic readjustments before for- ward growth is resumed, because that part of the plas- malemma of each cell which is responsible for the deposition of calcium carbonate starts secreting a cal- citic pad in continuity with the terminal face of an- other fiber. In the specimen examined, diurnal band- ing indicated that the pads were built up by a steady accretion for about two weeks, by which time individ- ual pads had joined up with one another to form a deposit about 5/x thick over the entire area of the former superficial synchronous surface of the second- ary layer. Two aspects of this phase of growth are noteworthy. The first is that these pads have a similar outline, at least in side-view, to outer epithelial cells engaged in the secretion of secondary shell. The likeness is accen- tuated by the presence of impersistent breaks between the pads, which lie, like tension gashes, more or less normal to the line of the fibers and occupy the same position as intercellular spaces in the outer epithelium. These breaks undoubtedly represent the lateral gaps that occur between the early growth stages of adjacent calcite pads before the pads make contact with one another by spreading beyond their original discrete centers of secretion (Figure 6c). It is even possible that they accommodate patches of protein exuded intercellularly from the outer epithelium at a time when secretion of protein from the external plasma- lemmas of the cells was in abeyance. Thus, as calcite deposition about each terminal face proceeded, suc- cessive layers spread towards the limits imposed by such breaks, and the pads, with dieir characteristic shape, came into being. The odier significant feature of recovery from re- traction is that the calcitic pads are secreted by all outer epithelial cells which have been brought into juxtaposition with the secondary shell surface irre- spective of the particular phase of the secretory regime they were in prior to retraction. Thus, cells occupying the outer surface of the outer mantle lobe, which normally deposit primary shell, secrete pads of the same thickness and shape as those deposited by sec- ondary outer epithelium. But once the pads have been deposited, cells revert to die normal secretory regime at precisely the point where they left off. The result is that, in shell sections, the isotopic boundaries dis- placed by a regression remain parallel to one another; and elimination of die effects of a regression enables the primary layer to be fitted neatly together in much the same way as faulted sedimentary rocks can be re- stored to stratigraphic continuity. 64 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY outer epithelium (B) FIGURE 6.—Stylized tracing of an electron micrograph showing a regression affecting the calcare- ous layers of a valve of Notosaria (A) ; with reconstructions of how calcite pads (c) and primary shell (D) are deposited on the exposed faces of fibers by a retracted mantle lobe after withdrawal of the outer epithelium responsible for the normal growth of the fibers (B). Mantle Retraction in Punctate Species Evidence for the retraction of the terebratulide mantle is worth considering because the terebratulide shell is permeated by densely distributed caeca of die mantle diat originate at the mantle edge and would appear to prevent any withdrawal whatsoever. The shells of many species, like Magasella sanguinea, how- ever, become rutted with coarse growth lines as they approach maturity, and sections show that these breaks in growth can be accompanied by mantle retraction penetrating deep into the secondary layer. In fact, die operation is exactly the same as that described for Notosaria except tiiat all outer epithelial cells secrete an organic cover for the superficial synchronous sur- face either before or during retreat (Figure 7 and Plate 3: figure 2). So far, this cover has been examined only under die scanning electron microscope, which has afforded no clue to its composition, although it is likely to be a protein or mucoprotein because it is continuous widi such substances lining the canals (punctae) that contain the caeca. Indeed, it is this continuity which is a key to the disengagement of those caeca involved in retraction. The protein lining not only separates every caecum from the calcareous sides of die puncta accommodating it but also forms NUMBER 3 65 FIGURE 7.—Stylized tracing of an electron micrograph, partly represented in figure 2 of Plate 3, showing a regression affecting the calcareous layers and caeca of a valve of Magasella (A) ; with reconstruction of how retraction of the mantle lobe leads to the transference of an anteriorly placed caecum (B) into a more posteriorly sited puncta (c) and the sealing off of the forward puncta during subsequent mantle growth (D) . the brush of fine tubules connecting with the periostracum. The microvilli covering the distal surface of die caecal head, which must have occupied the brush when it was being formed, are withdrawn shortly after com- pletion of the brush and remain so throughout the life of the animal.2 The withdrawal is accompanied by slide out from the lubricated organic lining of the punctae like fingers out of a glove. When retraction has ended and before shell deposition is resumed, some caeca probably become inserted in punctae other than those originally containing them. But it would also not be surprising to find that many caeca get trapped along the regression between the old superficial sur- face and the newly formed shell and eventually are pinched out, and that many punctal heads in die 2 G. Owen and A. Williams, "The Caecum of Articulate Brachiopoda," in press. immediate vicinity of the coarser growth lines are empty of living cells. Retraction of the Posterior Body Wall Causes of retraction have still to be explored, although obviously it may occur randomly as well as periodi- cally in response to environmental factors. Among genetically controlled retractions can be reckoned the sporadic sag of the posterior body wall away from the floor of pedicle valves of certain extinct species such as those of Richthofenia, Scacchinella, and Syringo- the exudation of mucopolysaccharide which fills the brush; and a similar arrangement is envisaged for caeca attached to the retreating part of the mantle that spira (see Cooper 1954, p. 326). In these unrelated stocks, the volume of the pedicle valves must have in- creased at a much faster rate than that of the coelomic cavity, so that the bounding outer epithelium moved 66 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY FIGURE 8.—Reconstructions to show the formation of a blister in the pedicle valve of Syringospira. suddenly in a dorsal direction from time to time and, on each occasion, deposited a partition (or blister) sealing off the vacated part of the valve floor. Exami- nation of Syringospira shows that these blisters are composed of orthodoxly stacked fibers lying at a very acute angle to the surface (Plate 3: figures 3, 4). There is no trace of any extra first-formed calcareous layer like the pads of Notosaria. Indeed, this is to be ex- pected because secretion of an organic seeding sheet must have preceded the deposition of a fibrous sheet away from the valve floor (Figure 8). Deposition of a blister, therefore, must have involved a sudden termi- nation of accretion on the valve floor, followed by a retraction of the outer epithelium of the body wall. The space created by this movement, possibly filled tem- porarily with fluid, must then have been sealed off by the exudation of a protein layer which, in turn, af- forded the foundation for the secretion of a stronger partition composed of fibers and their organic sheaths. Conclusions This review, which is necessarily only a cursory look at a few of the problems concerning exoskeletal growth in articulate brachiopods, has been devoted mainly to abnormalities of shell deposition. The intention was to present some new data that have come to light during a comprehensive study of the shell and mantle that currently is underway. Two aspects of exoskeletal growth are newly described. The first is tiiat interrup- tions in the processes of shell expansion, whether they involve forward or backward displacement of the mantle or just a simple suspension of deposition, do not upset the normal course of the secretory regime. As soon as the effects of interruption disappear, each outer epithelial cell reverts to its original phase in the sec- cretory regime. The second aspect is that when an inter- ruption involves a retraction of the mantle, disengage- ment from organic and mineral superficial synchronous surfaces usually is facilitated by the secretion of an or- ganic (probably proteinous) layer. The recognition of both features probably will be helpful in future palaeontological studies. It follows from the former that shell successions, no matter how much they are displaced in, for example, a regression, can be cor- related like strata on either side of a break. This rule will enable some realistic assessment to be made of the mobility of mantle in extinct stocks. The latter feature, i.e., exudation of an organic seal prior to mantle move- ment, also is important. Hydrolysis of such seals and removal of derived products during fossilization will leave only fabric evidence of their former presence, NUMBER 3 67 mainly as abrupt changes in the orientation of fibers. Yet, these are likely to be easily discernible and there- fore indicative of the vagaries of the mantle even in those fossil species for which there is no obvious living morphological model. Literature Cited Garter, R. M. 1967. On Lison's Model of Bivalve Shell Form, and Its Biological Interpretation. Proceedings of the Mal- acological Society of London, 37:265-278. Cooper, G. A. 1930. The Brachiopod Pionodema and Its Homeomorphs. Journal of Paleontology, 4:369-382. 1954. Unusual Devonian Brachiopods. Journal of Pale- ontology, 28:325-332. 1955. New Genera of Middle Paleozoic Brachiopods. Journal of Paleontology, 29:45-63. 1956. Chazyan and Related Brachiopods. Smithsonian Miscellaneous Collections, 127:1-1245. 1959. Genera of Tertiary and Recent Rhynchonelloid Brachiopods. Smithsonian Miscellaneous Collec- tions, 139:1-90. Krans, T. F. 1965. Etudes morphologiques de quelques spiriferes Devoniens de la Chaine Cantabrique (Espagne). Leidse Geologische Mededelingen, 33:73-148, 16 plates, 71 figures. Lison, L. 1949. Recherches sur la forme et la m6canique de d6- veloppement des coquilles des lamellibranches. Institut Royal des Sciences Naturelles de Belgique Memoirs, series 2, 34:3-87. Moore, R. C. (editor) 1965. Treatise on Invertebrate Paleontology, Part H, Brachiopoda. 927 pages, 747 figures. New York City and Lawrence, Kansas: Geological Society of America and University of Kansas Press. Muir-Wood, H. M., and G. A. Cooper 1960. Morphology, Classification and Life Habits of the Productoidea (Brachiopoda). Geological Society of America Memoir, 81:1-447, 135 plates, 8 figures. Rudwick, M. J. S. 1959. The Growth and Form of Brachiopod Shells. Geo- logical Magizine, 96:1-24. Schuchert, C, and G. A. Cooper 1932. Brachiopod Genera of the Suborders Orthoidea and Pentameroidea. Memoirs of the Peabody Museum of Natural History, 4(1): 1-270, 29 plates, 36 figures. Ulrich, E. O., and G. A. Cooper 1938. Ozarkian and Canadian Brachiopoda. Geological Society of America Special Paper, 13:1-323. Westbroek, P. 1967. Morphological Observations with Systematic Im- plications on Some Palaeozoic Rhynchonellida from Europe, with Special Emphasis on the Uncin- ulidae. Leidse Geologische Mededelingen, 41:1-82, 14 plates, 81 figures. Williams, A. 1956. The Calcareous Shell of the Brachiopoda and Its Importance to Their Classification. Biological Re- views, 31:243-287. 1968a. Evolution of the Shell Structure of Articulate Brachiopods. Special Papers in Palaeontology, 2 :1- 55, 24 plates, 27 figures. Palaeontological Associa- tion, London. 1968b. Significance of the Structure of the Brachiopod Periostracum. Nature, 218:551-554. 1968c. Shell Structure of the Billingsellacean Brachiopods. Palaeontology, 11:486-490. 1968d. A History of Skeletal Secretion Among Articulate Brachiopods. Lethaia, 1: 268-287. 372-386 O—71 6 PRECAMBRIAN-CAMBRIAN A. J. Rowell Suppos ed Pre- Cambri an Brachio pods ABSTRACT Examination of all reported occurrences of Pre-Cam- brian brachiopods reveals that all but one are most likely not brachiopods at all. Indian and Iranian mate- rial is most likely algal; the single Australian occur- rence is thought to be inorganic; and the supposed specimens of Lingulella from the Belt Series of Mon- tana are probably dolomitic segregations deformed by slippage. By way of contrast, well-preserved fossil brachiopods do occur in Arctic Canada, but they were collected from strata now thought to be of Cambrian rather than Pre-Cambrian age. Because of the diversity of shelled brachiopods that are known from the early Cambrian, the existence of Pre-Cambrian brachiopods with no mineralized shells is considered likely. However, exceptionally well-pre- served material would be needed to demonstrate con- vincingly that any such fossil remains truly represented the Brachiopoda. The nature of Pre-Cambrian life and the problems posed by the appearance of relatively abundant fossils in the Lower Cambrian have long been subjects of interest. During the past 15 years, intensive collecting in some areas of the world, coupled with the use of improved techniques and equipment, has enabled the known age of the oldest authenticated fossils to be greatly extended. Sufficient evidence has been accumu- lated to convince all but the most skeptical of the existence of primitive plants some 2,000 million years old from the Gunflint Chert of Ontario (Barghorn and Tyler, 1965; Cloud, 1965). Some simple bacteria may be even older than these plants, for minute (0.5ju), rodlike bodies, seemingly bacteria, have been recorded A. J. Rowell, Department of Geology, The University of Kansas, Lawrence, Kansas 66044. from the Fig Tree Series of South Africa in beds which are thought to be 3,100 million years old (Barghorn and Schopf, 1966). There have been many recorded finds of Pre-Cam- brian metazoan animals, but in several cases the or- ganic origin of the "fossils" has been questioned, and, in some instances, their Pre-Cambrian age is in dispute. Perhaps the most spectacular discovery has been the Ediacara fauna from Ediacara, South Australia, where Glaessner and Wade (1966) have reported the occur- rence of 25 species based on over 1,400 specimens. There has been no challenge of the organic nature of this diverse fauna, although there is room for differ- ences of opinion on its age, whether it is Late Pre- Cambrian or Early Cambrian (Cloud and Nelson, 1966; Ford, 1967). In part, these differences are semantic and reflect a lack of accord in the definition of the base of the Cambrian. Although the base of the Cambrian is not operationally well defined, the existence of elements of the Ediacara fauna in the Charnian of England (Ford, 1958) and the report of Cyclomedusa plana Glaessner from the Vendian of the Russian platform (Zaika-Novatskiy et al., 1968) suggest that at least part of the fauna was extant before Cambrian times. The possibility remains that the fauna persisted into the Cambrian, as Ford (1967) and Cowie (1967) have noted. It is not the purpose of this account to review all metazoan Pre-Cambrian fossils, as several such reviews have been published, with those of Glaessner (1966) and Cowie (1967) being the most recent. It will suffice to note that, apart from pennatulacean spicules, no commonly accepted Pre-Cambrran metazoan fossil is known to possess a mineralized skeleton. In this re- spect, the Ediacara fauna, although unusually rich, is typical of all reputedly late Pre-Cambrian faunas. 71 72 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Brachiopods are unknown at Ediacara and none have been reported in association with elements of the typical fauna from other regions. The Australian occurrences have been intensively collected, and the preservation is such that one would have expected representatives of the phylum to have been discovered if, indeed, they were ever present in significant num- bers. Reputedly, Pre-Cambrian brachiopods have been described from four major regions—India (Chapman, 1935), Iran (Stocklin, et ah, 1964), Australia (Chap- man, 1929), and North America (Fenton and Fenton, 1936; McNair, 1965) ; in every case, however, either their Pre-Cambrian age of their systematic assignment (or both) has been disputed. I am indebted to Dr. D. J. McLaren and Dr. A. H. McNair, who first invited me to study the Victoria Island material, and to Dr. McNair for discussions on the age of the specimens. I am grateful also to Dr. P. Kier and R. H. Hansman for arranging the loan of specimens of Protobolella minima and Lingulella montana from the collections of the United States Na- tional Museum and Princeton University, respectively. The following abbreviations of repositories are used in this paper: USNM, United States National Mu- seum; PU, Department of Geological and Geophysical Science, Princeton University; and GSC, Geological Survey of Canada. Indian and Iranian Occurrences Fossils have been known for 60 years from the Vind- hyan System of India—since the time that Middlemiss and Jones collected numerous, small black discs from the Suket shales. According to Holland (1909), Jones was uncertain of their affinity. Jones suggested that they might be referred to the brachiopod Obolella, or possibly they might be the operculum of Hyolithel- lus, or, alternatively, they might belong to Chuaria circularis Walcott. The difficulty in finding a defend- able systematic assignment of these fossils, apparent in their initial description, characterizes much of the history of later investigations of them. For a few years, the specimens generally were re- garded as brachiopods. They were examined by several paleontologists in the 1920s, and the consensus of Wal- cott, Ulrich, Bassler, and Resser was that they were definitely brachiopods and agreed most closely with Acrothele (Pascoe, 1928, p. 21). Chapman (in Fer- mor, 1932, p. 29; 1933, p. 20) also believed that these black discs were the remains of brachiopods; he ini- tially referred them to Neobolus and Obolella, but subsequently (Chapman, 1935) endeavored to erect two new genera, Fermoria and Protobolella to include them. Sahni (1936), who reinvestigated the type collec- tions of Fermoria and Protobolella, came to the con- clusion that such differences as were apparent between the specimens were differences of preservation, and he synonymized all of Chapman's species under the name Fermoria minima. At that time he was uncertain of their systematic position. Misra (in Rao, 1952), regarded the disks as ostracodes, but he soon aban- doned that view, believing them to be of inorganic origin (Misra, 1952). Although several investigators considered the pos- sibility that the specimens were the remains of fossil plants, Howell (in Pascoe, 1928, p. 21) was the first to express the opinion that they probably were plants, and later he drew further comparisons with other plantlike material (Howell, 1956). Some two years earlier Sahni and Shrivastava (1954) described addi- tional material, collected in 1950, that seemingly con- firmed the view that these discs were primitive plants, possibly algae. In these new specimens, the discs oc- cupied terminal positions on long filaments and were regarded as some form of spore sac. The collections of ^ Fermoria" in the United States National Museum that were made by H. C. Jones from the Vindhyans of Neemuch are, at least in a loose sense, topotypes of "F." minima. Like most of the described material, they reveal only the black disks (Plate 1: figures 1-3), with no indication of attached filaments. Superficially, these objects resemble the Ordovician inarticulate brachiopod Paterula; they are of a comparable size, have a flattened brim, and seem- ingly have the brim interrupted by a notch, which could be homologized with the pedicle notch of Paterula. These resemblances, however, are believed to be misleading. The black material of the disks is not an apatite and is probably carbonaceous; on two occasions when X-rayed, it failed to produce a picture, although the fragments of Paterula used as a control produced the characteristic lines. In some specimens the notch breaching the flattened brim (Plate 1: fig- ure 1) is demonstrably an artifact, and the carbona- ceous material is broken. There are no indications of musculature or mantle canal patterns; in fact, there is nothing other than outline that might lead one to accept these specimens as brachiopods. When con- NUMBER 3 73 sidered together with material figured by Sahni and Shrivastava, I lean to the view of those authors—that "Fermoria" is probably algal. Stocklin et al. (1964) have figured specimens which they refer to Fermoria from the Chapoghlu Shale Member of the Soltanieh Dolomite of Iran. From the available figures, the fossils are certainly similar to the Indian material, and I would regard them also as primitive plants. There is a nomenclatural point which, although briefly noted by Rowell (1965b, p. H864), has been generally overlooked. The specimens in question usually are referred to the generic name Fermoria of Chapman (1935), but the generic name is not avail- able with that author and date. As noted above, Chapman (1935) regarded the fossils as brachiopods and consequently worked under the International Code of Zoological Nomenclature. In his initial ac- count, Chapman (1935, pp. 115-117) described three new species—F. minima, F. granulosa, and F. capsella, but, unfortunately, he failed to make the generic name available as he neither designated nor indicated a type-species for the genus as required by Article 13(b) of the Code of Zoological Nomenclature. Under Article 17(3) of the Code, however, the three species- group names are available, and take date and author- ship of Chapman, 1935. It will be recalled that in the same paper Chapman also proposed the new genus Protobolella, which was erected as a monotypic genus, including only P. jonesi. As the latter is automatically the type-species by indication (Article 68(c) of the Code), the generic name Protobolella was made avail- able in 1935. In the following year, Sahni (1936) synonymized all of the species described by Chapman (1935), referring them to Fermoria minima. In his publication Sahni (1936), inadvertently perhaps, met all the conditions necessary to make Fermoria an avail- able generic name. Consequently, Fermoria takes authorship of Sahni, 1936, with type-species F. minima (Chapman), 1935, by indication (monotypy). If Sahni was correct in synonymizing the four species—and I accept his treatment—then Fermoria Sahni, 1936, although an available name, is, none- theless, not a valid name because it is a junior synonym of Protobolella Chapman, 1935. This is an un- fortunate situation, and one made even more untidy by the consequences of Sahni's action in synomymiz- ing Protobolella jonesi with Fermora minima, for both species names were published simultaneously in 1935 with Chapman as their author. Both names are available, but, as subjective synonyms, only one can be valid. In these circumstances, Sahni (1936) was acting as a "first reviser" (Article 24(a) (i) of the Code) and his action determined the relative priority of the two; regrettably, he chose Fermoria minima as senior. In summary, the legal aspects of the case are seem- ingly as follows: the valid name of the nominal taxon usually referred to as "Fermoria" is Protobolella Chap- man, 1935; and the type species of this nominal genus is Protobolella jonesi Chapman, 1935, which is a junior subjective synonym of Fermoria minima (Chap- man), 1935. There is also some doubt concerning the age of the beds containing Protobolella, both in India and Iran. The Indian fossiliferous horizons are in the Rhotas Stage at the top of the Semri Series of the Vindhyan System (Sahni, 1962a). The evidence of the age of the Vindhyans has been reviewed by both Howell (1956) and Sahni (1962b). Unfortunately, there are no undisputed faunal or floral elements that provide reliable information. Both Howell and Sahni consider the problem unsolved; both concede that at least the upper part may be Cambrian, but Howell, in par- ticular, is inclined to regard part of them as Pre- Cambrian. The stratigraphic control is only slightly better in Iran, where the oldest horizons with a diag- nostic fauna are Middle Cambrian in age (Stocklin et al., 1964). The beds yielding forms like Protobolella are around 2,000 meters lower in the section and are regarded as either Lower Cambrian or Upper Pre- Cambrian (Stocklin et al., 1964, fig. 4). Australian Occurrence In 1929, Chapman described two species which he referred to the inarticulate genera Lingulella and Obolella from the Blue Metal Limestone, Adelaide Series. The Blue Metal Limestone lies some 14,000 feet below the top of the Series (David and Browne, 1950) and its Pre-Cambrian age has not been dis- puted. The original material on which Chapman's account was based has not been seen, but the avail- able illustrations give little support to Chapman's opinion. Their form and described mode of preserva- tion strongly suggest that they are inorganic and pos- sibly were produced by weathering. 74 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY North American Occurrences Belt Series, Montana In 1932, the Fentons found remains of what they considered to be the pedicle valve of Lingulella (Fenton and Fenton, 1933). This material, from the Newland Formation, was formally described by Fenton and Fenton (1936) as Lingulella montana. The Pre-Cambrian age of the beds is not in question; indeed, if the remains are really those of a brachiopod they could lay claim to being the oldest known coelo- mate, with only the problematic Telemarkites Dons (1958) being a potential contender for the position. The question critical to the present study is whether the Montana specimens are brachiopods. Schindewolf (1956) had doubts that the specimens were Lingulella, observing that they were from 5 to 10 times larger than typical Cambrian species of the genus. They certainly are much bigger than common Cam- brian forms (Plate 1: figures 7, 9), but this argument alone is not very convincing. If one accepts that the beds from which they come are some 400 to 500 mil- lion years older than any Lower Cambrian species, then there seems no particular reason why the Belt speci- mens should not be Lingulella and belong to species substantially larger than Cambrian ones. Undoubtedly there are Lower Paleozoic lingulaceans that are sepa- rated from the Lower Cambrian by a time gap of much less than 400 million years yet are much larger than Cambrian species and noticeably larger than any of the specimens of Lingulella montana. Although the possible taxonomic position of these specimens cannot be refuted by merely considering size, there are other possible approaches to the prob- lem. Glaessner (1962) noted that the slippage which the specimens have clearly suffered would tend to destroy growth lines rather than enhance them. He was of the opinion that they were not brachiopods, but that they resembled stromatolites. A detailed reexamination of the type collection of Lingulella montana reveals features that I believe show, beyond reasonable doubt, that the "species" is not a brachiopod and probably is not even of organic origin. There are only two characteristics of the specimens that suggest they are brachiopods—their gross shape and the presence of concentric wrinkles which resem- ble growth lines. Fenton and Fenton (1936, p. 620) stated that the outline of the figured specimens had been emphasized, and this is readily apparent when one compares unfigured (and unprepared) paratypes with the illustrations in the original description (compare figures 7 and 9 of Plate 1, herein). The margins of the unprepared material are much more diffuse; it is, of course, impossible to say to what extent the sharp out- lines of the previously figured specimens reflect the activities of the preparator and to what degree they are natural. At least at one point on the holotype, the outline departs markedly from any organic structure (Plate 1: figure 10), for the dolomite forming the "shell" is seemingly extended out into the matrix as a veinlet. One can be more confident of the interpretation of the concentric wrinkles. In detail, these wrinkles are not continuous and show little resemblance to growth lines in any known lingulide. Their relation to the mar- gin of the specimens is also unlike that of any lingulide, or, indeed, that of any brachiopod. The wrinkles inter- sect the margin at high angles, whereas in lingulides they curve gently toward the margin, becoming asymp- totic to it near the beak. The wrinkles seemingly are deformation features consisting of a large number of small, steeply dipping, shear planes, all oriented in a comparable manner. Along the shear planes there is a preferred orientation of small elongate crystals, all of which lie parallel to the dip of the planes (Plate 1: figure 11). Thus, neither outline nor "ornament" can be utilized to support the claim that the specimens should be referred to the phylum Brachiopoda. The possibility that they are stromatolites remains open, but more probably they are inorganic segregations of dolomite that have been deformed by slippage. Victoria Island-Arctic Canada In 1965, McNair briefly described die occurrence of "primitive, very thin-shelled brachiopods" from what were thought to be beds of the Shaler Group, late Pre- Cambrian of western Victoria Island. This announce- ment of well-preserved fossils of seemingly Pre-Cam- brian age attracted, not surprisingly, considerable attention. It was, however, somewhat premature, for although the majority of the original specimens were forms which could provide little stratigraphic informa- tion, the material identified by McNair as Brachiopoda undoubtedly belongs to the phylum and strongly sug- gests an Early or Middle Cambrian age for the beds. After completion of more detailed field work, McNair (personal communication, 1968) now believes that the NUMBER 3 75 beds which yielded the fauna are Cambrian and are preserved in a local graben, being down-faulted into the Pre-Cambrian. Thus, in contrast to the previous cases that have been discussed, the supposed Pre-Cambrian brachio- pods of Victoria Island are sufficiently well-preserved and morphologically distinct that their taxonomic posi- tion is readily accepted, but they are not Pre-Cambrian in age. All of the identifiable specimens are paterina- ceans and seemingly are best referred to the genus Dictyonina. Order PATERINIDA Rowell, 1965 Superfamily PATERINACEA Schuchert, 1893 Family PATERINIDAE Schuchert, 1893 Genus Dictyonina Cooper, 1952 The brief diagnosis given by Rowell (in Moore, 1965, p. H295) is accepted here, but more information still is needed about many features of paterinacean genera. Dictyonina is being used for those members of the fam- ily that have a pitted ornament. It is recognized that this probably is a gross simplification of a complex phylogenetic pattern and that the ornament may have arisen independently in several stocks of the family. Indeed this seems probable, for at least two types of pitted ornament are recognizable: in some species the pits are arranged in quincunx; in others, they are hexagonal in outline and are disposed in rows radiating from the beak. It is unknown, at present, the extent to which a classification based on ornament is con- gruent with classifications based on other features. As yet we are unable to discern any probable evolu-1 tionary pathways within the group, and the simpler approach is preferred. TYPE-SPECIES.—Trematis pannulus White, 1874. Dictyonina sp. PLATE 1: FIGURES 4-6, 8 DESCRIPTION.—Specimens relatively large for the genus; maximum observed width, 13 mm. Ornament of fine pits arranged in quincunx, formed by inter- section of two sets of narrow ridges. Ornament never well preserved, but seemingly about 10 pits/mm at a distance of 5 mm from the ventral umbo. Concentric growth lines conspicuous anteriorly and anterolater- al^. Brachial valve low, gently convex, sub-semicircular in outline with straight posterior margin, maximum width 20 to 30 percent greater than length, occurring at about one-quarter valve length in front of beak. Pseudointerarea short, homeochilidium not observed. Outline of pedicle valve rounded subtriangular, apsacline. Length and maximum width approximately equal, maximum width occurring near midlength of valve. Pseudointerarea conspicuous, bounded laterally by abrupt change in slope of valve surface, ornamented only by growth striae. Homeodeltidium well devel- oped, covering apical one-third to one-half of delthy- rium, strongly convex externally, marked off from pseudointerarea by narrow groove. Internal structures unknown. DISCUSSION.—All the material is crushed to a vary- ing degree, and adequate biometric treatment is not possible. The above dimensions and ratios can be re- garded only as indicators of order of magnitude. The genus occurs in high Lower Cambrian beds and is known to range into the Upper Cambrian. The ornament of described Late Cambrian species, how- ever, is not closely comparable to that of the Victoria Island material; moreover, the Late Cambrian forms are all typically much smaller. Thus, the fauna is con- sistent with a late Early Cambrian to Middle Cam- brian age for the beds. Recognition of the Earliest Brachiopods Although it has been claimed that Pre-Cambrian brachiopods have been found on several occasions, all such claims seemingly are not justifiable. As discussed above, either the specimens are not brachiopods or, alternatively, they belong to the phylum but are not Pre-Cambrian. To date, we have no direct fossil evi- dence for the existence of the phylum prior to Cam- brian times; indeed, we still know remarkably little about the group from the lower beds of the Lower Cambrian. In several parts of the world there is a sequence of archeocyathid-bearing limestones that underlies the lowest beds with trilobites. In some areas, at least, brachiopods are reported to occur in these horizons carrying an archeocyathid fauna (Sokolov, 1968), but little work as yet has been done with them. We know that by late Early Cambrian times the phylum was represented by all four orders of inarticu- 76 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY late brachiopods together with kutorginaceans and primitive orthides. Seemingly, the ability to secrete a shell of calcium phosphate or calcium carbonate was acquired early in the Early Cambrian. The form and mode of life of any possible Pre- Cambrian ancestor is still in the realm of speculation, but the morphology of both fossil and living brachio- pods is thought to provide reasonable guidelines for such conjectures. Because of the diversity of late Early Cambrian brachiopods, it seems probable that com- mon ancestral forms existed that lacked mineralized skeletons. All known brachiopods have two valves that are underlain by very thin mantles, the valves being opened and closed by musculature of varying com- plexity. Almost by definition, our postulated shell-less brachiopod would need to possess such mantles; and for the musculature to be effective in controlling their movement they would have to possess a degree of rigidity. The rigidity conceivably could be produced by a hydrostatic skeleton; and the mantle canals, a feature of all known brachiopods, could have func- tioned secondarily in this manner. More probably, the rigidity was achieved by the secretion of an entirely organic integument. Williams (1968) has shown that the outer organic periostracum functions as a seeding sheet for calcite in Recent brachiopods and has sug- gested that it may have constituted the only skeletal cover in ancestral, pre-shell forms prior to attaining its present function. Regardless of how rigidity of the mantle was attained, consideration of brachiopod mus- culature and the torque stresses involved in its use suggest that in the absence of a mineralized shell the animals could attain only a small size. If such primitive brachiopods existed, one wonders how readily they would be recognized as members of the phylum in the fossilized state. In this condition, they would presumably form small, flattened carbo- naceous impressions, hopefully retaining some indica- tion of bilateral symmetry. It seems doubtful if such characterless objects would be universally accepted as brachiopods unless additional features, typical of the phylum, were also developed. Traces of two such features are known as fossils, for indications of the pedicle (Craig, 1952) and the form of the feeding organ, the lophophore (Steinich, 1963; Rowell and Rundle, 1967), have been described from fossil brachi- opods, but both are very rare. Amongst fossil forms, the lophophore is known only in the terebratulide Cancellothyrididae, a family which characteristically has very heavy development of calcareous spicules in the lophophore; these spicules are responsible for re- taining some indication of the nature and disposition of the lophophore after death. Also unusual is a pre- served pedicle, which is known only in Lingula, a genus whose pedicle is atypical of the phylum in its great length and unusual proximal thickness. A pedicle is typical of most brachiopods, but commonly it is relatively short and its position beneath the beak of the pedicle valve hinders its recognition in fossil speci- mens. Thus, preservation of either of these organs re- quires not only exceptional environmental conditions but also rather unusual development of the organs in the living animal. There are, however, additional structures which, although exceedingly fine, are characteristic of nearly all brachiopods. These are the setae, "chitinous" in composition, bristle-like in form, and projecting from the valve margins. They occur in living species of both classes, inarticulates and articulates, and it seems rea- sonable to believe that they may well have been present in their ancestors that lacked mineralized shells. Al- though they are not commonly preserved in fossilized PLATE 1: figures 1-3.—Protobolella minima (Chapman), from Vindhyan System, Neemuch, India. 1, Severely exfoli- ated disk, showing marginal brim; USNM 116016 ( X 15). 2, Specimen showing traces of concentric ornament on marginal brim; USNM 116012 (X15). 3, Specimen showing vague concentric marks probably due to crushing; USNM 116015 (X15). All specimens coated with ammonium chloride. Figures 4-6, 8, 12.—Dictyonina sp., from Cambrian, Minto Inlet, Victoria Island, Northwest Territories, Canada. 4, Damaged shell with pedicle valve displaced posteriorly rela- tive to brachial valve; GSC 24568 (X 4). 5, Posterior view of pedicle valve showing conspicuous homeodeltidium; GSC 24571 (X 9). 6, External view of brachial valve showing low pseudointerarea; GSC 24570 (X 4). 8, Ventral exterior of exfoliated pedicle valve, specimen crushed on right posterior; GSC 24569 (X 5). 12, Detail of ornament (X 25) of speci- men shown in figure 4. All specimens coated with ammonium chloride. Figures 7, 9-11.—Lingulella montana Fenton and Fenton, from Newland Limestone, Belt Series, Birch Creek, eight miles west of Sulphur Spring, Montana. 7, Unfigured para- type material showing four aligned specimens whose margins are locally diffuse; PU 40769 (X 1). 9, Holotype, PU 40767 (X 2), showing linguloid form and "ornament." 10, Detail (X 20) of right posterior margin of holotype, below vein cut- ting "shell"; the "shell" material is drawn out into the matrix. 11, Detail (X 20) of "ornament" of holotype, showing align- ment of crystals. Specimens in figures 7 and 9 coated with ammonium chloride; specimen in figure 10 photographed under water. NUMBER 3 77 ?- 'SMBM ;»,^3s W Br ^ ' tf^WjMw'W 5? :§L'-. t tJ^JHrMtoH WW' J '«P '''SljiatfiarOflEBQl Bs* i 4'ff Imflr '^Bon^l. L;jfr»W l i "• Jffit ? < >*l 9mSt>-JfiFtf ''? ' m m 11 PLATE 1 78 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY material, given very fine-grained sediments and free- dom from scavengers their form may be retained as casts—for example, Dictyonina pannula from the Middle Cambrian Burgess Shale of Mt. Stephen (Wal- cott, 1912, p. 362). It would appear that these struc- tures are the most likely ones to confirm the identifica- tion of a bilaterally symmetrical, carbonaceous impres- sion as the remains of a brachiopod as opposed to traces of plant material or the representative of another phylum. As yet we know nothing of the evolution of the Brachiopoda prior to the acquisition of a mineralized shell, and it is apparent that if such shell-less forms existed it would require exceptionally well-preserved material to demonstrate convincingy that their fossil remains indeed belong to the phylum. Literature Cited Barghorn, E. S., and S. A. Tyler 1965. Microorganisms from the Gunflint Chert. Science, 147: 563-577. Barghorn, E. S., and J. W. Schopf 1966. Microorganisms Three Billion Years Old from the Precambrian of South Africa. Science, 152: 758- 763. Chapman, F. 1929. Some Fossil Remains from the Adelaide Series of South Australia. Royal Society of South Australia Transactions and Proceedings, 53: 5-6, plate 2. 1935. Primitive Fossils, Possibly Atrematous and Neo- trematous Brachiopoda, from the Vindhyans of India. Geological Survey of India Records, 69: 109— 119, plates 1,2. Cloud, P. E., Jr. 1965. Significance of the Gunflint (Precambrian) Micro- flora. Science, 148: 27-35. Cloud, P. E., Jr., and C. A. Nelson 1966. Phanerozoic-Cryptozoic and Related Transitions: New Evidence. Science, 154: 766-770. Cowie, J. W. 1967. Life in Pre-Cambrian and Early Cambrian Times. In W. B. Harland, et al. (editors), The Fossil Rec- ord, pages 17-35. London Geological Society. Craig, G. Y. 1952. A Comparative Study of the Ecology and Palaeo- ecology of Lingula. Edinburgh Geological Society Transactions, 15: 110-120. David, T. W. E., and W. R. Browne 1950. The Geology of the Commonwealth of Australia. Volume 1, pages iii-xx, 3-746, 209 figures, 58 plates. London: Edward Arnold. Dons, J. A. 1958. Fossils(?) of Precambrian Age from Telemark, Southern Norway. Norsk Geologisk Tidsskrift, 39: 249-262. Fenton, C. L., and M. A. Fenton 1933. Oboloid Brachiopods in the Belt Series of Montana. (Abstract.) Geological Society of America Bulletin, 44: 190. 1936. Walcott's "Pre-Cambrian Algonkian Algal Flora" and Associated Animals. Geological Society of America Bulletin, 47: 609-620, plates 1-3. Fermor, L. L. 1932. General Report of the Geological Survey of India for the Year 1931. Geological Survey of India Records, 66: 1-150. 1933. General Report of the Geological Survey of India for the Year 1932. Geological Survey of India Records, 67: 1-82. Ford, T. D. 1958. Pre-Cambrian Fossils from Charnwood Forest. Yorkshire Geological Society Proceedings, 31:211- 217, plate 13. 1967. Pteridinium and the Precambrian-Cambrian Boundary. Science, 157:957-958. Glaessner, M. F. 1962. Pre-Cambrian Fossils. Biological Reviews, 37:467- 494, 1 plate. 1966. Precambrian palaeontology. Earth-Science Reviews, 1(1):29-50. Glaessner, M. F., and M. Wade 1966. The Late Precambrian Fossils from Ediacara, South Australia. Palaeontology, 9(4) :599-628, 2 figures, plates 97-103. Holland, T. H. 1909. General Report of the Geological Survey of India for the Year 1908. Geological Survey of India Records, 38:1-70. Howell, B. F. 1956. Evidence from Fossils of the Age of the Vindhyan System. Palaeontological Society of India Journal, 1:108-112. McNair, A. H. 1965. Precambrian Metazoan Fossils from the Shaler Group, Victoria Island, Canadian Archipelago. Geological Society of America Program, 1965 An- nual Meetings, p. 105. Misra, R. C. 1952. A New Collection and Restudy of the Organic Remains from the Suket Shales (Vindhyana), Rampura, Madhya Bharat. Science and Culture, 18(l):46-48. Moore, R. C. (editor) 1965. Treatise on Invertebrate Paleontology, Part H, Brachiopoda. 927 pages, 746 figures. New York City and Lawrence, Kansas: Geological Society of America and University of Kansas Press. Pascoe, E. H. 1927. General Report of the Geological Survey of India for the Year 1926. Geological Survey of India Reco-rds, 60:1-127. 1928. General Report of the Geological Survey of India for the Year 1927. Geological Survey of India Records, 61:1-140. NUMBER 3 79 Rao, S. R. N. 1952. News Report—India. The Micropaleontologist, 6(1):19. Rowell, A. J., and A. J. Rundle 1967. Lophophore of the Eocene Brachiopod Terebratu- lina wardenensis Elliott. University of Kansas Paleontological Contributions, 15:1-8, 2 figures. Sahni, M. R. 1936. Fermo-ria minima: A Revised Classification of the Organic Remains from the Vindhyans of India. Geological Survey of India Records, 69:458-468, plate 43. 1962a. The Vindhyan System of India. Geological Sur- vey of India Records, 91:271-278. 1962b. The Lower Palaeozoic in India and Burma with Observations on Its Faunal Anomalies and the Age of the Vindhyan System. Geological Survey of India Records, 91:357-364. Sahni, M. R., and R. N. Shrivastava 1954. New Organic Remains from the Vindhyan System and the Probable Systematic Position of Fermoria Chapman. Current Science, 23:39—41. Schindewolf, O. H. 1956. In Franz Lotze (editor), Uber Prakambrische Fos- silien, pages 455-480. Geotektonisches Symposium zu ehren von Hans Stille. Sokolov, B. S. 1968. Stratigraphic Boundaries of Lower Paleozoic Sys- tems. International Geological Congress 23rd Ses- sion Report, 9:31-41. Steinich, G. 1963. Fossile Spicula bei Brachiopoden der Rugener Schreibkreide. Geologie, 12:604-610, figures 1-9. Stocklin, J.; A. Ruttner; and M. Nabavi 1964. New Data on the Lower Paleozoic and Pre-Cam- brian of North Iran. Geological Survey of Iran Re- port, 1:1-29, plate 1. Walcott, C. D. 1912. Cambrian Brachiopoda. United States Geological Survey Monograph, 51(1) :5—872, 76 figures; 51(2):5-363, plates 1-104. Williams, A. 1968. Significance of the Structure of the Brachiopod Periostracum. Nature, 218:551-554. Zaika-Novatskiy, V. S., et al. 1968. Pervyy Predstavitel' Ediakarskoy Fauny v Vende Russkoy Platformy (Verkhniy Dokembriy). Paleontologicheskiy Zhurnal, 2:132—134, 1 figure. ORDOVICIAN Gertmda Biemat On Branched Surface Spines in Some Inarticulate Brachiopods ABSTRACT Multiple bifurcation of surface spines is considered one of the characteristic features of a Lower Ordovic- ian PSiphonotreta sp. The general morphology (the arborescent part excluded) and morphogenesis are judged to be similar to those of single spines of other spinose brachiopods. The function of branched spines is discussed in general. One of the most interesting, unusual and, as yet, almost unknown features of siphonotretids is their dichoto- mously bifurcating surface spines. That phenomenon within this group of inarticulate brachiopods was first recorded by Professor R. Kozlowski (1948, p. 6). Prior to 1939, while dissolving Lower Ordovician chalcedo- nites from Wysoczki, Holy Cross Mountains (Gory Swietokrzyskie), in diluted hydrofluoric acid he found a rich faunal assemblage containing, among otiiers, many valves of siphonotretids, numerous detached bi- furcating spines and (Kozlowski, personal communica- tion) one pedicle valve of a comparatively large siphonotretid with preserved branched spines. Unfor- tunately, all this material was lost during World War II. After the war, Professor Kozlowski continued dis- solving the chalcedonites from Wysoczki, obtaining in the residuum, among other microfossils, a number of separate valves of phosphatic inarticulate brachiopods and numerous, but only detached, branched spines. All this chalcedonic material of brachiopods was passed on to me by Professor Kozlowski in 1958. In 1965, with a view of enriching this material, I began dissolving chalcedonites from Wysoczki in di- Gertruda Biemat, Polska Academia Nauk, Zaklad Paleozoo- logii, Al. Zwirki i Wigury 93, Warsawa, Poland. 372-386 O—71 7 luted hydrofluoric acid and obtained complete speci- mens of siphonotretids with bifurcating spines. Two years of dissolving about 400 kilograms of rock yielded numerous and differentiated inarticulates—especially the genera Lingulella Salter, Oxlosia Cooper, Cono- treta Walcott, Schizambon Walcott, Helmersenia Pan- der, and Siphonotreta de Verneuil—and a number of separate dichotomous and multiple dichotomously bifurcating spines. It was only quite recently, however, that I succeeded in obtaining three almost complete brachial valves of PSiphonotreta sp., each with a few preserved dichoto- mously branched spines. These specimens supply con- crete evidence of the occurrence of this phenomenon within siphonotretids. The work here presented has been carried out in the Paleozoological Institute of the Polish Academy of Sciences in Warsaw. The studied collection is housed in that Institute and is abbreviated herein as Z. Pal. Bp. XIII. I sincerely thank Professor Kozlowski for having in- terested me in this field of investigation. Thanks also are due to Miss M. Witkowska, laboratory assistant, for dissolving the chalcedonites; to J. Kazimierczak for the photographs of specimens and spines; and to Mrs. K. Budzynska for the ink drawings—all of whom are with the Paleozoological Institute of the Polish Acad- emy of Sciences, Warsaw. Remarks on the Chalcedonites at Wysoczki The Lower Ordovician chalcedonites, the source of the siphonotretids, come from Wysoczki, about 51.5 kilo- meters southeast of Kielce (Figure 1). This outcrop was studied mainly from the lithological point of view by Czarnocki (1928), Samsonowicz (1948) and petro- graphically, with some considerations on the origins 83 84 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY 1M000000 I \ *i* X 4J ^/\ / £ , ( \ ?""%& yi o //~N "^ 1 \ \ C*s??S $ C$ ^^v,*-~-' 1 jCralro^tL p%g£ \ KIELCE e-;,-- ". ' < B ,., r^, .JIN PLATE 2 124 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY DISCUSSION.—The shape, proportions, and details of ornament of the shells in this suite of specimens have been considerably modified by tectonic deformation. Although all ventral valves preserve the resupinate profile distinctive of the genus, the place of resupina- tion and the degree of concavity of the anterior part of the shell vary widely. Such features as shell depth, outlines, length-width ratios, and the configuration of the dorsal sulcus also vary widely in response to defor- mation, as do the details of ornament, except in small parts of some specimens. Thus, species characteristics are known only in very general terms; a more precise species definition must await better preserved speci- mens. Nevertheless, the present material is sufficient to establish that these specimens possess a hitherto un- known combination of characteristics. Literature Cited Berry, W. B. N. 1967. Comments on Correlation of the North American and British Lower Ordovician. Geological Society of America Bulletin, 78: 419-428, 2 figures. Moore, R. C. (editor) 1965. Treatise on Invertebrate Paleontology. Part H, Brachiopoda. 927 pages, 746 figures. New York City and Lawrence, Kansas: Geological Society of America and University of Kansas Press. Neuman, R. B. 1964. Fossils in Ordovician Tuffs, Northeastern Maine. United States Geological Survey Bulletin, 1181— E: 1—38, 7 plates, 4 figures. 1968. Paleogeographic Implications of Ordovician Shelly Fossils in the Magog Belt of the Northern Appa- lachians, pages 35-48, 2 plates, 2 figures, in E-an Zen and others (editors), Studies of Appalachian Geology—Northern and Maritime. New York: Interscience. Opik, A. A. 1930. Brachiopoda Protremata der estlandischen Ordo- vizischen Kukruse-stufe. Tartu (Estonia) University Acta et Commentationes, series A, 17(1): 1-262, 26 figures. Poole, W. H. 1958. Napadogan Map-Area (1:63,360). Canada Geolog- ical Survey Preliminary Map Series 11—1958. 1963. Hayesville Map-Area (1:63,360). Canada Geologi- cal Survey Preliminary Map Series 6-1963 Robb, C. 1870. Report on the Geology of Part of New Brunswick. Canada Geological Survey Report for 1867—1869, pages 173-209. Spjeldnaes, N. 1961. Ordovician Climatic Zones. Norsk Geologisk Tids- skrift, 41: 45-77, 7 figures. Whittington, H. B. 1966. Phylogeny and Distribution of Ordovician Trilo- bites. Journal of Paleontology, 40: 696-737, 16 figures. Williams, H. 1963. Twillingate Map-Area, Newfoundland (1:63,360). Canada Geological Survey Paper 63-36, 30 pages. Reuben James Ross, Jr. A New Middle Ordovician Syntrophopsid Genus ABSTRACT The new genus Cuparius, a pitted syntrophopsid, is described from the Middle Ordovician Orthidiella and lower Anomalorthis zones of Nevada. The new genus is found also in Newfoundland and in the Mystic Conglomerate of Quebec. In 1965, Dr. G. A. Cooper, in the company of the author and several other biostratigraphers, collected silicified brachiopods from an old locality south of Frenchman Flat on the Nevada Test Site. From the insoluble residue Cooper obtained rare specimens of pitted syntrophopsids which he very kindly transmitted to the author for description. The species represented by these silicified specimens occurs also as calcareous shells in the great bioherm at Meiklejohn Peak, Nevada (Ross, 1964, pp. C25-C26). Specimens from both places are the basis for the description that follows. When Ulrich and Cooper (1938, p. 231) described the genus Syntrophopsis they called attention to two species that bore pitted exteriors like Porambonites. These species were classified with Syntrophopsis be- cause brachiophore plates converge dorsally and be- cause dental lamellae form a true spondylium. In Porambonites, the dental plates unite with the floor of the pedicle valve discretely, and brachiophore plates are parallel, without converging to form a septalium in the brachial valve; a spondylium can be simulated Reuben James Ross, Jr., Paleontology and Stratigraphy Branch, United States Geological Survey, Federal Center, Denver, Colorado 80225. Publication authorized by the di- rector, United States Geological Survey. because of deposition of secondary material in mature shells of Porambonites (Schuchert and Cooper, 1932, p. 102). The pitted surface and fine radial costellae distin- guish a group of shells of early Middle Ordovician age which cannot otherwise be separated from Syn- trophopsis. These species are named Cuparius in honor of Dr. Cooper, who first recognized tiiat they repre- sented a genus distinct from Porambonites. In June 1968, with the assistance of Dr. Valdar Jaanusson, the author was able to examine topotype specimens of Porambonites at the Naturhistoriska Riksmuseet, Stockholm, as part of a research project under National Science Foundation Grant GA-4020. Family SYNTROPHOPSIDAE Ulrich and Cooper, 1936 Cuparius, new genus Syntrophopsis Ulrich and Cooper (part), 1938, p. 231. Porambonites Pander (part), Ulrich and Cooper, 1938, pp. 242-243. Porambonites Pander (part), Cooper, 1956, pp. 609-610. DIAGNOSIS.—Shells transversely elliptical in outline, subequally biconvex in anterior profile. Brachial valve the deeper in lateral profile. Pedicle sulcus and brachial fold originate near midlength, variable in width and depth. Pedicle cardinal area of variable width, curved, apsacline; brachial cardinal area short, straight or very slightly curved, anacline. Surface marked by fine radial costellae and by fine pits. On surface, pits appear be- tween costellae for most part, but decorticated calcare- ous shells show concentric rows of shallow pits crossing radial rows to form closely packed series of quincuncial patterns. Shell substance impunctate. 125 126 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY Pedicle interior as in Syntrophopsis, except that median septum is reduced and does not extend for- ward in shell. Interior of brachial valve like Syn- trophopsis. TYPE-SPECIES.—Cuparius cardilatus, new species. DISCUSSION.—This new genus includes the follow- ing previously described species: Cuparius landmani (Ulrich and Cooper), 1938, p. 234. Boulder in Mystic Conglomerate, Quebec. Cuparius vetusta (Ulrich and Cooper), 1938, p. 239. Boulder in Mystic Conglomerate, Quebec. Cuparius umbonatus (Cooper), in Ulrich and Cooper, 1938, p. 242, pi. 53c, fig. 21.—Cooper, 1956, p. 609, pi. 107B, figs. 4—11. Antelope Valley Limestone, Ikes Canyon, Nevada. Cuparius sp. 1 (Cooper), 1956, pp. 609, 610, pi. 108H, figs. 40—42. Antelope Valley Limestone, lower part of middle member associated with Palliseria in unit 8 of McAllister (1952, p. 11), Quartz Spring area, Nevada (Ross, 1967, pi. 11). Cuparius sp. 2 (Cooper), 1956, p. 610, pi. 107A, figs. 1-3. Lower part of the Table Head Formation, at Table Point, Newfoundland. Cuparius sp. 3 (Cooper), 1956, pp. 610, 611, pi. 106D. (Prob- ably the same species as that described below from the Orthidiella zone, Ranger Mountains, Nevada.) Another species, Porambonites? sp. 4 of Cooper (1956, p. 611), is known only from a shell exterior and comes from the late Middle Ordovician of Penn- sylvania, presumably from strata younger than that of the other examples; it is excluded from this group until better known. According to Williams (1962, p. 230), Porambonites acutiplicata Reed resembles this shell from Pennsylvania and has dental plates that ally it with true Porambonites, not with Syntrophopsis. The genus Punctolira has the same shell structure and ornamentation, the same kind of spondylium, and very similar brachiophore plates. It differs in having a rudimentary cardinal process. Cuparius, new genus, may have special stratigraphic significance. It occurs in the Orthidiella zone of Ne- vada at the Nevada Test Site and at Meiklejohn Peak and is known in the same zone in the northern Moni- tor Range. It also is found in the lower part of the Table Head Formation of Newfoundland; recent cor- relations made by the present author show that the lower part of the Table Head is equivalent to the Orthidiella zone. The new genus also is known in the lower Anomalorthis zone in the Quartz Spring area of Nevada, and it probably is in the same zone at Ikes Canyon. The few species that occur higher stratigraphically may belong to true Porambonites rather than to Cuparius, new genus. Cuparius cardilatus, new species PLATE 1: FIGURES 1-8 DESCRIPTION.—Shell of medium size for the genus; transversely elliptical; length of pedicle valve between three-fourths and eight-tenths of width. Hinge width equals seven-tenths to eight-tenths of greatest width. Valves equally convex in anterior profile; brachial valve the deeper in lateral profile. Greatest convexity of pedicle valve near beak. Brachial fold and pedicle sulcus initiated at or in front of middle of shell; width at anterior margin close to half greatest width of shell. On outer surface very fine costellae separated by finer interspaces. At radius of 5 mm from brachial umbo, costellae spaced 5 per mm; at 10 mm radius spaced 6 per mm. Pits obvious in radial rows between costellae; however, decorticated specimens show pits closely spaced in quincuncial series, 6 per mm both radially and concentrically at radius of 10 mm from brachial umbo. In interior of brachial valve there is one pair of lanceolate canal patterns (vascula media?) along lat- eral flexure of fold. A second lanceolate pair lies about one-quarter of distance from flexure to hinge line. A lobate mantle impression lies between this pair and hinge-line. A pair of small elliptical scars in front of notothyrial cavity may be adductor muscle scars. In pedicle valve one pair of canal patterns extends from front of spondylium onto tongue, diverging slightly. Flabellate canal patterns cover much of posterolateral part of shell. Median septum under spondylium low and short. Measurements (in millimeters) of holotype: length of brachial valve, 12.5; length of pedicle valve, 13.6; width, 17.5; hinge width, 14.1; thickness, 9.7; length of fold, 5.2; anterior width of fold, 9.3; costellae at 5 mm, 5 per mm; costellae at 10 mm, 6 per mm; pits at 5 mm, 5 per mm; pits at 10 mm, 6-7 per mm. TYPES.—Holotype, United States National Museum (USNM) 162818; paratypes USNM 162819-21. OCCURRENCE.—The holotype and paratypes USNM 162819-20 were collected from Antelope Valley Lime- stone at same locality as United States Geological Sur- vey (USGS) collection D719-CO (Ross, 1964, p. C20; 1967, pi. 11). Paratype USNM 162821, from USGS collection D1968-CO, was collected from NUMBER 3 127 PLATE 1.—Cuparius cardilatus, new species. 1, Paratype, pedicle interior, fragmentary specimen; USNM 162819 (USGS coll. D719-CO), Nevada Test Site; stereophotograph (X 1.9). 2, Para- type, brachial interior, damaged specimen; USNM 162820 (USGS coll. D719-CO), Nevada Test Site; stereophotograph (X 1.9). 3, Paratype, brachial valve, dorsal surface largely decorticated (X 4.6); USNM 162821 (USGS coll. D1968-CO), bioherm, Meiklejohn Peak. 4-7, Holotype, complete specimen, dorsal, right lateral, ventral, and anterior view; USNM 162818 (USGS coll. D719-CO), Nevada Test Site; stereophotographs (X 1.9). 8, Same specimen as figure 3, showing quincuncial packing of pits on small area on anterior part of fold (X 9.3). Antelope Valley Limestone, in bioherm 167 feet above its base, Meiklejohn Peak (Ross, 1964, p. C26). DISCUSSION.—Cuparius cardilatus, new species, is readily distinguished from C. landmani (Ulrich and Cooper) by its greater hinge width. It is a larger species than C. vetusta (Ulrich and Cooper) ; C. vetusta seems to have a somewhat longer pedicle sulcus. Cuparius umbonatus (Cooper) is a wider and longer species and yet has a narrower hinge. The species designated Porambonites! sp. 1 by Cooper (1956, pp. 609, 610, pi. 108H) is about the same size but is much more coarsely pitted and seems to have a narrower hinge than C. cardilatus, new species; in both respects it resembles C. umbonatus. A Newfound- land species, Porambonites? sp. 2 of Cooper (1956, p. 610, pi. 107A) , has a much narrower hinge. The small specimens described as Porambonites? sp. 3 by Cooper (1956, pp. 610, 611, pi. 106D) prob- ably are immature individuals of C. cardilatus, from about one-third to one-half grown. 128 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY From the calcareous specimens, it is obvious that the pitted structure of the shell is not limited to the external surface and that the shell substance is es- sentially fibrous. The shell structure should be the sub- ject of study by electron microscope, using methods of Williams (1968). Other shells found at about the same stratigraphic level (R. J. Ross, Jr., unpublished data) show no evi- dence of pitting; these have been assigned to Syntro- phopsis, although they are above the expected strati- graphic range of that genus. The lack of pits cannot be attributed to silicification because calcareous speci- mens are found with the same characteristics. Literature Cited Cooper, G. A. 1956. Chazyan and Related Brachiopods. Smithsonian Miscellaneous Collections, volume 127: part 1 (text), pages 1-1024; part 2 (plates), pages 1025- 1245, 269 plates. McAllister, J. F. 1952. Rocks and Structure of the Quartz Spring Area, Northern Panamint Range, California. California Division of Mines Special Report, 25: 1—38. Ross, R. J., Jr. 1964. Middle and Lower Ordovician Formations in Southernmost Nevada and Adjacent California. United States Geological Survey Bulletin, 1180-C: 1-101. 1967. Some Middle Ordovician Brachiopods and Tri- lobites from the Basin Ranges, Western United States. United States Geological Survey Profes- sional Paper, 523-D: 1-43, 10 plates. Schuchert, C, and G. A. Cooper. 1932. Brachiopod Genera of the Suborders Orthoidea and Pentameroidea. Yale University Peabody Mu- seum of Natural History Memoir, 4(1): 1—270, 29 plates. Ulrich, E. O., and G. A. Cooper 1938. Ozarkian and Canadian Brachiopoda. Geological Society of America Special Paper, 13: 1—323, 58 plates. Williams, A. 1962. The Barr and Lower Ardmillan Series (Caradoc) of the Girvan District, South-West Ayrshire, with Descriptions of the Brachiopoda. Geological So- ciety of London Memoir, 3: 1—267, 25 plates. 1968. Evolution of the Shell Structure of Articulate Brachiopods. Special Papers in Palaeontology, 2:1-85, 24 plates, 27 figures. Palaeontological Association, London. H. B. Whittington A New Calymenid Trilobite from the Maquoketa Shale, Iowa ABSTRACT Parts of die exoskeleton, other than the thorax, of the trilobite species Calymene mammillata Hall, 1861, are redescribed from the original and new material. The latter comes from the cephalopod coquina beds in the Elgin Shaly Limestone Member, the lowest mem- ber of the Maquoketa Shale; this is probably the orig- inal locality. The Elgin Member is of Richmond, or possibly slightly older, age. The absence of buttresses from fixed cheek to glabellar lobes, the large basal glabellar lobe, papillation of the anterolateral angle of the glabella, and long, gently sloping preglabellar field with a lateral boss combine to distinguish the cranidium of C. mammillata from tiiat of any known species. It is regarded as the type-species of a new genus, Thelecalymene, and as most closely related to Gravicalymene, species of which have been described recently from contemporaneous and older rocks of the midcontinent and New York State. In 1936 Shirley discussed earlier work on calymenid trilobites and distinguished five new genera, basing his distinctions on characters of the glabella and ad- jacent fixed cheeks, and the preglabellar field (anterior border). He drew attention to the presence, in certain species, of papillation of particular glabellar lobes, and of corresponding buttresses projecting from the fixed cheek to meet such papillae, and he pointed out that these papillae and buttresses project over the 2xial fur- row to form a bridge. He divided calymenids into two series, based on presence or absence of the papillate- buttressed structures; however, whether these series H. B. Whittington, Department of Geology, University of Cambridge, Cambridge, England. are separate lines of descent or whether phylogeny is complicated by parallel series has yet to be demon- strated. Difficulties have been encountered in using form of the preglabellar field—whether it is flat, ridged or rolled in sagittal profile—in the way Shirley used it, and care must be taken to distinguish the form as seen in internal molds from that of the external surface of the exoskeleton. Nevertheless, the criteria Shirley used have been applied to discriminate spe- cies and genera in British (Dean, 1962, 1963) and North American (Whittington, 1954; Stumm and KaufTman, 1958; Ross, 1967) Ordovician material. There has been little recent work on Silurian caly- menids except for that of Tillman (1960) and Camp- bell (1967). Until more is known of exoskeletal characters of calymenids throughout their stratigraph- ical jange, it will not be possible to delineate phylo- genetic lines. The species from Iowa described here exhibits a group of cranidial and pygidial characters which set it apart from any other; consequently, it is made the type of a new genus. The age cannot be given more precisely than within the upper part of the Cincinnatian Series, and relationships to other genera are uncertain, though species of Gravicalymene may be the most closely related. The material from Iowa was collected in 1952 on a field trip undertaken with the advice of Dr. G. Arthur Cooper. It is a pleasure to acknowledge my indebted- ness to him for many hours of companionship in the field, for his wise guidance during my studies of Or- dovician rocks and fossils, and for his inspiring ex- ample as a practicing paleontologist. I am indebted also to Dr. Roger L. Batten, American Museum of Natural History (AMNH) for the loan of Hall's type mate- 129 130 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY rial, and to the late Professor A. K. Miller for the loan of specimens in the geology department of University of Iowa (UI). The material was prepared and photo- graphed in the Museum of Comparative Zoology, Har- vard University (MCZ). Dr. V. Jaanusson, State Museum of Natural History, Stockholm (RM), kindly loaned type and other material of Papillicalymene papillata, on which Figures \e-h are based. Family CALYMENIDAE Burmeister, 1843 Thelecalymene, new genus DIAGNOSIS.—Glabella bell-shaped in outline, three pairs of lateral lobes, anterolateral angle of frontal glabellar lobe papillate. No genal buttresses to lateral lobes l-3p. Preglabellar field broad (sagittal and exsagittal), sloping forward and upward from furrow, low boss at lateral margin projects towards corner of frontal glabellar lobe. Eye lobe opposite lateral glabellar lobe 2p, broad low eye ridge; anterior, inner corner of cheek slightly buttressed towards corner of frontal glabellar lobe. Hypostome without swelling or ridge on anterior lobe of middle body, posterior lobe with median posterior indentation; outline of posterior border bilobed. Pygidium with pleural region displaying deep pleural furrows which die out distally, leaving a border of width one-third that of the region; faint interpleural furrows. External surface of exoskeleton granulate, ex- cept in furrows. TYPE-SPECIES.—Calymene mammillata Hall, 1861. DISCUSSION.—The diagnosis presented above is in- tended to be read in conjunction with my diagnosis (Whittington, in Moore, 1959, pp. O450, 0451) of the family, lines two to four of which should read "widest across occipital ring or pre-occipital (i.e., lp) lateral lobes; may or may not project in front of genae." Dean (1962, pp. 111-118; 1963, pp. 216-228) de- scribed, from the Ordovician of Britain, species of Flexicalymene, Diacalymene, and Gravicalymene and erected the genus Onnicalymene, distinguished from Flexicalymene essentially by the relatively posterior position of the eye lobe. Ross (1967, pp. B7-B17) has discussed Flexicalymene and Gravicalymene and de- scribed species of them from the Ordovician of Ken- tucky and Ohio. Species of Flexicalymene in Britain (Figure li,j) and North America (see also Whitting- ton, 1954, pp. 147, 148; Stumm and Kauffman, 1958) have, as Ross (1967, p. B9) remarks, the axial furrow straight or slightly curved convexly outward, this curva- ture being related to the evenly graduated size of die lateral glabellar lobes. None of these lateral lobes is buttressed to the fixed cheek, the pleural region of the thorax is narrow (transverse) than the axis, and there is no smooth border on the pygidium. Of these char- acters, the lack of buttresses on the fixed cheek, is the only one that Flexicalymene has in common with Thelecalymene. The type-species of Diacalymene, D. diademala (Figures \c,d, herein and Whittington, in Moore, 1959, p. O450, figs. 353,2a,6) is from the Silurian, Wenlock, of Czechoslovakia. Lateral glabellar lobe 2p is opposed closely by a genal buttress, and the preglabellar field is strongly ridged, the posterior-facing slope having a low projection where it meets the anterior border opposite the axial furrow. Rostral suture is situated some dis- tance down the anterior-facing slope of the field. The inner, anterior corner of the fixed cheek is inflated and projects slightly, but the anterior angle of the frontal glabellar lobe is not papillate. The inflated inner corner of the fixed cheek and the slight distal projection (not a buttress) on the preglabellar field are as in Thelecaly- mene, but the genal buttress opposite lobe 2p and die ridged preglabellar field distinguish Diacalymene from that genus. The pygidium of Diacalymene has a narrow smooth border, but includes only five axial rings and pleural ribs. Thus, while Diacalymene and Thelecaly- mene have some characters in common, the two are not closely related if genal buttresses opposed to lateral glabellar lobes are considered important characters in phylogeny. In Gravicalymene (figure \k,l) the glabella is like that of Thelecalymene in having a relatively large basal lobe, so that the outline of the lateral margin is curved concavely outward. There are no genal buttresses to lateral lobes l-3p in either genus, but Gravicalymene does not have the outer angle of the frontal lobe papil- late. The preglabellar field is strongly ridged in Gravi- calymene, not gently upsloping medially as in Thele- calymene. In certain species of Gravicalymene (G. hagani, Gravicalymene sp. of Ross, 1967, pi. 3, figs. 1, 19) there is a slight development of a buttress on the outer part of the preglabellar field that resembles the much stronger boss in this position in Thelecalymene. The deep, trenchlike posterior border furrow in Gravicalymene hagani Ross, widest medially, is like NUMBER 3 131 FIGURE 1.—Outlines of cranidia of type species of calymenid genera, sagittal profile and dorsal views: a, b, Thelecaly- mene mammillata (Hall, 1861), based on MCZ 8672/2 (see Plate 1: figures 1-4); c. d, Diacalymene diademata (Bar- rande, 1846), based on MCZ 8346, Wenlock age portion of Liten Shale, St. Ivan, Prague district, Czechoslovakia; e, f, Papillicalymene papillata (Lindstrom, 1885), based on RM Ar 6208, Hemse Beds of Ludlow age (Martinsson, 1967), Ostergarn, Gotland; g, h, P. papillata, based on RM Ar 27199, Hemse Beds of Ludlow age, Hammars, Ostergarn, Gotland; i, j, Flexicalymene cf. caractaci (Salter, 1865), based on original of Whittington (1965, pi. 18, figs. 9, 10), Longvillian Stage, Caradoc Series, North Wales; k, 1, Gravi- calymene convolva Shirley, 1936, based on Sedgwick Museum A5877, Ashgill Series, Birdshill quarry, near Llandeilo, South Wales. that in Thelecalymene mammillata (Hall). The py- gidia of Gravicalymene from Kentucky (as described by Ross) appear to have one less pleural rib than in T. mammillata, and lack the distinct border; inter- pleural furrows are similar in being deeper proximally and distally. Only the pygidium described as Gravicaly- mene? sp. 5 by Ross (1967, p. B13, pi. 3, fig. 13) has a distinct border, but it has one less axial ring and pleural rib than T. mammillata. The cranidium described as 372-386 0—71 10 Gravicalymene sp. 4 by Ross (1967, pp. B12, B13, pi. 3, figs. 14-17), from the same formation, may belong with this pygidium; it has a preglabellar field which is less strongly ridged than that of many species of Gravi- calymene, and thus is more like that of Thelecalymene. Among Ordovician calymenids, Thelecalymene is thus most like Gravicalymene, and is approximately contemporaneous with it in the lower part of the Ma- quoketa Shale (G. aff. quadricapita of Ross, 1967, pi. 4, figs. 14, 15). An Ordovician species that is like Thelecalymene mammillata in having the anterolateral angle of the frontal glabellar lobe buttressed from the inner, anterior corner of the fixed cheek is Papillicaly- mene husseyi Stumm and Kauffman (1958, pp. 957- 959, pi. 123, fig. 18) from high Richmond strata of Michigan. This latter species has lateral glabellar lobes 2p and 3p opposed by low genal buttresses, and it shows a low projection at the outer edge of the preglabellar field; it differs from T. mammillata in having a deep preglabellar furrow and hence a strong flexure in the preglabellar field at the margin of this furrow, and the palpebral lobe may be situated slightly farther back. Stumm and Kauffman referred tiieir species to Papillicalymene Shirley (1936, p. 396, fig. 1), of which die type-species, Calymene papillata Lindstrom, 1885, is from strata of Ludlow age in Gotland. Through the kindness of Dr. V. Jaanusson I have been able to ex- amine the type and other material of this Swedish species, and it will be redescribed elsewhere. P. papil- lata (Figure \e~h) has prominent genal buttresses op- posed to lateral lobes 2p and 3p, the anterolateral mar- gin of the frontal glabellar lobe papillate, and a deep and narrow preglabellar furrow, from the anterior margin of which the preglabellar field curves upward close to the frontal slope of the glabella before arching dirough 180° to descend vertically to the rostral suture. The frontal glabellar lobe may be opposed by a but- tress from the inner corner of the fixed check (Figure 1/), or from the outer part of the preglabellar field (Figure \h). The pygidium is like that of T. mammil- lata, but it has a much less conspicuous sixth pleural rib and a narrower (transverse) border. Thus, Papilli- calymene is quite distinct from Thelecalymene. Stumm and Kauffman's species is based on a single, incomplete internal mold of the cranidium, and more material is needed to confirm its generic assignment, as those authors have recognized. 132 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Thelecalymene mammillata (Hall, 1861) FIGURE la, b; PLATE 1: FIGURES 1-7; PLATE 2: FIGURES 1-13. Calymene mammillata Hall, 1861, pp. 50, 51; Hall, in Hall and Whitney, 1862, p. 432, figs. 1, 2; Walter, 1927, p. 239- 241, pi. 19, figs. 1-3. MATERIAL, GEOLOGICAL HORIZON, AND AGE.—AM NH 1408a-g, originals of Hall (1861) ; 1408a, holo- type, incomplete cranidium, original of Hall (1862, fig. 1) ; 1408e, incomplete pygidium, original of Hall (1862, fig. 2) and a second pygidium on the same slab (Plate 2: FIGURES 7, 9, 19, 13) that has some of the exoskeleton attached and is not flattened; 1408b-d, three incomplete cranidia; and 1408f, g, incomplete pygidia; all are from "the shales above the Galena Limestone, Maquoketa Creek, 12 miles west of Du- buque, Iowa." UI 9069, 9077-8, originals of Walter (1927) from the same locality as Hall's specimens, and approximately from the well-known section, one-quar- ter mile southwest of Graf, Dubuque County, Iowa (Ladd, 1929, pp. 341-343). This section is in Ladd's graptolite zone of the Elgin Shaly Limestone Member, the lowest part of the Maquoketa Shale. When visiting this section in August 1952 I observed some 20 feet of dark gray shales, in the upper 6 feet of which were calcareous, sandy beds, 6-8 inches thick, with cephalo- pods (Miller and Youngquist, 1949). Dissociated exo- skeletal parts of the calymenid (MCZ 8672) occur in the lower part of the cephalopod beds, and there also occur gastropods, bivalves, and rarer linguloid brachi- opods. Hall's specimens are in similar calcareous sand- stones and contain cephalopods and gastropods. These beds must be approximately Beds 11-16 of Thomas's section, quoted by Ladd (1929, pp. 342-343). The only other trilobite that I have seen from the Graf section is the cranidium of Primaspis cf. crosotus (Locke, 1843), a species from the Eden of the Cincin- nati district; the specimens were collected by Mr. H. W. Tichenor. USNM 10263, from Maquoketa Creek, Iowa, presumably is from the same beds. Ladd (1929, pp. 384, 395) lists "C". mammillata only from his de- pauperate zone at the base of the Elgin Member, but evidently the species ranges higher, into Ladd's grapto- lite zone. The Maquoketa Shale is regarded (Twenhofel, et al., 1954) as being of Richmond age; the characteristic form in Ladd's graptolite zone of the Elgin Member being apparently Orthograptus truncatus peosta (Hall). Professor W. B. N. Berry (personal communi- cation, March 1968) has identified in collections from the Graf section this species and also Orthograptus truncatus abbreviatus (Elles and Wood), a form he considers confined to Richmond and younger strata. Glenister (1957) concluded from her study of cono- donts that the Elgin Member in Iowa was Richmond- ian or possibly slightly older. Professor Walter C. Sweet (personal communication, 28 January 1968) states that the Elgin of Iowa has yielded a number of specimens of the conodont Phagmodus undatus, which he considers is probably of Eden age, though it may be younger. The "graptolite zone" of Ladd embraces beds, at scattered localities, containing abundant graptolites; my investigations showed that different species oc- curred at particular localities, and cast doubt on the view that the age of these beds at different localities lay within the same limited span of time. Thus, the age of die Elgin Member appears to be Upper Ordovi- cian, possibly Richmond, possibly older. DESCRIPTION.—Occipital ring widest (sagittal) medially, distally curving forward to merge, across the shallow axial furrow, with the inner, posterior corner of the cheek. Occipital furrow deepens behind basal glabellar lobe into transversely elongate apo- demal pit. Basal (lp) lateral lobe oval in outline, inflated, separated by shallow longitudinal furrow from median lobe. The 2p lateral lobe similar in outline and inflation to lp lobe, length (exsagittal) half that of lp lobe; 3p lobe inflated, anterior angle of frontal glabellar lobe weakly to strongly papillate, the tip directed slightly upward and outward at some 45° to the sagittal line; anterior margin of frontal lobe curved convexly forward, the curvature varying ac- cording to the strength of the papillation of the lobe (compare figures 4 and 5 of Plate 1). Axial furrow a narrow, deep trench curving around the basal gla- bellar lobe, broadening progressively anteriorly, deep anterior pit opposite frontal lobe. Slope of frontal lobe of glabella is vertical adjacent to broad, shallow pre- PLATE 1.—Thelecalymene mammillata (Hall). 1-4, Crani- dium: anterior, left lateral, oblique, and dorsal views ( X 3); MCZ 8672/2, Elgin Shaly Limestone Member, Maquoketa Shale, 0.25 mile southwest of Graf, Dubuque County, Iowa. 7, Dorsal view ( X 6) of anterior portion of same specimen, show- ing details of external surface, glabellar lobes, and preglabellar field. 5, 6, Holotype, cranidium: dorsal and oblique views (X 3); AMNH 1408a, shales above the Galena Limestone, Maquoketa Creek, 12 miles west of Dubuque, Iowa (prob- ably same locality and horizon as original of figures 1-4, 7)- Original of Hall (1862, fig. 1). NUMBER 3 133 PLATE 1 134 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY glabellar furrow, on anterior side of which a broad (sagittal and exsagittal) preglabellar field slopes gently upward to margin, where it is abruptly flexed to slope downward and inward for a short distance to the ros- tral suture (Plate 2: figure 12). In dorsal view margin of preglabellar field is curved so that field projects as a lip; a broad, low boss at the inner, lateral margin of this field. Cheek subquadrant in outline, moderately convex, the portion in front of the eye lobe inflated so that it projects slightly over the shallow border fur- row; lateral border rounded, widening toward genal angle; posterior border widens progressively outward, posterior border furrow a steep-sided, flat-bottomed trench which is widest (exsagittal) medially. Eye lobe situated midway across cheek, the transverse line through midpoint runs through anterior portion of 2p lobe. Palpebral lobe with low, broad rim, continued across fixed cheek by broad, low, eye ridge, which runs down into axial furrow opposite lateral glabellar lobe 3p (Plate 1: figures 5, 7). Anterior branch of suture runs forward and slightly inward to reach margin well outside projected line of axial furrow; course of pos- terior branch angulate (Plate 1: figures 3, 4), the oblique angle situated a short distance out and back from the eye lobe, from whence the branch runs straight outward and backward to the rounded genal angle. Hypostome with median body gently convex, middle furrow deep at border furrow, extending in- ward and backward as a faint furrow; the two median furrows curve to meet and outline a faintly more con- vex, crescentic posterior lobe of the middle body; medi- an depression in posterior margin of posterior lobe. Anterior border continuous with anterior wing, narrow and flexed to slope downward and forward, sutural margin curved convexly forward. Lateral border nar- row, convex; posterior border broader, horizontal and bilobed in outline, a sagittal furrow bisecting the border. Thorax unknown. Axis of pygidium composed of seven rings (Plate 2: figures 6, 9), this portion taper- ing evenly, and a prominent, parallel-sided, posterior portion which is bluntly rounded posteriorly. First six ring furrows complete, deepest distally, seventh furrow not so deepened and not reaching axial furrow, shallower and narrower (transverse) eighth furrow on posterior portion. Pleural regions curved down so that the outer part slopes steeply, but not vertically. Six deep pleural furrows (Plate 2: figure 13), which, to- gether with the axial furrow along the margin of the seventh axial ring and posterior portion of the axis, define six prominent ribs. The pleural furrows die out on the outer portion to leave a border of width one- third that of the pleural region. Five interpleural fur- rows traverse the first five ribs, situated closer to the posterior than the anterior margin of the rib, and are deepest adjacent to the axial furrow, shallow medially, and deeper distally, where the first three extend on to the border. The sixth rib is unfurrowed. External surface, except in furrows, bearing scat- tered granules varying in diameter; no regular ar- rangement of larger granules. DISCUSSION.—The relations between Thelecaly- mene mammillata and species of other genera are dis- cussed above, and it is considered that Thelecalymene may be most closely related to Gravicalymene, though it has some characters in common with Diacalymene. A species of Gravicalymene is present in the lower part of the Maquoketa Shale (Ross, 1967, Plate 4: figures 14, 15), as is Flexicalymene fayettensis (Slocum, 1913; Ladd, 1929, p. 395; Whittington, 1954, p. 148), which appears to be a typical member of the genus. "F." gracilis (Slocum, 1913), from the highest mem- ber of the Maquoketa Shale (Ladd, 1929, p. 395), appears from Slocum's description to have the anter- olateral margin of the frontal glabellar lobe closely approached by a buttress from the inner, anterior angle of the cheek. Such a structure is similar to that in the PLATE 2.—Thelecalymene mammillata (Hall). 1, 5, Cranid- ium; dorsal and oblique views (X 2) ; AMNH 1408c, shales above the Galena Limestone, Maquoketa Creek, 12 miles west of Dubuque, Iowa (probably same locality and horizon as originals of figures 3 and 4 of this plate). 2, Glabella and part of left fixed cheek: dorsal view (X 2) ; AMNH 1408d, same locality as figures 1, 5 (above). 3, Incomplete hypo- stome: exterior view ( X 3) ; MCZ 8672/1, Elgin Shaly Lime- stone Member, Maquoketa Shale, 0.25 mile southwest of Graf, Dubuque County, Iowa. 4, Free cheek: exterior view (X 3) ; MCZ 8672/3, same locality as figure 3 (above). 6, Pygidium: dorsal view (X 2) ; AMNH 1408e, same locality as figures 1, 5 (above). Original of Hall (1862, fig. 2). 7, 9, 10, Pygidium: right lateral, dorsal, and posterior views (X 2); on same slab as AMNH 1408e. 8, Pygidium: first axial ring and adjacent parts of pleural regions broken off, dorsal view (X 4.5); AMNH 1408g, same locality as figures 1, 5 (above). 11, Holotype cranidium: anterior view (X 2); AMNH 1408a (see also Plate 1: figures 5, 6). 12, Incomplete cranidium: anterior view, showing ventrally-facing portion of preglabellar field ( X 2); AMNH 1408b, same locality as fig- ures 1, 5 (above). 13, Oblique view (X 6) of same specimen as in figures 7, 9, 10 (above), showing details of external sur- face ; first pleural furrow at extreme left, anterior band of first pleura broken off. NUMBER 3 135 PLATE 2 136 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY species Papillicalymene husseyi Stumm and Kauffman, 1958 (see above). However, Stumm and Kauffman (1958, p. 955), in discussing "F." gracilis, do not men- tion any buttress; evidently both gracilis and husseyi need further investigation. Literature Cited Barrande, J. 1846. Notice preliminaire sur le Systeme Silurien et les Trilobites de Boheme. vi + 19 pages. Leipzig. Campbell, K. S. W. 1967. Trilobites of the Henryhouse Formation (Silurian) in Oklahoma. Oklahoma Geological Survey Bulletin, 115:1-68, 7 figures, 19 plates. Dean, W. T. 1962. The Trilobites of the Caradoc Series in the Cross Fell Inlier of Northern England. Bulletin of the British Museum {Natural History) Geology, 7:65— 134, plates 6-18. 1963. The Ordovician Trilobite Faunas of South Shrop- shire, III. Bulletin of the British Museum {Natural History) Geology, 7:213-254, plates 37-46. Glenister, A. T. 1957. The Conodonts of the Ordovician Maquoketa For- mation in Iowa. Journal of Paleontology, 31:715— 736, plates 85—88. Hall, J. 1861. Report of the Superintendent of the Geological Survey, Wisconsin. 52 pages. Madison, Wisconsin. Hall. J., and J. D. Whitney 1862. Report of the Geological Survey of the State of Wisconsin, 1:1—455, maps. Ladd, H. S. 1929. The Stratigraphy and Paleontology of the Ma- quoketa Shale of Iowa, Part 1. Iowa Geological Survey, 34:305-448, plates 4-17, figures 64-76. Lindstrom, G. 1885. Forteckning pa Gotlands siluriska Crustaceer. Ofverst Konglig Vetenskaps-Akademiens Forhand- lingar Angdng, 42a(6) : 37-100, plates 12-16. Locke, J. 1843. Notice of a New Trilobite, Ceraurus crosotus. Amer- ican Journal of Science, 44:346. Martinsson, A. 1967. The Succession and Correlation of Ostracode Faunas in the Silurian of Gotland. Geologiska Foreningens i Stockholm Forhandlingar, 89:350— 386, 3 figures. Miller, A. K., and W. Youngquist 1949. The Maquoketa Coquina of Cephalopods. Journal of Paleontology, 23:199-204, plates 40-42. Moore, R. C. (editor) 1959. Treatise on Invertebrate Paleontology, Part O, Arthropoda, 1. 560 pages, 415 figures. New York City and Lawrence, Kansas: Geological Society of America and University of Kansas Press. Ross, R. J., Jr. 1967. Calymenid and Other Ordovician Trilobites from Kentucky and Ohio. United States Geological Sur- vey Professional Paper, 583—B: 1—19, 5 plates. Shirley, J. 1936. Some British Trilobites of the Family Calymenidae. Quarterly Journal of the Geological Society of London, 92:384-422, plates 29-31. Slocum, A. W. 1913. New Trilobites from the Maquoketa Beds of Fayette County, Iowa. Field Museum of Natural History Geological Series, 4(3) :43-83, plates 13-18. Stumm, E. C, and E. G. Kauffman 1958. Calymenid Trilobites from the Ordovician Rocks of Michigan. Journal of Paleontology, 32:943-960, plates 123, 124. Tillman, C.G. 1960. Spathacalymene, an unusual New Silurian Trilobite Genus. Journal of Paleontology, 34:891-895, plate 116. Twenhofel, W. H, et al. 1954. Correlation of the Ordovician Formations of North America. Geological Society of American Bulletin, 65:247-298. Walter, O. T. 1927. Trilobites of Iowa and Some Related Paleozoic Forms. Iowa Geological Survey, 31:167-400, plates 10-27. Whittington, H. B. 1954. Ordovician Trilobites from Silliman's Fossil Mount, in A. K. Miller, W. Youngquist, and C. Collinson, Ordovician Cephalopod Fauna of Baffin Island. Geological Society of America Memoir, 62:119— 149, plates 59-63. 1965. The Ordovician Trilobites of the Bala Area, Merioneth, Part 2. Palaeontographical Society (Monograph), pages 33-62, plates 9-18. London. SILURIAN Arturo J. Amos and S. JVoirat A New Species of Ancilloto echia from the Zapla Formatio n, Northern Argentin a ABSTRACT A new species of a trigonirhynchid brachiopod, An- cillotoechia cooperensis, is described from beds of probable Wenlockian age from the Province of Jujuy, Argentina. It is the first reported occurrence of the genus in the southern hemisphere. The genus Ancillotoechia Havlicek (1959) is a trig- onirhynchid brachiopod that McLaren (in Moore, 1965, p. H561) distinguished from similar genera— Cupularostrum Sartenaer (1961) and Rostricelhda Ulrich and Cooper (1942)—because of its strong sub- rounded plications, covered septalium supported by a median septum in the brachial valve, and dental plates in the pedicle valve. Species of Ancillotoechia were described for the first time by Havlicek (1959) from the Bohemian Silurian: A. ancillans (Barrande) and A. radvani Havlicek from the lower beds of the Kopanina Limestones (Budfiany) and A. minerva (Barrande) from the Liten beds (Wenlock). Recently, a North American species from the St. Clair Limestone (Wenlockian), Camarotoechia marginata (Thomas), was included by Amsden (1968) in this genus. The occurrence of Ancillotoechia in Argentina re- ported here is the first in the southern hemisphere. The specimens were collected in the northern part of the Sierra de Puesto Viejo near Las Chaquetas, or Chaqueta, east of San Juancito Chico, in the Province Arturo J. Amos and S. Noirat, Facultad de Ciencias Naturales y Museo Universidad Nacional de La Plata, La Plata, Argentina. of Jujuy. Poorly preserved pelecypods were also found with the brachiopods. The authors are indebted to Dr. R. Bellmann for the information given on the fossil locality and to Dr. C. Rayces for the loan of specimens of the Instituto de Geologia y Mineria (Jujuy). Family TRIGONIRHYNCHIIDAE McLaren, 1965 Genus Ancillotoechia Havlicek, 1959 DIAGNOSIS.—Subcircular in outline with slightly divergent dental lamellae in the pedicle valve and a low median septum extending anteriorly as a low and delicate ridge in the brachial valve. TYPE-SPECIES.—Rhynchonella ancillans Barrande, 1879. Ancillotoechia cooperensis, new species PLATE 1: FIGURES 1-14 DESCRIPTION.—The specimens are of medium size, subglobose with a subcircular outline. Maximum widtii at midlength of valves. Valves moderately convex. The ventral sulcus and brachial fold conspicuous anteriorly. A few well-marked costae present in the sulcus, fold, and flanks. Anterior commissure uniplicate and ser- rated. No other ornamentation can be observed except delicate concentric growth lines. Shell structure unknown. Pedicle exterior with moderately pointed umbo with apical angle of 118°. Sulcus well marked anteriorly, flanked by two well-rounded costae, with two costae originating at the umbo. In some specimens a third costa intercalates between the marginal and the sulcus 139 140 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY LOW. DEV OVERLYING. PRID0LI LUDLOW ZAPLA FMT WENL0CK MEC0YITA FMT LLANDOVERY costae near the posterior third of the valve. Flanks evenly convex with four costae, including the marginal. Brachial exterior not so convex with fold less defined than corresponding sulcus, conspicuous at midlength widi one costa. In some specimens a second costa inter- calates between the marginal and the central costa. Flanks evenly convex with four costae including the marginal. Pedicle interior with very short dental lamellae, dorsally divergent, leaving a large and deep umbonal cavity. Lateral cavities very small, usually filled with secondary growth. Teeth strong and subrounded. Brachial interior with short septalium supported by a relatively short and thick septum. Septalium covered by small cardinal plate. Median septum extends an- teriorly as a thin and low ridge to about midlength of valve. Short and strong crural processes directed ventrally. Dimension of holotype and paratype (measurements in millimeters) : Holotype (SJ 114d) Paratype (SJ 114a) Length of pedicle valve 14. 5 11.6 Length of brachial valve 12.2 10.8 Thickness 10.0 7.9 Maximum width 14.9 11.3 Width of fold at commissure 6.2 2. 1 Length/width ratio 0.97 1.0 Length/thickness ratio 1.45 1.46 MATERIAL.—Several steinkerns, external and in- ternal impressions of ventral and brachial valves. Most of the specimens were found in ferruginous concretions. LOCALITY.—Las Chaquetas, or Chaqueta, east of San Juancito Chico, Sierra de Puesto Viejo, Jujuy Province, Argentina. TYPES.—Holotype, SJ 114d, and paratype, SJ 114a, in Instituto de Geologia y Mineria, Jujuy (IGM). Other specimens: SJ 113-17 (in IGM) ; MLP 10761-2 (in Museo de La Plata). REMARKS.—This new species, in general, resembles the type of Ancillotoechia ancillans (Barrande), but it is subrounded and globose, with well-marked ribbing. Internally, the dental lamellae are slightly divergent, leaving a large umbonal cavity. Other described spe- cies are, however, very different in shape and ribbing. Internally, A. marginata (Thomas) from the St. Clair Limestone (Amsden, 1968) is similar, but externally is subtriangular in outline, with the surface covered by 4 or 5 costae in 3 mm. UNDERLYING CENTINELA FMT. FIGURE 1.—Silurian stratigraphic sequence, northern Argentina. Geologic Setting Formations of Silurian age are exposed in nortiiern Argentina along several sub-Andean ranges east of Jujuy City. Some exposures are also known near the Argentine-Bolivian border. Stratigraphically, the Silu- rian rocks lie unconformably on top of the Centinela Formation of Caradocian age (Harrington and Leanza, 1957, p. 8). The sequence starts with a marine glacial conglomerate, 30 meters thick, called Mecoyita Forma- tion (Turner, 1960) or "Horizonte Glacial de Zapla." This formation was formerly considered the top of the Centinela Formation of Caradocian age by Harrington and Leanza (1957, p. 8), of Silurian age by Schlagint- weit (1943). On top of the Mecoyita Formation lies die Zapla Formation, 640 meters thick, composed of dark gray to light blue, yellow and reddish shaly sandstones, and a ferruginous sandstone member about 5 to 25 meters thick near the base of the sequence. A paracon- formity is evident between the Zapla Formation and the overlying Lower Devonian rocks (Figure 1). Fossils were found in the Zapla Formation, collected during economic studies of high-grade iron ore found NUMBER 3 141 10 ft 13 14 PLATE 1.—Ancillotoechia cooperensis, new species. 1-5, Holotype, SJ 114d, (X 2) : 1, brachial exterior; 2, pedicle exterior; 3, anterior commissure; 4, lateral view, 5, posterior view. 6—9, Paratype, SJ 114a (X 2) : 6, brachial exterior; 7, pedicle exterior; 8, lateral view; 9, anterior commissure. 10-14, Other specimens: 10, pedicle exterior, SJ 115a; 11, brachial exterior, SJ 116a (X 2); 12, rubber mold of pedicle exterior, SJ 113c (X 2) ; 13, rubber mold of pedicle interior, MLP 10762 (X 2); 14, rubber mold of holotype, SJ 114d, posterior view ( X 3). 142 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY in the ferruginous sandstone member. Fossils also were found during oil exploration in the sub-Andean ranges. However, no descriptions of the coelenterates, brachio- pods, mollusks, trilobites, and graptolites have ever been made (Castellaro, 1966). Therefore, the age of the beds above the fossiliferous Ordovician was merely assumed or the age was assigned by comparison with known faunas of other parts of Argentina and Bolivia. For example, Angelelli (1946) mentioned the trilobite Calymene in his study of the Zapla ferruginous ore, and Cecioni (1953) mentioned finding a specimen of ^HomalonotusT The Calymene was compared with C. blumenbachi by Harrington (1956), who also men- tioned that the rocks of the Zapla area have yielded Scotiella, Clarkeia, and Monograptus cf. priodon. The trilobites mentioned above are known to oc- cur in formations from the Silurian up to the Middle Devonian in other parts of the world. The well-known Silurian guide fossil Clarkeia antisiensis (d'Orbigny) also has been mentioned in the geological literature of this area, but the authors were unable to verify this identification. This new species of Ancillotoechia is the first fossil ever to be described in Northern Argen- tina from the Silurian Period. Literature Cited Amsden, T. W. 1968. Articulate Brachiopods of the St. Clair Limestone (Silurian) Arkansas, and the Clarita Formation (Silurian) Oklahoma. The Paleontological Society Memoir 1, Journal of Paleontology, 42 (3, supple- ment) : 1-117, 20 plates, 83 figures. Angelelli, V. 1946. La Geologia y Genesis del Yacimiento ferrifero de Zapla. Revista de la Asociacion Geologica Argen- tina, 1(2): 117-147. Castellaro, H. 1966. Faunas Siluricas, in Guia Paleontologica Argentina. Consejo Nacional de Investigaciones y Cientificas y Tecnicas, part 1, section 3, pages 1-57. Cecioni, J. 1953. Contribucion al Conocimiento de los Nautiloideos Eopaleozoicos Argentinos. Part 1. Protocyclocera- tidae—Cyclostomiceratidae. Boletin Museo Nacional de Historia Natural Chile, 26(2) :57-110. Harrington, H. J. 1956. Argentina, in W. F. Jenks, Handbook of South America Geology. Geological Society of America Memoir, 65:131-165, 1 plate. 1962. Paleogeographic Development of South America. American Association of Petroleum Geologists Bulletin, 46(10): 1773-1814. Harrington, H. J., and A. F. Leanza 1957. Ordovician Trilobites of Argentina. University of Kansas, Department of Geology Special Publica- tion, 1: 1-276, 140 figures. Havlicek, V. 1959. Spiriferidae v ceskem siluru a devonu (Brachio- poda) . Ustfedniho Ustavu Geologickeho Rozpravy, 25: 1-275, 28 plates, 101 figures. 1961. Rhynchonelloidea des bomischen alteren Palao- zoikums (Brachiopoda). Ustfedniho Ustavu Geo- logickeho Rozpravy, 27: 1-211, 27 plates, 87 fig- ures. Moore, R. C. (editor) 1965. Treatise on Invertebrate Paleontology. Part H, Brachiopoda. 927 pages, 746 figures. New York City and Lawrence, Kansas: Geological Society of America and University of Kansas Press. Schlagintweit, O. 1943. La Posicion Estratigrafica del Yacimiento de Hierro de Zapla y la Difusion del Horizonte Glacial de Zapla en la Argentina y Bolivia. Revista Minera, 13(4): 115-127. Turner, J. C. M. 1960. Estratigrafia de la Sierra de Santa Victoria y Adyacencias. Boletin de la Academia Nacional de Ciencias en Cdrdoba, 41: 163-196. Thomas W. Amsden Triplesia alata Ulrich and Cooper ABSTRACT Ulrich and Cooper's description of Triplesia alata, 1936, was based on specimens from the "Brassfield" Limestone of Searcy County and the Batesville district of north-central Arkansas, but recent investigations indicate that the Batesville specimens are from die Cason Shale. This species is also present in the Black- gum Formation of eastern Oklahoma and in the Cochrane Formation of south-central Oklahoma. The strata bearing Triplesia alata are believed to be of early upper Llandoverian age (C1-2 zone) and slightly older than the Brassfield strata of Dayton, Ohio. This species has the typical, elongate, bifid triplesiid cardinal process with a hoodlike chilidium near its base. The chilidium fits into a shallow ventral delthyrium and the pseudo- deltidium has a median ridge produced by a thickening of the shell. All known North American representatives of Triplesia sensu stricto are Llandoverian or older; Wenlockian species formerly assigned to this genus are now referred to Placotriplesia. The biostratigraphic distribution of these genera in European strata is uncertain. Triplesia alata Ulrich and Cooper is distributed over a wide area in Arkansas and Oklahoma (Figure 1). It is present in strata of late Lower Silurian age (upper Llandoverian) and is thus one of the late representa- tives of the genus Triplesia sensu stricto. Miser (1922, p. 29, pi. 7a, figs. 1-3) first reported specimens from the "Brassfield" Limestone of the Batesville district in north-central Arkansas (referred to Triplesia ortoni). Later, Ulrich and Cooper (1936, p. 346, pi. 48, fig. 23, pi. 50, figs. 11, 14, 16, 17, 20, 21, 24) referred these shells to a new species, Triplesia alata, basing their description on specimens from the Batesville district, Thomas W. Amsden, Oklahoma Geological Survey, Norman, Oklahoma. Publication authorized by director, Oklahoma Geological Survey. and from Searcy County, Arkansas. In 1960,1 reported this species from the Cochrane Formation of the Arbuckle Mountains and Criner Hills in south-central Oklahoma (Amsden, 1960, p. 50), and later it was found in the Blackgum Formation of northeastern Oklahoma (Amsden and Rowland, 1965, pp. 15, 17, 21, 31, 93). The present study is based on specimens from the Searcy County and Batesville districts of Arkansas, including Ulrich and Cooper's type speci- mens, and on collections from the Blackgum and Cochrane Formations of Oklahoma. A study of these collections totaling over 100 specimens furnishes addi- tional details on external and internal morphology and on the biostratigraphic and geographic distribution of this species. This study also supplies additional informa- tion on the morphology and biostratigraphic range of the subfamilies Triplesiinae and Placotriplesiinae. I thank Dr. G. Arthur Cooper for the loan of the United States National Museum collections of Triplesia alata and T. ortoni. The repositories of the figured specimens are abbreviated as follows: United States National Museum, USNM; University of Oklahoma, OU. Stratigraphic and Geographic Distribution Ulrich and Cooper (1936, p. 346) based their de- scription of Triplesia alata on specimens from the "Brassfield" Formation near Gilbert, Searcy County, Arkansas. These authors also noted tiiat one speci- men from the Brassfield at Soldiers Home, Dayton, Ohio, "has the characters of this species." The speci- men which Ulrich and Cooper (1936, pi. 50, figs. 16, 17, 20, 21) illustrated from Cason Mine and which herein is designated the lectotype and refigured (Plate 2: figures 1-5) is the same specimen illustrated earlier by Miser (1922, pp. 28-29, pi. 7a, fig. 1) and reported to be from residual clays in the Montgomery Mine. The stratigraphic position of T. alata has been some- 143 144 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Boles.MIe . outcrop area V / 'rBRASSRELCT . \ LIMESTONE CASON SHALE I OKLAHOMA i 5 Cherokee County j CLARITA ,Arbuckle Mountains V^ _ FORMATION \ Seo,c ARKANSAS w.. / \^.'-\_-- ..J FIGURE 1.—Outcrop areas of the Cason Shale, "Brassfield" Limestone, Blackgum Formation and Cochrane Formation. thing of a puzzle because at Cason Mine, and at most places in the Batesville district, the Cason Shale is directly overlain by typical St. Clair Limestone. The only megafossils heretofore reported from the Cason Shale are specimens of Girvanella, which Ulrich (in Miser, 1922, p. 28) assigned to an Upper Ordovician age. The overlying St. Clair Limestone is richly fossil- iferous and yields many brachiopods, but includes no representative of this species, or of the genus Tri- plesia sensu stricto (Amsden, 1968, pp. 11, 12). It was, therefore, of considerable interest when, in September 1966, T. L. Rowland, O. A. Wise, Jr., and I visited Love Hollow Quarry and found a lens of limestone within the Cason Shale that yielded several specimens of T. alata. It now seems reasonable to assume that all of the specimens from the Batesville district originally came from some part of the Cason Shale (see Amsden, 1968, pp. 5-7). In the fall of 1967 I revisited Love Hollow and found that the limestone lens had been completely removed by quarry operations; however, its relationship to the Cason Shale and the overlying St. Clair Limestone has been illustrated (Amsden, 1968, fig. 5). The Searcy County specimens are from the "Brass- field: Tomahawk Creek (? = Buffalo River), 6 miles east of St. Joe, /2 to 1 mile west of Gilbert" (Ulrich and Cooper, 1936, p. 346). I have not examined this limestone in the Gilbert area, but Maher and Lantz (1953) describe it as pink, crystalline limestone. In this region the "Brassfield" Limestone is reported to be underlain by the Cason Shale and overlain by the St. Clair Limestone. No representative of T. alata or of any form of the genus Triplesia has been reported from the St. Clair Limestone in the Gilbert area (Ams- den, 1968, p. 9). Triplesia alata is present in the upper ten feet of the Blackgum Formation at Blackgum Landing on the south shore of Lake Tenkiller (NW/4SW/4SE/4 sec. 32, T 14 N, R 22 E), Cherokee County, eastern Okla- homa (upper limestone member, stratigraphic section Ch2-D, Amsden and Rowland, 1965, p. 93). About 20 specimens have been found in the Blackgum For- mation, where they are associated with Microcardi- nalia protriplesiana Amsden (1966, p. 1010). R 1 W RIE MILES Pl-C - fossil locality COCHRANE FORMATION FIGURE 2.—Distribution of the Cochrane Formation in the Arbuckle Mountains and Criner Hills of Oklahoma. Numbers (Pl-C, etc.) refer to collecting localities yielding specimens of Triplesia alata (see Amsden, 1960, appendix; modified from Amsden, 1960, fig. 15). NUMBER 3 145 Triplesia alata is widely distributed in the Cochrane Formation of the Arbuckle Mountains and Criner Hills of south-central Oklahoma (Figure 2). About 100 specimens, many of which are articulated shells, have been collected from this formation at the follow- ing localities: Cl-C (upper 5 ft.); Cal-B (8 ft. above base); Call-C; Cal4-C, B; J4-C (upper 2 ft.); J5-B (5 ft. below top); M14-A; Pl-C. Descrip- tions of these localities may be found in Amsden (1960, appendix). Triplesia alata also has been collected from loose blocks of Cochrane at the Ideal Cement Quarry (SE/4 sec 36, T 3 N, R 5 E), Pontotoc County, Okla- homa, and in Johnston County (SW*4 sec. 2, T 2 S, R 7 E), Oklahoma. The Cochrane has a maximum thickness in the Arbuckle Mountains-Criner Hills re- gion of about 60 feet, and T. alata appears to range through most of the formation. Age of the Triplesia alata Beds The brachiopod fauna associated with the strata bearing Triplesia alata is small. The Cason Shale lime- stone lens at Love Hollow Quarry in the Batesville district does include representatives of Plectodonta, Meristina, Streptis, and a few others, but based on present knowledge none furnishes critical age data. Undoubtedly the most significant association is the presence of Microcardinalia protriplesiana in the beds of the Blackgum Formation bearing Triplesia alata in eastern Oklahoma. In 1966 I described this species and discussed its age in terms of the stricklandiid phyl- ogenetic studies of Williams and of Boucot and Ehlers (Amsden, 1966, pp. 1010-1015, fig. 1). The structure of the dorsal apparatus of Microcardinalia protriplesiana is similar to that of Stricklandia lens progressa Williams from Llandovery strata in the Cl-2 zone, suggesting a correlation with the early upper Llandoverian. On the other hand, Microcardinalia triplesiana (Foerste) from the Brassfield Formation at Soldiers Home, Dayton, Ohio, has a different dorsal structure; and, on the basis of these pre- sumed phylogenetic differences, I concluded that Foerste's species was distinctly younger than M. protriplesiana and assigned it to zone C4-5. Carlson and Boucot (1967, p. 1122) have questioned this age relationship, pointing out, quite correctly, that M. triplesiana and M. protriplesiana have never been found in stratigraphic juxtaposition. The Brassfield at Soldiers Home also yields specimens of Triplesia ortoni (Meek), a species which is easily distinguished from T. alata. On the basis of this evidence it is here sug- gested that the beds bearing T. alata are correlative with one another, and are older than the T. ortoni UPPER SIL WENLOCK ARBUCKLE MTS CRINER HILLS EASTERN OKLAHOMA ? BATESVILLE DISTRICT ARKANSAS WEST CENTRAL OHIO CEDARVILLE DOL SPRINGFIELD DOL EUPHEMA DOL MASSIE SHALE LAUREL LS OSGOOD SHALE DAYTON LS CLARITA FORMATION ST CLAIR LIMESTONE LO W ER SI L U RI A N C LLANDOVERY B A ? 1 BRASSFIELD DOU ;:;:• M triplesiana,'.-. ? TENKILL ER FORMATION CjCHRANE FM 1 .BLACKGUM FM 1 \M prolnplesiona\ % CASON SHALE;!: M'fnestone lens with:; ? KEEL FORMATION BELFAST FORMATION ORD SYLVAN SHALE SYLVAN SHALE FERNVALE LS RICHMOND GR FIGURE 3.—Suggested correlation of the Triplesia-bearing beds (dotted) in Oklahoma, Arkansas, and western Ohio, and the brachiopod-bearing strata (toned) of the St. Clair Limestone and Clarita Formation. The ages of the other strata shown are not given detailed consideration in the present report. 146 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY beds as shown in Figure 3. Microcardinalia protriples- iana is found only in the Blackgum Formation, and its correlation with the Cochrane Formation and Cason Limestone lens is based almost entirely on the presence of Triplesia alata, there being little corroborative brachiopod evidence. It is emphasized that this correla- tion is proposed only for the beds of the Cason Shale bearing T. alata and that the other parts of this stratigraphic unit may have a different age. Rexroad (1967, pp. 12-14) has pointed out that there is a discrepancy between the age of the Brassfield Formation as determined by conodont correlations with the standard British section and the age based on brachiopod correlations. According to that author the Brassfield conodonts in the Cincinnati Arch region—and this includes the Brassfield at Soldiers Home, Dayton, Ohio (Rexroad, 1967, locality 15, p. 19), represent the upper part of Walliser's Bereich I zone. This part of the Bereich I zone, which includes the Monograptus cyphus, M. gregarius, M. convolu- tus graptolite zones, falls almost entirely within the middle Llandovery zone B. In contrast, M. triplesiana from the Brassfield at Dayton indicates a correlation with upper Llandoverian zone C4-5, and even M. protriplesiana, which is the oldest Microcardinalia zone now recognized in Ohio, Arkansas, Oklahoma, and Nebraska, falls in the upper Llandoverian zone C1-2 and is younger than Bereich I. The genus Triplesia, based on T. extans (Emmons) from the Middle Ordovician, generally has been de- fined to include species ranging into the Wenlockian. I suspect, however, that most if not all of the Wen- lockian species referred to Triplesia actually are repre- sentatives of Placotriplesia (Amsden, 1968, p. 40) and, at least in eastern North America, Triplesia sensu stricto appears to be confined to Llandoverian and older strata (see discussion under Triplesiinae and Placotriplesiinae). Internal and External Morphology As Ulrich and Cooper (1936, p. 346) have given a detailed description of the external characters of Triplesia alata, only a few points, concerned mainly with the effect of growth on relative shell proportions, need be discussed here. Immature specimens of T. alata, i.e., shells under ten millimeters long, have an erect ventral beak which stands well above the brachial umbo. The fold and sulcus on these small individuals are shallow and poorly defined, and the width is only slightly greater than the length (length/width ratio about 0.80; see Figure 4). With increased size the dorsal umbo becomes swollen so that it stands as high or higher than the pedicle beak, and the ventral sulcus and dorsal fold develop into strong, well-defined shell features. The lateral component of growth at this later stage of development was much stronger than the anterior component, so that large specimens are markedly transverse; the length/width ratio among specimens over 15 millimeters long generally falls be- tween 0.65 and 0.70 (Figure 4). All shells are essen- tially smooth, although a few of the largest individuals may show faint plications. Specimens from the Cason Shale (limestone lens) near Batesville, the "Brassfield" Limestone of Searcy County, and the Blackgum and Cochrane Formations of Oklahoma appear to be similar in all respects. Triplesia alata has the characteristic, prominent tri- plesiid cardinal process (Figure 5; Plate 1: figures 2-4). This process has a long shaft which bifurcates, producing two rodlike myophores extending into the posterior part of the ventral valve. Each myophore is partially cleft, probably to give the diductor muscles a better point of attachment. The base of the cardinal process is much thickened and expanded laterally to produce a stout structure out of which the sockets are excavated. A small collarlike or hoodlike structure is present on the dorsal side of the cardinal process near the base. The ventral interarea is well developed, and the stout teeth are supported on long dental plates. The pseudodeltidium is marked by a thicken- ing of the shell wall which produces a ridge on the external and internal side of the interarea; this ridge is clearly a part of the interarea, as the growth lines of the latter pass without interruption tirrough the pseudodeltidium (Figure 5; Plate 1: figure 1). Near the front margin the pseudodeltidium is notched back (i.e., posteriorly) to make a delthyrial opening into which the cardinal process collar fits (Figures 5c, 6). Presumably the collar corresponds to the chilidium, although this structure in the triplesiids is small be- cause of the extravagantly developed cardinal proc- ess. The adductor muscles make four moderately deep scars in the brachial valve (Plate 2: figure 3). The pedicle muscle scars have not been observed in T. alata, but they probably are similar to those in T. ortoni (Plate 2: figure 12). Triplesia alata is most similar to Triplesia anticosti- ensis Twenhofel (1914, p. 26; 1928, p. 198, pi. 18, figs. 1-5) from the lower part of the Jupiter Forma- NUMBER 3 147 / I / / J H D 0 n 0 > WENLOCKIAN c P woldronensis (Miller and Dyer) P rostellata (Ulrich and Cooper) r r c Ppraecipta (Ulrich and Cooper) j Pjuvenis (Ulrich and Cooper) LLANDOVERIAN T ortoni (Meek) T. anticostiensis Twenhofel T alata Ulrich and Cooper 7 'z H D D n 0 > ORDOVICI AN UPPER r r c ? I MIDDLE T extans (Emmons) T carmota Cooper* T subcarmata Cooper* T nucleus (Hall)* T cuspidota (Hall)* FIGURE 9.—Distribution of North American species of Triplesia and Placotriplesia. The structure of the pseudo- deltidium on species marked with an asterisk has not been determined. the Dayton region; moreover, the brachiopod faunas suggest that both the T ortoni and T. alata faunas should be assigned to the upper Llandovery Zone C, whereas Rexroad correlates the conodont faunas of the Brassfield of Dayton with the upper Bereich I zone, which he assigns to the middle Llandovery Zone B. The brachiopod fauna of the St. Clair Limestone that I recently described (Amsden, 1968) falls largely, perhaps entirely, within Craig's "sagitta zone," al- PLATE 1: figures 1-4, 13-15, 21.—Triplesia alata Ulrich and Cooper, Blackgum Formation, upper limestone member, south shore of Lake Tenkiller, SE/4SW/4, sec. 32, T 14 N, R 22 E, Cherokee County, Oklahoma (loc. Ch2-D, Amsden and Row- land, 1965, p. 93). 1-4, Transverse serial sections (see Figure 5 for other illustrations of this series) : 1, part of pedicle palintrope showing dental plates and pseudeltidial ridge (X 12), 0.9 mm in front of beak; 2, brachial cardinalia show- ing bifid tips of cardinal process (X 8) at 1.9 mm; 3, brachial cardinalia (X 8) at 2.1 mm; 4, brachial cardinalia (X 8) at 2.2 mm., OU 5681. 13, Posterior view of a nearly complete articulated shell (X 1), OU 5688. 14, Brachial view of a large specimen (X 1), OU 5689. 15, Pedicle view (X 1), OU 5690. 21, Brachial view (X 1), OU 5691. Figures 5-10, 12.—Triplesia alata Ulrich and Cooper, Cochrane Formation, Arbuckle Mountain-Criner Hills region, Carter County, southcentral Oklahoma. Specimen in figure 5 is from Rock Crossing, NE^SE/4, sec. 35, T 5 S, R 1 E (loc. Call-C. Amsden, 1960, p. 208) ; specimens in figures 6-10, 12 are from Henryhouse Creek, SEJ/4 sec. 30, T 2 S, R 1 E (eight feet above base of the formation). 5, Brachial valve (X 1), OU 5682; 6, 7, lateral view (X 2) and posterior view (X 1) of an incomplete shell, OU 5683; 8, anterior view (X 1), OU 5684; 9, pedicle view (X l),OU5685; 10, lateral view (X 2) of a small shell, OU 5686; 12, pedicle view (X l),OU5687. Figures 11, 16-20.—Triplesia alata Ulrich and Cooper, Cason Shale, limestone lens, Love Hollow quarry, SWJ4 sec 4, T 14 N, R 8 W, Batesville district, Izzard County, Arkansas (see Amsden, 1968, p. 6) ; all views natural size. 11, pedicle view of small shell, OU 5692; 16, pedicle view, OU 5693; 17, brachial view, OU 5694; 18, pedicle view, OU 5695; 19, pedicle view of a large specimen, OU 5697; 20, brachial view, OU 5696. Figures 22-25.—Triplesia alata Ulrich and Cooper, United States National Museum specimens from the "Brassfield" Limestone, Searcy County, Arkansas; all views natural size. Specimen in figure 22 is from Tomahawk Creek, six miles east of St. Joe; specimens in figures 23-25 are from one-half to one mile west of Gilbert. 22, brachial view of USNM 91874 (figured by Ulrich and Cooper, 1936, pi. 50, fig. 11); 23, pedicle view of USNM 91873, an unfigured paratype; 24, pedicle view of USNM 91873a (figured by Ulrich and Cooper, 1936, pi. 50, fig. 24); 25, brachial view of USNM 91873 (shell figured by Ulrich and Cooper, 1950, pi. 50, fig. 14). NUMBER 3 151 P "?'"' ''r-jf " fX "-V'. , 1/ 22 23 24 25 PLATE 1 152 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY 1 i 1 ^f * ? # 1 m - ? ^*>.a^> 7 1 - -. • ***?•**•" V'v 10 > 11 12 [ 14 v -?' 13 4 ' * / S 17 «'• 15 18 PLATE 2 NUMBER 3 153 though there could be some overlap on the underlying ^amorphognathoides^ and overlying "siluricus" cono- dont zones. Literature Cited Amsden, T. W. 1960. Hunton Stratigraphy, Stratigraphy and Paleon- tology of the Hunton Group in the Arbuckle Mountain Region, Part 6. Oklahoma Geological Survey Bulletin, 84:1-311, 17 plates, text-figures. 1966. Microcardinalia protriplesiana Amsden, a New Species of Stricklandiid Brachiopod, with a Dis- cussion on Its Phylogenetic Position. Journal of Paleontology, 40: 1009-1016, 3 plates. 1968. Articulate Brachiopods of the St. Clair Limestone (Silurian), Arkansas, and the Clarita Formation (Silurian), Oklahoma. The Paleontological Society Memoir, 1, Journal of Paleontology, 42(3, supple- ment) : 1-117, 20 plates, 83 figures. Amsden, T. W., and T. L. Rowland 1965. Silurian Stratigraphy of Northeastern Oklahoma. Oklahoma Geological Survey Bulletin, 105:1-174, 18 plates, and maps and text-figures. PLATE 2: figures 1-5.—Triplesia alata Ulrich and Cooper. "Brassfield Formation" (probably from the Cason Shale), Cason Mine ( ?Montgomery Mine), Batesville district, Arkan- sas. This steinkern was illustrated by Miser (1922, pi. 7, fig. 1) as Triplesia ortoni from the Montgomery Mine; it was re-illustrated by Ulrich and Cooper (1936, pi. 50, figs. 16, 17, 20, 21), who assigned it to their new species T. alata and cited the locality as Cason Mine, USNM 91872. This speci- men is here designated as the lectotype. Views are natural size. 1—4, anterior, posterior, pedicle, and brachial views; 5, rubber cast of the brachial valve showing muscle scars and incomplete cardinal process. Figures 6-17.—Triplesia ortoni (Meek), Brassfield Forma- tion, Soldiers Home, Dayton, Ohio. 6-8, Lateral, pedicle, and brachial views (X 1). USNM 85136a (figured in Shimer and Shrock, 1944, pi. 117, figs. 19-21); 9-11, pedicle (X 1.3), posterior (X 2), and posterior (X 1) views of a large shell, USNM 85136b; 12, 14, interior (X 2) and pos- terior (X 3) views of an incomplete pedicle valve, USNM 91871b (figured by Ulrich and Cooper, 1936, pi. 48, fig. 32); 13, 16, exterior and interior views (X 3) of the cardinal proc- ess, USNM 91871; 15, 17, external and lateral views of a dorsal cardinalia showing collar or chilidium, USNM 9187a (illustrated by Ulrich and Cooper, 1936, pi. 48, figs. 34, 35). Figure 18.—Placotriplesia praecipta (Ulrich and Cooper), Clarita Formation, Fitzhugh Member, Chimneyhill Creek, NEJ4SEJ4 sec 5, T 2 N, R 6 E, Pontotoc County, Okla- homa. Posterior view (X 7) showing palintrope with no evidence of pseudodeltidium. (This is specimen OU 6368, illustrated in Amsden, 1968, pi. 18, fig. 3k.) Barrande, J. 1879. Brachiopodes. Recherches paleontologiques, volume 5 in Systeme silurien du center de la Boheme. 226 pages, 153 plates. Prague. Craig, W. W. 1969. Lithic and Conodont Succession of Silurian Strata, Batesville District, Arkansas. Geological Society of America Bulletin, 80(8) : 1621-1628. Carlson, M. P., and A. J. Boucot 1967. Early Silurian Brachiopods from the Subsurface of Southeastern Nebraska. Journal of Paleontology, 41:1121-1125, text-figures. Davidson, T. 1871. Monograph of British Silurian Brachiopoda. Palaeontolographical Society, 7(4) : 249-397, plates 37-50. 1883. Monograph of British Fossil Brachiopoda; Supple- ment to the British Silurian Brachiopoda. Palaeontographical Society. 5(2):64-242, plates 8-17. Hall, J., and J. M. Clarke 1892. An Introduction to the Study of the Genera of Palaeozoic Brachiopods, Part 1. Palaeontology of New York (New York Geological Survey), 8:1— 367, 20 plates. Maher, J. C, and R. J. Lantz 1953. Geology of the Gilbert Area, Searcy County, Arkansas. United States Geological Survey Oil and Gas Investigation Map OM-132. Meek, F. B. 1872. Description of a Few New Species, and One New Genus of Silurian Fossils from Ohio. American Journal of Science, series 3, 4:274—28-1. Miser, II. D. 1922. Deposits of Manganese Ore in the Batesville Dis- trict, Arkansas. United States Geological Survey Bulletin, 734:1-273, 17 plates (including geologic map), text-figures. Nicoll, R. S., and C. B. Rexroad 1968. Stratigraphy and Conodont Paleontology of the Salamonie Formation and the Lee Creek Member of the Brassfield Limestone (Silurian) in South- eastern Indiana and Adjacent Kentucky. Indiana Department of Natural Resources, Geological Sur- vey Bulletin, 40:1-73. Reed, F. R. C. 1905. Ordovician and Silurian Brachiopods of the Girvan District. Transactions of the Royal Society of Edinburgh, 51:795-998, plates 1-24. Rexroad, C. B. 1967. Stratigraphy and Conodont Paleontology of the Brassfield (Silurian) in the Cincinnati Arch Area. Indiana Department of Natural Resources, Geologi- cal Survey Bulletin, 36:1-64, 4 plates. Shimer, H. W., and R. R. Shrock 1944. Index Fossils of North America. 837 pages, 303 plates. New York: John Wiley and Sons, Inc. 154 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Twenhofel, W. H. 1914. The Anticosti Island Faunas. Canada Department of Mines, Geological Survey, Museum Bulletin, 3:1-38, 1 plate. 1928. Geology of Anticosti Island. Canada Department of Mines, Geological Survey, Memoir, 154:1—48.1, 60 plates. Ulrich, E. O. and G. A. Cooper 1936. New Silurian Brachiopods of the Family Triplesii- dae. Journal of Paleontology, 10:331-347, plates 48-50, text-figures. Williams, A. 1951. Llandovery Brachiopods from Wales with Special Reference to the Llandovery District. Quarterly Journal of the Geological Society of London, 107: 85-136, plates 3-8. Wright, A. D. 1963. The Morphology of the Brachiopod Superfamily Triplesiacea. Palaeontology, 5:740-764, plates 109, 110, text-figures. Arthur J. Boucot A eYligmCLStrOphia, New Genus, a Difficult Silurian Brachiopod ABSTRACT A new genus of strophic shell, collected from Late Silurian (Ludlow) age beds in northern California and north-central Nevada, is described. The familial affinities of this shell are reviewed, and it is concluded that it cannot be placed in any presently defined family category/ The limited material precludes additional work on the problem at this time. During the summer of 1964 I visited several Late Silurian and Early Devonian fossil localities—discov- ered by Mr. Rodney Gregg, Gazelle, California, a former field assistant for Dr. Francis Wells of the United States Geological Survey—in the area between Gazelle and Callahan in the Klamath Mountains of northern California. At one of these localities—United States National Museum (USNM) locality 11162— a number of blocks of sandstone rich in plicate pen- tamerinids (Kirkidium sp.) were collected. During the preparation of these blocks, which had a total weight of about 100 pounds, there were obtained two speci- mens of a large, completely unfamiliar, strophic shell. Informal consultation with colleagues during the next few years produced no substantial suggestions as to the affinities of these two specimens, which were relegated to a drawer of unidentified brachiopods. During the summer of 1965;, in collaboration with Dr. Michael Murphy, of the University of California at Riverside, I began an extensive program of fossil Arthur J. Boucot, Department of Geology, Oregon State Uni- versity, Corvallis, Oregon 97331. collecting from the Roberts Mountains Formation in several areas on the north and west sides of the Roberts Mountains of Eureka County, Nevada. The subse- quent processing of many tons of Silurian limestone containing silicified brachiopods produced (during sorting by Dr. J. G. Johnson) three specimens from USNM localities 13257 and 13237 that were identical, in most respects, to the two unfamiliar specimens col- lected earlier with Mr. Gregg in the Klamath Mountains. The Klamath Mountain locality is difficult to date other than as Late Wenlock to Ludlow age. The fos- sils obtained from this locality are as follows: JCirki- dium sp. (flat type), Atrypella cf. A. prunum, Atrypa "recticularis," Howellella sp., and tetracorals (sent to Dr. C. W. Merriam for study). The occurrence of Atrypella in association widi Kirkidium is suggestive of a Ludlow rather than a Pridoli age. Kirkidium does not normally range above the Ludlow—with the exception of a few spots in the Old World—and is unknown in proved Pridoli age beds in North America. The Klamath Mountain lo- cality cannot, at present, be tied into the occurrences in the Peyton Ranch Limestone Member of the Gazelle Formation (which is exposed a few miles to the east) because of lack of mapping in this region. Relatively detailed mapping of the area of Silurian rocks exposed to the west of Gazelle during the summer of 1963 did not turn up any fossiliferous sandstone of the kind yielding fossils from Gregg's locality, which is a few miles farther west. All of the shelly fossils in the Gazelle area have been collected from either the Peyton Ranch Limestone Member or calcareous shales occurring no more than a few meters below the base of the lime- stone. It is possible that Gregg's locality represents a 155 156 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY lateral equivalent of the Gazelle Formation, which in- cludes a variety of rock types, but it also it possible that the locality represents a higher horizon than is preserved in the Gazelle area. The Peyton Ranch Lime- stone Member appears to be the highest Silurian unit preserved in that area. The age of the Roberts Mountains Formation occur- rence (USNM loc. 13257) is known with far more certainty. The new genus occurs with Ptychopleurella sp., Dolerorthis sp., Dalejina sp., Isorthis sp., Kirkidium sp. (both small fine-ribbed and large fine-ribbed), Cymbidium sp. (coarsely plicate and finely plicate forms), Gypidula sp., orthotetacid brachiopod, Atrypa sp., Atrypella sp., Spirigerina cf. S. marginalis, Gra- cianella crista, Nucleospira sp., Howellella? sp., and Alaskospira sp.—an association of undoubted Ludlow age. The second Roberts Mountains collection (USNM loc. 13237) has yielded tire following forms: Ptychopleurella sp., Dolerorthis sp., long-lobed Dicoe- losia sp., Isorthis? sp., Pentamerifera sp., Conchidium aft0, bilocularis, Conchidium sp., Severella aft", magni- ficaformis, Gypidula sp. (both smooth and plicate forms), Clorinda sp., aff. Gacella sp., Ferganella sp., Atrypa sp., Atrypella sp., Dubaria sp., Gracianella lis- sumbra, G. plicumbra, Nucleospira sp., Protathyris? sp., Hedeina aff. balchaschensis, and Alaskospira? sp.— which, taken together, indicate a Ludlow age. Superfamily and Family Uncertain Aenigmastrophia, new genus PLATE 1: FIGURES 1-15 DIAGNOSIS.—The unique features of this genus are a combination of gently concavoconvex cross section, laterally elongate strophic outline, and a relatively smooth exterior associated widi a laterally directed, narrow pair of cardinal process lobes l flanking a vari- ably developed ridgelike median septum in the brachial valve and a raised pedicle callist area in the posterior portion of the pedicle valve flanked by a pair of elon- gate, anterolateral^ directed hinge-teeth. 1 The designation of these structures is difficult. In other strophic shells, paired projections that have been inferred to have served as lophophore supports are termed "cardinal process lobes," but lateral to the cardinal process lobes there may occur additional paired projections termed socket ridges. The position of the projections in the herein-described shells is akin to that of socket ridges, but their function is inferred to be similar to that of cardinal process lobes. TYPE-SPECIES.—Aenigmastrophia cooperi, new spe- cies. REMARKS.—It is natural to attempt a comparison of Aenigmastrophia with other strophic shells. The genera of the Stropheodontidae are effectively re- moved from consideration by the absence of a denticu- late hinge-line, a differently organized pedicle valve interior insofar as the muscle field is concerned, and the differing organization of the cardinalia and muscle field in the brachial valve. The absence of well-im- pressed muscle fields in either valve is one of the dis- tinctive features of this novel genus. The isogrammids, as represented by Isogramma itself, at first glance pos- sibly would appear to be allied, but the absence of the critical trapezoidal-shaped attachment scar on the umbo of the pedicle valve as well as the differing form of the cardinalia make this comparison unlikely in view of our present knowledge. The complex internal structures of the leptaenids find no counterpart in this genus except in regards to the presence of cardinal process lobes in the brachial valve. It is probable that pseudopunctae are present in Aenigmastrophia owing to the pustulose nature of the interior of one of the better-preserved specimens from Nevada. The genera of the Chonetacea are effectively removed from com- parison because of their different cardinalia and pedi- cle valve interior, and also by the absence of spines or spine bases in Aenigmastrophia. The interiors of both valves in this peculiar new genus differ so greatly from what we know of Middle Devonian productids that a comparison of this Late Silurian form does not ap- pear profitable despite the strophic nature of both; the lack of spines is another defect as well. The plect- ambonitids of the Silurian possess little similarity to this unique shell except for the strophic outline and concavoconvex profile. The orthotetacids have a com- pletely different layout of their internal structures PLATE 1: figures 1-10.—Aenigmastrophia cooperi, new species. 1-6, Holotype, a brachial valve, USNM 160179 (USNM loc. 13257): 1, dorsal view; 2, internal view; 3, anterior view; 4, posterior view; 5, side view; 6, side view of valve tilted to show cardinal process lobe. 7-10, Paratype, a pedicle valve, USNM 160180: 7, dorsal view; 8, internal view; 9, side view; 10, posterior view. (All views X 4.) Figures 11-15.—Aenigmastrophia greggi, new species. 11, 13, 15, Holotype, a brachial valve, USNM 160182 (USNM loc. 11162) : 11, rubber replica of interior; 13, impression of interior; 15, impression of exterior. 12, 14, Paratype, a pedicle valve, USNM 160183: 12, a rubber replica of interior; 14, impression of interior. (All views natural size.) NUMBER 3 157 PLATE 1 158 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY which, combined with a prominent radial ornamenta- tion in all forms, removes them from effective con- sideration. All-in-all, there appears no recourse at the present time but to confess ignorance of this puz- zling taxon's familial affinities. The extreme rarity of the material does not assist the matter. The small specimens from the Roberts Mountains Formation are so similar to the rather large specimens from the Kla- math Mountains that future ontogenetic studies on better collections probably will not afford many clues about familial affinity. In summary, the best that can be done is to suspect that the new genus belongs to the general group of pseudopunctate, flat-to-concavocon- vex strophic shells so widespread in the Middle Paleozoic. Aenigmastrophia cooperi, new species PLATE 1: FIGURES 1-10 DESCRIPTION.—The exterior of both valves is rela- tively smooth except for the presence of weak, con- centric growth lines. The shells are laterally elongate in form, with the length being about one-half the width. The position of maximum width coincides with the straight hinge-line. The brachial valve is gently concave and the pedicle valve is gently convex. The interarea of the brachial valve is relatively short and orthocline; that of the pedicle valve about twice as long and apsacline. The delthyrium of the pedicle valve includes an angle of about 90° and does not appear to be modified by deltidial structures, although, in view of the small nature of the sample, this con- clusion may be premature. The notothyrial region in- cludes an angle of more than 120° and is unmodified by chilidial structures. The teeth of the pedicle valve and the cardinal process lobes of the brachial valve both project to the exterior margin along the hinge- line. The plane of commissure appears to be planar. The anterior margins of both valves are evenly rounded, as are the lateral margins. The interior of the brachial valve is relatively smooth except for minute pustules, which may reflect the presence of pseudopunctae. The cardinalia consist of a pair of laterally directed cardinal process lobes origi- nating almost at the apex of the notothyrial region and expanding in size laterally. They include an angle of about 140° anteriorly. Narrow, posteriorly directed hinge-sockets are included in the region between the interarea and the cardinal process lobes. These sockets are very narrow and expand distally hardly at all. In small specimens a rounded median septum extends from the base of the notothyrial cavity to about the midlength, but in large specimens this septum is re- stricted to the posterior fraction of the valve and is a relatively insignificant structure. The anterior edge of the interarea is raised up over the interior of the valve as a low ridge, which gradually weakens in a lateral direction. The interior of the pedicle valve is quite undis- tinguished and apparently smooth. The impress of a muscle field has not been discerned. In the delthyrial region, a pedicle callist is raised up off the floor of the valve. The hinge teeth consist of a pair of anterolater- ally directed ridges, including an angle of about 90°. The interarea extends somewhat anteriorly over the posterior portion of the interior as a narrow, low plat- form, but the poorly preserved nature of the material precludes any comment about its distal relationships. HOLOTYPE.—USNM 160179, from USNM loc. 13257, Roberts Mountains Formation, 672 feet above the base of the Formation, along a line of section east- northeast from a point 250 feet south and 3,699 feet west of hill 9219, west flank of Roberts Creek Moun- tain, Eureka County, Nevada. MATERIAL.—Figured specimens: USNM 160179 (holotype), 160180. Unfigured specimen: USNM 160181, from E37-5, USNM loc. 13237, Roberts Mountains Formation, Birch Creek area, elevation 6,520 feet, 550 feet south and 4,450 feet west of south- east corner of Sec. 22, T 24 N, R 50 E, northern Rob- erts Mountains, Eureka County, Nevada; collected by M. A. Murphy in 1966. REMARKS.—Aenigmastrophia cooperi, new species, is distinguished by its small size and relatively strong median septum in the brachial valve. Aenigmastrophia greggi, new species PLATE 1: FIGURES 11-15 HOLOTYPE.—USNM 160182, from USNM loc. 11162, unnamed Ludlow age sandstone, in middle of section 10, about one-half mile west of Mountain House Valley of East Fork of Scott River, China Mountain quadrangle, Siskiyou County, California. MATERIAL.—The holotype and the paratype, USNM 160183. REMARKS.—Aenigmastrophia greggi, new species, is similar to A. cooperi, new species, except for its large size and relatively weak median septum in the brachial valve. DEVONIAN Jean M. Berdan Some Ostracode s from the Schoharie Formation (Lower Devonian) of New York ABSTRACT Ostracodes are reported from the Schoharie Forma- tion for the first time. Descriptions and illustrations are given of three new genera and species—Schohariella grandis, Parabingeria cooperi, and Vietor josephinae— and of a new species of Phanassymetria Roth, P. col- lilupana. Tubulibairdia punctulata (Ulrich) is rede- scribed and figured, and related species are discussed. Three forms left in open taxonomy also are described and illustrated. None of these ostracodes occurs in the Camden Chert, which is correlated with the Schoharie on the basis of megafossils. Two of them, however, are similar to ostracodes described by Jones (1890) from drift boulders of the "Corniferous Limestone" near Canandaigua, New York. This similarity raises the possibility that Jones's material may have been de- rived from the Bois Blanc Formation (the Schoharie equivalent in western New York) rather than the Onondaga Limestone as hitherto supposed. Hitherto, ostracodes have not been reported from the Lower Devonian Schoharie Formation in New York, and it therefore seems desirable to describe these forms, even though only one locality has been sampled and few species are present. These descriptions are par- ticularly appropriate in a volume dedicated to Dr. G. Arthur Cooper because of his outstanding contribu- tions to the study of the Devonian of New York. Jean M. Berdan, United States Geological Survey, Room E—303, United States National Museum, Washington, D.C. 20242. Publication authorized by director, United States Geological Survey. The name Schoharie was first used by Vanuxem (1840, p. 378) for the "Schoharie layers," which were described as being rich in fossils. Later, Vanuxem (1842, p. 131) used the term "Schoharie grit," with the subheading "Shell grit." Mather (1843, p. 340), reporting on the geology of the type area, also called the formation the "Schoharie grit" and described it as being "a fine-grained calcareous grit rock, containing a great number of fossils .... The carbonate of lime gradually disappears from the rock where it is exposed to the weather, and the remaining mass is a rather porous and spongy fine-grained tough sandstone." The formation as originally conceived was thin and poorly exposed. Mather (1843, p. 340) stated that he had seen the formation in place at only two localities, one on "the mountain one or one and a half mile west and northwest of Schoharie," in Schoharie County, and the other on "the mountain one-half to one mile west and northwest of Clarksville," Albany County. Gra- bau (1906, pp. 180-181) indicated only two expo- sures in the Schoharie Valley, and he estimated the thickness of the Schoharie Grit in this area as being five or six feet. Ruedemann (1930, pp. 60-62), on the basis of nine outcrops, found the maximum thick- ness in the Albany area to be between six and eight feet. Goldring and Flower (1942) restudied the stratig- raphy of the Schoharie and Esopus Formations and named a new formation—underlying the Schoharie and overlying the Esopus—the Sharon Springs Forma- tion, later renamed the Carlisle Center Formation (Goldring and Flower, 1944, p. 340). Johnsen and Southard (1962, p. A8) considered the Carlisle Cen- 161 162 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY ter to be a member of the Schoharie and renamed the Schoharie Grit of previous authors the Rickard Hill Member of the Schoharie Formation. As understood by Johnsen and Southard (1962, p. A7), the Schoharie Formation is a complex of lithologic facies which ex- tends from Herkimer County, New York, to Monroe County, Pennsylvania, and which reaches a thickness of more than 200 feet. Their Rickard Hill Member, which is the most fossiliferous part of the formation, extends from East Springfield, New York, to southern Albany County, New York. The ostracodes described in this paper are all from their Rickard Hill Member of the Schoharie (or the Schoharie Grit of former usage). The nature of the contact of the Schoharie Forma- tion with the overlying Onondaga Limestone has been reviewed by Oliver (1967, pp. Al, A2), who concludes that there is a sharp faunal break between the two formations. The lower contact of the Carlisle Center Member of the Schoharie as used by Johnsen and Southard (1962) with the underlying Esopus Forma- tion is gradational. The contact is placed by Johnsen and Southard (1962, p. A8) at the lowest beds that are sufficiently calcareous to effervesce in cold dilute hydrochloric acid. Cooper (1942) and Boucot and Johnson (1968, pp. B1-B4) have discussed the correlation of the Schoharie with other North American formations. Boucot and Johnson (1968, pp. Bl, B3) consider the Schoharie to be early Emsian in age, in terms of the standard se- quence in the Rhineland. Most major groups of megafossils from the Scho- harie Grit were described by Hall (1867, 1879, 1884, 1885; in Hall and Simpson, 1887; in Hall and Clarke, 1888) in his series of monographs on the paleontology of New York. Some of the groups have undergone considerable revision by later paleontologists, but as yet no microfossils, except conodonts (Klapper and Ziegler, 1967), have been described from the Schoharie Formation in New York. Generally, however, the ostracodes described from the Camden Chert of west- ern Tennessee—which is correlated with the Schoharie on the basis of brachiopods by Cooper and others (1942, pp. 1748, 1779-1780) and by Boucot and John- son (1968, p. B3)—have been considered representa- tive of the ostracodes of Schoharie age. Thirty-eight species of ostracodes were described from the Camden Chert by Bassler (1941) and Swain (1953). The Camden assemblage is dominated by bolliids (six species or subspecies of Bollia) and thlipsu- rids (five species of Strepulites, three species each of Thlipsurella and Thlipsurina, and one species each of Stibus and Thlipsura) but is lacking in beyrichiaceans. In generic composition the Camden assemblage re- sembles the ostracodes described from the Onondaga Formation of Pennsylvania by Swartz and Swain (1941), as suggested by Bassler (1941, p. 22) and Swain (1953, pp. 259-261), although the assemblage in Pennsylvania includes one beyrichiacean. In con- trast, ostracodes from the Schoharie include two large beyrichiaceans, and bolliids and thlipsurids are scarce or lacking. In fact, the Schoharie assemblage described below has no species in common with the Camden assemblage. The differences between the Camden and Schoharie ostracode assemblages may be due to differences in age, facies, or faunal provinces, or perhaps to all three. With regard to difference in age, Bassler (1941, p. 22) implied that the ostracodes he" described came from near the base of the Camden, but Cooper (in Swain, 1953, p. 258) has noted: "No definite horizon can be stated for the specimens because they came from pure white clay found on the floor of the quarry. They evidently came from somewhere in the quarry wall, but I never located any of the material in place." In view of the uncertainty as to the exact stratigraphic position of the ostracode fauna within the Camden, and the possibility suggested by Swain (1953, p. 259) that the assemblage is mixed, it is conceivable that the Camden ostracodes are not precisely contemporaneous with those of the Schoharie and may be slightly younger. With regard to difference in facies, although Boucot and Johnson (1968, fig. 2) include both formations in a limestone and chert belt on their generalized litho- facies map, the ostracodes from the Camden appar- ently came from a very fine-grained sediment indica- tive of a low-energy environment whereas the Scho- harie beds from which the ostracodes are described are primarily a sand composed of quartz and calcite grains with calcareous cement, which suggests a high- energy environment. With regard to difference in faunal province, the described ostracode faunas from the Lower Paleozoic of the Mississippi Valley area appear to differ from ostracode faunas of those formations in the Appa- lachian area that are dated as contemporaneous on the basis of other groups of fossils. Whether this faunal difference is due to lithologic facies or to some less tangible factor such as temperature or depth is yet to NUMBER 3 163 be determined. Benton County, Tennessee, where the Camden ostracodes were collected, is about 900 miles from Albany County, New York, and some differences might be expected. The possibility exists that the difference between the Camden and Schoharie ostracode assemblages may be more apparent than real. Future collecting from other parts of the Schoharie Formation may show more ele- ments in common than are now known. Although no ostracodes have as yet been described from the Schoharie Formation, Jones (1890) described six species of ostracodes from die "Chert of the Cornif- erous Limestone" in Ontario County, New York. These ostracodes have been considered to come from the Onondaga Limestone (Bassler and Kellett, 1934, p. 75; Warthin, 1937) and, if so, are the only described ostracodes from the Onondaga of New York. The specimens, which are deposited in the New York State Museum, are molds and casts preserved in yellow buff- weathering rotted chert, and the accompanying labels indicate that they come from "Onondaga Ls., (drift), Canandaigua, Ontario County, NY." In 1964 I tried to find this assemblage in place in the Onondaga in the vicinity of Canandaigua, but without success. How- ever, two of the species described by Jones, "Eury- chilina reticulata'' and "Moorea kirkbyi," resemble species from the Schoharie described in this paper. In addition, Oliver (1967, p. A7, fig. 2) has indicated that an outcrop of the Bois Blanc Formation, which is the equivalent of the Schoharie in western New York, occurs in the Phelps 7/2-minute quadrangle, north and east of Canandaigua. Because the ostracodes described by Jones came from a boulder of glacial drift, there is a possibility that they actually may be from the Bois Blanc Formation rather than the Onondaga. As yet no ostracodes are known from the Bois Blanc, but the ostracodes that I have collected from the Onondaga do not include Jones's species. Further study of the assem- blages of both formations is necessary before any firm conclusions can be drawn. The original sample of Schoharie from which ostra- codes were studied is a part of a collection made by G. A. Cooper for megafossils. The label with the ostra- code sample reads "Schoharie, 134 mi. NNW of Clarks- ville, New York, G A. Cooper, 1938, loc. 119f." This locality was described by Goldring and Flower (1942, p. 679) as follows: "Above New Salem along the Wolf Hill highway, one-quarter of a mile east of the New Scotland town quarry in the Onondaga limestone, the Schoharie is exposed in the woods above a road cut in 372-386 O—71 12 the Esopus and Sharon Springs formations. Here only the upper part of the Schoharie grit is present, meas- uring two feet six inches, exposing the pelops bed, the upper layers and the Onondaga contact." This locality was examined specifically for ostracodes in 1964 and 1968, when two additional collections for the United States Geological Survey (USGS) were made—USGS 7267-SD and USGS 8211-SD. The first of these (USGS 7267-SD) was taken from loose blocks apparently quarried by some previous paleontologist; the second (USGS 8211-SD) was collected in place about ten inches below the contact of the Schoharie with the Onondaga. The locality is overgrown, but about two feet of Schoharie is still exposed. The contact with the Onondaga is clear and appears to be some- what irregular; the Schoharie just beneath the contact is a brown-weathering leached sandstone packed with brachiopods and other fossils. This bed is six or seven inches thick, and is underlain by ten inches to a foot of very hard, steel-gray, sandy limestone with fewer fossils; it was from this bed that the ostracodes in USGS 8211-SD were obtained. The locality is on the Clarks- ville 7/2 -minute topographic quadrangle, on the west side of Wolf Hill Road (New York Route 85), in the woods above the road, 1,500 feet south of benchmark 781 and 900 feet north of the point where a gas pipe- line crosses the road. The ostracodes described in this paper are few in both number of species and number of individuals. This is due in part to the siliceous character of the Schoharie matrix, which renders both collection and preparation of the material extremely difficult. The original sample containing ostracodes from Dr. Cooper's collection is small; it consists of pieces of rock which just fill a 3-inch by 4-inch tray. The two addi- tional collections made by me specifically for ostra- codes are larger but together do not represent more than half a standard 3-inch-deep drawer, or about 924 cubic inches of rock. Ostracodes are not abundant in the matrix, and one collection in particular, USGS 7267-SD, yielded very few specimens for the amount of rock processed. The matrix consists of sand grains, of both quartz and calcite, in a calcareous cement and includes aggregates of crystalline pyrite. The most abundant groups of fossils associated with the ostra- codes are trilobites, brachiopods, and corals, which are randomly oriented in the rock and are in many in- stances fragmentary, especially the trilobites. The ostracode specimens are calcareous and were prepared by the use of a vibratool and needle on broken 164 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY pieces of rock. In some instances preparation is facili- tated by heating the rock and quenching it in cold water; this technique was used on some of the mate- rial, but as it tended to make the shells spall off, some of the material was prepared without heating and quenching. Because of the character of the matrix it is diffi- cult to obtain unbroken specimens. In addition to the species described below, fragmentary ostracodes that are too poorly preserved to illustrate have been found. These fragments indicate the presence in the fauna of a beyrichiopsid; a bolliid; one or more species of hollinids, possibly including Parabolbina; a bairdio- cyprid; a condracyprid; and a form resembling Baschkirina Rozhdestvenskaja, 1959. It is obvious that the ostracodes described in this paper represent only a small part of the total assemblage in the Schoharie Formation. The classification used in this paper is based on the Treatise on Invertebrate Paleontology, Part Q, Ostra- coda (Moore, 1961) except for the Beyrichiidae, which is taken from Martinsson (1962, 1963). The type specimens are in the United States Na- tional Museum (USNM). All photographs were taken by Robert H. McKin- ney and printed by Haruo E. Mochizuki, both of the United States Geological Survey. Order PALAEOCOPA Henningsmoen, 1953 Superfamily BEYRICHIACEA Matthew, 1886 Family CRASPEDOBOLBINIDAE Martinsson, 1962 PSubfamily CRASPEDOBOLBININAE Martinsson, 1962 Schohariella, new genus DIAGNOSIS.—Unisulcate to nonsulcate craspedobol- bininids with broad, complete, tubulous velum, drop- shaped external muscle spot; crumina ventral, bluntly acuminate in lateral outline, protrudes markedly above velum. Dolonoid scar obscure, ventral in posi- tion. No torus. TYPE-SPECIES.—Schohariella grandis, new species. DISCUSSION.—This new genus is questionably placed in the Craspedobolbininae because of an obscure groove interpreted to be the trace of a dolonoid scar at the most ventral part of the crumina. In addition, it resembles die craspedobolbinine genera Apatobol- bina and Leptobolbina in being essentially nonsulcate and having a prominent external muscle spot. Schohariella may be distinguished from these genera by the form of its crumina and by its broad, complete, tubulous velum. Hyrsinobolbina Martinsson, 1962, has a broad complete velum but is decidedly unisulcate and has a globular crumina. The genus Schohariella differs from typical treposelline genera in its broad tubulous velum and lack of the treposelline "bridge" at the contact of the crumina with the velum. It differs from described amphitoxotidine genera in essential lack of sulcation and in the character of the subcruminal field. No tecnomorphs of Schohariella grandis have been found in the collections of the Schoharie Formation from Wolf Hill. The tecnomorphic specimen described by Jones (1890, p. 535, pi. 20, figs. 13a, 13b) as "Eurychilina reticulata Ulrich"—later renamed Treposella reticulosa by Warthin (1937, card 20) — appears, however, to belong to Schohariella. The material figured by Jones consists of an internal mold and external cast of the same tecnomorphic specimen, and is deposited in the New York State Museum (NYSM 4452). A latex cast of the original of Jones' specimen (Jones, 1890, pi. 20, fig. 13b) is illustrated herein as figure 24 of Plate 1, which shows the broad velum and drop-shaped muscle spot of Schohariella. The surface reticulation is coarser and the muscle spot is less well defined, however, than in Schohariella grandis, and it seems probable that the two specimens are not conspecific. As previously noted, Schohariella reticulosa was reported to have come from the Onon- daga Limestone, but there is a possibility that it came from float from the Bois Blanc Formation, which is the Schoharie equivalent in western New York. The name Schohariella is derived from a latinized diminutive of Schoharie, the formation in which the type-species occurs. Schohariella grandis, new species PLATE 1: FIGURES 25, 26 DESCRIPTION.—The lateral outline is semielliptical and amplete. A low, indistinct median lobe is bounded posteriorly by a very shallow median sulcus occupied by a slightly raised drop-shaped muscle spot. The pos- terior lobe is very slightly cuspate and is swollen in NUMBER 3 165 profile above the median and anterior lobes. A narrow, shallow fissus extends about two-thirds of the length of the valve beneath the muscle spot and the median lobe. The surface of the domicilium is finely punctate, except for the muscle spot. The velum is broad, tubu- lous, and radially striate with superimposed, very fine, concentric filae. It extends from the posterior cardinal angle around the free margin of die valve and ap- parently to the anterior cardinal angle, although as the anterior angle is broken it is difficult to be certain of this. The crumina is protuberant, bluntly triangular with broad base against the domicilium, and very faintly striatopunctate. In end view the crumina pro- trudes laterally as a linguiform extension at nearly a 45° angle to the plane of the valve. In ventral view the velum meets the crumina at about half its height but does not cross it. There is a probable dolonoid scar at the ventral edge of the crumina. Holotype measurements (in millimeters) : length in- cluding the velum, 3.40, not including the velum, 2.25; height including the crumina, 1.9, to the base of the crumina, 1.2. MATERIAL.—The holotype only. HOLOTYPE.—USNM 163663. DISCUSSION.—Although only one specimen of this species has been found, it seems to be sufficiently dis- tinctive to warrant description. It differs from Scho- hariella reticulosa (Warthin) not only in having the finer, less distinct reticulation and the more distinct muscle spot mentioned above but also in having a more conspicuous median lobe and fissus—the latter being almost lacking in S. reticulosa. Some of these differences might be due to the difference in the type of preservation, or might be sexual, as the only speci- men of S. grandis is a heteromorph and the only specimen of S. reticulosa is a tecnomorph. It seems unlikely, however, that all the differences are due to either of these factors, and therefore the two species are considered distinct. The specific name is indicative of the large size of the holotype. Family BEYRICHIIDAE Matthew, 1886 Subfamily BEYRICHIINAE Matthew, 1886 Parabingeria, new genus DIAGNOSIS.—Unisulcate to weakly bisulcate beyri- chiid ostracodes; anterior and posterior lobes with cuspidal crest; median lobe set below hinge-line; dis- tinct velar ridge; nearly horizontal zygal ridge extend- ing onto posterior lobe, bounded ventrally by weak fissus. Crumina anteroventral, not distinctly set off from domicilium, inflating and obliterating velar ridge. TYPE-SPECIES.—Parabingeria cooperi, new species. DISCUSSION.—This genus closely resembles Bingeria Martinsson, 1962, from the Tofta Beds (Middle Silurian) of Gotland. However, Bingeria lacks the horizontal zygal ridge and subjacent fissus extending onto the posterior lobe and also has a more distinctly delimited crumina. Arikloedenia Adamczak, 1968, from the Middle Devonian of Poland is more distinctly bisulcate and lacks the distinct velar ridge, fissus, and zygal ridge of Parabingeria. The zygal ridge and fissus suggest some species of Beyrichia (Simplicibeyrichia) Martinsson, 1962, but in that subgenus the crumina is more distinctly developed and spines or pustules are present on the valves. The name is derived from the Greek prefix para (near) and from Bingeria, indicating the similarity of this genus to Bingeria. Parabingeria cooperi, new species PLATE 1: FIGURES 27-30 DESCRIPTION.—The lateral outline is amplete and subelliptical. The anterior and posterior margins are smoothly curved; the ventral margin is nearly straight on tecnomorphs; the dorsal margin is slightly sinuate to straight; and the dorsal and ventral margins are subparallel. The anterior and median lobes are fused; the anterior lobe has a low cuspidal crest. The pos- terior lobe has a distinct cuspidal crest at the posterior angle. The median sulcus is distinct, curved anteriorly, and extends about two-thirds of the distance from the hinge to the ventral margin. A zygal ridge extends posteriorly from the median lobe, beneath the median sulcus, to the middle of the posterior lobe. A shallow fissus occurs beneath the zygal ridge. The velar ridge is distinctly separated from the domicilium, reflexed away from the contact margin of the valves; it is con- tinuous around the free margin on tecnomorphs, and it is occupied by crumina and is obliterated antero- ventrally on heteromorphs. The crumina is an indis- tinct swelling of the domicilium, anteroventral in position. The surface of the valves is smooth except for widely spaced, shallow punctae which are more numerous and deeper on the crumina. 166 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY The holotype, a heteromorphic right valve, is 3.3 mm in length and 2.2 mm in height. Two paratypes, a right and a left tecnomorphic valve, are 3.3 mm in length and 2.1 mm in height and 2.8 mm in length and 1.7 mm in height, respectively. MATERIAL.—One complete right and one broken left heteromorphic valve, one complete and two broken right tecnomorphic valves, and one complete left tec- nomorphic valve, all from G. A. Cooper collection 119f; also, two broken specimens in collection USGS 8211-SD. TYPES.—Holotype, USNM 163660; paratypes, USNM 163661—2. DISCUSSION.—Parabingeria cooperi appears super- ficially similar to other large subquadrate beyrichia- ceans from the Lower Devonian of North America, such as Kloedenia? newbrunswickensis Copeland, 1962, and "Beyrichia'' occidentalis Walcott, 1884, from the Dalhousie beds and the Nevada Limestone, respec- tively. However, P. cooperi differs from both these species in being essentially unisulcate and in having a conspicuous velar ridge. The zygal ridge and fissus also are distinctive. Heteromorphs of the other taxa are not as yet known. The beyrichiacean described by Jones (1890, p. 538, pi. 21, fig. 1) as "Beyrichia kloedeni M'Coy var." is a heteromorphic right valve with a prominent median lobe and a conspicuous crumina sharply delimited from the domicilium and thus is distinctly different from Parabingeria cooperi. Swartz and Swain (1941, p. 428) suggested that Jones's specimen might be a fe- male of their Kloedenia rectangularis, which species also differs from Parabingeria in having a more promi- nent median lobe. The specific name is given in honor of Dr. G. Arthur Cooper, who collected the specimens described. Suborder KLOEDENELLICOPINA Scott, 1961 Superfamily PARAPARCHITACEA Scott, 1959 Family PARAPARCHITIDAE Scott, 1959 Genus Neoaparchites Boucek, 1936 Neoaparchites sp. aff. N. mesleri (Bassler, 1941) PLATE 1: FIGURES 11-13 DESCRIPTION.—The lateral outline is elliptical and preplete. The hinge is less than half the length of the shell and very slightly incised. The greatest thickness is in the posterior third and dorsal half of the shell. The right valve slightly overlaps the left around the free margin. The shell surface is smooth. A carapace measures 1.4 mm in length, 1.1 mm in height, and 0.7 mm in width. DISCUSSION.—One broken carapace (USNM 163664), found in collection USGS 8211-SD, appears to belong in Neoaparchites as diagnosed by Boucek (1936, p. 39), and Krandievs'ky (1963, pp. 17-18). This specimen has been compared to "Paraparchites" mesleri Bassler but it differs from that species in being preplete instead of amplete in lateral outline. "Para- parchites" mesleri is represented only by single valves, so the overlap is not clear, but it probably should also be assigned to Neoaparchites. Similar species are Apar- chites mesleri] or mis Polenova, 1960, Aparchites rozh- destvenskajae Polenova, 1968, and Antiparaparchites primaevus Kesling, 1958; however, judging from the illustrations of those species, especially the last, they may belong in Pseudoaparchites Krandievs'ky, 1963, because of the character of the overlap. Order PODOCOPIDA Muller, 1894 Suborder PODOCOPINA Sars, 1866 Superfamily BAIRDIACEA Sars, 1888 Family BAIRDIIDAE Sars, 1888 Genus Bairdiacypris Bradfield, 1935 Bairdiacypris? sp. PLATE 1 : FIGURES 20-23. DESCRIPTION.—The lateral outline is reniform. The dorsal margin is curved; the anterior margin is evenly but sharply curved; the ventral margin is concave; and the posterior margin is bluntly acuminate. The maximum length is below midheight; the maximum width is in the posterior half; and the maximum height is in the anterior half of the carapace. The left valve overlaps the right around the free margins and over- reaches the right along the dorsal margin. The hinge is short, straight, and slightly incised. The shell surface is smooth. The muscle scar is not known. The figured specimen measures 0.85 mm in length and 0.40 mm in height and width. NUMBER 3 167 MATERIAL.—Only the figured specimen (USNM 163665), a complete carapace. DISCUSSION.—The single specimen is tentatively as- signed to Bairdiacypris because of the character of the overlap and outline. Although it appears similar to some specimens of Camdenidea Swain, 1953, the pos- terior is not as acuminate and the ends are not com- pressed as they are in that genus. It differs from specimens of Silus Polenova, 1968, in having a more acuminate posterior and in having the maximum height anterior. Suborder METACOPINA Sylvester-Bradley, 1961 Superfamily HEALDEACEA Harlton, 1933 Family PACHYDOMELLIDAE Berdan and Sohn, 1961 Vietor, new genus DIAGNOSIS.—Pachydomellid ostracodes with flat- tened venter and conspicuous groove near and parallel to ventral margin on the right valve; left valve without groove. TYPE-SPECIES.—Vietor josephinae, new species. DISCUSSION.—This new genus resembles other pachydomellid genera in being distinctly inequivalved and asymmetrical, with left over right overlap, a short incised hinge, and a thick shell that has conspicuous tubules which do not open to the exterior of the shell but may be seen when the shell is wetted. However, other genera of this family that develop longitudinal grooves have them situated close and parallel to the hinge-line instead of near the ventral margin as in the genus described. The generic name is taken from the Latin vietor, one who makes baskets of osier to be covered with leather— the Roman equivalent of one who makes barrels, a cooper. Vietor josephinae, new species PLATE 1: FIGURES 1-5 DESCRIPTION.—The lateral outline is subovoid, and the ventral and dorsal outlines are irregularly elliptical. The left valve is larger than the right, overlapping it around the free margins but not overreaching it along the hinge. The left valve has a slight anterodorsal sag or depression, and the ventral surface is flattened so that it makes an acute angle with the lateral slope of the valve. The right valve has a more pronounced an- terodorsal sag, posterior to which is a posterodorsal bulge which is bounded ventrally by a cleft or groove parallel to and just above the ventral angulation. This groove extends from near the posterior margin to about one-third of the length from the anterior margin. The ventral surface of the right valve is flattened and angu- lated like that of the left valve. The shell surface is smooth. Tubules may be seen on wetted specimens, especially on the ventral surfaces, where they appear to be more numerous and larger than those on the rest of the shell. Measurements, in millimeters, of holotype (from coll. USGS 8211-SD): length 1.35, height 0.75, width 0.85. Two paratypes from the same collection measure, respectively, 1.4 and 1.35 mm in length; both measure 0.75 mm in height and 0.85 mm in width. MATERIAL.—Three carapaces (including the holo- type) from collection USGS 8211-SD and one cara- pace from collection USGS 7267-SD. TYPES.—Holotype, USNM 163666; paratypes, USNM 163667-9. DISCUSSION.—This species is somewhat more abun- dant than would appear from the number of specimens listed above, for several additional specimens were seen in the matrix but broke in the course of preparation. The species is named in honor of Mrs. G. A. Cooper. Genus Phanassymetria Roth, 1929 Phanassymetria collilupana, new species PLATE 1: FIGURES 6-10 DESCRIPTION.—The lateral outline is subtrape- zoidal. The anterior margin is bluntly rounded; the ventral and dorsal margins are nearly straight and sub- parallel; and the posterior margin is rounded to ob- tusely angulate. The left valve overlaps the right around the free margins and has an indistinct longi- tudinal groove parallel to and just below the dorsal margin, which defines a sharp, angular ridge along the dorsum. The right valve has a similar groove re- stricted to the posterior half of the valve. The ventral surface of both valves is flattened. There is a low longitudinal ridge on the posterior half of the right valve at the junction of the ventral and lateral surfaces. 168 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY The ventral and dorsal outlines are subpyriform. In dorsal view the two dorsal ridges enclose a drop-shaped depression, which is open posteriorly and in which the hinge is incised. The width is greater than the height. The shell surface is smooth. Tubules are visible on wetted specimens and appear to be evenly distrib- uted in both valves. The holotype measures 1.4 mm in length, 0.8 mm in height, and 1.1 mm in width. MATERIAL.—One carapace (the holotype), one right valve, and one left valve, all from collection USGS 8211-SD. TYPES.—Holotype, USNM 163670; paratypes, USNM 163670a, b. DISCUSSION.—This species appears to be inter- mediate between Phanassymetria Roth, 1929, and Pachydomella Ulrich, 1891, in some respects. It lacks the angular cross section of Phanassymetria triserrata Roth, the type species of the genus, but differs from Pachydomella tumida Ulrich in having a dorsal ridge on the right valve. It differs from the types of both these genera in having dorsal grooves on both valves instead of on the left valve only. Sohn (1961, p. 76) considered the most important criterion for distinguish- ing Phanassymetria from Pachydomella to be a longitu- dinal ridge below midheight on one or both valves of Phanassymetria, this feature not being present in Pachydomella. Accordingly, this new species is assigned to Phanassymetria. This is the first species of the genus reported outside the Midcontinent region. The specific name is based on the Latin collis (hill) and lupus (wolf) with the suffix ana (denoting place) to indicate that the species is found near Wolf Hill Road. Genus Tubulibairdia Swartz, 1936 Tubulibairdia punctulata (Ulrich, 1891) PLATE 1, FrcuRES 14-18 Bythocypris punctulata Ulrich, 1891, p. 196, pi. 17, figs. 2a-c. Microcheilinella punctulata (Ulrich).—Bassler and Kel- lett, 1934, p. 412. Tubulibairdia punctulata (Ulrich).—Sohn, 1961, p. 75, pi. 5, figs. 7, 10, 11, 14-17. DESCRIPTION.—The lateral outline is subovate; the dorsal outline is ovate; and the anterior outline is in- dented subcircular. The greatest height and the great- est width are posterior to the midpoint of the cara- pace, and the greatest length is below the midpoint. The left valve overlaps the right around the free mar- gins, especially on the ventral margin where a flap, or lappet, protrudes over the right valve in the anterior half of the carapace so that the ventral commissure is sinuate. The left valve is ovate in lateral outline and overreaches the right valve dorsally. The right valve is smaller than the left and subtrapezoidal in lateral outline. The hinge is straight, half or slightly more than half the greatest length of the carapace, and deeply incised. The shell surface is smooth or has ex- tremely fine longitudinal striations. Tubules that are visible on lightly calcined specimens appear to be evenly distributed through the shell except for the muscle scar and overlapping parts of the ventral mar- gin. The muscle scar is circular and located at or slightly below midheight on the left valve and at mid- height on the right valve. It is composed of radiating muscle flecks. Measurements (in millimeters) of the two fig- ured specimens: USNM 163671 (Plate 1: figures 14- 17), length 1.40; height 0.80, width 0.95; USNM 163672 (Plate 1: figure 18), length 1.60, height 0.95, width 1.10. The measurements of six other complete carapaces are as follows: USNM Specimen Length Height Width 163673 1.75 1. 10 1. 15 163674 1.70 1.05 1. 15 163675 1.65 1.00 1. 10 163676 1.60 1.00 1. 10 163677 1.60 1.00 1. 10 163678 1.45 0.85 0.90 MATERIAL.—Eight complete carapaces (USNM 163671-78), nine broken carapaces (USNM 163678a), and six separate valves (USNM 163678b). DISCUSSION.—Although the genus is easily recog- nizable, species of Tubulibairdia are difficult to dis- criminate because they have few reliable distinctive characteristics. The most obvious difference between many of them is size, but size alone is hardly a reliable specific character in ostracodes. The specimens from the Schoharie are here assigned to T. punctulata (Ul- rich) because, although they average slightly larger in size than typical T. punctulata, they seem to agree with that species in other respects. With regard to the size of the species, it should be noted that Ulrich (1891, p. 196) gave the dimensions of T. punctulata as 1.93 mm in length, 1.15 mm in height, and 1.14 mm in thickness. However, none of the specimens in the NUMBER 3 169 type lot (USNM 41823) attains this size. The holo- type, figured by Sohn (1961, pi. 5, figs. 14-17), is 1.25 mm long, 0.75 mm high, and 0.85 mm wide; and the paratype figured by Sohn (1961, pi. 5, figs. 10, 11) is 1.50 mm long, 0.90 mm high, and 0.95 mm wide; a topotype from the Falls of the Ohio is 1.30 mm long, 0.80 mm high, and 0.85 mm wide. It must be assumed that either the specimen measured by Ulrich was not included in the type lot or that the measurements were in error. In addition to T. punctulata, nine other taxa of Tubulibairdia have been described from North Amer- ica. T. tubulifera Swartz, 1936, the type-species of the genus, was based on molds and casts, and not all of its dimensions can be determined. The holotype of T. tubulifera (USNM 94195), from the Shriver Chert of Curtin, Pennsylvania, is represented by an internal and an external mold of a left valve which measures 1.30 mm in length and 0.85 mm in height. According to Swartz (1936, p. 581), an average left valve meas- ures 1.40 mm in length and 0.98 mm in height, and an average right valve measures 1.40 mm in length and 0.88 mm in height. These dimensions are compara- ble to those of T. punctulata, but a latex squeeze of the external impression of the holotype of T. tubulifera shows a high dorsal surface—which indicates tiiat the valve was more umbonate than that of T. punc- tulata—and a weak longitudinal groove or sag parallel to and just below the dorsum. This weak sag is not comparable to the well-developed subdorsal groove of Pachydomella tumida Ulrich, 1891, which is the characteristic feature of the genus Pachydomella Ulrich, 1891. Tubulibairdia simplex (Roth, 1929), as described by Lundin (1968, pp. 70-72), may also have a slight subdorsal groove like that of T. tubulifera. The fol- lowing measurements (in millimeters) of the holotype (USNM 80646) of T. simplex are given by Lundin (1968, p. 71) : length, 1.45; height, 0.88; and width, 0.80. Lundin's (1968, fig. 32) size-dispersion diagram shows, however, that T. simplex may attain consider- ably larger dimensions. That species differs from the holotype of T. tubulifera. in being more acuminate posteriorly. Two species, Tubulibairdia paucitubulis Swartz and Swain, 1941, and T. multitubulis Swartz and Swain, 1941, from the Onondaga beds of Pennsylvania were described only from internal molds in shale. Swain (1953, p. 282) considered the smaller of these, T. paucitubulis (length, 0.82 mm; height, 0.45 mm), to be juvenile specimens of T. multitubulis, which is based on one broken internal mold that is 1.45 mm in length and 0.78 mm in height (Swartz and Swain, 1941, p. 446). T. multitubulis is unrecognizable by present criteria and will remain so until topotypic material is found and studied. Swain (1953, p. 280) considered Tubulibairdia to be a junior synonym of Pachydomella Ulrich, 1891, and assigned specimens from the Camden Chert to Pachydomella multitubulis. The Camden specimens lack the prominent subdorsal groove of true Pachydomella (compare P. dorsoclefta Swain, 1953, also from the Camden) and are here considered to belong to Tubulibairdia. These Camden specimens, which are more acuminate posteriorly than most other species of Tubulibairdia, may represent a new species. Of the remaining five taxa assigned to Tubulibairdia in North America, T. longula (Ulrich and Bassler), 1913, from the Keyser Limestone in Maryland and Tubulibairdia sp. cf. T. longula (Ulrich and Bassler) described by Lundin (1965, pp. 65, 66) from the Henryhouse Shale of Oklahoma both appear to be somewhat smaller than the species previously discussed. The following measurements of T. longula are given by Ulrich and Bassler (1913, p. 542) : length 1.15 mm, height 0.60 mm, and width 0.70 mm. The left valve of the holotype (USNM 53289) proportionately is not as high as that of the other species, so the specimen appears to be longer. Lundin (1965, p. 65) gave meas- urements of 1.20 mm in length, 0.72 mm in height, and 0.75 mm in width for the average of his adult speci- mens, although his diagram (Lundin, 1965, p. 66, fig. 37) shows two specimens attaining a length of 1.30 mm. Tubulibairdia decaturi (Wilson, 1935) from the Birdsong Shale of Tennessee and T. windomensis Swartz and Oriel, 1948, from the Windom Member of the Moscow Shale of New York are both considerably smaller than any of the other species. Wilson (1935, p. 646) gave the measurements of T. decaturi as 0.62 mm in length, 0.36 mm in height, and 0.37 mm in width. The holotype (USNM 112905) is a small, corroded complete carapace, but single valves and fragments of Tubulibairdia in other collections from the Birdsong suggest that the species may attain larger dimensions. It will be necessary to study a large suite of specimens before the species can be adequately defined. 170 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Swartz and Oriel (1948, p. 563) gave the following dimensions for Tubulibairdia windomensis: length 0.89 mm, height 0.62 mm, and width 0.65 mm. This species differs from all others of the genus in being finely punctate and in having the greatest height of the cara- pace anterior rather than median or posterior in posi- tion. The latter character suggests that the species may be based on juvenile instars rather than on adult specimens and raises the possibility that the punctation may be a juvenile characteristic. Of all the North American species of Tubulibairdia, T. chaleurensis Copeland, 1962, is most similar to T. punctulata, and at one time it was thought that the specimens from the Schoharie might belong to T. chaleurensis. Copeland (1962, p. 48) gave measure- ments for one carapace as 1.70 mm in length, 1.10 mm in height, and 1.10 mm in width—dimensions which are comparable to those of the larger specimens here assigned to T. punctulata. Copeland (1962, p. 48), however, stated that the left valve overlaps the right "except anteriorly." In this respect T. chaleurensis differs from T. punctulata, in which the left valve over- laps the right around the entire free margin, including the anterior part. The Schoharie specimens agree with T. punctulata in this character. Sohn (1961, pp. 75, 76) placed seven European species in Tubulibairdia, of which three were only questionably so assigned. One of these, Cythere corbu- loides Jones and Holl, 1869 (the type-species of Da- leiella Boucek, 1937), while probably a pachydomellid, may not belong in Tubulibairdia. Of the other six species, all but Tubulibairdia? fecunda (Pribyl and Snajdr, 1950) are smaller than T. punctulata, and in T.? fecunda the hinge is not parallel to the ventral margin, as it is in most species of Tubulibairdia. Becker (1965, p. 178) has placed Microcheilinella seminalis Kummerow, 1953, in Tubulibairdia, but this is a very small species. In addition, Microcheilinella regularis Polenova, 1968, may also belong to Tubulibairdia, but this species also is smaller than T. punctulata. In sum- mary, the species of Tubulibairdia require further study, but until this is done the specimens from the Schoharie are believed to be most correctly identified with T. punctulata. Tubulibairdia punctulata is the most abundant ostra- code species represented in the Schoharie Formation; and it has been found in all three of the collections studied. Superfamily QUASILLITACEA Coryell and Malkin, 1936 Family BUFINIDAE Sohn and Stover, 1961 Genus Parabufina Smith, 1956 Parabufina? sp. PLATE 1: FIGURE 19 DESCRIPTION.—The lateral outline is elliptical and amplete; dorsal and ventral margins are smoothly PLATE 1.—Vietor josephinae, new species. Figures 1-5, views of holotype, USNM 163666, from coll. USGS 8211-SD: 1, right lateral, showing groove parallel to ventral margin; 2, left lateral; 3, dorsal, showing incised hinge-line; 4, ventral, showing flattened venter; 5, anterior. (All views X 15.) Phanassymetria collilupana, new species. Figures 6—10, views of holotype, USNM 163670, from coll. USGS 8211-SD: 6, right lateral; 7, left lateral; 8, anterior, showing grooves on both valves below dorsum which set off angular ridges; 9, dorsal, showing hinge-line incised and enclosed by ridges; 10, ventral, left valve broken. (All views X 15.) Neoaparchites sp. aff. N. mesleri (Bassler). Figures 11-13, views of USNM 163664, from coll. USGS 8211-SD: 11, right lateral; 12, left lateral; 13, dorsal. (All views X 15.) Tubulibairdia punctulata (Ulrich). Figures 14—17, views of USNM 163671, from G. A. Cooper loc. 119f: 14, dorsal, showing incised hinge-line; 15, anterior (projections on anterior margin are adventitious) ; 16, ventral, showing lap- pet on left valve; 17, right lateral, showing overlap. Figure 18, left lateral view of USNM 163672, also from G. A. Cooper loc. 119f. (All views X 15.) Parabufina? sp. Figure 19, USNM 163679, from collection USGS 8211-SD (X 15.) Bairdiacypris? sp. Figures 20-23, views of USNM 163665, from collection USGS 8211-SD: 20, left lateral; 21, right lateral; 22, dorsal (anterior end toward top); 23, ventral (anterior end toward top). (All views X 15.) Schohariella reticulosa (Warthin). Figure 24, view (Xl5) of latex cast of impression of a tecnomorphic right valve (NYSM 4452) from "Onondaga limestone (drift), Canandai- gua, N.Y." Schohariella grandis, new species. Figures 25, 26, views of holotype, USNM 163663, from G. A. Cooper loc. 119f: 25, lateral view of heteromorphic right valve (X 15) ; 26, ven- tral view of same specimen (X 30) after preparation, show- ing indistinct dolonoid (?) scar near contact margin; matrix not removed from right side. Parabingeria cooperi, new species. Figures 27-30, views of USNM 163660-62, all from G. A. Cooper loc. 119f: 27, lat- eral view of tecnomorphic left valve (USNM) 163661), show- ing velar ridge, fissus and zygal ridge, and scattered punctae; 28, ventral view of heteromorphic right valve (holotype, USNM 163660), showing inflation of velar bend by cru- mina; 29, lateral view of heteromorphic right valve (holo- type, USNM 163660) ; 30, lateral view of tecnomorphic right valve (USNM 163662, showing cuspate syllobium. (All views X 15.) NUMBER 3 171 PLATE 1 172 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY curved; and anterior and posterior margins are more sharply curved. A vertical outward-facing ridge is pres- ent at either end of the valve. The shell surface is smooth or minutely punctate. One valve measures 1.05 mm in length and 0.55 mm in height. MATERIAL.—One valve, from collection USGS 8211-SD. DISCUSSION.—A single specimen (USNM 163679) of a right (?) valve was obtained for study; it is illus- trated and described here to demonstrate the presence of this group of ostracodes in the Schoharie fauna. This specimen agrees with the generic description of Para- bufina given by Smith (1965, pp. 6, 7) but lacks the spinelets on the anterior and posterior margins present on some specimens from the Ludlowville Formation. The genus Parabufina was placed in synonymy with Bufina Coryell and Malkin, 1936, by Sohn and Stover (in Moore, 1961, p. Q375), but it is here considered a valid genus. The single specimen resembles, to some extent, the form described and illustrated by Jones (1890, p. 542, pi. 20, figs. 9a,b, 10a,b,) as Moorea kirkbyi from the "Corniferous chert" (Onondaga Limestone), although the vertical ridges of M. kirkbyi are more arcuate and the dorsal margin is straighter. Jones (1890, p. 542) described two specimens which he stated were "an inner and an outer cast." The internal cast that is deposited at the New York State Museum (NYSM 1592, new number; NYSM 13760/1, old number) is probably the original of Jones's figure 10, as his figure 9 is shown to be partly broken on the dorsal margin. Warthin (1945, index card) has suggested that this internal cast is unrecognizable but that it probably is of an Octonaria or Strepulites. However, in view of the two vertical ridges, it seems more likely that it belongs in the genus Parabufina. Literature Cited Adamczak, F. 1968. Palaeocopa and Platycopa (Ostracoda) from Mid- dle Devonian Rocks in the Holy Cross Mountains, Poland. Stockholm Contributions in Geology, 17:1-109, 46 plates. Bassler, R. S. 1941. Ostracoda from the Devonian (Onondaga) Chert of West Tennessee. Washington Academy of Sci- ences Journal, 31 (1) : 21-27, 1 plate. Bassler, R. S., and B. Kellett 1934. Bibliographic Index of Paleozoic Ostracoda. Geo- logical Society of America Special Paper, 1 :1-500. Becker, G. 1965. Revision Kummerow'scher Ostracodenarten aus dem deutschen Mitteldevon. Fortschrifte der Geol- ogie Rheinland und Westfalen, 9:151-188, 9 plates. Boucek, B. 1936. Die Ostracoden des bohmischen Ludlows. Neues Jahrbuch fiir Mineralogie, Geognosie, Geologie und Petrefactenkunde, 76(B) :31—98, 5 plates. 1937. Uber einige Ostrakoden aus der stufe ea des bohmischen Silurs. Krdlovske Ceske Spolecnosti Nauk, Vestnik, 1936 (2): 1-11, 5 figures. Boucot, A. J., and J. G. Johnson 1968. Brachiopods of the Bois Blanc Formation in New York. United States Geological Survey Professional Paper, 584-B:Bl-B27, 8 plates, 2 figures. Cooper, G. A., and others 1942. Correlation of the Devonian Sedimentary Forma- tions of North America. Geological Society of Amer- ica Bulletin, 53:1729-1794, 1 plate. Copeland, M. J. 1962. Canadian Fossil Ostracoda, Conchostraca, Euryp- terida and Phyllocarida. Geological Survey of Can- ada Bulletin, 91:1-57, 12 plates. Goldring, W., and R. H. Flower 1942. Restudy of the Schoharie and Esopus Formations in New York State. American Journal of Science, 240 (10):673-694. 1944. Carlisle Center Formation, a New Name for the Sharon Springs Formation of Goldring and Flower. American Journal of Science, 242:340. Grabau, A. W. 1906. Guide to the Geology and Paleontology of the Scho- harie Valley in Eastern New York. New York State Museum Bulletin, 92:1-386. Hall, J. 1867. Descriptions and Figures of the Fossil Brachiopoda of the Upper Helderberg, Hamilton, Portage and Chemung Groups. New York Geological Survey, Palaeontology of New York, 4(1): 1-428, 63 plates. 1879. Descriptions of the Gasteropoda, Pteropoda, and Cephalopoda of the Upper Helderberg, Hamilton, Portage, and Chemung Groups. New York Geologi- cal Survey, Palaeontology of New York, 5(2): 1— 492, 113 plates. 1884. Lamellibranchiata. 1. Descriptions and Figures of the Monomyaria of the Upper Helderberg, Hamil- ton, and Chemung Groups. New York Geological Survey, Palaeontology of New York, 5(1): 1—268., 45 plates. 1885. Lamellibranchiata. 2. Descriptions and Figures of the Dimyaria of the Upper Helderberg, Hamilton, Portage, and Chemung Groups. New York Geologi- cal Survey, Palaeontology of New York, 5(1): 269- 561, 51 plates. Hall, J., and J.M. ,Clarke 1888. Descriptions of the Trilobites and Other Crustacea of the Oriskany, Upper Helderberg, Hamilton, Portage, Chemung, and Catskill Groups. New York NUMBER 3 173 Geological Survey, Palaeontology of New York, 7:1-236, 36 plates. Hall, J., and G. B. Simpson. 1887. Corals and Bryozoa: Descriptions and Figures of Species from the Lower Helderberg, Upper Helder- berg, and Hamilton Groups. New York Geological Survey, Palaeontology of New York, 6:1-292, 66 plates. Johnsen, J. H., and J. B. Southard 1962. The Schoharie Formation in Southeastern New York. New York State Geological Association, 34th Annual Meeting, Port Jervis, New York, Guide- book, pages A7-A15. Jones, T. R. 1890. On Some Devonian and Silurian Ostracoda from America, France and the Bosphorus. Quarterly Journal of the Geological Society of London, 46: 534-556, 2 plates. Jones, T. R., and H. B. Holl 1869. Notes on Palaeozoic Bivalved Entomostraca, No. 9. Some Silurian Species. Annals and Magazine of Natural History, series 4, 3: 211-227, 2 plates. Kesling, R. V. 1958. A Middle Devonian Species of the Ostracod Genus Antiparaparchites. University of Michigan Con- tributions from Museum of Paleontology, 14 (12)' 191-200, 1 plate. Klapper, G., and W. Ziegler 1967. Evolutionary development of the Icriodus lateri- crescens Group (Conodonta) in the Devonian of Europe and North America. Palae onto graphic a, 127(A): 68-83, 4 plates. Krandievs'ky, V. S. 1963. Fauna Ostrakod Siluriis'kikh Vidkladiv Podillya. Akademiya Nauk Ukrains'koi RSR, 148 pages, 12 plates. Kiev. Kummerow, E. 1953. Uber oberkarbonische und devonische Ostracoden in Deutschland und in der Volksrepublik Polen. Geologie, 2(7) : 1-75, 7 plates. Lundin, R. F. 1965. Henryhouse Ostracodes. Oklahoma Geological Sur- vey Bulletin, 108:1-104, 27 plates, 45 figures. 1968. Haragan Ostracodes. Oklahoma Geological Survey Bulletin, 116:1-121, 22 plates, 51 figures. Martinsson, A. 1962. Ostracodes of the Family Beyrichiidae from the Silurian of Gotland. Uppsala University Palaeon- tological Institute, (41) : 1-369, 203 figures. 1963. Kloedenia and Related Ostracode Genera in the Silurian and Devonian of the Baltic Area and Britain. Uppsala University Palaeontological In- stitute, (42): 1-63, 36 figures. Mather, W. W. 1843. Geology of New York. Part 1. Comprising the Geology of the First Geological District. 653 pages. Albany, New York. Moore, R. C. (editor) 1961. Treatise on Invertebrate Paleontology, Part Q, Arthropoda 3, Crustacea, Ostracoda. 442 pages, 334 figures. New York City and Lawrence, Kansas: Geological Society of America and University of Kansas Press. Oliver, W. A., Jr. 1967. Stratigraphy of the Bois Blanc Formation in New York. United States Geological Survey Professional Paper, 584-A:Al-A8, 1 plate, 5 figures. Polenova, E. N. 1960. Devonskie ostrakod'i kuznetskogo basseina i minusinskoi kotlovin'i. Vsesoyuznogo Neftyanogo Nauchno-Issledovatel'skogo Geologorazvedochnogo Instituta {VNIGRI) Trudy, 152: 1-139, 13 plates. 1968. Ostrakod'i Nizhnego Devona Salaira, Tom'chum'- ishksii gorizont. Akademiya Nauk SSSR, Sibirskoye Otdeleniye, Institut Geologii i Geofiziky, 152 pages. 26 plates. Moscow. Pfibyl, A. and M. Snajdr 1950. On new Ostracoda from the Chotec" limestones— gy2 (Middle Devonian) of Holyne near Prague. Ceskoslovenske Stdtniho Geologickeho Ustavu, Sbornik, volume 17 (Paleontology), pages 101— 179, 5 plates. Roth, R. 1929. Some Ostracodes from the Haragan Marl, Devo- nian, of Oklahoma. Journal of Paleontology, 3(4) : 327-372, 4 plates. Ruedemann, R. 1930. Geology of the Capital District (Albany, Cohoes, Troy and Schenectady Quadrangles). New York State Museum Bulletin, 285 : 1-218. Smith, M. L. 1956. Some Ostracods from the Middle Devonian Led- yard and Wanakah Members of the Ludlowville Formation in Western New York. Journal of Paleontology, 30 (1) : 1-8, 1 plate. Sohn, I. G. 1960. Paleozoic Species of Bairdia and Related Genera. [1961] United States Geological Survey Professional Paper, 330-A:Al-A105, 6 plates. Swain, F. M. 1953. Ostracoda from the Camden Chert, Western Ten- nessee. Journal of Paleontology, 27(2) : 257-284, 3 plates. Swartz, F. M. 1936. Revision of the Primitiidae and Beyrichiidae, with New Ostracoda from the Lower Devonian of Penn- sylvania. Journal of Paleontology, 10 (7) : 541— 586, 12 plates. Swartz, F. M., and S. S. Oriel 1948. Ostracoda from Middle Devonian Windom Beds in Western New York. Journal of Paleontology, 22 (5):541-566, 3 plates. Swartz, F. M., and F. M. Swain 1941. Ostracodes of the Middle Devonian Onondaga Beds of Central Pennsylvania. Geological Society of America Bulletin, 52:381-458, 8 plates. 174 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Ulrich, E. O. 1891. New and Little Known American Paleozoic Os- tracoda. Cincinnati Society of Natural History Journal, 13:173-211, 8 plates. Ulrich, E. O., and R. S. Bassler 1913. Systematic Paleontology of the Lower Devonian Deposits of Maryland; Ostracoda. Maryland Geo- logical Survey, Lower Devonian:513-542. Vanuxem, L. 1840. Fourth Annual Report of the Geological Survey of the Third District. New York State Geological Survey, Annual Report, 4:355-383. 1842. Geology of New York, Part III. Comprising the Survey of the Third Geological District. 306 pages. Albany, New York. Walcott, C. D. 1884. Paleontology of the Eureka District. United States Geological Survey Monograph, 8:1-298, 24 plates. Warthin, A. S. 1937. Beyrichiacea, in Type Invertebrate Fossils of North America (Devonian). Wagner Free Institute of Science, Unit 9a, cards 1-106. 1945. Thlipsuridae, in Type Invertebrate Fossils of North America (Devonian). Wagner Free Institute of Sci- ence, Unit 9c, cards 1-82. Wilson, C. W., Jr. 1935. The Ostracode Fauna of the Birdsong Shale, Helderberg, of Western Tennessee. Journal of Pale- ontology, 9(8) :629-646, 2 plates. Preston E. Cloud, Jr., and Arthur J. Boucot Dzieduszyckia in Nevada ABSTRACT The rare bisulcate Famennian rhynchonelloid brach- iopod Dzieduszyckia, a homeomorph of the Trias- sic genus Halorella, occurs in barite of the Slaven Chert of Nevada. The rocks are western facies Devo- nian and the occurrence in them of Dzieduszyckia is the first reported for North America. Previously known occurrences of Dzieduszyckia are in Eurasia and North Africa. The presence of Dzieduszyckia implies lhat Slaven strata are, at least in part, of Famennian age, and odier fossils suggest that they may range in age from Siegenian to Famennian. One of the rarest of the cosmopolitan Late Devonian shelly invertebrates of the world is the brachiopod Dzieduszyckia. Previously recognized only in Poland, Morocco, and several localities in the Union of Soviet Socialist Republics (Biernat, 1967, p. 146), it is here recorded from Nevada. The Nevada occurrences have been the source of problems in affinities and age rela- tions (Gilluly and Gates, 1965, pp. 40, 41) since their discovery, as have also those from Old World localities (Biernat, 1967). The object of this note is to diminish these problems. Dzieduszyckia is a medium-size, strongly costate, laterally elongate, biconvex, strongly bisulcate shell, which, like Halorella, is asymmetrical about the median axis in degree of convexity and strength of ribbing. Unlike Halorella, however, which has only an incon- spicuous median septum, it possesses a prominent Preston E. Cloud, Jr., Department of Geology, University of California, Santa Barbara, California 93106. Arthur J. Boucot, Department of Geology, Oregon State University, Corvallis, Oregon 97331. median septum in the brachial valve; and, contrary to Siemiradzki (1909), it does not have spiralia. Its affinities have been reviewed critically by Biernat (1967), and we concur in her assignment of the genus to the Rhynchonelloidea. To discriminate unerringly between the Late Devonian Dzieduszyckia and its Triassic rhynchonelloid homeomorph Halorella, how- ever, still can be a problem where preservation is poor. We are grateful to the following for help with this study: G. A. Cooper, United States National Museum (USNM), for the loan of reference specimens of Dzieduszyckia from the Famennian of Morocco and from the Union of Soviet Socialist Republics and of Halorella from the Triassic of Europe; Henri Termier, University of Paris, for specimens of Dzieduszyckia from the Famennian of Morocco; D. V. Nalivkin, Academy of Sciences of the Union of Soviet Socialist Republics, for the gift (to J. G. Johnson) of the speci- men of D. baschkirica herein illustrated (Plate 1: fig- ures 1-5) ; N. J. Silberling and Mackenzie Gordon, Jr., for directing our attention to the problem, and to Silberling also for advice about field relations; James Gilluly for help in the field and advice about field relations; Nelson Shupe for photography of the speci- mens; J. G. Johnson for assistance in the laboratory; and Roman Kozlowski for informative correspondence relating to his and Gertruda Biernat's work on Polish Dzieduszyckia. Occurrences in Nevada The Nevada occurrences of Dzieduszyckia have been described by Silberling (in Gilluly and Gates, 1965, p. 40) as coming from four localities within the Shoshone Range, Lander County, in and adjacent to the Mount Lewis quadrangle. We have on hand seven lots of fossils, including thirty-four specimens, of which 175 176 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY six lots, and perhaps all seven, are from the same four localities. Although poor, the preservation of the ma- terial is strikingly unusual in that it consists of imprints of the interior and exterior of the shell in gray barite. No shell-material is preserved. Gilluly and Gates (1965) consider that the barite deposits of the Shoshone range have a replacement origin. It is not known ex- actly how the imprints of Dzieduszyckia happened to be preserved, but one explanation could be that the shell material was dissolved away before replacement of the matrix. The fossils were collected mainly by geologists of the United States Geological Survey (USGS) be- tween 1950 and 1963. Later, a collection made by H. E. Wheeler in 1938 was discovered in the Stanford collections. Locality information for some of the col- lections is not available in detail, but the unique preser- vation precludes reasonable doubt as to its provenance. In the following list of known occurrences the letter "M" before the numeral indicates that the collection is located in the United States Geological Survey laboratories at Menlo Park, California. 1. USGS Green loc. M-135. Carico Claim. Greystone barite mine, NEJ4 sec 26, T 28 N, R 45 E, Mt. Lewis (1/62,500) quadrangle, Nevada. Dzieduszyckia in dark crystalline limestone on strike with barite in easternmost quarry. Collected in 1957 by K. B. Ketner, N. J. Silber- ling, and James Gilluly. 2. Stanford University loc. 36430. Valley View barite mine, west side of Shoshone Range, north of Hilltop, north of Mt. Lewis quadrangle, approximately 17 miles southeast of Battle Mountain, Nevada. Collected in 1938 by H. E. Wheeler. 3. USGS Green Iocs. M-134 and 25023. Mound Springs barite quarries, 2 miles east and 1 mile south of NW corner T 28 N, R 44 E, Mt. Moses quadrangle, Nevada. Collected in 1957 by K. B. Ketner (his loc. 410) ; also by M. R. Mudge, Olcott Gates, and H. R. Gould; and again, in 1960, by Gordon Estes. 4. USGS Green loc. 25453, Hilltop, east of Mound Springs. T 28 N, R 45 E, Sonoma Range quadrangle, Nevada. Collected in 1950 by P. E. Cloud, Jr. 5. Carico barite prospect, Carico Lake, Shoshone Range, Lander County, Nevada. Perhaps the Carico claim at Greystone mine or perhaps another "Carico" claim south of the Mt. Lewis quadrangle. Specimen given to Ralph Roberts, United States Geological Survey, by an acquaintance in Battle Mountain. Problems of Identification and Age Despite its importance for regional geologic history, the taxonomy both of the Shoshone Range Dzie- duszyckia and of similar forms from other continents has been confused. This confusion started widi the description of the genus by Siemiradzki (1909), who mixed in with the spire-free type lot from the Devonian of the Kadzielnia quarry at Kielce, Poland, a spire- bearing shell from the Carboniferous of Maas, Belgium (Biernat, 1967, p. 136), and described the genus as a spire-bearer. The specimens from Morocco originally were assigned to the spire-free Triassic genus Halorella by Termier and Termier (1936, 1948) without reser- vation, although they were well aware of the age anomaly. In 1949, however, at the suggestion of G. A. Cooper, the Termiers considered the possibility of affinities with Dzieduszyckia. They decided to re- tain the name Halorella for the Moroccan shells but to call attention to the need for a general revision of the Paleozoic rhynchonelloids. The first specimen from Nevada to come to our at- tention was one collected by Cloud in company with James Gilluly in 1950. This specimen (Plate 1: fig- ure 13) is only a fragmentary impression of the anterior of a valve, and it was retained only because megafos- sils were so rare in the Slaven Chert. Since the beds were thought to be of upper Paleozoic age, this speci- men was referred for study to Mackenzie Gordon, Jr., who made the perceptive suggestion that it was very like the Triassic genus Halorella (internal report of USGS). Such an identification, however, would have been inconsistent with the field evidence and with the Middle Devonian age later suggested on the basis of conodonts by W. H. Hass and ostracodes by Jean Ber- PLATE 1: figures 1-5.—Dzieduszyckia baschkirica (Tscherny- schew), western part of Terek River area, southern Ural Mountains, Upper Devonian (D3J) ; brachial, pedicle, and anterior, posterior, and lateral views ( X 1) of a steinkern, USNM 160193A. Figures 6—17.—Dzieduszyckia sp., Slaven Chert, Shoshone Range, Nevada. All illustrated specimens except those of fig- ures 8 and 13 are from USGS Green loc. M-134 (=25023). 6, Internal impression of a pedicle valve (X 2), USNM 160185; 7, 9, 12, pedicle and lateral and anterior views (X 1) of a steinkern, USNM 160186; 8, internal impres- sion of a brachial valve (X 1.5), USNM 16087, from un- numbered locality at Carico Lake (No. 5 in list of occur- rences) ; 10, partial external impression (X 1), USNM 160188; 11, exterior of a pedicle valve (X 1), USNM 160189; 13, partial external impression (X 1) of first Ne- vada specimen studied, USNM 160190, from USGS Green loc. 25453; 14, impression of a brachial valve (X 1), USNM 160191; 15, internal impression of a brachial valve (X 2), USNM 160192; 16, internal impression of a brachial valve ( X 1), USNM 160193; 17, internal impression of a brachial valve (X 1),USNM 160184. NUMBER 3 177 PLATE 1 178 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY dan (Gilluly and Gates 1965, p. 40)—unless Halorella had a much longer range than was believed likely. Norman Silberling, therefore, became interested in the problem as a student of the Triassic, and he sum- marized the stratigraphic and paleontologic evidence available to December 1957 in an internal (USGS) report of which the critical parts later were published by Gilluly and Gates (1965, pp. 40-41). Silberling advised caution in choosing between a Triassic age, as favored by an assignment of the brachiopods to Halorella, and a Devonian age, as suggested by the microfossils of the Slaven Chert. Moreover, he already suspected that the brachiopods might be most similar to forms of "Halorella" described from the Upper Devonian of North Africa by Termier (1936). Following up on Silberling's suggestion, Cloud (in- ternal report, USGS, 23 September 1958; and in Gil- luly and Gates, 1965, p. 41) compared the Slaven Chert brachiopods then available with both Triassic Halorella and similar Moroccan Famennian rhyn- chonelloids assigned to H. intermedia and H. crassi- costa by Termier and Termier. He concluded that the Nevada and Moroccan specimens were similar, that they might be generically distinct from true Halorella, and that the Nevada occurrence should be considered to be of "Devonian? and perhaps Late Devonian" age. At that time Cloud was dubious of an assignment of these specimens to Dzieduszyckia because of the spiralia that were supposed to be found in this genus. The age of Dzieduszyckia has been somewhat less of a problem. Originally assigned by Siemiradzki (with a query) to the Hypothyridina cuboides beds of Frasnian and upper Givetian age, the Polish specimens now are regarded as of Famennian age (Biernat, 1967, p. 140). A Famennian age also is established for the Moroccan and Russian specimens. As a result of our restudy, and of the work of Biernat, we conclude that the Nevada specimens belong to the genus Dzieduszyckia and that a Famennian age is indicated by the ages of known occurrences of that genus in Morocco, Poland, and the Union of Soviet Socialist Republics. Dzieduszyckia sp. PLATE 1 : FIGURES 6-17 The description below is based on thirty-four incom- plete specimens from Nevada, of which only ten are good enough to illustrate. Unfortunately, the preserva- tion even of these ten is too poor to warrant assignment or description at the species level. EXTERIOR.—Both valves are strongly convex, the pedicle valve being the deeper, and both bear a broad, well-developed sulcus. The position of maximum width is near the midlengdi. The hinge-line is rela- tively short; it seems to be roughly about one-third of the maximum width. The beak of the pedicle valve is incurved. The posterolateral and lateral margins round evenly into the gently curved anterior margin. The surface of commissure is flat and medially incised in plan view. Ornamentation consists of numerous costae, crossed by a few concentric growth lines. The number of costae is quite variable among specimens studied, ranging from 12 to about 30. The number of costae in the sulci is also variable, with four to eight being observed. The prominent sulci present in both valves originate at a relatively early growth stage. The internal impressions of the costae and the interspaces separating them are U-shaped in cross section and in- crease in size anteriorly. The shell is asymmetrical about the median axis, one side appearing to be slightly more convex and having a larger number of costae than the opposite side—a feature that Dzieduszyckia and Halorella share in common. INTERIOR OF BRACHIAL VALVE.—A well-developed median septum is present (Plate 1: figures 15-17). The septum extends anteriorly to about the mid- length or beyond. The outline of the muscle field is not well displayed on the Nevada specimens. The impress of the external ornamentation is strong. INTERIOR OF PEDICLE VALVE.—Short dental lamellae are present (Plate 1: figure 6). Secondary shell ma- terial was deposited in the area of the umbonal cavities, and the impression of the muscle field projects well beyond the impression of the umbonal chambers. Al- though the muscle field is poorly preserved, it appears to have consisted of paired, elongate, lateral elements separated by a median area that is raised off the floor of the valve relative to the lateral areas. It could not be determined whether the median area was oc- cupied entirely by the adductor muscle attachments or not. AFFINITIES.—Cloud (in Gilluly and Gates, 1965, p. 41) has already contrasted the external form of the Nevada material with that of Triassic Halorella. He noted particularly that the Nevada specimens seemed to lack the "sharply defined beak ridges and palin- NUMBER 3 179 tropes" of the Triassic forms, and also that the beak of Triassic Halorella is suberect, as opposed to the incurved condition observed in its Devonian homeo- morph from Nevada and Morocco (and also Poland, Biernat, 1967, p. 146). Internally, the presence of a prominent median septum in the brachial valve dis- tinguishes our specimens from Triassic Halorella. Well-preserved specimens of Dzieduszyckia basch- kirica (Plate 1: figures 1-5) are similar to the Nevada material in outline and general external form. How- ever, the costae of the Russian specimens (D. basch- kirica) are flatter across their crests and the interspaces are relatively narrow. The median septum is not con- spicuous in the brachial valve of the Russian specimens that we examined, but Biernat (1967, p. 137)assures us that it is conspicuous in other specimens. The type- species, D. kielcensis, from Poland, described and illus- trated in detail by Biernat (1967), closely resembles the Nevada material in its span of variation as regards costae and interspaces, and it displays a prominent sep- tum in the brachial valve. Similarity is also observed between the Nevada specimens and the Moroccan species D. intermedia and D. crassicosta, which like other Dzieduszyckia, possess a median septum in the brachial valve (Termier and Termier, 1950, pi. 100, fig. 17). In seeking to establsh suprageneric affinities, we sectioned one of the Russian specimens, D. baschkirica, and one of the Nevada specimens to see if we could find the spiralia reported by Siemiradzki (1909) to characterize the genus. Our results were negative, con- sistent wtih those of Biernat (1967), who also observes (Biernat, 1967, p. 19) that the spiralium-bearing specimen illustrated by Siemiradzki (1909, fig. 3b) did not belong to the type lot of D. kielcensis from the Devonian at Kielce but came instead from the Car- boniferous of Maas, Belgium. Termier and Termier (1950) also found no spiralia in the Moroccan speci- mens. All evidence, therefore, is now consistent with the assignment of Dzieduszyckia to the Rhynchnel- loidea. It probably belongs to the Dimerellidae, in the subfamily Halorellinae. It is evidently not a member of the Athyrisinidae, as assigned by Boucot (in Moore, 1965, p. H654). DISTRIBUTION.—The rarity of the genus Dziedus- zyckia raises puzzling questions about how it became distributed so widely (Biernat, 1967, p. 146). Late Devonian brachiopods are among the most cosmo- 372-386 O—71 13 politan marine invertebrates known in the northern hemisphere, and many of them are very abundant. With the open connections and free genetic exchange thus implied, it is hard to see why equally cosmopolitan forms like Dzieduszyckia should be so rare. One can only guess that they might have been ovoviviparous or limited to an extremely specialized environment. Age of the Slaven Chert Earlier reports on the Slaven Chert (Gilluly and Gates, 1965, pp. 40-41) enable a Middle Devonian and pos- sibly Late Devonian age to be assigned to the unit. The presence of Dzieduszyckia is now interpreted to demon- strate that the Slaven Chert includes strata of Late Devonian age. In addition, however, Berdan (in letter dated 9 January 1967 to J. G. Johnson) concludes that some of the ostracodes from the Slaven Chert at Gil- luly's locality F-106 are very similar to forms from the Siegennian Quadrithyris zone of the Windmill Lime- stone (Johnson, 1965). This suggests that part of the Slaven Chert also may be of Early Devonian age. Thus, it appears that the Slaven Chert spans much of the Devonian and that Dzieduszyckia comes from its upper part. On present evidence Dzieduszyckia ap- pears to be distinctive of a Famennian age. Literature Cited Biernat, G. 1967. New Data on the Genus Dzieduszyckia Siemirad- zki, 1909 (Brachiopoda). Palaeontologica Acta Polonica, 12(2) : 133-156, plates 1, 2. Gilluly, J., and O. Gates 1965. Tectonic and Igneous Geology of the Northern Shoshone Range, Nevada. United States Geological Survey Professional Paper, 465:1-153, 6 plates. Johnson, J. G. 1965. Lower Devonian Stratigraphy and Correlation, Northern Simpson Park Range, Nevada. Bulletin of Canadian Petroleum Geologists, 13(3) : 365-381, illustrated. Moore, R. C. (editor) 1965. Treatise on Invertebrate Paleontology, Part H, Brachiopoda. 927 pages, 746 figures. New York City and Lawrence, Kansas: Geological Society of America and University of Kansas Press. Siemiradzki, J. 1909. Zbiory L. Zejsznera z Kieleckiego Dewonu. Acad. Umiejetnosci W. Krakowie, Sprawozdanie Kom. Fizyograficznej, 43:85-86, plate 3, figures 3a, 3c- 180 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY g. [Also, later in 1909, in French, in Bulletin In- ternationale de VAcademe Science de Cracovie, Annee 1909, pages 768-769, plate 13, figures 3a, 3c-g.] Termier, H. 1936. Cinquieme partie (Paleontologie), in Etudes Geo- logiques sur le Maroc central et le moyen Atlas septentrional. Morocco Service des Mines et de la Carte Geologique, Notes et Memoires, 33(3): 1196-1201, plates 11, 12. Termier, H., and G. Termier 1948. Les Phenomenes de speciation dans le genre Halorella. Morocco Service Geologique Notes et Memoires, 71:47—63. 1949. Sur les Genres Halorella et Dzieduszyckia. Morocco Service Geologique Notes et Memoires, 74:113-115. 1950. Invertebres de l'ere primaire—Bryozoaires et Brachiopodes. Paleontologie Marocaine. 2. Morocco Service Geologique Notes et Memoires, 77:1-253, plates 52-122. J. Thomas Dutro, Jr. The Brachiopod Pentagonia in the Devonian of Eastern United States ABSTRACT Pentagonia is a specialized meristellid brachiopod widi a restricted stratigraphic and geographic distribution in the Devonian of eastern North America. The genus is characterized by its external pentagonal shape, bodi in plan view and in posterior outline. Internally, the pedicle valve has a large, flabellate muscle area bounded by relatively prominent dental plates. The brachial interior is dominated by a massive cardinal process and a low median septum, both of which are interpreted as evolving from meristellid structures. Pentagonia has no functional pedicle, and shell shape is apparently related to a free-living condition on the sea floor. An extraordinary feature of this brachiopod is a pair of pouchlike structures, near the postero- lateral margins, which may have been either brood pouches or the sites of the sexual organs. Pentagonia was recognized as a separate genus more than a century ago. The genus was named by Isaachar Cozzens in 1846, with P. peersii as the type-species based on material obtained by the Rev. Benjamin O. Peers at the Falls of the Ohio River, Louisville, Ken- tucky. The oldest described species now assigned to the genus is Atrypa unisulcata Conrad, 1841. Conrad's fossils came from the Onondaga Limestone at Scho- harie, New York. James Hall (1861), apparently unaware of Cozzens' paper, proposed the genus Goniocoelia based on Conrad's species. Later, Hall (1894) stated that Goniocoelia was an exact synonym /. Thomas Dutro, Jr., United States Geological Survey, Room E-325, United States National Museum, Washington, D.C. 20242. Publication authorized by the director, United States Geological Survey. of Pentagonia and therefore should be stricken from the list of brachiopod genera. Other species ascribed to Pentagonia have been described and illustrated by Hall (1867), Butts (1941), and Cooper (1944). Pentagonia peersii was described and illustrated by Nettleroth (1889) from Kentucky as Meristella unisulcata Conrad. Meristella lenta, assigned herein to Pentagonia, was described by Hall in 1867. Pentagonia? goldringae (Flowell), from the Esopus Formation, was inadequately described and illustrated by Howell (1942). Occurrences of the genus in Ontario were listed by Stauffer (1915) and those in Kentucky were cited by Savage (1931). Stauffer (1909) did not report Penta- gonia from Ohio, but the genus later was listed by Stumm (1942) as occurring in the fauna of the Prout Limestone. Devonian occurrences of the genus in Ten- nessee were discussed by Dunbar (1919), and those from rocks of Onondaga age in Virginia were fisted and illustrated by Butts (1941). A summary of the distribution of Pentagonia in rocks of Schoharie age was provided by Boucot and Johnson (1967, 1968). They reported the species P. unisulcata from the Kanouse Sandstone and Woodbury Creek Member of the Esopus Formation in southeast- ern New York, the Frog Mountain Sandstone of Alabama, the Camden Chert of western Tennessee and Devonian strata of the James Bay lowland in Canada. The specimens from the Schoharie itself are listed as Pentagonia cf. P. unisulcata (Conrad). The only report of the genus outside eastern North America known to me is that by Caster (1939) from the Lower Devonian Floresta fauna of Colombia. Caster's species, P. gemmisulcata, appears to be very close to P. unisulcata. 181 182 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Pentagonia? sp., reported years ago by Williams and Breger (1916) from the Chapman Sandstone of Maine, was only doubtfully assigned to that genus. The assign- ment has been questioned by Boucot (in Boucot and others, 1964) and the illustrated specimens in the collections of the United States National Museum (USNM) were examined again in the course of this study. A positive generic identification is not possible on the basis of these poorly preserved internal molds. Because of the apparent absence of a septum in the brachial valve, these specimens are removed from Pentagonia. Phylogenetic Relationships With the exception of the highly doubtful Chapman Sandstone occurrence, the earliest true Pentagonia occurs in rocks of Schoharie age. During that time, in the late Early Devonian, the genus, although a rare element of the total brachiopod assemblage, apparently was rather widely distributed. Pentagonia lenta (Hall) from the "Oriskany" of Ontario is actually from the Springvale Sandstone, as pointed out by Stauffer (1915). This small species was apparently a local variant related to the much more widely distributed P- unisulcata. The genus is well represented by P. unisulcata in Onondaga age rocks from New York to Virginia, but that species has not yet been reported from the western equivalents of the Onondaga in Ohio, Kentucky, Indiana, and Illinois. The Centerfield Limestone of New York and its correlatives in Ontario, Ohio, Pennsylvania, Kentucky, and Indiana have yielded many specimens of die youngest known species, Pentagonia peersii Cozzens. The species is especially abundant in the Beechwood Limestone of southern Indiana. M I C H I G A N* / //' ( x y / + Pentagonia lenta (Hall) x Pentagonia unisulcata (Conrad) ?x Pentagonia cf P unisulcata (Conrad) O Pentagonia peersii Cozzens A B Main Onondaga outcrop Possible shoreline Edge of Devonian outcrop FIGURE 1.—Distribution of localities yielding Pentagonia in eastern North America. NUMBER 3 183 Z Z GIVETIA HAMILTC P. peersii i i i i < z o < _i s P. unisulcata UJ z LL o LU z o P. cf. P. j i Ld 1ARI unisulcata P. lenta MSIA N SCHOr (Camden) . \ \ 1 1 l I co Ld z> \ Q. \ 1 O CO \ 1 Ld \i Z >- Meristella 3ENI A ISKAN lentiformis Ld cr CO O large, robust species Pentagonia peersii had a more re- stricted range, and the genus became extinct before the end of the Middle Devonian. The history of the genus, then, is one of increasing species size from Schoharie through Centerfield time (see Figures 3-5). Although other genera among the Meristellinae lasted until the end of the Middle Devo- nian (Tully time), no representative of Pentagonia has been reported from post-Centerfield rocks. The reason for the disappearance of the meristellids is not known, but their ecologic position in the brachiopod assemblage apparently was taken over by the closely related athy- rids, which had external stabilizing mechanisms. Pos- sibly, the unanchored condition of the Pentagonia shell made survival difficult in the relatively high- energy environments that resulted in well-sorted clastic carbonate rocks. No representative of the genus has been collected from mud or silt rocks; doubtless the free-living position of die shells would have resulted in rapid clogging and death of individuals in such an environment. There must have been a fine balance between current action and firm substrate for the genus to have survived at all. FIGURE 2.—Suggested phylogenetic relationships among the species of Pentagonia and Meristella lentiformis Clarke. In searching for an ancestor of Pentagonia, species of Meristella in beds older than the Schoharie were examined. The most likely progenitor to the specialized Pentagonia appears to be Meristella lentiformis Clarke, 1900. This species was described from the Glenerie Limestone of southeastern New York. Meristella lentiformis retained a foramen that sug- gests a functional pedicle throughout its lifespan. In Pentagonia, reduction of a functional pedicle resulted in, or was the response to, a free-living condition on the sea floor. The flattened pedicle valve probably re- flects an adjustment to stabilize the shell on the bot- tom. Pentagonia lived with its pedicle valve down, and the development of a thick callosity in the posterior part of the pedicle valve undoubtedly aided in the stabilizing process (Rudwick, in Moore, 1965, p. H201; Grant, 1965, p. 27; Ager, 1967, p. 160). The genus was widely dispersed during Schoharie time (see Figures 1,2) but became more restricted and more specialized during the Middle Devonian. The / 32^ 3D- ' Ppeersii / / "=37 / Plenta 11=14 / 22 24 26 28 30 32 FIGURE 3.—Comparative-size polygons for Pentagonia lenta (Hall), P. unisulcata (Conrad), and P. peersii Cozzens, fitted visually. 184 Life Habit In contrast to the other genera that are included in the subfamily Meristellinae, the adult Pentagonia had no foramen; hence, it apparently had no functional pedicle. Some very small specimens show what may be a tiny opening that could have accommodated a hair- like pedicle during early growth stages. With age, how- ever, this opening was closed and the animal must have lived free on the sea floor. The flattened pedicle valve would have contributed to the stability in this position of life, as would the secondarily thickened posterior regions of the valves, especially the heavy umbonal region of the pedicle valve. A number of specimens in the collections from the Beechwood Limestone of southern Indiana have organisms attached to the shell. Common epifaunal elements are the tabulate coral Aulopora and various kinds of bryozoans (Plate 2: figures 11, 12). These organisms are invariably found either on the brachial valves or on that part of the pedicle valve, lateral to the prominent carinae, where they would have been elevated above the water-sediment interface. Similar as- sociations were reported by Ager (1961) in his paper on the epifauna of a Devonian spiriferid. Many of the shells from the Beechwood locality near Charlestown, Indiana, have been drilled or bored by other organisms (Plate 2: figures 5, 8, 9). Three kinds of borings have been observed. One roughly circular cylindrical type of boring is by far die most common. A second type, commonly attributed to bar- nacles (Tomlinson, 1963), is elliptical to teardrop- shaped in plan. The third variety, an elongate tubular boring, perhaps of sponge origin, is quite rare. Tabulated below are the kinds and number of bor- ings observed in a single large collection consisting of 93 brachial valves, 410 pedicle valves, and 12 articulated shells. Brachial Pedicle Articulated Borings Valves {93) Valves {410) Shells {12) Circular /cylindrical Single, per shell 5 35 1 Multiple, per shell — 9 1 Elliptical/teardrop Single, per shell 1 1 — Multiple, per shell 1 2 — Elongate/tubular Single, per shell — — — Multiple, per shell — 2 — Clearly, because of the disproportionate number of disarticulated shells and the preponderance of the SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY heavier pedicle valves, this collection represents an ac- cumulation of shells as gravel on the sea floor. The petrography of the Beechwood Limestone supports its interpretation as a lag deposit at many localities. No doubt these particular shells were drilled after their disarticulation. However, such a large number of in- dividuals of Pentagonia must reflect the existence of a considerable community of this species in the neigh- borhood of the Falls of the Ohio during Centerfield time. No other recorded locality of the genus has yielded more than a few specimens of what must have been a minor element of the total brachiopod assemblage. Superfamily ATHYRIDACEA M'Coy, 1844 Family MERISTELLIDAE Waagen, 1883 Subfamily MERISTELLINAE Waagen, 1883 Genus Pentagonia Cozzens, 1846 1846. Pentagonia Cozzens, p. 158, pi. 10, figs. 3a,b.—Hall, 1867, pp. 309-311, pi. 50, figs. 18-35.—Hall and Clarke, 1894, p. 80, pi. 42, figs. 22-32.—Schuchert and LeVene, 1929, p. 95.— Caster, 1939, p. 171.— Cooper, 1944, p. 333.—Boucot, Johnson, and Sta- ton (in Moore, 1965), p. H656. 1861. Goniocoelia Hail, p. 101. ORIGINAL DIAGNOSIS.—"Shell bivalve, inequivalve, having five sides, somewhat gaping; lower valve with three sides, upper with two; beaks contiguous." (Coz- zens, 1846, p. 158.) EMENDED DIAGNOSIS.—Specialized, strongly bicon- vex meristellids with pentagonal shell form; pedicle valve with broad, flat sulcus bounded by angular cari- nae; brachial valve with high fold, generally having narrow median groove. Pedicle interior with striate, flabellate muscle area and flaring dental plates ex- tending about halfway to anterior margin of muscle field. Brachial interior with enlarged brachial process that has rotated posteriorly to fill delthyrial cavity; process with concave anterior surface and a cardinal plate that extends posteriorly as a scoop-shaped con- cavity; low median septum extends downward across the anterior face of the cardinal process and anteriorly nearly to the anterior margin of the valve. No func- tional pedicle in the adult growth stages. Pouchlike developments in posterolateral regions. TYPE-SPECIES.—Pentagonia peersii Cozzens, 1846. NUMBER 3 185 INCLUDED SPECIES.—Pentagonia peersii Cozzens, P unisulcata (Conrad), P. lenta (Hall), P. gemmisulcata Caster. REJECTED SPECIES.—Pentagonia? sp. Williams and Breger, Pentagonia? goldringae Howell. DISCUSSION.—In Cozzens' original diagnosis (1846), two significant points should be noted. The five-sided nature of the shell clearly refers to its ap- pearance in posterior outline, as is shown by his figured specimen (Cozzens, 1846, pi. 10, fig. 3a). The second significant point is the emphasis on contiguous beaks. A significant feature of the genus, it would appear, is the absence of a functional pedicle in the adult animal. In reverting to Pentagonia peersii as the type-species, I depart from Hall (1867) and all subsequent workers. Pentagonia peersii was based on material from the Falls of the Ohio, now recognized as coming from the Beechwood Limestone. Following Hall's suggestion, later usage referred to this Middle Devonian form as P. biplicata. Conrad's species P. unisulcata was de- scribed on material from the Onondaga Limestone. Nearly all workers have recognized the distinct specific differences between the Schoharie-Onondaga form and the one from the Centerfield and its equivalents. How- ever, P. biplicata Hall is a synonym of P peersii Cozzens. Pentagonia peersii Cozzens, 1846 FIGURES 3, 4; PLATE 1: FIGURES 9-12; PLATE 2: FIGURES 1-3,5-12 1846. Pentagonia peersii Cozzens, p. 158, pi. 10, figs. 3a,b. 1862. Meristella? unisulcata (Conrad) var. biplicata Hall, pi. 2, p. 158, fig. 18. 1862. Meristella? unisulcata (Conrad) var. uniplicata Hall, pi. 2, p. 158, figs. 19, 24, 25. 1867. Meristella {Pentagonia) biplicata Hall, p. 311, pi. 50, figs. 30-35. 1889. Meristella unisulcata (Conrad).—Nettleroth, pp. 99, 100, pi. 15, figs. 9-16. 1894. Pentagonia unisulcata (Conrad).—Hall and Clarke, pi. 42, figs. 25-32.—Savage, 1930, pi. 4, figs. 24-26; 1931, pi. 30, figs. 17, 18. 1936. Pentagonia sp. Cooper, p. 9, fig. 8.8. 1942. Pentagonia bicostata Stumm, p. 556, pi. 84, fig. 46. 1944. Pentagonia bisulcata (Cooper), p. 333, pi. 127, figs. 32, 36.—Rickard, 1964, chart.—Oliver and others, 1969, chart. ORIGINAL DESCRIPTION.—"Shell somewhat gaping, with five sides and three carinae; two of the carinae on the lower valve commence at the beak, and diverge toward the margin, and end at the opening, the valve being concave between them; the lateral margins small and nearly vertical, an elevated carina on the middle of the upper valve, rendering its sides somewhat concave. This carina has a shallow furrow in it, commencing at the beak and running more than halfway along the shell towards the opening. On each side of the upper valve and contiguous to the beaks, are two angular protuberances, giving the shell when viewed at the beaks, a pentagonal appearance, and at the same time a visage-form look; length 1.1 inch, breadth 0.9 inch." (Cozzens, 1846, p. 158.) EMENDED DESCRIPTION.—Shell large for the genus, unequally biconvex, pentagonal in plan view and pos- terior profile; wider than long with greatest width near anterior margin; shell structure fibrous, impunctate; external ornament consisting only of concentric growth lines. Brachial valve strongly arched with high fold ex- tending to anterior margin, fold with shallow furrow extending to midlength; sharp to rounded carinae variably developed on either side of beak, at acute angles to the plane of commissure, extending about one-third to one-half distance to anterolateral mar- gins; generally one carina on either side of die beak, very rarely two carinae on either side; carinae are external reflections of internal pouches. Interior dom- inated by massive cardinal process with steep concave anterior face, cardinal plate depressed to form con- cavity that lies in the plane of the valve, process ex- tending posteriorly to completely fill umbonal cavity, often with bilobed appearance in posterior view; nar- row, low median septum extends across anterior face of process and about one-half distance to anterior margin; adductor muscle tracks, deeply impressed on either side of septum, reflect anterior movement of narrow, ellip- tical adductor marks; process bounded by deep sockets; crural bases project posteriorly from anterior face of cardinal process on either side of median septum; pouchlike depressions developed laterally from sockets, many with pitted inner surfaces. Pedicle valve broadly sulcate with sharp carinae extending from beak to anterolateral margins, dividing sulcus from steep lateral slopes; thin costa in middle of sulcus, complimenting furrow in brachial fold. In- terior with broad, flabellate, striate muscle area bound- ed by extensions of dental plates for about one-half 186 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY 1959. W 1 mm | Nettleroth's -—*•<&--. 32 illus. spec. ?^ \ -—° 30 y ./ ? -\ —' 1 \ °4 - ? - P.peersii 28 / / ?* L (Ontario) / n=9 26 24 P.peersii (Beechwood) n=29 1/ *s /* Cozzen's type 22 / 20 " / t / 18 / // m^ 16 '/ 14 - o^ 12 12 i i 1 I —L L i , i 10 14 16 18 20 22 24" 26 28 30 32 L FIGURE 4.—Length-width relationships of collections of Penta- gonia peersii Cozzens from the Hungry Hollow Formation of Ontario and the Beechwood Formation of southern Indiana. its length; prominent teeth surmount the dental plates; pitted pouchlike areas lateral to the dental plates resemble those in the brachial valve. DISCUSSION.—The character by which Hall orig- inally suggested that this stratigraphically younger species be differentiated was the development of two carinae on either side of the brachial beak. This fea- ture, however, appears to be only a very rare variant of the normal condition. Only three of 105 brachial valves in the collection from the Beechwood show these carinae. The main character which separates this species from P. unisulcata is average size. On the whole, the shells from the Beechwood and other Cen- terfield equivalents are much larger than those of P. unisulcata from the Onondaga (Figure 3). Immature shells, those up to about half the size of adults in P. peersii, are almost identical in form to adults of P. unisulcata. It is a curious coincidence that two later species designations appear to fall into the category of unjusti- fied emendations (Stumm, 1942; Cooper, 1944; Rickard, 1964; Oliver and others, 1969). These can be no more than inadvertent oversights because "bipli- cata" was clearly the trivial name advocated by Hall. ILLUSTRATED SPECIMENS.—USNM 163807-13 from Beechwood Limestone, near Charlestown, Indiana, Greene Collection, USNM Ace. No. 232542. Pentagonia lenta (Hall) FIGURES 3, 5; PLATE 1: FIGURES 5-8; PLATE 2: FIGURE 4 1867. Meristella lenta Hall, p. 420, pi. 63, figs. 19-22.— Stauffer, 1915, pp. 79-82. Pentagonia unisulcata (Conrad)—Boucot, p. 748, pi. 94, figs. 9, 10. ORIGINAL DESCRIPTION.—"Shell small, broadly ovate, or transversely oval, with a slightly projecting beak and very unequally convex valves. Ventral valve nearly flat in the upper part, with sharply angular cardinal margins becoming deeply and subangularly sinuate towards the front, where it is slightly bent upwards. Dorsal valve very ventricose in the umbonal portion, and subangular along the center, with the sides some- what flattened. This species differs from any described form of the genus so far as it can be ascertained. In the form of the ventral valve it approaches somewhat the M. (Pentagonia) unisulcata, but the sinus is nar- row instead of embracing the greater part of the valve, and it differs so materially in other respects that there is no danger of confounding the two." (Hall, 1867, p. 420.) EMENDED DESCRIPTION.—Shell small, unequally biconvex, subovate, about as wide as long widi the greatest width near midlength; triangular to flattened pentagonal shape in posterior outline; shell structure fibrous, impunctate; exterior smooth. Brachial valve strongly arched with high fold ex- tending to anterior margin; fold unmodified, without lateral carinae. Interior with massive cardinal process, as in P. peersii, and median septum extending about halfway to anterior margin; narrow adductor muscle tracks faintly impressed on either side of septum. Pedicle valve flattened, sulcate, faintly carinate with two rounded ridges bounding sulcus, which occupies about one-third the width of the valve. Interior with broad, flabellate striate muscle field and dental plates like those of the type-species. DISCUSSION.—Hall (1867) suggested the possible relationship of this species to Pentagonia, but he did not make the assignment because of differences in external shape. In most respects, however, this small species is an ideal progenitor of the two larger forms of Pentagonia. In addition, the interior, with its mas- sive cardinal process, is identical to Pentagonia. Meris- tella lenta is herein assigned to that genus. The stratigraphic position of this species needs clari- fication. According to Hall (1867), M. lenta was found "In rocks of the age of the Oriskany sandstone or NUMBER 3 187 Upper Helderberg limestone, near Cayuga, C.W." Stauffer (1915) was able to pinpoint the occurrence of the species to the Springvale Sandstone. Oliver (1967, p. A3) suggests that the Springvale is of Bois Blanc age. Thus, the stratigraphic position of Penta- gonia lenta is approximately the same as that of the earliest P. unisulcata reported from eastern New York. Although die two species probably represent geo- graphically isolated communities in Schoharie lime, it is possible that P. lenta represents a slightly older horizon. Pentagonia unisulcata appears in Onondaga rocks higher in the Ontario section and could be con- sidered a descendant of P. lenta. The P. unisulcata of eastern New York can be derived from Meristella lentiformis in the same geo- graphic area through the small species of Pentagonia described by Boucot (1959) from the Woodbury Creek. Boucot's P. unisulcata is assigned to P lenta. ILLUSTRATED SPECIMEN.—USNM 28038 from "Upper Oriskany" (now the Springvale Sandstone), 6 miles west of Cayuga, Ontario, Canada. Collector: Charles Schuchert. USNM Ace No. 30039. Pentagonia unisulcata (Conrad) FIGURES 3, 5 1841. Atrypa unisulcata Conrad, p. 56. 1861. Atrypa uniangulata Hall, p. 101. 1862. Meristella? unisulcata (Conrad), Hall, pi. 2, p. 158, figs. 17, 20-23 (not figs. 19, 24, 25). 1867. Meristella {Pentagonia) unisulcata (Conrad)—Hall, 1867, p. 309, pi. 50, figs. 18-29 (not figs. 30-35). 1889. Not Meristella unisulcata (Conrad), Nettleroth, p. 99, pi. 15, figs. 9-16. 1915. Pentagonia unisulcata (Conrad), Stauffer, p. 104, 245 (not pp. 160, 171, 175, 234).—Dunbar, 1919, pp. 87, 89.—Goldring, 1935, p. 148, figs. 53B-D.— Butts, 1940, pp. 300, 301, 304, 305; 1941, pi. 115, figs. 17-21, 35.—Cooper and others, 1942, chart- Cooper, 1944, p. 333, pi. 127, fig. 37.—Oliver, 1954, pp. 633, 634, 638-640; 1956, pp. 1452, 1456, 1462, 1469.—Rickard, 1964, chart.—Boucot and others, in Moore, 1965, p. H656, pi. 533, figs. 2a-d (not figs. 2e-f).—Oliver and others, 1969, chart. 1930. Not Pentagonia unisulcata (Conrad), Savage, p. 47 50, 53, 62; 1931, p. 242, pi. 30, figs. 17, 18. ORIGINAL DESCRIPTION.—"Trigonal, superior valve with a broad, prominent middle, sulcated longitudi- nally, the sulcus being obsolete towards the base; sides concave; the depression giving the margins a cari- nated appearance; inferior valve deeply concave, sub- angulated in the middle; umbonal slope carinated, and the area between it and the margin much depressed. Locality.—Schoharie, in Onondaga limestone." (Con- rad, 1841, p. 56.) EMENDED DESCRIPTION.—Shell medium-sized, un- equally biconvex with triangular to pentagonal shape in plan view and in posterior outline, greatest width near anterior margin; shell impunctate with fibrous structure and ornamented only with concentric growth lines. Brachial valve with prominent sulcated fold and two lateral carinae that extend at acute angles from the umbo across the posterolateral slopes; tendency towards deeply concave slopes between the fold and the lateral margins produces a corrugated appear- ance; sulcus in fold extends to or nearly to the ante- rior margin. Interior with prominent cardinal process, as in Pentagonia peersii, and median septum extend- ing about two-thirds distance to anterior. Details of muscle scars and presence or absence of posterolateral pouches not determined. Pedicle valve broadly sulcate with two prominent sharp carinae extending from umbo across lateral slopes to anterolateral margins. Interior with broad, w mm. I / j P. unisulcata P cf. unisulcata J (Camden) / n=9 , I (Virginia) P. unisulcata Onondaga) n=!2 ^ P. unisulcata " (Woodbury Creek) / P. lenta (Springvale) n=!2 n=5 / x 6 8 10 12 14 16 18 20 22 24 |_ mm FIGURE 5.—Length-width relationships of several collections of Pentagonia lenta (Hall) and Pentagonia unisulcata (Con- rad) . 188 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY striate, flabellate muscle area bounded by dental plates that extend nearly half the distance to the anterior edge of the muscle field. DISCUSSION.—Much confusion has surrounded this species since its original description by Conrad. In the 1800s, and even as late as Savage's (1930, 1931) papers on the Devonian of Kentucky, there was a tendency to place all specimens of Pentagonia in P. unisulcata and to regard P. peersii as a synonym of Conrad's species. In 1862, Hall discussed the variations within this group of species and proposed two varietal names for the forms that differ from Conrad's unisul- cata of the New York Onondaga. Hall suggested that the "Hamilton form" be called Meristella? unisulcata var. biplicata and that the "western form" be desig- nated M? unisulcata var. uniplicata. By implication, the typical form would have been called M? unisulcata var. unisulcata. I have been able to confirm the essential conspecifity of the Hamilton (i.e., Centerfield) and the western (i.e., Beechwood) forms. These are, in fact, included in Cozzen's Pentagonia peersii. The residue of forms left in Hall's original concept of the species thus be- comes P. unisulcata (Conrad). This species has been well illustrated by Hall (1862,1867). Genus Meristella Hall, 1859 1859. Meristella Hall, p. 78; 1860, p. 74; 1867, p. 295.— S. A. Miller, 1889, p. 353.—Hall and Clarke, 1889, p. 73.—Schubert and LeVene, 1927, p. 82.— Cooper, 1944, p. 331.—Amsden, 1958, p. 128.— Boucot, Johnson, and Staton, 1964, p. 820.—-Bou- cot and others, in Moore, 1965, p. H656. DIAGNOSIS.—"Unequally biconvex shells, commonly longer than wide; interarea obscure; ventral beak strongly incurved at maturity, commonly concealing foramen; deltidial plates may be exposed in early growth stages; dorsal fold and ventral sulcus may oc- cur, or sulcation may affect only anterior commissure, or valves may be nonsulcate. Dental plates obsolescent; ventral muscle scar flaring strongly laterally, commonly deeply impressed into secondary shell material; cardi- nal plate strong, variable, triangular to subquadrate in outline; commonly concave on upper surface and de- pressed to form broad septalium; median septum originating beneath cardinal plate and extending part way to anterior margin; jugum produced backward as stem bifurcates and recurves dorsally, then ante- riorly to reunite with stem." (Boucot and others, in Moore, 1965, p. H656). TYPE-SPECIES.—Atrypa laevis Vanuxem, 1842, p. 120; by subsequent designation, S. A. Miller, 1889, p. 354. DISCUSSION.—I accept the diagnosis given in Trea- tise on Invertebrate Paleontology (Moore, 1965, p. H656) along with the discussion of the genus and its relationship to Meristina by Boucot, Johnson, and Staton (1964, pp. 820, 821). The synonymy listed above is abbreviated: Hall and Clarke (1894) give a full synonymy up to 1889, so only the pertinent, more recent references are included. Meristella lentiformis Clarke, 1900 PLATE 1: FIGURES 1-4 1900. Meristella lentiformis Clarke, 1900, pp. 44, 45, pi. 6, figs. 5-11.—Van Ingen and Clark, 1903, pp. 1203, 1208.—Schuchert and others, 1913, pp. 126-129.— Stauffer, 1915, pp. 60-66.—Dunbar, 1919, pp. 99, 101.—Woodward, 1940, p. 147.—Boucot and John- son, 1967, pp. 80, 81. ORIGINAL DESCRIPTION.—"Shell unequally convex; outline transversely oval; pedicle valve with short in- curved beak and narrow cardinal slopes bordered by obtuse cardinal ridges diverging from beak. Umbo slightly convex or flat, the surface sloping to the sides very gradually. A median sinus starts at the umbo and rapidly broadens, producing a general depression in the pallial region, which is rather sharply deflected. The sinus is produced into a linguate extension at the anterior margin. The general flatness of this valve is more marked in young shells, the anterior deflection becoming prominent with the increase of age. The brachial valve has a full beak curved into the delthy- rium of the opposite valve, and is elevated medially into a broad ridge-like concavity terminating in a fold on the anterior margin. The lateral slopes are gently concave. The surface of both valves is smooth or bears only concentric growth lines. "On the interior the apophyses and impressions are those characterizing the genus. Particularly well de- veloped is the broad, flabellate muscular scar of the pedicle-valve and the median septum of the brachial valve, extending for more than half the length of the shell." (Clarke, 1900, pp. 44, 45.) EMENDED DESCRIPTION.—Shell unequally biconvex, generally wider than long, greatest width about at midlength, pedicle valve flattened with broad sulcus; NXTMBER 3 189 PLATE 1: figures 1-4: Meristella lentiformis Clarke. 1, 3, External and internal views of a brachial valve, USNM 163805 (X 2) ; 2, 4, external and internal views of a pedicle valve, USNM 163806 (X 2). Figures 5-8: Pentagonia lenta (Hall). Lateral, posterior, brachial, and pedicle views of an internal mold, USNM 28038 (X 3). Figures 9-12: Pentagonia peersii Cozzens. 9, 11, External and internal views of a brachial valve, USNM 163807 (X 2); 10, 12, external and internal views of a pedicle valve, USNM 163808 ( X 2). 190 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY PLATE 2: figures 1-3, 5-12.—Pentagonia peersii Cozzens. 1-3, 10, Posterior, brachial, pedicle, and lateral views of an average specimen, USNM 163809 (X 2). 5, 8, 9, Pedicle valves with two circular borings, USNM 163810; barnacle(?) borings, USNM 163811; and sponge (?) borings, USNM 163812 (all views same size). 6, 7, Posterior and anterior views ( X 2) of brachial valve showing details of cardinal process and septum, USNM 163807. 11, 12, Lateral and brachial views ( X 2) of shell with Aulopora in growth position, USNM 163813. Figure 4.—Pentagonia lenta (Hall). Latex cast of specimen USNM 28038 ( X 5) showing dental plates, cardinal process, and median septum. NUMBER 3 191 shell fibrous, impunctate; external ornament consisting only of concentric growth lines. Pedicle valve with broad, indistinct sulcus; interior with flabellate striate muscle area, commonly deeply impressed; short dental plates with prominent teeth; open delthyrial area, pedicle opening as a notch at the posterior end of the delthyrium. Brachial valve strongly arched with high fold ex- tending full length of the valve; interior with promi- nent cardinal process having a flat hingeplate; lateral sockets accommodate the teeth of the pedicle valve; narrow, sharp medium septum extends about halfway to anterior margin. DISCUSSION.—Internal characters of this species are transitional between Meristella and Pentagonia. The cardinal process is rather variable, in some specimens being very similar to the process in Pentagonia. Ex- ternally, the shell is a Meristella, although the broad, flattened pedicle valve gives a hint of the exaggerated form of the stratigraphically younger Pentagonia. Lateral carinae are not developed in M. lentiformis, nor is there any indication of the internal pouches of Pentagonia. Clarke (1900, p. 44) discussed the similarity of this species to Pentagonia lenta, but indicated clearly the differences that set the two species apart. The species are certainly very close, suggesting that the slightly younger P. lenta probably evolved out of the Meristella lentiformis stock. ILLUSTRATED SPECIMENS.—USNM 163805-6, from Glenerie Formation, on New York highway 9W, 1 mile south of Glenerie, New York, and 1 mile north of Cockburn, New York. Collector: B. Zimm. USNM Ace. No. 167820. Literature Cited Ager, D. V. 1961. The Epifauna of a Devonian Spiriferid. Quarterly Journal of the Geological Society of London, 117: 1-10, 1 plate. 1967. Brachiopod Palaeoecology. Earth Science Reviews, 3(3):157-179. Amsden, T. W. 1958. Stratigraphy and Paleontology of the Hunton Group in the Arbuckle Mountain Region, Part II, Haragan Articulate Brachiopods. Oklahoma Geo- logical Survey Bulletin, 78:1—144, plates 1—14, figures 1-38. Beecher, C. E. 1892. Notice of a New Lower Oriskany Fauna in Co- lumbia .County, New York, with an Annotated List of Fossils by J. M. Clarke. American Journal of Science, series 3, 44:410-414. Boucot, A. J. 1959. Brachiopods of the Lower Devonian Rocks at High- land Mills, New York. Journal of Paleontology, 33(5) :727-769, plates 90-103. Boucot, A. J., and others 1964. Reconnaissance Bedrock Geology of the Presque Isle Quadrangle, Maine. Maine Geological Survey Quadrangle Mapping Series, 2:1-123, 4 plates, 3 figures, 7 tables. Boucot, A. J.; J. G. Johnson; and R. D. Staton 1964. On Some Atrypoid, Retzioid, and Athyridoid Brachiopoda. Journal of Paleontology, 38(5) :805- 822, plates 125-128, 6 text figures. Boucot, A. J., and J. G. Johnson 1967. Paleogeography and Correlation of Appalachian Province Lower Devonian Sedimentary Rocks. Tulsa Geological Society Digest, 35 :35-87, 2 plates, 5 figures, 5 tables. 1968. Brachiopods of the Bois Blanc Formation in New York. United States Geological Survey Professional Paper, 584-B:l-27, 8 plates, 2 figures, 2 tables. Butts, C. 1940. Geology of the Appalachian Valley in Virginia, Part I. Geologic Text and Illustrations. Virginia Geological Survey Bulletin, 52: 1-568, plates 1-63, 10 figures. 1941. Geology of the Appalachian Valley in Virginia, Part II. Fossils, Plates and Explanations. Virginia Geological Survey Bulletin, 52: 1-271, pis. 64-135. caster, K. E. 1939. A Devonian Fauna from Columbia. Bulletins of American Paleontology, 24(83): 1-218, 14 plates. Clarke, J. M. 1900. The Oriskany Fauna of Becraft Mountain, Colum- bia County, New York. New York State Museum Memoir, 3(3): 1-128, 9 plates. Conrad, T. A. 1841. Descriptions of New Genera and Species of Or- ganic Remains. New York Geological Survey, Fifth Annual Report on the Paleontology of the State of New York, pages 25-57. Cooper, G. A. 1936. Studies of Middle Devonian Rocks in the Mid- West. Smithsonian Institution Exploration and Field-Work in 1935, pages 9—12, figure 8. 1944. Phylum Brachiopoda. In H. W. Shimer and R. R. Shrock, Index Fossils of North America, pages 277— 365, plates 105-143. New York: John Wiley and Sons. Cooper, G. A., and others 1942. Correlation of the Devonian Formations of North America. Geological Society of America Bulletin 53: 1729-1794, 1 plate, 1 figure. Cozzens, I. 1846. Descriptions of Three New Fossils from the Falls of the Ohio. New York Lyceum of Natural History Annals, 4: 157-159. 192 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY Dunbar, CO. 1919. Stratigraphy and Correlation of the Devonian of Western Tennessee. Tennessee State Geological Survey Bulletin, 21: 1-127. Grant, R. E. 1965. The Brachiopod Superfamily Stenoscismatacea. Smithsonian Miscellaneous Collections, 128(2) : 1— 192, 24 plates, 34 text figures. Goldring, W. 1935. Geology of the Berne Quadrangle. New York State Museum Bulletin, 303: 1-238, 72 figures, map. Hall, J. 1859. Contributions to the Palaeontology of New York. New York State Cabinet of Natural History, 12th Annual Report, 110 pages, illustrated. 1860. Contributions to Palaeontology. New York State Cabinet of Natural History, 13th Annual Report, pages 55-125, illustrated. 1861. Descriptions of New Species of Fossils from the Upper Helderberg, Hamilton and Chemung Groups; with Observations upon Previously De- scribed Species. New York State Cabinet of Natu- ral History, 14th Annual Report, pages 99-109. 1862. Contributions to Paleontology. New York State Cabinet of Natural History, 15th Annual Report, page 158, plate 2. 1S6 7. Descriptions and Figures of the Fossil Brachiopoda of the Upper Helderberg, Hamilton, Portage and Chemung Groups. New York Geological Survey, Paleontology, 4(1): 1-428, 63 plates. Hall, J., and J. M. Clarke 1894. An Introduction to the Study of the Genera of Paleozoic Brachiopoda, Part II. New York Geolog- ical Survey, Paleontology, 8: i-xvi, 1-394, plates 21-84. Howell, B. F. 1942. New Localities for Fossils in the Devonian Esopus Grit of Ulster County, New York. New York State Museum Bulletin, 327: 87-92, figure 15. Liddle, R. A.; G. D. Harris; and J. W. Wells 1943. The Rio Cachiri Section in the Sierra de Perija, Venezuela. Bulletins of American Paleontology, 27 (108): 1-100, 10 plates. Miller, S. A. 1889. North American Geology and Paleontology. 718 pages, 1194 figures. Cincinnati, Ohio. Moore, R. C. (editor) 1965. Treatise on Invertebrate Paleontology, Part H, Brachiopoda. 927 pages, 746 figures. New York City and Lawrence, Kansas: Geological Society of America and University of Kansas Press. Nettleroth, H. 1889. Kentucky Fossil Shells, a Monograph of the Fossil Shells of the Silurian and Devonian Rocks of Ken- tucky. Kentucky Geological Survey, 245 pages, 36 plates. Oliver, W. A., Jr. 1954. Stratigraphy of the Onondaga Limestone (Devo- nian) in Central New York. Geological Society of America Bulletin, 65(7) : 621-652. 1956. Stratigraphy of the Onondaga Limestone in East- ern New York. Geological Society of America Bul- letin, 67 (11): 1441-1474. 1967. Stratigraphy of the Bois Blanc Formation in New York. United States Geological Survey Profes- sional Paper, 584-A: 1-8, 1 plate, 5 figures. Oliver, W. A., Jr., and others 1969. Correlation of Devonian Rock Units in the Appa- lachian Region. United States Geological Survey Oil and Gas Investigations Chart, OC—64. Rickard, L. V. 1964. Correlation of the Devonian Rocks in New York State: New York State Museum and Science Serv- ice Geological Survey Map and Chart Series No. 4, chart with accompanying descriptive text. Savage, T. E. 1930. The Devonian Rocks of Kentucky. Kentucky Geo- logical Survey, series 6, 33: 1-161, 52 figures. 1931. Devonian Fauna, in The Paleontology of Ken- tucky. Kentucky Geological Survey, series 6, 36: 217-246, plates 27-32. Schuchert, C, and others 1913. Lower Devonian. Maryland Geological Survey, 560 pages, 98 plates. Schuchert, C, and C. M. LeVene 1929. Fossilium Catalogus I: Animalia, Part 42: Brachi- opoda. 140 pages. Berlin: W. Junk. Schumann, D. 1967. Die Liebenweise von Mucrospirifer Grabau, 1931 (Brachiopoda). Palaeogeography, Palaeoclimatol- ogy, Palaeocology, 3: 381-392. Stauffer, C. R. 1909. The Middle Devonian of Ohio. Ohio Geological Survey Bulletin, series 4, 10: 1-204, 17 plates. 1915. The Devonian of Southwestern Ontario. Canada Geological Survey Memoir, 34: 1-341, 20 plates. Stumm, E. C. 1942. Fauna and Stratigraphic Relations of the Prout Limestone and Plum Brook Shale of Northern Ohio. Journal of Paleontology, 16(5): 549-563, plates 80-84. Tomlinson, J. T. 1963. Acrothoracan Barnacles in Paleozoic Myalinids. Journal of Paleontology, 37(1) : 164-166, plate 23, 1 text figure. Van Ingen, G., and P. E. Clark 1903. Disturbed Fossiliferous Rocks in the Vicinity of Rondout, New York. New York State Museum Bulletin, 69:1176-1227, map. Vanuxem, L. 1842. Geology of New York. Part III, Comprising the Survey of the Third Geological District. 306 pages, illustrated. Albany, New York. Williams, H. S., and C. L. Breger 1916. The Fauna of the Chapman Sandstone of Maine. United States Geological Survey Professional Paper, 89: 1-347, 27 plates. Woodward, H. P. 1940. Devonian System of West Virginia. West Virginia Geological Survey, 15: 1-655, 69 plates. William A. Oliver, Jr. The Coral Fauna and Age of the Famine Limestone in Quebec ABSTRACT Rugose corals belonging to species of Cylindrophyllum, Heliophyllum, Aulacophyllum, Heterophrentis, Si- phonophrentis and Cystiphylloides are present in the limestone part of the Famine Formation in southeast- ern Quebec. The corals indicate a Middle Devonian, probably late Onondaga or Hamilton, age for the limestone. Cylindrophyllum stummi, new species, is formally described and specimens of eight other rugose coral species are illustrated and briefly described and dis- cussed. Representatives of four tabulate coral species also were found in the Famine Formation. The Famine Formation in southeastern Quebec is a small, isolated mass of Devonian sedimentary rocks surrounded by Ordovician clastic and volcanic rocks, the whole tightly folded and cleaved by post-Famine structural activity. A limestone unit in the middle part of the Famine is very fossiliferous and has been recog- nized as Devonian since it first was studied over 100 years ago. Kindle (in MacKay, 1921) and Clark (1923) considered the Famine Formation to be Onon- daga in age, but this assignment has never been sup- ported by published descriptions or illustrations of key fossils; also, an Onondaga age has seemed too young to some workers because of paleogeographic considera- tions. Recent restudy of the corals and other fossil groups has indicated, however, that the Famine is William A. Oliver, Jr., United States Geological Survey, Room E—305, United States National Museum, Washington, D.C. 20242. Publication authorized by the director, United States Geological Survey. Middle Devonian, Onondaga or younger, in age and that paleogeographic prejudices must be set aside. The present paper reviews previous work, illustrates and de- scribes the more important rugose corals, and discusses the evidence of these forms. I am indebted to A. J. Boucot, Oregon State Uni- versity, and to N. R. Gadd and L. M. Cumming, Ge- ological Survey of Canada, for collecting and sending to me the corals on which this study is based. All thin sections were prepared by W. C. Pinckney, Jr. Photographs are by D. H. Massie. Previous Work Logan (1863, p. 428) noted "dark fossiliferous lime- stones . . . which form a low short ridge, overlooking the Famine [River], in eastern Quebec." He listed sev- eral corals (Table 1, herein) and brachiopods from the limestones and noted that the fossils "have a Devonian aspect." Logan's list was based on identifications by Elkanah Billings, as noted by Ells (1888, p. 9). Ells (1888, pp. 9-11) mentioned the same limestone and discussed its distribution in somewhat greater de- tail. In addition, he provided a slightly different list of corals (Table 1), brachiopods, pelecypods, and tri- lobites as identified by H. M. Ami. The Famine "Series" was named and described by MacKay (1921, pp. 31-33) from the small area south of the Famine River near St. George, Quebec, that is referred to in the above reports. The formation includes a fossiliferous limestone "about 40 feet thick" (Mac- Kay, 1921), an underlying basal conglomerate (50 feet thick), and overlying shales and limestones (great- er than 120 feet thick) (L. M. Cumming, personal communication, 1968). MacKay considered all the 193 194 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY beds as probably of Onondaga age on the basis of fos- sils from the middle unit determined by E. M. Kindle (MacKay, 1921, p. 32). Kindle's list includes two rugose and four tabulate corals (Table 1), several brachiopods, and trilobites. TABLE 1.—Previous identifications of corals from the lime- stone of the Famine Formation. (A), Billings, in Logan 1863; (B), Ami, in Ells, 1888; (C), Kindle, in MacKay, 1921; (D), Clark, 1923. Numbers in paren- theses following species names indicate probable corre- spondence with the new identifications in Table 2. RUGOSE CORALS: Diphyphyllum arundinaccum {1) Diphyphyllum (1) Cyathophyllum? (2) Cyathophyllum sp. (2) Heliophyllum oneidaense (2?) Heliophyllum sp. (2) ^aphrentis sp. (5 to 8) Amplexus sp. cf. A. hamiltoniae (none) Cystiphyllum vesiculosum (9) TABULATE CORALS: Alveolites sp. (none) Favosites gothlandica{us) (11) F. basaltica (11?) Favosites sp. (11) Favosites sp. cf. F. limitaris (12) Syringopora hisingeri (13) S. tabulata (13) (A) (B) (C) (D) X X X X X — X — — — X x — — — x — X x — — — — — X — — — X X — X X — — X — X X — X — — — — X — — X X — — — X Onondaga age," stating that it "contains the corals Siphonophrentis and Heliophyllum." Corals and Age New identifications of corals from the limestone in the Famine Formation, based on collections made by A. J. Boucot (then with the United States Geological Survey) in 1955 and by N. R. Gadd (Geological Sur- vey of Canada) in 1965 are listed in Table 2. The listed species are numbered, and these numbers are inserted parenthetically in Table 1 to indicate probable cor- respondence between these species and the earlier identifications. The remarkable similarity of the new and old lists suggests that the collections are repre- sentative, although relatively small and containing few species. The specimens listed in Table 2 are now in the collections of the United States National Mu- seum (USNM) and the Geological Survey of Canada (GSC). Cylindrophyllum stummi, new species, is probably the Diphyphyllum of the earlier lists, because it is com- mon and is the only phaceloid rugose coral in the assemblage. TABLE 2.—Corals identified from the Famine Limestone, St. George, Quebec. Numbers of Specimens in collections made by A. J. Boucot in 1955 (USNM) and N. R. Gadd in 1965 (GSC) are listed. Asterisk indicates apparently common forms. Clark (1923) presented additional information on the limestone and an annotated list of fossils. "The limestone is dark gray, almost black; and is crowded with fossils wherever it outcrops . . . upon weathering breaks down into a rubbly mass. . . ." (Clark, 1923, p. 218). Clark annotated several species of corals (Ta- ble 1) and brachiopods and noted the presence of bryozoans and gastropods. He considered the corals to indicate an Onondaga age. Boucot and Johnson (1967, p. 54) mentioned the Famine Limestone as of "Middle Devonian and prob- ably Eifelian (Onondaga) age" on the basis of a spe- cies of Brevispirifer. An "Athyris of the A. undata- type and a pholidostrophiid of the Teichostrophia- type" are also present and suggest Old World faunal affinities (Boucot and Johnson, 1967, p. 54). Cumming (1968, p. 1046) briefly mentioned the Famine Formation as "an isolated outlier of probably RUGOSE CORALS: *1. Cylindrophyllum stummi, new species *2. Heliophyllum halli E. and H. *3. Heliophyllum sp. cf. H. pro- liferum Hall 4. Aulacophyllum sp. 5. Heterophrentis sp. 1 6. Heterophrentis sp. 2 7. Siphonophrentis sp. cf. S. yandelli (E. and H.) 8. Siphonophrentis sp. 2 9. Cystiphylloides sp. TABULATE CORALS: 10. Cladopora sp. 11. Favosites sp. 1 12. Favosites sp. 2 *13. Syringopora sp. 1 USNM GSC Total 10 6 17 8 7 1 2? 1 3 1 2 NUMBER 3 195 It is surprising that the earlier workers made so little of the excellently preserved and common Helio- phyllum halli, though this is presumably the Helio- phyllum sp. and Cyathophyllum sp. of tiieir lists. Logan's list (by Elkanah Billings) does include Helio- phyllum oneidaense, but this must have been intended to refer to the form now known as Acrophyllum onei- daense (Billings), which has quite different internal structure. However, A. oneidaense is limited to rocks of Schoharie-Bois Blanc age (Emsian; Oliver, 1967) and is unlikely to occur in association with the other Famine corals. Possibly the Billings identification is based on Heliophyllum halli, even though this would seem an unlikely error. Zaphrentis on the earlier lists can refer to either Heterophrentis or Siphonophrentis, but Clark's descrip- tion is certainly of a Heterophrentis as presently under- stood. Clark's Cystiphyllum vesiculosum is on my list as Cystiphylloides sp., but I have nothing tiiat can be compared to his Amplexus sp. cf. A. hamiltoniae. Similarly, the tabulate corals on the earlier lists corre- spond to my identifications. The only significant ex- ceptions are Logan's two species of Favosites, which imply greater numbers and diversity of specimens than are present in the new collections, and Kindle's Alveo- lites, a genus that is not present in the new collections. The corals listed in Table 2 indicate a Middle De- vonian (Eifelian-Givetian) age for the fossiliferous part of the Famine Formation. In a following section, the significance of each species is discussed separately, but die results of this analysis can be summarized here. The apparent abundance and variety of Helio- phyllum and the presence of Onondaga and Hamilton- like species of Heterophrentis, Siphonophrentis, and Cystiphylloides indicate that the Famine faunule be- longs to the eastern North American, Middle Devonian coral assemblage of Oliver (1968). These forms col- lectively preclude a pre-Onondaga assignment for the Famine Limestone. An early Onondaga (Edgecliff and equivalents) coral assemblage is very widespread in eastern North America (Oliver, 1968, p. 741 and table 4, A,B,F,G) . The lack of characteristic elements of this assemblage in the Famine Formation suggests that the Famine is post- Edgecliff in age. However, the corals that are present do not indicate a closer correlation of the Famine with the standard New York section than Onondaga to Hamilton. Boucot and Johnson (1967, p. 54) suggest that the Famine brachiopods are a mixture of European and eastern North American types. The corals, on the con- trary, are distinctly of eastern North American aspect and most of them probably belong to species already described from New York or Ontario. Cylindrophyllum stummi, new species, is an exception on the species level, but the genus is basically an eastern North American element and is rare or unknown in Europe. Conodonts from the same collections as the corals have been studied by Gilbert Klapper (University of Iowa) and T. T. Uyeno (Geological Survey of Can- ada) . They both found small, poorly preserved cono- dont faunas composed exclusively of members of the Icriodus nodosus group, indicative of Middle Devonian age. According to Klapper (personal communication, 1968) they "are comparable, in terms of evolution within Icriodus in the Middle Devonian, to specimens of Icriodus corniger from the Dundee Limestone of northwestern Ohio." Locality All collections recorded in the literature or discussed herein have come from a general area just north of St. George, Quebec, east of the Chaudiere River but south of the Famine River. St. George is approximately 60 miles southeast of Quebec City. MacKay (1921, geo- logic map no. 1835) shows his collecting locality as 0.6 mile (1 km) north of the east end of the bridge in St. George. Limestone outcrops extend eastward from this point (see MacKay map), and Clark's collections may have come from some of the other exposures as well as from MacKay's locality. The two collections that serve as a basis for this paper are from the MacKay locality. Systematic Description and Annotations Cylindrophyllum stummi, new species PLATE 1: FIGURES 1-6; PLATE 2 : FIGURES 4, 5. DIAGNOSIS.—Phaceloid coralla with cylindrical corallites commonly in lateral contact with each other so as to form short "chains." Septa short, wedge-shaped, radially arranged; minor septa nearly equal to major septa in length; septal carinae present or very common. Dissepimentarium narrow with one to four rows of nor- mal, globose dissepiments. Tabulae complete or incomplete, variable in form. 372-386 0—71- -14 196 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY MATERIAL.—Seventeen incomplete coralla, including the holotype, two illustrated paratypes, and fourteen unillustrated paratypes. Preservation is mediocre; parts of all coralla are distorted and slightly crushed. EXTERNAL FEATURES.—Phaceloid (partly cateni- form) colonies are as much as 17 cm or more in height and 30 cm or more in diameter; they are composed of cylindrical corallites 6 to 11 mm in diameter. Coral- lites are essentially parallel in astogenetically advanced parts of the corallum but are radiating in early stages. Corallite spacing was apparently less than one diam- eter, and many corallites grew in lateral contact with one or two neighboring corallites to form short "chains" (Plate 1: figure 3). Average or maximum distance between corallites is unknown, as all specimens are from a structurally disturbed area, and some crushing and distortion is evident in parts of most coralla. Off- setting is lateral. INTERNAL FEATURES.—The major septa are radi- ally arranged and very short, commonly extending from one-fifth to one-half the distance to the axis; minor septa are nearly as long. Major septa number 19 to 27 in cylindrical parts of studied corallites. Septa are wedge-shaped, somewhat thickened peripherally, and composed of trabeculae, the axes of which are confined to a single plane; trabecular fibers diverge at a high angle in the axial part of the septa, but the angle be- comes less peripherally and the septa seem to merge imperceptably into the fibrous wall. Zig-zag carinae are inconspicuous in most transverse sections but are prom- inent in parts of most longitudinal sections. Locally, septa expand at their axial ends; in such places the trabecular structure of the septa is particularly clear (Plate 2: figure 5). The dissepimentarium is narrow, commonly occupy- ing less than three-tenths the corallite radius. Dissepi- ments are normal, commonly globose, and arranged in one to four rows. Inner dissepiments where present are as large or larger than peripheral ones. Size and shape variation is considerable, as can be seen from the illustrations. The tabularium is wide and composed of complete tabulae supplemented by incomplete peripheral tabu- lae. Tabula shape varies from nearly flat to irregularly, but gently, convex or concave. TYPES.—Holotype, USNM 163236; illustrated para- types, GSC 24639 and USNM 163237; unillustrated paratypes, GSC 24640, 24797, 24798, and three others; USNM 13238, 163396-8, and four others. DISCUSSION.—Cylindrophyllum stummi is charac- terized by its very short, relatively thick septa. The type and other known species of the genus have longer, very attenuate septa and more prominent septal carinae. In transverse section C. stummi is similar to Acinophyllum crassiseptatum (Ehlers and Stumm 1949, pp. 28-29, pi. 6, figs. 1-6), but A. cr assise ptatum differs in having a peripheral row of dissepiments that are larger than any inner dissepiments and in having very coarse sep- tal trabeculae. Acinophyllum rectiseptatum Rominger (Stumm, 1955, card 269) is much smaller in diameter and has fewer rows of dissepiments. DISTRIBUTION.—Cylindrophyllum stummi is known only from the Famine Limestone, St. George, Quebec, at MacKay's (1921) locality (see locality section). Middle Devonian; Eifelian? Heliophyllum halli Edwards and Haime, 1850 PLATE 3 : FIGURES 1, 7, 9 Heliophyllum halli E. and H. Wells, 1937, pp. 9-18, pi. 1, figs. 1-15. DISCUSSION.—Specimens of this species are in both the USNM and GSC collections (see Table 2). They are typical in showing wide variation in growth form, in having very attenuate septa and long cross-bar carinae in mature stages, and in having dilated septa in early stages. In New York and Ontario the species is very com- mon from the base of the Onondaga Limestone (Edge- cliff Member) through the Tully Limestone. The genus, and possibly the species, occurs in the older Bois Blanc Formation (Emsian; Oliver, 1967) and die younger West Falls Formation (Frasnian) but is not common. I consider abundant H. halli to indicate a Middle Devonian age in eastern North America. MATERIAL.—USNM 163239, the illustrated speci- men, and seven unillustrated specimens. PLATE 1.—Cylindrophyllum stummi, new species. 1—3, Holo- type corallum, USNM 163236: 1, 3, transverse thin sections ( X 5 and X 1.5, respectively) ; 2, longitudinal section ( X 5). 4-6, Paratype corallum, USNM 163237: 4, longitudinal sec- tion (X 5) ; 5, 6, transverse sections (X 2.5 and X 5, respec- tively) . NTJMBER 3 197 at™ PLATE 1 198 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY Heliophyllum sp. cf. H. proliferum Hall, 1877 PLATE 3: FIGURES 3, 4. Cf. Heliophyllum proliferum Hall, 1877, pi. 26, figs. 1-2. Not Heliophyllum proliferum Nicholson, 1874, p. 27—28-. DISCUSSION.—Seven fragments of Heliophyllum colonies show peripheral offsetting and are morpho- logically similar to Heliophyllum proliferum Hall (not Nicholson) from the Hamilton Group in New York. The largest fragment, 15 cm in diameter, includes 20 or more corallites. Individual corallites differ from the associated H. halli, only in showing greater variability in septal cari- nae. In the colonial form the carinae are zig-zag or cross-bar and very long to almost absent. Hall (1877) illustrated H. proliferum as a new spe- cies but the name had already been introduced for an Onondaga species by Nicholson (1874). Wells (1937, pp. 14, 15) considered Hall's species to be a form of H. halli. I agree that individuals of H. halli produced peripheral offsets, but consider it unlikely that the large, bushy colonies found in the higher parts of the Hamilton and represented here were no more than variants of the solitary species. Detailed study might show that the Onondaga-Hamilton sequence contains a record of the evolution of H. proliferurn-like species from H. halli twice—an earlier Famine species and the later upper Hamilton species of Hall. Hall's name is a junior homonym but it would be inappropriate to re- name the species until a detailed stratigraphic and morphologic study of the New York material has been completed. MATERIAL.—GSC 24641 the illustrated specimen, and six unillustrated specimens. Aulacophyllum sp. PLATE 2: FIGURES 2, 3. DISCUSSION.—One specimen has the characters of this genus, but it apparently differs from described species in having zig-zag carinae developed on the peripheral parts of many septa. Without more material and detailed comparison with named species, identi- fication is impractical. In North America, specimens of Aulacophyllum species are common in rocks of Onondaga and Hamil- ton age but extend downward into the Bois Blanc For- mation and equivalents (Emsian). MATERIAL.—USNM 163242. Heterophrentis spp. PLATE 2: FIGURES 1, 8; PLATE 3: FIGURES 6, 8 DISCUSSION.—A fragment of a horn coral, USNM 163243 (Plate 3: figures 6, 8), is similar to Hetero- phrentis simplex Hall, 1843 (Hall, 1877, pi. 21, figs. 5-11), in general morphology but shows much more septal dilation than is common in that species. Hetero- phrentis ferronensis and H. curviseptata (both Stumm, 1962b) show greater dilation than does H. simplex but they differ from the Famine form in tabula shape. A second, more complete horn coral, GSC 24799 (Plate 2: figures 1, 8), also is referable to Heterophren- tis but probably to a different species. This specimen has attenuate septa extending four-fifths the distance to the axis. The degree of individual variation in these corals or in described species of Heterophrentis is not known, and identification of the Famine specimens is not prac- tical. The genus, as presently understood, may range both above and below the Middle Devonian in eastern North America. Siphonophrentis sp. cf. S. yandelli (Edwards and Haime) PLATE 2: FIGURES 9, 10 DISCUSSION.—Three specimens represent a species of small Siphonophrentis 18-20 mm in diameter and with approximately 32 to 36 major septa. The Famine specimens are quite similar to S. yandelli (E. and H.) from the Jeffersonville Limestone at the Falls of the Ohio and the Onondaga Limestone in New York. In New York the species ranges from the Edgecliff through Moorehouse Members but is most common in the Edgecliff. MATERIAL.—GSC 24642, the illustrated specimen, and two unillustrated specimens. PLATE 2: figures 1, 8.—Heterophrentis sp. 2. Longitudinal and transverse thin sections (X 1-5); GSC 24799. Figures 2, 3.—Aulacophyllum sp. Longitudinal and trans- verse thin sections (X 1.5); USNM 163242. Figures 4, 5.—Cylindrophyllum stummi, new species. Para- type corallum, GSC 24639; transverse thin sections (X 5 and X 50, respectively). Figures 6, 7.—Cystiphylloides sp. Longitudinal and trans- verse thin sections (X 1.5) ; USNM 163245. Figures 9, 10.—Siphonophrentis sp. cf. S. yandelli (E. and H.). Transverse and longitudinal thin sections (X 1.5); GSC 24642. NUMBER 3 199 200 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY PLATE 3 NUMBER 3 201 Siphonophrentis sp. 2 PLATE 3: FIGURES 2, 5 DISCUSSION.—A fragment of a specimen apparently referable to this genus is in the USNM collection. Septa are short, and a prominent fossula is marked by a short and attenuate cardinal septum. Tabulae are complete, concave axially, and strongly arched periph- erally. The fragment is too incomplete for species-level identification, but similar forms occur in both the Onondaga Limestone (S. gigantea) and the Hamilton Group (S. halli). MATERIAL.—USNM 163244, the illustrated speci- men. Cystiphylloides sp. PLATE 2: FIGURES 6, 7 DISCUSSION.—Single specimens of this genus are in both the USNM and GSC collections. They re- semble the described specimen of C. alpenense Stumm (1962a, pp. 220-221, pi. 2, figs. 3, 4), but may be simply a variant of the common Onondaga-Hamilton C. americanum (=vesiculosum of some authors). The Famine specimen has zones of thickened dis- sepiments that appear as concentric rings in the trans- verse section and as broad V-shaped zones in the longitudinal section. The genus strongly suggests a Middle Devonian age. MATERIAL.—USNM 163245, the illustrated speci- men, and one unillustrated specimen. Literature Cited Boucot, A. J., and J. G. Johnson 1967. Paleogeography and Correlation of Appalachian Province Lower Devonian Sedimentary Rocks. Tulsa Geological Society Digest, 35: 35—87. Clark, T. H. 1923. The Devonian Limestone at St. George, Quebec. Journal of Geology, 31: 217-225. PLATE 3: figures 1, 7, 9.—Heliophyllum halli E. and H.; USNM 163239: 1, 9, transverse thin sections (X 1.5 and X 5, respectively) ; 7, longitudinal thin section (X 1.5). Figures 2, 5.—Siphonophrentis sp. 2. Longitudinal and transverse thin sections (X 1.5) ; USNM 163244. Figures 3, 4.—Heliophyllum sp. cf. H. proliferum Hall. Transverse and longitudinal thin sections ( X 1 and X 1-5, re- spectively) ; GSC 24641. Figures 6, 8.—Heterophrentis sp. 1. Longitudinal and transverse thin sections (X 1.5); USNM 163243. Cumming, L. M. 1968. Devonian of Canadian Appalachians and New England States. Alberta Society of Petroleum Geologists, International Symposium Devonian Sys- tem, 1967, Calgary, Canada, 1: 1041-1055. Ehlers, G. M., and E. C. Stumm 1949. Corals of the Devonian Traverse Group of Michi- gan, Part 2. Cylindrophyllum, Despasophyllum, Disphyllum, Eridophyllum, and Synaptophyllum. University of Michigan, Museum of Paleontology Contributions, 8(3) : 21-41. Ells, R. W. 1888. Second Report on the Geology of a Portion of the Province of Quebec. Geological Survey of Canada, Annual Report, 3(K) : 1-120. Hall, James 1877. Illustrations of Devonian Fossils; Corals of the Upper Helderberg and Hamilton Groups. New York Geological Survey, 7 pages, 136 plates. Logan, W. E., and others 1863. Geology of Canada. Geological Survey of Canada Report of Progress to 1863. 983 pages. MacKay, B. R. 1921. Beauceville Map-Area, Quebec. Geological Survey of Canada Memoir, 127: 1—105, 2 maps. Nicholson, H. A. 1874. Report upon the Paleontology of the Province of Ontario, 133 pages. Toronto: Hunter, Rose and Company. Oliver, W. A., Jr. 1967. Stratigraphy of the Bois Blanc Formation in New York. United States Geological Society Professional Paper, 584-A: 1-8. 1968. Succession of Rugose Coral Faunas in the Lower and Middle Devonian of Eastern North America. Alberta Society of Petroleum Geologists, Interna- tional Symposium Devonian System, 1967, Cal- gary, Canada, Proceedings, 2: 733-744. Stumm, E. C. 1955. Tetracoralla, Part C. Type Invertebrate Fossils of North America {Devonian). Division 1, unit IF, cards 128-337. Philadelphia: Wagner Free Insti- tute of Science. 1962a. Corals of the Traverse Group of Michigan, Part 7. The Digonophyllidae. University of Michigan, Museum of Paleontology Contributions, 17(9): 215-231. 1962b. Corals of the Traverse Group of Michigan, Part 8. Stereolasma and Heterophrentis. University of Michigan, Museum of Paleontology Contributions, 17(10): 233-240. Wells, J. W. 1937. Individual Variation in the Rugose Coral Species, Heliophyllum halli, E. and H. Pale onto graphic a Americana, 2(6) : 1-22. Paul Sartenaer Redescri ption of the Brachio pod Genus Tunnane lla Grabau, 1923 (Rhynch onellida ABSTRACT The genus Yunnanella Grabau, A. W., 1923, and its type-species Y. hanburii (Davidson, T., 1853), are redescribed. A considerable number of the forms that have been named "Yunnanella" actually belong to other genera. As a result, the stratigraphic position and the geographic distribution of the genera Yunnanella and Nayunnella Sartenaer, P., 1961, undergo drastic modifications, although information from China is not sufficient to permit definite conclusions. RESUME—Le genre Yunnanella Grabau, A. W., 1923 est redecrit ainsi que son espece-type Y. hanburii (Davidson, T., 1853). Une grande partie des formes denommees "Yunnanella" a un moment ou l'autre ap- partiennent en fait a. d'autres genres. II en resulte des modifications profondes dans la position stratigraphi- que et la distribution geographique des genres Yunna- nella et Nayunnella Sartenaer, P., 1961, encore que l'insuffisance des informations en provenance de la Chine n'autorise pas des conclusions definitives. "So little is known of Chinese fossils that every fresh discovery interests the palaeontologist." Is it not strange that 117 years after this sentence was written by T. Davidson (1853, p. 353) it has lost very little of its value? Although widely mentioned in the literature over the past 40 years, the genus Yunnanella Grabau, A. W., 1923, is still unsatisfactorily known as regards its sys- Paul Sartenaer, Institut Royal des Sciences Naturelles de Belgique, Rue Vautier, 31, Bruxelles 4, Belgium. tematic definition, stratigraphic range, and geographic distribution. Some very interesting papers published in the last decade, while making important new infor- mation available, have indirectly emphasized the need for a better understanding of the genus. I have been faced with the alternative of leaving the problems in which the genus Yunnanella is involved as they stand—an attitude suggested by the lack of precision of data—or of trying to solve them. The con- fusion has reached such a level, however, that the latter positions seems inescapable. The negative elements rendering it difficult to deal with these problems are: (1) because most of the material is Chinese and Russian, its availability is either nil or limited; (2) stratigraphic information on Chinese collections is imprecise, vague, and sometimes conflicting; (3) I have had no field experience in the various regions considered. On the other hand, there are several favorable cir- cumstances. First, the primary types of Y. hanburii (Davidson, T., 1853), the type-species of the genus Yunnanella, are deposited in the British Museum of Natural History (BM) in London, where I studied them in 1961 and in 1968. During my first visit, all facilities, including photography, were put at my dis- posal by Dr. H. M. Muir-Wood; during the second visit similar courtesies were extended by Dr. H. Brun- ton. Secondly, specimens of the type-species as well as of the species Y. triplicata Grabau, A. W., 1931, and Y. uniplicata Grabau, A. W., 1931 (two species I do not consider valid), were examined in the United States National Museum (USNM) in Washington, D.C., in 1966. Moreover, Dr. G. A. Cooper generously 203 204 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY sent this precious collection to Brussels for an extended period and allowed serial sections to be made of one specimen of the type-species. Thirdly, as a guest of the Russian Academy of Sciences, and under the auspices of the Belgo-Russian cultural agreement, I had the opportunity to visit some colleagues in Moscow and Leningrad in 1963-1964. Due to the kindness of Acad- emician D. V. Nalivkin, Drs. E. A. Ivanova, M. V. Martynova, M. A. Rzhonsnitzkaia, Kh. S. Rozman, T. V. Sverbilova, S. V. Tcherkesova, and P. N. Varfolomeev, not only could most of the Russian spe- cies relevant to this paper be observed and form the subject of fruitful discussions but specimens were do- nated and permission was granted for making latex molds of any primary type or figured specimen in the collections, including some Chinese specimens received by Russian scientists. Fourthly, Dr. Herta Schmidt kindly showed to me, in 1968, all the material on which her study of the genus Schnurella Schmidt, H, 1964, was based. Thus, in spite of many drawbacks, it is my opinion that I am in the best position for dealing with this intricate question with the hope of indicating the direction toward a solution and of stimulating controversy. Genus Yunnanella Grabau, A. W., 1923 NOMENCLATORIAL NOTE.—In two papers dealing with nomenclatorial problems, Sartenaer (1961a, 1962) remarked that the nominal genus Yunnanella has been valid since 1923, when it was erected by monotypy by Grabau, and that the generic name Yun- nanellina Grabau, A. W., 1931, is a junior objective synonym of Yunnanella. The second paper was written in answer to a publication by Ivanova (1961) oppos- ing the views expressed in the first paper. I introduced the nominal genus Nayunnella Sartenaer, P., 1961, with N. synplicata (Grabau, A. W., 1931) as the type- species. Consequently, in the present paper, Yunna- nella = A. W. Grabau's Yunnanella (before 1931) and Grabau's and some other authors' Yunnanellina, while Nayunnella = A. W. Gabau's (from 1931 on) and some other authors' Yunnanella. Some generic descriptions of the past have included characters of the genus Nayunnella considered as a subgenus, while others were not applicable to the genus Yunnanella because the species described did not belong to that genus. Most of the descriptions have concentrated chiefly on the ornament, however, and very little information has been given on internal characters. Therefore, it seems advisable to give a new complete description. DIAGNOSIS.—Small to medium size. Uniplicate. An- terolateral commissures low in the shell. Moderately deep to deep sulcus and high fold starting at some dis- tance from the beaks. Sulcus wide at front. Tongue high, trapezoidal. Suberect beak. Ventral interarea clearly delimited. Top of shell usually at the frontal commissure. Few wide and simple costae, generally restricted to the anterior part of the shell, strongly indenting the commissure. Parietal costae have been observed in one specimen only. Surface covered by divided, and sometimes intercalated, costellae. Values of shoulder angle: 94° to 110°. Deltidial plates strong. Dental plates and teeth stout and long. Ventral um- bonal cavities clearly separated. Septum stout and short. Septalium short, deep, wide, amphora-shaped, and uncovered. Dental sockets long and moderately deep. Crura long and slender. DESCRIPTION.—Shell small to medium size, unipli- cate, with a transversly subelliptical to subcircular and subpentagonal contour. Anterolateral commissures low in the shell, strongly indented by the costae, as is the frontal commissure, and seldom sharp on account of the verticality of the extreme margins of the flanks. Top of the shell usually at the frontal commissure, sometimes somewhat posterior to it. Width is the great- est dimension. Values of width and length sometimes approach each other. Height sometimes greater than length. Values of the shoulder and apical angles, re- spectively, between, 94° and 110°, and 105° and 120°. Few wide and simple costae, generally restricted to the anterior part of the shell. Median dorsal costae and internal ventral lateral costae are the only high costae. Parietal costae have been observed in one specimen only. Surface completely covered by costellae (3 to 8 per mm at midlength of the shell), which increase in number anteriorly by bifurcation or, exceptionally, by intercalation. Pedicle valve generally slightly convex in the um- bonal region with flat to slightly concave anterolateral parts; these two characters, together with the low position of the anterolateral commissures, explain why the umbonal region is the only part showing in lateral views. Flanks steep in their posterolateral parts, some- times concave near the commissure. Sulcus starting imperceptibly between 30 and 55 percent of the length of the shell forward of the beak, or between 25 and 40 percent of the unrolled length of the valve. Sulcus moderately deep to deep, usually rapidly widening; NUMBER 3 205 its width, where it starts, varies between 30 and 50 percent of its maximum width at the front, which varies between 60 and 75 percent of the width of the shell. Floor of the sulcus generally flat and sometimes slightly convex. High trapezoidal tongue with sharp borders. Upper part of the tongue tending to become vertical and even usually recurved posteriorly. Ventral median costae often protruding anteriorly beyond the sharp borders of the tongue. Greatest thickness located in the posterior quarter of the unrolled length of the valve. Very well-marked suberect beak. Interarea clearly delimited. Small half-circular foramen. Strong deltidial plates. Stout and long dental plates divergent posteriorly, becoming progressively parallel, then con- vergent anteriorly. Divergence and convergence are in relation to the median line of transverse serial sections. Clearly separated umbonal cavities. Teeth long and stout widi undulated brachial side. Brachial valve strongly convex, but never inflated. Flanks steep in their posterolateral parts, where they become concave near the commissure. Well-marked high fold starting imperceptibly at some distance for- ward of the beak. Top of the fold flat to slightly arched, exceptionally strongly arched. Stout and short septum supporting a short, deep, wide, and uncovered am- phora-shaped septalium. Dental sockets long and mod- erately deep. Outer hinge plates concave. Moderately thick and wide crural bases passing to long, slender crura, crescent shaped in transverse serial sections. TYPE-SPECIES.—Rhynchonella hanburii Davidson, 1853. COMPARISONS.—The strikingly different pattern of ornamentation of the genera Yunnanella and Nayun- nella makes it easy to separate them at first sight. In the genus Nayunnella, as emphasized by many scien- tists (Grabau, 1931a, p. 141, 1932, pp. 92, 94, 95; Likharev, 1934, p. 518; Tien, 1938, pp. 43, 44, 74; Roger, 1952, p. 90; Basic Invertebrate Fossils of China, 1957; Rozman, 1959, pp. 94, 95, 96, 98; 1962, pp. 130, 137; Rzhonsnitzkaia, in Rzhonsnitzkaia, Likharev, and Makridin, 1960, p. 243; McLaren, 1962, pp. 65, 66; in Moore, 1965, p. H586; Schmidt, 1965, p. 16), some costellae enlarge forward and become costae, others join in forming costae, and still others become obsolete before reaching the frontal and anterolateral commis- sures. Furthermore, the costellae undergo few divi- sions. In the material at my disposal, divisions have been observed only in the posterior third of the shell. In the genus Yunnanella the costellae not only are independent of the costae and cover the whole surface but they divide more often. I accept these differences, with the restriction that if costellae are present in the genus Yunnanella I am not ready to call costellae the narrow costae present in the genus Nayunnella. The term "costellae" is used here in a general way for des- ignating fine costae. The term "capillae" would be more appropriate but it is rejected because it has been strictly defined in another order of brachiopods. The unavailability of rich collections prevents the determining of the validity of many other differences. The material at hand, however, suggests that, com- monly, in the genus Yunnanella height is more im- portant, contour is transversely less elongated, costae are less restricted to the anterior part of the shell, sulcus and fold start nearer to the beaks, anterolateral margins of the flanks are abrupt, and the upper part of the tongue is recurved posteriorly. Insufficient material has prohibited the making of serial sections in species of the genus Nayunnella to investigate the internal characters. Although not precise enough, the serial sections found in the literature do not suggest major differences between the two genera. Finally, it must be noted that Grabau's (1931a, pp. 141, 142, 157) statement that Yunnanella was a subgenus of Nayunnella was later disregarded by Grabau (1932) himself. The differences between the genus Yunnanella and the Givetian genus Schnurella are important and numerous. Some of them were recognized as far back as 1853 by Davidson, who compared the type-species of both. The genus Schnurella differs in the contour; the smaller size; the steep frontal and lateral sides with typical excavations in the flanks reminiscent of the lunules in the pelecypods; the commissure not sticking out; the greatest thickness somewhat posterior to the frontal commissure; the l./w., t./w., t./l ratios fluctuating around one, with width seldom being the greatest dimension; the greater thickness; the smaller apical angle; the very shallow sulcus excavating the pedicle valve and occupying almost the whole width of the shell at the front; the nontrapezoidal tongue without sharp borders and not recurved posteriorly in its upper part; no clear ventral interarea; the very low fold with its top always convex; more median costae; lower costae more restricted anteriorly and flattened on the frontal and lateral sides; the fine costae (different from costellae) not reaching the commissure; and in most of the internal characters. 206 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Some of these differences have been indicated by Schmidt (1964, p. 506; 1965, p. 17; in Moore, 1965, p. H587). The lower Famennian genus Eoparaphorhynchus Sartenaer, P., 1961, is very easily separable, as noted by Sartenaer (1961b, p. 2). Yunnanella hanburii (Davidson, T., 1853) PLATE 1; FIGURES 1-5; Plate 2 NOMENCLATORIAL NOTE.—Conditions of Article 32 (a) of the International Code of Zoological Nomen- clature being fulfilled, Yunnanella hanburii is the only correct spelling, although it is quite clear, according to Davidson (1853, p. 353), that the person who pre- sented four or five specimens of that species to the British Museum of Natural History was D. Hanbury. Nevertheless, since the problem is of little importance, I have no strong objection to the spelling "Y. han- buryi" if Recommendation 31A of the second edition of the Code (= Article 31 of the first edition) is fol- lowed and given retroactive power. TYPES.—Lectotype, BM B42506; Province of Kwangsi, China; Devonian. The lectotype was pur- chased in a drug warehouse in Shanghai and was des- ignated by Grabau (1931, pp. 157, 158) on the sug- gestion of F. A. Bather. It is illustrated herein as Plate 1: figures la-3 (=pl. 15, figs. 10, lOa-b in Davidson, 1853; fig. 137, p. 195, in Grabau, 1923; fig. 13, p. 158, in Grabau, 1931a; fig. 29, p. 92, in Grabau, 1932; fig. 462, 2a-c of McLaren in Moore, 1965). Syntypes (=paralectotypes), same locality and same mode of acquisition as the lectotype: A, BM B42507 (Plate 1: figures 3a-b [=pl. 15, fig. 11, in Davidson, 1853]); B, BM B42508 (Plate 1: figure 2); c, BM B42509; D, BM B5290. Only four specimens, designated as syntypes, are in the box containing the specimens presented to the British Museum of Natural History by D. Hanbury. During the last visit to London, in going through T. Davidson's collections, I managed to locate a fifth specimen with two labels, on one of which was writ- ten, in Davidson's own hand: "publ. in the Proc. of Geol. Soc. for June 1853." It is believed, therefore, that this specimen is one of the primary types (Syntype D), and that the expression "four or five" used by Davidson might be due to the fact that he had some doubt about the attribution of Syntype A to the species. This juvenile specimen is somewhat deformed and gives the false impression, at least on its right side, of having a long hinge line. A sixth specimen, from the D. Hanbury collection, was presented to the British Museum of Natural History in 1911 and now bears the registered number B82307. Hypotypes, Province of Hunan(?), China; Yaoso Group, Devonian; USNM Cat. No. 60950; probably purchased in Canton: A, USNM 154996 (Plate 1: figures 4a-e) ; B, USNM 165708 (Plate 1: figures 5a- e) ; c, USNM 165709; D, USNM 165710; E, USNM 165711; F, USNM 165712; G, USNM 165713 (Plate 2). Locus TYPICUS AND STRATUM TYPICUM.—Accord- ing to Davidson (1853, pp. 353, 354, 356), the four or five specimens of the species described from the Pro- vince of Kwangsi were purchased in a drug ware- house in Shanghai and sent by W. Lockhart to D. Hanbury, who presented them to the British Museum of Natural History. A Devonian age has been as- signed to them (Figure 1). MATERIAL.—The following description is based on the 5 primary types; 16 specimens from the British Museum of Natural History (one in the D. Hanbury collection, B5290, and 15 in the R. Swinhoe collec- tion, 46787) ; 36 specimens from the United States National Museum (29 identified as Yunnanella han- burii, 3 as Y. triplicata, and 4 as Y. uniplicata), 4 specimens from the Geological Institute of the Acad- emy of Sciences, Union of Soviet Socialist Repub- lics, presented to Kh. S. Rozman by Chi-pou Yang (2 identified as Y. hanburii and 2 as Y. uniplicata); 4 specimens figured by Kayser (1883) ; 14 specimens figured by Grabau (1931a-1933) (5 identified as Y. hanburii, 3 as Y. hanburii lata, 3 as Y. triplicata, and 3 as Y. uniplicata) ; 7 specimens figured by Tien (1938) (1 identified as Y. hanburii, 3 as Y. hanburii sublata, 1 as Y. cf. triplicata, 1 as Y. uniplicata, and 1 as Y. triplicata latiformis). EXTERNAL CHARACTERS.—Pedicle valve is convex— generally slightly—in the umbonal region, which is the only part showing in lateral views. Flanks are steep in the posterolateral parts, and (sometimes) concave near the commissure. Anterolateral parts of the valve are flat to gently concave, but their extreme margins often are abrupt. Sulcus starts usually imperceptibly between 41 and 55 percent of the length of the shell forward of the beak (but values as low as 30 percent have been measured) or between 28 and 38 percent of the un- rolled length of the valve. Sulcus widens sometimes NUMBER 3 207 slowly, usually rapidly; its width, where it starts, varies between 30 and 50 percent of its maximum width at the front, which varies between 61 and 75 percent of the width of die shell. The sulcus is moderately deep to deep—3 to 5 times the height of the low costae at the front, where it is deepest. Floor of the sulcus is generally flat, and sometimes slightly convex. Sulcus passes progressively into a high trapezoidal tongue, the upper part of which tends to become vertical and commonly is even recurved posteriorly. The borders of die tongue are not indented by the costae. Median costae often protrude anteriorly beyond the sharp borders of the tongue. Greatest thickness of the shell is located posteriorly between 15 and 25 percent of die unrolled length of the valve forward of the beak. Sub- erect, well-marked beak ends with a small half-cir- cular foramen. Interarea is clearly delimited; it may reach a height of 2 mm, and its length varies between 35 and 50 percent of the width of die shell. Large del- tidial plates are best seen in transverse serial sections. Brachial valve is strongly convex, although never inflated. Flanks slope gently to abruptly, but their extreme margins are always abrupt; they also are steep in the posterolateral parts, where they become concave near the commissure. Like the sulcus, the fold starts imperceptibly at some distance forward of the beak. It is strongly marked, and is higher than the sulcus is deep; its top is flat to exceptionally strongly arched. Greatest thickness of the valve commonly is at the frontal commissure but sometimes occurs somewhat posterior to it. ORNAMENT.—All costae are simple and round- flattish in their posterior part. Median costae start generally between 50 and 66 percent of the unrolled length of the valves forward of the beaks, but they may start at 25 percent of this length. Their average width at the front is 3 mm to 3.5 mm with a range from 2.5 mm to 4 mm. The dorsal costae are high and acute in their anterior part, and usually are irregular because of one or two costae being lower and narrower than the others. The ventral costae are low and obtuse. Parietal costae have been observed in a single specimen (Syntype c). Lateral costae are restricted to the an- terior part of the shell, and the most external ones are mere identations of the commissure; these indentations are considered and counted as costae. Internal ventral lateral costae are high and acute in their anterior part. All other lateral costae are obtuse with rounded top, but, because of the angle under which they are cut by the commissure, they sometimes are acute or right at the commissure. Furrows have the same characteristics as the costae. Round-flattish costellae, starting at the umbones, cover the surface of the shell. They increase in number by successive bifurcation. Some costellae originate by intercalation. Four to six costellae per mm is an average number at mid-length. Costellae are superimposed on the costae. Fine growth lines are seldom seen. Ratios of median and lateral costae are shown in Table 1. The general costal formula, which in rhyncho- nelloid brachiopods is a formula grouping at least 75 percent of the specimens in each category, is as follows in this species: 2 to 4, 3 to_5 lto3' '4 to 6* GENERAL CHARACTERS.—The species is of medium- size and uniplicate. The contour in ventral and dorsal views is variable. Most of the specimens are trans- versely subelliptical to subcircular. Some specimens have a subpentagonal contour. The commissure is strongly indented by the costae. The anterolateral com- missures are low in the shell; they are not sharp, be- cause the extreme margins of the flanks are often ver- tical, but still they are indented in high zigzags by the costae. TABLE 1.—Ratios of median and lateral costae. MEDIAN COSTAE Number of costae Number of specimens Percent 2/1 3/2 4/3 12 63 10 14. 1 74. 1 11.8 85 100.0 LATERAL COSTAE 2/3 3/4 4/5 5/6 6/7 1 27 26 17 7 1.30 34.60 33.35 21.80 8.95 78 100. 00 The top of the tongue usually represents the greatest thickness of the shell. The greatest width of the shell is generally between 60 and 66 percent of the length from the beak. DIMENSIONS.—Measurements (in millimeters) are given in Table 2 (1, length; w, width; t, thickness; p.v., pedicle valve; b.v., brachial valve). The figures 208 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY SOUTH- KWANGSI EASTERN YUNNAN EASTERN CENTRAL AND SOUTHERN KWEICHOW Central Kwangs, SIKANG Northeaste n Kwangs. After D. S. SOKOLOV. After A w. After A. W After A W 1950 (1956) After D. S After A. W. After D. S. After A. C After A. W. After A. W. After AW. After D. S. After D. S. After A. C. GRABAU. GRABAU. GRABAU, After A. C. SOKOLOV, GRABAU. SOKOLOV. CHANG. GRABAU. GRABAU. GRABAU. SOKOLOV. SOKOLOV. CHANG. 1923 1931. 1931. and CHANG. 1958. and C. W. Ku. 1960 (1956). 1931.and I960 (1956). 1958. and 1931. 1931,and 1931. and I960 (1956). 1960 (1956). 1958. and pp. 186-188. pp. 149. V. K. TING. p. 302. V. K TING. pp. 458. 553. S. S. YOH pp. 149. 150. C. C. TIEN. S. S. YOH. p. 469. pp. 483. Y. C. SUN, 152. 153. pp. 135. 137. 139. Y. C. SUN. pp. -128-429. pp 71-72.73. pp.91 129- 130. 134. 139. P 73 152. 153. 155, 157 V. K. TING. pp.98. 134, 135. 139. pp. 130. 135. 139. 506-507. T. C.Lu, K K. CHAO. p. 74. CO _ 1 £ ~ rn E T a J, E E ;= | i _ £ s ? £ c -c " o | II E with Na ine//a Be E 1 E 6 E E 2: li I E I -3 h E E e » 1 1' B. & - g | 1 1 E ~ r£ 1 § 1 ~ 1 _ 1 E 1 -£ S E 1 I > E ] £ 1 — > > + U J E ,| X A ~t I E - o < /) s J. £ ~ | £ S = - D 3 E § £ Q 1 m 2 _c E 's, J ~ ^ 2 - E o J= ^2 > y > - * " js •?" g U in = - ? E t (2 | N Q I N c c Q J I 5 £ Q H I N t N Q - Y c H 1 Y N = | G 3 2 3 E. c E i S % c D o — D 0 ~ 2 G | l | ? G E 5 & G V, 6 J 1 m E j 1 I r 3 *'_ ^ HJ J 1 6 1 E 3 I 5 1 i J " & li r^ "j ?s l £ a 1/5 7 in 1 CT. ° u Member (Sh a ?J J — u Y N ^ J FIGURE 1.—Various interpretations on the age of the Upper Devonian formations in the area considered in this paper. When the genera Yunnanella and Nayunnella have been mentioned in the literature, they are indicated by "Y" and "N," respectively; the position of these letters on the figure has nothing to do with the genera's stratigraphic positions within the formation itself. This is not a correlation chart. in parentheses in the table represent reasonable esti- mates only, because of damage to specimens. Length and thickness of the primary types are different from those given by Davidson (1853, p. 356) and Grabau (1931a, p. 158) because they are measured with the plane of commissure being horizontal. Width is the greatest dimension, but the l./w. ratios show that length and width sometimes have close values; t./l. ratios show that the height may be greater than length. The majority of specimens have shoulder angles be- tween 94° and 110°. INTERNAL CHARACTERS.—Internal characters are given under the description of the genus. GROWTH.—Juvenile characters, observed on Syntype A: absence of costae, sulcus and fold not developed, height not developed. REMARKS.—The original description by Davidson is briel and incomplete. In particular, almost no informa- tion is given on the internal characters, and the figures are somewhat stylized drawings. This indicates the necessity of a new study based on more material, in- cluding the primary types. The species Y. triplicata and Y. uniplicata are con- sidered as entering the range of variability of Y. han- burii. Therefore, they are included in die description, as are the "mutation lata" of A. W. Grabau (1931a- NUMBER 3 209 HUNAN Southern Hunan Central Hunan After D. S. Alter A. W. Alter C.C.TIEN. 1938. After D. S. SOKOLOV. Alter AC. SOKOLOV, GRABAU, and 1960 (1956). CHANG. 1960(1956). 1931, and Y. L. WANG. p. 475. 1958, and p480. C. C. TIEN. pp.98. 131- 132. 134, 136. 137. 139. pp. 2, 3-4. 6, 7, 8-9, 14, 74-75. Y. C. SUN. p. 74. » o B - J 3 o mesto 3-300 (/i / E one (2 § Makunao J o 1 "5 / 0 S o Limestone E gsh ? g J c (200-220 m) § a u! X J + N LB B a. X / / a per Member = Hsiki Makun N N kuangsh an Seri N angsh an Forma Nit.iron hed(l- 2m) ZTW- son Nitangli iron bed (l-2m) N E 1 ^ ohlu Tut: Ls. [ 10-20 m) X Tutzutang Y Hsik l - 6 0 0 c / 3 Vun.Beds , . V Limestone( 10-20 m ) / ^ n / » nj 5 8 0 - 5 / "*? 0 / J = Chang!ung- 9 Changlungchic l B q S Q chieh Shale Q Shale (50-80 m 1 s UL Q > p e r u + N 3 Upper Shetienchiao Formation (500 m n . Shetienchiao Series — Sinosptnfer Shetienchiao Suite (100-200 m) Shetienchiao Formation Lungkouchung Suite (150-200 m) SOUTHWESTERN HUPEH After D S. SOKOLOV. 1960 (1956). p. 377-378. p. 541. After K. C YANG. 1964. p. 26. Upper Devonian Hsiehkiangsu Suite (3-36 m) -< Upper Devonian Hsiehkiangsu Formation -< Huangiadang Suite (2-65 m) SOUTHERN SHENSI After A. W. GRABAU. 1931. p. 162. Upper Devonian Shaly limestones of Hanchung NORTHWESTERN SZECHWAN After A. W. GRABAU. 1923. pp. 193-195. After D. S. SOKOLOV. 1960(1956) pp. 313. 520. After A. C CHANG. 1958. and S. S. YOH. pp. 71-72. 75. Upper Devonian (Tsinglingian Series) Hanchung Formation Upper Devonian Tangwangchai Series Maopa Limestone (360 ml Frasnian Tangwangchai Series Maopa Limestone B o CO Shawotze Dolomite INNER MONGOL. AUT. REG. After A. C. CHANG. 1958 andY. WANG. C. S. NING. L. LEE. M. L. CHANG pp. 76, 78. 79. 01 NY FIGURE 1.—Continued. TABLE 2.— -Dimensions of type specimens (measurements in millimeters). Syntypes are paralectotypes. Specimen I. w. l.p.v. t. t.p.v. t.b.v. l./w. t.fw. /.//. Shoulder Apical {unrolled) angle angle Hypotype B 17.5 20.7 26.5 17. 1 3.2 13.9 0.85 0.83 0. 98 95° 120° Syntype B (17.2) 18.6 23.5 (13.8) (2.8) 11.0 (0. 93) (0. 74) (0. 80) 105° 114° Syntype c 17.4 (17.4) 26.2 16.5 3.7 12.8 (1.00) (0.95) (0. 95) 94° 7 Hypotype A 17. 1 18.4 25.0 13.9 3.4 10.5 0.93 0.75 0. 81 100° 108° Hypotype E (16.8) 20.5 (29.5) 15.3 3.8 11.5 (0. 82) 0.75 (0.91) 110° 120° Lectotype, BM B42506 (16.6) (20.1) (28.5) 17.6 3.2 14.4 (0. 83) (0. 88) (1.06) ? 115° Hypotype c 16.4 21.0 30.7 16.8 3.5 13.3 0.78 0.80 1.02 103° 117° Hypotype D 14.4 19.6 25.0 17. 1 3.5 13.6 0.73 0.87 1. 19 100° 116° Hypotype F S 12.2 14.5 20.5 12.5 2.5 10.0 0.84 0.86 1.02 105° 110° Syntype A 3 8.3 9.2 9.2 3.8 1.7 2. 1 0.90 0.41 0.46 ? 102° 1 A very high specimen. 2 A young specimen having already developed adult features. 3 A juvenile specimen. 210 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY 1933), the "mutation sublata" and the Y. triplicata var. latiformis of C. C. Tien (1938) ; the first of these three forms, according to my analysis, includes adult specimens which did not develop the usual height of the species. Discussion of the Forms Labelled Yunnanella To my knowledge, 68 species and subspecies, some of which are introduced in the literature as mutations or varieties, have been labelled Yunnanella at one time or another; species mentioned merely as "F. sp." are not included in this figure. The 68 forms include: (1) the species originally and correctly attributed to the nominal genus Yunnanella; (2) the species mentioned under the generic name Yunnanellina and which should, according to the Code (see above), have been assigned to the nominal genus Yunnanella; and (3) the species mentioned under the nominal genus Yun- nanella that should be assigned, according to the Code, to the nominal genus Nayunnella. Except for Yun- nanella hanburii, Y. hanburii lata, Y. hanburii sublata, Y. triplicata, Y. triplicata latiformis, and Y. uniplicata, which have been discussed under the description of Y. hanburii, the validity of these species and subspecies is not considered here. (1) Seven species and subspecies (all Chinese) be- long to the genus Yunnanella: Y. hanburii (Davidson, T., 1853) ; Y. hanburii lata Grabau, A. W., 1931; Y. triplicata latiformis (Tien, C. C, 1938) ; Y. obesa (Tien, C. C, 1938) ; Y. hanburii sublata (Tien, C. C, 1938); Y. triplicata Grabau, A. W., 1931; and Y. uniplicata Grabau, A. W., 1931. It is possible that the two species described in Novaya Zemlya under the following names by Tcherkesova (1961) could belong to the genus, but they are represented, respectively, only by three and two specimens, and more informa- tion is needed: Junnanellina karina Tcherkesova, S. V., 1961, and /. triaequalis (Nalivkin). The latter form has already been attributed to the genus Yunnanella by Rozman (1959, p. 97), and some specimens from the Tarbagatai Mountains of Kazakhstan that have been identified as Paraphorhynchus triaequalis by Sverbilova (in Litvinovitch and Sverbilova, 1963, pp. 277-278) also could belong to the genus Yunnanella. (2) Eight Givetian species and subspecies belong to the genus Schnurella or to an undescribed genus: Yun- nanella custos Schmidt, H., 1941; Y. incisa Schmidt, H., 1941; Y. innae Ivanova, E. A., 1962; Y. olgae Ivanova, E. A., 1962; Terebratula schnurii de Vern- euil, E., 1840; Rhynchonella schnurii transversa Reed, F. R. C, 1908; Yunnanella transversiformis Tiajeva, A.P., 1962; and Terebratula voltzii d'Archiac, A., and de Verneuil, E., 1842. Terebratula schnurii and T. voltzii are correct spellings, as the conditions of Article 32(a) of the Code are met with. The names T. sch- nuri and T. voltzi also may be accepted (see the nom- enclatorial note at the beginning of the description of Y. hanburii). Four species are known from Germany, two from the Kouznetzk Basin, one from the western flank of the South Ural Mountains, and one from Burma and Armenia. One species belonging to the group has been mentioned in the Givetian of the Kouz- netzk Basin by Ivanova and Tchoudinova (1959, p. 612) under the name Yunnanella aff. triloba. (3) Six names are nomina nuda, and the generic assignment may be fixed for five of them: Schnurella schnurii kuzbassica (Rzhonsnitzkaia, M. A., 1962) ; Nayunnella multiplicata Grabau, A. W., 1931); N. tieni (Grabau, A. W., 1931) ; Yunnanella pentaplicata (Grabau, A. W., 1931); and Y. quadriplicata (Gra- bau, A. W., 1931) ; these forms are found in China and in the Kouznetzk Basin. The sixth name is Yun- nanellina markovskii Rozman, Kh. S., 1959, a Frasnian form from the Ural Mountains. (4) The following six species do not belong to the genus Yunnanella and are not present in the area in- dicated by various scientists. The Yunnanese species Leiorhynchus deprati Mansuy, H., 1912 is present neither in Kazakhstan (contrary to Nalivkin, 1930, pp. 67, 68; Brongouleev, 1957, pp. 20, 22; and Sidiat- chenko and Alekseeva, 1958, p. 159), nor in the Tian Chan Mountains (contrary to Poiarkov, 1960, p. 30, table 2), nor in Armenia (contrary to Abramian, 1957, p. 8, tables 2, 3, and Arakelian, 1964, pp. 61, 94), nor in the Altai Mountains (contrary to Komar, 1957, p. 36, 39). The Chinese species Nayunnella ericksoni (Grabau, A. W., 1931) is present neither in Kazakh- stan (contrary to Rozman, 1959, pi. 6, figs. la,b,v; 1962, pp. 77, 80, table 11, pp. 90, 131, 132), and Alek- seeva and Sidiatchenko, 1959, table 2), nor in Novaya Zemlya (contrary to Tcherkesova, 1961, p. 45). The Chinese species N. grandis (Grabau, A. W., 1931) is not present in Kazakhstan (contrary to Sidiatchenko and Alekseeva, 1958, p. 159). The Belgian species Ptychomaletoechia gonthieri (Gosselet, J., 1887) is not present in Kazakhstan (contrary to Nalivkin, 1937, pp. 78, 79; Simorin, 1956, pp. 241, 242; Martynova, 1956, p. 97; 1962, table 10; and Alekseeva and Sidiat- NUMBER 3 211 chenko, 1959, table 3, p. 27). The Chinese species Yunnanella hanburii (Davidson, T., 1853) is present neither in Kazakhstan (contrary to Nalivkin, 1930, p. 66, and Brongouleev, 1957, pp. 19, 20), nor in the Tian Chan Mountains (contrary to Poiarkov, 1960, p. 30, table 2). The Belgo-French species Eoparaphor- hynchus triaequalis (Gosselet, J., 1877) is present neither in Kazakhstan (contrary to Nalivkin, 1937, pp. 79, 80; Simorin, 1956, pp. 239-241; Martynova, 1956, p. 92; 1961, pp. 28, 30, 40, 42, 44, 68, 100-102; Sidiatchenko and Alekseeva, 1958, p. 159; Rozman, 1959, pp. 92, 97-99; 1962, p. 18, tables 7, 10, pp. 67, 69, 78, 80, table 11, pp. 90, 138-140; Alekseeva and Sidiatchenko, 1959, p. 19, table 2, pp. 20, 23, 25, table 3, p. 27; Sidiatchenko, 1961, p. 1159, table 1, p. 1161; and Sverbilova in Litvinovitch and Sverbilova, 1963, pp. 277, 278), nor in Novaya Zemlya (contrary to Tcherkesova, 1961, pp. 45, 49, 50). (5) The two English species Terebratula anisodonta Phillips, J., 1841, and Rhynchonella (Camarotoechia) partridgiae Whidborne, G. F., 1897, do not belong to the genera Yunnanella and Nayunnella (contrary to Tien, 1938, p. 43, and Reed, 1943, p. 134), respectively. (6) The following species constitute a very hetero- genous group: Paraphorhynchus badura Nalivkin, D. V., 1937; P. celak Nalivkin, D. V., 1937; P. fatima Nalivkin, D. V., 1937; Yunnanellina karat- auensis Rozman, Kh. S., 1960; Y. kasakhstanica Roz- man, Kh. S., 1960; Y. kurgandjarica Rozman, Kh. S., 1960; Y. mugodjarica Rozman, Kh. S., 1960; Para- phorhynchus zobeida Nalivkin, D. V., 1937; and P. zuleika Nalivkin, D. V., 1937. The diacritical mark on the word "celak" is dropped in accordance with Article 27 and Article 32(c) (i) of the Code. It is not certain that these species include all forms that have been given their names. To this group may be added some forms identified as P. gonthieri (including a va- riety) and P. triaequalis, or Yunnanellina gonthieri and Y. triaequalis, by Russian scientists. None of these twelve species and forms belongs to the genus Yun- nanella, from which they differ by shape, sulcus (shape, depth, delimitation, point where it begins), fold (point where it begins), longer costae, divided median costae, etc. Furthermore, these species and forms do not be- long to a single genus; a small species like Paraphor- hynchus zuleika has very little in common with a large species like Yunnanellina kasakhstanica. In fact, these species and forms belong to genera such as Evanesciros- trum Sartenaer, P., 1965; Porostictia Cooper, G. A., 1955; possibly Rugaltarostrum Sartenaer, P., 1961; 372-386 0—71 15 and, perhaps, one new genus if not two new genera. Each of the species and forms should be investigated thoroughly, but the lack of material does not permit me to proceed with such a study. It is of primary im- portance that a detailed investigation should solve these systematics problems and thus permit the in- clusion of the regions concerned in the Famennian correlation scheme based on rhynchonellid zones pro- posed by Sartenaer (1967). Except for Yunnanel- lina kurgandjarica and Y. mugodjarica, which come from the Mugodjary Mountains, all these species and forms are found in Kazakhstan. (7) The forms identified in Kazakhstan as Camaro- toechia hanburyi by Nalivkin (1930, p. 66) and as Yunnanella ericksoni by Rozman (1959, pi. 6, figs. la,b,v; 1962, pp. 77, 80, table 11, pp. 90, 131, 132) and by Alekseeva and Sidiatchenko (1959, table 2) do show some similarity with the genus Nayunnella but, among various other differences, the median costae in those forms start nearer to the beaks. Although too little is known about these forms, the possibility can- not be dismissed that a new genus could include some Russian species that have been assigned to the genus Nayunnella. (8) The following two species from Kazakhstan most probably belong to the genus Eoparaphorhynchus Sartenaer, P., 1961: Yunnanella acutiplicata Rozman, Kh. S., 1962; and Y. nalivkini Rozman, Kh. S., 1960. The latter species includes in its synonymy the forms attributed by Nalivkin (1930, pp. 67, 68) to Leiorhyn- chus deprati Mansuy, H., 1912. In addition, some specimens from Kazakhstan identified by Russian sci- entists as Paraphorhynchus gonthieri or Yunnanellina gonthieri may belong to the genus Eo paraphorhynchus. This attribution has been suggested by Sartenaer (1967, p. 1048; 1969, pp. 62, 63), who also had in mind some specimens from the Mugodjary Mountains included in species validly assigned to other genera. (9) The following 26 Chinese species and subspecies, some of which have been introduced in the literature as mutations or varieties, belong to the genus Nayun- nella: N. abrupta (Grabau, A. W., 1931); N. erick- soni (Grabau, A. W., 1931) ; N. abrupta globosa (Tien, C. C, 1938); N. grandis (Grabau, A. W., 1931); N. hsikuangshanensis (Tien, C. C, 1938) plus four varie- ties; N. hunanensis (Tien, C. C, 1938) plus one vari- ety; N. abrupta media (Tien, C. C, 1938) ; N. meso- plicata (Grabau, A. W., 1931) ; N. abrupta rostrata (Tien, C. C, 1938) ; N. abrupta schnurioides (Tien, C. C, 1938) ; N. simplex (Tien, C. C, 1938) ; N. 212 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY abrupta subcuboides (Tien. C. C, 1938) ; N. uncinu- loides subpentaplicata (Tien, C. C, 1938) ; N. synpli- cata subtriplicata (Tien, C. C, 1938) ; N. supersyn- plicata (Tien, C. C, 1938) ; N. synplicata (Grabau, A. W., 1931) plus one mutation; N. triloba (Tien, C. C, 1938) ; N. tieni (Grabau, A. W., 1931) ; N. uncinu- loides (Tien, C. C, 1938) ; N. wangi (Tien, C. C, 1938). The only specimen described by Tcherkesova (1961, pp. 46, 47) in Novaya Zemlya as Junnanella ericksoni polaris Tcherkesova, S. V., 1961, could well belong to the genus Nayunnella, as could some speci- mens identified as Paraphorhynchus triaequalis in the Karaganda Basin of Kazakhstan. (10) The genera Yunnanella and Nayunnella have been reported several times as present in North America, but this view is rejected. Warren and Stelck (1950, p. 64) state that Leiorhyn- chus walcotti from the Mackenzie River Valley "ap- pears to belong to the sub-genus Yunnanellina of the genus Yunnanella"; Sartenaer (1969, pp. 62, 63) in- cludes this form in the lower Famennian species Eo paraphorhynchus maclareni Sartenaer, P., 1961. In the same publication Sartenaer (1969) attributes, with doubt, the Nayunnella cf. N. mesoplicata, cited by C. H. Crickmay (1952, p. 593) and collected at about 530 feet below the top of the Palliser Formation, to the species Sinotectirostrum medicinale Sartenaer, P., 1961. Yunnanella? sp. has been mentioned recently by Dutro (in Jones, Hernon, and Moore, 1967, p. 22; 1967, personal communication) as occurring in the Percha Shale of southwestern New Mexico and the Upper Devonian rocks of northwestern Alaska. These two occurrences most probably can be referred to the genus Rugaltarostrum Sartenaer, P., 1961. Thanks to the kindness of Dr. Dutro, I had the opportunity to see the material in Washington in 1966, but a detailed study is needed before a more definite statement can be made. Stratigraphic Position of Yunnanella and Nayunnella One of the greatest handicaps in dealing with the genera Yunnanella and Nayunnella is the lack of pre- cise stratigraphic information on Chinese outcrops. The Chinese literature is very vague on the subject, and contacts with colleagues in China are impossible for the time being; thus, stratigraphic problems can only be touched upon. Still, a few points, some of them already mentioned (Sartenaer, 1967, p. 1056; 1969, pp. 62, 63), may be discussed. (1) There are two completely opposed lines of thought among Chinese scientists on the age of the beds containing the genera Yunnanella and Nayun- nella: Famennian or Frasnian. In the latter case, the late Upper Devonian is generally considered absent in South China. One of the strongest arguments in favor of the Frasnian age of the beds containing these genera is used by Wang and Ning (1957) and by Chang (1958) in the Great Khingan (Inner Mongolian Au- tonomous Region), where the Upper Devonian is divided into an Upper Suhuho Formation and a Lower Suhuho Formation, separated by an hiatus. A Sporadoceras-Prolobites fauna, to which a lower to middle Famennian age is given, has been discovered in the Upper Suhuho Formation, in which the trilobite Phacops granulatus also has been found. The Lower Suhuho Formation contains the Hsikuangshan fauna (Frasnian, according to Chang). When the thickness (800 m to 1,200 m) of the Upper Suhuho Formation is considered, the fact that the Sporadoceras-Prolobites fauna has been collected in an unspecified limestone bed does not permit far-reaching conclusions. These views are strongly opposed, however, by Yao (1959), who accepts ideas expressed by Tien (1938). Hamada (1960, pp. 228, 229), who accepts the Frasnian age of the beds with Yunnanella and Nayunnella, does not agree with the absence of the Famennian in South China advocated by A. C. Chang (1958) and believes that the hiatus between the Lower and Upper Suhuho Formations is small. There is not much an outsider can do in this changeable situation, roughly illustrated on Figure 1, in which the various regions are separated in order to point out that the figure is not a correlation chart. Obviously, in some instances in China, either the genus Yunnanella or the genus Nayunnella is asso- ciated with a Frasnian fauna (e.g., in Grabau, 1923, pp. 194, 195), but not much importance can be put on this kind of information found in the literature, because the thickness of rocks involved may be great. In some instances (e.g., in Sokolov, 1960; 1956, p. 428) this thickness is as small as 5 meters. (2) Although representatives of the genera Yun- nanella and Nayunnella may occur together in China, the latter genus ranges higher in the sequence. (3) The genus Nayunnella has been mentioned er- roneously by Tien (1938, p. 49) in the Lower Car- boniferous of China. The fact that Reed (1943, p. NUMBER 3 213 134) has attributed to the genus Nayunnella the Brit- ish species of the Pilton Beds, Rhynchonella (Camaro- toechia) partridgiae, may also explain why the genus sometimes has been mentioned from the Lower Car- boniferous. (4) Neither genus is present in the Frasnian de- posits of Kazakhstan. The mention of the genus Yunnanella in the South Ural Mountains is due to the attribution by Rozman (1959, pp. 92, 97) of the Pug- noides triaequalis mentioned by Markovskii (1948, p. 34) in the Barma Beds. I reject this attribution. Rozman (1959, p. 93, fig. 1, p. 98) also mentions the genus Yunnanella in the Upper Frasnian of the Mugodjary Mountains, but this is not substantiated by any information known to me. At any rate, these two references were later omitted by Rozman (1962). (5) Similarities between the fauna of the Famen- nian of western Europe and the fauna associated with the genera Yunnanella and Nayunnella are difficult to establish. Still, it must be born in mind that the Chinese fauna is still very poorly described. (6) All Givetian species attributed in the past to the genus Nayunnella belong either to the genus Schnurella or to an undescribed genus. (7) Although most of the species attributed in the Union of Soviet Socialist Republics to the genus Yun- nanella do not belong to it, the presence of that genus in the Famennian cannot be dismissed. (8) The stratigraphic range of both Yunnanella and Nayunnella is still unsettled in the type area. FIGURE 2.—Known geographic distribution in China of the genera Yunnanella and Nayunnella. U Vx S s fZZ \ R KAZAKHSTAN V^ /y^a^-, Tarbagatai Mts/^ ^Balkhash Lake P^ /y-sT /HEILUNGKlW J ^"^ _j> ^? /SINGKIANG W MONGOLIA s^ cJ U_^ ^ \ J^^X~\KIHGIZIYA ^"^-y ) v (\~^v-^3 Tian_pahan Mts. ^J TA0ZH>V^T_^-^^ SINKIANG j-Swm MONGOUANW /"^AUTONOMOUS REGION KIRIN ^^J/^-'^ —vr-O t_ AUTONOMOUS REGION AFGHJ^J KASHMIR C /~^~~^-^—— PAKISTAN^ Y j 1 KANSU TSINGHAI ^^-^^ /-" V^^^CHENGTEH Pfty ~\ NINGSIA ^-j SUIYUAN JX ^—^/ / \__/^\ IV1 jTy M HOPEH £ V. V ~\_y PHANSI / s? ^-* S \ ^-\T^ ( L-.\ SHANTUNG f m J / yOREA\ ^ ^ TIBET ^ INDIA \^^_ / SIKANG ( -^^ fiSHENSI \ HONAN 7o^^A*IGSU\ 2? PACIFIC ANf OCEAN I " "^ Ullnru^( ANHWEI\T~. V SZECHWAN ~M HUPEH 1 r * 3 £^j— ^_^ /J^V^vtHEK L yNj^^V ^= ) l(IANGSI r^Vr, ^_^_^ SPAKISTAip / BURMA V^ YUNNAJF^\^~ ^= ^ J <^~~\J = S_^ KMNfiSl yinfWANGTUNG s^ TAIWAN 74 \VIETNAM V>-v*fC^^ LAOO X /siAM 214 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY wWr TV ' w ^ mm &% 5c PLATE 1.—Yunnanella hanburii (Davidson, T., 1853). Views are not necessarily oriented with the plane of commissure parallel or perpendicular to the plane of the plate. (Photos for figures 1-3' furnished by Museum of Natural History.) la-e, Lectotype, BM B42506; Province of Kwangsi, China; Devonian. Ventral, dorsal, lateral, apical, and frontal views (X 2). Costal formula: 3/2; 0; 4/5. 2, Syntype (= Paralectotype B), BM B42508; Province of Kwangsi, China; Devonian. Dorsal view ( X 2). Costal formula: 3/2; 0; 5/6. 3a, b, Syntype (= Paralectotype A), BM B42507; Province of Kwangsi, China; Devonian. Dorsal and ventral views (X 3). 4a-e, Hypotype A, USNM 154996; Province of Hunan(?), China; Yaoso Group, Devonian. Ventral, dorsal, lateral, apical, and frontal views ( X 1 )? Costal formula: 3/2; 0; 5/6. 5a-e, Hypotype B, USNM 165708; Province of Hunan (?), China; Yaoso Group, Devonian. Ventral, dorsal, lateral, apical, and frontal views (X 1). Costal formula: 3/2; 0; 5/6. NUMBER 3 215 O Lt tmt iot ioi JoH 0.7 0.75 0.9 0.92 0.95 1 1.1 1.15 1.25 1.35 1.4 1.45 1.5 1.65 o a o 2.1 2.15 2.25 /^r>ooo *-» -» 2.7 2.95 3.3 3.5 3.6 3.65 r—^ >—^ \^-^ X 3 Yunnanella hanburii (DAVIDSON, T., 1853) PLATE 2.—Yunnanella hanburii (Davidson, T. 1853). Hypotype G, USNM 165713; Province of Hunan(?), China; Yaoso Group, Devonian. Camera lucida drawings of serial transverse sections ( X 3) ; distances are in mm forward from the crest of the umbo. Dimensions of specimen: length, 17.50 mm; width, 20.85 mm; thickness, 14.80 mm. 216 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Geographic Distribution of Yunnanella and Nayunnella The regions of China where species and subspecies of the genera Yunnanella and Nayunnella have been de- scribed or cited in the literature are indicated on Figure 2. Outside these areas, the genera could be present in Kazakhstan (Tarbagatai Mountains, Karaganda Basin) and in Novaya Zemlya, but additional studies are needed. The genera are not present in other parts of the Union of Soviet Socialist Republics, and certainly not in the Ural and Mugodjary Mountains. The at- tribution of various species and subspecies to the Givetian genus Schnurella and to an undescribed genus eliminates the genus Nayunnella from Armenia, Burma, Germany, the Kouznetsk Basin, and the western flank of the South Ural Mountains. As no English species belongs to either the genus Yunnanella or the genus Nayunnella, these genera are absent in Great Britain. Conclusions An effort to clarify and sort the problems involving the genus Yunnanella has been attempted. A clear definition of the genus has permitted the elimination of various species and subspecies attributed to it in the past and thus to clarify its systematic significance. The immediate result has been to restrict its geographic distribution to China, and perhaps Kazakhstan and Novaya Zemlya, and its stratigraphic range to the Upper Devonian. The lack of good information, and particularly of good stratigraphic sections from China, prevents a more precise statement about the strati- graphic range within the Upper Devonian, although it is quite obvious that the genus Yunnanella did not range through the entire late Devonian. Once again, it is demonstrated that a considerable stratigraphic range (Givetian to Carboniferous) for a rhynchonellid genus immediately indicates that it is in need of revision and that it encompasses other genera. It will be possible to deal fruitfully with the problems of paleogeography, classification, and evolu- tion only when the various genera are more precisely defined. References Abramian, M. S. 1957. Brakhiopody verkhnefamenskikh i etrenskikh otlo- jenii iougo-zapadnoi Armenii. Akademiya Nauk Armyanskiy S.S.R., Institut Geologicheskiy Nauk, 142 pages. Erivan. Alekseeva, R. E., and A. I. Sidiatchenko 1959. Biostratigrafiia famenskikh otlojenii tzentralnogo i iougo-vostotchnogo Karataou (Ioujnyi Kazakh- stan). Izvestiya Vysshikh Uchebnykh Zavedeniy Geologiya i Razvedka, 2:15—29. Arakelian, R. A. 1964. Devon, in Geologiia Armianskoi S.S.R., t.II: Strati- grafiia. Akademiya Nauk Armyaniskiy S.S.R., In- stitut Geologicheskiy Nauk, pages 46—96. Basic Invertebrate Fossils of China. 1957. 3 volumes. Peking: Academia Sinica, Institute of Paleontology. Brongouleev, V. V. 1957. Osnovnye tcherty stroeniia i razvitiia srednepaleo- zoiskogo strouktournogo etaja tzentralnogo Kara- taou. Akademiya Nauk Izvestiya, Seriya Geologis- che skay a, 2:15-41. Chang, A. C. 1958. Stratigraphy, Palaeontology and Palaeogeography of the Ammonite Fauna of the Clymeneenkalk from Great Khingan with Special Reference to the Post Devonian Break (Hiatus) of South China. Acta Palaeontologica Sinica, 6(1): 71—33 (in Chinese), 83-89 (in English). Chao, K. 1947. Stratigraphical Development of Kwangsi. Bulletin of the Geological Society of China (A. W. Grabau Memorial Volume), 27:321-346. Crickmay, C. H. 1952. Discrimination of Late Upper Devonian. Journal of Paleontology, 26(4) : 585-609. Davidson, T. 1853. On some fossil brachiopods of the Devonian age, from China. Quarterly Journal of the Geological Society of London, 9(33) : 353—359. Grabau, A. W. 1923, Stratigraphy of China, Part I. Palaeozoic and 1924. Older. Geological Survey of China, pages 1-200 [1923], pages 201 to end [1924]. 1931a, Devonian Brachiopoda of China. I. Devonian 1933. Brachiopoda from Yunnan and Other Districts in South China. Palaeontologica Sinica, Series B, 3(3) : 1-454 [1931], 54 plates [1933]. 1931b. Problems of Chinese Stratigraphy. Science Quar- terly of the National University of Peking, 2(2): 91-162. 1932. Studies for Students. Studies of Brachiopoda, III, Science Quarterly of the National University o\ Peking, 3(2): 75-112. Hamada, T. 1960. The Middle Palaeozoic Formations in China and Korea. I. Korea and Northeast China. II. North- west and South China. Japanese Journal of Geology and Geography, 31 (2-4) : 165-183, 219-239. Ivanova, E. A. 1961. Yunnanella Grabau, 1931, ili Nayunnella Sar- tenaer, 1961?. Paleontologicheskiy Zhurnal, 1961, 4:151-152. NUMBER 3 217 1962. Ekologiia i razvitie brakhiopod siloura i devona Kouznetzkogo, Minousinskogo i Touvinskogo bas- seinov. Akademiya Nauk S.S.S.R., Institut Paleon- tologii Trudy, 88:1-151, 20 plates. Ivanova, E. A., and I. I. Tchoudinova 1959. Novye dannye po faoune devona Kouznetzkogo basseina. Akademiya Nauk S.S.S.R., Doklady, 125(3): 611-613. Jones, W. R.; R. M. Hernon; and S. L. Moore 1967. General Geology of Santa Rita Quadrangle, Grant County, New Mexico. United States Geological Survey Professional Paper, 555:1-144, 3 plates, 50 figures. Kayser, E. 1883. Devonische Versteinerungen aus dem siidwestlichen China, in F. P. W. von Richthofen, volume 4, part 5, pages 75-102, in China. Ergebnisse eigener Reisen und darauf gegrundeter Studien. 1877-1912. 5 volumes. Berlin: D. Reimer. Komar, V. A. 1957. Stratigrafiia devonskikh otlojenii Roudnogo Altaia in Materialy po geologii i metallogenii Roudnogo Altaia. Trudy Vsesoyuznojo Aerogeolicheskaya T rest a, 3:15^5. Lee, J. S. 1939. The Geology of China. 528 pages, 93 figures. London: Thomas Murby and Company. Likharev, B. K. 1934. Klass Brachiopoda pererabotano B. K. Likharevym. In, Osnovy Paleontologii {Paleozoologiia) (K. A. von Zittel) pod redaktziei A. N. Riabinina, pages 458-552. Litvinovitch, N. V., and T. V. Sverbilova 1963. Brakhiopody verkhnego devona in Opisanie faouny i flory. In, Stratigrafiia i faouna paleozoiskikh otlojenii khrebta Tarbagatai, pod redaktziei A. A. Bogdanova, pages 253-292. Moskoviskiy Uni- versitut. Markovskii, B. P. 1948. Otcherk stratigrafii devonskikh otlojenii zapadnogo sklona srednego i ioujnogo Ourala. Materialy V.S.E.G.E.I. {All-Union Geological Scientific Re- search Institute), obshchya seriya, Sbornik 8, pages 22-38. Martynova, M. V. 1956. Famenskii iarous verkhnego devona zapadnoi tchasti tzentralnogo Kazakhs tan a. Sovietskaya Geologiya, 52:85-98. 1960. Famenskii iarous verkhnego devona zapadnoi tchasti tzentralnogo Kazakhstana. In, Troudy sovecht- chaniia po ounifikatzii stratigrafitcheskikh skhem dopaleozoia i paleozoia vostotchnogo Kazakhstana, volume 2, pages 23-25. Akademiya Nauk Kazakh- skoy S.S.R., Institut Geologicheskiy Nauk. 1961. Stratigrafiia i brakhiopody famenskogo iarousa zapadnoi tchasti Tzentralnogo Kazakhstana. In, Materialy po geologii Tzentralnogo Kazakhstana, volume 2, 151 pages. Tzentralno-Kazakhstanskoge Geologicheskoge Upravlenige Ministerstva Geo- logii i Okhrany Nedr Kazakhskoy S.S.R. McLaren, D. J. 1962. Middle and Early Upper Devonian Rhynchonel- loid Brachiopods from Western Canada. Geologi- cal Survey of Canada Bulletin, 86:1-122, 18 plates, 29 figures. Moore, R. C. (editor) 1965. Treatise on Invertebrate Paleontology, Part H, Bra- chiopoda. 927 pages, 746 figures. New York City and Lawrence, Kansas: Geological Society of American and University of Kansas Press. Nalivkin, D. V. 1930. Brakhiopody verkhnego i srednego devona Tour- kestana. Geol. Komit. Trudy, new series, 180:1— 221. 1937. Brakhiopody verkhnego i srednego devona i ni- jnego karbona severo-vostotchnogo Kazakhstana. Tzentralnya Geologo-razvedochni Institut, Trudy, 99:11-200. Ngo, Z.-K., and C.-C. Yang 1959. The Devonian Stratigraphy of the Tzelishihmen Area, North-Western Hunan. Acta Geologica Sinica, 39(1):94-101 (in Chinese), 101-102 (in English). Poiarkov, B. V. 1960. O stratigrafii famenskikh i nijnetourneiskikh otlo- jenii zapadynkh otrogov Tian-Chania. Akademiya Nauk Kirgizskoy S.S.R., Izvestya Seriya Estest- vinnyy i Teknicheskiy Nauk, 2(9): 23-48. Reed, F. R. C. 1908. The Devonian Faunas of the Northern Shan States. Memoirs of the Geological Survey of India, Pale- ontologica Indica, new series, 2(5): 1-183, 20 plates. 1953. Notes on Certain Upper Devonian Brachiopods Figured by Whidborne. Geological Magazine, 80 (2): 69-78; 80(3): 95-106; 80(4): 132-138. Rzhonsnitzkaia, M. A. 1948. Devonskie otlojeniia Zakavkazia. Akademiya Nauk S.S.S.R, Doklady, 59(8) : 1477-1480. 1962. Devonskie otlojeniia glavneichikh razrezov Sibiri i ikh korreliatziia s devonom Evropy. Sdvietskaya Geologiya, 10: 16-27. Rzhonsnitzkaia, M. A.; B. K. Likharev, and V. P. Makridin. 1960. Otriad Rhynchonellida, Klass Articulata, Tip Bra- chiopoda. In, Osnovy paleontologii, spravotchnik dlia paleontologov i geologov SSSR, Mchanki, Bra- khiopody. Prilozhenige: Foronidy, Otvetst. redak- tor T. G. Sarytcheva, pages 239-257. Roger, J. 1952. Classe des Brachiopodes. In, Traite de Paleonto- logie, volume 2, publie sous la direction de J. Pivetaeu, pages 3-160, 12 plates, 121 figures. Paris. Rozman, Kh. S. 1959. O predstaviteliakh podsemeistva Yunnanellina iz Kazakhstana i Mougodjar. Paleontologicheskiy Zhurnal, 2: 91-100. 218 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY 1960a. Stratigrafiia famenskikh i nijnetourneiskikh otlojenii Mougodjar i smejnykh raionov Ourala. Akademiya Nauk S.S.S.R., Izvestiya, Seriya Geologischeskaya, 12: 42-51. 1960b. Novye vidy devonskikh kamarotekhiid Mougodjar in Novye vidy drevnykh rastenii i besozvonotch- nykh SSSR, Part 1, V.S.E.G.E.I. {All-Union Geological Scientific Research Institute), pages 352-360. 1962. Stratigrafiia i brakhiopody famenskogo iarousa Mougodjar i smejnykh raionov. Akademiya Nauk S.S.S.R., Trudy, Geologicheskogo Instituta, 50: 1- 195. Sartenaer, P. 1961a. Note nomenclatoriale: Yunnanella Grabau, Yun- nanellina Grabau, Nayunnella nom. nov. (Rhyn- chonelles). Institut Royal des Sciences Naturelles de Belgique Bulletin, 37(2) : 1—3. 1961b. Late Upper Devonian (Famennian) Rhynchonel- loid Brachiopods. Institut Royal des Sciences Naturelles de Belgique Bulletin, 37(24): 1-10, 2 plates. 1962. A propos de l'espece-type du genre Yunnanella Grabau, A. W., 1923. Institut Royal des Sciences Naturelles de Belgique Bulletin, 38(19). 1967. Famennian Rhynchonellid Brachiopod Genera as a Tool for Correlation. Proceedings International Symposium on the Devonian System, pages 1043— 1060. Calgary, Canada. 1969. Late Upper Devonian (Famennian) Rhynchonel- lid Brachiopods from Western Canada. Geological Survey of Canada Bulletin, 169: 1-269, 19 plates, 47 figures. Schmidt, Herta 1941a. Die mitteldevonischen Rhynchonelliden der Eifel. Senckenbergische Naturforschende Gesellschaft Ab- handlungen, 459:1—65, 7 plates, 1 figure. 1941b. Rhynchonellidae aus rechtsrheinischem Devon. Senckenbergiana Lethaea, 23(4/6): 277-290. 1964. Neue Gattungen palaozoischer Rhynchonellacea {Brachiopoda). Senckenbergiana Lethaea, 45(6): 505-506. 1965. Neue Befunde an palaozoischen Rhynchonellacea (Brachiopoda). Senckenbergiana Lethaea, 46(1): 1-25. Sidiatchenko, A. I. 1961. Zonalnoe rastchlenenie famenskogo iarousa khr. Karataou, zapadnykh otropov Tian-Chania i zapad- noi tchasti tzentralnogo Kazakhstana po brakhio- podam. Akademiya Nauk S.S.S.R., Doklady, 140 (5):1159-1162. Sidiatchenko, A. I., and R. E. Alekseeva 1958. Brakhiopody i osnovnye voprosy stratigrafii famen- skikh otlojenii tzentralnoi i iougovostotchnoi tchas- tei khrebta Kara-Taou. Bioul. Byulleten' Moskov- skogo Obshchestva Ispytateley Prirody, new series, volume 43; Otdal Geologicheskiy, 33(1) : 159-160. Simorin, A. M. 1956. Stratigrafiia i brakhiopody Karagandinskogo bas- seina. 300 pages, illustrated. Akademiya Nauk Kazakhskay SSR, Institut Geologii. Sokolov, D. S. 1956. Regionalnaia stratigrafia Kitaia. [In Chinese; translated in Russian in 1960 under the editorship of and with a preface by D. S. Sokolov.] Tcherkesova, S. V. 1961. O nekotorykh predstaviteliakh semeistva Camaro- toechiidae v famenskikh otlojeniiakh Novoi Zemli. In, Sbornik statei po paleontologii i bio stratigrafii, Nauchno-Issledovatelskii Institut Geologii Arktiki, 25:44-51. Tiajeva, A. P. 1962. Brakhiopody srednedevonskikh otlojenii zapadnykh i tzentralnykh raionov zapadnogo sklona ioujnogo Ourala in Brakhiopody, ostrakody i spory srednego i verkhnego devona Bachkirii. Akademiya Nauk SSSR, Bashkirskii Filial, Institut Gornago-Geologii, pages 5—165. Tien, C. C. 1938. Devonian Brachiopoda of Hunan. Palaeontologia Sinica, new series B, 4: 1-192, 22 plates. Wang, Y., and C. S. Ning 1957. Devonian Formations at the Upper Part of the Halahaho River, Southern Central Khingan Range, and Their Lithic Characters. Geological Informa- tion, 9. Warren, P. S., and C. R. Stelck 1950. Succession of Devonian Faunas in Western Canada. Transactions of the Royal Society of Canada, series 3, 44(4):61-78. Yang, K.-C. 1964. Some Bryozoans from the Upper Devonian of Changyang, Western Hupeh. Acta Paleontologica Sinica, 12(1) : 26-29, 32-33 (in Chinese), 29-31 (in English). Yao, S. C. 1959. Some Opinions on "Stratigraphy, Palaeontology and Palaeogeography of the Ammonite Fauna of the Clymeneenkalk from Great Khingan with Special Reference to the Post Devonian Break (hiatus) of South China by A. C. Chang." Geologi- cal Review, 19(5). Yoh, S. S. 1956. Subdivision, Zonation and Correlation of the De- vonian Formations in Lungmenshan Area, North- western Szechuan. Acta Geologica Sinica, 36(4): 443-470 (in Chinese), 470-476 (in English). J. Stewart Williams The Beirdneau and Hyrum Formations of North- Central Utah ABSTRACT The Hyrum and Beirdneau Formations of north-cen- tral Utah have their best exposures in Blacksmith Fork Canyon, Cache County. Sections at this locality are described in detail. Three sections to the northwest and two to the southeast are compared with the Black- smith Fork section. The Hyrum Formation, about 1,000 feet thick, is composed of thin to medium beds of dark and medium gray dolomite with some calcitic dolomite, dolomitic limestone, and limestone. There are a few thin inter- beds of sandstone and dolomitic sandstone. The fos- siliferous Samaria Member at the base is of Middle Devonian (Givetian) age. The Hyrum Formation cor- relates with the Jefferson Formation to the north and east. The Beirdneau Formation, 1,000 feet or less in thick- ness, is composed of thin-bedded dolomitic sandstone and argillaceous dolomite. It correlates closely with the Three Forks Formation to the east and north and is upper Upper Devonian (Famennian) in age. In a paper on the geology of the Paleozoic rocks of the Logan quadrangle, the writer (Williams, 1948) described briefly three stratigraphic units in the local Devonian: the Water Canyon Formation of Lower Devonian age and the Hyrum Dolomite and Beirdneau Sandstone members of the Jefferson Formation. The "Contact Ledge" at the top of the section was assigned to the Mississippian. Shortly thereafter Holland (1952) defined the Leatham Formation beneath the Madison Limestone in Blacksmith Fork Canyon and reported that the "Contact Ledge" contained Devonian fossils. Later, Rigby (1959) related the high detrital content /. Stewart Williams, Department of Geology, Utah State University, Logan, Utah 84321. in the Beirdneau Sandstone Member to Late Devonian diatrophism in the Stansbury Range and adjacent areas, southwest of the Bear River Range outcrops. In 1964 Mullens and Izett published a geologic map of the Paradise quadrangle, which includes the Devo- nian outcrops in Blacksmith Fork Canyon, and with the accompanying text is a figure of a measured section of the Hyrum and Beirdneau members. In the same year, the writer and Michael Taylor restudied the Water Canyon Formation and added considerable detail about the Bear River Range sections of that formation (Williams and Taylor, 1964). In the past three years, Beus (1965, 1968) has re- ported his studies on the Paleozoic section in Samaria Mountain, northwest of the Bear River Range expo- sures, and has defined the Samaria Limestone Member of the Jefferson Formation. Recent regional studies of the Devonian—namely, those of Benson (1966), Poole and others (1967), and Sandberg and Mapel (1967)—have shown that the Hyrum Dolomite and Beirdneau Sandstone are recog- nizable over a considerable area in southeastern Idaho and north-central Utah and that they probably are sufficiently different from time-equivalent sections to the northeast and north in Wyoming and Montana to warrant their elevation to formational rank. Because of the growing interest in the Hyrum Dolo- mite and Beirdeau Sandstone (hereafter to be called the Hyrum Formation and the Beirdneau Formation), it seemed desirable to examine in detail the well- exposed sections of these strata in Blacksmith Fork Canyon, to describe them more accurately, to search them more diligently for fossils, and, on the basis of the information obtained, to attempt to describe the conditions under which they were deposited. The section of the Hyrum Formation reported in this paper was measured in sees 1 and 12, T 10 N, R 219 220 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY 1 E, and that of the Beirdneau Formation in sees 7 and 8, T 10 N, R 2 E (SLB&M). Other sections that received the same examination are in Logan Canyon, sec 23, T 12 N, R 2 E (SLB&M); in old Laketown Canyon, Rich County, Utah, sees 17 and 18, T 12 N, R 6 E (SLB&M); in Mahogany Canyon north of Morgan in Morgan County, Utah, sec 24, T 4 N, R 2 E (SLB&M) ; along the norfh side of Portage Canyon, Box Elder County, Utah, sees 33 and 34, T 15 N, R 4 W (SLB&M) ; and north of Wide Hollow, west of Grace, Caribou County, Idaho, sec 35, T 9 S, R 39 E, and sec. 2,T10S,R39E(BB&M). The beds in these sections were measured and then described in field notes, with chips taken for laboratory study and analysis. All samples taken from the Black- smith Fork section were analyzed for calcium, mag- nesium, and acid-insoluble residue, with the residues retained for microscopic examination. Rock chips from the other sections that appeared to offer differences in kind also were analyzed. The color terminology is that of the "Rock-Color Chart" distributed by the Geological Society of Amer- ica. The nomenclature used in describing the rocks, which are mostly carbonate with a varying admixture of detrital material, is that of Guerrero and Kenner (1955) and Mather (1955). The certain presence of Devonian rocks in northern Utah has been known since E. M. Kindle (1908) pub- lished his study of the Jefferson Limestone from its type section in Montana to the outcrops in the Bear River Range. Geologists at Utah State University long have been aware that the local section contains thick and well-exposed representatives of the Devonian, and one of the earliest master's theses completed in the department was devoted to these rocks (Cooley, 1928). The writer is indebted to Mr. James Eliason for his excellent work as a field assistant; to the University Re- search Council, Utah State University, for funds to support the field and laboratory work; to Dr. G. Arthur Cooper of the United States National Museum for assistance in identifying brachiopods; to Dr. Gil- bert Klapper of the University of Iowa for processing samples and identifying the conodonts; and to Mr. Robert H. Denison of the Field Museum of Natural History for identifying the fishes. Hyrum Formation GENERAL LITHOLOGY.—The Hyrum Formation is generally a cliff-former, in contrast to the weaker FEET 932 UNIT UPPER CARBONATE- 18 DETRITUS MEMBER UNIT UPPER DOLOMITE 17 MEMBER UNITS LOWER CARBONATE- 10-16 DETRITUS MEMBER UNITS LOWER 00L0MITE 5-9 MEMBER TTT UNITS I "4 SAMARIA MEMBER HYRUM FORMATION FIGURE 1.—Stratigraphic diagram of the Hyrum Formation in Blacksmith Fork Canyon showing lithology and members. Beirdneau Formation, which, in most localities, forms smooth slopes. However, there are few massive beds in the formation, and thin to medium bedding is com- mon. Dark and medium gray are the common colors, with the argillaceous and sandy beds showing such colors as light gray, yellowish gray, and light olive gray. Dolomite is the common rock in the formation but there are beds of limestone, dolomitic limestone, and calcific dolomite. The detrital component is fine- grained, being generally clay, silt, or fine sand. Only two relatively thin beds of sandstone occur in the for- mation, and one of these is dolomitic. The Blacksmith Fork section is readily divisible into five members (Figure 1), which may be traced northwestward into the Portage Canyon section. NUMBER 3 221 40.8 25.6 29.0 16.8 27.9 9.3 3. 5 3.2 The Hyrum Formation, exposed in Blacksmith Fork Canyon, Cache County, Utah, Sees 1, 12, T 10 N, R 1 E (SLB&M)—by far the best section exposed in the region—is as follows: Thickness Feet Meters 134.0 Unit and Member UPPER CARBONATE-DETRITUS MEMBER 18. Limestone: massive, forming ledges up to 10 feet thick, medium gray to medium light gray, very finely crystal- line to compact, some with oolitic tex- ture; grading above and below into interbeds of dolomitic limy sandstone; thin-bedded, wavy-bedded, yellowish gray, very fine-grained; halite casts, ripple marks, dessication cracks. De- formation and brecciation common, with limestone-sandstone breccias. UPPER DOLOMITE MEMBER 17. Dolomite: thin- to medium-bedded, medium dark gray to medium gray, 84.0 very finely crystalline. LOWER CARBONATE-DETRITUS MEMBER 95.0 16. Dolomitic limestone: thin- to medium- bedded, medium gray to medium light gray, very finely crystalline to aphanitic; near base of unit several beds of dolo- mite, medium dark gray, very finely crystalline; some thin beds of limestone; considerable intraformational breccia; ledges weather with rounded profiles. 55.0 15. Calcific dolomite: thin- to medium- bedded, dark gray to medium gray, finely crystalline; some thin, more sandy beds of arenaceous calcitic dolomite. 91.5 14. Same as 13. Some of the carbonate beds are limestone. Some interbeds are silty limestone. 30.5 13. Dolomite, in massive ledges, which in detail are laminated with contorted bedding and intraformational breccia; medium dark gray to medium gray, very finely crystalline; interbeds (silty dolomite) which separate ledges thin- bedded to laminated, medium gray. Interstratal contortions including "fists" and "roll-ups" developed here. 11.5 12. Sandstone: laminated to medium- bedded, light gray to yellowish gray, generally fine-grained; subparallel cross-lamination, sharp contacts to beds above and below. 10.5 11. Dolomite: thick-bedded, medium gray weathering light gray, medium crys- talline. 10. Silty or argillaceous dolomite: lami- nated, wavy bedding, light gray weath- ering dusky olive; thin interbeds of sandstone; sharp wavy contacts to dolomites above and below. 3.0 0.9 LOWER DOLOMITE MEMBER 9. Dolomite: medium- to thick-bedded, medium gray, finely crystalline; lami- nations not visible in thick beds. 50. 0 15. 2 8. Calcitic dolomitic sandstone: thin- bedded to laminated, light olive gray, very fine-grained; grades into dolomite below through medium light gray silty dolomite and into dolomite above through laminated, shaly, light olive gray silty dolomite; edgewise conglom- erate and intraformational breccia common. 7. 7 2. 3 7. Dolomite: cliffs massive, but in detail the beds appear thin-bedded or lam- inated; medium dark gray weathering medium gray, finely crystalline. 110.0 33.5 6. Arenaceous calcitic dolomite: thin- bedded, yellowish gray, fine-grained; grades above and below into adjacent dolomites through beds of laminated, light gray, silty or argillaceous dolo- mite. Ripple marks; microcontortions in some laminated beds. 4.0 1.2 5. Dolomite: medium- to thick-bedded, medium dark gray, finely crystalline. 111. 0 33. 8 SAMARIA MEMBER 4. Silty calcitic dolomite: thin- to medium- bedded, medium dark gray weathering medium gray. 53. 0 16. 2 3. Dolomite: medium-bedded, medium gray weathering light gray, finely crystalline: thin partings of silty or argillaceous dolomite, laminated, light gray; with microdeformation. 26.3 8.0 2. Dolomitic limestone: thin- to medium- bedded, dark gray to medium dark gray, weathering light olive gray, com- pact to coquinoidal. In lower part silty or argillaceous dolomite: thin- to medi- um-bedded, medium gray weathering light olive gray. Rennselandia midway up unit. 45.0 13.7 1. Limestone: medium- to thick-bedded, medium gray, brecciated in part. 10. 4 3. 2 932.4 284. 1 TOTAL LOWER BOUNDARY.—The basal unit of the Hyrum Formation, a rounded ledge of gray limestone, rests upon the lighter-colored sandy sediments of the Grassy Flat Member of the Water Canyon Formation. The contact, generally fairly sharp, is to be seen in the al- 222 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY cove beneath the basal limestone, where the weaker Water Canyon sediments have weathered back. At some localities, however, where the basal limestone is brecciated and where there is also intraformational breccia in the uppermost beds of the Water Canyon, it is more difficult to select the boundary, and addi- tional study of the breccias above and below the con- tact is needed. One such locality is at Grassy Flat in Logan Canyon (Williams and Taylor, 1964, p. 47). Another is the Portage Canyon section where Williams and Taylor (1964, p. 50) and Beus (1968, p. 789) se- lected different positions for the boundary that are 134 feet apart in the section. The difficulty of selecting the formational boundary at Portage is heightened by the fact that the upper part of the Grassy Flat Member contains well-bedded dark gray dolomites, with traces of marine fossils, that resemble the beds of the Hyrum Formation. SAMARIA MEMBER.—The basal member of the Hy- rum Formation can be distinguished by its generally lighter color and thinner beds, which contrast with the darker dolomite beds above that form the principal cliffs on any outcrop of the formation. The lower part of the member is limestone or calcitic limestone, and here fossils are common. Beus (1965) has described part of the fauna from this member, which he named the Samaria Limestone Member of the Jefferson For- mation (Beus, 1968). The writer proposes to extend the use of this name to the sections in the Bear River Range. In Blacksmith Fork Canyon the lower half of the member is limestone and dolomitic limestone, which grades upward into silty calcitic dolomite. In Logan Canyon the member has essentially the same develop- ment as in Blacksmith Fork Canyon except that the exposures are somewhat better and fossils are more abundant. In Portage Canyon the member reaches its maximum thickness (338 feet) in the study area, being about twice as thick as in Logan and Blacksmith Fork Canyons. Here the member appears as a succession of three limestone ledges (Plate 1: figure 6) separated by two weaker units of thin-bedded silty calcitic dolo- mite or silty dolomitic limestone, dusky yellow or yel- lowish gray in color. These erode back, and bring the limestone beds into relief. The ledges of limestone grade from limestone into silty dolomitic limestone. As stated above, the writer has not drawn the bound- ary between the Water Canyon Formation and the Hyrum Formation in the Portage Canyon section at the same position as has Beus (1968, p. 789, fig. 5). According to the writer's interpretation, the Samaria Limestone Member of Beus would be 508 feet thick in the State Line section. In Portage Canyon the writer's measurement of the Samaria Member is 338 feet. The Samaria Member is not recognizable in the Laketown Canyon section. AGE OF THE SAMARIA MEMBER.—The Samaria Limestone Member at the base of the Hyrum Forma- tion is by far the most fossiliferous part of the forma- tion. Fossils are rare in the sandstone-carbonate and in the dolomite members above it. Some fossils have been recovered from the Samaria Limestone Member in each of the sections studied except the Laketown Can- yon section and the Morgan section, where the Hyrum is represented by the Lower Dolomite Member only. The Portage Canyon section, some two miles south of Beus's (1965) State Line section, produced the most abundant fossils. These came, of course, from the southward continuation of the type Samaria Lime- stone. Beus (1965, pp. 23, 24), who discussed at some length the probable age of the fauna of the Samaria Limestone, concluded that the Middle Devonian- Upper Devonian boundary might well occur in the fos- siliferous limestone unit, making the lower 300 feet of the Hyrum Formation Middle Devonian in age. In summary reports, Sandberg and Mapel (1967, p. 857) and Poole and his co-authors (1967, p. 902) agreed. The writer's faunal list for the Samaria Limestone Member is not long, despite assiduous collecting: Thamnopora sp. Zaphrentid coral Lingula sp. Rennselandia cf. R. missouriensis (Swallow) Schuchertella sp. Atrypa oneidensis Beus Tenticospirifer utahensis (Meek) Allanaria engelmanni (Meek) Ambothyris utahensis Beus Cyrtina sp. Reticularia sp. Raphistomina (?) sp. Hormotomina (?) sp. Bembexia (?) sp. Naticopsis sp. Aviculopecten cf. A. cancellatus (Hall) Proetus sp. Icriodus sp. Icriodus nodosus, sensu lato Polygnathus decorosus, sensu lato Atopacanthus dentatus Hussakof and Bryant Ptyctodus sp. Dipterus sp. NUMBER 3 223 The item in the above list that appears to be im- portant is Rennselandia cf. R. missouriensis (Swal- low) . According to Cloud (1942, p. 96), Rennselandia and Stringocephalus are coexistent in Europe and parts of North America, and Rennselandia is essentially as good a marker of the Stringocephalus zone as String- ocephalus itself. This is confirmed by Dr. G. Arthur Cooper (personal communication), who states that Rennselandia is unknown above the Middle Devonian. With Rennselandia occurring in the same beds as Allanaria allani in both Logan Canyon and Portage Canyon, it appears that a strong argument can now be made for the Middle Devonian age of the Samaria Limestone Member. LOWER DOLOMITE MEMBER.—The Lower Dolomite Member of the Hyrum Formation is nearly 300 feet thick and forms the conspicuous dark gray cliffs that characterize the formation (Plate 1; figure 1). Three- fourths of the way up this member is an eight-foot bed of calcitic dolomitic sandstone, probably representing one of the detrital units recognized by Benson (1966, p. 2572) to the north and east. A four-foot bed of arenaceous calcitic dolomite, a third of the way up the member, also interrupts the expanse of dark gray dolomite, and it may be correctable in other sections as a detrital unit. In Logan Canyon the Lower Dolomite Member has approximately the same expression and thickness as in Blacksmith Fork Canyon, and the same is true in Portage Canyon. The member is well-developed in Laketown Canyon. In Mahogany Hollow, north of Morgan, there is 135 feet of mostly dark gray dolomite between the unconformity with the Cambrian and the base of the Beirdneau Formation. These dolomite beds appear to represent the Lower Dolomite Member. LOWER CARBONATE-DETRITUS MEMBER.—The Lower Carbonate-Detritus Member of the Hyrum Formation is notable in the Blacksmith Fork section (Plate 1: figure 2) for the presence of an eleven-foot sandstone bed, the most conspicuous detritus unit in the Hyrum Formation; for the presence of rounded, lighter-gray cliffs separated by less-resistant units of silty or argillaceous light gray, yellowish gray, or dusky olive, dolomite or limestone, that are involved in in- trastratal deformation on a large scale; and for the presence of a considerable amount of limestone. This same member is readily recognized in Logan Canyon (Plate 1: figure 3), between the upper and lower dolomite members, but there the "fists" and "roll-ups" are not so well developed. In Portage Canyon the same beds appear to be present, but the amount of work the writer has been able to devote to exposures in that sec- tion has not resulted in their positive identification. The Laketown Canyon section is partly covered, and the boundaries between the Lower Dolomite Member and the Upper Dolomite Member and the interven- ing detritus-rich beds are not clearly seen, but the presence of the Lower Carbonate-Detritus Member is evidenced by the presence of a 10-foot sandstone bed, evidently the same as Unit 12 of the Blacksmith Fork section. No beds that can be equated to this member appear in the thin Mahogany Canyon section. INTRASTRATAL DEFORMATION.—Intrastratal defor- mation is present to some degree throughout the Hy- rum and Beirdneau Formations, from beds in the Sam- aria Member to those just below the "Contact Ledge." The structures vary in size from a few inches to tens of feet. Perhaps the most common form is that of a broken-through, slightly overturned fold, a "fist," but there are also Z-folds, and "roll-ups," where the beds have been turned into a vertical position, and then truncated, perhaps produced when a fist-like structure moved forward and away, leaving only the lower part of the truncated fold (Plate 1: figures 2, 3). The largest of these structures in the Blacksmith Fork section are present in the Lower Carbonate-Detritus Member, and where the section was measured (Plate 1: figure 2), particularly in Unit 13, but there are similar structures in the Upper Carbonate-Clastic Member—well seen in Logan Canyon, and, on a smaller scale, in the Samaria Member at Portage and in the Beirdneau Formation in Blacksmith Fork and Logan Canyons. These structures are marked in gray, medium-thick beds of carbonate rock that are separated by thin- bedded units of silty limestone or dolomite weathering light gray or yellowish gray. The transfer of material necessary to produce the deformation appears to have taken place in these thinner beds, while the thicker- bedded, more richly carbonate rocks above accom- modated themselves to the flowage underneath. When the contorted beds are exposed to erosion, the thin beds with greater detrital content are recessed, and the thicker carbonate ledges are brought into relief. The greater mobility of the thin beds probably is due 224 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY both to a greater number of bedding planes and the greater detrital content. UPPER DOLOMITE MEMBER.—The Upper Dolomite Member repeats the dark gray cliffs of the Lower Dolomite Member, but not so high and with a sharp rectangular profile at the top, in contrast to the rounded profiles of the cliffs in the Carbonate-Detritus members above and below it (Plate 1: figure 3). The Upper Dolomite Member appears to be the equivalent of the Birdbear Member of the Jefferson as defined by Sandberg and Hammond (1958) and traced into the Blacksmith Fork section by Benson (1966, p. 2588). The thickness of 84 feet measured for this member in the Blacksmith Fork section agrees in general magnitude with the average thickness of 60 to 115 feet reported for the Birdbear Member. UPPER CARBON ATE-DETRITUS MEMBER. — The lighter gray, rounded ledges of the Upper Carbonate- Detritus Member which top the Hyrum Formation in the Blacksmith Fork section (Plate 1: figure 4) are well represented in Logan Canyon and have been recognized in both the Laketown and Portage Canyon sections. In this member the ledge-forming rock is limestone, and the weaker interbeds are dolomitic limy sandstone. Although in lithology this member repeats essentially the Lower Carbonate-Detritus Member, in- trastratal deformation is not developed to the same extent as in the lower member. Since the Birdbear Member marks the top of the Jefferson Formation over a wide area, and since the angular profile of the top of the Upper Dolomite Member in the study area is a readily recognized hori- zon that could be mapped without difficulty, there are reasons for drawing the upper boundary of the Hyrum Formation at the top of the Upper Dolomite Member. But considering the make-up of the two formations, the Upper Carbonate-Detritus Member essentially re- peats the lithology of the Lower Carbonate-Detritus Member. At the same time it is unlike any part of the Beirdneau Formation and, further, it was included in the Hyrum Formation in the original description of the Hyrum and Beirdneau members (Williams, 1948, p. 1140) where Units 7, 8, and 9 are the Upper Car- bonate-Detritus Member and Unit 6 is the Upper Dolomite Member. Also, the uppermost limestone bed at the top of the member is as mappable as the top of the Upper Dolomite Member. For these reasons the writer prefers to stay with the original definition of the members, now called formations. REGIONAL RELATIONSHIPS.—The Samaria Member, the Lower Dolomite Member, and the Upper Dolomite Member appear to have, in a general way, the same lithology as the type Jefferson Formation (Sandberg, 1965, p. 6), and in total they have about the same thickness. The difference between the type Jefferson Formation and the Hyrum Formation might be con- sidered to consist of the introduction of the Lower and Upper Carbonate Detritus members, members that are transitional to the overlying Beirdneau Formation. Mullens and Izett (1964, p. 28) and I seem to have drawn the boundary between the Beirdneau and Hy- rum formations at the same position; Benson (1966) appears to have drawn it at the *op of the upper dark dolomite cliff and to have thrown the writer's Unit 18 into the Beirdneau Formation. The carbonate-detritus members may represent surges of detrital material from the Stansbury and Uinta Uplifts to the southwest and southeast. Examination of the section in the Fish Creek Range north of Wide Hollow, Caribou County, Idaho, was disappointing in that the section is so badly faulted and so poorly exposed as to be of little use in strati- graphic studies. The light-colored dolomite in the sec- tion (Benson, 1966, fig. 12) which appears to be re- peated three times is the Card Member of the Water Canyon Formation. No fossiliferous limestones repre- senting the Samaria Member were seen. CONDITIONS OF DEPOSITION.—The Hyrum Forma- tion accumulated in a basin at the edge of the miogeo- syncline north of the Uinta Uplift (Poole and others, 1967, fig. 9). Other basins in the miogeosyncline farther west accumulated greater thicknesses of sediment dur- ing Frasnian time, perhaps with more influence from the Antler Peak orogenic belt. The sea that deposited the formation spread southward and eastward from the miogeosyncline, as reflected in the greater thick- ness of the Samaria Member in the Portage section. However, the two carbonate-detritus members in the Bear River Range sections appear to represent a more rapid accumulation of detrital material than occurred farther north in the vicinity of the type section of the Jefferson Formation. In the advancing sea, calcium carbonate mud was the principal sediment, accumulating below wave base, with frequent small impulses of fine detrital material that marked thin beds in the depositing sediment. In these waters lived an epineritic fauna of corals, brachi- pods, pelecypods, and snails; their remains were buried NUMBER 3 225 after transportation by bottom currents for short dis- tances, with the resulting rocks constituting the Sam- aria Member. A decrease in the arrival of detrital sediments ap- pears to have coincided with an increase in salinity of the sea water, resulting in the dolomitization of the detritus-free carbonate muds and the production of the Lower Dolomite Member. Dolomitization largely de- stroyed the fossils, which may have been fairly abun- dant in these beds, as they are in the undolomitized beds of the Samaria Member. Even under this new regime, there were short periods of large detrital im- port, producing two thin but notable detrital interbeds. A relatively long period of more rapid detritus arrival interrupted the deposition of the Lower Dolomite Member and originated the Lower Carbonate-Detritus Member. Near-shore depositional conditions are recorded in the 11-foot sandstone bed, Unit 12, near the base of this member. The sea had become much shallower. Detrital sediment arrived in much larger quantity, and units of carbonate mud were deposited alternately with detritus-rich interbeds. The presence of a slight gradi- ent on the shallow sea floor toward the east set the stage for the intrastratal deformation that marks this member. Before the deposition of the Hyrum Forma- tion had been completed deeper-water conditions re- turned to create the upper Dolomite Member, and shallow-water conditions were repeated to deposit the Upper Carbonate-Detritus Member. The latter indi- cates conditions transitional to those that produced the shallow-water Beirdneau Formation. Beirdneau Formation GENERAL LITHOLOGY.—The Beirdneau Formation was first conceived and described by the writer as con- sisting of thin-bedded sandstone (Williams, 1948, p. 1140). The type locality was the Beirdneau Peak trail in sees 16 and 21, T 12 N, R 2 E (SLB&M), north of Logan Canyon. In describing the section in Blacksmith Fork, Mullens and Izett (1964, p. 27) and Benson (1966, p. 2596) pointed out that dolomite is an im- portant part of the formation. The Leatham Forma- tion disconformably overlies the Beirdneau Formation in Blacksmith Fork Canyon. The Beirdneau Formation (Figure 2), exposed in Blacksmith Fork Canyon, Cache County, Utah, Sees 7, 8, T 10 N, R 2 E (SLB&M), is as follows: FEET 2000 "CONTACT LEOGE" UNITS UPPER CARBONATE 29-30 MEMBER UNITS SANDSTONE 25-28 MEMBER UNITS LOWER CARBONATE 19-24 MEMBER BEIRDNEAU FORMATION FIGURE 2.—Stratigraphic diagram of the Beirdneau Forma- tion in Blacksmith Fork Canyon showing lithology and members. Thickness Feet Meters Unit and Member UPPER CARBONATE MEMBER 7.6 25 30. "Contact Ledge" of limestone, inter- bedded with some arenaceous calcitic dolomite. 61.0 200 29. Silty or argillaceous calcitic dolomite: thin-bedded, medium light gray and light olive gray, very fine-grained; ripple marks, desiccation cracks; about midway in unit, calcitic dolomite, me- dium gray to medium dark gray. SANDSTONE MEMBER 41. 1 135 28. Dolomitic sandstone and arenaceous dolomite as below, but with some beds of limestone, medium gray, compact with stringers of white chert; also, some beds of dolomite, medium gray, finely crystalline. 226 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Thickness Feet Meters Unit and Member 27 Dolomitic sandstone and arenaceous dolomite as in Units 25 and 26, with ripple marks and desiccation cracks common. In parts dolomite, laminated, medium gray, very fine-grained. 160 48. 8 26. Dolomitic limy sandstone in beds of varying thickness, yellowish gray, medium grained, with strongly cross- bedded strata up to one foot thick; ripple marks common; some arenaceous calcitic dolomite as in Unit 25. 93 28. 4 25. Dolomitic sandstone: thin-bedded, light olive gray weathering yellowish gray, ripple marks common. Upper 15 feet arenaceous dolomite, medium light gray weathering yellowish gray. 155 47. 2 LOWER CARBONATE MEMBER 24. Calcitic dolomite: thin-bedded with bedding planes wavy in part, straight in part; medium gray weathering light olive gray, very fine-grained to compact. 10 3.0 23. Arenaceous dolomite: uniformly thin- bedded, light gray weathering yellow- ish gray, very fine-grained with scattered larger grains. Makes blocky cliffs; upper portion medium-bedded. 105 32. 0 22. 21 20. 19. Calcitic dolomite: thin to medium- bedded, medium light gray, fine- grained. Parts show fiowage and contor- tion. 18 5. 5 Calcitic dolomitic sandstone: medium to thick-bedded, light gray weathering yellowish gray, grains a mixture of coarse and medium with fine. Marked cross-bedding, some on a scale of a few inches and some on a scale of several feet. Forms prominent cliff. 56 17. 1 Arenaceous calcitic dolomite: thin- bedded, wavy-bedded, light olive gray weathering yellowish gray, very fine- grained to silty; halite casts, ripple marks, desiccation cracks common. 65 19. 8 Arenaceous calcitic dolomite: thin- bedded, wavy-bedded, light olive gray weathering yellowish gray, very fine- grained to silty; halite casts, iipple marks, desiccation cracks common. Unit is topped by 2-foot bed of dolo- mite: medium gray; finely crystalline. 65 19.8 1087 331.3 The thickness of the Beirdneau Formation is of immediate interest because it varies substantially in the course of a few miles along the north side of Black- smith Fork Canyon. Mullens and Izett, measuring in sec 1, T 10 N, R 1 E, obtained 699 feet; Benson, at the same locality, recorded 777 feet. The writer, seeking excellent exposures lower in the canyon, measured the lower part of the section in NE^4 sec 7 and the remain- der in NW/4 sec 8, T 10 N, R 2 E. Here the total thickness is 1,087 feet. Because of these differences, the writer's measurements were carefully checked. East- ward at the forks of Blacksmith Fork Canyon, in sec 3, T 10 N, R 2 E the thickness is 937 feet. Since at all three localities along the north side of Blacksmith Fork Canyon the formation is capped by the "Contact Ledge" beneath which are other parts of the Upper Carbonate Member, it must be concluded that the formation actually was deposited to a greater thickness in the middle section—opposite the reservoir in the canyon—than to the west or the east. Other thicknesses (by the writer's measurement) of the Beirdneau Formation in the study area are Logan Canyon, 651 feet; Laketown Canyon, 524 feet; and Gardner Canyon, Idaho (Beus, 1968, p. 791), 808 feet. At the latter two localities the "Contact Ledge" lime- stone was not recognized, but there was an increase in carbonates in the upper part of the section, indicating the presence of at least part of the Upper Carbonate Member. LOWER AND UPPER CARBONATE MEMBERS.—The lower 319 feet of the formation is distinguishable from the thicker middle member (Figure 2) by the presence of several units of medium gray or medium light gray dolomite or calcitic dolomite that are thicker-bedded and more resistant to erosion than the thin-bedded arenaceous calcitic dolomite or calcitic dolomitic sand- stone that makes up most of the formation. These differences justify the designation of the member as the Lower Carbonate Member (Plate 1: figure 4). Again in the upper 225 feet of the formation the rocks become generally grayer and the ledges steeper and more resistant as the carbonate content of the formation increases (Plate 1: figure 5). Hence, the upper member has been designated the Upper Car- bonate Member. Here the characteristic rock type is silty or argillaceous calcitic dolomite, medium light gray or light olive gray. SANDSTONE MEMBER.—The middle portion of the formation, where the clastic content is highest, has been designated the Sandstone Member (Plate 1: figure 5). Here thin-bedded dolomitic sandstone and dolomitic limy sandstone predominate over arenaceous dolomite and over the few thin beds of dolomite and NUMBER 3 227 PLATE 1.—Members of Hyrum and Beirdneau Formations. Figure 1.—Blacksmith Fork Canyon. Between markers, from bottom to top, are Lower Dolomite Member, Lower Carbonate-Detritus Member, Upper Dolomite Member, Upper Carbonate-Detritus Member of Hyrum Formation. 2. Blacksmith Fork Canyon. Between markers is Lower Carbonate-Detritus Member of Hyrum Formation with intrastratal deformation. 3. Logan Canyon, Hyrum Formation. Between markers is Upper Dolomite Member; beneath it is Lower Carbonate-Detritus Member with intrastratal deformation. 4. Blacksmith Fork Canyon. Between markers is Upper Carbonate- Detritus Member of Hyrum Formation; above it is Lower Carbonate Member of Beirdneau Formation. 5. Blacksmith Fork Canyon. Between markers is Upper Carbonate Member of Beirdneau Formation (with "Contact Ledge" at top); below it is Sandstone Member. 6. Portage Canyon. The three ledges of the Samaria Member. 372-386 0—71- -16 228 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY limestone that are present. Unit 26, with its medium- grained sandstone, strongly cross-bedded in part, marks the maximum deposition of clastic material in the formation. This unit probably correlates with the 60-foot orthoquartzite unit that Beus (1968, p. 790) recognized from Gardner Canyon, Idaho to Little Mountain, Utah. REGIONAL RELATIONSHIPS.—Despite variations in thickness, the Beirdneau Formation appears to main- tain its tripartite character from the Bear River Range localities northward and westward into Idaho. All three members are recognizable in the Portage and Gardner Canyon sections. In the Laketown Canyon section to the east, the midpart of the section representing the Sandstone Member of the Beirdneau Formation is occupied by massive beds of limestone intraformational breccia. The limestone is silty and light brown to pale yellowish brown, and it may well be a solution breccia repre- senting the Logan Gulch Member of the Three Forks Formation. At the upper and lower parts of the section, the lithology is more like that in the Blacksmith Fork section. The Laketown Canyon section then is transi- tional between the facies of western Wyoming and those of the Beirdneau Formation. From this point northeastward, the term Three Forks is appropriate, but westward the term Beirdneau can replace it. In Mahogany Canyon north of Morgan, the Beird- neau Formation is only about 200 feet thick, but it consists largely of sandstone and probably represents the shore zone of the late Upper Devonian sea at its maximum transgression onto the Uinta Uplift. It appears that the Lower Carbonate Member followed by the Sandstone Member of the Beirdneau Formation may reflect the filling of the intracratonic basin, as it happened in Montana (Sandberg and Mapel, 1967, p. 873), and the spread of evaporite precipitation from the type area of the Logan Gulch Member of the Three Forks Formation. Evaporite precipitation reached as far south as Laketown Can- yon. The climax of the recession may be represented in the strongly cross-bedded sandstone of Unit 26. The upper Carbonate Member may reflect the wide- spread transgression that produced the Trident Mem- ber of the Three Forks Formation (Sandberg and Mapel, 1967, p. 874). More particularly, the two lower members of the Beirdneau Formation would represent the argillaceous dolomite facies of the Logan Gulch Member (Benson, 1966, p. 2595) modified by the addition of more and coarser clastic material. The Upper Carbonate Member would represent the carbonate facies of the Trident Member (Benson, 1966, p. 2597). CONDITIONS OF DEPOSITION.—The thin-bedded ar- gillaceous and silty dolomites and dolomitic sandstones that make up the Beirdneau Formation appear to rep- resent generally fine-grained, probably partly autoc- thonous and partly allocthonous sediments deposited in a very shallow, epineritic area which sank steadily but slowly. The bottom of this very shallow sea was occasionally exposed by neap tides and off-shore winds, causing the formation of dessication cracks, and the growth of skeletal halite crystals in the fine muds. Most of the time, moderate to weak currents in the shallow water, probably largely of tidal origin, formed ripple marks and from time to time received greater amounts of noncarbonate clastic material, resulting in more silty and sandy beds. Disturbance of the bottom produced intraformational breccias and microdeformation dur- ing storms or particularly strong tides. Continuous ar- rival of sedimentary material and frequent disturbance of the bottom by tidal currents probably account for the lack of fossils. The bottom of the Beirdneau sea was generally a marine desert (Bowsher, 1967, p. 341). During the deposition of Unit 26 the currents were much stronger, the supply of sand was much greater, and marked cross-bedding was produced. In this unit thin beds of light-colored, structureless, medium- grained sand may represent discontinuous deposits of aeolian origin and mark the presence of the littoral zone. The fine clastic material probably came to this shal- low coastal area from the south and southwest, where and areas existed in Beirdneau time (Rigby, 1959, p. 213) and where the shoreline was never more than a few tens of miles away. AGE OF THE BEIRDNEAU FORMATION.—The Beird- neau and Hyrum Formations are comformable throughout the study area, with the possible exception of the Morgan section. At all other localities deposition appears to have been continuous from Hyrum into Beirdneau time and the lithology of the Upper Carbonate-Detritus Member is transitional to that of the Beirdneau Formation. The Middle Devonian (Givetian) age of the Sa- maria Member of the Hyrum Formation has been dis- NUMBER 3 229 cussed. Above this basal member the writer has not collected any fossils except at the very top of the Beird- neau Formation, where Cyrtospirifer and a rhynchonel- lid brachiopod are common. These represent a fauna which, according to Sandberg and Mapel (1967, p. 869), is equivalent to that of the Cyrtospirifer monti- cola Zone. Thus, the top of the Beirdneau is lower Famennian; Frasnian and earliest Famennian times are represented in the Hyrum Formation above the Samaria Member and in the Beirdneau Formation. Literature Cited Benson, A. L. 1966. Devonian Stratigraphy of Western Wyoming and Adjacent Areas. American Association of Petroleum Geologists Bulletin, 50:2566-2603, 16 figures. Beus, S. S. 1965. Devonian Faunule from the Jefferson Formation, Central Blue Springs Hills, Utah-Idaho. Journal of Paleontology, 39:21-30, plates 9, 10, 4 figures. 1968. Paleozoic Stratigraphy of Samaria Mountain, Idaho-Utah. American Association of Petroleum Geologists Bulletin, 52:782-808, 10 figures. Bowsher, A. L. 1967. Ocean Tides as a Geologic Process, in Curt Teichert, and E. L. Yochelson, Essays in Paleon- tology and Stratigraphy. Department of Geology, University of Kansas, Special Publication, 2:319- 348, 12 figures. Cloud, P. E., Jr. 1942. Terebratuloid Brachiopoda of the Silurian and Devonian. Geological Society of America Special Papers, 38:1-182, 26 plates, 17 figures. Guerrero, R. G, and C. T. Kenner 1955. Classification of Permian Rocks of Western Texas by a Versenate Method of Chemical Analysis. Journal of Sedimentary Petrology, 25(1) :45-50. Hague, A., and S. F. Emmons 1877. Descriptive Geology. United States Geological Ex- ploration of the 40th Parallel, 2: 1-890, illustrated. Holland, F. D., Jr. 1952. Stratigraphic Details of Lower Mississippian Rocks of Northeastern Utah and Southwestern Montana. American Association of Petroleum Geologists Bul- letin, 36:1697-1734, 17 figures. Kindle, E. M. 1908. The Fauna and Stratigraphy of the Jefferson Lime- stone in the Northern Rocky Mountain Region Bulletins of American Paleontology, 4:3—39. Klapper, G. 1966. Upper Devonian and Lower Mississippian Cono- dont Zones in Montana, Wyoming and South Da- kota. University of Kansas Paleontological Con- tributions, Paper 3:1-43, 6 plates, 2 figures. Mather, K. K. 1955. Terminology of Limestone and Related Rocks—An Interim Report. Journal of Sedimentary Petrology, 25 (4): 304-305, 1 figure. Mullens, T. E., and G. A. Izett 1964. Geology of the Paradise Quadrangle, Cache County, Utah. United States Geological Survey Bulletin, 1181-S:S1-S32, 1 plate, 1 figure. Pettijohn, F. J. 1957. Sedimentary Rocks. 718 pages. New York: Harper and Brothers. Poole, F. G; D. L. Baars; H. Drewes; P. T. Hayes; K. B. Ketner; E. D. McKee; C. Teichert; and J. S. Williams 1967. Devonian of the Southwestern United States, in D. H. Oswald (editor), International Symposium on the Devonian System, Calgary, Alberta, Sep- tember 1967, 1: 879-912, 10 figures. Calgary, Al- berta, Society of Petroleum Geologists. Rigby, J. K. 1959. Upper Devonian Unconformity in Central Utah. Geological Society of America Bulletin, 70:207- 218, 5 figures. Sandberg, C. A. 1965. Nomenclature and Correlation of Lithologic Sub- divisions of the Jefferson and Three Forks Forma- tion of Southern Montana and Northern Wyoming. United States Geological Survey Bulletin, 1194—N: NI-N18, 3 figures. Sandberg, C. A., and C. R. Hammond 1958. Devonian System in Williston Basin and Central Montana. American Association of Petroleum Geol- ogists Bulletin, 42:2293-2334, 8 figures. Sandberg, C. A., and W. J. Mapel 1967. Devonian of the Northern Rocky Mountains and Plains, in D. H. Oswald (editor), International Symposium on the Devonian System, Calgary, Al- berta, September 1967, 1:843-877, 10 figures. Calgary, Alberta, Society of Petroleum Geologists. Shrock, R. R. 1948. Sequence in Layered Rocks, 507 pages. New York: McGraw-Hill. Williams, J. Stewart 1948. Geology of the Paleozoic Rocks, Logan Quadrangle, Utah. Geological Society of America Bulletin, 59: 1121-1163, 5 plates, 2 figures. Williams, J. Stewart, and M. E. Taylor 1964. The Lower Devonian Water Canyon Formation of Northern Utah. University of Wyoming Contribu- tions to Geology, 3(2): 38-53, 3 plates. Ellis L. Tochelson A New Late Devonian Gastropod and Its Bearing on Problems of Open Coiling and Septation ABSTRACT Nevadaspira cooperi is described from beds of late Frasnian age in the upper Devils Gate Limestone exposed in the southern part of the Spotted Range, Nye County, Nevada. The genus, so far, is known from a single species, confined to the type locality. Specimens of N. cooperi have characteristics atypical of gastropods in that the whorls are not in contact and they contain numerous septa. Open coiling may come about through one or more different geometries of growth. There may be a mini- mum limit to the logarithmic spiral in open coiling, because shell forms that expand rapidly are unknown. The septa in N. cooperi are more closely spaced than in any other known genus. The whole question of why septa are produced is debatable. It may be that septa strengthen the shell, but there are so many septa in Devonian specimens that they may have served to keep the body relatively small as the shell grew. The open- coiled gastropods could have been sedentary in habit. In 1965, during stratigraphic investigations at the Nevada Test Site, F. G. Poole, United States Geolo- gical Survey (USGS), discovered a most interesting fossil mollusk. The specimens superficially resemble cephalopods, but are actually atypical gastropods which are open coiled and multiseptate. These characters are unusual enough to warrant the naming of a new genus. Ellis L. Yochelson, United States Geological Survey, Room E-317, United States National Museum, Washington, D.C. 20242. Publication authorized by the director, United States Geological Survey. Appreciation is expressed to Prof.-Dr. H. K. Erben, Institut fiir Palaontology, Bonn, Germany; to Prof.- Dr. Ulrich Jux, Geologisches Institut, Cologne, Ger- many; and to Dr. William Ball, Keeper of Palaeon- tology, British Museum (Natural History), for lend- ing specimens of Straparollus (Serpulospira) under their charge for comparison. Drs. Robert M. Linsley, Colgate University, and David M. Raup and Steven Stanley, both of the University of Rochester, discussed the problems associated with this unusual genus. It is particularly appropriate that a Devonian form be named for G. Arthur Cooper, of the United States National Museum (USNM). Occurrence and Age A few specimens of this genus were first collected in the Spotted Range, within the Mercury 7/2 -minute quadrangle, in Nye County, Nevada, during 1965. Because the area contains virtually no human habita- tions or artificial landmarks, a locality description is difficult. Geographic coordinates of the locality, USGS 7601-SD, are 36°39'05" N and 115°54'44 W. Speci- mens are in a dense, dark-gray, coarse-grained lime- stone in the Devils Gate Limestone, 110 to 115 feet below the top of the formation. They are exposed on a dip slope and practically none weathers free. The shell adheres tightly to the matrix and exfoliates from the steinkern during weathering or preparation. In spite of the difficulty in obtaining specimens from this massive limestone, Poole in 1966 was able to obtain sufficient material on which to base a meaningful con- cept of the species. Without his efforts, this study would 231 232 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY not have been possible. Specimens of this taxon have not yet been collected outside the type locality. This fossil is so distinctive that the genus may be easily identified in the field, and Poole (personal communica- tion, 1968) reports its probable occurrence in the Pahranagat Range, Lincoln County, Nevada; eventu- ally it may be a useful fossil for local correlation. Stratigraphically higher beds at this locality contain numerous brachiopods. J. Thomas Dutro (personal communication, 1968) identifies Cyrtospirifer cf. C. portae Merriam as the principal species in these beds. Rare athyrids, "Cleiothyridina" cf. "C." angelicoides (Merriam), also are present. Dutro further states: "This assemblage represents the Cyrtospirifer zone of Merriam in a broad sense. It is most likely of late Frasnian age." According to Poole (personal commu- nication, 1968), Cyrtospirifer occurs throughout a "clayey limestone unit" at the top of the Devils Gate Limestone. He notes that the multiseptate gastropods occur near the boundary between the Cyrtospirifer- bearing unit and an underlying unit of cliff-forming limestone (see Poole and others, 1967, p. 885, fig. 26, column 12.) Superfamily EUOMPHALACEA de Konnick, 1881 Family EUOMPHALIDAE de Konnick, 1881 Nevadaspira, new genus DIAGNOSIS.—Large, very slightly hyperstrophic, open-coiled euomphalid gastropods; whorl section suboval with a greatly thickened upper angulation; early part of whorls containing numerous concave septa. TYPE-SPECIES.—Nevadaspira cooperi new species. DISCUSSION.—The hyperstrophic open coiling im- mediately differentiates this new genus from all other described Devonian gastropods. It shows a slight re- semblance to Straparollus (Serpulospira), but that taxon is orthostrophic in coiling; for comparison, see Plate 1: figures 1,2. Nevadaspira bears more similarity to Ordovician open-coiled forms but can be distinguished readily from two of the three genera that show this feature. Calaurops has a flattened whorl profile and a more rapidly expanding coil. Lytospira is more strongly trochoid in coiling. Differentiation from Ecculiom- phalus is exceedingly difficult, except for die differ- ence in age between Early Ordovician and Late Devonian. Ecculiomphalus may be conveniently dis- tinguished as having a smaller number of septa— though this may be a spurious character—and in hav- ing a generally smaller average size for specimens; the inclination of the lower part of the outer whorl face also may be greater in that genus. Rate of logarithmic expansion may be different in the two genera, but little is known of the taxonomic significance or variability of such a feature. Most Ecculiomphalus specimens are steinkerns, thus making comparison of whorl profiles impossible. This new genus is known only from the type-species. Nevadaspira cooperi, new species PLATE 1: FIGURES 3-8; PLATE 2: FIGURES 1-3 DESCRIPTION.—Hyperstrophic open-coiled gastro- pods with a prominent acute angulation in the upper whorl profile and numerous partitions within the early part of the whorl. Earliest growth stages unknown. Juvenile through mature stages open-coiled, the angle of tangency of the logarithmic spiral greater than 80°. Upper surface of shell distinctly concave; lower surface very slightly convex, expanding in essentially one plane for most of the growth and being only very slightly hyperstrophic. Upper whorl surface profile straight, inclined up- ward and outward (at an angle of nearly 30° from horizontal) to the wide bluntly rounded outer angula- tion; lower edge of angulation probably forming the periphery, outer whorl face profile below nearly verti- PLATE 1: figures 1,2.—Straparollus {Serpulospira) centrifuga (F. A. Roemer). Two specimens (X 1-5) showing variation in coiling; from Upper Givetian, Biicheler beds, at Biichel, Herrenstrunden (Bergisch Gladbach), Germany; in collection of the "Naturfreunde," Cologne, Germany. Figures 3—8.—Nevadaspira cooperi Yochelson, new species. 3a,b, Two sides of a rock (X 1) broken at essentially right angles to the plane of coiling (the white dot near the center is the calcite filling of an inner whorl) : holotype, USNM 164181. 4, Fragment of a steinkern (X 1.5) showing the whorl profile (specimen has been tilted slightly with refer- ence to figure 3); paratype, USNM 164182. 5, A natural section ( X 1) in which some septa may be observed at lower right; paratype, USNM 164183. 6, A large natural section (X 1) in which the earlier whorls are filled with recrystal- lized calcite; paratype, USNM (164184. 7, Polished section (X 1.5) of a whorl fragment; paratype, USNM 164185. 8, Thin section (X 2) of a juvenile specimen; paratype, USNM 164186. NUMBER 3 233 PLATE 1 234 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY cal for approximately half of total length, curving downward and inward below midwhorl; outer basal angulation obtuse, near 140°, basal surface essentially straight; inner basal angulation also obtuse, near 120°; profile in inner whorl face a gentle curve deviating little from a straight line inclined inward and upward to its abrupt juncture with the upper whorl face, the total length of this inner face about half of the outer face. Profile of steinkern deviating only slightly from cir- cularity, but reflecting the inner basal angulation. Shell probably of two layers: inner layer thin and relatively constant in thickness; outer layer varying markedly, being thin on the inner whorl face and the basal face, except at the basal angulation, thick along the outer face, especially along the upper one-third of its length, and very greatly thickened below the upper angulation. Penultimate and earlier whorls filled with calcium carbonate, interrupted by dark-colored septa; individ- ual septum exceedingly thin in section, concave for- ward following the arc of a circle for about one-third of its circumference, presumably a truncated hem- isphere when viewed in three dimensions; septa closely, but irregularly, spaced. Growth lines and apertural details unknown; orna- ment unknown, but presumed absent. DISCUSSION.—Although this species originally was known from half a dozen specimens, F. G. Poole subsequently collected more than two dozen additional specimens at the original locality. Because both col- lections were made by the same person from essentially one bed through a five-foot interval, all material has been combined under a single locality number. Most specimens apparently occur base downward on the bedding surface, in a matrix of dense light-gray lime- stone. Many individuals have the upper part of the whorls worn away, exposing the prominent septation seen in some weathered specimens. A few fragmentary steinkerns weather free on the outcrop (Plate 1: figure 4), but these are not particu- larly useful for study. The specimens are best studied by polished and thin sections, which give a good idea of internal details but do not indicate the cross section. Fortunately, two specimens do show the cross section (Plate 1: figure 3a,b), and much of the description of this important feature is based upon them. The specimens vary widely in preservation. For the most part, those that have the upper surface weathered away show only coarse calcite filling.the inner part of the whorl (Plate 1: figure 6; Plate 2: figure 2). However, exceptional specimens do show septa stand- ing in relief (Plate 1: figure 5). Even in the sectioned material there is considerable variation in preserva- tion. Although almost every specimen of the early whorls shows some partitioning, the number and dis- tribution of septa vary. One is led to the conclusion— expressed earlier when a study was made of a limited number of specimens of Omphalocirrus goldfussi (Yochelson, 1966, p. 44)—that diagenetic changes may alter the number of septa that may be observed, or may even destroy all trace of such features (Plate 2: figure 3). Even though no operculum is known, the species is considered to be hyperstrophic rather than sinistrally coiled. In general shape and whorl profile it is similar to most euomphalaceans, and these are conventionally oriented as normally coiled, nonsinistral gastropods. TYPES.—Holotype, USNM 164181. Paratypes: USNM 164182-8 (figured) ; USNM 164189a-z (not figured). As 17 additional specimens are too poorly preserved to have played a part in this study, they are not designated as paratypes. All of the specimens of this new species illustrated in Plates 1 and 2 are from USGS loc. 7601-SD, described above. Review of Literature No described species referable to the new genus Nevadaspira were found in the American and European literature examined. References to early and mid- Paleozoic open-coiled gastropods are scattered. With the exception of this Nevada material, all described Devonian species might be placed within Straparollus (Serpulospira), although they presently are under a variety of generic names. All the species appear to have in common early whorls in contact, orthostrophic coil- ing, and a generally oval to circular cross section. The one apparent exception is the recently named Strapa- PLATE 2.—Nevadaspira cooperi Yochelson, new species, la-c, Paratype, USNM 164187: la, thin section (X 2) showing numerous septa; lb, enlargement showing some details of the walls and septa (X 5) ; lc, the septum at the body chamber and five earlier septa (X 10). 2, A broken specimen (X 1) that has been partly smoothed to show the calcite in the early whorls; paratype, USNM 164188. 3, Enlargement (X 15) of figure 8 on Plate 1; near the center on the left wall is a septum that has been destroyed in the recrystallized calcite between more prominent septa. NUMBER 3 235 PLATE 2 236 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY rollus (Sinistrispira) (Jhaveri, 1969, p. 163), from the Early Devonian of the Carnic Alps; it is described as having an oblique oval cross section, but this feature is not illustrated and no close comparison to Nevada- spira can be drawn. Any projected revision of open- coiled species, however, should include examination of specimens as well as literature. A problem may result when coiling is close to a plane of bilateral symmetry. So far as known, open-coiled nautiloid cephalopods invariably show bilateral sym- metry, and open-coiled euomphalaceans invariably show asymmetry. The differences between these two modes of coiling may be slight, and both shell forms may become deformed after death. Some authors illus- trate only a top view of specimens, so the degree of coiling in the third dimension, if any, cannot be deter- mined. It is quite possible that a few Paleozoic open- coiled "gastropods" actually are bilaterally symmetrical cephalopods and are not correctly assigned on the class level. A second problem is loss of one or both shell layers from a fossil; this is particularly common among the euomphalaceans. In at least one species the outer shell layer is calcite and the inner layer aragonite (Yochel- son, White, and Gordon, 1967) ; if this is a superfamilial characterisitc, it might explain some of the peculiarities of preservation encountered within the group. In a classic case, Philip and Talent (1959, pp. 50-52) were able to duplicate a shell referable to the Early Devonian open-coiled Liomphalus Chapman by breaking off the outer shell layer of a specimen of Straparollus (Euomphalus) northi (Etheridge), thus demonstrating the spurious nature of Chapman's generic concept. The effect of loss of a potentially great thickness of outer shell should always be considered before assuming that a specimen is open coiled. In addition to open coiling, an inner shell may show a profile different from that of the outer shell (Yochelson, 1956, p. 216). Open Coiling of Gastropods In many introductory courses in invertebrate paleon- tology, the hard parts of gastropods are treated as shells characteristically trochoid in shape, that is, coiled in a third dimension without bilateral symmetry. In partial contrast, cephalopods are considered as bilaterally symmetrical shells which may change from straight orthocones, through curved shells to those tightly coiled. The exceptions to both generalizations seldom are emphasized. For example, shells of most bellerophon- tacean gastropods are bilaterally symmetrical; as a consequence, they occasionally are confused with cephalopods. Also, shells that deviate only slightly from this plane of symmetry may be misinterpreted as symmetrical. Raup (1966) has discussed and clarified geometry of coiling in detail. For this paper, it is appropriate merely to mention some simple generalizations concerning gas- tropod coiling. The gastropod shell may be considered as a cone expanding at a uniform rate. The cone (shell) is curved, and the curvature follows a logarith- mic function in almost every case. The cone may be bilaterally symmetrical (isostrophic). More typically, however, the cone is also inclined into a third dimen- sion with this deviation from a plane of symmetry also varying at a logarithmic rate. Following conventional English and American orientation, almost all shells curve downward (orthostrophic) to the right (dextral) or, less commonly, to the left (sinistral). Far less com- monly they may curve upward (hyperstrophic) and to the right or to the left. All the possibilities are illus- trated by Knight (1952, p. 8), but it should be em- phasized that hyperstrophic coiling is atypical. Gastropod shells generally have the individual whorls in contact, and it is the exceptional shell in which the latest (body) whorl does not impinge upon, or at least touch, the preceding (penultimate) whorl. It is also the exceptional shell that does not following a logarith- mic function throughout most of its growth. Once the postjuenvile stage is past, allometric change is rare, especially among the more primitive Paleozoic gastropods. Some gastropods, for example the Holocene Ver- metus, grow irregularly once the very early stages are passed. This is a consequence of their habits of cement- ing the apex of the shell and, in part, of growing in dense colonies. In addition, occasional extremely large individuals in many diverse groups may have part or all of the body whorl free; and irregularity in growth lines, thickened shells, and other features indicate that the free body whorl may be a gerontic condition. In de- scriptions of either the exceptionally old shells or the profound deviations caused by changes in life habit, terms such as "disjunct" or "uncoiled" should be used as adjectives to describe such changes in shape. On the other hand, I propose to use the term "open- coiled" to designate those gastropod shell forms that fail to have some or all of the whorls in contact but NUMBER 3 237 that do not obviously deviate from logarithmic factors in rate of coiling. Unfortunately, only a few well-pre- served specimens exist, and, so far as I know, no de- tailed mathematical studies have been made of such forms. As a consequence, the opinions given here are unsupported by any except the most qualitative of data. Raup (1966) distinguishes four variables in coiling. The shape of the whorl (S) is basic to gastropod mor- phology, but this feature may be ignored here because it has little significance in interpreting the phenomenon of open coiling. The other three geometric components are: expansion rate of the whorl (W), distance of each whorl from the axis (D), and translation into the third dimension around the axis of coiling (T). In bilater- ally or nearly bilaterally symmetrical shells, the angle of tangency of the logarithmic spiral varies inversely to(W). Open coiling may occur if any of these components is large enough to prevent contact of the whorls. For convenience, the most obvious case is discussed first. Many Paleozoic Platyceratacea show relatively simple open coiling. This is especially evident in species of Platyceras (Orthonychia). The shells are all high- spired and are coiled through more than one whorl. The whorls are not in contact because the factor (T) is so large. In effect, the form has simply been elongated so that whorls do not touch. Were the speci- mens lower spired, the whorls would be in contact. In this factor of coiling, the Paleozoic Orthonychia show geometric similarity to the unrelated Vermicularia through part of its growth, except that representatives of the Holocene genus have a far greater number of whorls. In isostrophic shells, such as the bilaterally symmetri- cal Bellerophontacea, translation (T) is not a factor, and open coiling could come about through either an increase of each whorl in distance from the axis (D) or in expansion rate of the whorl (W), or by a combi- nation of both factors. However, no open-coiled Bel- lerophontacea are known. Within the Macluritacea, the members of which are not characteristically open-coiled, the genus Macluri- tella has whorls that are not in contact. The whorl pro- file of Macluritella is very low, and for practical pur- poses the shell may be considered as nearly isostrophic. The rate of coiling of Macluritella is normal for a gastropod, and the only unusual feature is that the whorls do not quite touch. If the whorls were only slightly larger in cross section—that is, if (W) were de- creased slightly—they would be in contact, and Mac- luritella would be an otherwise undistinguished mem- ber of the Macluritacea. Among the Paleozoic gastropods, the greatest variety of open-coiled forms occur within the Euomphalacea. In particular, the orthostrophic coiled Straparollus (Serpulospira) contains species which appear to have variations of both (D) and (T) components. An ex- cellent example has been described by Linsley (1968, pp. 375, 376) from the Middle Devonian Anderdon Limestone Member of the Lucas Formation in Michi- gan. Judging from other gastropods, his specimens of Serpulospira show enough variation in shape to be placed in a minimum of four species, but because they all were obtained from a small outcrop area and there is a suggestion of intergradation one may alternatively conclude that they all belonged to a single biologic population. The large variation in height of spire and openness of coil may be the result of relatively trivial changes in geometric components which are not, in themselves, significant. The few European specimens of Serpulospira that I have examined also show a re- markably high degree of variability and reinforce Lins- ley's observation. Clearly, variation within gastropod species tends to become stretched beyond normal limits when open coiling becomes a factor in gastropod morphology. Hyperstrophic coiling seldom leads to the same amount of variation in shape as does orthostrophic coiling. Hyperstrophic gastropods with a higher spire than Serpulospira are very rare in the fossil record. Open-coiled hyperstrophic genera have a base that is essentially flattened, or very nearly so. The factor (T) cannot be completely ignored, but it is so small in Nevadaspira that it cannot possibly account for much of the openness of coil. In fact, so far as one can tell in Nevadaspira and its allies, (T) is so small and constant that it is not responsible for any of the whorl disjunctness. Although most hyperstrophic open-coiled specimens are incomplete, enough is known of Nevadaspira to permit some measurements. By ignoring (T) and as- suming that all coiling is in a single plane, Raup (per- sonal communication, 1968) was able to measure the angle of tangency of the logarithmic spiral as near 83° In his terminology, assuming (T) as zero, (D) is 0.64 and (W) is 2.1. However, if (D) were decreased to 0.48, the whorls would be in contact; if (W) were decreased to 1.56, the whorls again would be in con- tact. In more general terms, assuming isostrophic coil- 238 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY ing and a circular cross section, if (D) is greater than (1/W) or if (W) is greater than (1/D), the shell will be open-coiled. By way of comparison with Nevadaspira, several species of typical tight-coiled Permian Euomphalidae have a logarithmic spiral angle that varies between 79° and 84° (Yochelson, 1956) ; these earlier measure- ments are not as accurate as those of Raup. Other Ordovician open-coiled gastropods such as Lytospira and Ecculiomphalus—which were particularly well illustrated by Knight (1941, pi. 71) to show hyper- strophic coiling—also have logarithmic spirals close to 80°. The early Carboniferous Phanerotinus also has a similar angle of tangency. It is evident, therefore, that in some factors of coiling Nevadaspira and similar forms are not distinguished from other Euomphalidae. Nevadaspira is open-coiled not because it curves in an atypical manner but be- cause the whorls are so narrow in cross section that they do not touch. If the whorls were just a bit wider, the external morphology of this gastropod would hardly be worthy of a second glance. In essence, Nevadaspira is too narrow a cone to expand at the rate at which it does and still have the whorls in contact. The "er- ror," if that word is proper, lies with the inner edge of the whorl, not the outer and certainly not with a dra- matic change in the logarithmic spiral. In his review of the curvature of nautiloid shells, Flower (1955) suggested that some sort of "saltation" was involved. All known nautiloid shell forms are straight or slightly curved on one hand, or fairly strongly curved with the whorls varying from no con- tact to tightly coiled, on the other. Flower attributed the absence in the fossil record of any nautiloid shell with intermediate curvature to possible difficulty the animal might encounter in contact with the bottom. Although this important idea has been virtually ig- nored, Raup (1967) has provided strong support for the concept by noting that of the many theoretically possible geometric shapes available for coiled am- monoids, there is a distinct preference for a limited number of general shapes and an apparent absence in the fossil record of some of the theoretically possible shapes. Another remarkable Devonian open-coiled genus is Mastigospira (La Rocque, 1949). Specimens are large and show such a slight degree of curvature that for many years they were assumed to be straight and were assigned to Hyolithes, a mollusk unrelated to the Gas- tropoda. This slight logarithmic curvature stands in marked contrast to all the open-coiled but strongly curved shells discussed above. The German Devonian Odontomaria also might be a slightly curved euompha- loid, or a calcareous worm tube. I have found no gastropods that show logarithmic curvatures bridging the pronounced morphologic gap between very slightly curved and strongly curved shells. Except for the strikingly atypical Mastigospira and the less well known Odontomaria, open-coiled Paleozoic gastropods are like other gastropods in being strongly curved. There has been a tendency in classification (Knight, Batten, and Yochelson, in Moore, 1960) to differenti- ate between tightly coiled and open-coiled gastropods so that such diverse forms as Nevadaspira and Masti- gospira are lumped together. The range of curvature theoretically available between that of a few degrees, as in Mastigospira, and that of 75° or greater may not be represented in the shell-bearing gastropods. The concept of a "saltation" in logarithmic curvature has not hitherto been applied to interpretation of gastro- pod shapes, but I suspect that this idea may provide a fruitful field for future investigation. Whorl Blockage in Paleozoic Gastropods Another myth perpetuated in beginning paleontology courses is that cephalopods are readily distinguished from gastropods by the presence of septa. Actually, it has long been known that some gastropods contain septa. Only one feature distinguishes a cephalopod shell from a gastropod shell—the "key" characteristic of having the septa pierced by a siphuncle. Because the prime significance of this feature was not properly emphasized, d'Orbigny and other early workers con- fused bilaterally symmetrical bellerophontaceans with cephalopods. Nevadaspira is nearly bilaterally sym- metrical and has septa, but it definitely is not a cepha- lopod because it lacks a siphuncle. Septa occur in a variety of gastropod shell shapes in specimens from rocks at least as old as Early Ordovician. So far, they are known from members of the Euomphalacea, Pleurotomariacea, and Loxone- matacea. There is an increasing suspicion that simple presence or absence of septa may be a particularly poor taxonomic criterion at all levels of classification. In septate specimens assigned to the last two super- families, the septa appear to be confined to the early whorls. So few good examples are known that one cannot be certain whether a single septum occurs, NUMBER 3 239 whether a few septa are present, or whether the early whorl is filled entirely with calcite, ending at a rounded partition. No particular advantage appears to have been gained by sealing off the soft parts from the early whorls. So far as known, extensive septation in the Paleozoic gastropods is confined to the Euomphalacea, although some typical tight-coiled Straparollus (Euomphalus) appear to show only juvenile whorl septation. This stands somewhat in contrast to the hyperstrophic open- coiled specimens which, when well preserved, show more than half-a-dozen septa. Any generalization must be approached with caution, for it has been demon- strated that septa originally present in a tight-coiled euomphalid may be destroyed during diagenesis and not appear in an average specimen (Yochelson, 1966). Virtually nothing is known about ontogenetic changes in the number and spacing of septa, let alone about individual variation in these features. Several specimens of the Early Ordovician Ecculiomphalus give an impression of uniform or logarithmic spacing of septa. The numerous specimens of Nevadaspira do not give such an impression of regularity, even if one makes allowance for possible erratic destruction of septa after the death of the individuals. Although the evidence is not sufficient to build a firm interpre- tation of septal distribution, I suggest that the data from the Nevadaspira specimens point to randomness in distribution of septa in all known open-coiled gastropods. An open-coiled form appears to be more fragile than a tightly coiled one. Accordingly, one might conclude that the multiseptate condition is a device for strengthening the shell, an opinion that has been advanced by some authors as the explanation for cephalopod septation. The sediment in which Nevada- spira occurs, and that of many of the occurrences of the Early Ordovician Ecculiomphalus, gives no indi- cation, however, of a high-energy environment that would require strengthened shells. At least one example of a large number of septa in a tightly coiled shell is known. The paratype of Arct- omphalus grandis Tolmachoff, a Middle Devonian species, shows at least ten septa irregularly spaced at intervals averaging about 2.5 millimeters (Yochelson, 1966). These septa occur in a part of a whorl inter- mediate in size; dozens may be present in the earlier whorls. If strengthening an open-coiled shell is the primary reason for multiple septa, there should be no need for the closely spaced septa in a tightly coiled form. This does not, in itself, disprove the need for strengthening, but it does cast some doubt on the suggestion. Finally, gastropod shells commonly are thicker than cephalopod shells and probably do not need internal strengthening. I suggest that strengthen- ing of the shell is at best a secondary result and that a more basic reason for septation lies elsewhere. It is placing the cart before the horse to think of growth of body mass as keeping up with the rate of tube growth, for it is the soft parts that build shell, not vice versa. However, this change in outlook may illus- trate a point. When septation occurs, the body no longer occupies the early part of the shell. Septation allows a smaller than average body mass to function within the shell. For septation confined to the early part of the shell, this is a trivial effect, but it is not a minor feature at more advanced growth stages. Re- peated septation obviously is not detrimental to life, for the existence of the multiseptate fossils proves this. It is well known from the most elementary growth studies that volume increases more rapidly than surface area. One result of the body mass moving forward in the tube is that expansion of its mass need not follow anything approaching a circumference to area ratio. Rather, the volume addition would be smaller than this ratio and would tend to approach the rate of shell addition. Eliminating the need for a greatly elongated body mass actually may allow for greater efficiency in construction of the shell. Septation or partitioning carries with it the concept of an open space between partitions. In cephalopods, camerae automatically develop each time a partition is inserted. It is reasonable to question whether septa- tion in gastropods is really the same as septation in cephalopods or simply appears to be the same. If sep- tation is taken to include both the partition and a chamber behind, there is no doubt that a few living gastropods are truly septate in the apical part of the whorl. There is also no doubt that some others deposit solid calcium carbonate in this shell area without any development of chambers. In fossils diagenetic changes may complicate the shell material and mask evidence for either interpretation. There has always been a general assumption that Paleozoic gastropods are septate as the term is gen- erally understood, the evidence for septation com- monly being a steinkern ending abruptly at a rounded 240 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY surface, though the latter could be equally as well the result of a solid apical filling. Partitions are especially evident in a few eroded Early Ordovician specimens. The same phenomenon of partitioning may be seen in a few weathered specimens of Nevadaspira (Plate 1: figure 5) ; others show a solid calcite mass in the early part of the whorl (Part 1: figure 6). Polished specimens of Nevadaspira show many more septa than do weathered surfaces. Weathering differ- entially affects the partitions, some of which, but not all, stand in relief. The partitions are clear in section and begin with a dark line, following the arc of a circle in thin section (Plate 1: figure 8). Anterior to this is an area of calcite. The sequence is closed by another dark line. As might be expected, no clastic impurities occur between the dark lines; calcium car- bonate seems to be the only material present in the interspaces (Plate 2 : figure 1). The large size of the calcite crystals suggests re- crystallization between partitions. Recrystallization is emphasized, for in some regions a dark line may extend from one wall part way into the mass of calcite and then be lost (Plate 2: figure 3). The alternatives are that either the early whorls were filled solidly by calcium carbonate, interrupted irregu- larly by deposition of dark material, which has now recrystallized, or that camerae were open and were subsequently filled after death. The available evidence supports the interpretation of later filling, for some areas between partitions show small crystals near the partition faces and larger ones in the intervening area. (Steven Stanley, personal communication, 1968.) See Plate 2: figure lc. Interpretations of septation, as contrasted with solid filling of whorls, have been made in the past on in- sufficient evidence. Fortunately, information from Nevadaspira supports septation, and it now rests on a much firmer basis. There seems little doubt that the open-coiled euomphaloid genera formed partitions as a life process. The partitions were formed by secretion of calcium carbonate by the apical part of the body mass, though the precise mechanism is as much a mys- tery as it is in the cephalopods. Interpreted Functional Morphology The open coil of Nevadaspira and similar gastropods appears poorly adapted for locomotion when compared with the more common tightly-coiled gastropod shells. The numerous partitions that shorten the body mass suggest that these gastropods would have even greater difficulty than normal gastropods in balancing the shell on the foot. It seems quite reasonable to rule out an active benthonic existence for all Paleozoic open-coiled gastropods. It seems almost as reasonable to assume a sedentary life habit. This may have been sessile, but morphologic evidence of any attachment or cementation is lacking. However, if the animal spent most of its adult life rest- ing in one area with the flattened profile of the shell on the substrate, an increase in area of contact with the bottom, as would come through open coiling, should have an advantage in maintaining this position. In- creasing weight with increasing size may have caused a slight sinking of the shell into even a fairly firm sub- strate. Hyperstrophic coiling would keep the aperture out of the sediment and thus would appear to be a natural response in shape change for a coiled animal living a sedentary life on a mud bottom. Many kinds of fossils deposited partitions of some sort within their exterior hard parts. Groups which come immediately to mind are rugose corals, richto- fenid brachiopods, and rudistid pelecypods. These all have in common a sessile mode of life. Septation may be a consequence of a sessile habit, for there may have been a physiological requirement to remove calcium carbonate from the soft parts; however, septation in cephalopods shows the danger of overgeneralization. Nevadaspira may not have had a concomitant reason to extend the shell at a maximum rate. Internal dep- osition gets rid of much calcium carbonate but does not materially enlarge the shell. If the concept of a sedentary life habit is valid, it may give some suggestion as to feeding habits. Heretofore, it has been tacitly assumed that the Euomphalacea, as primitive archaeogastropods, were probably herbi- vorous. It may be that members of this group were, in part, deposit feeders. The atypical open-coiled repre- sentatives may have further specialized toward ciliary feeding and evolved toward simply moving in and out of a tube to gather food, rather than foraging. Nevadaspira may have been a "sedentary worm," in a functional sense, rather than a typical gastropod. The hyperstrophic, low-spired, open-coiled euom- phalid gastropods are extinct. None of the living gastropods is similar either in presumed phylogenetic position or shell form. Some sort of analogy can be drawn with the unrelated Vermetidae. In a summary of the family, Morton (1965, p. 616) notes that NUMBER 3 241 septation is most frequent in the genus Dendropoma. As a consequence, the body is "plump and finger-shaped, no longer occupying the earlier convolutions of the tube." The animal has an operculum as wide as the aperture and "its quickness of response in darting back and closing the tube with the operculum is very strik- ing." (Morton, 1965, p. 620.) Because the body is relatively short, the columellar muscle is short and straplike rather than elongate. Many of Vermicularidae have developed a most remarkable method of feeding by secreting strings and sheets of mucus into the water to entrap other organisms. This habit apparently is an outgrowth of ciliary feeding. In Dendropoma, ciliary feeding is more common than in genera having the body more elongate and wormlike. I have not had the opportunity to observe a living member of the Vermetidae, and my conclusions regarding a sedentary life for Nevadaspira were formed prior to reading Morton's paper. Literature Cited Flower, R. H. 1955. Saltations in Nautiloid Coiling. Evolution, 9(3): 244-260. Jhaveri, R. B. 1969. Unterdevonische Gastropoden aus den Karnischen Alpen. Palaeontographica, Abteilung A, 133(4-6) : 146-176, plates 19-23. Knight, J. B. 1941. Paleozoic Gastropod Genotypes. Geological Society of America Special Paper, 32:1-510, 96 plates. 1952. Primitive Fossil Gastropods and Their Bearing on Gastropod Classification. Smithsonian Miscellane- ous Collections, 117(3): 1—56, 2 plates. La Rocque, A. 1949. New Uncoiled Gastropods from the Middle De- vonian of Michigan and Manitoba. University of Michigan Museum of Paleontology Contributions, 7(7): 113-122, 3 plates. Lindsley, R. M. 1968. Gastropods of the Middle Devonian Anderdon Limestone. Bulletins of American Paleontology, 54(244): 333-465, plates 25-39. Moore, R. C. 1960. Treatise on Invertebrate Paleontology, Part I, Mollusca (/). 351 pages, 216 figures. New York City and Lawrence, Kansas: Geological Society of America and University of Kansas Press. Morton, J. E. 1965. Form and Function of the Evolution of the Verme- tidae. British Museum {Natural History) Bulletin of Zoology, 2:586-630. Philip, G. M., and J. A. Talent 1959. The Gastropod Genera Liomphalus Chapman and Scalaetrochus Etheridge. Journal of Paleontology, 33(l):50-54, plates 7, 8. Poole, F. F.; D. L. Baars; H. Drewes; P. T. Hayes; K. B. Ketner; E. D. McKee; C. Teichert; and J. S. Williams 1967. Devonian of the Southwestern United States, in D. H. Oswald (editor), International Symposium on the Devonian System, Calgary, Alberta, Sep- tember 1967, 1:879-912. Calgary, Alberta, Society of Petroleum Geologists. Raup, D. M. 1966. Geometric Analysis of Shell Coiling: General Prob- lems. Journal of Paleontology, 40(5) : 1178-1190. 1967. Geometric Analysis of Shell Coiling: Coiling in Ammonoids. Journal of Paleontology, 41 (1):43^65. Yochelson, E. L. 1956. Permian Gastropoda of the Southwestern United States: I. Euomphalacea, Trochonematacea, Pseudophoracea, Anomphalacea, Craspedostoma- tacea, Platyceratacea. American Museum of Nat- ural History Bulletin, 110(3) : 173-276, plates 9-24. 1966. A Reinvestigation of the Middle Devonian Gastro- pods Arctomphalus and Omphalo cirrus. Norsk Polarinstitutt, Arbok 1965, pages 37-48, 2 plates. Yochelson, E. L.; J. S. White, Jr.; and M. Gordon, Jr. 1967. Aragonite and Calcite in Mollusks from the Penn- sylvanian Kendrick Shale (of Jillson) in Kentucky. United States Geological Survey Professional Paper, 575-D:D76-D78. 372-386 0—71 17 CARBONIFE ROUS John L. Carter New Early Mississippian Silicified Brachiopods from Central Iowa ABSTRACT Two new Mississippian brachiopod genera, based on new species, and a new species of the conservative rhipidomellid genus Perditocardinia Schuchert and Cooper, 1931, are described from a small silicified faunule found in the Eagle City Limestone Member of the Hampton Formation of central Iowa. This faunule appears to be latest Kinderhookian in age but precise means of correlation or age determination are not presently available. The brachiopod fauna of the upper Hampton For- mation of central and north-central Iowa has not been described. Undoubtedly this has been due to the difficulty in collecting well-preserved specimens in quantities adequate for accurate identification. The more fossiliferous beds in north-central Iowa fre- quently contain only molds and casts and the available shell material commonly fares unusually poorly in "crack-out" collecting from these beds, especially in central Iowa. The sum total of several weeks collecting can compare very poorly indeed with the materials of similar age readily obtainable in southeastern Iowa and northeastern and central Missouri. It was my good fortune to discover, after some days of arduous rock-breaking in the Eagle City Limestone Member sequence near LeGrand Iowa, several small blocks of clean limestone that appeared to contain a number of well-preserved silicified brachiopods. After etching with dilute hydrochloric acid I recovered a small faunule consisting entirely of brachiopods. The purpose of this paper is to describe several interesting John L. Carter, Department of Geology, University of Illinois, Urbana, Illinois 61801. new taxa discovered in this faunule. My identification of these and the other silicified brachiopod taxa in this faunule are listed below: Rhipidomella cf. R. dalyana (Miller) Perditocardinia iowensis, new species Schuchertella sp. Streptorhynchus sp. "Rhynchopora" sp. Eumetria sp. Composita sp. Cleiothyridina aff. Athyris crassicardinalis White Planalvus, new genus Planalvus gibberosa, new species Unispirifer cf. U. minnewankensis (Shimer) Mirifusella, new genus Mirifusella fortunata, new species Spirifer sp. Verkhotomia? sp. Punctospirifer solidirostris (White) Beecheria sp. Dielasma sp. Most of the species are represented by only one or two specimens, mostly single valves, making accurate specific identification difficult. The three species de- scribed below are the most abundant elements in the faunule. Age of the Faunule The limestone bed of Eagle City Limestone from which most of the blocks of silicified material were taken oc- curs near the top of a large, actively worked quarry about one mile north of LeGrand, Marshall County, Iowa. This quarry is north and directly across the Iowa River from the abandoned quarry that yielded the famous LeGrand crinoid fauna of Laudon and Beane (1937). The silicified faunule occurs stratigraphically above that crinoid zone and probably would be placed in the "Spiriferina Zone" of Laudon (1931, pp. 423, 245 246 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY 424). The several small blocks containing silicified specimens were loosened and exposed by a bulldozer in preparation for further drilling and blasting. Addi- tional blocks may be difficult to obtain. Small fist-size lumps containing silicified specimens were collected in the abandoned south "crinoid" quarry and similar small collections were made in a quarry two miles east of LeGrand in Tama County. Blocks of adequate size, however, were not found at any quarry other than the first one described above. The brachiopod faunule contains no convincing horizon markers for delimiting the youngest possible age for the highest beds of the LeGrand quarries. The presence of Eumetria sp., Verkhotomia? sp., Rhipido- mella cf. R. dalyana, and the new species of Perdito- cardinia in the silicified materials certainly suggests that these beds are younger than any part of the Chouteau Limestone of northeastern Missouri, and their strati- graphic position above the Maynes Creek Dolomite makes them younger than the highest Kinderhook of southeastern Iowa, the Wassonville Dolomite, which is a correlative of the Maynes Creek. Without more data I am forced to rely on the work of others and tentatively accept Laudon's (1931, p. 412) assignment of a Kinderhookian age for the upper Hampton For- mation. Nevertheless, there is no evidence from the brachiopod fauna, including the nonsilicified speci- mens, that would preclude an early Osagian age for these beds, especially considering the possibility that the upper Hampton Formation sequence of central Iowa could conceivably represent a nearly continu- ous record of latest Kinderhookian-earliest Osagian sedimentation. Preservation This small silicified assemblage represents a thanatocoe- nose. Many of the spiriferid valves and some other shells are very much abraded, and no complete large shells were found with both valves intact. Many of the spiriferid pedicle valves are selectively penetrated by borings. A few of the shells have repaired boring in- juries, so these brachiopods must have been alive when attacked by the boring organism. Most of the pene- trated shells showed no signs of repair, however, and the large number of holes per valve suggests that the valves were bored after the death of the brachiopods. Localities UI (University of Illinois) locality Z-1F Top beds exposed in SE corner of Martin-Marietta Cor- poration quarry, just north of Iowa River, about one mile north of LeGrand, NW^4 sec 1, T 83 N, R 17 W, Mar- shall County, Iowa. Eagle City Limestone. UI locality Z-2 Upper beds exposed in abandoned quarry just south of Iowa River, about three-fourths mile north of LeGrand, SW/4 sec 1, T 83 N, R 17 W, Marshall County, Iowa. Eagle City Limestone. UI locality Z-3 Top beds of northwest corner of B. L. Anderson Company, Montour quarry, NE/4 sec 8, T 83 N, R 16 W, Tama County, Iowa. Eagle City Limestone. All types are in the collections of the University of Illinois, Urbana. Phylum BRACHIOPODA Dumeril, 1806 Class ARTICULATA Huxley, 1869 Order ORTHIDA Schuchert and Cooper, 1932 Suborder ORTHIDINA Shuchert and Cooper, 1932 Superfamily ENTELETACEA Waagen, 1884 Family RHIPIDOMELLIDAE Schuchert, 1913 Genus Perditocardinia Schuchert and Cooper, 1931 Perditocardinia iowensis, new species PLATE 1: FIGURES 21-41; PLATE 2: FIGURES 6-11 DESCRIPTION.—Shell average size for the genus, un- equally biconvex, the brachial valve being slightly to moderately more inflated than the pedicle valve; out- line variable, commonly subcardiiform with a rostrate posterior or, more rarely, subcircular to subovate, and either longitudinally or transversely elongated; greatest width usually attained in the anterior half of the shell or, more rarely, at or near midlength; profile lenticular; hinge-line very narrow; cardinal extremities rounded; sulcus moderately developed, fold low and obscure; anterior commissure weakly uniplicate, anterior margin commonly slightly emarginate or gently rounded; orna- ment consists of numerous fine costellae and strong, irregularly spaced growth varices. NUMBER 3 247 Pedicle valve thick-shelled, moderately and almost evenly convex in lateral profile; beak small, bluntly to sharply pointed and of moderate size, slightly incurved; interarea extremely small, often reduced to low ridges bounding the delthyrium, or completely absent; del- thyrium filled by the cardinal process, leaving a small apical foramen; beak ridges per se poorly developed, leaving a rounded palintrope; lateral edges of valves commonly flattened and differentiated from the ventral surface by a ridge; sulcus narrow, originating in the beak region, becoming considerably impressed anteriorly. Pedicle valve interior with stout, elongate, slightly divergent teeth; crural fossettes well developed on the anteromedial edges of the teeth; pedicle callist clearly differentiated between the tooth bases; lateral and anterior margins minutely crenulated by the inner edges of the surface costellae; muscle field large, fla- bellate, deeply impressed, extending anteriorly about one-half to two-thirds the length of the valve; adduc- tor field narrow, completely enclosed by the diduc- tor-adjustor field; low thick median ridge commonly divides the muscle field but especially prominent in anterior portion of the diductor field. Brachial valve thinner-shelled than pedicle valve and more strongly convex in both anterior and lateral profile; beak rounded, inconspicuous; interarea obso- lete; fold very weakly developed or absent; anterior commissure commonly flexed dorsad to match the small tongue of the pedicle valve sulcus. Brachial valve interior with large thick cardinal process; myophore often appears trilobed in anterior or posterior profile; shaft not discernible, fused to brachiophores by adventitious shell tissue; brachio- phores elongate, bladelike, fused to floor of valve and cardinal process, with a ventral process that articulates with the crural fossettes on the teeth of pedicle valve; short, slender, anteriorly directed crura developed on anterior edge of the brachiophores; sockets long and deep; fulcral plates not developed but the sockets may be slightly constricted anteriorly by a low ridge devel- oped on the exterior or lateral faces of the brachio- phores; dorsal adductor scars moderately impressed, separated by a wide low rounded median ridge. Measurements (in millimeters) of the types of Perd- itocardinia iowensis, new species, from the Hampton Formation (Eagle City Limestone Member) of LeGrand, Marshall County, Iowa: Length 15.5 12.6 12.4 12.5 12.7 12.9 11.6 11.2 Width 14.9 15.2 13.7 12.7 10.8 14.6 13.2 10.5 Locality Z-3 Z-3 Z-1F Z-1F Z-1F Z-3 Z-1F Z-1F UI No. PEDICLE VALVES X-3425 X-3427 X-3432 X-3426 X-3428 BRACHIAL VALVES X-3429 X-3430 X-3431 HOLOTYPE.—UI No. X-3426 (Plate 1: figures 24- 26; Plate 2, figure 10). DISTINGUISHING CHARACTERS.—This species is characterized by its subcardiiform to subovate out- line; poorly developed rostration; greatly reduced or obsolete ventral interareas and obsolete dorsal inter- areas; and narrow sulcus originating in the beak re- gion. Internally, the large lobate cardinal process is fused to the brachiophores by a thick mass of adventi- tious shell tissue, and the ventral muscle field extends anteriorly about one-half to two-thirds the length of the valve. COMPARISONS.—Perditocardinia iowensis, new spe- cies, is related to P. dubia (Hall, 1858), identified from various late Osagian and Meramecian horizons, but it can be differentiated from that species by its less ros- trate, more variable outline; usually thinner-shelled valves; and narrower, shorter ventral muscle field. Rhipidomella thiemei (White, 1860), from the Starr's Cave Limestone of Iowa and upper Chouteau Group of northeastern Missouri, often is similar in size, outline, and profile to P. iowensis, new species, but is never rostrate. It further differs in possessing a wider hinge-line, larger ventral interarea, and in having the sulcus originate anterior to the beak region of the pedicle valve. REMARKS.—Since the establishment of the genus Perditocardinia by Schuchert and Cooper in 1931 no species, other than the type P. dubia (Hall), has been recognized in the Mississippian or Lower Carbonifer- ous. That species has been reported, however, from many horizons ranging in age from Osagian through Chesterian. As Schuchert and Cooper (1932, p. 135) pointed out, the type species, as thus identified, is a composite "and it will probably prove desirable to separate it into several specific groups." Rhipidomella Oehlert, 1890, and its close relative Perditocardinia represent an extraordinarily conserva- tive stock of punctate orthids in the late Paleozoic. Tracing the unusually slow-paced morphological 248 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY changes that took place in these brachiopod lineages seems to me to be possible under these circumstances. It is my view that, morphologically, Perditocardinia dubia is sufficiently distinct from the more typical Carboniferous rhipidomellid stock that it probably was preceded by earlier intermediate species. Perdito- cardinia iowensis seems to fit this requirement because it falls morphologically between the Meramecian ros- trate forms lacking interareas and the nonrostrate rhipidomellid forms with well-developed interareas. I interpret P. iowensis, then, to represent an early perditocardinid offshoot from the rhipidomellid line- age, perhaps even being the direct ancestor of P. dubia and its unnamed younger relatives. OCCURRENCE AND ABUNDANCE.—This is one of the commonest brachiopod species in the "Spiriferina Zone" of Laudon in the LeGrand vicinity. The silici- fied collection alone consists of several hundred dis- articulated valves, many of them well preserved. Order SPIRIFERIDA Waagen, 1883 Suborder ATHYRIDIDINA Boucot, Johnson, and Staton, 1964 Superfamily ATHYRIDACEA M'Coy, 1844 Family ATHYRIDIDAE M'Coy, 1844 Subfamily ?ATHYRIDINAE M'Coy, 1844 Planalvus, new genus DIAGNOSIS.—Longitudinally to transversely ovate, small to medium size, unequally biconvex athyridids with weakly developed ventral sulcus and dorsal fold; pedicle valve thick-shelled, weakly convex, nearly flat posteriorly; ventral beak small, almost obsolete; beak ridges subangular, defining pseudoareas on either side of the delthyrium; pedicle foramen small, possibly obsolete in adults; dental plates lacking; surface lamel- lose, the lamellae being finely striated and apparently fringed with minute spines. Brachial valve gibbous and much more inflated than the pedicle valve; dorsal beak angular, pointed, incurved, filling the delthyrium of the pedicle valve completely; cardinal plate large, athyridid, nonper- forate, nearly filling the ventral umbo; brachidium unknown. TYPE-SPECIES.—Planalvus gibberosa, new species. DISTINGUISHING CHARACTERS.—This genus is char- acterized by its unequally biconvex valves, the brachial valve being gibbous and the pedicle valve weakly convex. The pedicle valve is thick-shelled; the ventral beak is inconspicuous; and ventral pseudoareas are developed. Internally, the pedicle valve lacks dental plates, and the cardinal plate in the brachial valve is large and nonperforate. Surface ornament consists of fine, closely spaced, striated lamellae fringed with minute spines. COMPARISONS.—Planalvus occurs with Composita Brown, 1849, but can be differentiated externally from that genus by its weakly convex, nonrostrate pedicle valve with planareas, strongly inflated brachial valve with an extended, acute, incurved dorsal beak, and finely lamellose ornament. Composita is rostrate, un- equally biconvex, and smooth. Internally, the pedicle valve of Planalvus lacks dental plates and the cardinal plate is nonperforate, whereas Composita has stout dental plates and a perforate cardinal plate. Athyris M'Coy, 1844, overlaps Planalvus in its stratigraphic range but can be distinguished easily by its rostrate, subequally biconvex form and broad smooth lamellar frills. Internally, Athyris possesses dental plates and a perforate cardinal plate. Cleiothyridina Buckman, 1906, is similar to Athyris except it has smooth lamellar frills fringed with well- differentiated, flattened spines. Spirigerella Waagen, 1883, is similar to Planalvus in its thick shell and apparent lack of dental plates, but it differs in being smooth and having a more convex pedicle valve with a pronounced incurved beak, and the pedicle foramen is occluded by the dorsal umbo. Internally, the cardinal plate is perforate. Actinoconchus M'Coy, 1844, is similar to Planalvus in possessing striated lamellar frills but the grooves or striations on the lamellae are widely spaced and do not form a spinose fringe at the margins. Furthermore, Actinoconchus is subequally biconvex, possesses a moderately developed ventral beak, and, internally, possesses dental plates. REMARKS.—Although Planalvus probably belongs in the subfamily Athyridinae M'Coy, 1844, along with the genera mentioned above, I have assigned it to this subfamily tentatively because the brachidium is un- known. The lamellose nature of the type-species is not apparent in most of my specimens. I had thought ini- tially that the surfaces of the valves were more or less smooth in the manner of Composita and, hence, that possibly Planalvus was a close relative of that genus. NUMBER 3 249 Ultimately, however, several lamellose specimens etched out, and it became apparent that P. gibberosa, at least, has a more typical athyridid ornament of spine- fringed lamellae. I have not been able to determine whether Athyris densa possesses a lamellose ornament but place it in Planalvus with fair confidence due to its close similarity to the type species in most other respects. The name is derived from the Latin plan (flat) and alvus (belly). SPECIES ASSIGNED.—Planalvus gibberosa, new spe- cies; Athyris densa Hall and Clarke, 1894. RANGE.—Late Kinderhookian and Meramecian of North America. Planalvus gibberosa, new species FIGURE 1; PLATE 1: FIGURES 1, 2; PLATE 2: FIGURES 12-41 DESCRIPTION.—Shell small for the subfamily; un- equally biconvex, the brachial valve being much more inflated than the pedicle valve; longitudinally subel- lipsoidal to subcircular in outline, rarely wider than long; greatest width attained near midlength or slightly posterior to midlength; profile subovate; hinge-line short, curved, subterebratulid; cardinal extremities well rounded; fold and sulcus weakly developed in adult shells only; growth varices irregularly spaced; surface finely lamellose, the lamellae being finely striated and anteriorly fringed with very fine spines. Pedicle valve thick-shelled, moderately to weakly and evenly convex in lateral profile; beak inconspicu- ous, with a small curved notch formed by the pedicle foramen; delthyrium almost completely filled by the dorsal umbo; beak ridges subangular, forming flattened pseudoareas on either side of the foramen; sulcus, if present, broad and shallow, giving the anterior portion of the shell a sublingual outline in many shells. Pedicle valve interior lacking dental plates, with a moderately impressed muscle field posteriorly; teeth large, blunt. Brachial valve thinner-shelled than pedicle valve; gibbous; much more inflated than the pedicle valve; most strongly convex in the umbo but nearly evenly curved in lateral profile; dorsum very high, the flanks sloping rapidly to the lateral margins but not forming an actual fold; beak well developed, incurved into the delthyrium of the pedicle valve. Brachial valve interior with large nonperforate ven- trally concave cardinal plate that bends posterodorsally to unite with the floor of the valve between stout socket ridges; posterior edges of cardinal plate on either side of the dorsal beak slightly recurved, forming two low flanges that may fuse medially to form a small perforation (however, this perforation is in no way homologous with the visceral perforation in Composita, Cleiothyridina, or Athyris); crural bases arise either on the outer anterior edge of the cardinal plate or on the inner anterior edges of the socket ridges (this is essentially a histological problem and can best be handled with thin section techniques applied to calcar- eous or essentially unaltered specimens) ; brachidial structures unknown; low myophragm present in pos- terior third of valve. Measurements (in millimeters) of Planalvus gib- berosa, new species, from the Eagle City Limestone Member of the Hampton Formation, LeGrand, Iowa; U.I. locality Z-1F: UI No. Length Width Thickness X-3435 10. 2 9. 5 5.6 X-3436 8.0 7.8 4.0 X-3437 5. 7 6. 1 2. 8 X-3438 4.1 4.1 1.9 X-3439 2.9 3.6 1. 5 HOLOTYPE.—UI No. X-3435 (Plate 2: figures 17-21). DISTINGUISHING CHARACTERS.—This species is characterized by its small size, longitudinally sub- ellipsoidal to subcircular outline, inconspicuous ventral beak, and poorly developed fold and sulcus. COMPARISONS.—Planalvus gibberosa, new species, is most similar to P. densa (Hall and Clarke, 1894), from the Salem Limestone of Indiana and Kentucky. Planalvus densa is transversely subovate in outline and attains a much greater size than P. gibberosa. Further- more, P. densa may have a weak median sulcus on the dorsal fold. GROWTH.—Although a large collection of well- preserved pedicle valves is not available for measure- ment and comparison, there is a clear trend toward acceleration in length relative to width, as seen in growth lines of several of the larger well-preserved shells. These growth lines, measured for three of the types, are plotted on Figure 1. This observation is substantiated to some extent by the length-width meas- urements available for several well-preserved juvenile specimens. All of these specimens are transverse except for one which is equidimensional, indicating that mod- erate allometry probably occurs in later growth stages. 250 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY / 10 LENGTH(MM) FIGURE 1.—Plots of length/width measurements of Planalvus gibberosa, new species, taken from growth varices of three pedicle valves illustrating allometric growth in late ontoge- netic stages that results in elongation of the valves. OCCURRENCE AND ABUNDANCE.-—Found in the upper Hampton Formation (Eagle City Limestone Member) of LeGrand, Iowa, and vicinity, and the Gilmore City Limestone of north-central Iowa. The silicified collec- tion from the LeGrand localities consists of over 50 specimens. Only eight specimens have been recovered from the Gilmore City Limestone. Suborder SPIRIFERIDINA Waagen, 1883 Superfamily SPIRIFERACEA King, 1846 Family uncertain Mirifusella, new genus DIAGNOSIS.—Medium size for the superfamily; ma- ture shells often longer than wide, greatest width usu- ally anterior to the hinge-line, often near or anterior to midlength; cardinal extremities rounded or sub- angular, never mucronate; outline subovate to guttate; ventral umbo long and inflated, extending posteriorly beyond the hinge-line about one-fourth to one-third the length of the valve; fold and sulcus poorly devel- oped; hinge-line denticulate; ornament usually con- sists of a few simple rounded costae on the flanks, a median sulcal costa that may bifurcate anterior to the beak, and other sulcal costae, which are simple and bifurcate from the sulcus bounding costae; surface weakly capillate, forming a faintly reticulate pattern with fine growth lines; shell substance impunctate. TYPE SPECIES.—Mirifusella fortunata, new species. DISTINGUISHING CHARACTERS.—Mirifusella is char- acterized by its commonly elongate subovate to guttate outline and relatively narrow hinge-line (for spirifers) in mature specimens, with few low rounded simple costae on the flanks, a sulcal median costa that may bifurcate, and simple lateral sulcal costae that bifurcate from the sulcus-bounding costae. COMPARISONS.—Mirifusella is most similar to Fusella M'Coy, 1844, Anthracospirifer Lane, 1963, Unispirifer Campbell, 1957, and Imbrexia Nalivkin, 1937. It can be distinguished from all of these genera by its elongate outline in most mature specimens. In addition, its maximum width is usually anterior to the hinge-line and the cardinal extremities are rounded or subangular. Fusella and Unispirifer are usually strongly transverse genera, commonly having angular ears and the maximum width at or very near the hinge-line. Also, Fusella has a simple median sulcal costa, and Unispirifer has numerous lateral costae. Anthra- cospirifer and Imbrexia commonly have rounded or subangular ears but are, nevertheless, transverse, and rarely, if ever, elongate. Further, Anthraco spirifer has lateral costae which may bifurcate, and its median sulcal costa is consistently simple; in Mirifusella the lateral costae are invariably simple, but the median sulcal costa may bifurcate. Imbrexia (although that genus has not been properly established due to our limited knowledge of its type-species, Spirifer imbrex Hall, 1858) apparently has a strongly imbricate ornament covering the entire surface of the shell, except for the interareas, and a simple median costa in the sulcus. The simple costation and tendency toward elonga- tion of the shell of Eochoristites Chu, 1933, from the Kinling Limestone of northeastern China, tempt one to compare it with this new genus. However, Eochoristites was clearly described as possessing basal plates ("crural plates" of others) in the brachial valve, and possibly has a nondenticulate hinge, although I am not con- vinced of that. REMARKS.—The grouping of genera into families of spiriferid brachiopods is a particularly difficult and challenging occupation. Unfortunately, the Treatise on Invertebrate Paleontology, Part H, Brachiopoda (R. C. Moore, editor), which appeared in 1965, did not alleviate our miseries in this regard. The classifica- NUMBER 3 251 y/r- • /? LENGTH/ MAX. WIDTH RATIO /?• / LENGTH (MM) i l 15 20 25 LENGTH/ HINGE WIDTH RATIO I ? I *% ,' tion of the spiriferids remains a perplexing problem to the student of spiriferid systematics. This is not the time and place to attempt a new synthesis for the spiriferids. I merely want to point out the impossibility of being able to assign genera confidently to families, other than types. Mirifusella, new genus, possesses both spiriferid and brachythyridid characters, as do many genera assigned to one family or the other in the Treatise. Per- haps a good dose of numerical taxonomy will unsnarl the knot, perhaps not. In any case, I am unable at this time to place Mirifusella in a family in such a manner that the assignment in any way elucidates its true relationships. The name is derived from the Latin mirus (strange) and fusella (little spindle). SPECIES ASSIGNED.—Mirifusella fortunata, new species, from the Eagle City Limestone Member of the Hampton Formation of south-central Iowa, and Spiri- fer indianensis Weller, 1914, from the Harrodsburg Limestone of Indiana. Weller (1914, p. 353) noted the peculiar elongation of his species and considered it to be unique. In various collections, however, I have seen undescribed representatives of that genus ascribed to be from the Keokuk Limestone and Warsaw Forma- tion of the upper Mississippi Valley. RANGE.—Late Kinderhookian through early Mera- mecian. Mirifusella fortunata, new species FIGURE 2; PLATE 1: FIGURES 3-20; PLATE 2: FIGURES 1-5 DESCRIPTION.—Medium size for the family; sub- equally biconvex; moderately inflated; proportions variable due to allometry but usually considerably longer than wide in large adults; outline longitudinally guttate to transversely subovate in most specimens; hinge-line straight and less than the maximum width in most specimens, often narrow; maximum width usually near or anterior to midlength; cardinal extre- mites rounded in about half the known specimens, sub- angular in the rest; anterior commissure uniplicate; fold and sulcus of moderate width, weakly developed; macro-ornament consists of about 10 to 13 simple rounded costae per flank in adult specimens; sulcus with about 3 to 6 costae, the median costa usually bi- furcating in the umbonal region and the lateral sulcal costae bifurcating from the sulcus-bounding costae; growth varices are irregularly spaced; micro-ornament consists of fine capillae and closely spaced growth lines, FIGURE 2.—Scatter diagrams of dimensions of 73 pedicle valves of Mirifusella fortunata, new species, from UI locality Z-1F. the latter becoming slightly imbricate near the an- terior margin and forming a weak reticulate pattern. Pedicle valve with greatest convexity near the beak, considerably more inflated than the brachial valve; beak small and incurved; umbonal region often ex- tending posteriorly beyond the hinge-line for more than one-fourth the total length of the valve; ventral inter- area apsacline, usually narrower than the maximum width, strongly concave, triangular, often vertically striated, well-defined ventrolaterally by subangular beak ridges; hinge-line denticulate; delthyrium narrow, usually forming an angle at the apex of less than 60 degrees, often partially closed in mature specimens by coalescing of the thickened dental plates; sulcus usually very shallow, or even indistinct in some specimens. 252 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY "':' i 31 41 k 1 fM J 39 40 NUMBER 3 253 Interior of pedicle valve with strong, divergent dental plates that laterally enclose the posterior part of the muscle field; dental plates may become greatly thickened in larger specimens and often flare out anter- olateral^ to accommodate the muscle field; in some adults and juveniles a small subdelthyrial closure is formed in the apex of the delthyrium by the initial fusion of the thickening dental plates, however, such a closure is not a true plate; muscle field moderately impressed in large adult specimens; ovarian markings well developed between the dental plates and the pos- terolateral margins; teeth small, bladelike, directed laterally. Brachial valve with greatest convexity in the um- bonal region of most shells, usually moderately in- flated, never as inflated as the pedicle valve; usually wider than long, much shorter than the pedicle valve, PLATE 1: figures 1, 2—Planalvus gibberosa, new species. Mi- cro-ornament of anterior portions of two pedicle valves, show- ing very finely radially striated, closely spaced lamellae fringed with minute spines at their free edges (all views X8) : 1, A medium-size pedicle valve, UI X-3433, from UI loc. Z—3, Tama County, Iowa; 2, a larger pedicle valve, UI X—3434, from the north quarry, UI loc. Z—IF, near LeGrand, Marshall County, Iowa. Figures 3-20.—Mirifusella fortunata, new species, from the north quarry, UI loc. Z—IF, near LeGrand, Marshall County, Iowa (all views X1.5): 3-5, ventral, lateral, and anterior views of a pedicle valve, the holotype, UI X—3445; 6-8, anterior, lateral, and dorsal views of a brachial valve, UI X-3446; 9-11, ventral, lateral, and anterior views of a pedicle valve, UI X-3447; 12-14, posterior, lateral, and dorsal views of a brachial valve, UI X-3448; 15-17, ventral, lateral, and anterior views of a pedicle valve, UI X-3449; 18-20, ante- rior, lateral, and dorsal views of a brachial valve, UI X— 3450. Figures 21-41.—Perditocardinia iowensis, new species (all views X2). 21-23, ventral, lateral, and anterior views of an unusually large pedicle valve, UI X-3425, from UI loc. Z-3, Tama County, Iowa; 24—26, ventral, lateral, and ante- rior views of a pedicle valve, the holotype, UI X-3426, from the north quarry, UI loc. Z-1F, Marshall County, Iowa; 27- 29, ventral, lateral, and anterior views of a wide pedicle valve, UI X-3427, from UI loc. Z-3, Tama County, Iowa; 30-32, ventral, lateral, and anterior views of an elongate pedicle valve, UI X-3428, from the north quarry, UI loc. Z—IF, near LeGrand, Marshall County, Iowa; 33—35, dorsal, lateral, and anterior views of a large, wide brachial valve, UI X-3429, from UI loc. Z-3, Tama County, Iowa; 36-38, dor- sal, lateral, and anterior views of an emarginate brachial valve, UI X-3430, from the north quarry, UI loc. Z-1F, Marshall County, Iowa; 39-41, dorsal, lateral, and anterior views of an elongate brachial valve, UI X-3431, from the north quarry, UI loc. Z-1F, Marshall County, Iowa. and with a subelliptical outline in most specimens; dor- sal beak tiny, inconspicuous; interarea extremely low, orthocline; fold low, of moderate width, often some- what flattened; ornament similar to and complemen- tary with that of the pedicle valve. Brachial valve interior with moderate-size striate cardinal process which commonly is medianly divided by a narrow groove, almost becoming bilobed in large specimens; cardinal process rests on floor of valve posteriorly but is supported laterally and anterolateral^ by infolded shell tissue that wraps around the crural bases from the dorsal surfaces of the inner socket ridges (lacking thin section material, it cannot be determined at present whether this shell tissue is fibrous or colum- nar in nature); weak median ridge developed between the moderately impressed adductor scars; sockets wide, typically spiriferid; crura originate on the dorsolateral surfaces of the inner socket ridges but do not extend to the floor of the valve; however, they are commonly enclosed dorsally by the same shell tissue that wraps around from the inner socket ridge to support the cardinal process; other brachidial details unknown; ovarian markings present on the posterolateral por- tions of the valve. Measurements (in millimeters) of the types, from the Eagle City Limestone Member of the Hampton Formation, UI locality Z-1F, near LeGrand, Iowa: Hinge Maximum UI No. Length Width Width PEDICLE VALVES X-3451 23.9 18.9 21.8 X-3445 24.0 16.3 22.3 X-3447 25.8 15.3 19.9 X-3452 17.0 18. 1 18.3 X-3449 15.6 12.8 16. 1 BRACHIAL VALVES X-3455 17.5 16.2 20.3 X-3446 16.3 17.7 20.6 X-3448 18.4 11. 7 16.3 X-3450 13.2 14.0 16.3 X-3454 13.4 13.4 14.9 HOLOTYPE.—UI No. X-3445 (Plate 1: figures 3-5). DISTINGUISHING CHARACTERS.—This species is characterized by its longitudinally guttate to trans- versely subovate outline; usually narrow hinge-line with the maximum width at or anterior to midlength; rounded or subangular cardinal extremities; weakly developed fold and sulcus; about 10 to 13 low rounded simple costae per flank; about three to six sulcal costae, the median one usually bifurcating; capillate micro- 254 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY PLATE 2 NUMBER 3 255 ornament; strong, divergent dental plates; and com- monly in large specimens, a weakly bilobate striate cardinal process. COMPARISONS.—Mirifusella fortunata, new species, is most similar and closely related to Spirifer indianen- sis Weller, 1914, from the Harrodsburg Limestone of Indiana and possibly the Keokuk and Warsaw Forma- tions of the upper Mississippi Valley. The latter species possesses fewer but stronger lateral costae and the median costa in the sulcus rarely bifurcates. The only other discernible difference between these species is that the dorsal interarea in M. fortunata is orthocline, whereas in M. indianensis it is anacline. Spirifer marshallensis Weller, 1914, was described from two brachial valves and, according to Weller (1914, pi. 50, figs. 13-14), is from the Kinderhook beds of LeGrand, Iowa. I have examined Weller's types and must admit that the matrix lithology can be only of the Hampton Formation, probably Maynes Creek Dolomite. As Weller noted, the types have a granulose surface ornament; hence, this species belongs to a genus other than Mirifusella and is probably punc- tate, possibly a Punctothyris Hyde, 1953. GROWTH.—The allometric growth form of the pedicle valve of this species is illustrated in the scatter diagrams of Figure 2. Elongation of the shell is ac- celerated in relation to both hinge width and maxi- PLATE 2: figures 1—5.—Mirifusella fortunata, new species, from the north quarry, UI loc. Z-1F, near LeGrand, Marshall County, Iowa (all views X 2): 1, 2, slightly oblique views of the interiors of two pedicle valves, UI X-3451-2, re- spectively; 3, interior of large dorsal umbo, UI X-3453, show- ing the thickened bilobed cardinal process; 4,5, interior views of two brachial valves, UI X-3454-5, respectively. Figures 6—11.—Perditocardinia iowensis, new species (all views X 2) : 6, interior view of a large wide brachial valve, UI X-3429, from a quarry, UI loc. Z-3, Tama County, Iowa; 7,8, interior views of two brachial valves, UI X-3430-1, re- spectively} from the north quarry, UI loc. Z-1F, near Le- Grand, Marshall County, Iowa; 9-11, interior views of three pedicle valves, UI X-3432, 3426 (holotype), and 3428, re- spectively, from the north quarry, UI loc. Z-1F, near Le- Grand, Marshall County, Iowa. Figures 12-41.—Planalvus gibberosa, new species, from the north quarry, UI loc. Z-1F, near LeGrand, Marshall County, Iowa (all views X 3) : 12, 13, exterior and interior views of a pedicle valve, UI X-3440; 14, 15, interior views of two brachial valves, UI X-3441-2, respectively; 16, oblique anterior view of a pedicle valve, UI X-3443, with the car- dinalia of the dorsal umbo articulated with the teeth; 17- 41, ventral, dorsal, lateral, anterior, and posterior views of a growth series of five complete specimens, UI X-3435-9, respectively, including the holotype (figures 17-21). mum width. Acceleration seems to occur when the shells are between about 7 and 15 mm long. Over 90 percent of the pedicle valves available attained maxi- mum width anterior to the hinge-line in all observable growth stages, and elongation of the valve was evident in about two-thirds of the valves 15 mm wide and larger. Internally, the dental plates are extremely variable in length and thickness in small and medium-size shells, the thickening that is characteristic of the larger mature shells apparently not being related to shell size alone. The cardinal process in the smaller brachial valves is usually small, not well supported by callus. In larger valves, however, the lateral edges of the process grow anteriorly, and, in conjunction with deposition of callus, give the process a decidedly bilobed appearance. REMARKS.—Many of the pedicle valves and several brachial valves in the silicified collection were pene- trated by boring organisms, possibly sponges. Some of these borings were sealed by secretion of additional shell tissue, indicating that the boring organism at- tacked the shell while the brachiopods were still alive. However, many of the holes penetrate the valves com- pletely. These borings either were fatal or were made on already lifeless shells. OCCURRENCE AND ABUNDANCE.—Mirifusella fortu- nata is one of the most common species in the Eagle City Limestone Member in the vicinity of LeGrand, Iowa. The silicified collections consist of several hun- dred disarticulated valves, and unaltered or weakly silicified shells are common in the beds just below the highest bed exposed in the LeGrand quarries. Literature Cited Laudon, L. R. 1931. The Stratigraphy of the Kinderhook Series of Iowa. Iowa Geological Survey Annual Report, 35:335-451, 24 figures. Laudon, L. R., and B. H. Beane 1937. The Crinoid Fauna of the Hampton Formation at LeGrand, Iowa. University of Iowa Studies in Natural History, 17(6) : 227-272, plates 15-19. [New series number 345.] Schuchert, C, and G. A. Cooper 1932. Brachiopod Genera of the Suborders Orthoidea and Pentameroidea. Yale University Peabody Mu- seum of Natural History Memoir, 4(1): 1—270, plates A and 1-29. Weller, Stuart 1914. The Mississippian Brachiopoda of the Mississippi Valley Basin. Illinois State Geological Survey Monograph, 1:1-508, plates 1-83. Mackenzie Gordon, Jr. Carlinia, a Late Mississippian Genus of Productidae from the Western United States ABSTRACT Carlinia, new genus (type-species: Productus phillipsi Norwood and Pratten, 1855) occurs in rocks of late Chester age in Nevada, Utah, and Wyoming. It is a late derivative of Diaphragmus Girty in which the multiple diaphragms and trails in the brachial valve become modified into a series of concentric frills, very closely spaced along the valve margin. The diaphragm area is not differentiated within the interior of the brachial valve. Carlinia also differs from Diaphragmus in lacking spines on the brachial valve and in having a thicker shell. Besides the type-species, two other spe- cies, Carlinia amsdeniana, new species, and C. dia- bolica, new species, are described. All three species have a limited stratigraphic range near the top of the Mississippian System. Diaphragmus cestriensis (Wor- then), D. fasciculatus (McChesney), and D. monte- sanae Ulrich are discussed in relation to Carlinia and its species. Three distinctive species of diaphragm-bearing pro- ductoids have a limited stratigraphic range near the top of the Mississippian System. These shells have char- acters in common that differentiate them from other members of the family Productidae and are believed to belong in a new genus, herein named Carlinia for its occurrence in Carlin Canyon, Nevada. The relatively short time-span of this genus renders these species use- ful for stratigraphic correlation in that part of the western United States where they occur. Mackenzie Gordon, Jr., United States Geological Survey, Room E—315, United States National Museum, Washington, D.C. 20242. Publication authorized by the director, United States Geological Survey, Washington, D.C. In the United States the family Productidae ranges through the upper part of the Meramec Series to the top of the Chester Series (see Table 1). Its earliest known representative is Productus tenuicostus Hall, which occurs in the St. Louis Limestone and in the Arroyo Penasco Formation of northern New Mexico. This species bears a single diaphragm along the margin of the visceral disc in the brachial valve. It is suc- ceeded, in the St. Genevieve Limestone and in early Chester beds, by Diaphragmus montesanae Ulrich, a small, finely costellate species with two or three dia- phragms and delicate trails. In middle early Chester time the familiar and wide- spread D. cestriensis (Worthen), the type-species of the genus Diaphragmus Girty, appeared along with the closely related D. fasciculatus (McChesney). In mid- dle and late Chester time these species spread from the Mississippi Valley and ranged rather widely over the United States. D. cestriensis is recognized from western Wyoming to northern Alabama. Several other species of Diaphragmus also can be recognized in various parts of the United States, but they have more restricted geographic ranges than D. cestriensis. The tendency in the evolution of the brachial valve in Diaphragmus was for the initial disc to become proportionately smaller, the diaphragm area wider, and individual diaphragms and trails more numerous. Typical Diaphragmus has as many as seven diaphragms and trails in the brachial valve. The outer surface of the initial trail is commonly covered with scattered spines along costae and costellae; spines generally are most numerous on the brachial valve opposite the spine patch commonly present on the flanks and umbonal slopes of the pedicle valve. 257 258 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY TABLE 1.—Distribution of species of the family Productidae in the Upper Mississippian of the United States. Species Meramec Chester Lower Middle Upper Lower Middle Upper Productus tenuicostus X Diaphragmus montesanae D. cestriensis X X - - X X X - - D. fasciculatus Carlinia diabolica - X X X - - XX- C. phillipsi C. amsdeniana - X - - X - D iaphragmus, undescribed species - - X In fairly late Chester time the modified form herein designated as Carlinia appeared on the scene in the western United States. The multiple diaphragms and trails of this genus are more numerous than in Dia- phragmus but seem to have served mainly as external ornamentation rather than having the function of separating the visceral cavity from the trail, as in Diaphragmus. The trails in Carlinia do not appear to have grown to any appreciable length, and the dia- phragms were covered with callus internally so that they cannot be distinguished on the interior of the shell. The spines on the ventral valve became sparser and coarser in Carlinia and those on the brachial valve disappeared entirely. The acme of the genus Carlinia was near the end of Chester time. Nevertheless, the youngest member of the Productidae in the Great Basin, in the highest Mississippian beds along the Nevada-Utah border, is a species of Diaphragmus as yet undescribed. Carlinia is not found in these highest beds. The family Pro- ductidae, as such, appears to have died out suddenly at the end of Chester time, but probably gave rise to other forms in the later Carboniferous. I thank A. J. Unklesbay and R. L. Etherington of the Department of Geology of the University of Mis- souri for the loan of Branson and Greger's figured spec- imens, which herein are designated holotype and paratype of Carlinia amsdeniana, new species. I am grateful also to J. T. Dutro, Jr., and R. E. Grant for their review of this paper and helpful suggestions. The following abbreviations are used in this paper: USGS, United States Geological Survey; USNM, United States National Museum; UM, University of Missouri; UW, University of Wyoming. Superfamily PRODUCT ACE A Gray, 1840 Family PRODUCTIDAE Gray, 1840 Carlinia, new genus DIAGNOSIS.—Productids with rather coarse uneven costation, bearing scattered spines on pedicle valve. Brachial valve nearly flat to moderately concave, with large diaphragm area bearing numerous concentric frills externally; spines absent. Diaphragm area not clearly delimited internally. DESCRIPTION.—Shell of moderate size, somewhat elongate. Pedicle valve fairly evenly arched longitudi- nally; curvature decreasing gradually from umbo anteriorally; venter elevated, somewhat flattened medially, with or without sulcus; flanks fairly steep; valve flaring rather widely toward anterior margin. Pedicle valve ornamentation of costae and costellae, weak indistinct rugae, and spines. Costae and costellae somewhat irregular in strength, with subequal to narrow intercostal sulcae; spines normally rather coarse, suberect, occurring: (1) scattered over shell on costae and commonly disposed in rows along thicker costae, directed forward on anterior slope; (2) in ran- dom group on flanks and on ears; and (3) usually sparsely and weakly along hinge. Pedicle valve interior having narrow, shallow muscle scar platform flanked by relatively small weak diductor muscle scars; without identifiable endospines. Brachial valve having small concave initial part bounded by narrow concentric ridge; diaphragm area large, concentric frills spaced 2 to 3 mm apart posteri- orly and bunched anteriorly. Costae weak over initial NUMBER 3 259 unfrilled part of valve, absent over diaphragm area except locally on frills; spines absent on valve. Brachial interior having visceral disc not clearly di- vided into prediaphragm and diaphragm areas; cardi- nal ridges moderately strong; median septum stout posteriorly; posterior platform supporting short sessile cardinal process, like that of Diaphragmus. Anterior and posterior pairs of adductor muscle scars well differ- entiated, on prominent raised platforms at either side of median septum, quadrilobate in overall aspect; brachial ridges enclosing slightly raised areas; no identifiable endospines. TYPE-SPECIES.—Productus phillipsi Norwood and Pratten, 1855. DISCUSSION.—This genus differs from Diaphragmus in having a very short trail, in having more irregular and usually coarser costation, in lacking a well-defined spine cluster on the flanks, in lacking spines on the brachial valve, and in having more than 10 concentric frills over the diaphragm area becoming more numer- ous and bunched together anteriorly. Carlinia is a late offshoot of Diaphragmus, in which the trails developed so fast and so weakly that they never extended very far beyond the floor of the brachial valve but either broke off or just did not grow long. Spines also failed to develop on the brachial valve. This genus appears to have been a mud-lover and is generally found in calcareous shale and siltstone, rarely in impure limestone. The name Carlinia is derived from Carlin Canyon near Carlin, Nevada, which is almost certainly the type-locality of its type-species, and where a second species, Carlinia diabolica, new species, is fairly common. STRATIGRAPHIC CONSIDERATIONS.—This genus is re- stricted largely, and perhaps entirely, to late Chester rocks in the western United States. It is particularly abundant in those Late Mississippian rocks that overlie the Caninia excentrica coral zone (K zone of Dutro and Sando, 1963). In the Great Basin region of Utah and Nevada, however, Diaphragmus ranges a little higher than Carlinia. Carlinia phillipsi (Norwood and Pratten) PLATE 1: FIGURES 1-8, 21 Productus phillipsi Norwood and Pratten, 1855, p. 8, pi. 1, figs. 2a-c. Productus(?) phillipsi Norwood and Pratten. Sutton, 1938, p. 559. 372-386 0—71 -18 DIAGNOSIS.—Slightly elongate-oval Carlinia with hinge-line equal to three-fourths to four-fifths greatest width of shell; 20 to 26 costae and costellae on pedicle valve. DESCRIPTION.—Shell generally slightly longer than wide, subcircular to suboval in plan; widest at or just anterior to middle. Pedicle valve strongly curved pos- teriorly; curvature lessening gradually anteriorly; umbo protruding 1 or 2 mm across hinge, moderately inflated, its sides diverging initially at approximately a right angle; flanks steep. Venter flattened, either with or without broad shallow sulcus on anterior slope; ears rather small, set off weakly from umbonal slopes by broad sulci. Valve somewhat flattened and tending to flare anteriorly. Pedicle valve ornamentation consisting typically of 7 or 8 coarse ribs with wide and fairly deep intercostal sulci over ventral part of valve, commonly with nar- rower costellae between and on flanks, increasing by bifurcation; 20 to 26 costae and costellae present near anterior margin; considerable variation in arrangement and strength of costation among specimens. Rugae very indistinct, confined to posterior third of valve. Spine bases fairly prominent, suberect: (1) distributed in alternating pattern on stronger costae, 7 to 15 mm apart on single costa, (2) scattered in weak row or rows along hinge, strengthening toward ears, and (3) in two or three diagonal rows on ears and along base of umbonal slope, forming group of 6 to 9 spines just anterior to ears and on flanks. Interior of pedicle valve having slightly raised mid- ventral adductor muscle scar platform, about 9 mm long and 3 l/u mm wide, beginning 5 mm in front of umbo. Posterior pair of adductor muscle scars elongate, lenticular, weakly dendritic behind and smooth in front, separated by shallow median groove; anterior pair of adductor muscle scars indistinct, occupying V- shaped area enclosing anterior end of posterior pair and joining in oval area at anterior end of muscle scar platform. Diductor muscle scars occupying short fan- shaped areas at either side of adductor muscle scar platform, beginning 7 mm in front of umbo, rather coarse posteriorly, weaker and more finely radial an- teriorly. No endospines observed. Brachial valve gently concave; initial disc slightly more so, extending to approximately one-half the length of valve, edged by narrow raised ridge or lamella; broad shallow medial ridge arising anterior to initial disc, strongest near margin of valve. Part of 260 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY valve outside of initial disc, valve consisting of frilled multiple diaphragm area, frills becoming more closely spaced anteriorly and crowded near margin. Costae narrow, weak, indistinct on initial disc, even weaker over diaphragm area but expressed by radial rows of frill convolutions. Trail very short on most brachial valves but moderately long and curving from visceral disc on some. Interior of brachial valve with initial disc not dif- ferentiated from diaphragm area; ears pointed and set off from rest of valve by marginal embayment just in front of them. Cardinal ridges 1 prominent, acute, con- tinuing outward and forward in curve across ear to valve margin, separating ear from visceral disc. Pos- terior platform massive, supporting cardinal process of marginiferid type; shaft short; myophore bifid above, divided by narrow median groove, trifid poste- riorly. Median septum stout posteriorly, narrowing sud- denly between adductor muscle scar platforms and extending about three-fifths of way across valve. Ad- ductor muscle scars on prominent elevated subtriangu- lar platforms; posterior pair dendritic, sprawling, anterior pair smooth, on club-shaped elements with narrow ends converging slightly anteriorly. Brachial ridges rather obscure, given off from middle of anterior pair of adductor muscle scars at angle of 25 degrees to hingle-line, curving outward, then inward and back- ward to enclose two subtriangular areas about 5 by 7 mm across, their farthest anterior extremities along line 7 mm behind anterior edge of valve. Dimensions (in millimeters) of two hypotypes, both occurring near the top of the Chainman Shale, are as follows: USNM USNM Dimension 144014* 144015] Maximum length 28. 5 30. 8 Length along arch 51.0 59. 0 Maximum width 26.0 24. 6 Width of hinge 19.0$ 17. 8 Depth 15.0 17.5 *From USGS loc. 20460-PC. tFrom USGS loc. 17217-PC. JTwice the half-width. DISCUSSION.—The description and illustrations of Carlinia phillipsi are based upon specimens from the upper part of the Chainman Shale at Conger Spring in the Confusion Range and at Jensen Wash in the Burbank Hills, both in Millard County, Utah. As the 1 The term "cardinal ridges" is used in preference to "lat- eral ridges." partially silicified specimens at these localities did not yield well-preserved interiors suitable for illustration, the brachial valve interior is illustrated by a specimen from the type section of the Manning Canyon Shale in Soldier Canyon, Oquirrh Mountains, Tooele County, Utah (Plate 1: figure 21). Norwood and Pratten's type-specimen, the where- abouts of which is not known to the writer, was well illustrated in the original report and shows the char- acteristic sculpture of this species. The specimens from Conger Spring have about the same relative propor- tions as those given for the holotype but tend to be coarsely costate. Those from Jensen Wash tend to be narrower than normal and ornamented by alternating costae and costellae. Those at Soldier Canyon are broad and tend to have costae of subequal strength. In general, the very coarse ribs and spines of this species are its identifying characters. TYPE-LOCALITY.—According to Norwood and Prat- ten (1855, p. 8) the holotype of this species came from "Big Canyon of Humboldt River, Utah Territory." Only two large canyons are present along the Hum- boldt River: Carlin Canyon, a few miles east of the town of Carlin, and Palisades Canyon, a few miles southwest of Carlin. No upper Paleozoic rocks are ex- posed in Palisades Canyon, but in the eastern part of Carlin Canyon, near Tonka Siding, limestone of Late Mississippian age in the upper part of the Diamond Peak Formation crops out on both sides of the Hum- boldt. Light gray limestone in three fault blocks on the south side of the river has yielded fossils normally as- sociated with Carlinia phillipsi. In two of the fault blocks, at the west side of a small north-trending tribu- tary valley, 0.1 and 0.2 mile south of the river, respec- tively, Carlinia diabolica, new species, is included in this association. Although the writer has not found C. phillipsi in Carlin Canyon, he believes that the holo- type must have come from these beds. TYPES.—Hypotypes, USNM 144014-144020, inclu- sive (7 specimens). DISTRIBUTION.—Carlinia phillipsi ranges from near Stockton, Utah, to near Eureka, Nevada. It is limited to a stratigraphic zone roughly 50 to 60 feet thick in the upper part of the Chainman Shale and strati- graphically equivalent beds in Utah in the lower part of the Manning Canyon Shale and in Nevada in the upper part of the Diamond Peak Formation. The top of this zone lies at levels of from roughly 60 to 150 feet below the top of the Mississippian System. NUMBER 3 261 OCCURRENCE.—Chainman Shale, upper part, USGS loc. 17217-PC, fossils weathering from calcareous shale on slope in measured section in Jensen Wash, Burbank Hills, 80 to 135 feet below top of Chainman Shale in SW^SE^NW^ sec 35, T 22 S, R 18 W; USGS loc. 20460-PC, fossils in 5-foot calcareous shale bank, 100 feet north of Conger Spring, Confusion Range, Millard County, Utah. Manning Canyon Shale, USGS loc. 14515-PC, 10-foot platy limestone bed 405 feet above base in measured section in Soldier Canyon, one-half mile upstream from mouth, Oquirrh Mountains, Tooele County, Utah. Carlinia amsdeniana, new species PLATE 1: FIGURES 9-13, 22, 23 Diaphragmus phillipsi (Norwood and Pratten). Branson and Greger, 1918, p. 314, pi. 19, figs. 5, 6. Marginifera muricatina Dunbar and Condra. Burk, 1954 p. 10, 11, pi. 1, figs. 26-28. DIAGNOSIS.—Rounded-subpentagonal Carlinia of subequal length and width; pedicle valve normally having V-shaped median sulcus, surface ornamented by 24 to 35 costae and costellae and numerous scat- tered spines including about 20 in area of hinge and ear. Brachial valve transversely subrectangular. DESCRIPTION.—Pedicle valve having greatest width usually at ears just in front of hinge. Valve convex throughout longitudinal profile; curvature greatest at umbo and decreasing a little unevenly but gradually toward anterior margin. V-shaped median sulcus (not quite central in lectotype) present in some specimens, commonly beginning less than 5 mm in front of umbo, which protrudes slightly across hinge; in other speci- mens, area Of sulcus is merely flattened or slightly de- pressed. Anterior and lateral slopes diverging toward margins; umbonal slopes steep and curving gradually outward so that ears are only moderately set off by broad diagonal sulci. Umbonal slopes diverging an- teriorly at angle of roughly 110 degrees. Surface of pedicle valve ornamented by somewhat irregular and subequal costae interspersed locally with costellae, averaging 30 in all but ranging from 25 to 35 in various specimens. Costae increase by bifurcation and some tend toward fasciculation anteriorly. Rugae absent or merely suggested by two or three faint con- centric ridges in umbonal region of some specimens; concentric growth lirae not prominent. Spine bases abundantly scattered over valve: (1) on top of costae, commonly 1 to 2 mm apart over anterior slopes and 2 to 2J/* mm apart along individual costae; (2) in two rows of 6 to 7 spine bases at either side of umbo, one along hinge margin, other just in front at a very low angle diagonal to hinge and extending on to ears, merging with (3) cluster of about 12 to 15 spines scat- tered over umbonal and lateral slopes near ears. Interior of pedicle valve with low medial platform- like structure beginning 2 to 3 mm anterior to umbo and up to 3 mm wide, consisting of two elongate smooth lobes divided by low median ridge; adductor muscle scar area begins at its anterior end, occupying space 4 mm long and 1.3 to 1.7 mm wide between diductor muscle scars. Posterior pair of adductor muscle scars narrowly dendritic, diverging slightly pos- teriorly; anterior pair nearly smooth, situated in elongate parallel depressions opposite anterior parts of diductors. Diductor muscle scars transversely oval, longitudinally Urate, 4 mm by 5 mm across, beginning 4 to 5 mm in front of umbo. Rest of pedicle interior weakly costate; no endospines visible. Brachial valve transversely subrectangular in outline, gently concave medioposteriorly, somewhat less con- cave in lateral areas just in front of ears; also concave near anterior margin, except at shallow fold, commonly present, corresponding to sulcus of pedicle valve. Sur- face ornamentation of fine radial costellae limited to initial part of valve, bounded by single tiny concentric frill approximately at middle of valve. Two or three more such frills having wide but gradually narrower interspaces outward surrounding first one; band of 7 to 10 concentric wavy frills just within valve margin; all appear to represent trails that never fully developed. Interior of brachial valve not known. Dimensions (in millimeters) are as follows: Dimension Holotype UM 2645 Paratype USNM 163760 Paratype UM 2645 Length 25. 1 23.5 18.5 Length along curvature 41.0 40.5 32.5 Greatest width 25.0 23.5 20.9 Width of hinge 22.5 18.5 19.2 Depth 13.3 12.5 9.7 DISCUSSION.—Branson and Greger (1918) de- scribed and figured this species as Diaphragmus phillipsi (Norwood and Pratten). Their two figured specimens (UM 2645) have been examined and are refigured (Plate 1: figures 9-13, 22, 23) in this 262 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY report. The larger has been selected as the holotype of Carlinia amsdeniana and the smaller has been des- ignated a paratype. Dimensions of these two specimens are given above. Included among the paratypes is a lot of 23 speci- mens collected by Keyte near Lander, Wyoming. The dimensions of one of these specimens (USNM 163760) are given above. Burk (1954) described and figured two broken speci- mens from South Pass, Fremont County, Wyoming, as Marginifera muricatina Dunbar and Condra. Both specimens (UW IT-197 and IT-198) have been compared with the primary types of Carlinia amsdeniana and are judged to be conspecific. Carlinia amsdeniana is a distinctive species that differs from C. phillipsi (Norwood and Pratten) in having a slightly wider shell, commonly with a V- shaped median sulcus in the pedicle valve; a larger number of costae; much wider hinge and ears; a much greater number of spine bases scattered over the surface, particularly along the hinge and near the ears; and a transversely subrectangular brachial valve. Carlinia diabolica, new species, differs from C. amsdeniana in lacking a V-shaped median sulcus, in its finer surface ornamentation of 35 to 50 costellae, and in having fewer and coarser spine bases. TYPES.—Holotype, UM 2645 (part) ; paratypes, UM 2645 (part, 1 specimen) ; paratypes, USNM 163760-3. DISTRIBUTION.—Carlinia amsdeniana apparently is restricted to the Spirifer welleri zone of Shaw (1955), which is very late Chester (Late Mississippian) in age, in Fremont County, Wyoming. OCCURRENCE.—Amsden Formation, Horseshoe Shale Member, near middle; UM 2645 (USGS loc. 19283-PC), Cherry Creek, sec 19, T 31 N, R 99 W; USNM 163760-1, near Lander, Fremont County, Wyoming. Carlinia diabolica, new species PLATE 1: FIGURES 14-20, 24, 25 Productus elegans Norwood and Pratten. Hall and Whit- field, 1877, p. 268, 269, pi. 5, figs. 3, 4. Diaphragmus elegans (Norwood and Pratten). Girty, 1920, pi. 53, fig. 8. DIAGNOSIS.—Subcircular to subpentagonal Carlinia with greatest width near hinge. Pedicle valve orna- mented by 35 to 50 costae and costellae; coarse spines scattered sparsely over surface, 6 to 7 on each flank near ear. Brachial valve subtrapezoidal in plan. DESCRIPTION.—Shell subcircular in plan, except for protruding ears giving subpentagonal aspect; visceral cavity of moderate depth. Pedicle valve moderately arched; curvature decreasing rather gradually from umbo, except for slight geniculation 7 to 8 mm in front of umbo. Venter and flanks merging in gradual convex curve, venter either rounded or somewhat flattened; some specimens possessing broad shallow median sul- cus on anterior slope. Ears rather large, not well differentiated from sloping lateral and umbonal slopes but merging in broad concavity. Umbo narrow, pointed, protruding not quite 2 mm in front of hinge, sides diverging anteriorly at approximately a right angle; umbonal region somewhat depressed. Pedicle valve ornamented by fine irregular costae or costellae, weak rugae, and scattered spines. Costae low, rounded, with narrow intercostal sulci, increas- ing by bifurcation and tending toward fasciculation, particularly near spine bases; some costae pairs merg- ing into single costa anteriorly; normally 14 to 15 costae occurring in space of 10 mm in middle of PLATE 1.—All views natural size unless otherwise indi- cated. Figures 1-8, 21.—Carlinia phillipsi (Norwood and Pratten) : 1-3, Pedicle, side, and brachial views of USNM 144015, from USGS loc. 17217-PC, Burbank Hills, Utah; 4, enlarged posterior view (X V/i) of USNM 144017, from the same locality; 5-7, pedicle, side, and brachial views of USNM 144014, from USGS loc. 20460-PC, Confusion Range, Utah; 8, interior of pedicle valve from rubber cast of an in- ternal mold (X l'/2); USNM 144016, from USGS loc. 17217-PC; 21, interior of brachial valve from USNM 144019, from USGS loc. 14515-PC, Oquirrh Mountains, Utah. Figures 9-13, 22, 23.—Carlinia amsdeniana, new species: 9-11, 22, 23, pedicle, side, and brachial views and enlarged posterior and side views (X 2) of the holotype UM 2645, Wind River Range, Wyoming; 12, 13, pedicle and side views of a paratype, UM 2645, from the same locality. Figures 14-20, 24, 25.~Carlinia diabolica, new species: 14-16, 25, pedicle, side and brachial views and enlarged side view (X 2) of the holotype USNM 144021, from USGS loc. 16993-PC, Confusion Range, Utah; 17, enlarged interior view of a specimen collected by the 40th Parallel Survey and paratype, USNM 144024, from USGS loc. 20460-PC, Confu- sion Range, Utah: 18, 19, side and pedicle views of a para- type, USNM 144025, from USGS loc. 16993-PC; 20, pedicle view of a specimen collected by the 40th Parallel Survey and figured by Hall and Whitfield (1877, pi. 5, figs. 3, 4) as Pro- ductus elegans and by Girty (1920, pi. 53, fig. 8) as Diaphrag- mus elegans, from Oquirrh Mountains, Utah; 24, enlarged view (X V/i) of interior of a brachial valve, a rubber cast from a paratype, USNM 144024, from USGS loc. 20460-PC. NUMBER 3 263 PLATE 1 264 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY anterior slope; costae averaging about 40 in all, weak to absent near ears. Rugae fine, weak, confined to flanks and ears on posterior half of valve. Spines rather stout, suberect; spine bases occur: (1) scattered over venter and anterior and lateral slopes, 2 to 7 mm apart, mainly along rows of costae where fasciculation is greatest, 3 to 11 mm apart on single rows; (2) in row of 5 or 6 along hinge margin at either side of umbo and generally two vague rows of 2 or 3 in sulcus at base of ear and also on ear at considerable angle to hinge; and (3) in group of 5 or 6 large spines on flanks in front of ears. Interior of pedicle valve having prominent bilobate raised adductor muscle scar platform beginning 5 mm in front of umbo, 4 mm wide and 8 mm long; posterior and anterior adductor scar pairs dendritic. Diductor scars deeply grooved, fan-shaped, beginning 7 mm in front of umbo and extending about 6 mm. Anterior part of valve interior finely and irregularly lirate, with numerous tiny endospines. Brachial valve very shallowly concave; initial part of valve shallowly concave, bordered with acute raised ridge; area along hinge about 2 mm wide, smooth from ear to ear; rest of valve consisting of multiple dia- phragm area, ornamented by concentric frills 1 or 2 mm apart near initial part and becoming ever more closely spaced anteriorly; frilled area with broad shal- low medial ridge anteriorly. Fine costae and rugae indistinct but present on initial part of valve; some costae discernible in diaphragm area. Interior of brachial valve having rather thick rounded cardinal ridges supporting short sessile cardinal process, weakly bifid above and trifid poste- riorly and externally; posterior part of septum fairly thick, marked by shallow median groove; septum narrowing markedly in front of adductor muscle scar platforms and extending to point about two-thirds of way across valve. Pair of adductor muscle scar plat- forms forming four pear-shaped lobes standing in low relief anteriorly and sloping posteriorly; posterior pair of muscle scars dendritic, anterior pair appearing smooth. Brachial ridges anterior to and at either side of muscle scars enclosing suboval areas roughly 3 mm by 4 mm across near lateral margins of valve and be- hind line 4 mm from anterior end of valve. Dimensions (in millimeters) of holotype and a para- type: Holotyp e USNM 144021 Paratype USNM 144022 23.6 23.5 45.0 45.0 25.7 *28. 0 24.0 *25. 0 14.0 14.5 Dimension Length Length along arch Maximum width Width of hinge Depth *Twice the half-width DISCUSSION.—The description of the external char- acters of this species is based on specimens from Wallet Gulch on the west slope of the Confusion Range, Utah. The internal characters are from a paratype from Conger Spring, two miles southeast of Wallet Gulch, where this species occurs in association with C. phillipsi (Norwood & Pratten). Variation in Carlinia diabolica most frequently in- volves the number of costae and costellae, number and distribution of spines, and presence or absence of a shallow ventral sulcus. The number of costae and costellae on individual specimens ranges, roughly, from 35 to 50. Those with finer costation generally have more rows of spines. Some specimens have more spines along the hinge and also on the flanks near the ears, but the spines are distributed randomly and never form an isolated spine cluster as in species closely re- lated to Diaphragmus cestriensis (Worthen). Carlinia diabolica can be mistaken for shells of Diaphragmus cestriensis because the number and strength of the costae are similar; it may be distin- guished, however, by its wide ears, less steeply sloping flanks, less protruding umbonal region, absence of a prominent spine cluster on the flanks just in front of the ears in the pedicle valve, the absence of a long tail or of spines on the brachial valve and the presence, in their stead, of concentric frills on the brachial valve. A productid from the Oquirrh Mountains, Utah, described and illustrated by Hall and Whitfield (1877) as "Productus elegans Norwood and Pratten" [=Diaphragmus cestriensis (Worthen)] has been com- pared with the primary types of C. diabolica and found to be conspecific. The same specimen was figured by Girty (1920). This specimen, recorded by geologists of the 40th Parallel Survey as coming from north of Snowstorm Hill, Dry Canyon, Oquirrh Mountains, is refigured here (Plate 1: figure 20). The new species is readily distinguished from C. phillipsi by its finer sculpture, wider hinge, and promi- nent ears. This species can be distinguished from C. amsdeniana by its more numerous costae and costellae, NUMBER 3 265 fewer and coarser spines, and by its general lack of a narrow median sulcus in the pedicle valve. TYPES.—Holotype, USNM 144021; paratypes, USNM 144022-5, inclusive (7 specimens). Figured specimen, USNM 14216. DISTRIBUTION.—This species is recognized in the Great Basin from the Oquirrh Mountains, at the east, to the Carlin region, Nevada, at the west. It is found throughout at least 150 feet of the section at Wallet Gulch in the upper part of the Chainman Shale. In the Oquirrh Mountains it probably comes from the Manning Canyon Shale, and in the Carlin region it has been collected in the upper part of the Diamond Peak Formation. OCCURRENCE.—Primary types from the Chainman Shale, upper part (USGS loc. 16993-PC), 50 feet stratigraphically below and east of base of Ely Lime- stone. Additional paratypes from 5-foot calcareous shale bank (USGS loc. 20460-PC) 100 feet north of Conger Spring, Confusion Range, Millard County, Utah. Figured specimen from unidentified formation, probably the Manning Canyon Shale, Dry Canyon, Oquirrh Mountains, Utah. Literature Cited Branson, E. B., and D. K. Greger 1918. Amsden Formation of the East Slope of the Wind River Mountains of Wyoming and Its Fauna. Geo- logical Society of America Bulletin, 29: 309-326, 2 plates. Burk, C. A. 1954. Faunas and Age of the Amsden Formation in Wyoming. Journal of Paleontology, 28.(1): 1-16, plate 1. Dutro, J. T., Jr., and W. J. Sando 1963. New Mississippian Formations and Faunal Zones in Chesterfield Range, Portneuf Quadrangle, Southeast Idaho. American Association of Petro- leum Geologists Bulletin, 47(11): 1963-1986,6 figures. Girty, G. H. 1920. Carboniferous and Triassic Faunas [of Utah]. United States Geological Survey Professional Paper, 111: 641-648, plates 52-57. Hall, J., and R. P. Whitfield 1877. Paleontology. United States Geological Exploration of 40th Parallel, 4(2): 197-302, 7 plates. Muir-Wood, H. M., and G. A. Cooper 1960. Morphology, Classification and Life Habits of the Productoidea (Brachiopoda). Geological Society of America Memoir, 81: 1-447, 135 plates, 8 figures. Norwood, J. G, and H. Pratten 1855. Notice of Producti Found in the Western States and Territories, with Descriptions of Twelve New Species. Academy of Natural Sciences Philadelphia Journal, series 2, 3: 5-22, 1 plate. (Preprint dated 1854.) Shaw, A. B. 1955. The Amsden Formation in Southwestern and South-Central Wyoming. Wyoming Geological Association Guidebook, 10th Annual Field Confer- ence, Green River Basin, 1955, pages 60-63. Sutton, A. H. 1938. Taxonomy of Mississippian Productidae. Journal of Paleontology, 12(6): 537-569, plates 62-66, 2 figures. M. J. S. Rudwick The Functional Morpholog y of the Pennsylvan ian Oldhaminoi d Brachiopod Poikilosako s ABSTRACT The morphology of Poikilosakos, the earliest known oldhaminoid, is analyzed functionally. It is inferred that (1) much of the ventral mantle tissue was perma- nently exposed beyond the edge of the pinnate dorsal valve; (2) the normal external growth lines on the dorsal valve indicate that it was not sheathed in mantle tissue but (3) it grew according to the same morpho- genetic "rules" as later oldhammoids; (4) the hinge, although aberrant, indicates habitual movement of the dorsal valve rotating vertically; (5) the right di- ductor was atrophied except in a rare variant form; (6) the axis of a ptycholophe was borne near the edge of the dorsal valve; and (7) the feeding mechanism may have been of a rhythmic type analogous to that known in septibranch mollusks and inferred in rich- thofeniid brachiopods. The phyletic significance of this functional reconstruction is considered briefly; it is concluded that the oldhaminoids may have been de- rived neotenously (and hence cryptogenetically) from davidsoniacean ancestors, and that their extinction should not be attributed too readily to over-narrow adaptive specialization. The oldhaminoids are of exceptional interest in the study of brachiopod evolution. In form and structure they are more peculiar and aberrant than any other members of the phylum, and they seem to indicate the fullest extent to which the basic organization of the brachiopod body has ever been modified in the course M. ]. S. Rudwick, Department of Geology, University of Cambridge, Cambridge, England. of evolution. Their structural peculiarities suggest that their mode of life may have been correspondingly un- usual; therefore, they are important in any functional or adaptive interpretation of brachiopod evolution. Such aberrant fossils have naturally attracted much attention, and interpretations of their structure and mode of life have been varied and controversial. Pa- pers by Williams (1953) and Stehli (1956) summarize some of the points at issue, and refer to earlier works. An evaluation of divergent views has been hampered, however, by the small number of well-preserved speci- mens which have been adequately illustrated. For the later, Permian, members of the group this will no doubt be rectified by the publication of Dr. G. A. Cooper's description of the magnificent silicified material from the Glass Mountains in western Texas. Meanwhile, it seems worthwhile to analyze the earliest member of the group, which is simplest in structure and, by com- mon consent, likely to be the ancestor of all later mem- bers; for in its functional interpretation there may be found a key to the understanding of the aberrant struc- ture of the group as a whole. Poikilosakos was first described half a century ago by D. M. S. Watson (1917), from 20 specimens collected in the Graham Group (late Pennsylvanian) of north- central Texas. These specimens are now in the British Museum (Natural History), abbreviated herein as BMNH. They came from "a well-known locality in the west bank of the Salt Creek at Graham, Young County, Texas." This was probably in the Wayland Shale, the topmost member of the Graham Group, which consists of clay shales with partings of earthy limestone. In the United States National Museum 267 268 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY (USNM) there are other specimens from the Wayland Shale, from Gunsight, Stephens County, Texas, and also specimens from the Gonzalez Formation, in the lower part of the Graham Group, from near Finis, Jack County, Texas. Also, I have restudied the speci- mens described by Dunbar and Condra (1932) from the base of the Plattsmouth Limestone Member of the Douglas Group, near Lawrence, Kansas, and now in the Peabody Museum, Yale University (YPM 13559) ; and a single important specimen from the same lime- stone at Williamsburg, Kansas, in the United States National Museum (USNM 137498). All these local- ities are at approximately the same horizon in the late Pennsylvanian. Poikilosakos was cemented throughout life to some hard material, on a substrate which must have been predominantly muddy. The surfaces used for attach- ment include large fragments of a massive shell with prismatic structure, productoid shells, large cylindrical corals, and crinoid stems. These organisms merely pro- vided debris suitable for the attachment of the spat of Poikilosakos: there is no evidence to suggest that any of them were alive at the time of attachment. I am grateful to Dr. H. Brunton and Dr. G. A. Cooper for access to specimens of Poikilosakos in the British Museum (Natural History) and in the United States National Museum, respectively, and for the loan of some specimens. Most of this work was done under a grant from the Science Research Council of Great Britain. Shell and Mantle GENERAL MORPHOLOGY OF THE SHELL (FIGURES 1, 4A).—Poikilosakos is the earliest and structurally the simplest of the oldhaminoids. But it already possessed most of the peculiar characteristics of the group, and has little immediate resemblance to most ordinary brachiopods. The ventral valve consists of a very thin sheet of calcite. Generally its outer surface is entirely adherent to the substratal surface, and the shell material is so thin that in parts of some specimens it is almost transparent. Its outer edge, therefore, is difficult to locate precisely, except where it has grown a little away from the substrate. For this reason, the true size of the valve is difficult to determine in most specimens. This thin sheet of shell material is thickened locally into a long sinuous ridge, termed by Watson as the "flange," which coalesces posteriorly with a short trans- verse "hinge ridge." The sinuous course of the flange marks out a series of blunt-ended "lobes" separated by narrower "indentations." There is a long and relatively conspicuous "median indentation," projecting back- wards towards the hinge ridge. The "lateral indenta- tions" are much less regular. Generally, the course of the flange is highly variable and shows only a very rough approximation to bilateral symmetry (Figure 2). Its irregularity is not due to that of the surface of at- tachment, since many highly irregular specimens are attached to almost smooth plane surfaces. Both outside and inside the flange, the surface of the valve is marked faintly with fine pustules (Plate 1: figure 8; Plate 2: figure 5). These probably indicate a pseudopunctate shell structure, as on other old- haminoid brachiopods. The area outside the flange may be termed the "peripheral zone" on the valve; the area inside, the "inner zone." The dorsal valve 1 is also a very thin plate, almost flat, and with a pinnate outline corresponding to the course of the flange on the ventral valve. Watson's specimens included no dorsal valves at all, but he correctly inferred that its pinnate edge would be found to rest on the narrow "shelf or rabbet" which runs around the inner side of the flange. Specimens since discovered, in which the dorsal valve is preserved in place, show that there is a tight fit between the edge of the valve and the shelf on which it rests. This may be termed a "secondary commissure": there is no true commissure, since the ventral valve extends on all sides beyond the edge of the dorsal. Between the dorsal valve and the inner zone of the ventral valve, within the flange, there is a very shallow "shell cavity." When 1 Williams (1953) has argued that the dorsal valve of oldhaminoids was, in fact, an "internal plate," homologous with the internal "lophophore platform" of some earlier brachiopods, and that the true dorsal valve is vestigial. But the term "dorsal valve," as ordinarily understood, includes not merely the external surface of the valve (generally cov- ered with primary-layer shell) but also all its various internal modifications and processes, such as the cardinalia, and in- cluding any so-called "lophophore platform." Even if Wil- liams' anatomical reconstruction is correct, it is therefore legitimate to continue to use the conventional term "dorsal valve'' for oldhaminoids, rather than the confusing and over- interpretative term "internal plate." Similarly, I prefer to use the clearly understood term "ventral valve" rather than "pedicle valve" for a structure in which no trace of a pedicle is known. NUMBER 3 269 accessory ridge median hollow hinge ridge diductor trough flange shelf peripheral zone FIGURE 1.—Ventral valve of Poikilosakos, diagrammatic, to illustrate morphology, A-B, Approximate line of section shown in Figure 4A. the dorsal valve is resting on the shelf, the shell cavity is completely isolated from the exterior. On the posterior side of the dorsal valve, there is a broad rectangular projection. Its straight posterior edge abuts against the anterior face of the hinge ridge. No specimen in which the internal surface of the dorsal valve is exposed has been available to me. RECONSTRUCTION OF MANTLE.—In order to secrete the peripheral zone, the ventral mantle must have extended outwards beyond the edge of the dorsal valve and mantle. Williams (1953, and in Moore, 1965) has argued convincingly that the ventral mantle of some later oldhaminoids, which was responsible for secreting a complex "posterior flap" in the peripheral zone, must have been highly contractile. If the tissue of Poikilo- sakos shared this property, the ventral mantle might normally have been retracted within the flange, and thus protected by the dorsal valve; and it might have been extended periodically to secrete new increments to the peripheral zone. But it is difficult to imagine how the ventral mantle could have been retracted from the posterior sector of the peripheral zone, because its retraction across the hinge ridge would have been barred by the abutting posterior edge of the dorsal valve. It is more satisfactory to conclude, with Williams, that the ventral mantle, although perhaps somewhat contractile, was permanently exposed beyond the edge of the dorsal valve. The dorsal valve would thus have lost its normal—and probably original—function of protecting all the soft tissues from the external environ- ment (cf. Sarycheva, 1964). In this respect Poikilosakos and other oldhaminoids resemble richthofeniids: the spines developed in the peripheral zone ("outer shell cavity") of Prorichthofenia would have prevented the ventral mantle from retracting within the protection of the dorsal valve (Rudwick 1961a). Williams (1953) postulated that the dorsal valve of oldhaminoids was entirely sheathed in mantle tissue, covering its external as well as its internal surface. This reconstruction was based on the known relation be- tween the mantle edge and the external primary-layer shell in living brachiopods, and on the absence of primary-layer (except a small area at the hinge) on the dorsal valve of oldhaminoids. This was criticized FIGURE 2.—Outlines of dorsal valves of Poikilosakos, arranged in approximate order of size, to show variability, and correla- tion of complexity with size. In most specimens (lightly stippled) the dorsal valve is not preserved, and its outline is inferred from the course of the shelf on the ventral valve flange. Specimens with dorsal valve preserved are more darkly stippled. The diductor troughs also are shown where visible. Key to specimen numbers: a, BMNH BB.11266 (holotype, as in Plate 1:figure 1); b, USNM 148051d; c, YPM 13559a; d, BMNH BB.l 1267a (as in Plate 2:figure 3); e, BMNH BB.54734a; f, BMNH BB.l 1269a (as in Plate 2:figure 6); g, BMNH BB.l 1269b; h, BMNH BB.54734b; i, USNM 148051c; ;, USNM 165838; k, USNM 165839 (as in Plate 1: figure 4); /, BMNH (unnumbered) ; m, USNM 148051b (as in Plate 1: figure 5; note paired diductor troughs); n, USNM 137498 (as in Plate 1:figure 7 and Plate 2:figures 1, 2); o, YPM 13559b; p, YPM 13559c; q, USNM 165840 (as in Plate 1: figure 3 and Plate 2: figure 5; lateral lobes foreshortened by being curved around crinoid stem, outlined by dashed lines) ; r, USNM 148051e;.r, USNM 165841 (as in Plate l:figure8; outline modified by circular base of crinoid cirrus) ; t, USNM 165842; u. USNM 148051a; v. BMNH BB.l 1267b (as in Plate 1: figure 2); w, USNM 165843 (as in Plate 1:figure 6); x, USNM 165844 (as in Plate 1: figure 4). 270 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY by Stehli (1956), who suggested that the absence of primary-layer more probably reflected some differences in the physiology of shell secretion in the oldhaminoids. To some extent this criticism has been justified by the discovery, emphasized recently by Williams (in Moore, 1965; 1968), that primary-layer is sporadically devel- oped, and sometimes absent, on strophomenide and other early brachiopods. Yet, Williams' interpretation is strongly supported, in my opinion, by the distinctive pustular external surface of the dorsal valve of most oldhaminoids (Rudwick and Cowen, 1968). This sur- face closely resembles the adjacent internal surface of the ventral valve. It is quite different from the external surface of the ventral valve, which, like the external valve surfaces of most other brachiopods, is marked with strong growth lines. In this context the presence of pustules or of growth lines on a valve surface may be a more reliable criterion than the absence or presence of primary-layer for de- termining which surfaces were covered with mantle tissue and which were not. It would hardly be doubted that the criterion is valid for the ventral valve of old- haminoids, so it may also be valid for the dorsal. There is a striking difference between the dorsal valve of Poikilosakos and that of later oldhaminoids such as Leptodus and Oldhamina. In the later old- haminoids, as already mentioned, the external surface of the dorsal valve is pustular, and resembles the adja- cent internal surface of the ventral valve. In Poikilo- sakos it is not pustular but smooth, and shows clear growth lines, indicating the course of its development during ontogeny (Plate 1: figures 6, 7; Plate 2: figures 1, 2, 4). It closely resembles the external surface of the ventral valve of other oldhaminoids (the ven- tral valve of Poikilosakos itself cannot be used for comparison, since its external surface is wholly adher- ent to the substratum). On any less-aberrant brachio- pod, such a surface with growth lines would be accepted without hesitation as a true external surface, which was not covered by mantle tissue. Specimens of Poikilosakos are too rare to justify sectioning a dorsal valve to determine its shell structure; but it appears to have primary-layer on its external surface. Even if primary-layer is absent, the clear growth lines indicate that the surface grew only by peripheral accretion, and this is strong circumstantial evidence that it was not covered by mantle tissue. It therefore is probable that the dorsal valve of Poikilosakos was lined with mantle tissue only on its internal surface, as in any normal brachiopod, and that it was not wholly sheathed in mantle tissue. If Wil- liams's interpretation is correct for some later old- haminoids, as I believe it is, the dorsal valve must have become sheathed with mantle tissue at some point in oldhaminoid evolution later than the stage represented by Poikilosakos. GROWTH-PATTERN OF THE DORSAL VALVE.—The growth lines on the dorsal valve show that fairly early in ontogeny it had a bilobed outline (Plate 2: figure 4) : the earliest stages of all are not clearly preserved. Beyond this stage the analysis rests perforce on a single good specimen. In this specimen (Plate 1: figure 7), the bilobed outline developed by differential secretion into a four-lobed form. In detail (Plate 2: figures 1, 2), each lobe can be seen to have grown principally by a somewhat irregular addition of arcuate increments at its distal end; there was little or no accretion on the sides of the lobe, with the result that it maintained a fairly uniform width while growing in length. The posterior pair of lobes grew laterally almost parallel to the hinge-axis; the anterior pair grew parallel to the posterior, maintaining a uniform width of indentation between them. At some point in the growth of these four lobes, a third pair seems to have been initiated by the budding of new accretionary material from the anteromedian corners of the anterior pair. These new lobes, like their predecessors, seem to have grown in such a way as to maintain slits of uni- form width both medially and anterolaterally. This growth-pattern has been analyzed elsewhere, in conjunction with evidence from later and more com- plex oldhaminoids, as the result of the operation of simple morphogenetic "rules" (Rudwick, 1968b). These "rules" would include two basic parameters, a standard lobe-width width (W) and a standard slit- width (S). The valve would develop, probably from an initial subcircular form, in such a way as to main- tain those parameters, each lobe growing parallel to a pre-existing lobe. There may then have been a critical lobe-length (K), on attaining which the anterior lobes would produce anteromedian buds, thus initiating further lobes. On the assumption of parameters such as these, the growth of a wide variety of oldhaminoid dorsal valves can be simulated realistically. The dorsal valve of Poikilosakos is amenable to analysis in these terms. In particular, the widths of the lobes and slits (or indentations) are comparable to NUMBER 3 271 those of all later oldhaminoids, and the dorsal valve growth lines show positive evidence for localized "budding," which can only be inferred circumstantially in other genera. But Poikilosakos is distinctive in two important respects: the dorsal valve is both simpler and more irregular in form than in most later genera. These features can be interpreted as the result of the same morphogenetic "rules" as in later genera, operat- ing on a valve of smaller absolute size, and also operat- ing with much less precision. RECONSTRUCTION OF BODY AND MANTLE CAVITIES.— The limits of the body cavity or coelom of Poikilosakos can be reconstructed approximately by homology with living brachiopods. In living articulates the body wall is closely wrapped around the anterior and lateral sur- faces of the muscles, and the coelom is confined to a very small posteromedian portion of the shell cavity. The muscles of Poikilosakos are difficult to interpret (see below), but they almost certainly were confined to the posteromedian portion of the shell cavity. In any case the coelom must have been greatly confined by the extreme shallowness of the space available between the valves (Figure 3). The remainder of the shell cavity would have formed an equally shallow "mantle cavity." When the dorsal valve was closed and resting on the shelf, the tight secondary commissure would have insured that the mantle cavity was completely isolated from the ex- terior. Thus, the dorsal valve must have retained at least a part of its original protective function. When closed it would have sealed off from the external en- vironment not only the "body" of the brachiopod (i.e., muscles, gut, etc.) but also the lophophore and any other organs in the mantle cavity. Only the peripheral tissue of the ventral mantle would have been perma- nently exposed, and unprotected by the dorsal valve. Hinge and Musculature STRUCTURE OF THE HINGE (FIGURE 3).—? The pos- terior edge of the dorsal valve abuts the anterior face of the hinge ridge. This anterior face is sharply defined and concave in the vertical plane (Plate 2: figure 6). Ventrally, it curves around into a very fine sharp ridge that runs parallel to the hinge ridge itself. This may be termed the "accessory ridge" (Plate 1: figure 8; Plate 2: figure 3). Dorsally, it curves around to the crest of the hinge ridge, behind which the ridge slopes away gradually and merges into the posterior sector of the peripheral zone. Although no dorsal valve interior has been seen in the course of this study, an isolated dorsal valve (USNM 137498) shows that the posterior edge is con- vex in the vertical plane, so that it could have fitted accurately against the concave surface of the hinge ridge. This implies that the actual articulation of the valves occurred between these cylindrical surfaces. Thus, the hinge axis of the rotation of the dorsal valve during its movements would have passed through the posterior edge of the dorsal valve. The anterior surface of the hinge ridge is marked with faint growth lines (Plate 2: figure 6) which show that during ontogeny it increased in size only by mar- ginal accretion. This suggest that it was not covered by mantle tissue. If it was an articulating surface, as it appears to have been, it is not surprising that it should have been an area of naked shell. The anterior face of the hinge ridge of Poikilosakos was identified by Watson as the vestigial "area," i.e., ventral interarea. This homology seems to be difficult to sustain. A true interarea is a part of the external surface of the valve, modified by being the growth track of the hinge line during ontogeny. The homol- ogous area in Poikilosakos would be the posterior sec- accessory axis rid ? adductor diductor ? cardinal process coelom ventra l mantle FIGUR E 3.—Reconstruction of hinge mechanism and muscles of Poikilosakos: left side of hinge region, cut in median section and seen from right. Form of cardinal process and position of adductor muscle hypothetical. 272 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY tor of the peripheral zone of the ventral valve, but this is adherent to the substratum and has lost all func- tional and spatial connection with the hinge mech- anism. Likewise, a faint median notch in the hinge ridge of some specimens was identified by Watson as the homologue of the delthyrium, but it too can only be regarded as an analogue. Anterior to the hinge ridge is a relatively deep de- pression, which may be termed the "median hollow." It is deepest and also narrowest immediately in front of the hinge ridge, where it interrupts the fine accessory ridge (Plate 3: figure 3). Anteriorly it becomes broader and shallower, and merges into the rest of the inner zone of the ventral valve surface. The posterior part of the median hollow probably is occupied fully by the cardinal process. Since no dorsal valve interior has been seen, the form of the cardinal process is uncertain. Flanking the median hollow laterally are the two triangular areas which Watson termed "dental areas." By giving them that name, he evidently regarded them as homologues, or at least analogues, of normal teeth and sockets. He noted that they are very slightly con- cave (not convex as stated by Williams in Moore, 1965), but it is not known whether there are corre- sponding convex areas on the dorsal valve. In any case they do not make convincing homologues for the con- vex teeth of normal brachiopods. Watson noted that they are marked with faint longitudinal striations (Plate 2: figure 5). These cannot be equivalent to the striations of the teeth of some other brachiopods, nor can they be vestigial denticles. Anteriorly they can be seen in some specimens to swing laterally, following the course of the rest of the flange; and this indicates that they are merely growth lines showing the stages of growth of the most posterior part of the shelf and flange. Watson inferred, from the asymmetrical muscle at- tachments (see below), that the dorsal valve did not rotate vertically but slewed laterally. But the edge of the dorsal valve, when closed, rests on the shelf within the flange. Therefore no lateral movement would have been possible unless the valve were first raised vertically enough to bring its edge clear of the crest of the flange. Even after this slight vertical rotation, a subsequent lateral movement seems highly improbable. Any such movement must have been centered around some sec- ondary, vertical pivotal axis, which would surely be reflected in some asymmetry of the hinge ridge and dental areas. In fact, the symmetrical and strictly transverse structure of the hinge region suggests strongly that the pivotal axis was exclusively horizontal and transverse in orientation, and that the movement was strictly vertical, as in any normal articulate brachi- opod. In any case, as both Williams and Stehli have pointed out, no lateral movement would have been possible in most of the later oldhaminoids, for in these brachiopods the peripheral zone of the ventral valve generally curves upwards on either side of the dorsal valve and would have confined its movement strictly to the vertical plane. The degree to which the dorsal valve of Poikilosakos habitually opened cannot easily be inferred. Judging by the orientation of the articulating surfaces, the maxi- mum possible angle of opening might have been as great as 45 degrees, but the usual angle may have been much less. In some later oldhaminoids, e.g., "Lyttonia conic a," the ventral valve developed into a narrow cone within which the maximum possible angle of opening was certainly quite small (Rudwick and Cowen, 1968). The articulation of Poikilosakos is highly modified from the hinge structures of more normal brachiopods, but it does not support Williams' suggestion that the dorsal valve of oldhaminoids was immobile. Given the existence of suitably arranged muscles, the dorsal valve of Poikilosakos would have been physically capable of rotating vertically through a small or large angle. The valves were not locked together by teeth and sockets, but in this respect the oldhaminoids merely resemble the many productoids in which teeth and sockets were lost. RECONSTRUCTION OF MUSCLES.—On a ventral valve of Poikilosakos, one muscle attachment is conspicuous. It is generally a narrow parallel-sided trough with raised margins, extending obliquely forwards from the posterior part of the median hollow. In detail its form is variable, but its outer edge is generally over- hanging, so that the opening of the trough faces obliquely inwards (Plate 2: figures 3, 6). Within the trough there are faint growth lines (Plate 2: figure 5), showing the stages of its enlargement away from the hinge region during ontogeny. Watson identified this trough as the ventral attach- ment of an adductor muscle, but in form and position it more closely resembles a diductor attachment and generally has been accepted as such. From the apex of the median hollow, where the cardinal process would have been situated, the muscle would have NUMBER 3 273 extended obliquely along the axis of the trough as a slender bundle of fibers. The raised margins of the trough would have increased the area for attachment of the fibers without increasing the area occupied on the floor of the valve, thus saving space in the con- fined body cavity. This attachment invariably lies on the anatomical left side of the ventral valve (i.e., right side of illustra- tions in conventional orientation). With one important exception (to be discussed below), there is no trace of any corresponding attachment on the anatomical right side. Watson marked two small areas on this side as being muscle scars, but I have been unable to detect them on his original specimens or on any others. In well-preserved specimens the whole of this area is covered uniformly with a finely pustular sur- face (Plate 1: figure 8; Plate 2: figure 5). It is diffi- cult to avoid the conclusion that the right diductor muscle of Poikilosakos had atrophied altogether. If so, it is likely that when a specimen with cardinal process is discovered it will be found to have corresponding asymmetry, i.e., a single lobe instead of the usual bilobed structure.2 This conclusion is supported by the existence of a single specimen in which the left diductor attachment is matched by another attachment of equal size on the right side. That specimen is from the same locality as other specimens having the usual single asymmetrical attachment. It may be interpreted as a variant, evi- dently rare, in which the right diductor was not atrophied. No specimen has been found in an inter- mediate condition, with a right diductor attachment reduced in size. Though the available sample is small, this suggests that the atrophy of the right diductor may have affected the population in an all-or-none manner. The pull of the diductor during contraction would have been nearly horizontal, but on a line running ventral to the inferred hinge-axis (Figure 3). This would have enabled the animal to rotate and open the dorsal valve by the normal leverage, provided only that the posterior edge of the dorsal valve was con- strained from sliding forward. In most brachiopods this constraint is given by the interlocked teeth and sockets. In Poikilosakos the accessory ridge may have been able to hold the posterior edge of the dorsal valve 2 Since completing this paper, I have been informed by Dr. Cooper that he has found specimens with a single-lobed asymmetrical cardinal process, as predicted here. in position; less probably, there may have been some specially toughened strip of mantle tissue along the inner side of the hinge line. Provided that the hinge was in some way held in position, the asymmetrical pull of a single diductor would not have affected the vertical nature of the rotation of the dorsal valve. Only the forwardly directed component of the contractile power would have been effective; the minor lateral component, although unbalanced by an equal and opposite component from a right diductor, could have been held by the articulation of the cardinal process against the side of the median hollow. As in most brachiopods, the moment of the diductor's leverage on the dorsal valve would have been extremely small. The extremely restricted shell cavity of Poikilosakos suggests a functional explanation for the atrophy of the right diductor muscle. A comparison with the unique symmetrical specimen shows that this atrophy was not accompanied by any substantial increase in the size of the remaining muscle (compare figures 5 and 4 of Plate 1). Therefore the power available for raising the dorsal valve must have been almost halved. On the other hand^the atrophy of one diductor would have made available an appreciably greater space within the very small body cavity. Apart from the muscles, the organ of greatest bulk within the coelom of a living brachiopod is the digestive tube, and, in particular, the stomach and its associated ramifying digestive diver- ticula ("liver"). The atrophy of the right diductor in Poikilosakos could have enabled the digestive tube to spread from its usual medial position into the whole of the right half of the coelom, and might have allowed it to increase in size with a corresponding increase in its metabolic efficiency. The evidence for adductor muscles in Poikilosakos is circumstantial. By homology with more normal brachiopods, a pair of small attachments, perhaps fused into a single scar, should be found in the midline on the ventral valve. If the right adductor had atrophid like the right diductor, a small scar might still be expected on the medial side of the left diductor attachment. No clear trace of any such scar can be seen, but the attachment might not be visible on such a thin shell. Most scars owe their visibility to a slightly lower rate of shell thickening of the site of the muscle attachment, relative to the rate on the surrounding surface. On a very thin shell there is little chance of such a differential becoming apparent. The diductor attachment is con- spicuous only because it has strongly raised edges. 274 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY By itself, the absence of any visible scar would suggest that the adductors of Poikilosakos were atrophied This is possible, but the existence of definite adductor scars on the dorsal valves of other oldhaminoids makes it more likely that adductors (or at least a left adductor) were present in Poikilosakos. Circumstantial evidence for the existence of an adductor can also be derived from the form of the hinge ridge. The chief resistance to the contraction of an adductor would have been the resistance of the water that had to be expelled from beneath the dorsal valve. This would have produced a backward thrust on the dorsal valve itself. If part of the adductor consisted of "quick" fibers for rapid closure of the shell, as in living articulates (Rudwick, 1961b), this thrust might have been quite considerable, relative to the size of the shell, but it would have been countered effectively by the apposition of the hinge ridge. The articulating anterior face of the hinge ridge is buttressed strongly from behind by the mass of the ridge itself, sloping away posteriorly. With this form, the hinge ridge would have been well adapted to counter the thrust of the dorsal valve during its closure. Lophophore and Feeding Mechanism Ever since oldhaminoids were first recognized as aber- rant brachiopods, it has been commonly agreed that their lophophore was probably schizolophous to ptycho- lophous. The pinnate outline of the oldhaminoid dor- sal valve has an obvious resemblance to the pinnate grooves on the inner surface of thecideoid dorsal valves. In living thecideoids (Thecidellina, Lacazella) these grooves are known to accommodate the brachial axes of a schizolophous to ptycholophous lophophore. The interior of the dorsal valve of Poikilosakos is unknown, but in other oldhaminoids there is a fine ridge running parallel to the pinnate outline of the valve, separated from the valve edge by a narrow groove. This submarginal ridge has been identified by Stehli (1956) and Sarycheva (1964) as the site of attachment of the brachial axes; but the marginal groove is a more probable site, for it is closely analo- gous to the axis-bearing grooves of thecideoids (Rud- wick, 1968a). But in any case the brachial axis of Poikilosakos may be reconstructed in a submarginal position, near the edge of the valve on its inner side (Figure 4B, C). Then, by homology with the ptycho- lophes of living brachiopods, the growing tips of the brachial axes would have been situated at the tip of the median indentation. Proximally, the brachial axes would have left the edge of the dorsal valve, probably at the base of the rectangular hinge projection, and would have joined each other at the mouth, in the midline on the anterior side of the coelom. The frontal side of the axes would have faced inward, and the abfrontal side outward, all around the lobate course of the brachia. The outline of the dorsal valve evidently increased in relative complexity with increasing absolute size during ontogeny; it is not practicable to express this quantitatively, but in qualitative terms the allometric relation between the overall shell size and the length of the dorsal valve edge (or of the shelf) is obvious enough (Figure 2). If the edge of the dorsal valve in fact represents the course of the brachial axes, this allometry becomes intelligible: in living brachiopods the relative complexity of the brachia increases in the same way during increase in absolute size. This is a necessary consequence of the dimensional relation be- tween the linear brachia and the metabolic require- ments of the body (cf. Rudwick, 1962). Although the morphology of oldhaminoids is aber- rant, it is methodologically important to assume, as a provisional working hypothesis, that their feeding mechanism was comparable to that of living brachio- pods. Only if their morphology proves difficult to ex- plain on this assumption should an aberrant feeding mechanism be postulated. On this provisional assumption, therefore, lateral cilia on the lophophoral filaments would have driven water from the frontal side to the abfrontal whenever the dorsal valve was raised and the filaments relaxed. Food particles colliding with the frontal surfaces of the filaments would have been enmeshed in mucus and con- veyed by frontal cilia to the food groove at the base of the filaments, and thence by other cilia along the groove to the mouth (cf. Rudwick, 1962, and references therein). The most effective orientation of the filaments would be for all of them to project obliquely outward and ventrally all around the edge of the dorsal valve. Then the tips of the filaments on either side of each indentation would meet and interlock, covering each indentation with a complete screen of filaments. A similar interlocking of filaments occurs in each in- dentation of the ptycholophe of the living Megathiris (Atkins, 1960). Thus, if the dorsal valve of Poikilosa- kos was habitually opened through some moderately peripheral zone hinge ridge dorsal valve flange valve edge peripheral mantle coelom mantle cavity filament (contracted) nhalant current jt^-.rn-v^trcrrccoxcozo.xoxco^o:.^ V sxhalant current jg^—^h^Y-^ 4#, #( ^r ::_:;--:,:.::: ::::.-::.:.::::. v:,::: :::: - : - ? yyy yy.yy—yy FIGURE 4.—Possible feeding mechanisms of Poikilosakos. A, Oblique longitudinal section through shell (along line A-B in Figure 1), as preserved, to illustrate morphology, B, C, Steady-flow feeding mechanism with brachial axis reconstructed submarginally on dorsal valve: B, dorsal valve open, filaments in feeding position; c, dorsal valve closed, filaments contracted. Orientation and action of filaments reconstructed by analogy with living megathirids. C-E, Rhythmic-flow feeding mechanism: c, dorsal valve closed, filaments contracted; D, diductor contracting, dorsal valve rising, water drawn into mantle cavity; E, adductor contracting, dorsal valve falling, water expelled from mantle cavity through screen of filaments. small angle, the pumping action of the lateral cilia would have drawn an inhalant water current inward under the edge of the dorsal valve and then dorsally through the screen of filaments. The water, thus fil- tered, would have been released as an exhalant cur- rent from the dorsal side of the dorsal valve (Figure 4B ). When feeding ceased temporarily, or the brachi- 372-386 O—71 19 opod was disturbed, all the filaments would have been quickly contracted inward, and the dorsal valve would have snapped shut (Figure 4c). In living articulate brachiopods this reaction is due to the contraction of "quick" muscle fibers on the frontal side of each fila- ment, followed immediately by the contraction of the similar fibers in the posterior adductor muscles. 276 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY This hypothetical feeding mechanism provides a reasonable functional explanation of the morphology of Poikilosakos, but it is not without difficulties. Most living brachiopods are able to operate similar steady- flow filtering systems with great effectiveness, but only because the filaments are so arranged that they divide the mantle cavity into separate inhalant and exhalant chambers, and also divide the gape between the valve edges into separate inhalant and exhalant apertures. By this means, recirculation of water around the tips of the filaments is reduced to a minimum, and the fil- tered exhalant water is thrown clear of the shell as a fairly strong exhalant jet from the exhalant aperture. But with the steady-flow system postulated for Poikilo- sakos, on the other hand, a high proportion of all the filaments would have projected freely, and recircula- tion around their tips might have detracted seriously from the effectiveness of the filtering; there would have been no clear channeling of exhalant water away from the shell. It is true that there is a similar lack of effectively separate apertures in the living Megathiris, but it is noteworthy that this genus and its relatives are invariably much smaller in size than even the sim- plest oldhaminoids such as Poikilosakos. The thecideoids probably operate a system similar to megathirids, although they are not closely related; but they too are always small in size, even in fossil representatives with ptycholophes almost as complex as those of later oldhaminoids (Rudwick, 1968a). This disparity in the absolute size of the lobate outlines of thecideoids and megathirids on the one hand and of oldhaminoids on the other is strong evidence for a functional discontinuity between them. There are other strong contrasts in the morphology of the two groups. The dorsal valves of thecideoids and megathirids are always moderately thick, and sometimes massive, and are clearly protective in function. In oldhaminoids, as already mentioned, the prime protective function seems to have been lost, and the dorsal valve is extremely thin and delicate; yet the muscle scars, where they are visible at all, show that the musculature was not correspondingly reduced. A similar combination of strong musculature with a very light dorsal valve exists in another aberrant group of late Paleozoic age, the richthofeniids; this has been interpreted as evidence for a rhythmic-flow feeding system comparable to that of living septibranch bi- valves, and the feasibility of this reconstruction has been confirmed by experiments with working models of richthofeniids (Rudwick, 1961a). Without pre- judging the question of the phyletic relation between richthofeniids and oldhaminoids, it is worth consider- ing whether a rhythmic-flow system might be applicable to Poikilosakos. Such a system need not have been exactly analogous to that of richthofeniids: indeed, there are enough differences in morphology to make a close similarity unlikely. In particular, there is no evidence for the form or even the existence of a lophophore in richtho- feniids; whereas in oldhaminoids, as already argued, the lobate dorsal valve probably reflects the course of a ptycholophous lophophore. But both groups have a notably thin and delicate dorsal valve, resting on a shelf within a larger ventral valve, and both groups combine aberrant hinge mechanism with a relatively strong musculature. This suggests that both groups might have derived the motive power for feeding from a rhythmic movement of the dorsal valve, even if their means of food collection were different. In living bra- chiopods the lophophore performs two distinct func- tions: it creates the movement of water for filtering, and it catches the particles out of the current so created. There is no reason a priori why Poikilosakos should not have lost the first of these functions while retaining the second. A possible rhythmic-flow system for Poikilosakos can be reconstructed as follows. The feeding cycle would have begun with the dorsal valve resting on the ven- tral (Figure 4c). A strong contraction of the diductor muscle would have raised the dorsal valve (Figure 4D) . Since the valve was light, most of the resistance would have come from the water displaced. Because the mo- ment of leverage was very small, the movement could have been fairly rapid, even if the diductor was a "slow" muscle (as in living brachiopods), provided only that it was powerful enough. With this upward movement, water would have swirled under the edge of the dorsal valve, past the contracted filaments. The mantle cavity would then have been entirely filled with unfiltered water, apart from the very small volume re- maining from the phase when the valve was closed (this would give functional advantage to the excep- tionally shallow mantle cavity). If the filaments then relaxed and projected ventrally, they could have sur- rounded this mass of unfiltered water with a complete screen. The contraction of the adductor muscle (or muscles) would then have forced this water outwards through the screen of filaments as the dorsal valve NUMBER 3 277 closed toward the ventral (Figure 4E). The resistance of this water would tend to thrust the dorsal valve backward; but as already mentioned, this thrust would have been held against the buttressing posterior ridge of the ventral valve. As the water was forced through the screen of filaments, suspended particles would have been caught on their frontal surfaces and transported to the mouth in the usual way. The most important feature of this hypothetical feeding cycle is that the relatively powerful though intermittent inhalant current could have swept rela- tively large suspended particles into the mantle cavity. Poikilosakos might therefore have been able to exploit a wider range of food material than more normal brachiopods living in the same environment. Even those particles (including small animals) that were too large to pass to the mouth in the normal way might have been utilized. They would have been trapped within the mantle cavity when the dorsal valve closed at the end of the cycle. The small and tightly sealed mantle cavity would have been highly effective as a site for external digestion of such material, if digestive juices were extruded from the gut through the mouth. The living septibranchs provide a very close analogue here (Yonge, 1928), and such a mechanism of external digestion would also help to explain the tight closure of the secondary commissure of Poikilosakos. It might also explain the exceptionally flat form of Poikilosakos, which was so closely adherent to its substrate that it must have been highly inconspicous. Small animals crawing across the ventral mantle edge (which in liv- ing brachiopods is highly sensitive) might have acti- vated the contraction of the diductor and have been swept inward into the mantle cavity. It is also possible that the peripheral zone of the ventral valve might have acted as a subsidiary site of particle collection, if it was covered in ciliated mucus-secreting mantle tissue like that of living brachiopods: in richthofeniids the homologous zone often developed fluted spines projecting into the path of the strongest currents (Rudwick, 1961a, Rudwick and Cowen, 1968). These suggestions are unavoidably hypothetical, but they are based on homologies with the mantle and muscles of living brachiopods and on analogies with the quasi-carnivorous habits of living septibranchs; and they do provide a coherent functional explanation of most of the peculiar morphological features of Poikilo- sakos. On the whole, a conventional steady-flow mechanism seems a less satisfactory explanation of the morphology than an aberrant rhythmic-flow system, notwithstanding a proper methodological bias in favor of the former. Evolution and Phylogeny The ancestor of Poikilosakos, and hence of the old- haminoids as a whole, must be sought among other Pennsylvanian brachiopods sharing at least some of the same distinctive characters. Among the characters which should be taken into account are (1) its cemented attachment and lack of pedicle, (2) its lack of tubular spines assisting attachment, (3) its loss of a true commissure, and extension of the ventral valve beyond the edge of the dorsal, (4) its peculiar hinge without true teeth, sockets, or denticles, and (5) its probably marginal or submarginal sessile lophophore. Other important characters, not positively demon- strable in Poikilosakos itself but which have rightly been stressed in earlier discussions of oldhaminoid af- finities, are (6) the pseudopunctate shell and (7) the small bilobed cardinal process. The cemented attachment of Poikilosakos was re- tained by all later oldhaminoids, at least for the earlier part of their ontogeny. In articulate brachiopods there are no well-authenticated instances of cemented attach- ment except among the Strophomenida (Cowen and Rudwick, 1967). (The pseudopunctate shell structure also points to the Strophomenida.) It has been argued that within that order cementation may have been evolved several times, but it seems unnecessary to postulate that the oldhaminoids had evolved it sepa- rately. Contemporary groups with cementation include the strophalosiaceans and the davidsoniaceans. The strophalosiaceans, however, are characterized by their tubular, external spines, some varieties of which (the "rhizoid" spines of Muir-Wood and Cooper, 1960) are used to supplement the cemented attachment of the ventral valve itself. Some strophalosiaceans resembled Poikilosakos in being adherent to the substrate by almost the whole external surface of the ventral valve. Yet, even in these, rhizoid spines were developed, spreading out across the surface of attachment. The total absence of such spines in Poikilosakos and almost all later oldhaminoids is, as Williams and Stehli stressed, suggestive evidence that they were not de- rived from strophalosiacean ancestors; and a david- soniacean origin seems more probable. 278 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY But Sarycheva (1964) has described the new genus Spinolyttonia, in which rhizoid spines are said to be developed, from the late Permian of Transcaucasia; and she argues that this discovery points to a strophal- osiacean affinity for all oldhaminoids. If this is so, it is curious that such an important and effective func- tional device was held in abeyance throughout the earlier history of the group and only developed in one species of late date. It might be argued that Spinolyttonia more probably represents an independent development of the ability of the mantle edge to se- crete tubular spines—if, indeed, the spines are genu- inely growing out of the oloUiarninoid, and are not merely productoid spines fortuitously attached to it. On the other hand, there are some features of Poikilosakos and later oldhaminoids that might favor a strophalosiacean affinity. Sarycheva has pointed to the resemblance between oldhaminoids and the richthofeniids, suggesting that both were derived from teguliferinid ancestors. This suggestion is sup- ported by the extension of the ventral valve beyond the dorsal, the lack of teeth and sockets and develop- ment of an aberrant "internal" articulation, and by the form of the cardinal process. In any case, the distinctive pinnate dorsal valve of oldhaminoids would be derived most simply from an ancestral form having a circular or bilobed dorsal valve similar to the early growth stages of Poikilosakos. A paedomorphic origin of the oldhaminoids may thus account for their cryptogenetic appearance in the fos- sil record. Functionally, the feeding mechanism postulated for Poikilosakos could have been derived by gradual stages from a normal mechanism. Rapid closure of the dorsal valve could have been purely a cleansing mechanism at first, as it is in living brachiopods. It could have taken over the motive function of the lateral cilia by very gradual stages; in an intermediate stage it would be possible for the water to be driven between the fila- ment partly by the cilia and partly by the closure of the shell. The existence of one specimen of Poikilosakos in which the right diductor attachment is as large and conspicuous as the left is of great interest. It suggests that in this population the atrophy of the right diductor muscle may have been controlled by a gene with simple all-or-none effect. There are two possible explanations for its occurrence as a rare variant: (a) it may be a surviving representative of a condition that was general in the population at an earlier time but which had become submerged by the selective advantage of the asymmetrical form, or (b) the asymmetrical condi- tion may have long been stabilized in the population; but the gene for symmetrical muscles may have been held latent in the gene pool, and may have been actual- ized phenotypically from time to time as a rare variant. This would be analogous to the occasional narwhal with paired "unicorn's horns" at the present day, as a sporadic variant from the normal form with only a single hypertrophied incisor. (I owe this analogy to Dr. K. A. Joysey.) The latter alternative seems the more probable, in view of the later history of the oldhaminoids. Later oldhaminoids, as Stehli has stressed, can be divided into those with symmetrical diductor scars and those in which one diductor is missing. In the asymmetrical group it is always the right diductor that has atrophied, and not the left; yet these two conditions would be PLATE 1.—Poikilosakos petaloides Watson, Pennsylvanian. All views X 3 except figure 8. Figure 1.—-Ventral valve (holotype), showing broad lobes and narrower indentations of flange, hinge ridge, and asym- metrical diductor sheath (BMNH BB.l 1266: "P. petaloides Watson, West bank of the Salt Creek at Graham, Young Co., Texas"; figured by Watson (1917, fig. 1). Photo courtesy E. F. Owen. Figure 2.—Ventral valve of small (Pjuvenile) specimen with roughly bilobed flange (BMNH BB.l 1267b; "P. peta- loides Watson, West bank of the Salt Creek at Graham, Young Co., Texas"). Figure 3.—Ventral valve of specimen with unusually ir- regular flange, attached to a crinoid stem (USNM 165840; "P. sp., Graham (Wayland), 1.2 m S. of Gunsight, Texas"). Figure 4.—Ventral valve of specimen showing edge of valve, probably broken (USNM 165839; "P. sp., Graham (Gonzales Creek Shale), roadside, 3.9 m. N.W. of Finis, Young Co. [actually, Jack Co.], Texas"). Figure 5.—Ventral valve of specimen widi paired diductor sheaths (USNM 148051b; "P. petaloides Watson, Gonzales Shale, 6 m. N.E. of Finis, Jack Co., Texas"). Figure 6.—Specimen widi bilobed dorsal valve, crushed but showing clear growth lines, preserved in position resting on the ventral valve flange. (USNM 165843; "P. sp., Graham (Wayland), 1.2 m. S. of Gunsight, Texas"). Figure 7.—Dorsal valve with external growth lines (USNM 137498; "P. aff. petaloides Watson, basal Plattsmouth lime- stone, 1 m. S. of Williamsburg, Kansas"). Figure 8.—Part of a ventral valve (X 6) showing detailed structure of hinge ridge, diductor sheath, flange, and pustular valve surface both within and outside the flange (USNM 165841; "P. sp., Graham (Wayland), 1.2 m. S. of Gunsight, Texas"). NUMBER 3 279 - ? w* 'J— -* : '- ? JFIv'i ks**^ -•'*F^ d- ~*>jr. J'^-J .<3S97 -'".fi't.- ' ssyii ;-, 280 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY PLATE 2 NUMBER 3 281 functionally identical, and probably also selectively equivalent. This is strong evidence that all species in the asymmetrical group are derived from Poikilosakos. But the occurrence of oldhaminoids with symmetrical diductors is more difficult to interpret. On the alterna- tive (a) above, it would be necessary to postulate that these symmetrical oldhaminoids were derived from a Poikilosakos population, at present undiscovered, in which the symmetrical condition had survived. Unless and until such a species is discovered in late Penn- sylvanian strata, it seems preferable to postulate that the symmetrical oldhaminoids evolved in the Permian from an ancestral population that was predominantly asymmetrical, under conditions which gave the rare symmetrical variants a selective advantage. If the asymmetrical condition originally had an ad- vantage in relation to the very shallow shell form of Poikilosakos, the evolution of populations in which the symmetrical condition predominated could have been connected with the abandonment of that shell form for other forms adapted to other environments: the diversity of later oldhaminoids strongly suggests a minor adaptive radiation of this kind. The evolution of the later oldhaminoids can be interpreted in terms of the further development of the postulated aberrant feeding mechanism. Although some later species are no larger than Poikilosakos, there was a general trend towards a larger overall size, cou- PLATE 2 —Poikilosakos petaloides Watson, Pennsylvanian. All views X 9. Figures 1, 2.—Enlargements of parts of external surface of a dorsal valve, to show growth lines indicating localized "bud- ding.'' (Specimen as in Plate 1: figure 7.) Figure 3.—Hinge region of a ventral valve, showing thick hinge ridge, slender accessory ridge, and diductor sheath (BMNH BB.l 1267a; "P. petaloides Watson, West bank of the Salt Creek at Graham, Young Co., Texas"). Figure 4.—Small (?juvenile) specimen showing dorsal valve with clear growth lines, preserved in position on ventral valve (USNM 165844; "P. sp., Graham (Gonzales Creek Shale), roadside, 3.9 m. N.W. of Finis, Young Co. [actually, Jack Co.], Texas"). Figure 5.—Hinge region of a ventral valve, showing acces- sory ridge, so-called dental areas, growth lines within diductor sheath, and pustular valve surface. (Specimen as in Plate 1: figure 3.) Figure 6.—Oblique view of hinge region of a ventral valve, showing growth lines on anterior face of hinge ridge, and oblique medial orientation of diductor sheath. (BMNH BB.l 1269a; "P. petaloides Watson, West bank of the Salt Creek at Graham, Young Co., Texas"; figured by Watson, 1917, fig. 4.) pled with an increase in the number of lobes on the dor- sal valve. The lobes and indentations retained the same standard dimensions throughout, but increased in number with increasing overall shell size. This implies a further relative increase in the complexity of the ptycholophe, similar to that which occurs in the on- togeny of Poikilosakos itself. As in Poikilosakos, it would be due to the necessary dimensional relation be- tween the collecting capacity of the linear brachia and the metabolic requirements of the whole body. The later evolution of the group is not, however, wholly explicable in terms of adaptation to enlarging body size. There is also evidence of increasingly effec- tive overall adaptation: the irregularly lobate outline of Poikilosakos was replaced by a more highly orga- nized system of regular lobes and indentations. These would have served to arrange the maximum length of filament-row, and hence the maximum filtering capacity, within a shell of any given size. The extinction of the oldhaminoids in late Permian time was relatively sudden in geological terms. Al- though it is part of a much larger problem, affecting many other groups of brachiopods and many other phyla, it is important that it should be explained and not merely explained away. A functional analysis of some of the later oldhaminoids does not suggest that they were so narrowly adapted as to be ecologically vulnerable (Rudwick and Cowen, 1968). On the con- trary, they seem to have acquired extremely "promis- ing" new features, both anatomically and physiologi- cally. These innovations might have been expected to open up new and less restricted possibilities in brachio- pod evolution, and to lead to some new phase of adap- tive radiation. Their failure to do so, or even to survive at all, must be given a convincing and not merely facile explanation. Literature Cited Atkins, D. 1960. The Ciliary Feeding Mechanism of the Megathyri- dae (Brachiopoda), and the Growth Stages of the Lophophore. Journal of the Marine Biological Asso- ciation of the United Kingdom, 39:459-479. Cowen, R., and M. J. S. Rudwick 1967. Bittnerula Hall and Clarke, and the Evolution of Cementation in the Brachiopoda. Geological Maga- zine, 104:155-159. Dunbar, C. O., and G. E. Condra 1932. Brachiopoda of the Pennsylvanian System in Ne- braska. Nebraska Geological Survey, series 2, 5:1- 377,44 plates, 25 figures. 282 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY Moore, R. C. (editor) 1965. Treatise on Intvertebrate Paleontology, Part H, Brachiopoda. 927 pages, 746 figures. New York City and Lawrence, Kansas: Geological Society of Amer- ica and University of Kansas Press. Muir-Wood, H. M., and G. A. Cooper 1960. Morphology, Classification and Life Habits of the Productoidea (Brachiopoda). Geological Society of America Memoir, 81:1-447, 135 plates. Rudwick, M. J. S. 1961a. The Feeding Mechanism of the Permian Brachio- pod Prorichthofenia. Palaeontology, 3:450—471. 1961b. "Quick" and "Catch" Adductor Muscles in Brachi- opods. Nature, 191:1021. 1962. Filter-Feeding Mechanisms in Some Brachiopods from New Zealand. Journal of the Linnean Society of London, Zoology, 44:592-615. 1968a. The Feeding Mechanisms and Affinities of the Triassic Brachiopods Thecospira Zugmayer and Bactrynium Emmrich. Palaeontology, 11(3): 329— 360, plates 65-68, 12 figures. 1968b. Some Analytic Methods in the Study of Ontogeny in Fossils with Accretionary Skeletons. In, Paleo- biological Aspects of Growth and Development, a Symposium. Paleontological Society Memoir, 2:35-49, 17 figures. Rudwick, M. J. S., and R. Cowen 1968. The Functional Morphology of Some Aberrant Strophomenide Brachiopods from the Permian of Sicily. Bollettino delta Societa Paleontologica Italiana, 6:113—176. Sarycheva, T. G. 1964. Oldhaminid Brachiopods from the Permian of Transcaucasia [English translation of title]. Paleontologicheskiy Zhurnal, 3:58—72, 2 plates, 2 figures. [Published in English, 1965, International Geology Review, 7(10) : 1826-1839, 2 plates, 2 figures.] Stehli, F. G. 1956. Notes on Oldhaminoid Brachiopods. Journal of Paleontology, 30:305-313. Watson, D. M. S. 1917. Poikilosakos, a Remarkable New Genus of Brachio- pods from the Upper Coal Measures of Texas. Geological Magazine, 54:212-219. Williams, A. 1953. The Morphology and Classification of the Old- haminoid Brachiopods. Washington Academy of Sciences Journal, 43:279-287. 1968. Evolution of the Shell Structure of Articulate Brachiopods. Palaeontological Association Special Papers in Palaeontology, 2:1-55, 24 plates, 27 figures. Yonge, C. M. 1928. The Structure and Function of the Organs of Feeding and Digestion in the Septibranchs Cuspidaria and Poromya. Philosophical Transac- tions of the Royal Society of London, series B, 216:221-263. PERMIAN Robert M. Finks Sponge Zonation in the West Texas Permian Type Section ABSTRACT Five sponge zones are recognized, corresponding to the Neal Ranch-Lenox Hills, Skinner Ranch, Cathe- dral Mountain-Road Canyon, Word-Cherry Canyon, and Gapitan Formations. The second and third zones can be subdivided into two subzones each: the Decie Ranch and post-Decie Ranch parts of the Skinner Ranch Formation, and the Cathedral Mountain and Road Canyon Formations. The most pronounced turn- over of species is at the base of the Cathedral Mountain Formation, or about 100 feet above the base of the Bone Spring Formation. It is characterized by radia- tion of the sphinctozoan lineages of Stylopegma and Guadalupia, as well as by extinction of Heliospongia and Stereodictyum. The base of the Skinner Ranch is also an important datum, marked by the apparent immigration of the earliest pharetronid calcisponges. Eight "acme-associations" are also recognized, each characterized by a unique combination of species that reach a simultaneous peak of abundance. The type section of the Permian, in western Texas, is unusually rich in fossil sponges. Such abundance raises the hope that they may prove useful in correlation, and the present article is an essay in that direction. The conclusions presented here are based on a mono- graphic study of the late Paleozoic sponges of Texas, now nearing completion, in which thousands of speci- mens from several hundred localities have been exam- ined. It is my hope that stratigraphers working in the Permian of the Southwest may be able to test the proposed zone fossils in a practical way. The richest sponge collections of the lower part of the section, from the base of the Permian to the mid- Robert M. Finks, Department of Geology, Queens College, Flushing, New York 11367. die of the Word Formation, are from the Glass Moun- tains. Most of the collections were silicified specimens removed en masse by means of acid from blocks of limestone. Supplementary collections of calcified speci- mens showed no striking differences in faunal com- position and the silicified faunas seem to be representa- tive. The upper part of the section, the Cherry Canyon and Capitan Formations, is known almost entirely from the Guadalupe Mountains, 150 miles to the northwest, from both silicified and calcified faunules. The nearby Sierra Diablo has provided a con- siderable series of faunules, mostly weathered material off the outcrop, from the Hueco, Bone Spring, and Vic- torio Peak Formations, which correspond to the Glass Mountains section. Some interesting differences in sponge faunas are apparent, between the Sierra Diablo and the Glass Mountains, due either to environmental differences, geographic separation, or both. Neverthe- less, the main outline of the Glass Mountains sponge zonation is recognizable, and sustains the hope that the zonal distribution will be recognizable over a wider area. Ideally, biostratigraphic zones should completely subdivide a span of time and have contiguous bound- aries. Even on a worldwide basis, however, there are likely to be gaps in the sedimentary and fossil record. Moreover, one is unlikely to find coterminous ranges of zonal guide species; they either overlap or are separated by gaps. The ideal is approached when one has a maximum of worldwide information, but this is probably never quite reached. Nevertheless, the pres- ent paper attempts a complete subdivision of the type section in terms of sponges. The species chosen as guides do seem to have coterminous ranges in many instances, though it remains for future work to test to what extent this is due to insufficient information. 285 286 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY to gaps in the section, or to truly simultaneous immigra- tion, evolution, and extinction of species. Without the vast treasures of fossils from the Texas Permian gathered over the last three decades by Dr. G. Arthur Cooper at the Smithsonian Institution, and gathered by Dr. Norman D. Newell at the Amer- ican Museum of Natural History, this attempt at a comprehensive sponge zonation would not have been possible. The very large samples from many localities are the primary data on which this study has been based, and the stratigraphic framework provided by the field studies of these men has been absolutely essential. It is a pleasure to acknowledge my indebted- ness to them also in a more personal way: to Dr. Newell who started me on the study of fossil sponges when I was his student, and to Dr. Cooper who has been a friend, guide and mentor over the years at the United States National Museum. I am pleased to be able to dedicate this paper to Dr. Cooper. The large collections of fossils from the Sierra Diablo made by S. J. Kfiz, then of Princeton University, and by P. B. King and J. B. Knight of the United States Geological Survey, have also formed a major part of the data presented here. Continued access to the col- lections of the United States National Museum, the United States Geological Survey, the American Museum of Natural History, the Yale Peabody Museum, and Princeton University, have been an essential element in the pursuit of this study. I can only thank my colleagues at these institution en masse for the many stimulating discussions, material aid, and bits of advice during the past 16 years. They are, of course, not to be blamed for my mistakes. Evolution and Zonation Species that form part of a single lineage and its branches, undergoing progressive evolutionary change, are at once the most reliable for stratigraphic zonation and the most difficult to use. They are reliable because a continuously evolving lineage makes it more likely, though it does not prove, that we are dealing with a true sequence in time and not with contemporaneous, environmentally controlled species that merely succeed one another locally by the chances of distribution. Lineages that show a progressive change at a more or less constant rate also favor the conclusion that no major time gaps are present in the record. For these reasons I have chosen, insofar as possible, zonal guide species from such continuously developing lineages. At the same time, such species are difficult to use because, being part of a gradually changing lineage, they are difficult to separate from one another. Boundaries between species are therefore somewhat ar- bitrary, and the sharpness of resolution of the corre- sponding zonal boundary is thereby diminished. This is somewhat mitigated by using concurrent overlap- ping ranges of species from different lineages. For a given zonal boundary a second lineage may show a sharper break between species or, better still, a species may begin or end at the boundary by immigration or extinction, thereby avoiding problems of transitional forms. The existence of a lithologic break can also be used to provide a precise point of division between zonal species in a lineage, though of course it must coincide approximately with a recognizable morpho- logic change in the fossils themselves. Needless to say, the species must be initially sorted out and delimited without reference to their stratigraphic position if they are to be at all useful in correlation. The known stratigraphic sequence, however, must be used subse- quently to choose which morphological differences are significant in terms of time. Such methods have been used here. The principal evolutionary lineages among the Texas Permian sponges are two. One is a line of develop- ment that seems to stem from the Pennsylvanian sphinctozoan calcisponge Maeandrostia, which was al- ready present in the Texas area. Two species of the genus Stylopegma in the Neal Ranch Formation seem to have arisen from it. The more advanced of the two, S. stenaulos, is found in the reefy facies, the other in the shales. Stylopegma stenaulos seems to have given rise to the Skinner Ranch species, S. dulcis by a con- tinuation of the same trends that brought it out of Maeandrostia. The Cathedral Mountain and later species, S. wordensis, continues the same trend still fur- ther. Two other Cathedral Mountain species, S. iso- pora and Polysiphonella flabellata, represent quite dif- ferent lines of development, but are clearly related morphologically to the S. dulcis type. Partly contempo- rary with S. dulcis, but appearing slightly later in the upper Skinner Ranch, is the rather different S. an- nulata. It could have arisen from S. dulcis by a more NUMBER 3 287 abrupt transition, in a direction parallel to that of S. wordensis. Early Word (China Tank) representatives of S. annulata show a greater resemblance to the later S. getawayensis of the Cherry Canyon, and could have developed into it, though one might also regard S. getawayensis as a further development of S. wordensis. The major adaptive radiation of the Stylopegma line- age appears to have taken place around the Skinner Ranch-Cathedral Mountain boundary. This same period seems to be that at which the ini- tial radiation of the second lineage took place, namely, that of the sphinctozoan calcisponge Guadalupia. Un- like the Maeandrostia—Stylopegma lineage, that of Guadalupia has no known Pennsylvanian history in the area, unless it be derived from the related, but rather different, Pennsylvanian genus Cystauletes. Guadalupia first appears as the species G. williamsi in the middle of the Skinner Ranch Formation, long after the last known Cystauletes, and it seems more likely to have evolved elsewhere and migrated into the Texas region during Skinner Ranch time. By late Skinner Ranch time a second species, G. auricula, has appeared, which is closely related to a complex of species in the Ca- thedral Mountain and Road Canyon Formations, and in the China Tank Member of the Word Formation; namely, G. lepta, G. cupulosa, G. ramescens and G. vasa. These forms, including G. auricula, intergrade with one another and they may well be environmen- tally controlled growth forms of one species, but they show a certain stratigraphic separation, and for this reason I have given them separate names. They also show a progressive trend in the sequence G. auricula through G. cupulosa to G. vasa, which is directly con- tinued into the Cherry Canyon and Capitan species G. cylindrica. Meanwhile, at the base of the Cathedral Mountain Formation two other species of Guadalupia appear, G. zitteliana and G. "Polyphymaspongia" ex- planata, which like G. auricula may have arisen from G. williamsi. Guadalupia zitteliana may have given rise in turn to the Word species G. cystauletoides and probably did give rise to the Word species Cystothal- amia nodulifera. The last is almost certainly ancestral to the Capitan C. megacysta. Another Word species, G. microcamera, was probably derived from G. cupu- losa (transitional forms occur in the Road Canyon Formation), and in turn leads to the Capitan G. favosa. The Guadalupia family tree is more continuously branching than that of Stylopegma, but both show a marked radiation at the beginning of Cathedral Moun- tain time, and both show progressive trends that reas- sure one concerning temporal sequences. Systematics of Zonation The boundaries between the zones have been drawn at horizons of major turnover in the sponge faunas. They have also been chosen, insofar as possible, so that the overlapping ranges of at least two common and easily recognizable species will identify the zone. It is convenient to arrange the zones in a hierarchi- cal scheme according to the magnitude of the faunal breaks between them and according to the apparent breadth of distribution of the zonal guide fossils. One may assess the magnitude of the faunal breaks by com- paring what may be called the turnover index of the useful zonal species, that is, the sum of species dis- appearing and appearing at that boundary. One may also examine turnover indices for genera (first and last appearance of genera at the zonal boundary) as well as for families and for the Class or Order level. The indices for the formational boundaries of the type section are given in Table 1. The principal breaks (apart from the base of the Neal Ranch) are at the top of the Neal Ranch and the base of the Skinner Ranch, at the base of the Cathedral Mountain, at the base of the Word (as revised by Cooper and Grant, 1966), and at the base of the Capitan. They are marked not only by a high turnover index for species, but also for the higher categories. These breaks serve to delimit five biostrati- graphic zones of sponges. If one judges them further according to the breadth of geographic distribution of the zonal guides, and their ease of recognition, one of these breaks is some- what more prominent than the rest; namely, that at the base of the Cathedral Mountain Formation, or the top of the Skinner Ranch Formation. At his horizon, the widespread, abundant, large, and easily recogniz- able genera Heliospongia and Stereodictyum disappear from West Texas, and so far as is now known, from the earth. This horizon is recognizable not only in the Glass Mountains but also in the Sierra Diablo- Guadalupe Mountains area, where the equivalent horizon lies about 100 feet above the base of the Bone Spring Limestone. This horizon may serve to delimit two megazones. 288 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY The remaining breaks can be arranged so as to de- fine subzones within the major zones, and what may be called microzones within some of the subzones. The two megazones can be defined on the range of a single species each within the Glass Mountains. The zones, subzones, and microzones are best identified on the basis of numerous concurrent, overlapping ranges (shown on Figures 1 and 2) and may take their names, but not their ultimate definition, from two of the species whose concurrent overlapping ranges help to fix the zone. The complete classification of the zones is as follows: Zones Top of Section II. Stylopegma isopora Megazone 5. megacysta-cylindrica Zone {Cystothalamia megacysta and Guadalupia cylindrica) 4. microcamera-zitteliana Zone {Guadalupia microcamera and G. zitteliana) 4b'. nodulifera-typ kale Microzone {Cystothalamia nodulifera and Stromatidium typicale) 4a'. vasa-dunbari Microzone {Guadalupia vasa and Girtyocoelia dunbari) 3. flabellata-robusla Zone {Polysiphonella flabellata and Virgola robust a) 3b. ramescens-cupulosa Subzone {Guadalupia ramescens and G. cupulosa) 3a. auricula-lepta Subzone {Guadalupia auricula and G. lepta) I. Heliospongia vokesi Megazone 2. vokesi-defuncta Zone {Heliospongia vokesi and Dejordia defuncta) 2b. williamsi-vokesi Subzone {Guadalupia williamsi and Heliospongia vokesi) 2b'. vokesi-auricula Microzone {Heliospongia vokesi and Guadalupia auricula) 2a. megalochetus-agaricus Subzone {Haplistion megalochetus and Catenispongia agaricus) 1. vokesi-prosseri Zone {Heliospongia vokesi and Amblysiphonella prosseri) Base of Section Formations of type section Capitan Word and Cherry Canyon Cherry Canyon China Tank Member of Word Cadiedral Mountain and Road Canyon Road Canyon Cathedral Mountain Skinner Ranch post-Decie Ranch Member base of Taylor Ranch Member of Hess to top of Skinner Ranch Decie Ranch Member of Skinner Ranch Neal Ranch (and Lenox Hills?) The megazones, zones, and subzones are intended to completely partition the sequence and to have con- tiguous boundaries. The microzones lie wholly within zones or subzones, and their boundaries are not neces- sarily contiguous with other units. The two micro- zones within the microcamera-zitteliana Zone may ultimately prove to be regular subzones with contiguous boundaries. NUMBER 3 289 TABLE 1.—Indices for the formational boundaries of the type section as recognized by Cooper and Grant (1966) and based on the ranges shown in Figures 1 and 2- Turnover Indices Formational Boundary Species Genera Families Classes or Orders (1) Base of Neal Ranch (2) (a) Top of Neal Ranch (b) Base of Skinner Ranch 12 7 4 1 3 2 0 1* 1 0 1* 1 Combined 11 5 2* 2* (3) Top of Decie Ranch (4) Base of Taylor Ranch (5) Base of Cathedral Mountain 6 2 12 2 0 4 1 0 3 0 0 0 (6) Base of Road Canyon (7) Base of Word (China Tank) (8) Top of China Tank (9) Base of Willis Ranch 5 10* 3* 1* 0 2* 1* 1* 0 1* 1* 0 0 0 0 0 (10) Base of Cherry Canyon (11) Base of Capitan 4 10 1 1 1 0 0 0 *Includes taxon whose range may extend farther elsewhere. The vokesi-prosseri Zone This is the fauna of the Neal Ranch Formation. Be- sides Amblysiphonella prosseri, which ranges up from the Virgilian, and Heliospongia vokesi, which continues into higher beds, the zone contains a number of other species which are restricted to this horizon, mainly oc- curring in the reefy facies, and mainly constituting a radiation of the family Fissispongiidae and to a lesser extent of the Maeandrostiidae. Reefy faunas of sponges are not well known from the Virgil. It is possible that some of these species were already present in Virgilian reefs; it is reasonably certain, however, that they are not present in post-Wolfcampian reefs. The base of the Neal Ranch Formation is marked by the disappearance of the common Pennsylvanian genera Coelocladia, Girtycoelia (but not Girtyocoe- lia!), and Maeandrostia. The first appears to have died without issue. Girtycoelia seems to have died out locally, but Ott (1967) considers the Triassic Colospongia Laube to be a senior synonym of Girtycoelia. They are indeed very similar, and probably related, but in view of the time difference, of the absence of Girtycoelia from the Texas Permian, and of certain differences in form, the possibility of homeomorphy is suggested, and I believe the two genera should not be combined at the present time. The Permian Steinmannia Waagen and Wentzel, from the Salt Range of Pakistan, is also considered by Ott to be a synonym of Girtycoelia, but it is at present very poorly known. The third Pennsylvanian genus, Maeandrostia, ap- pears to have evolved into the genus Stylopegma, of which two species characterize the vokesi-prosseri Zone. One of these, S. turbinata, occurring in the shaly, or nonreefy, beds is very close indeed to Maeandrostia, and could be included in that genus if it was inter- preted broadly. Wewokella (Talpaspongia) also dis- appears at the base of the Neal Ranch in the Glass Mountains, but is extremely common in the southern calcitic facies' (Finks, in King, 1965) of the Hueco Formation of the Sierra Diablo and Hueco Mountains, which contains Heliospongia vokesi and appears to lie within the vokesi-prosseri Zone. The distribution of Wewokella appears to be environmentally controlled. Stereodictyum orthoplectum, another common sponge of the vokesi-prosseri Zone, extends into the overlying vokesi-defuncta Zone, and has also been found in reefy facies of the Virgilian Holder Formation of New Mexico (J. M. Parks, J. L. Wilson, and D. F. Toomey, personal communications). It is not clear whether the Lenox Hills Formation, which directly overlies the Neal Ranch in the Glass Mountains, falls within the vokesi-prosseri Zone be- cause the only sponges so far found in it are H. vokesi and Fissispongia species, neither of them diagnostic. The diagnostic sponge association of the vokesi- prosseri Zone is as follows (Page 292) : 290 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY one JK— 7K 2i ' TK T 7T—* A <7>* * #— i i i * | 1 ' i nca z O 1 i 1 ! ? ensis «= sopora Megazone megacysta-cylin CAPITAN i "1 I: 1 OI G. favosa alupia cylindHca Amblyslphone guadaiupen Stromatldium typic i =?' I a.1 'St I 1 Si &: o ' U | 3.1 c i te / , A| !^S * * ^ 3 2* Styl ope gm a i zitte lian a zon e nod ulif er typi cale 'OR D < : 0 Jl| C ' -16 o ig V a> .el*: Is ... t- ? i ? i i i I 1 X 1 1 1 c r o 1 > p o r 5 t ! I 1 TJ A A S fi" A X I ii E CHINA TANK jO< > o ° .?+ » E x > \ „ /?V| flabellate- auricula-lep subzone romescens - cupulosa sub CATHEDRAL MTN. \ Ivsiphonella f ongia conus Q, S' o a. 3 O 3i » f o a. 3 D 3 r i williamsi , G. cupulos £ G. lepto nocoelia verr < • _ Virgolo rob -1 :inctuta nbari xl A at fi- o A?5> (S al u pi j c- «.. o 1 3 — CO \ TAYLOR RANCH c i s s s i i ii i II c < . o Til e t a z o n e w il li a m v o k e s i R A N C H Pileofites "° baccatus w j ^ G u a d < i a d e f g a r i c u s a t i s p o n g y o c o e li a / o I 4 k \ / V u CO vokesi -def un egalochetus- ancus subz. I s 3S O o i i o SKINNER DECIE RANCH Cv —• actum D e f o • a t e n i s p o n g i a Str Gir •" u u CP o \/ $& E ? a. 3 r E & o c • a a. < CO s. £ i 3 - o ys ip ho ne li vokesi- Stylop Stylo Rssi Fissis < CoalocI g i 3 r ' t \u ^r_j / i C V N t \t v ! o 1 A VIRGILIAN i ? > a <-. 5 - I M a e a n drosti < cladic * Girty - coelic FIGURE 1.—Distribution of sponge zonal-guide taxa in the standard section of the Permian in the Glass Mountains, Texas, showing sponge zones and key rock units. Symbols: < first appearance of a genus; > last appearance of a genus; 0» nrst or l**t appearance of a family; ?, limit of range not clear-cut because of intergradation with related species; X, range may extend farther else- where; , known range outside this area, absent here due to lack of collecting or facies difference; #, horizon of unusual abundance (acme). NUMBER 3 291 —>\— w A f A 7v /^ f\ /1\1 JO o CO a « .2 c CO o & 0 TJ Q. a • oner CO o o 1 iso • E c •3 C la n a ipt un CAPITAN BELL CANYON 3 capit typic D E CO CO pegm c lam ia ilupia zlttel indrica honella pensis ola n« i Corynelli omotidium Styloi Stylo ^Cystot ha y Guadalupia uadolupla cyl ^ Amblysip guadalu v> DA.= " A <>• O ^. GOAT CHERRY e g m yens •— s ° •ll 11 lo n al CO c O CO SEEP CANYON S t y l o p etawa llfera rocan G. exp Virgo 'o 1 a o V ! VCT - 1 — n p o u -* nf \| A i_ CO CV- T S o sT 3 cl <-!•£ Q. s ; O SAN BRUSHY CANYON r i k ! \ 1 vS ° £ 2 § 1* Guadali 6y/ i nispongia agar ANDRES ] a flal ? b l A- n u l O | A CO c ? o "5 XZ V o ! ° «>• I °M VICTORIO E 9-1 I e 11 PEAK 3 9 d 0 Polyi A i a d e m idrini 3 C o Styl a Q. JO iford :tino o CO o BONE SPRING o 1 Ui • S < E 1 o n gi a i. v ^i/c- v. \y \l/ \!/ V M/ \/ ' < * p * S J ?r ^ cv. iv a. CO CO 1 o | S S j l c i ii ?c LOWER 100' OF u. ip Olil D \^ ia cine dunbe BONE SPRING *5 ctum Pile « box 8 olites- g » £ xatus Strotispong i GIrtyocoalia 1 > / o th o pl e x: -*> >/ Mfrr 1 ^ «y. o o c o E a. 3 CO >N l i o T3 \ HUECO CO X 5 Stereodl Wewokella —^__^__—__^_ FIGURE 2.—Distribution of sponge zonal-guide taxa in die Permian section of die Sierra Diablo- Guadalupe Mountains area, including the standard section of the Guadalupian. Symbols as in Figure 1. 372-386 0—71- -20 292 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Calcarea Sphinctozoa Fissispongiidae Fissispongia tuba Finks Fissispongia species cf. F. conus Finks Aulocyclus Finks Aulospongia Finks Maeandrostiidae Stylopegma stenaulos Finks S. turbinata Finks Sebargasiidae Amblysiphonella prosseri Clarke Heteractinida Wewokellidae Wewokella {Talpaspongia) clavata (King) Demospongea Heliospongiidae Heliospongia vokesi King Coelocladiella philoconcha Finks Haplistiidae Haplistion megalochetus Finks Hexactinellida Stereodictyidae Stereodictyum orthoplectum Finks Base of the vokesi-prosseri Zone: Last appearances; By local extinction: Coelocladia Girtycoelia By evolution: Maeandrostia (into Stylopegma) First appearances: By immigration: Autocyclus (or by evolution from Fissispongia) Haplistion megalochetus (or by evolution from other Haplistion) By evolution: Stylopegma stenaulos (from Maeandrostia) S. turbinata (from Maeandrostia) Heliosponga vokesi (from H. ramosa) Exclusive to zone: Stylopegma stenaulos S. turbinata Aulospongia Coelocladiella philoconcha Species having acmes within zone: Stylopegma stenaulos Stereodictyum orthoplectum Recognition of the base of the vokesi-prosseri Zone may be rendered somewhat difficult by the fact that the boundary between H. vokesi and the earlier H. ramosa is somewhat arbitary, and the two may occur together (Finks, 1960, p. 45). The presence of vokesi, rather than the absence of ramosa, should be considered diagnostic. The whole association of species is the best guide to this zone. The top is best placed at the incur- sion of the characteristic species of the overlying vokesi- defuncta Zone. In the present state of our knowledge this occurs at the base of the Skinner Ranch Formation in the Glass Mountains, and at the base of the Bone Spring in the Sierra Diablo. The entire Hueco of the Sierra Diablo will thus fall into this zone, as will the Neal Ranch and Lenox Hills in the Glass Mountains. It should be noted that the exclusion of both the Hueco and Lenox Hills Formations from the higher vokesi-defuncta Zone is based only on negative evidence. The vokesi-defuncta Zone This zone coincides with the Skinner Ranch Formation of the Glass Mountains and with approximately the lower 100 feet of the Bone Spring Formation in the Sierra Diablo. It is defined by the co-occurrence of Heliospongia vokesi or Stereodictyum orthoplectum, which become extinct at the top, with Catenispongia agaricus, Stratispongia cinctuta or Defordia defuncta, which start at the base. Catenispongia and Strati- spongia appear to represent the first pharetronid cal- cisponges in the fossil record (they are the "Pharetronid Genera A and B" of Finks, 1960). The top of this zone is the last appearance of the families Heliospon- giidae and Stereodictyidae as well- as the first appear- ance of many new forms of the succeeding zone. The upper and lower boundaries of the vokesi-defuncta Zone are the most clear-cut breaks in the sequence of sponge faunas of the Texas Permian. In the Glass Mountains a lower subzone (megalo- chetus-agaricus) can be recognized, bearing an attenu- ated fauna (Aulocyclus, Haplistion megalochetus, Fissispongia tuba) from the preceding vokesi-prosseri Zone, along with the new elements. This lower subzone is coincident with the Decie Ranch Member of the Skinner Ranch Formation. It is marked by a strong abundance of Defordia defuncta and Heliospongia vokesi. Immediately above it, in the middle of the Skinner Ranch Formation, die first Guadalupia appears, and defines the base of die upper, or williamsi-vokesi Sub- zone. Towards the top of this subzone, the Taylor Ranch Member of die Hess Formation has a very characteristic faunal association, described below under the section on "Acme-associations," and seems to be marked by the first appearances of Guadalupia auricula and Stylopegma annulata. Because of these first appear- ances it may have a time significance, and has been NUMBER 3 293 named the vokesi-auricula Microzone. On the other hand, this may be merely a richer fauna of the vokesi- williamsi Subzone. In the Sierra Diablo area, the known sponge fauna is much less varied and the subzones and microzone cannot be recognized. Nevertheless, the lower part of die Bone Spring contains Heliospongia vokesi, Stereo- dictyum orthoplectum, and Stratispongia cinctuta, which together are sufficient to diagnose the vokesi- defuncta Zone. The correlation is strengtiiened by the occurrence in both die mid-Skinner Ranch and the lower Bone Spring of Stylopegma dulcis, which is intermediate in an evolutionary progression between the Neal Ranch S. stenaulos and the Cathedral Moun- tain and die later S. wordensis. Also the peculiar hex- actinellid Pileolites baccatus is known, and ratiier abundantly, from but two localities; one in the middle of the Skinner Ranch, die other at the base of the Bone Spring. Wewokella (Talpaspongia) clavata is common in the Clyde Formation of central Texas, which has been correlated approximately with die Skinner Ranch hori- zon on the basis of die presence in it of Pseudo- schwagerina crassitectoria (Dunbar and others, 1960, p. 1789). Wewokella, absent from the Glass Mountains Permian entirely, is not present in the Sierra Diablo above the Hueco Formation. It may be tiiat its absence from the Sierra Diablo at the Skinner Ranch horizon is due to environmental differences, or, perhaps, the Clyde is actually older than the Skinner Ranch, or perhaps again, the Clyde correlates with only the lower part of the Skinner Ranch of the Glass Moun- tains, and that in turn with the post-Hueco, pre-Bone Spring erosion interval in the Sierra Diablo, or even with the upper Hueco. In connection with the last possibility, a Defordia species in the Upper Hueco suggests Skinner Ranch affinities. In any case, Wewo- kella is not presently known to be a part of the fauna of the vokesi-defuncta Zone of western Texas. The diagnostic faunal association of the vokesi- defuncta Zone is as follows: Calcarea Sphinctozoa Fissispongiidae Fissispongia conus Finks F. tuba Finks (lower subzone only) Aulocyclus Finks (lower subzone only) Maeandrostiidae Stylopegma dulcis (King) S. annulata Finks (uppermost microzone only) Celyphiidae Girtyocoelia dunbari King Guadalupiidae Guadalupia williamsi King (upper subzone only) G. auricula Finks (uppermost microzone only) Pharetronida Stellispongiidae Catenispongia agaricus Finks Stratispongia cinctuta Finks Demospongea Helispongiidae Heliospongia vokesi King Chiastoclonellidae Defordia defuncta King Hexactinellida Stereodictyidae Stereodictyum orthoplectum Finks Pileolitidae Pileolites baccatus Finks Base of vokesi-defuncta Zone: Last appearances (top of Neal Ranch): By local extinction: Aulospongia Coelocladiella Amblysiphonella prosseri Wewokella (Talpaspongia) clavata (in Hueco only; persists in Clyde Forma- tion of central Texas) Stylopegma turbinata By evolution: Stylopegma stenaulos (into S. dulcis and possibly S. annulata) First appearances (base of Skinner Ranch): By immigration: Catenispongia Stratispongia Defordia defuncta (or by evo- lution) Base of williamsi-vokesi Subzone (upper subzone): Last appearances: By local extinction: Aulocyclus Haplistion megalochetus Fissispongia tuba First appearances: By immigration: Guadalupia williamsi G. auricula (base of upper- most microzone) (or by evolution from G. willi- amsi) By evolution: Stylopegma annulata {base of uppermost microzone) (from S. stenaulos?) S. dulcis (possibly earlier) (from S. stenaulos) Stylopegma dulcis ^Pileolites baccatus Exclusive to vokesi-defuncta Zone: 294 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Acmes within vokesi-defuncta Zone: Heliospongia vokesi (lower subzone and uppermost microzone) Defordia defuncta (lower sub- zone) Pileolites baccatus Stereodictyum orthoplectum (uppermost microzone) Fissispongia conus (upper- most microzone) Girtyocoelia dunbari (upper- most microzone) Catenispongia agaricus (up- permost microzone) The flabellata-robusta Zone This zone includes the Cathedral Mountain and Road Canyon formations of the Glass Mountains and the upper part of the Bone Spring and Victorio Peak Formations of the Sierra Diablo. Its base is marked by a great expansion of the Guadalupia and Stylopegma lineages, giving rise to four new species of Guadalupia and two new species of Stylopegma, as well as the related genus Polysiphonella. It marks also the first appearance of the genus Virgola. Most of the new forms persist into higher zones, but progressive changes in a group of species seemingly descended from Guada- lupia auricula give some assurance of temporal con- tinuity and progression. The extinction of the formerly abundant Heliospon- gia vokesi and Stereodictyum orthoplectum at the base of this zone, together with the appearance of the new forms cited above, makes this boundary an important punctuation in the sequence of sponge faunas. For this reason, it has been selected as the break between two megazones. Many forms persist from earlier zones so that the total sponge fauna of the flabellata-robusta Zone is perhaps the richest in the entire section of the Texas Permian. This is particularly true when one includes the many hexactinellid and lithistid species that have been omitted from the lists of zonal guides because their ranges are either too great or too poorly known. The top of this zone is less well defined than the base. It is probably best defined by the appearance of the diagnostic species of the succeeding zone: Guada- lupia microcamera, G. vasa, and G. cystauletoides, al- though a transitional form to the first species is present in the Road Canyon Formation (the upper subzone) and may give trouble. In the Glass Mountains the genera Fissispongia and Polysiphonella are not known above this zone, but in the Sierra Diablo—Guadalupe Mountains area they are associated with species of the succeeding zone. Several species appear to be confined to this zone; namely, Guadalupia cupulosa, G. lepta, G. ramescens, Actinocoelia verrucosa, Defordia densa, and Virgola robusta. The Guadalupia species appear to be the most reliable guides because they form part of a general evolutionary progression extending from the preceding zone to the following one. Within the zone, Guadalupia lepta appears to be confined to the lower (Cathedral Mountain) part and G. ramescens to the upper (Road Canyon) part, permitting the recognition of two sub- zones in the Glass Mountains. The disappearance of Guadalupia auricula, Defordia defuncta, and Actino- coelia verrucosa at the boundary between the two subzones tends to reinforce it. The two subzones have distinctive associations of species that have acmes within the subzone. The lower subzone (auricula-lepta) is characterized by the abun- dant occurrence of Guadalupia lepta, G. explanata, Polysiphonella flabellata, and Actinocoelia verrucosa. The upper subzone (ramescens-cupulosa) is character- ized by the acmes of Stylopegma wordensis, S. an- nulata, Polysipnonella flabellata, Fissispongia conus, Guadalupia cupulosa, G. ramescens, Virgola robusta, Catenispongia agaricus, Stratispongia cinctuta, and Girtyocoelia dunbari. The two subzones cannot be recognized in the Sierra Diablo region. The distinctive lower subzonal species Defordia defuncta and Actinocoelia verrucosa are re- placed by Defordia densa and Actinocoelia maean- drina. They occur with Guadalupia lepta, but die upper subzonal guide, G. ramescens, is unknown. It may be that tthe entire upper Bone Spring and Victorio Peak are equivalent only to the lower subzone (Cathe- dral Mountain). The lower part of the San Andres has provided the holotype of Actinocoelia maeandrina, which ties it to the Victorio Peak. It does not contain Guadalupia ramescens, though nothing in the lower San Andres sponge fauna would rule out an upper subzone age. The presence of Cystothalamia nodulifera suggests an even later age, for this is a species of the next higher zone, but at the same time the presence of Fissispongia, which in the Glass Mountains lies wholly beneath Cystothalamia, issues a warning that the Glass Mountains picture, at least as presently known, is not always applicable elsewhere. NUMBER 3 295 Actinocoelia maeandrina is one of the most widely distributed of the Permian sponge species, and is a common sponge in the Kaibab Formation of Arizona, and its equavalents in Nevada and Utah. Since pub- lication of an article calling attention to its wide dis- tribution (Finks, Yochelson, and Sheldon, 1961), it has been reported in Kaibab equivalents from other locali- ties in southern Arizona and in Utah (various personal communications). It also is known from the top of the Franson Member of the Park City Formation in western Wyoming, possibly out of line with its strati- graphic position in western Texas. An occurrence re- ported from die Goat Seep Formation of the Sierra Diablo (Finks, in P. B. King and others, 1965, p. 76) may be a different species of Actinocoelia having finer trabeculae. The diagnostic sponge association of the flabellata- robusta Zone is as follows: Calcarea Sphinctozoa Fissispongiidae Fissispongia conus Finks Maeandrostiidae Stylopegma isopora Finks S. wordensis Finks S. annulata Finks Polysiphonella flabellata Finks Guadalupiidae Guadalupia williamsi King G. auricula Finks (lower subzone only) G. cupulosa Finks G. lepta Finks (lower subzone only) G. ramescens Finks (upper subzone only) G. zitteliana Girty G. explanata (King) Celyphiidae Girtyocoelia dunbari King Pharetronida Stellispongiidae Catenispongia agaricus Finks Stratispongia cinctuta Finks Virgola robusta Finks Demospongea Chiastoclonellidae Defordia defuncta King (lower subzone only) D. densa Finks Actinocoelia verrucosa Finks (lower subzone only) A. maeandrina Finks Base of the flabellata-robusta Zone: Last appearances: By local extinction: Heliospongia Stereodictyim By evolution: Stylopegma dulcis (into S. wordensis and S. isopora) First appearances: By immigration: Virgola robusta By evolution: Stylopegma isopora (from S. dulcis?) S. wordensis (from S. dulcis) Polysiphonella flabellata (from S. dulcis or by immigration) Guadalupia zitteliana (from G. wil- liamsi or by immigration) Guadalupia (Polyphymaspongia) explan- ata (from G. zitteliana or by immi- gration) G. cupulosa (from 5". auricula) G. lepta (from G. auricula) ?Actinocoelia maeandrina (related form known from preceding zone) ?A. verrucosa (from A. maeandrina?) ?Defordia densa (from D. defuncta?) Base of the ramescens-cupulosa Subzone: Last appearances: By local extinction: Defordia defuncta Guadalupia auricula By evolution: G. lepta (into G. ramescens?) First appearances: By evolution: Guadalupia ramescens (from G. cupulosa or G. lepta) Species confined to the flabellata-robusta Zone: Guadalupia lepta (lower subzone) G. ramescens (upper subzone) G. cupulosa Virgola robusta ?Actinocoelia maeandrina (may extend into zone above) A. verrucosa Defordia densa ?Polysiphonella flabellata (in Glass Mountains) Acmes within flabellata- robusta Zone: Stylopegma wordensis (upper subzone) S. annulata (upper subzone) Polysiphonella flabellata Fissispongia conus (upper subzone) Guadalupia explanata (lower subzone) G. lepta (lower subzone) G. cupulosa (upper subzone) G. ramescens (upper subzone) Actinocoelia verrucosa A. maeandrina Defordia densa Virgola robusta (upper subzone) Catenispongia agaricus (upper subzone) Stratispongia cinctuta (upper subzone) Girtyocoelia dunbari (upper subzone) The microcamera-zitteliana Zone This zone includes the sponge faunas of Word Lime- stones 2, 3, and 4 in the Glass Mountains and of the 296 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Getaway and Goat Seep Limestones in the Guadalupe Mountains. It is coextensive with the range of Guada- lupia microcamera, which is an easily recognizable species and has been found in both the Glass Mountains and the Guadalupe Mountains. Transitional forms be- tween it and G. cupulosa occur in the Road Canyon Formation, however, and may give a bit of trouble. Guadalupia cystauletoides occupies all of this zone in the Glass Mountains, but has not been found outside that area. The distinctive species Cystothalamia nod- ulifera is abundant in the Getaway Limestone of the Guadalupe Mountains, and so far has not been found below the Willis Ranch Member of the Word (Word 3) in the Glass Mountains. There is a possibility, how- ever, that its range may extend to the base of this zone, or even lower, because it occurs in the lower part of the San Andres Formation along with Fissispongia conus, Stylopegma annulata, and Girtyocoelia dunbari. In the Glass Mountains, Fissispongia conus goes no higher than Road Canyon; the other two species no higher than the China Tank. Consequently, either the range of Cystothalamia must be lowered or those of the other three raised, or both. The best guide for temporal relationships within this zone may prove to be the lineage Guadalupia cupu- losa-G. vasa-G. cylindrica. Guadalupia vasa of the China Tank Member of the Word (Word 2) appears to be a direct development out of G. cupulosa of the preceding zone and to lead into G. cylindrica of the Getaway, Goat Seep, and higher formations. Guada- lupia vasa, Stylopegma annulata, and the genus Girty- ocoelia, seem to disappear at the top of the China Tank Member. The fauna of the China Tank is a distinctive assemblage of some temporal significance, and may be identified at the vasa-dunbari Microzone. The fauna of the Cherry Canyon Formation (Geta- way Limestone and Goat Seep Limestone in particu- lar) is also distinctive, and includes two species possibly peculiar to it (Stylopegma getawayensis and Virgola getawayensis) as well as the first appearances of Guad- alupia cylindrica and Stromatidium typicale, both of which continue into the overlying zone. This fauna, at least partly of temporal significance, is set off here as the nodulifera-typicale Microzone. Future collecting, especially in the upper part of the Word, from which relatively few sponge-bearing collections have been available for study, may ultimately show that the two microzones have a contiguous boundary, in which case they could serve as regular subzones. The genera Fissispongia, Polysiphonella, and Girty- ocoelia disappear from the Texas Permian record some- where within the microcamera-zitteliana Zone. The first two genera are absent in the Glass Mountains above the top of the preceding zone and Girtyocoelia is absent above the China Tank Member. In two locali- ties the Guadalupe Mountains-Delaware Mountains area, however, one or another of them occurs together with Cystothalamia. Either the last-named genus ranges lower than it does in the Glass Mountains or the first three range higher, or both. This question cannot be settled at the moment. Polysiphonella occurs in one of these localities with Guadalupia vasa. If we are to use the Guadalupia evolutionary lineage as an Ariadne-thread in this labyrinth of incompatible ranges, then at least the range of Polysiphonella must be raised somewhat, and the range of Cystothalamia perhaps, but not necessarily, lowered. In the very large etched collections from the Getaway Limestone, there is no trace of Fissispongia, Polysiphonella, or Girtyocoelia, nor are they known from the Goat Seep, Capitan, or Bell Canyon Formations. It is likely, there- fore, that they disappear before the beginning of Cherry Canyon time (the nodulifera-typicale Micro- zone) . The diagnostic faunal association of the micro- camera-zitteliana Zone is as follows: Calcarea Sphinctozoa Maeandrostiidae Stylopegma isopora Finks S. wordensis Finks S. annulata Finks (China Tank only) S. getawayensis Finks (Getaway Limestone only) Guadalupiidae Guadalupia zitteliana Girty G. microcamera Finks G. explanata (King) G. cystauletoides Finks G. vasa Finks (China Tank only?) G. cylindrica Finks (Getaway-Goat Seep only) Cystothalamia nodulifera Girty (post-China Tank only?) Celyphiidae Girtyocoelia dunbari King (China Tank only?) Pharetronida Stellispongiidae Catenispongia agaricus Finks Virgola getawayensis Finks (Getaway only) Hexactinellida Stromatidiidae Stromatidium typicale Girty (Getaway only) NUMBER 3 297 Base of the microcamera-zitteliana Zone: Last appearances: By local extinction: Polysiphonella flabellata (or higher) Fissispongia (or higher) Guadalupia ramescens G. williamsi (or higher) By evolution: Guadalupia cupulosa (into G. vasa and G. microcamera) Virgola robusta (into V. getawayensis) First appearances: By evolution: Guadalupia vasa (from G. cupulosa) G. microcamera (from G. cupulosa) G. cystauletoides (from G. zitteliana) Top of the vasa-dunbari Microzone: Last appearances: By local extinction: Girtyocoelia (or higher) By evolution: Stylopegma annulata (into S. geta- wayensis) Guadalupia vasa (into G. cylindrica) Base of the nodulifera-typicale Microzone: First appearances: By immigration: Stromatidium typicale By evolution: Stylopegma getawayensis (from S. annulata or S. wordensis) Guadalupia cylindrica (from G. vasa) Virgola getawayensis (from V. ro- busta) Exclusive to the microcamera-zitteliana Zone: Guadalupia microcamera G. cystauletoides G. vasa (China Tank only?) Cystothalamia nodulifera (post-China Tank?) Stylopegma getawayensis (Getaway only?) Virgola getawayensis (Getaway only?) Acmes within the microcamera-zitteliana Zone: Stylopegma isopora (China Tank) S. annulata (China Tank) S. getawayensis (Getaway) Cystothalamia nodulifera (Getaway) Guadalupia zitteliana (Getaway) G. microcamera (Getaway) G. vasa (China Tank) Virgola getawayensis (Getaway) The megacysta-cylindrica Zone The Bell Canyon, Capitan, and Carlsbad Formations of the Guadalupe Mountains have a distinctive sponge fauna which constitutes this zone, the highest recog- nizable in the section. In it Cystothalamia megacysta replaces C. nodulifera, Guadalupia favosa replaces G. microcamera, and Virgola neptunia replaces V. getawayensis. These would all seem to be evolutionary develop- ments and suggest that the lower boundary of the zone may ultimately prove to be difficult to place exactly. Amblysiphonella guadalupensis Girty and Corynella capitanensis (formerly Anthracosycon ficus var. capi- tanense Girty, 1909) are distinctive species now seem- ingly confined to this zone. In a practical way the zone may be recognized by the great abundance of Guadalu- pia cylindrica and Virgola neptunia together with the diagnostic Cystothalamia megacysta. On the whole, the difference between this zone and the preceding, espe- cially the reefy facies of the Goat Seep, in which Guadalupia cylindrica has already become prominent, is less than between the earlier zones. The diagnostic species of the megacysta-cylindrica Zone are as follows: Calcarea Sphinctozoa Maeandrostiidae Stylopegma isopora Finks S. wordensis Finks Guadalupiidae Guadalupia zitteliana Girty G. explanata (King) G. cylindrica Girty G. favosa Girty Cystothalamia megacysta Finks Sebargasiidae Amblysiphonella guadalupensis Girty Pharetronida Stellispongiidae Catenispongia agaricus Finks Stratispongia cinctuta Finks Virgola neptunia (Girty) Corynella capitanensis (Girty) Hexactinellida Stromatidiidae Stromatidium typicale Girty Base of the megacysta-cylindrica Zone: Last appearances: By local extinction: ?Stylopegma getawayensis ?Guadalupia cystauletoides By evolution: Guadalupia microcamera (into G. favosa) Cystothalamia nodulifera (into C. megacysta) Virgola getawayensis (into V. nep- tunia) First appearances: By immigration: ?Amblysiphonella guadalupensis ?Corynella capitanensis By evolution: Guadalupia favosa (from G. micro- camera) Cystothalamia megacysta (from C. modulifera) Virgola neptunia (from V. getaway- ensis) 298 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Exclusive to the megacysta-cylindrica Zone: All species listed under first apperaances. Acmes within the megacysta-cylindrica Zone: Guadalupia zitteliana Guadalupia cylindrica Virgola neptunia Acme-Associations Many of the longer-ranging species show distinct periods of markedly greater abundance, not neces- sarily at the midpoints of their ranges, and often at more than one horizon (see Figures 1 and 2). Such flowerings can usually be recognized in more than one locality at the same horizon, and therefore are not a strictly local matter. The environment undoubtedly played a major part in determining the time and place of a bloom of a particular species. At the same time, an evolutionary, and therefore a temporal, ele- ment is also present, as can be seen by examining those species that have more than one period of abundance, such as Heliospongia vokesi or Caten- ispongia agaricus. Each of their blooms is accom- panied by the bloom of a somewhat different set of other species (see Figures 1 and 2). At particular horizons, the association of species enjoying a period of florescence gives a characteristic stamp to the sponge fauna. Such an association may prove to be more use- ful for correlation within a limited area than the zonal-guide fossils. The characteristic acme-associa- tions (I use the word "acme" perhaps somewhat loosely) are listed below. I use the same type of two- species name for the acme-associations as are used for the other biostratigraphic units. 1. The stenaulos-orthoplectum acme-association (Neal Ranch reef facies ): Stylopegma stenaulos Stereodictyum orthoplectum 2. The defuncta-vokesi acme-association (Decie Ranch Member of the Skinner Ranch Formation): Defordia defuncta Heliospongia vokesi 3. The vokesi-dunbari acme-association (Taylor Ranch Member of the Hess Formation): Stylopegma annulata Fissispongia conus Heliospongia vokesi Stereodictyum orthoplectum Guadalupia auricula Catenispongia agaricus Girtyocoelia dunbari 4. The lepta-flabellata acme-association (Cathedral Mountain Formation): Polysiphonella flabellata Guadalupia explanata G. lepta Actinocoelia verrucosa 5. The wordensis-robusta acme-association (Road Canyon Forma- tion): Stylopegma wordensis S. annulata Polysiphonella flabellata Fissispongia conus Guadalupia cupulosa G. ramescens Virgola robusta Catenispongia agaricus Stratispongia cinctuta Girtyocoelia dunbari 6. The isopora-vasa acme-association (China Tank Member of th* Word Formation): Stylopegma isopora S. annulata Guadalupia vasa 7. The nodulifera-zitteliana acme-association (Getaway Limestone Member of the Cherry Canyon Formation): Stylopegma getawayensis Cystothalamia nodulifera Guadalupia zitteliana Guadalupia microcamera Virgola getawayensis Catenispongia agaricus 8. The cylindrica-neplunia acme-association (Capitan Limestone): Guadalupia cylindrica G. zitteliana Virgola neptunia Extinction and Immigration The introduction of new forms by apparent immigra- tion may be suspected when they appear without known relatives in the area in earlier deposits. The principal such cases occur at the base of the vokesi-defuncta Zone, when the pharetronids appear for the first time in the form of Catenispongia and Stratispongia. This may also be the first appearance of the pharetronids anywhere. The related pharetronid Virgola appears at the base of the next higher flabellata-robusta Zone. It is sufficiently different from the other two for it to be likely to have evolved outside the area and to have entered by immigration. The genus Guadalupia ap- pears at the base of the williamsi-vokesi Subzone of the vokesi-defuncta Zone. It is seemingly without ante- cedent in the area and also appears to have entered as an immigrant. It has undergone a considerable adap- NUMBER 3 299 / -2B FIGURE 7.—Camerisma (Callaiapsida) arctica (Holtedahl). Sections of peripheral flange and covered groove system for ducting of feeding currents and sealing of commissure; all from specimen USNM 163651 (shown on Plate 3: figures 7-10), from Wrangell Mountains, Alaska, USGS loc. 7109. A, C, G, sections of right side, between 20 and 25 mm from ventral beak; E, section of left side, about 20 mm from beak; B, D, F, positions of flanges as shell gaped, reconstructed by tracing from the actual sections shown in A, C and E. Section G is shown, uninked on Plate 3: figure 21; compare with Plate 3: figure 22, showing similar system in C. (Callaiapsida) kekuensis. the distal edge of the curl that forms the peripheral groove. Thus, the shell could gape at the anterior while the posterior part remained sealed by the over- lapping flanges, and the peripheral groove remained closed by means of its thin cover. The segment of the commissure anterior to the flanges and up to about half the height of the fold also remained effectively sealed by the thin edge of the dorsal valve when the shell gaped wide enough to raise the dorsal valve above the peripheral groove. This mechanism is best ex- plained by the illustrations (Figures 7A-F; Plate 3: figures 21, 22). These peripheral grooves have not been discussed previously, but they can be seen in illustrations of C. pentameroides by Tschernyschev (1902, pi. 23, figs, lb-d, 2d) and Sarycheva and Sokolskaja (1952, pi. 48, fig. 268) and of C. arctica by Holtedahl (1911, pi. 21, figs. 1, 2). Many specimens are preserved as steinkerns that do not show these structures clearly, but the cited illustrations confirm their widespread presence and provide evidence that they are consistent features of the subgenus. About halfway from the turn of the flanks to the crest of the fold each lateral peripheral groove termi- nates in a shallow notch that provided an opening to the seawater. (Plate 3: figures 1, 5). This notch is essentially parallel to the midline of the shell, and opens toward the posterior, thus turning the course of each groove through a sharp angle. The line of juncture of the valves then continues on its slightly posteriorly di- rected course, bringing it well inside a broad flap of the dorsal valve. The ventral valve, from this point on each side of the fold to the crest, forms another set of shallow grooves that appear to meet at the crest. In- stead of being covered by a flap of the ventral valve, however, these short segments are covered by exten- sions of the dorsal valve, so that they opened when the valve gaped (Plate 3: figures 1,3). The net effect of this elaborate flap and groove system was to keep the shell very tightly sealed every- where when the shell was closed, and to keep it just as effectively sealed everywhere except at the top half of the fold when the shell opened. If the circulation was similar to that in some modern rhynchonellid brachio- pods, namely with incurrents lateral and the excur- rents median (Rudwick, 1962), the water would have flowed in through the notches in the sides of the folds, then through the essentially cylindrical pipes formed by the lateral grooves with their thin covers, and entered the visceral chamber just anterior to the broad posterolateral flanges. Excurrents would have flowed directly out near the crest of the fold, through the opening provided when the thin dorsal flap lifted off the shallowly grooved edge of the ventral valve. This shallow recessing of the ventral valve edge seems more likely to have provided a slight tongue-and-groove effect for tight sealing of the closed shell, rather than a channel for currents entering or leaving the gaping shell. According to Ager (1967, p. 163), the question of the direction of flow of feeding currents in fossil 326 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY brachiopods cannot be considered settled. If incur- rents were median in C. (Callaiapsida), waste and outward-flowing water would have left the shell through the pipes provided by the lateral grooves, and then would have been projected downward toward the substrate by the notches on the fold. Circulation could have been maintained entirely by ciliary action of the lophophore. Expelling water through the fold would suck it in through the lateral pipe system, or conversely, bringing it in through the open fold would force it out through the pipes. The pipes themselves, however, had to have been lined by epithelium which could have been ciliated. Active cilia within the pipes would have contributed to mainte- nance of flow through the system, but nothing that was observed in the shells offers solid evidence for or against this speculation. The gape of average-size adults could not have been wider than about 2 mm in order for the flange and peripheral groove system to have functioned in ex- cluding sediment. This narrow gape would have kept the dorsal flap over the opening on the fold in such a position as to continue to provide protection while the shell was open. A gape of 3 or 4 mm seems likely for the very large specimens that attained a shell width near 60 mm. McCammon (1968, p. 193) cites evidence that liv- ing brachiopods can survive on dissolved nutrients alone, without ingesting particles of any kind. At least this is true of species that now live in cold waters con- taining high concentrations of dissolved nutrients. If this boreal form C. (Callaiapsida) lived on dissolved nutrients alone, the peripheral grooves would have been sufficient to carry the inflowing nutrient-bearing currents as through a pipeline. This brachiopod was admirably adapted for living free on the sea floor, able to withstand the effects of a rain of sediment or of oscillating currents that caused the shell to settle into the substrate. The probably ideal position of the shell was with the deepest part of the ventral valve lowermost, buried to the level of the edge of the fold (Figure 8A) . Thus, the dorsal flap over the A FIGURE 8.— Camerism a {Callaiaps ida). Possible living positions under nearly complete burial. The covered grooves and overlappin g flanges around the commissur e enabled the shell to admit and expel water currents while excluding sediment.T he system would work best to the depth of burial depicted, but probably could have functioned with the shell completely buried as long as sediment remained uncompact ed. A, Shell lying on ventral valve in position similar to that illus- trated by Ivanova (1949, fig. 30) and Ager (1967, fig. If) ; B, shell buried with ventral beak lowermost, current system remaining protected in this position by extended valve edges. NUMBER 3 327 gape at the fold would have been in a position to pro- tect it from falling sediment, and the openings to the two lateral groove-flap systems would have faced down- ward. Thin sections show, however, that the posterior part of the dorsal valve is thickened also, but not quite to the thickness of the ventral valve. This pos- terior weight, and the heavy beak of the ventral valve curled around the posterior end of the shell, could have caused the shell to settle into the substrate wedgelike, beak lowermost, the flared fold and anterolateral flanks providing stability and a slowdown in shell settling by their displacement effect on the surrounding sediment (Figure 8B) . The complicated valve edges and protected water-current system undoubtedly could have allowed the animal to function while completely buried in sediment that was loose or only mildly com- pacted, as deeply as nutrient-bearing water, could pene- trate. The narrow gape, pipes for lateral currents, and the roofed opening at the fold crest, all worked to allow the animal to bring in and expel water while excluding unwanted sediment. Such an adaptation would have enabled this group of brachiopods to survive as a con- stituent of the infauna. ENCLOSING ROCK.—The largest collection of Camerisma (Callaiapsida) is from the Halleck Forma- tion of southeastern Alaska (Muffler, 1967). The speci- mens are from Kuiu Island and the northernmost of the Keku Islets just off the northeast point of Kuiu. The dominant rock type of the Halleck Formation on Kuiu Island is described by Muffler (1967, p. 22) as, "dark gray very calcareous siltstone that grades into silty limestone." The formation contains several other rock types at Halleck Harbor, but the specimens of C. (Callaiapsida) kekuensis (new species described be- low) from there were collected from dark gray argilla- ceous limestone of nearly uniform grain size near 0.02- mm diameter, containing quartz, biogenic calcite, dark rock fragments, pyrite, micrite, sparite cement, and about 50 percent clay. On a wave-cut terrace on the northernmost of the Keku Islets, the Halleck Forma- tion contains many beds of conglomerate in dark argil- laceous matrix, but specimens of C. kekuensis from there were found in a matrix of dark gray calcareous mudstone. Specimens identified as C. (Callaiapsida) arctica (Holtedahl) from the lower Permian near Skolai Pass, Alaska Range, are in black argillaceous limestone with authigenic pyrite. The section there also contains basalts and tuffs, as well as massive limestones (Moffit, 1938a). The same species was reported by Holtedahl (1911, p. 34) from Novaya Zemlya in "beds of a grey limestone, finely crystalline and not unlike the mosquensis-limestone, yet of a less compact and homogeneous character." Ivanova (1949, p. 109) re- ported C. pentameroides from the Upper Carbonif- erous of the Moscow Basin "exclusively in argillaceous facies." Tschernyschev (1902, pp. 15, 448) reported this species from white and light gray limestone, with- out further lithic description. Species of C. (Callaiapsida) seem to have inhabited muddy sea bottoms in northern waters; those in Alaska, at least, lived near areas of volcanic activity. The mudstones and shales that enclose these brachio- pods are calcareous, and massive limestones occur above or below them, suggesting that, although the latitude was high, conditions were not Arctic. The interbedded conglomerates suggest intermittent tur- bidite activity, possibly in conjunction with volcanism, and the jumbled occurrence of many shells in these beds indicates that they are not in the exact place they inhabited. Some shells, especially those of C. (Callaiapsida) and other Stenoscismatacea remain articulated although somewhat crushed; others are single valves, especially ventral valves. It may be inferred that they lived upslope from where they now are found, but probably in the same kind of sediment. Camerisma (Callaiapsida) arctica (Holtedahl) FIGURES 7, 9; PLATE 3: FIGURES 1-10, 18-21 Camarophoria pentameroides Holtedahl, 1911 [not Tscherny- .schev, 1902], p. 19, pi. 2, figs. 5, 6. Camarophoria sella var. arctica Holtedahl, 1924, p. 34, pi. 21, figs. 1, 2.—Likharev and Einor, 1939, pp. 69, 208, pi. 14, figs. 5a,b. ?Laevicamera cf. L. arctica (Holtedahl) Gobbett, 1963, p. 128, pi. 16, figs. 13-16. Levicamera pentameroides (Tschernyschev) Ustritskiy, in Ustritskiy and Chernyak, 1963, p. 100, pi. 27, fig. 14, pi. 28, figs. 3, 4. DESCRIPTION.—Shell large, attaining length near 50 mm, width near 60 mm, average length about 30 mm; outline wedgelike in all views; commissure strongly uniplicate with sharp fold producing median crest on brachial valve, median groove in sulcus, pro- ducing effect of narrow Gothic arch, crest normally deflected slightly left or right; sulcus wrapped strongly around anterior, thus turning commissure through about 90°, with strongest curvature at widely flared 328 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY flanks; ornament very weak, consisting of radial shell fibers crossed by light concentric growth lines, produc- ing faintly reticulated effect. Ventral valve moderately strongly convex in um- bonal region; beak strongly curved over dorsal umbo, without pedicle foramen; broad flanges on each side anterior to hinge teeth (strongly overlapped by cor- responding flanges on dorsal valve), giving way anteriorly to deep peripheral groove; very thin flap of shell covering peripheral groove; shallow notch on each side of fold (about midway between lateral flank and fold crest) where peripheral groove crosses commis- sure; groove then follows edge of sulcus in slight reflexion toward posterior at crest. Dorsal valve very strongly convex, swollen in umbonal region, profile nearly straight along sharp crest of fold; flanks sloping very steeply from crest of fold. Interior of ventral valve thickened at posterior; spondylium large, thick-walled, sessile near beak or partly buried by secondary shell thickening, supported farther forward by low median septum duplex. In- terior of dorsal valve less thickened; hinge plate short; cardinal process large, bulbous; crura beginning as ridges along sides of hinge plate, extending forward as slender apophyses; camarophorium buried in thick- ened apex of beak, extending forward as deep, nearly semicircular spoon supported by increasingly high and slender median septum duplex; intercamarophorial plate present in posterior of camarophorium, obscured in observed sections by shelly filling between hinge plate and camarophorium. COMPARISONS.—Camerisma (Callaiapsida) arctica is characterized by its large size, proportionately wide FIGURE 9.—Camerisma {Callaiapsida) arctica (Holtedahl). Section about 8 mm anterior to beak, showing thickened camarophorium; USNM 163656, from Wrangell Mountains, Alaska, USGS loc. 7109. Same sections are shown, uninked, on Plate 3: figure 18. outline, sharp fold that projects about the same dis- tance forward from the beak as do the two anterolateral flanks, and its moderately convex ventral valve. It at- tains the same large size as C. (Callaiapsida) kekuensis, but its average size seems to be somewhat smaller. The sharp crest of the fold begins farther back than in C. kekuensis, but farther forward than in the species identified as C. pentameroides (Tschernyschev) by Sarycheva and Sokolskaja (1952, pi. 48, fig. 268). The sharp crest of the fold and the lesser convexity of both valves distinguish it from C. pentameroides (Tschernyschev, 1902, p. 100, pis. 22, 23). Callaiapsida arctica differs from C. quadrata (Likharev and Einor, 1939, p. 70, pi. 14: figs. 4a-e) in its higher fold that does not extend as far forward, and the sharp crest begins farther back. PLATE 1.—[All figures natural size]. Figures 1-20.—Septa- camera stupenda, new species, from the Pybus Formation in Halleck Harbor, Saginaw Bay, Kuiu Island, Alaska. 1-4, Specimen of average size and shape; ventral, posterior, side, and anterior views (USGS loc. 2452; USNM 163627). 5-8, Large specimen with grossly thickened dorsal valve, pedicle foramen typical; ventral, anterior, side, and posterior views (USGS loc. 2452; USNM 163628, holotype). 9-12, Large specimen; ventral, side, posterior, and anterior views (USGS loc. 2451; USNM 163629). 13-14, Small specimen with typically inflated dorsal valve, dorsal and side views (USGS loc. 2450; USNM 163630). 15-16, Silicified specimen etched in acid, posterior view showing pedicle foramen, and interior showing septalium and spondylium imbedded in shelly thicken- ing; pedicle adjustor muscle attachments visible (USGS loc. 2452; USNM 163631). 17, Interior of broken silicified speci- men lacking secondary thickening, showing septalium, crura, and spondylium (USGS loc. 2452; USNM 163632). 18, In- terior of dorsal valve, posterior mesial part showing hinge plate, sockets, crural bases, septalium, septum, and adductor muscle area (USGS loc. 2451; USNM 163633). 19, Same part of a different shell showing similar features as in figure 18, but with slight differences in the hinge plate and some- what more of the crura (USGS loc. 2452; USNM 163634). 20, Interior of ventral valve with attached fragment of dorsal cardinalia, showing hinge plate and distinct eminences in posterior part of spondylium (USGS loc. 2450, USNM 163635). Figure 21.—Septacamera, species undetermined, from Hal- leck Formation on northernmost Keku Islet, USGS locality 3700. Anterior part of dorsal valve showing concave anterior margin; USNM 163636. Figures 22, 23.—Septacamera kutorgae (Tschernyschev). Side and interior views of hollow calcareous specimen; USNM 163637. Locality is unknown, but the specimen was identified by Tschernyschev and agrees with his illustrations (Tschernyschev, 1902), so it can be considered an authentic example of the species, probably a topotype. NUMBER 3 'ill- 1 !i ^i * Mi f- \\w W -a >jw -'I1*"', '.'? I t'i Mi-/ ^'t - \ 'v*U,-ttX.J ;!'uv\ ?fig1-''- •'/( \\vx I ?hi mf^^J KVVA,, 329 PLATE 1 330 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY OCCURRENCE.—One nearly complete and well- preserVed specimen and two fragments that show essential details of the peripheral grooves were collected from Permian rocks near Skolai Pass, Alaska Range, Alaska, by A. Knopf (USGS loc. 7109 grn: see Moffit, 1938a, p. 36). These rocks have not received a formal stratigraphic name but were considered by Girty (in Moffit, 1938a) to be Artinskian in age. Their detailed correlation with the Halleck Formation of the south- eastern islands of Alaska, which has been assigned a probable Leonard age (in Muffler, 1967), is uncertain and awaits further study. The species was described originally from Lower Permian rocks of Novaya Zemlya (Holtedahl, 1924) and subsequently confirmed in that region by Likharev and Einor, (1939). More recently it has been recog- nized in rocks of Taimyr that are considered Upper Permian (Paikhoian Stage) in age (Ustritskiy and Chernyak, 1963). If identification of the Alaska spec- imens with C. arctica is correct, it suggests an age for the rocks near Skolai Pass somewhat younger than the Halleck Formation, perhaps correlative with the Pybus Formation that overlies the Halleck in south- eastern Alaska. Camerisma (Callaiapsida) kekuensis new species FIGURE 10; PLATE 3: FIGURES 11-17, 22 DESCRIPTION.—Shell large, typically about 35 mm in length, 40 mm in width, maximum lenth and width 50 mm and 60 mm respectively (Table 1); outline FIGURE 10.—Camerisma {Callaiapsida) kekuensis, new spe- cies. Sections of two specimens from Halleck Formation, Kuiu Island, Alaska, USGS loc. 3683: A, section 10 mm anterior to ventral beak, confirming presence of intercamarophorial plate, USNM 163658; B, section of fragmentary specimen, about 12 mm anterior to ventral beak, USNM 163659. diamond shaped; profile wedge shaped; commissure strongly uniplicate with fold rounded for much of length, crest becoming sharp at anterior crest de- flected slightly right or left; median groove in sulcus correspondingly shallow except at anterior end; sulcus curving around anterior, but not for full 90°, thus projecting somewhat anteriorly at crest of fold; curva- ture of commissure rather gentle at anterolateral flanks; growth lines distinct but weak. PLATE 2.—[Figure 13, X 2; figure 29, X 1.5; all other views natural size]. Figures 1-16.—Septacamera pybensis, new species, from the Pybus Formation, Pybus Bay, Admiralty Island, Alaska. 1-4, Unusually wedge-shaped specimen, slightly compressed, ventral, side, posterior, and anterior views (USGS loc. 18345; USNM 163638). 5, Ventral valve, interior showing spondylium and lateral buttress plates (USGS loc. 2547; USNM 163688). 6, Dorsal valve, interior showing half of hinge plate, septalium, median septum, and muscle area (USGS loc. 2547; USNM 163639). 7-13, Silicified shell, partly broken showing interior, ventral, posterior, dorsal, side oblique showing spondylium and crus, anterior, and side views at natural size; 13, interior view of interlocking anterior marginal spines (USGS loc. 2547; USNM 163640, holotype). 14-16, Decorticated shell show- ing low number of costae in sulcus and typically low and rounded profile (USGS loc. 2541; USNM 163641). Figures 17, 18.—Septacamera opitula, new species. Speci- men USNM 163642, from Assistance Formation, Grinnell Peninsula, Devon Island, Arctic Archipelago, Canada, USNM loc. 769770; essentially at Geological Survey of Canada loc. 26406. 17, Ventral valve, ventral view; USNM 163642. 18, Dorsal interior, partly excavated, showing hinge plate and attached fragment of ventral spondylium; USNM 163644. Figures 19-21.—Septacamera opitula? from Pybus Forma- tion, on island near cannery in Saginaw Bay (numbered 70 on map in Muffler, 1967), Kuiu Island, Alaska. Ventral, side, and anterior views (USGS loc. 2385; USNM 163645). Figures 22-29.—Septacamera stupenda, new species, from the Pybus Formation, Halleck Harbor, Saginaw Bay, Kuiu Island, Alaska. 22, 23, Posterior half of silicified shell, etched, showing interior features; septalium, hinge plate, septum and muscle area of dorsal valve, spondylium with lateral buttress plates in ventral valve, and short lateral marginal spines interlocking (USGS loc. 2451; USNM 163646). 24, 25, Fragment of posterior parts of both valves, with unusu- ally well preserved crura and septalium and good view of hinge teeth and socket (USGS loc. 2451; USNM 163647). 26, 27, Silicified posterior part of thickened shell, with septalium and spondylium imbedded, muscle area deeply impressed in dorsal valve (USGS loc. 2450; USNM 163648). 28, Interior of posterior part of silicified shell showing septalium, hinge plate, and spondylium (USGS loc. 2450; USNM 163649). 29, Interior view of posterior part of dorsal valve (shown on Plate 1: figure 18) showing hinge plate, septalium, and posterior end of muscle area (USGS loc. 2451; USNM 163633). NUMBER 3 331 PLATE 2 332 TABLE 1.—Measurements in millimeters of Camerisma (Callaiapsida) kekuensis. Locality Length Width Thickness 5452 35 38 5452 36 38 — 5452 39 36 — 10276 36 44 — 3701 34 45 — 68AFsl8 47 55 50 Ventral valve flatly convex in umbonal region; beak fairly straight, but yet pressed against dorsal umbo without room for pedicle foramen; broad posterolateral flanges overlapped by dorsal flanges, leading into rela- tively shallow peripheral groove; sides of fold notched shallowly where peripheral groove crosses. Dorsal valve swollen in umbonal region, flatly convex along crest of fold, strongly convex transversely. Interior of ventral valve moderately thickened; spondylium large, wide, with thick walls, buried in thickened shell at posterior, supported on low spondy- lium duplex toward anterior; pallial lines slender, dichotomous, typically stenoscismatacean. Interior of dorsal valve with bulbous cardinal process; camaro- phorium thick at posterior end near hinge plate (Figure 10) becoming high and more delicate, curving strongly toward spondylium; intercamarophorial plate thick, normally buried in secondary shell filling be- tween hinge plate and camarophorium. HOLOTYPE.—USNM 163655 (Plate 3: figures 15-17). COMPARISONS.—Camerisma (Callaiapsida) keku- ensis is characterized by its large size, comparatively flat ventral valve, fold that becomes sharp along the crest farther forward than in C. arctica (Holtedahl), and its fold and sulcus that project forward, rather than wrapping around the anterior to become perpendicu- lar to the commissure as in C. arctica. In this respect it resembles C. quadrata (Likharev and Einor, 1939, p. 70, pi. 14, figs. 4a-e), although the forward projec- tion is stronger in that species, the ventral valve very strongly convex, and the proportionate width nar- rower. Its sharp-crested fold differentiates it from C. pentameroides (Tschernyschev, 1902, pp. 100, 510, pi. 22, fig. 1; pi. 23, figs. 1-3) as do its less-swollen ventral valve and the less strongly convex profile of its ventral valve. The fold of C. kekuensis continues SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY nearly straight forward from the umbonal region, where in C. pentameroides it curves strongly ventrally, producing a more globular profile for the entire shell. In the Carboniferous species identified as C. pentame- roides by Sarycheva and Sokolskaja (1952, p. 172, pi. 48: fig. 268) the sharp crest of the fold begins nearly at the dorsal beak, the fold becomes very sharp and narrow toward the front, and the ventral valve is thick and convex throughout its length. This species is clearly different from C. pentameroides of Tscherny- schev, as well as from the other Alaska species. OCCURRENCE.—This species is named for its occur- rence in the Halleck Formation of the Keku Islets of PLATE 3.—[All figures natural size except where indicated.] Figures 1-10.—Camerisma {Callaiapsida) arctica (Holte- dahl) from Permian near Skolai Pass, Wrangell Mountains, Alaska (USGS locality 7109). 1-6, Well-preserved shell (USNM 163650). 1, dorsal view showing shallow groove on fold, covered by flap of dorsal valve (arrow points to notch where peripheral groove opens) ; 2, oblique view showing peripheral groove on flank; 3, side view showing peripheral groove still partly covered by flap of ventral valve; 4, ventral view with groove visible on left; 5, anterior view showing peripheral groove terminating at notch marked by arrow, also remains of notch on broken side; 6, posterior view. 7-10, Posterior half of broken and slightly compressed shell (USNM 163651) which was cut and polished to show relationships of shell sections inked on figure 7 and shell margins shown on Plate 3: figure 21. Figures 11-17.—Camerisma {Callaiapsida) kekuensis, new species, from the Halleck Formation, Kuiu Island and Keku Islets, Alaska. 11-12, Ventral valve showing peripheral grooves and fold, ventral and anterior views (USGS loc. 10276; USNM 163652). 13, Section (X1.75) about 10 mm anterior to ventral beak, showing thickened spondylium and camarophorium (USGS loc. 3683; USNM 163653). 14, Rubber mold of shell (X2) broken longitudinally, showing relation of spondylium to long and curved camarophorium, dorsal septum missing (USGS loc. 3683; USNM 163654, for comparison with Tschernyschev's, 1902, pi. 23, fig. 3). 15-17, Large specimen, almost entirely decorticated, from Halleck Formation on northernmost Keku Islet, showing form of shell, high fold, remnants of peripheral grooves, ends of septalium and spondylium, dorsal, anterior, and side views (USGS loc. 68AFsl8; USNM 163655, holotype). Figures 18-21.—Camerisma {Callaiapsida) arctica, from USGS loc. 7109. 18-20, Sections of posterior part of broken shell, 9 mm, 11 mm, and 13 mm from ventral beak; USNM 163656. 21, Section (X6) of right margin of shell shown in Figure 7G and in figures 7—10 on this plate. Figure 22.—Camerisma {Callaiapsida) kekuensis. Section (X 7) through lateral margin showing relation of dorsal flap to ventral peripheral groove (USGS loc. 3683; USNM 163653). Same shell as in figure 13 of this plate, but about 22 mm anterior to beak. NUMBER 3 333 PLATE 3 334 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY southeastern Alaska, just off the northeast shore of Kuiu Island where it also is found in the same forma- tion along the tidal flats of Halleck Harbor in Saginaw Bay. The geology of this area was mapped by Muffler and others, and the formation described by Muffler (1967, p. 22). Most of the Kuiu Island, Saginaw Bay, specimens studied here were collected by Buddington in 1924 (USGS loc. 5452), Waters and Girty in 1918 (USGS loc. 3683), Girty in 1918 (USGS loc. 3701), and Williams in 1940 (USGS loc. 10276). Those from the northern Keku Islet were collected by Waters and Girty in 1918 (USGS Iocs. 3685, 3700), and Stehli and Grant in 1968 (USGS loc. 68A-Fs-18). The age of the Halleck Formation has been recog- nized as Permian since the faunas were studied by Girty (as reported in Buddington and Chapin, 1929, pp. 118- 126). A probable Leonard age was estimated by Dutro (in Muffler, 1967, p. 23). Literature Cited Ager, D. V. 1967. Brachiopod Paleoecology. Earth Science Reviews, 3:157-179. Amsterdam: Elesevier. Buddington, A. F., and T. Chapin 1929. Geology and Mineral Deposits of Southeastern Alaska. United States Geological Survey Bulletin, 1241-C:C1-C52, 1 plate. Cooper, G. A. 1957. Permian Brachiopods from Central Oregon. Smithsonian Miscellaneous Collections, 134 (12): 1— 79, 12 plates. Cooper, G. A., and R. E. Grant 1966. Permian Rock Units in the Glass Mountains, West Texas. United States Geological Survey Bulletin, 1244-E:E1-E9, 2 plates. Dunbar, C. O. 1955. Permian Brachiopod Faunas of Central-Eastern Greenland. Meddelelser om Gronland, 110(3) : 1- 169, 32 plates. Gobbett, D.J. 1963. Carboniferous and Permian Brachiopods of Sval- bard. Norsk Polarinstitutt Skrifter, 127:1-201, 25 plates. Oslo. Grabau, A. W. 1936. Early Permian Fossils of China. Part 2: Fauna of the Maping Limestone of Kwangsi and Kweichow. Palaeontologia Sinica, series B, 8(4): 1—441, 31 plates. Grant, R. E. 1965. The Brachiopod Superfamily Stenoscismatacea. Smithsonian Miscellaneous Collections, 148(2) : 1- 192, 24 plates. 1968. Structural Adaptation in Two Permian Brachiopod Genera, Salt Range, West Pakistan. Journal of Paleontology, 42(1): 1-32, plates 1-9. Harker, P., and R. Thorsteinsson 1960. Permian Rocks and Faunas of Grinnell Peninsula, Arctic Archipelago. Geological Survey of Canada Memoir, 309:1-89, 25 plates. Holtedahl, O. 1911. Zur Kenntnis der Karbonablagerungen des west- lichen Spitzbergens. 1: Eine fauna der moskauer stufe. Videnskapsselskapet I Kristiania Skrifter Matematisk-Naturvidenskabelig Klasse, 10:1-46, plates 1-5. 1924. On the Rock Formation of Novaya Zemlya. In Re- port of the Scientific Results of the Norwegian Expedition to Novaya Zemlya, 1921. 22:1-183, 44 plates. Ivanova, E. A. 1949. Usloviya sushchestvovaniya, obraz zhizni i istoriya razvitiya nektorikh brakhiopod srednego i ver- khnego Karbona Podmoskovnoi Kotlovini. Aka- demia Nauk SSSR, Trudy Paleontologicheskogo Instituta, 21:1-148, 20 plates [seen in English translation]. Kutorga, S. 1844. Zweiter Beitrag zur Paleontologie Russlands. Ver- handlungen der Russisch-Kaiserlichen Mineralo- gischen Gesellschaft zu St. Petersburg, 4:62-104, plates 1-10. Likharev, B. K. 1960. Upper Paleozoic Rhynchonellida. Pages 239-257 in Sarycheva [q.v.], Osnovi Paleontologii. Likharev, B. K. and O. L. Einor 1939. Materiali i poznaniyo Verkhnepaleosoiskikh faun Novoi Zemli Brachiopoda. Trudy Arcticheskogo Nauchno-Issledovatelskogo Instituta, 127:1-245, 28 plates. Loney, R. A. 1964. Stratigraphy and Petrography of the Pybus- Gambier Area, Admiralty Island, Alaska. United States Geological Survey Bulletin, 1178:1-103, 2 plates. McCammon, H. M. 1968. Distribution of Articulate Brachiopods with Rela- tion to Organic Nutrients. [Abstract.] Geological Society of America, Program with Abstracts for 1968 Meeting, page 193. Moffit, F. H. 1938a. Geology of the Chitina Valley and Adjacent Area, Alaska. United States Geological Survey Bulletin, 894:1-137, 13 plates. 1938b. Geology of the Slana-Tok District, Alaska. United States Geological Survey Bulletin, 904:1-54, 4 plates. Moore, R. C. (editor) 1965. Treatise on Invertebrate Paleontology, Part H, Brachiopoda. 927 pages, 746 figures. New York City and Lawrence, Kansas: Geological Society of America and University of Kansas Press. NUMBER 3 335 Muffler, L. J. P. 1967. Stratigraphy of the Keku Islets and Neighboring Parts of Kuiu and Kupreanof Islands, Southeastern Alaska. United States Geological Survey Bulletin, 1241-C:C1-C52, 1 plate. Nassichuk, W. W.; W. M. Furnish; and B. F. Glenister 1965. The Permian Ammonoids of Arctic Canada. Geo- logical Survey of Canada Bulletin, 131:1-56, 5 plates. Rudwick, M. J. S. 1962. Filter-Feeding Mechanisms in Some Brachiopods from New Zealand. Linnean Society of London, Zoology, 44(300): 592-615, 17 figures. 1964. The Function of Zigzag Deflexions in the Commis- sures of Fossil Brachiopods. Palaeontology, 7(1): 135-171, plates 21-29. Rzhonsnitskaja, M. A. 1956. Systematization of Rhynchonellida [Abstract.] 20th Congreso Geologico Internacional, Mexico; Res- umenes de los Trabajos Presentados. Pages 125, 126. 1958. K sistematike Rinkhonellid. 20th Congreso Geo- ISgico Internacional, Mexico. Seccion 7, pages 107— 121. Sarycheva, T. G. (editor) 1960. Osnovi Paleontologii; spravochnik dlya Paleontol- ogov i Geologov SSSR: Mshanki, Brakhiopodi. 324 pages, 82 plates. Akademii Nauk SSSR. Sarycheva, T. G., and A. N. Sokolskaja 1952. Opredelitel Paleozoickikh brakhiopod Podmos- kovnoi Kotlovini. Akademia Nauk SSSR, Trudy Paleontologicheskogo Instituta, 38:1—307, 71 plates. Schellwien, E. 1900. Die fauna der Trogkofelschichten in den Karnis- chen Alpen und den Karawanken. Part 1, Brachi- opoden. Kaiserlich-kbnigliche Geologischen Reich- sanstalt Abhandlungen, 16(1) :1—122, 15 figures, plates 1-15. Schmidt, H. 1937. Zur Morphogenie der Rhynchonelliden. Sencken- bergiana Lethaea, 19: 22—60. Sokolskaja, A. N., and A. D. Grigorieva 1968. Order Rhynchonellida. In Sarycheva (editor), Brakhiopodi Verkhnego Paleozoya Vostochnogo Kazakhstana. Akademii Nauk SSSR, Trudy Paleontologicheskogo Instituta, 121:1-212, 38 plates. Stehli, F. G. 1957. Possible Permian Climatic Zonation and Its Im- plications. American Journal of Science, 255:607— 618. 1964. Permian Zoogeography and Its Bearing on Climate. Pages 537-549 in A. E. M. Nairn (editor), Prob- lems in Paleoclimatology: North Atlantic Treaty Organization Paleoclimates Conference, Newcastle upon Tyne and Durham, England, 1963 Proceed- ings. Stepanov, D. L. 1937. O Nekotorikh Verkhnekamennougolnikh brakhio- podakh Urala (On Some Upper Carboniferous Brachiopods of the Ural). Seria Geologo-Poch- venno-Geographicheskaya, 3(4): 114—50 [in Rus- sian, with English summary]. Tschernyschev, T. 1902. Verkhnekamennougolniya brakhiopodi Urala i Timana (Die obercarbonischen Brachiopoden des Ural und des Timan). Trudy Geologicheskago Komiteta (Memoires du Comite Geologique), 16(2): 749, 63 plates. [Two volumes, text and plates; in Russian with German summary.] Ustritskiy, B. I., and G. E. Chernyak 1963. Biostratigraphiya i brakhiopodi Verkhnego Paleo- zoya Taymira. Trudy Nauchno-Issledovatelskogo Instituta Geologii Arktiki, 134:1-139, 42 plates. Weller, S. 1914. The Mississippian Brachiopoda of the Mississippi Valley Basin. Illinois Geological Survey Mono- graph 1, 508 pages, 36 figures, 83 plates. Westbroek, P. 1967. Morphological Observations with Systematic Im- plications on Some Paleozoic Rhynchonellida from Europe, with Special Emphasis on the Uncinulidae. 82 pages, 14 plates. Leiden: Groen en Zoon. [Thesis, Leiden University.] Wright, F. E., and C. W. Wright 1908. The Ketchikan and Wrangell Mining Districts, Alaska. United States Geological Survey Bulletin, 347:1-210, 12 plates. Yanishevskiy, M. E. 1935. Opisanie fauni izh osnovaniya Uglenosnoi Tolshchi Kuznetskogo Basseina. Seria Geologo-Pochvenno- Geographicheskaya, 1(1) : 53-76, 6 plates. Francis G. Stehli Tethyan and Boreal Permian Faunas and Their Significan ce ABSTRACT The Tethyan and Boreal faunal provinces of the Permian are shown to be temperature controlled. In terms of brachiopod families, the Boreal fauna con- sists of cosmopolitan forms and is devoid of endemics. The Tethyan fauna contains in addition to the cosmo- politan families many endemic families. The brachio- pod families occurring in the Boreal realm are significantly older and more prosaic than the endemic families of the Tethyan realm, where evolution is evidently more rapid, and more distinctive forms occur. Correlation should be easier within the Tethyan realm than either between it and the Boreal realm or within the Boreal realm, especially during intervals of glacial climate and thus may account for some of the difficulties experienced in intra-Permian correlation. It has been recognized for many years that Permian faunas seem generally separable into two major, and to a large extent mutually exclusive, assemblages which have been termed Boreal and Tethyan. The present study evaluates the significance of the Boreal and Te- thyan faunas of the Permian in terms of temperature, composition, evolution, and effect upon correlation. Acknowledgement is made to G. A. Cooper and R. E. Grant for assistance in identification and to P. O. Banks, R. G. Douglas, and W. B. Clapham, Jr. for critical reading and editorial comment. Francis G. Stehli, Department of Geology, Case Wester Reserve University, Cleveland, Ohio 44106. General Considerations Because the fossil record, as presently known, is mani- festly imperfect, this study is confined to simple, strongly expressed phenomena which may be expected, under favorable circumstances, to be recognizable and interpretable despite perturbing effects. Preference is given to data which can be expressed in terms suscepti- ble to statistical evaluation. Data relate principally to the northern hemisphere because of the superior ex- pression of marine Permian rocks and the more thor- ough collecting and study accorded the fossil remains incorporated in them. Marine invertebrates are em- ployed because of their abundance. Emphasis is given to brachiopods because they are the group best known to me and thus most easily and reliably treated. Faunal Discontinuities The application of the distinguishing terms, "Boreal" and "Tethyan," to Permian faunas implies the exist- ence of some kind of faunal discontinuity. There is, in fact, an element of exclusivity in the distribution of. these two assemblages which, while not complete, re- sults in the dominance of Tethyan faunas in low and intermediate latitudes and of Boreal faunas in high latitudes (Newell, 1955, p. 8). This relationship ap- pears to suggest temperature control. It must be re- called, however, that such a conclusion is supported only if we are justified in assuming that the present earth is a geographically suitable model for Permian time. Since the possibility of continental displacements across latitude is widely asserted, we may not be justi- fied in using such a model. The question of whether the Tethyan-Boreal faunal province boundary is tem- 337 338 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY JACCARDS COEFFICIENT perature controlled should, if possible, be answered without involving latitude or committing ourselves to a particular configuration of the Permian earth. A Model We may begin by considering the causes of faunal discontinuities in the present seas as a model for in- terpretation of the Permian. Gunter (1957, p. 159) notes that "Temperature is the most important single factor governing the occurrence and behavior of life." This is true since, in a simplistic sense, organisms are physico-chemical machines whose vital reactions are temperature controlled. All ecologists are aware, how- ever, that some faunal discontinuities are not wholly or even primarily temperature controlled, and one problem will be to find a means to distinguish between those which are controlled primarily by temperature and those which are not. We can use the distribution of present-day organisms as an example. Data on the occurrence of families of modern clams at 39 globally distributed stations accu- mulated by A. L. McAIester for a study of diversity (Stehli and others, 1967) have been examined by means of a cluster analysis (Bonham-Carter, 1967) which determines the level of association between the sample stations, in this case in terms of their clam family composition. Latitude of the stations is not in- volved in the analysis, so commitment to any model of the earth is avoided. A dendrogram (Figure 1) por- trays the successively lower levels of association (greater differences) among these stations. When temperature or latitude information is intro- duced, it becomes apparent that the cluster program has recognized discontinuities in the clam family dis- tribution, the more profound of which can be related to the latitude-controlled planetary temperature gradient. The primary dichotomy (0.25 association level) sepa- rates the fauna of the south polar region from that of the remainder of the world.1 The second dichotomy (0.35) separates the fauna of the north polar region from the warm-water faunas of the world. These low- est branchings probably represent temperature control in its purest form, because branchings at higher associa- tive levels will be determined increasingly by factors other than temperature. To show, however, that the 1 The south polar clam family fauna is homogeneous around the Antarctic continent; thus, in the interest of economy, it has been entered as a single station. FIGURE 1.—Dendrogram showing levels of association be- tween globally distributed stations in terms of the assemblage of Recent clam families occurring at each as expressed in terms of Jaccard's Coefficient. The stations are as follows: 1, Barents Sea; 2, Kara Sea; 3, Laptev Sea; 4, East Siberian Sea; 5, Chukotka Sea; 6, Orkney Islands; 7, British Isles; 8, Ireland; 9, Portugal; 10, South France; 11, French West Africa; 12, South Africa; 13, Arabian Sea; 14, Japan; 15, Queensland; 16, New South Wales; 17, Victoria; 18, South Australia; 19, New Zealand; 20, Point Barrow; 21, Puget Sound; 22, Monterrey; 23, Baja California; 24, Sinaloa; 25, Panama-Peru; 26, Chile-Argentina; 27, Brazil; 28, Puerto Rico; 29, West Florida; 30, Texas; 31, South Caro- lina; 32, Eastern Canada; 33, East Greenland; 34, Iceland; 35, Antarctica; 36, Hawaii; 37, Philippines; 38, Mississippi Delta; and 39, Spain: (A map of station locations and the sources of data used may be found in Stehli et al., 1967.) faunal clustering is fundamentally temperature con- trolled, the value of Jaccard's Coefficient of association for each station relative to one on the equator is plotted against temperature. Figure 2 shows that the relation- ship to temperature is very clear. In the data for present-day clam families we thus find that the single most fundamental discontinuity in NUMBER 3 339 each hemisphere is that between the cold-water and warm-water fauna. In a less objective way, the basic nature of these discontinuities has long been recognized. For instance, Ekman (1953, p. 186) indicates that the sharpest faunal discontinuity in the oceans is that be- tween temperate and subtropical faunas. Hence, no new information has been added by the present anal- ysis, but an objective technique has been used for recognizing breaks which requires no assumptions about latitude or temperature, yet can be interpreted in terms of both. The Permian Tethyan-Boreal Discontinuity Examination of clam family data yields a model which we can use to objectively analyze Permian faunal data for its primary temperature controlled discon- tinuity—provided that we can get sufficient data of adequate quality. Before attempting to utilize the geo- logic record, we should consider some of the problems which inevitably become involved if, instead of data like that for the clams derived from an instant in time, we deal with data representing a significant span of time such as the whole Permian Period. 25 5 75 10 125 15 175 20 225 25 275 30 TEMPERATURE OF SURFACE WATER IN DEGREES CENTIGRADE FIGURE 2.—Jaccard's Coefficient of association for Recent clam family composition, relative to equatorial stations, plotted against surface temperature of each station for the warmest month. Temperature data for August was used in the northern hemisphere and that for February in the south- ern hemisphere. Data have been averaged for classes of 2.5° C. and plotted as a 3-class running average. Solid points at the end of the curve represent raw data since a 3-class average is, for them, impossible. In the first place, evolution may cause some prob- lems, for if we are examining the entire Permian, but use organisms that persisted for only a part of the period, their presence in one area and absence in another might be due to different ages of the rocks in the two areas rather than to temperature control. Clearly this is a serious problem, but can be dealt with easily by restricting attention to groups of organisms that persisted through all or almost all of the Permian. This approach involves dealing with high taxonomic categories, and it was for this reason that families of modern clams, rather than genera or species were used in deriving the model. A second serious problem involves short-term cli- matic variation. There is evidence to indicate that at least a part of the Permian represents an interval of glacial climate. We know that in the Pleistocene, im- pressive latitudinal compression of climatic belts occurred during glacial intervals. We know also that despite these marked changes, the earth's climate re- mained cold in polar regions and warm in equatorial regions throughout the Pleistocene. Thus, although the slope of the latitudinally controlled temperature gradi- ent varies through time, its direction remains constant. The effect of variation in slope is to increase the noise level in the belt between the continually warm region and the continually cold region, for here either cold or warm conditions may be recorded depending on the glacial or interglacial climate at the time of deposition of any particular rock unit. Because the direction of the gradient is constant, however, it will continue to furnish a constant and strong signal in a globally distributed set of data. Thus, if the temperature-related signal is sufficiently strong, we can find the average position of a faunal discontinuity regardless of considerable fluctuation in climate. With the clam family analysis as a model for inter- pretation, let us attempt a similar analysis of Permian faunal data. The data to be used have been obtained from the literature, from examination of collections, and from information generously provided by col- leagues. It pertains to families of articulate brachio- pods, as well as a few corals and fusulinids. To achieve a measure of uniformity in taxonomic treatment, the generic assignments of brachiopods were made from specimens or plates by G. A. Cooper, R. E. Grant, or myself, and in some questionable cases, by all three of us. Family assignments for the genera present were then made following the system used in Part H of the 372-386 0—71- -23 340 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Treatise on Invertebrate Paleontology. The coral data refer to the families Lophophyllidae, Durhaminidae, and Waagenophyllidae. Data for the first family were taken from many sources in the literature and from collections. The studies of Minato and Kato (1965) on the Waagenophyllidae and on the Durhaminidae were used as the primary data source for these two groups, which are treated as a unit. Data for the fam- ily Fusulinidae were taken from many sources in the •r w uo ; : » SAMPLE NUMBER .9 .8 !: 6 ICIEh OEFF w ... 1 a 4_ < u u I r .!-?- FIGURE 3.—Dendrogram showing levels of association between northern hemisphere stations in terms of the assemblage of Permian families (mainly brachiopods) occurring at each, expressed in terms of Jaccard's Coefficient. The stations clus- ter in two main groups, one Boreal, the other Tethyan. Sta- tions in the Boreal cluster are: 3, Taimyr; 9, Alaska-Yukon Border; 10, Grinnell Peninsula; 11, Central-East Greenland; 17, Kolyma; and 19, Spitzbergen. Stations in the Tethyan cluster are: 1, Carnic Alps; 2, Japan; 4, Southern Urals; 5, Jisu Honguer; 6, Ussuriland, etc.; 7, Kiangsi; 8, Novaya Zemlya; 12, Central Oregon; 13, Glass Mountains, Texas; 14, Huehuetenango, Guatemala; 15, Kuiu Island and Keku Islets, Alaska; 16, Sicily; 18, Ko Muk, Thailand; and 20, Salt Range, Pakistan. literature while those for the family Verbeekinidae follow the recent study by Gobbett (1967). Data were obtained for 20 reasonably well-known and well-distributed localities in the present northern hemisphere and were subjected to cluster analysis (Figure 3). They separate into two fundamental clus- ters at the 0.24 level. According to the model, this primary division should represent that between warm- water and cold-water fauna (presumably—but not necessarily—of the northern hemisphere). The pri- mary dichotomy recognized by the cluster program separates the fauna into the long recognized Boreal and Tethyan subdivisions. The noise level in the fossil data is doubtless higher than that in the Recent Model because collection and study are less perfect and be- cause of the time-related problems mentioned earlier. No attempt is made, therefore, to interpret any branches of the Permian dendrogram that occur at higher associative levels. The probability appears strong, on the basis of comparison to the Recent Model, that the Tethyan-Boreal faunal discontinuity is temperature controlled. It would be desirable to have some independent evidence by means of which to test this conclusion.2 A study of the same clam family data used in our model has shown a pronounced relationship between the number of clam families present at a sample station and its temperature (Stehli et al., 1967.) This effect is known to be nearly universal in large groups of marine organisms and has been shown to be closely related to both temperature and latitude (Fischer, 1960; Stehli, 1968). As a very simple test, then, let us use this well-known temperature-dependent effect to confirm or deny our conclusion regarding the tempera- ture control of the Tethyan-Boreal faunal disconti- nuity. We can use the present-day clam family data as a model and determine the average number of families present in typical stations of the warm- and cold-water clusters (Table 1). " Lowenstam (1964) discusses the few O18/010 paleotem- perature determinations for the Permian and shows that they probably are from the high noise region between the warm- and cold-water realms. Globally distributed paleotempera- ture results for the Permian would be of the greatest interest, but are at present not available and probably never will be. The results given by Lowenstam probably reflect the changes in shallow water between glacial and interglacial intervals, though our knowledge of paleogeography is not yet such as to rule out the possibility that the observed differences were caused by temporary upwellings or some similar phenomena. NUMBER 3 TABLE 1.—Recent clam family data. Name of Station Number of Recent Clam Families Present NORTH COLD-WATER CLUSTER 1. Point Barrow, Alaska 161 2. East Greenland 3. Laptev Sea, U.S.S.R. I Average 18 4. Chukotka Sea, U.S.S.R. lj SOUTH COLD-WATER CLUSTER 1. Antarctica 24 Average 24 WARM-WATER CLUSTER 1. Persian Gulf 52] 2. Philippines 52 3. Texas Coast 45 Average 49 4. Sinaloa, Mexico 47 5. South France 50 J This simple comparison shows a strong difference between the diversity of the warm- and cold-water clusters and shows us a means of testing the conclusion regarding the temperature significance of the Tethyan- Boreal discontinuity. To make a test, the diversity of Permian brachiopods at the family level, as shown in the Boreal and Tethyan clusters, may be considered. Though collecting and study of Permian brachiopod faunas has been vigorous, it certainly has been less pre- cise than that for Recent clams. It follows, therefore, that diversity data for Permian brachiopod families will be less perfect than that for Recent clams. Imper- fect collecting, which is certainly the principal source of noise, will have the effect of lowering the apparent level of diversity, because one commonly finds some- what fewer taxa than were actually present. This diffi- culty can be minimized if we consider only those sta- tions from each cluster showing the highest diversity, since these will give the closest approximation to the true diversity (Table 2). Comparison of the family diversity of brachiopods in the Tethyan and Boreal clusters shows that diversity in the Tethyan facies is higher by about 50 percent. This test corroborates the temperature dependence, and we may thus be virtually certain that the faunal boundary is that between the warm- and cold-water faunas. The Permian data demonstrate that the Tethyan fauna was a warm-water facies and the Boreal fauna a cold-water facies. None of the data involved latitude or commitment to any continental configuration. We 341 TABLE 2.—Permian brachiopod fami ly data. Name of Station Number of Permian Brachiopod Families Present ''Boreal Cluster'' 1. Grinnell Peninsula, Devon Island 21) 2. Spitzbergen 3. Central-East Greenland 19 2Q/Average 21 4. Alaska-Yukon Border 24j "Tethyan Cluster" 1. Carnic Alps 2. Southern Urals 29' 32 3. Jisu Honguer, Mongolia 4. Glass Mountains, Texas 27 35 Average 30 5. Salt Range, Pakistan 30J are thus free to use this temperature-controlled faunal boundary to select the best fitting model of the Per- mian earth in terms of latitude. No extensive test is made in this study, but Figure 4 shows the stations of the Boreal and Tethyan Permian brachiopod clusters and those of the cold-water and warm-water clam family clusters plotted against pres- ent-day latitude. Both the Permian and the Recent data show only a minimal degree of overlap between the two groups. In the case of the Recent clams this overlap is seen to be due to ocean currents, particularly the Gulf Stream which accounts for the warm-water fauna of England and Ireland (50°-55° N) and the Labrador Current which accounts for the cold-water fauna of Eastern Canada (45°-50° N). In the Per- mian data the zone of overlap is farther north, but may again relate to currents. The occurrence of a warm-water station in Novaya Zemlya north of 70° N, for instance, may be due to a warm-water current running north through the Ural Seaway and into the Arctic Ocean during lower Permian time (Ustritskij, 1962). In general it may be said that while data are PERMIAN FAMILIES T CLAMS BECEN N 80 70 60 50 40 30 20 10 10 20 30 40 50 60 70 80 S FIGURE 4.-—Occurrence of Permian brachiopods, etc., and Recent clams as a function of present-day latitude. Open circles represent stations comprising the cold-water cluster from the dendrograms of Figures 1 and 3, while the solid points are the stations comprising the warm-water clusters in the same dendrograms. Southern Hemisphere data were not used in the Permian analysis. 372-386 0—71- -24 342 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY still limited for the Permian, those which exist fit a present earth-latitude framework about as well as the Recent clam data fit it. Composition of the Tethyan and Boreal Faunas The Tethyan fauna, rich in peculiar genera, has re- ceived much attention, and it has been widely noted that certain groups seem characteristic of this faunal facies (Newell, 1955; Stehli, 1957; Ustritskij, 1962; Rudwick and Cowen, 1967). The Boreal fauna, having a somewhat more prosaic character, has received less attention. An attempt is made from the data at hand to examine the basic character of both the Tethyan and Boreal brachiopod faunas and to determine at the family level whether the Boreal assemblage contains endemic elements or whether it is distinguished simply by the lack of Tethyan elements. Definition of Boreal and Tethyan Brachiopod Facies For the purposes of this investigation, the two Permian faunal facies can be defined in terms of the two pri- mary clusters obtained in the cluster analysis in Fig- ure 3. In this way, it becomes possible to designate various stations as belonging to one faunal association or the other and to examine in some detail the makeup of the brachiopod family fauna characterizing the stations of each cluster. Figure 5 shows the family composition of the two faunas in terms of the brachiopods and the few other faunal elements employed in the cluster analysis (Fusu- linidae, Verbeekinidae, Lophophyllidae, Waageno- phyllidae-Durhaminidae). We may assume once more that the stations giving the highest family diversity values for brachiopods are those which have been most thoroughly collected and studied and which therefore provide the most reliable data. The stations of each cluster showing the highest diversity have for this reason been chosen for analysis. There is in this pro- cedure a certain danger because it assumes diversity to be a step function in which two discrete levels of diver- sity are represented, while in fact diversity is known to be a relatively smooth function of temperature and thus latitude. As an approximation, however, the procedure is justified because the error involved is much smaller than that resulting from poor collecting. Using what appear to be the most reliable data, it is evident that 19 brachiopod families and one fusuline <0 N « 9> O •- — W n^tntOt*'iBO>0~~**t^*t,*t'0 N 8 O O ?- •O K so O" © *- ^- ^ ^ ^ *»- v ^- *r ^ ^ 10 m BOREAL CLUSTER • • • 2 4 ALASKA-YUKON • • • • • • • • • • • • • • • • • 2 1 GRINNELL PENINSULA • • • • • • • • • • • • • • 2 0 CENTRAL EAST GREENLAND • • • • • • • • • • • • • • 1 9 SPITZBERGEN TETHYAN CLUSTER • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 2 9 CARNIC ALP5 • • • • • • • • • • • • • • • • • • • • • • • • • • • 3 2 SOUTHERN URALS • • • • • • • • • • • • • • • • • • • • • • • • • • 2 7 JISU HONGUER • • • • • • • • • • • • • • • • • • • • 2 5 USSURILAND ETC. • • • • • • • 2 6 NOVAYA ZEMLYA • • • • • • • • • • • • • • • • 3 5 GLASS MOUNTAINS • • • • • • • • • • • • • • • 2 6 KUIU ISLAND • • • * • • • • • • • • • • • • • • • • • • • • • • • 2 5 SICILY • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 3 0 SALT RANGE FIGURE 5.—Permian brachiopod and other families occurring in what are believed to be the most reliable stations of the Boreal and Tethyan clusters of Figure 3. The families involved are keyed to the numbers along the top of the figure as follows: 1, Enteletidae; 2, Rhipidomellidae; 3, Isogrammidae; 4, Meekellidae; 5, Schuchertellidae; 6, Orthotetidae; 7, Chonetidae; 8, Stro- phalosiidae; 9, Teguliferinidae; 10, Aulostegidae; 11, Chonetellidae; 12, Spyridiophoridae; 13, Tschernyschewiidae; 14, Scacchincllidae; 15, Richthofeniadae; 16, Leioproductidae; 17, Overto- niidae; 18, Marginiferidae; 19, Echinoconchidac; 20, Buxtoniidae; 21, Dictyoclostidae; 22, Linoproductidae; 23, Lyttoniidae; 24, Poikilosakidae; 25, Spinolyttoniidae; 26, Uncinulidae; 27, Camarotoechiidae; 28, Rhynchotetraidae; 29, Wellerellidae; 30, Atriboniidae; 31, Stenoscisma- tidae; 32, Rhynchoporidae; 33, Retziidae; 34, Athyrisinidae; 35, Athyrididae; 36, Ambocoeliidae; 37, Cyrtinidae; 38, Syringothyrididae; 39, Spiriferidae; 40, Brachythyrididae; 41, Spiriferinidae; 42, Reticulariidae; 43, Elythidae; 44, Martiniidae; 45, Mutationellidae; 46, Labaiidae; 47, Dielasmatidae; 48, Notothyrididae; 49, Heterelasminidae; 50, Cryptonellidae; 51, Durhaminidae- Waagenophyllidae; 52, Lophophyllidae; 53, Verbeekinidae; 54, Fusulinidae. Shown in column 55 is the diversity at the family level for brachiopods at each of the stations used in this analysis. NUMBER 3 343 TETHYAN UNIQUES family dominate the Boreal assemblage, occurring at practically all stations. These dominant Boreal families occur in the Tethyan facies as well, and 17 of the 19 brachiopod families continue as dominant elements, as does the family Fusulinidae. On the basis of the data at hand, we find that only the Strophalosiidae and the Syringothyrididae among the brachiopods seem pos- sibly to be more important in Boreal than in Tethyan assemblages. In short, there is a cosmopolitan fauna common to both the Tethyan and Boreal facies and an endemic fauna restricted to and characteristic of the Tethyan facies. In addition to the dominant forms of the Boreal assemblage, there are 13 families of brachiopods and one of corals (Durhaminidae) which occur sparingly in the stations analyzed or in nearby stations not in- volved in the cluster analysis. The remaining 18 brachi- opod families together with the Lophophyllidae and Verbeekinidae are known only from the Tethyan facies. Many of the families unique to the Tethys are repre- sented by few genera in the Permian or are monotypic or are excessively rare so that their usefulness is im- paired. The families which appear to be dominant among those examined and found unique in the Teth- yan facies are relatively few. They are: Meekellidae, Richthofinidae, Lyttoniidae, Lophophyllidae, and Verbeekinidae. Presence of these families may be con- sidered diagnostic of the Tethyan facies in the Permian. In warm-water faunas of the present day, it is found that some groups show greater tolerance to tempera- ture variation than others and consequently have more extensive ranges. This seems also to be true among the Tethyan index families noted above with the Meekel- lidae, in particular, suggesting by their range less rigor- ous restriction to warm water. The Verbeekinidae, on the other hand—at least as their distribution is shown by Gobbett (1967)—are very strongly restricted and probably occupied only the warmest water region of the Tethyan facies. Evolution and the Tethyan and Boreal Discontinuity In analyzing the brachiopod family distribution in the Boreal and Tethyan facies, it was noted that the primary distinguishing characteristic of the Boreal facies was its complete dependence on cosmopolitan families. Unlike the Tethyan facies, it lacks endemic groups at the family level. It is of interest to inquire somewhat further into this faunal differentiation. A 0 20 40 60 80 100 120 140 160 180 200 TIME SINCE FAMILY ORIGIN IN 106 YEARS FIGURE 6.—Percentage of brachiopod families that occur in the Boreal facies and of those that are endemic to the Tethyan facies plotted against the number of millions of years since the origin of each family. The time scale is very rough, for it combines the uncertainties of the absolute age scale with those of family ranges given at best in terms of upper, middle, and lower divisions of periods and frequently lacking even that detail. The open circle control points refer to the dotted curve; the solid points to the solid curve. Each point repre- sents an average value for the 20-million year class in which it occurs. comparison can be made if we consider the relative antiquity of the Boreal (cosmopolitan) and Tethyan (endemic) families. Figure 6 compares the percentage of families in each group as a function of their time of origin prior to the Permian. The families of the Boreal (cosmopolitan) assem- blage show a pronounced mode in the region about 30 X 106 years. The number of families younger than this is small, and none occur which are first known from Permian rocks. A relatively small percentage are very much older than the model class. The Boreal as- semblage thus seems to be composed predominately of well-established middle-aged brachiopod families to- gether with a few successful and more ancient stocks. The curve representing the endemic Tethyan families shows a pronounced mode in the range of 10X106 years and in addition indicates the presence of a large number (22 percent) of families that actually appear first in the Permian itself. Although relatively few forms are materially older than the mode, there is a 344 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY higher percentage of ancient family stocks present in the Tethyan facies than in the Boreal. In short, the Tethyan fauna shows excesses relative to the Boreal fauna principally in very young groups, but to a lesser extent in very ancient families as well. The excess in ancient families in the Tethyan fauna can probably be interpreted to mean that successful and stable families were able to persist in the relatively constant warm-water environment almost indefinitely, whereas they were either unable to occupy the Boreal environment or could not persist under its more de- manding conditions. Of far more interest than the excess of ancient families is the excess of very young families in the Tethyan facies. This excess appears to afford clear evidence that evolution proceeds most rapidly in the warm-water region of the world. It appears to be in this region that major new adaptive thresholds are crossed, and it is from this region that their occupants sometimes spread into the cold-water region. It is of considerable interest to deterimne whether this pattern in evolution is a general one or is unique to the Permian brachiopods. Tethyan and Boreal Faunas and Correlation Problems Most of the data presented in this study have tended to show that a major temperature-controlled faunal dis- continuity separates the Boreal and the Tethyan facies. The Boreal fauna consists of cosmopolitan families which are relatively old. The Tethyan fauna in con- trast contains many endemic families which tend to be young. In other words, the Tethyan fauna contains the kind of rapidly evolving, short-lived, and distinc- tive groups which facilitate correlation, while the Boreal fauna is composed of prosaic, long-lived, and slowly evolving families. It is quite evident from these considerations that in the Permian, correlation within the Tethyan region should be a relatively simple matter. Correlations within the Boreal realm and between it and the Tethyan realm will tend to be relatively difficult, since in both cases, one must work with those families least favorable for the recognition of short time intervals. It may be argued that the problem is not serious because, after all, very little correlation is done on the basis of families, which are the units used in this in- vestigation. The problem is serious, however, for fami- lies are, after all, abstractions based on the genera and species comprising them, and they reflect the most pronounced properties of these lower ranking groups. It appears probable that north-south correlation will generally prove more difficult than east-west correla- tion and that biostratigraphy will always be able to produce more refined and useful subdivision of rocks and time in the warm-water regions of the world than in those characterized by cold waters. The problem will undoubtedly be greatest for times such as the Permian and the Present, when, due to the existence of glacial conditions, the earth's planetary temperature gradient is compressed. One might suspect that much of the confusion that has occurred in the recognition and correlation of Upper Permian rocks and fossils is in part a result of this problem. Literature Cited Bonham-Carter, G. F. 1967. Fortran IV Program for Q-mode Cluster Analysis of Nonquantitative Data Using IBM 7090/7094 Computers. Kansas Geological Survey Computer Contributions 17:1-28. Ekman, S. 1953. Zoogeography of the Sea. 417 pages. London: Sedgwick and Jackson, Ltd. Fischer, A. G. 1960. Latitudinal Variations in Organic Diversity. Evo- lution, 14:64-81. Gobbett, D. J. 1967. Palaeozoogeography of the Verbeekinidae (Per- mian Foraminifera). In Aspects of Tethyan Bio- geography. Systematics Association Publication, 7:77-91. Gunter, G. 1957. Temperature. Chapter 8 in Treatise on Marine Ecology and Paleoecology, Geological Society of America Memoir, 67:159—184. Lowenstam, H. A. 1964. Paleotemperatures of the Permian and Cretaceous Periods. Pages 227-252 in Problems in Paleocli- matology. New York, London, Sydney: Interscience Publications. Minato, M., and M. Kato 1965. Waagenophyllidae. Journal Faculty of Science, Hokkaido University, series 4, Geology and Miner- alogy, 12(3-4): 1-241. 1965. Durhaminidae. Journal Faculty of Science, Hok- kaido University, series 4, Geology and Mineralogy, 13(1) : 11-86. Newell, N. D. 1955. Permian Pelecypods of East Greenland. Meddelelser om Grenland, 110(4): 1-36. Rudwick, M. and R. Cowen 1967. The Functional Morphology of Some Abberant Strophomenidae Brachiopods from the Permian of NUMBER 3 345 Sicily. Bollettino delta Societh. Paleontologie a Itali- ana, 6(2): 113-176. Stehli, F. G. 1957. Possible Permian Climatic Zonation and Its Impli- cations. American Journal of Science, 255:607-618. 1968. Taxonomic Diversity Gradients in Pole Location: The Recent Model. Chapter 6 in Evolution and Environment. New Haven: Yale University Press. Stehli, F. G., A. L. McAlester; and C. E. Helsley 1967. Taxonomic Diversity of Recent Bivalves and Some Implications for Geology. Geological Society of America Bulletin, 78(4) :455-466, 10 figures. Ustritskij, V. I. 1962. Principal Stages in the Permian Evolution of Asian Marine Basins and Brachiopod Fauna. Interna- tional Geology Review, 4(4) :415-426. J. B. Waterhouse The Permian Brachiopo d Genus Terrakea Booker, 1930 ABSTRACT The linoproductinid genus Terrakea Booker is re- viewed. Two lineages are recognized, one with numer- ous ear spines, one without. The former, represented by T. pollex Hill in the early Permian (late Sakmarian, early Artinskian) of eastern Australia and New Zea- land developed into the Wordian (Kazanian) T. brachythaerum, a very prolific species, or artenkreis including also solida Etheridge and Dun and elongata Etheridge and Dun. Allies penetrated the northern hemisphere at a slightly early stage, and are repre- sented in Pai Hoi and eastern Siberia, and also in the Canadian Arctic and Yukon as Terrakea arctica new species, (basal Word) with specimens from higher Word equivalents, and probably in the Texas Moun- tains as Grandaurispina kingorum Muir-Wood and Cooper, and G. signata (Girty). The other lineage, with few ear spines, is first known in basal Wordian beds of New Zealand and eastern Australia as T. con- cavum Waterhouse, and reappears as a new species in Capitanian beds of New Zealand. It is also represented in east Siberia from a fauna not yet described. A poorly known species also occurs in high Permian beds (Ochoan-Djulfian equivalents) in New Zealand, but no specimens are known from the topmost Changsing equivalent of either China or New Zealand. Terrakea arctica is described. Terrakea is a Permian member of the linoproductinid subfamily Linoproductinae Stehli, 1954, closely allied to Lino productus and Cancrinella, and distinguished most readily by its posteriorly prolonged spine bases in the pedicle valve, and by its cardinal process, as well as other criteria involving shape, ornament, and internal details. /. B. Waterhouse, Department of Geology, University of Toronto, Toronto 5, Canada. The genus was proposed by Booker (1930) for Productus brachythaerus Sowerby, 1844, as interpreted from a productacean specimen so named and figured by Morris (1845, pi. 14, fig. 4c) from the Illawarra district, near Wollongong, New South Wales (Figure 1). The only specimen extant known to have been examined and described as brachythaerus by Sowerby is strophalosiacean (Hill, 1950; Maxwell, 1956), but the International Commission for Zoological Nomen- clature in Opinion 486, 1957, suppressed the specific name brachythaerus Sowerby, 1844, in favor of brachy- thaerus Morris (1845), making P. brachythaerus Mor- ris the type species of Terrakea (see Waterhouse, 1964a). Thanks are due Drs. E. W. Bamber and W. W. Nassichuk, Geological Survey of Canada, for the loan of Arctic and Yukon collections, and Dr. Bruce Runne- gar, Department of Geology, University of Queensland, for the loan of a specimen of Productus solidus. Dr. G. A. Cooper, Smithsonian Institution, Washington, D.C, kindly provided facilities for examining types of Grandaurispina. Dr. Alan McGugan, Department of Geology, University of Alberta, Calgary, showed me specimens from the Ranger Canyon Formation, and Dr. Victor Gamelin showed me Siberian Terrakea at the Paleontological Institute, Moscow. The figures were drawn with the help of Mr. M. Jurgeneit, and the photographs were prepared by Mr. D. O'Donovan, both at the Department of Geology, University of Toronto. One photograph was supplied by Mr. S. N. Beatus from the New Zealand Geological Survey, Lower Hutt. Discussion of Species All of the species so far ascribed to Terrakea come from eastern Australia and New Zealand, where a 347 348 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY number of forms are extremely useful for detailed cor- relation. Species are recognized primarily from the dis- tribution of the spines, particularly over the ears, and also the nature of the spine bases and minor internal criteria. Shape is highly variable, particularly for the two best known species T. brachythaerum and T. con- cavum, Waterhouse (1964a, figs. 26-30) demonstrating that populations of these species ranged from narrow elongated and highly incurved forms to very transverse sulcate and slightly incurved little inflated specimens. Local populations of T. pollex and elongatum appear to have more consistence in shape, suggestive of iso- lated subspecies. Early marine Permian (lower Sakmarian) beds with the "Eurydesma" fauna in the Yarrol Basin of Queensland have not yielded Terrakea (Maxwell, 1964), and no occurrences are known for sure from correlative beds of New South Wales and Tasmania, though modern studies on Productida are lacking for these areas. The first species appears in the overlying Cattle Creek and equivalent late Sakmarian or early Artinskian faunas as a typically narrow "thumb"-like species, with a strongly incurved pedicle valve, and numerous ear spines, described as Terrakea pollex Hill (1950). Correlative beds of the Telfordian Stage in New Zealand contain a probable subspecies, with iden- tical costellae and spinepattern, but a flatter broader visceral disc and heavily impressed muscle scars, re- flecting adaption to a current-swept environment (Wa- terhouse, 1964a). In both regions the species is often dominant. The next species to appear is a totally different form, called Terrakea concavum Waterhouse, 1964a, with a few ear spines, and body spines inclined from the disc at a high angle. Never more than common, this species characterizes the lower Braxtonian Stage of New Zealand (approximately Road Canyon of the Glass Mountains (Cooper and Grant, 1966) and per- haps Ufimian of the Pre-Urals), and apparently occurs in the upper Grange and basal Malbina beds of Tas- mania, and in the Ulladulla beds of New South Wales along the coast south of Wollongong, to judge from visits to these localities with M. R. Banks, University of Tasmania in 1967, and with K. S. W. Campbell, Australian National University, in 1963. First appear- ance of this plexus with few spines is in the underlying Berriedale Limestone of Tasmania, to judge from col- lections made by the writer. The slightly younger Mangarewa Formation, re- ferred to the upper Braxtonian Stage of New Zealand, and equivalent to high (or restricted) Wordian and Kazanian, is characterized by Terrakea brachythaerum (Morris) a large extremely abundant form descended from T. pollex, with numerous large ear spines and very long spine bases. It occurs also in the correlative Flowers Formation of Nelson, New Zealand, and across the Tasman Sea in the upper Malbina beds of Tasmania and correlative beds at Wollongong, New South Wales, in the Muree and Mulbring Formations of the Hunter Valley, northern New South Wales, and in the Peawaddy (Springsure) and correlative beds of Queensland. The species has been recorded in various Geological Survey reports from older horizons, but while such records are not necessarily unreliable, they have yet to be substantiated. Four other specific names are in use for Terrakea of this age. Terrakea fragile (Dana 1847, 1849) is based on a suite of specimens much better preserved than the original type, but seems to be identical with it within the limits of preservation (Waterhouse, 1964a). Terrakea brachy- thaerum elongatum (Etheridge and Dun, 1909) is a large elongate T. brachythaerum, with a very long trail, suggestive of considerable maturity (Waterhouse, 1964a), and found in New Zealand at least to have inhabited a slightly different environment. Productus solidus Etheridge and Dun 1909 from Queensland is based on fragments of very large specimens probably allied to T. brachythaerum elongatum (Waterhouse, 1964a). It was treated as valid species by Hill and Woods (1964), but needs detailed study to establish its validity, because specimens loaned me from the Uni- versity of Queensland by B. Runnegar show ear spines as in T. brachythaerum. Small and little inflated Ter- rakea figured by Etheridge and Dun (1909, pi. 43, figs. 6, 9, 11) were referred to a new species leve by Booker (1930, p. 70, fig. la, pi. 2, figs. 3, 4), but Fletcher and others (1952, p. 12) and Waterhouse (1964a, p. 81) considered leve to have been based on immature shells of brachythaerum. Younger species are limited to New Zealand. A widespread limestone (Glendale Limestone, AG 4 limestone) of Capitan age (Puruhauan Stage of New Zealand) contains a new species probably developed from T. concavum, with few ear spines, and very short spine bases. Still younger beds of the New Zealand Waiitian Stage (Djulfian of Armenia, Wuchiaping of China) contain rare elongate and incurved Terrakea, NUMBER 3 349 tt pollex n. subsp. X FIGURE 1.—Sequence and distribution of Terrakea in Australia and New Zealand. Species are arranged vertically in order of appearance (T. pollex coming in first) and horizontally according to geographic distribution. Those on the left are restricted to Australia; those on the right are restricted to New Zealand; and those in the middle are present in both countries (indicated by double side lines). Two lineages are suggested by the density of ear spines. The density is not known for the youngest species. with the spine pattern poorly known (Waterhouse, 1967a). No species of Terrakea has been found in the ap- proximately correlative fauna of a tillite horizon found below the Cygnet Coal Measures at the top of the Permian succession in Tasmania. Nor has the genus been recorded in topmost Permian beds of New Zea- land (Waterhouse, 1967a), ascribed to the Makarewan Stage, and equivalent to the Changsing fauna of south China, and to part of the so-called basic Triassic of Djulfa, Armenia, and Azerbaidzhan. Distribution Beyond Australia and New Zealand ARCTIC.—As stated by Waterhouse (1969), Victor Gamelin of the Geological Institute at Magadin has discovered two species of Terrakea in the Upper Per- mian deposits of eastern Siberia. I have examined these 350 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY specimens at the Paleontological Institute, Moscow, and fully agree with Gamelin's identifications. The specimens figured from the upper Talatin Suite and reported from the Vorkutsk Suite of Pai Hoi by Solo- mina (1960, pi. 8, figs. 8-12) as Cancrinella koninck- iana also look somewhat like Terrakea. Permian faunas of Siberia are so similar to those of Canada that it is therefore of no surprise to discover Terrakea near the base of the Upper Permian in the Richardson Mountains of the Yukon, and in the Arctic Archipel- ago on Melville Island. They have been found by the writer in collections submitted for examination by the Geological Survey of Canada. A single species occurs in the Assistance Formation of Melville Island, and approximately correlative beds in the Richardson Mountains, and is here described as a new species, Terrakea arctica. Specimens are also known in slightly younger beds (Plate 2: figure 15) in glauconitic sand- stone above the Assistance (Formation B of Nassichuk, 1965), and in the correlative Ranger Canyon Chert of British Columbia (McGugan et al., 1965; Logan and McGugan, 1968) to judge from collections seen at the Department of Geology, University of Alberta, Cal- gary, and kindly shown to me by A. McGugan. Ter- rakea, therefore, appears to be one of those genera with disjunct distribution noted by Waterhouse (1967c, 1969) as characterizing the Arctic realm of Siberia and Canada (at present latitudes of 65° to 70°), and the East Australasian realm of eastern Australia and New Zealand (at present latitudes of 20° to 45°). See Figure 2. OCCURRENCE IN TEXAS.—As hinted in Waterhouse (1969), it is by no means certain that Terrakea does have a more or less bipolar distribution, because speci- mens from the Glass Mountains of Texas closely ap- proach Terrakea in many respects. Specimens were referred by Muir-Wood and Cooper (1960, p. 305) to a new genus Grandaurispina, with type species G. kingorum Muir-Wood and Cooper (1960, pi. 121, figs. 1-13) from the Word Limestone Number 3—see also Avonia signata non Girty of King (1931, pi. 20, figs. 16, 17—not 18-24) and the Cherry Canyon Formation of the Guadalupe Mountains. They also referred Productus signatus Girty (1909, p. 263, pi. 22, figs. 4a, b) from the Cherry Canyon Formation to Grandauri- spina. Initially placed in the Linoproductinae Stehli 1954, Grandaurispina was transferred to the Overtoni- idae by Muir-Wood (in Moore, 1965, p. 341). Exami- nation of the type-species at the Smithsonian Insti- tution suggests that the genus is linoproductinid, with typical ornament and cardinalia. In some specimens the costellae are rather faint, and the nature of the spine bases varies within single specimens, from a rather aprupt emergence of the spine from the costella which may be wider than the spine, to swollen bases suggestive of the Overtoniidae. Frequently the spine bases are prolonged posteriorly for 4 or 5 millimeters into the shell (e.g., USNM 149994, loc. 706e) exactly as in Terrakea. The cardinal process is like that of Terrakea figured by Waterhouse (1964a) and the sep- tum is very long, and doubled posteriorly, and the marginal ridge well developed in the brachial valve. Juvenile pedicle valves are deeply pited internally in the type species, and grooves appear later in ontogeny just as in Terrakea. Dimples occur over the dorsal exterior as in Terrakea. One apparent difference lies in the cluster of spines over the inner ears of the pedicle valve in Grandauri- spina kingorum, each spine typically about twice as thick as the body spines, in a band between the posterior and lateral margins, though less sturdy and numerous in some individuals. Because of the different preserva- tion of Australian and New Zealand material, workers may not have realised that a similar cluster of sturdy spines on the inner ears is typical of Terrakea brachy- thaerum (Morris), as described by Waterhouse (1964a, p. 76) for New Zealand material and in the Australian material (Etheridge and Dun, 1909, pi. 42, fig. 8; Waterhouse, 1964a, p. 79). It seems probable that the two are congeneric, though no doubt Grandaurispina could be treated as a subgenus. It clearly occupied a very different niche and geographic realm. Paleogeographic Distribution Assuming that Grandaurispina is closely allied, we then have a crudely circum-Pacific distribution for Terrakea, a most unusual one as far as the Permian is concerned, for no other similar brachiopod distribu- tion comes to mind. Its apparent absence from even western Australia, apart from Timor, Himalyas, Salt Range, China, and Japan suggests that the genus, though highly tolerant in terms of latitude and there- fore presumably of temperature, was restricted by other factors from attaining world-wide colonization. From present knowledge, one might speculate that the genus arose in the early Permian (Artinskian, or perhaps up- per Sakmarian) in eastern Australia (from Cancrinel- NUMBER 3 351 kingorum, signata FIGURE 2.—Sequence and distribution of Terrakea throughout the globe. Species with few ear spines to left; with many, to right. Arrows indicate possible migration route. The ages indicated in column (1—9 stages) are not known for eastern Siberia. Extension of the arrow from British Columbia into Texas is tentative, in view of uncertainty on the validity of Grandaurispina and its relationship to Terrakea. Some evidence suggests that this genus was present in the Leonardian Stage. la?) and rapidly spread to New Zealand. Free inter- change of species between these regions persisted throughout most of the period in two lineages, one with densely spinose ears, the other without such spi- nose ears. At some time near the close of the Lower Permian, or beginning of the Upper Permian, both lineages penetrated east Siberia (presumably along the Tethys from New Zealand through Indonesia though no trace remains), and the spiny lineage continued into the Canadian Arctic, and on into cen- tral-southern United States, where rapid speciation probably occurred. Such a reconstruction is hardly well-supported by the fossil record, and further discov- eries are bound to show a more complicated pattern. Nevertheless, the distributional history accords well with our present knowledge of Permian climates. The 352 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY TABLE 1.— Occurren ces of Terrakea * and Inter con elation Stage Russian Platform New Zealand Tasmania Queensland New South Wales 9 Makarewan Cygnet Newcasde 8 Tatarian *Waiitian Ferntree PBlackwater 7 *Puruhauan Black Alley Tomago 6 Kazanian *Upper Braxtonian *Malbina D-E *Peawaddy *Mulbring *Muree 5 PUfimian *Lower Braxtonian *Malbina A Ingelara *Branxton 4 Baigendzinian Mangapirian *Berriedale Albebaran Greta 3 PAktastinian Sakmarian *Telfordian Nassau Golden Valley * Cattle Creek *Farley 2 Asselian Mourlonia beds Quamby Reids Dome Allandale 1 PAsselian or pre- Asselian Atomodesma Lochinvar Stage Verchoyansk Siberia Arctic Islands Richardson Mountains British Columbia Texas 9 8 7 Ochoan Hivach ? Atomodesma Capitan 6 Gijigin Omolon *Formation B *Licharewia *Ranger Canyon *Word 5 Djigdalin *Assistance Formation *Lissochonetes *Road Canyon 4 Djeltin Munugudjak Belcher Channel Yakovlevia-Liosotella ?Cathedral Mountain 3 Yasachnm Takutoproductus- Tornquista Skinner Ranch 2 Irbichansk Attenualella- Takovlevia-T omiopsis Ross Creek Wolfcamp 1 Burgali Ortholichia Telford Johnson Canyon likelihood that Terrakea arose in eastern Australia sug- gests that it was at first adapted to dwelling in waters severely cooled by ice of the great early Permian glacia- tion, centered south of Australia, and the fossil record throughout the remainder of the Permian in Australia and New Zealand suggests that the genus preferred cool rather than warm waters. As Waterhouse (1964b, 1967b; in Brown et al., 1967, pp. 211-214) and Banks (1968) have shown, there were three major glacial intervals each followed by a recovery period, then a lengthy warm interval in the Permian of Australia and New Zealand. Terrakea is particularly prolific in the rocks formed during the recovery intervals when the ice, though shrinking, still exerted a considerable cool- ing effect (e.g., Stages 3 and 6 in Table 1). Specimens are fewer during the warm and to lesser extent the cold episodes. The time that Terrakea appears to have successfully colonized the parts of the northern hemi- sphere appears to coincide with the cold episode in the middle Permian (Stage 5), when much of the world suffered cooling, though of course not as severe as in eastern Australia. This major cooling may have al- lowed Terrakea, hitherto restricted to the south by the warm-water equatorial barrier, to expand its world coverage considerably, penetrating the equator by means of an ephemeral invasion to enter the more favorable, cooler waters of Siberia and Arctic Canada. The genus persisted in this realm during the ensuing "recovery interval" of Stage 6 in Table 1. It is believed that the Wordian of Texas does possess a number of "cool-water," otherwise-Arctic genera, but it is never- theless surprising that Terrakea flourished in the low NUMBER 3 353 latitudes of Texas. Perhaps the genus at this stage changed or improved its ecological tolerance, unless I am mistaken in suggesting that Grandaurispina is al- lied to Terrakea. The possible occurrence of Terrakea in the Talatin Suite of Pai Hoi may not accord with the preceding reconstruction, because the Talatin faunas are re- garded as Baigendzinian (Licharev, 1966). But the Talatin faunas described by Solomina (1960) come from two levels, the lower appearing to be Artinskian, the upper in my view being younger, with forms such as Stepanoviella (Linoproductus ex grupo cora), Megousia (Linoproductus kulikii), Cancrinella jani- schewskiana, and other forms suggestive of Stage 5 or 6 in the Canadian and other Siberian sequences. Habitat The two lineages in Terrakea reflect adaptation to slightly different niches. Terrakea brachythaerum and its close allies, including pollex, arctica and presum- ably kingorum, posseessed large and numerous ear spines, that functioned somewhat like the lateral halteroid spines of Marginifera described by Grant (1968) for steadying the shell, while the body spines curve freely and evenly away from the disc, suggest- ing growth into water rather than sediment (Grant, 1968). New Zealand and Australian members of this lineage are generally found in sandy sediment, gen- erally arkosic or quartzitic or volcanic graywackes, and even in breccia and conglomerate. It seems highly likely that they dwelt in a sandy or finely graveled sea- floor, well winnowed by currents. The sharply genicu- lated trail in both pollex and brachythaerum suggests that the brachial valve became partly buried in sedi- ment, so that the trail had to grow, sometimes rapidly, at right angles to the disc in order to maintain contact with the water. A similar habit probably obtained for T. arctica from the sandy Assistance Formation of Melville Island. Yet in the Richardson Mountains what appears to be the same species is found in a very limy silt. Shells here are so numerous that the rock was a breccia of shell in a fine matrix. The amount of geniculation is not known for these shells. The habitats of Grandaurispina and the Russian Terrakea are yet to be described. Terrakea concavum, with few ear spines, inhabited muddy siltstone, thickly colonized by Echinalosia at its type-locality, and also in muddy limestone. Ear spines are scarce from early ontogeny in every known specimen, though the trail varies in geniculation. Since there is no other consistent difference in shape from brachythaerum, the shells must have been able to maintain stability without the numerous ear spines, thanks perhaps to quieter bottom conditions, and per- haps a high angle of inclination of the body spines. The younger New Zealand species, not as yet described, is found in a similar calcareous mud in the Glendale Limestone. In the limestone of the Arthurton Group, it is found in a muddy matrix with mats of bryozoa, betokening quiet sedimentation. Attention should also be drawn to the very sturdy spines along the anterior margin of the mature brachial valve, illustrated by Waterhouse (1964a) for T. aff. pollex (pi. 10, fig. 5), T. concavum (pi. 11, figs. 4, 5) and T. elongatum (pi. 15, fig. 15), and less commonly seen in T. brachythaerum (Figure 3). A specimen of a new subspecies of T. pollex is illustrated (Plate 2: figure 16) to show several rows of the fine spines over the pedicle disc and rows of very coarse spines over the trail. These spines are fully as thick as those over the anterior trail of the pedicle valve, in contrast to the slender posterior brachial spines. Though without adequate material to determine the exact course and length of the spines, I believe that these spines func- tioned as strainers in the same fashion as the posterior spines, and that their size may have reflected avail- ability of shell material and the fact that, by develop- ing at maturity when the shell was no longer increasing much in length, a long period of time was available FIGURE 3.—Terrakea brachythaerum (Morris). Pedicle valve showing generalized distribution of fine inclined body spines, passing posteriorly into hollow tubules within the shell, and massive erect spines over the inner ears. A, Gen- eralized from BR 158 figured by Waterhouse (1964a, pi. 12, fig. 8); B, from BR 66 figured by Waterhouse (1964a, pi. 13, fig. 12). Both from Mangarewa Formation, Letham Burn, Southland, and kept at the New Zealand Geological Survey, Lower Hutt. All views approximately X 1.5. Abbreviations: es, ear spines; k, umbo; sp, body spines. 354 for the anterior spines to remain functional. It seems difficult to imagine that these spines had any halteroid function. Elongated Spine Bases Spines at the front of the visceral disc, and over the ears, and in juvenile specimens pass straight through the shell. This remains true throughout ontogeny for spines over the brachial valve, and ears of both valves. But as the pedicle valve is thickened internally by layers of calcite over the floor, channels or gutters are left behind each open spine for the tissue that main- tained contact with the isolated epithelial lobe grow- ing at the tip of the spine. When the spine ceased to grow, the groove became covered over and sealed off. Similar open channels are occasionally seen in Can- crinella (Campbell, 1953, p. 7), thus providing some evidence for relationships to this genus, as suggested by Waterhouse, 1964a, p. 63). Examples of Cancrinella with the grooves are figured by Campbell (1953, pi. 1, figs. 5, 6), Hill and Woods (1964, pi. P 6, fig. 17a) and Waterhouse (1964a, pi. 9, figs. 5, 6). Genus Terrakea Booker 1930 DIAGNOSIS.—Concavoconvex Linoproductinae, usu- ally flattened across the venter, and commonly genicu- late. Ears of moderate size, ornament of costellae, rugae inconspicuous or absent, both valves spinose, arranged in quincunx, with many species bearing strong erect spines over the inner ears or adjoining lateral slopes, spines over visceral disc and trail of pedicle valve pass posteriorly into grooves, later enclosed by secondary shell, leaving hollow tubules. Spines of brachial valve erect. Muscle scars productiform, cardinal process usu- ally erect, with high median lobe subdivided by notch, and curving lateral lobes, sometimes with alveolus at base. Septum in brachial valve long, broad immediately in front of process, often subdivided by groove. TYPE-SPECIES.—Productus brachythaerus Morris 1845 non G. B. Sowerby 1844. Terrakea arctica, new species PLATE 1: FIGURES 1-15; PLATE 2, FIGURES 1-14, 17 DIAGNOSIS.—Small Terrakea, varying in shape from transverse with nonsulcate disc to highly arched with narrow flattened or sulcate disc, small ears, maximum SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY width anteriorly placed, costellae fine, spines numerous over ears, and close set, emerging at high angle from visceral disc of pedicle valve. Cardinal process low with deep median cleft in immature specimen. DESCRIPTION.—Size small for genus, pedicle valve varying in shape from high elongated vaulted speci- mens with sulcate of flattened venter and steep sides to transverse specimens with rounded venter, gently in- clined sides, and less incurved umbones, especially in specimens from the Assistance Formation at C-464 and C-1872. Specimens from the Richardson Moun- tains usually have rounder discs and arched venters. Ears small, with slightly acute cardinal extremities maximum width of shell generally near anterior third of length; brachial valve gently concave over venter, curving with strong geniculation into trail which ex- tends nearly half length of shell in mature specimens. Mature specimen apparently sulcate anteriorly, sug- gestive of "tubiform" trail. Ornament characteristic. Costellae fine, generally 10-12 in 5 mm anteriorly on pedicle valves, with up to 15 to 5 mm in specimens from C 464 and up to 17 in 5 mm on a brachial valve from C 462. Spines best preserved on silicified specimens, emerg- ing over visceral disc of pedicle valve at 35° to 45°, with slightly swollen bases, extended posteriorly for 0.5 mm in some instances, arranged roughly in quin- cunx, in rows nearly 2 mm apart in mature shells, spaced over 2 mm apart along a row, compared with spacing of only 1 mm apart for the first 10 mm or so PLATE 1.—Terrakea arctica, new species. All views are twice natural size. Specimens are from Assistance Formation of Melville Island (see Table 3 for locality details) and are at Geological Survey of Canada, Ottawa. 1, 2, Ventral and dorsal views of internal mold specimen 24487, from C-462, Sabine Bay, Melville Island. 3, Dorsal aspect of external mold specimen 24488 from same locality. 4, External mold of pedicle valve showing costellae and spine bases; specimen 24489, from same locality. 6, 9, 13, Posterior, lateral, and ventral aspects, respectively, of internal mold of pedicle valve 24491 from same locality. 5, Internal mold of pedicle valve 22490 from C 463, Sabine Peninsula, Melville Island. 7, 15, Lateral and ventral views of internal mold of pedicle valve 24492, from same locality. 8, 14, Posterior-ventral and ven- tral aspects of internal mold of pedicle valve 24493 from C 464, Melville Island. 10, Internal mold of pedicle valve GSC 24484 from C 463, Melville Island. 11, Ventral aspect of internal mold of pedicle valve 24485, from C 464. 12, Internal mold of pedicle valve 24486 from C 1872 from Mel- ville Island. NUMBER 3 355 • ??'£ 'Aifa; ?? ? ? sen, <2^ s ? wm ?A - jr^T *':' J ~* ^rfii f— 15 ' / ' ,••/. 8 5.0 9. 1 NUMBER 3 359 HORIZON.—Beds approximately equivalent to the Assistance Formation of Harker and Thorsteinsson, on Melville Island, and in a "Lissochonetes" zone of Richardson Mountains (GSC 53848, 53850), and slightly younger beds (GSC 53822, 53823, 53834). RESEMBLANCES.—In the well-defined radial orna- ment and high angle of emergence of its pedicle spines, this species comes close to Terrakea concavum Water- house (1964a) from the approximately correlative Letham Formation of New Zealand, also believed to occur in equivalent beds of Tasmania (Grange Mud- stone, Malbina A) and southern New South Wales. Though the southern species is generally not genicu- late and is ovally transverse without a median sulcus, some specimens, including topotypes, are geniculate, and sulcate. But this species is distinguished by its pau- city of spines over the ears of the pedicle valve. A closer ally in this respect is Terrakea brachythaerum (Mor- ris) from the Malbina D-E beds of Tasmania, and re- ported widely from correlative beds of New South Wales (Wollongong) and Hunter River—Muree and Mulbring Formations, and the "Fauna IV" of Queens- land, as well as correlative Mangarewa and Flowers Formations of New Zealand. A few Canadian speci- mens have the identical narrow shape, incurved urn- bones and flat or sulcate venter often seen in the Australasian species, and the arrangement of spines over the ears is much the same, as well as the geniculate trail. On the other hand, visceral spines emerge at a higher angle in the Canadian form and are much more closely set. Internally the cardinal process differs in being more sessile, with a deeper median cleft, though an alveolus and anterior double ridge are present as in some New Zealand specimens of T. brachythaerum. No other species from the southern hemisphere come very close, T. pollex Hill generally having an elon- gated outline with a moderate number of ear spines, but fewer visceral spines, and a younger New Zealand form having few ear-spines. Grandaurispina kingorum Muir-Wood and Cooper (1960, p. 306, pi. 121, figs. 1-13; also figured as Avonia signata non Girty of King, 1931, p. 83, pi. 20, figs. 16, 17?), from Word Limestone Number 3 of the Hess Canyon Quadrangle, has a somewhat similar density of spines over the disc, but its ear spines are much larger, and its cardinal process close to that of Austra- lian and New Zealand Terrakea. Closer resemblances to Texan forms might be revealed when the descrip- tions are published by G. A. Cooper and R. E. Grant. TABLE 3.—Localities with Terrakea arctica, new species. (A) Northern Richardson Mountains, collected by E. W. Bamber, Geological Survey of Canada, 1962. Locality number Description GSC 53822 Three miles north of Horn Lake, Aerial Photograph A 14363-14, 1362 feet below top of measured section (67°477136°02'). GSC 53823 Three miles north of Horn Lake, Aerial Photograph A 14363-14, 1375 feet below top of measured section (67°477136°02')- GSC 53834 Symmetry Mountain, Aerial Photograph A 14361-68, in talus 680 feet above top of Devonian (67°427136°15')- GSC 53848 Same section, 2425 feet above top of Devonian. GSC 53850 Same section, 2655 feet above top of Devonian. (B) Melville Island, Assistance Formation, collected by W. W. Nassichuk, 1964. C 462 Three miles NW of Tingmisut Lake, bearing 335°, Sabine Peninsula NF 64-2-11. C 463 Four miles NW of center of Tingmisut Lake, Sabine Peninsula, bearing 310°, NF 64-2-25. C 464 Photo A 16763-171, 16 miles SW of center of Tingmisut Lake, Sabine Peninsula, bearing 235°, NF 64-2-35. C 1872 Five miles SE of Tingmisut Lake; on the east side of the west arm of Weatherall Bay, Melville Island. C 465 [Terrakea sp.] Photo A 16763, 6 miles west of Tingmisut Lake NF 64-2-27. 360 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Literature Cited Banks, M. R. 1968. The Upper Paleozoic Ice Age and Permian Cli- mates in Australia. Geological Society of Australia, Specialist Symposium, May 1968, Permian of Aus- tralia, page 18. Booker, F. W. 1930. A Review of Some of the Permo-Carboniferous Productidae of New South Wales, with a Tentative Reclassification. Journal of the Royal Society of New South Wales, 64:65-77, plates 1-3. Brown, D. A.; K. S. W. Campbell; and K. A. W. Crook 1968. The Geological Evolution of Australia and New Zealand. 409 pages. Oxford: Pergamon Press. Campbell, K. S. W. 1953. The Fauna of the Permo-Carboniferous Ingelara Beds of Queensland. University of Queensland De- partment of Geology Papers, 4(3) : 1-44, plates 1-7. Cooper, G. A., and R. E. Grant 1966. Permian Rock Units in the Glass Mountains, West Texas. United States Geological Survey Bulletin, 1244-E:E1-E9, 2 maps. Dana, J. D. 1847. Description of Fossil Shells of the Collections of the Exploring Expedition under the Command of Charles Wilkes, U.S.N., Obtained in Australia from the Lower Layers of the Coal Formation in 111a- wara, and from a Deposit Probably of Nearly the Same Age at Harper's Hill, Valley of the Hunter. American Journal of Science, 54:151-160. 1849. In United States Exploring Expedition During the Years 1838, 1839, 1840, 1841 under the Command of Charles Wilkes, U.S.N. Geology, 10:681-713. Etheridge, R., and W. S. Dun 1909. Notes on the Permo-Carboniferous Producti of East- ern Australia. Geological Survey of New South Wales Report, 8:293-304, plates 41-43. Fletcher, H. O.; D. Hill; and R. W. Willett 1952. Permian Fossils from Southland. New Zealand Geol- ogical Survey Palaeontological Bulletin, 19:1—17, plates 1, 2. Girty, G. H. 1909. The Guadalupian Fauna. United States Geological Survey Professional Paper, 58:1-651, plates 1—31. Grant, R. E. 1968. Structural Adaptation in Two Permian Brachiopod Genera, Salt Range, West Pakistan. Journal of Paleontology, 42: 1-32, plates 1-9. Hill, Dorothy 1950. The Productinae of the Artinskian Cracow Fauna of Queensland. University of Queensland Depart- of Geology Papers, 3(2) : 1-36, plates 1-9. Hill,D., and J. T.Woods 1964. Permian Index Fossils of Queensland. Queensland Palaeontographical Society, Brisbane, pages 1—32, plates P1-P15. King, R. E. 1931. Geology of the Glass Mountains. Part 2: Faunal Summary and Correlation of the Permian Forma- tions with Description of the Brachiopoda. Uni- versity of Texas Bulletin, 3042:1-245, plates 1^-4. Licharev, B. K. (editor) 1966. The Permian System. Stratigraphy of the USSR. 536 pages, 62 figures. Moscow. Logan, A., and A. McGugan 1968. Biostratigraphy and Faunas of the Permian Ishbel Group, Canadian Rocky Mountains. Journal of Paleontology, 42(5) : 1123-1139, plates 141-144, 3 figures. Maxwell, W. G. H. 1956. Designation of a Type Species for Terrakea in Harmony with Present Usage. Bulletin of Zoologi- cal Nomenclature, 11(11):333—336. 1964. The Geology of the Yarrol Region. Part 1: Bio- stratigraphy. University of Queensland Department of Geology Papers, 5(9) :3-79, plates 1-14. McGugan, A.; H. K. Roessingh; and W. R. Danner 1965. In Permian Geological History of Western Canada, Chapter 8, pages 103-112, figures 8-1-8-14. Alberta Society of Petroleum Geology. Moore, R. C. (editor) 1965. Treatise on Invertebrate Paleontology, Part H, Brachipoda. 927 pages, 746 figures. New York City and Lawrence, Kansas: Geological Society of America and University of Kansas Press. Morris, J. 1845. Descriptions of Fossils. In P. E. de Strzelecki, Phys- ical Description of New South Wales and Van Diemen's Land, pages 270-291, plates 10-19. Lon- don: Longman, Brown, Green, and Longmans. Muir-Wood, H. M., and G. A. Cooper 1960. Morphology, Classification and Life Habits of the Productoidea (Brachiopoda), Geological Society of America Memoir, 81:1-447, plates 1-135. Nassichuk, W. W. 1965. Pennsylvanian and Permian Rocks in the Parry Islands Group, Canadian Arctic Archipelago. Geo- logical Survey of Canada Paper, 65-1:9-12. Ruzhencev, V. E. 1952. Biostratigraphy of the Sakmar Stage in the Aktyu- bin Region of the Kazakhstan SSR. Akademii Nauk SSSR, Paleontologicheskogo Instituta Trudy, 42: 1-87, plates 1-6. 1956. Lower Permian Ammonites of the Southern Urals. II. Ammonites of the Artinskian Stage. Akademii Nauk SSSR, Paleontologicheskogo Instituta Trudy, 60:1-275, plates 1-39. Solomina, R. V. 1960. Some Permian Brachiopods from Pai-Hoi. Nauchno- Issledovatelskii Institut Geologii Arktiki, Sbornik Statei po Paleontologii Biostratigrafii, 19:24-73, plates 1-12. NUMBER 3 361 Sowerby, G. 1844. Paleozoic Shells from Van Diemen's Land. In C. Darwin, Geological Observations on the Volcanic Islands Visited During the Voyage of HMS Beagle, pages 158-160. London: Smith, Elder and Company. Stehli, F. G. 1954. Lower Leonardian Brachiopods of the Sierra Diablo. American Museum of Natural History Bulletin, 105:257-358, plates, 17-27. Waterhouse, J. B. 1964a. Permian Brachiopods of New Zealand. New Zea- land Geological Survey Paleontological Bulletin, 35:1-289, plates 1-35. 1964b. The Permian of New Zealand. Twenty-Second In- ternational Geological Congress, India, Abstracts, page 142. 1967a. Upper Permian (Tatarian) Brachiopods from New Zealand. New Zealand Journal of Geology and Geophysics, 10(1): 74-118, figures 1-47. 1967b. Proposal of Series and Stages for the Permian of New Zealand. Transactions of the Royal Society of New Zealand, Geology, 5(6) : 161-180. 1967c. Cool-water Faunas from the Permian of the Canadian Arctic. Nature, 216(5110) :47—49, fig- ures 1, 2. 1969. The Paleoclimatic Significance of Permian Pro- ductacea from Queensland, pages 226-235, figures 40-43. In K. S. W. Campbell (editor), Palaeon- tology and Stratigraphy Essays in Honor of Dorothy Hill. Garner L. Wilde Phylogeny of Pseudofus ulinella and Its Bearing on Early Permian Stratigrap hy ABSTRACT The fusulinid genus Pseudofusulinella Thompson is represented in the McCloud Limestone of northern California by 40 species that can be separated into four phylogenetic (phenetic) groups: Group I, Elongate- fusiform species with perched chomata; Group II, Small, rare, very slender-fusiform species with perched chomata; Group III, Intermediate, thickly fusiform species with perched chomata; and Group IV, Large, thickly fusiform species with massive chomata. Recog- nition of these four groups within Permian (Wolfcam- pian) rocks of the McCloud Limestone provides a basis for long-range correlation of significant strati- graphic and physical breaks, not only throughout the Pacific Northwest, but with important exposures in Texas and Kansas. Systematic paleontologists have, until only recently, confined their efforts largely to the description of new genera and species, and to the erection of ontogenetic and phylogenetic schemes for classification purposes. Much of this kind of descriptive paleontology still re- mains to be done despite the hue and cry from those who feel that too many species have already been named. Even so, criticism is certainly warranted unless every effort is made to apply systematics and phylogen- esis to stratigraphic problem solving. It is the purpose of this paper to deal with that objective. Garner L. Wilde, Humble Oil and Refining Company, P.O. Box 120, Denver, Colorado 80201. Published with permission of Humble Oil and Refining Company. Description of numerous species of the fusulinid genus, Pseudofusulinella Thompson (1951) from the McCloud Limestone of northern California by Skinner and Wilde (1965b), provides an excellent starting point for dealing with phylogeny and its applicability to Early Permian stratigraphic problems. It is generally believed that the Early Permian (Wolfcampian) experienced periods of crustal insta- bility, accompanied by widespread transgressions and regressions of the seas. In many areas around the world unconformities are apparent in Wolfcampian se- quences. Because of dense drilling and excellent ex- posures in surrounding mountainous outcrops, the Greater Permian Basin of western Texas and south- eastern New Mexico offers some of the best documented evidence of widespread unconformable relationships in the Wolfcampian. Coincidental relationships between these physical and faunal breaks have also been observed in the phy- logenetic history of 40 species of Pseudofusulinella, all of which occur in the McCloud Limestone of the Shasta Lake Region of northern California. Two of the species occur in Leonardian strata. Recently, Ozawa (1967) has also dealt with the phy- logenetic relationships within Pseudofusulinella. Ozawa's paper was not yet available to the present writer in early 1968 when some of the material pre- sented here was given in a paper read before the Permian Basin Section of the Society of Economic Paleontologists and Mineralogists in Midland, Texas. In the present study, therefore, Ozawa's contribution is considered. 363 364 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Fusulinid Zonation of the McCloud Limestone Skinner and Wilde (1965b) divided the McCloud Limestone into eight fusulinid faunal zones with in- formal letter designations for each zone. Zone A, the oldest, was only tentatively considered to be Permian, Zones B through G were identified as Wolfcampian and Zone H, Leonardian. It is believed now that Zone A is definitely Permian (Wolfcampian) in age. This age assignment also is important to the thesis developed in the present paper. Hoare (1963) has described a fusulinid fauna from the Sunflower Reservoir area of northern Nevada which contains Zone A elements, but which also contains species of Schwagerina. This genus is absent in Zone A of the McCloud. Along with. Schwagerina, Hoare described Pseudofusulinella sym- metrica Hoare and Pseudofusulinella nevadensis Hoare. Both of these species obviously belong to the genus Thompsonella Skinner and Wilde, which is common in Zone A. Indeed, later study might prove that identical species occur in the two localities. Species of Triticites described by Hoare are not incompatible with those of Zone A of the McCloud. Interbedded dark brown to black, silty, siliceous lime- stone and dark gray siltstone and silty shale character- istic of Zone A of the McCloud is in rather sharp contrast to most of the overlying sequences of rocks which are generally light-colored, thick-bedded, and cherty limestones. Accompanying these lithologic differ- ences is a distinct faunal contrast (Figure 1). From Zone A nine species of Triticites have been described along with twelve species of Pseudufusulinella, two species of the new genus Thompsonella, and rare un- described forms of Schubertella. The typical Early Permian genus, Schwagerina, has not been seen in Zone A although a few of the species of Triticites ex- hibit Schwagerina features. Triticites has not been found higher. In the next higher Zone B only three species of Pseudofusulinella are present, accompanied by four species of Schwagerina, three species of Pseudofusulina (Rugosofusulina of earlier authors), and a single spe- cies of Paraschwagerina. None of the latter three gen- era have been seen in Zone A. Again, rare specimens of Schubertella are present in Zone B, but apparently are different species. An apparent angular discordance between Zones A and B is seen in at least one large exposure on the west- ern face of Ellery Mountain. It is suggested that this break marks a regionally important unconformity which will be considered in greater detail later. Zones B and C of the McCloud Limestone are es- sentially identical lithologically, and six species of fusulinids are common to the two zones. In Zone C, however, the typical Middle Wolfcampian genus, Pseu- doschwagerina, makes its first appearance. It is repre- Correlation of Related Localities Series Genus No. I 2 3 4 5 6 7 * Indicates some overlapping of identical species. 8 10 12 m o> c £- OI ,ros o Ore ^ _ c tra l a > a> a r n\ WIDTHS 2 3 ?V^V^V\/V^ u_ o 4 W W Y \v Ul u Z < 5 - FIGURE 6.—Range of width in species of Pseudofusulinella from McCloud Limestone. SPECIES NUMBER 1 2 3 A 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 FIGURE 7.—Form ratio L/W in species of Pseudofusulinella from McCloud Limestone. widths (Figure 6); consequently, form ratios became smaller (Figure 7). Other measurable features have been plotted graphi- cally, but they seem less meaningful; therefore, those plots are not included here. Generally, however, the spirotheca became thicker in time, with the one ex- ception of Pseudofusulinella tumida in Zone D which has the thickest spirotheca of all species. The changes in proloculus diameter are interesting in that species which generally have the larger proloculi occur in Zone A and then again in Zones F, G and H. The smaller proloculi occur in the middle zones. Reasons for this are not clear unless it is related to the greater numbers of species in the lowest and highest zones coupled with the above-mentioned fact that species generally are larger higher in the section. Finally, there seemingly was a tendency for species to develop greater numbers of septa per whorl through time al- though this was not a consistent trend. Phenetic Versus Phylogenetic Classification The writer is well aware of the supposed controversy taking place today between so-called "classical taxon- omists" and the numerical taxonomists. Cain and Harrison (1960) have defined a phenetic classification as an "arrangement by overall similarity, based on all available characters without any weighting." These authors contrast phenetic with phyletic, a classification supposedly designed to show the course of evolution with weighting of characters implied. Actually, weighting is used by both the "classical" and the numerical taxonomist. The former selects, from experience, the most obviously meaningful characters available with which to tell his story. Weighting, then, takes place at the outset, but the characters chosen are usually given equal weighting. The numerical taxon- omist selects a group of characters with which to work also, purely for practical reasons, because in the fast approaching day of the optical scanner selections must be made to cope with the infinite number of characters available. The numerical taxonomist also gives equal weighting to the characters chosen. The great value in his approach, however, lies in the fact that the char- acters receive a more exacting treatment, so that the subjective, or human element is very nearly eliminated. 370 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY It is absolutely impossible to define species among extinct fossil groups on the same basis as with living forms; that is, in the sense of interbreeding populations. Good approximations are possible, but that is all. Blackwelder (1964, p. 22) has pointed out that "one of the tasks of taxonomy is to identify specimens, and identification consists of determining the correct place of the specimen in the prior classification by means of its comparative attributes. A reproductive definition of a species would nearly always make identification impossible. . . Our entire taxonomic system is based on the assumption that species can be segregated and classified on the basis of the attributes of the in- dividuals. No other basis for species can be mixed into that system, and all efforts so far have been futile." Data presented in the present study are not intended to represent proof of phylogenetic relationships in Pseudofusulinella. After all is said, species, or phylo- genetic schemes, are never sacred, they are man-made. Homo sapiens was determined by Homo sapiens, and this might be the strongest argument for separating ourselves from other, earlier humanoids! The best, still, that can be suggested is that a species is what a competent taxonomist decides to call a species. The "phylogenetic" relationships shown in the pres- ent paper might just as easily be called "phenetic," but the important point for consideration is not what something is called, but what one does with the information. Correlation of McCloud Events With Other Impor- tant Wolfcampian Sequences in the United States Attempts to correlate breaks within the stratigraphic record should be accompanied always with a great deal of caution. It should be emphasized here that cata- strophic diastrophism is not intended to explain the seemingly widespread nature of the stratigraphic breaks noted in the McCloud. Possibly none of the correlative events occurred at precisely the same time; and, in many areas, certain fusulinid species continued living much as they had before. But the end effect of major tectonic disturbances which, no doubt, occurred over a relatively short span of geologic time, left un- deniable evidence in the sediments and the fossils. Correlative Exposures in the Pacific Northwest Exposures of Permian limestone containing fusulinid faunas identical with portions of the McCloud Lime- stone (Figure 1) have been reported by Skinner and Wilde (1966a) from near Quinn River Crossing in northwestern Nevada, the Coyote Butte Limestone of the Suplee area, east-central Oregon, (1966b) and from exposures near Kettle Falls and Republic, in northeastern Washington (1966c). Exposures near Quinn River Crossing correlate with Zones E, F, and G of the McCloud. Base of the Quinn River sequence corresponds roughly with a suggested stratigraphic break between Zones D and E in the McCloud, and its top seemingly corresponds to the Wolfcampian-Leo- nardian boundary of the McCloud section. Exposures near Kettle Falls and Republic in north- eastern Washington are actually large limestone pods or lenses within the Permian Mission Argillite; con- sequently, tops and bottoms may mean very little. The contained fusulinids indicate that these exposures correlate roughly with Zones F and G of the McCloud of northern California. The base of these exposures does not seem to correspond with any of the suggested major breaks of the McCloud sequence, but the top appears to be related to the Wolfcampian-Leonardian boundary. Exposures of the Coyote Butte Limestone of east- central Oregon contain fusulinids that correlate with Zone G of the McCloud. Thus, the base corresponds roughly with a major stratigraphic break between Zones F and G, and the top apparently is close to the Wolfcampian-Leonardian boundary. Here, then, one sees some evidence of widespread stratigraphic and physical breaks in Early Permian strata over portions of the Pacific Northwest. Can such breaks be recognized in more remote regions where fusulinid faunas are quite different, especially at the specific level? Glass Mountains, Texas Because of the evidence that has been accumulated in recent years by Dunbar and Skinner (1937), Ross (1959, 1963), and Wilde (1962), rather precise age dating of the type Wolfcampian of the Glass Moun- tains of Texas is now possible. It is here also that im- portant stratigraphic breaks have been determined. Ross (1959, 1963) divided the Wolfcampian into a lower Neal Ranch Formation separated by uncon- formities at its base and top, and an upper Lenox Hills Formation, which he considered to lie also in uncon- formable relationship to the Leonardian rocks above (Figure 8). NUMBER 3 371 McCLOUD Ls. FUSULINID ZONES GLASS MTNS, WEST TEXAS DIABLO PLATFOR M, WEST TEXAS NORTH- CENT RAL TEXAS KANSAS SERIES SKINNER RANCH Fm. (Res.) ALACRAN MTN. Fm. (Upper) CLYDE Fm NIPPEWALLA Gr STONE CORRAL Fm. SUMNER GROUP LENOX HILLS Fm. ALACRAN MTN. Fm. (Lower) ADMIRAL Fm Basal Lenox Hills Congl. NEAL RANCH Fm. CERRO ALTO- HUECO CANYON Fms. MORAN- PUTNAM Fms CHASE- COUNCIL GROVE GROUPS King's Bed 3 ? Powwow Conglomerate - s e e t e x t ^ Salt Creek Bend Shale Eskridge Shale KING'S BED 2 of "GRAY LIMESTONE" BURSUM Fm. PUEBLO Fm. ADMIRE GROUP FIGURE 8.—Correlation of McCloud fusulinid zones with other lower Permian sequences. In recognizing these breaks, Ross placed Bed 2 of the "Gray Limestone" of King (1930) in the under- lying Gaptank Formation of Pennsylvanian age, al- though he recognized an unconformity also at the base of the "Gray Limestone." The writer has observed these same unconformable relationships; but because of the presence of Wolfcampian species of Triticites and Schwagerina in Bed 2, a Wolfcampian age assign- ment for the unit seems most logical. Apparently Ross had none of these in his collections. Bed 2 of the "Gray Limestone" is considered at present to correlate with some portion of Zone A of the McCloud Limestone. It is of interest to note that undescribed species of Pseudofusulinella have been found in the "Uddenites zone" of the underlying Gap- tank Formation at Wolf Camp, but none have been found in Bed 2 of the overlying "Gray Limestone." The "Uddenites zone" is of Pennsylvanian (Virgilian) age; and species of Pseudofusulinella are known in the Great Basin Region of the western United States in beds of the same age. The Neal Ranch Formation correlates roughly with Zones B, C, and D of the McCloud sequence. Interest- ing similarities and differences may be noted among the fusulinids of the two sequences. In the McCloud sequence, Triticites is not known above Zone A, but species of the genus are present in the Neal Ranch of the Glass Mountains. Similarly, Pseudofusulinella is abundant in McCloud Zones B, C, and D, but it has not been found in the Neal Ranch. In fact, Pseudo- fusulinella has not been found above the "Uddenites zone" of the Gaptank Formation in the Glass Moun- tains. Genera in common, however, are numerous. Species of Schwagerina, Pseudofusulina, Paraschwa- gerina, Pseudoschwagerina, and Schubertella occur in both, and a number of similar species may be noted in Table 1. Correlation of the Lenox Hills Formation with Zones E and F of the McCloud Limestone is also judged to be quite acceptable, although certain local problems arise which need further explanation. The Lenox Hills Formation was referred to as constituting the "assemblage-zone of Monodiexodina linearis" by Wilde (1962) because of the widespread occurrence of the species in rocks of an equivalent age elsewhere as well as its dominance in the Lenox Hills of the Glass Mountains. 372 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY TABLE 1.— Similar species in Neal Ranch Formatio n and McCloud %ones B, C, D. Neal Ranch Formation McCloud Zones B, C, D Paraschwagerina gigantea Dunbar and Skinner Pseudoschwagerina uddeni Dunbar and Skinner Pseudofusulina huecoensis Dunbar and Skinner Schwagerina pugunculus Ross Paraschwagerina magna Skinner and Wilde Pseudoschwagerina calijornica Skinner and Wilde Pseudofusulina soluta Skinner and Wilde Schwagerina acuminata Skinner and Wilde Monodiexodina has not been discovered in the Mc- Cloud Limestone, but a very closely related genus, Eoparafusulina, is represented by twenty species, and these species occur only in Zones E and F. Zone E con- contains five species and Zone F contains fifteen species.1 The main zone fossils of the Lenox Hills Formation, other than Monodiexodina linearis (Dunbar and Skin- ner), are Pseudoschwagerina gerontica Dunbar and Skinner, Pseudoschwagerina convexa Thompson, Schwagerina laxissima Dunbar and Skinner, Schwa- gerina diversiformis Dunbar and Skinner, and Schwagerina nelsoni Dunbar and Skinner. The latter species more rightly belongs in the genus Chalaro- schwagerina Skinner and Wilde, and is here so placed. Schubertella kingi Dunbar and Skinner ranges throughout the Wolfcampian of the Glass Mountains and, therefore, is not confined to the Lenox Hills Formation. It occurs however, only in Zone E of the McCloud of California. As mentioned previously, Triticites does not occur higher than Zone A of the McCloud sequence. Numerous abraded specimens of Triticites have been found in Lenox Hills strata, but in all known instances the specimens appear to have been reworked from Neal Ranch and older beds. By reference to some of the fusulinids listed above, comparisons are shown for correlation purposes in Table 2. The question arises logically as to whether Zone G of the McCloud Limestone, the highest Wolfcampian of the sequence, has an equivalent counterpart in the Glass Mountains. Admittedly, the correlation might be considered rather hazardous, but it is suggested that the Decie Ranch Member of the Skinner Ranch For- 1 The species described recently by Ross (1967) as Eopara- fusulina allisonensis Ross from the Neal Ranch Formation near Gaptank, eastern Glass Mountains, does not belong in that genus, but in Alaskanella Skinner and Wilde, which Ross considered a synonym of Eoparafusulina. mation (Cooper and Grant, 1964) correlates with at least a portion of Zone G of the McCloud Limestone. Wilde (1962) considered the sequence now known as the Decie Ranch Member at Dugout Mountain in the western Glass Mountains to belong in the Wolf- campian rather than the Leonardian. The Decie Ranch Member is dominated by the following species of fusulinids: Schwagerina hessensis Dunbar and Skinner, Schwagerina hawkinsi Dunbar and Skinner, and Schwagerina dugoutensis Ross. The latter is quite similar to Schwagerina amoena Skinner and Wilde and Schwagerina eximia Skinner and Wilde of the McCloud. Schwagerina hawkinsi is a transitional species between Schwagerina and Chalaroschwagerina and it compares favorably with Schwagerina corpu- lenta Skinner and Wilde as well as Chalaro schwagerina tumentis Skinner and Wilde both of which occur in Zone G of the McCloud. Monodiexodina linearis (Dunbar and Skinner) has been cited also as occurring in the Decie Ranch Mem- ber, as well as higher in the Leonard (Ross, 1962). The writer has made similar observations, although re- working and faulting leave doubts about some occur- rences. One would not expect the highest collection reported by Ross to be indicative of an indigenous occurrence, but presence of the species at a number of localities in the Decie Ranch Member only strength- ens the argument that this member is of Wolfcampian age. Diablo Platform, Texas For purposes of discussion, Diablo Platform is used here to refer to the area of the Sierra Diablo, Hueco, and Franklin Mountains of western Texas. Thompson (1954) and Williams (1963) have discussed the Wolf- campian stratigraphy of the Hueco Mountains and they have described the fusulinid faunas. Wilde (1962) discussed the Permian stratigraphy of the Sierra Diablo and suggested correlations with the Glass Mountains NUMBER 3 373 TABLE 2.—Similar species in Lenox Hills Formation and McCloud %ones E and F. Lenox Hills Formation McCloud Zones E and F Monodiexodina linearis (Dunbar and Skinner) Pseudoschwagerina gerontica Dunbar and Skinner Chalaroschwagerina nelsoni (Dunbar and Skinner) Eoparafusulina thompsoni Skinner and Wilde Pseudoschwagerina robusta (Meek) Chalaroschwagerina obesa Skinner and Wilde sequence on the basis of fusulinid assemblage zones. Stewart (1963) described some of the fusulinids in the Sierra Diablo section. Dunbar and Skinner (1937) de- scribed fusulinids from the Sierra Diablo, Hueco, and Franklin Mountains; and Williams (1966) recently discussed the stratigraphy and described fusulinids from the Franklin Mountains Wolfcampian section. The oldest Permian rocks of the Diablo Platform are exposed in the Hueco Mountains and have been referred to the Bursum Formation, described originally by Wilpolt, McAlpin, Bates, and Vorbe (1946) from exposures in the Hansonburg Hills, Socorro County, New Mexico. At the type locality the Bursum repre- sents a zone of transition between dominantly marine Pennsylvanian strata and predominantly continental redbeds of Permian age (Bachman, 1968). Beds oc- cupying this same faunal and similar lithologic transi- tional position between the Pennsylvanian and Per- mian of other areas in New Mexico have been given local formation names. For example, Otte (1959) gave the name Laborcita Formation to such beds in the northern Sacramento Mountains; and Stark and Dap- ples (1946) introduced the name Aqua Torres Forma- tion for a similar unit in the Los Pinos Mountains. Thompson (1954) reported a number of species from the Bursum at several localities in New Mexico. From Abo Canyon he cited Triticities (Leptotriticites) eoextentus (Thompson), Triticities creekensis Thomp- son, Schwagerina pinosensis Thompson, and Schwa- gerina grandensis Thompson[?]. All but S. pinosensis were reported from the Oscura Mountains. From the Robledo Mountains, Thompson also reported Pseudo- fusulina robleda Thompson, Leptotriticities aff. glen- ensis (Thompson), and L. hughesensis (Thompson). In the Hueco Mountains, the Bursum beds contain Triticities cellamagnus Thompson and Bissell and Schwagerina grandensis Thompson[?] (Thompson, 1954, Williams 1963). From the Laborcita Formation of the Sacramento Mountains, Steiner and Williams (1968) have recently identified Triticites creekensis Thompson, T. ventricosus (Meek and Hayden), T. (Leptotriticites) americdnus (Thompson), Schwa- gerina campensis Thompson, and Schwagerina ema- ciata (Beede). All of these species are widespread in Early Wolfcampian rocks from many areas of the United States. Already mentioned is the fact that Pseudofusulinella is not known in Permian strata of the southwestern and mid-continental United States. Leptotriticites Skinner and Wilde (1965a), fulfills a role in the Permian of these mentioned regions similar to that of Pseudofusu- linella in the western provinces; that is, the two ap- parently occur exclusive of each other throughout the Wolfcampian in their respective areas. Species of Triticites in the Bursum equivalents are well-developed members of the genus, but they have not attained the evolutionary development of some of the species of Zone A of the McCloud Limestone, which, as mentioned earlier, exhibit features transi- tional between Triticites and Schwagerina. Simple spe- cies of Schwagerina do occur in Bursum equivalents, so that the conclusion seems justified that Bursum beds and Zone A of the McCloud are at least approximately equivalent. Williams (1963) has given formational status to the Lower, Middle and Upper divisions of the Hueco Limestone of the Hueco Mountains, recognized origi- nally as mappable units by King, King, and Knight (1945). In ascending order these are the Hueco Can- yon, Cerro Alto, and Alacran Mountain Formations. In the same paper, Williams described the fusulinid faunas of the Hueco Group. Later Williams (1966) recognized the same threefold separation of the Hueco Group in the Franklin Mountains and described the fusulinids. The lowest of the three formations in the Hueco Mountains, Hueco Canyon, is separated from the Bur- sum beds by an unconformity. This surface of un- conformity is marked by the Powwow Member of the 374 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Hueco Canyon Formation, a basal conglomerate and red-bed sequence. The Powwow is widespread at the surface on the Diablo Platform, and is recognized in the subsurface, especially over major regional highs such as the Central Basin Platform, which separates the Midland and Delaware basins. By its very nature, the Powwow Member varies a great deal in thickness, lithology, and stratigraphic relationship to underlying Bursum equivalents (Thompson, 1954). In the vicinity of the Hueco Mountains, the Powwow surface of un- conformity truncates older and older beds from north to south. Williams (1963) reports the presence of beds containing a Pseudoschwagerina fauna beneath the uncomformity at one locality in the Hueco Mountains. In the Sierra Diablo the Powwow at the base of the Hueco Group is about the same age as the basal Lenox Hills conglomerate in the Glass Mountains (Wilde, 1962). This fact prompted Wilde to suggest an ap- proximate time equivalency for the two clastic units, while at the same time recognizing differences in age from place to place. On the accompanying correlation chart (Figure 8), the time-transgressive character of the Powwow is not indicated, but the Powwow is shown in the correct 'position of its type section in Powwow Canyon, Hueco Mountains. Fusulinids from the Powwow Member were de- scribed first by Dunbar and Skinner (1937), and more recently by Williams (1963). Williams identified Schwagerina emaciata (Beede), Schwagerina bellula Dunbar and Skinner, and Triticites powwowensis Dunbar and Skinner. The latter is considered a primi- tive species of Schwagerina by the writer, and is not greatly different from transitional forms that occur in lower, McCloud beds in California. It occurs also in Bed 8 (King, 1930) of the type section of the Neal Ranch Formation in the Glass Mountains. At least four other well-known species of fusulinids occur in the Hueco Canyon Formation which also occur in the Neal Ranch Formation of the Glass Mountains. These are Pseudoschwagerina uddeni (Beede and Kniker), P. beedei Dunbar and Skinner, P. texana Dunbar and Skinner, and Pseudofusulina huecoensis (Dunbar and Skinner). Correlation of the Neal Ranch fauna with Zones B, C, and D of the McCloud has already been discussed. Further evidence for correlation of the Hueco Can- yon Formation with the above-mentioned zones of the McCloud is shown in Table 3. In the present paper the overlying Cerro Alto Lime- stone of the Hueco Mountains is correlated also with some portion of McCloud Zones B, C, and D. It is recognized, however, that there are difficulties in mak- ing such a correlation. Only two species of fusulinids, Schwagerina eolata Thompson, and S. neolata Thomp- son, occur in the Cerro Alto, and direct comparisons cannot be made with accuracy to other described species. On the other hand, the Cerro Alto is separated from most of the overlying Alacran Mountain Forma- tion by a prominent series of red beds, the Deer Moun- tain Member of the Alacran Mountain. Intrusion of this lithic unit across older beds of normal marine deposition denotes a major regression. Above the Deer Mountain Member the Alacran Mountain contains a fine Lenox Hills fusulinid fauna. It seems obvious, therefore, that the Deer Mountain marks the strati- graphic break between the Neal Ranch and Lenox Hills formations of the Glass Mountains, and, in turn, be- tween Zones D and E of the McCloud Limestone of northern California. Species very similar to, if not identical with Schwagerina neolata and S. eolata oc- cur in King's (1930) Beds 8 and 15, respectively, in the type Neal Ranch Formation of the Glass Mountains. All of the main zone species of the Lenox Hills ex- cept Monodiexodina linearis (Dunbar and Skinner) occur in the Alacran Mountain; Pseudoschwagerina gerontica Dunbar and Skinner, P. convexa Thompson, Chalaroschwagerina nelsoni (Dunbar and Skinner), and Schwagerina diversiformis Dunbar and Skinner. It has been stated earlier that the formations de- scribed in the Hueco Mountains have been recognized also in the Franklin Mountains (Williams, 1966). The fusulinid faunas are also the same. Similarly, the fusulinid faunas are known in the Sierra Diablo TABLE 3.—Similar species in Hueco Canyon Formation and McCloud ^pnes B, C, D. Hueco Canyon Formation McCloud Zones B, C, and D Schwagerina bellula Dunbar and Skinner Schwagerina knighti Dunbar and Skinner Paraschwagerina shuleri Williams Schwagerina pseudoprinceps Skinner and Wilde Schwagerina pseudoprinceps Skinner and Wilde Paraschwagerina fairbanksi Skinner and Wilde NUMBER 3 375 (Wilde, 1962), but the formational units have not been separated. It is probable, however, that only Alacran Mountain equivalents are present in the Sierra Diablo. North-Central Texas Many attempts to correlate rock units of the north- central Texas Wolfcampian with those of other areas have generally proved to be unsatisfactory insofar as fusulinid faunas are concerned. The conclusions pre- sented here, for example, are at variance with those of Ross (1963, p. 43), his work being about the most recent to deal with the problem. As regards the Bursum equivalents in north-central Texas sequence, the writer generally agrees with Ross that the Pueblo Formation is approximately coeval. The presence of Triticites creekensis Thompson, T. ventricosus (Meek and Hayden), and Schwagerina campensis Thompson, in the Pueblo Formation, along with numerous species of Leptotriticites, is reminiscent of the Laborcita fauna (Williams, 1966) mentioned previously. Top of the Pueblo Formation as defined does not seem to be the most logical place to draw the boundary between Bursum and McCloud Zone A events on a purely lithologic basis. A prominent sandstone overlain by an equally prominent red-shale sequence some 25 to 50 feet below the top may mark this boundary. These beds comprise the upper half of the Salt Creek Bend Shale Member. Comparison of the Pueblo fauna and that of Zone A is difficult. It is true, however, that both contain a dominant Triticites fauna. Moreover, the few species of Schwagerina in the Pueblo Forma- tion are quite simple and not highly developed. Ross (1963) believed that the Neal Ranch was rep- resented in north-central Texas by a hiatus. This is difficult to reconcile with the fusulinid evidence. The Gouldbusk Limestone Member of the Moran Forma- tion contains Pseudoschwagerina texana Dunbar and Skinner, Schwagerina complexa Thompson, Pseudo- fusulina[?] moranensis Thompson,2 and other unde- 2 A number of species referred to Pseudofusulina Dunbar and Skinner by Thompson (1954) obviously do not belong to that genus. Skinner and Wilde (1965b, 1966d) have offered conclusive evidence that Pseudofusulina sensu stricto has it rugose spirotheca, and is a senior synonym of Rugoso- fusulina Rauser. The species described by Thompson (1954) as Pseudofusulina possess certain unique characters which place them between typical Schwagerina Moeller and Pseudo- scribed species. The two first-named species are com- mon to abundant in King's beds 13 to 15 of the type Neal Ranch in the Glass Mountains, and Pseudo- schwagerina texana occurs in the Hueco Canyon Formation of the Hueco Mountains. The Putnam Formation, which overlies the Moran Formation in the north-central Texas sequence con- tains few species of fusulinids. In the top of the Put- nam, however, the Coleman Junction Limestone Mem- ber contains Schwagerina colemani Thompson, a spe- cies comparable in development to Schwagerina ema- ciata (Beede). Fusulinids have not been reported from the Admiral Formation of the north-central Texas section. Recently Myers (1968) described Schwagerina crassitectoria Dunbar and Skinner and S. guembeli Dunbar and Skinner from the Leonardian Clyde Formation, which is separated from the Admiral by the Belle Plains Formation. Because the Schwagerina crassitectoria fauna marks an important zone near the base of the Leonardian over a wide area (Figure 8), it seems en- tirely possible that the Admiral is equivalent to Zones E and F of the McCloud (Lenox Hills, Alacran Moun- tain) ; and that the Belle Plains occupies the position of McCloud Zone G (Decie Ranch). This view is similar to that of Roth (in Dunbar et al., 1960, p. 1788) who believes that the Wolfcampian-Leonardian boundary falls within the Belle Plains Formation rather than at its base. Kansas Section Very little difference exists between the eastern Kansas Wolfcampian and that of north-central Texas. Identi- cal species of fusulinids occur throughout, and the cy- clic sedimentational pattern is similar. Moore (1936) placed the boundary between the Pennsylvanian and Permian of Kansas just above the schwagerina Dunbar and Skinner. Low, rounded septal folds, tightly coiled juvenaria followed by some inflation of outer whorls, and chomata in the two inner volutions are similar to Pseudoschwagerina texana Dunbar and Skinner. Lack of well-defined inflation and thickness of spirotheca are more like Schwagerina. Such forms are hereby assigned a new generic name, Stewartina, in honor of Wendell J. Stewart of Midland, Texas. Type species is chosen as Pseudofusulina[?] moranensis Thompson, 1954 (p. 69, pi. 39: figs. 1-7; pi. 40: figs. 1-9). Other species assigned here to the new genus, Stewartina, are Pseudofusulina loringi Thompson (1954) Pseudofusulina robleda Thompson (1954), Schwagerina laxissima Dunbar and Skinner (1937). 372-386 0—71- -26 376 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY ZONES PLATE 1.—Conjectured evolutionary trends in species of Pseudofusulinella in McCloud Lime- stone, Northern California. Group I: elongate-fusiform species with perched chomata; range, Zones A-D. Group II: small, rare, very slender-fusiform species with perched chomata; range, Zones A-F. Group III: intermediate, thickly fusiform species with perched chomata; range, Zones A-F. Group IV: large, thickly fusiform species with massive chomata; range, Zones A-H. Species number 22, P. occidentalis, and number 30, P. montis, were described by Thompson, Wheeler, and Hazard (1946) ; all others shown here were described by Skinner and Wilde (1965). Brownville Limestone. The choice was somewhat ar- bitrary inasmuch as channel sandstones mark discon- formities above and below the Brownville. Recently, however, Douglass (1962) has described Triticites (Leptotriticites) brownvillensis (Douglass) from the Brownville. Accumulated evidence (Skinner and Wilde, 1965) indicates that Leptotriticites is appar- ently restricted to the Wolfcampian. It, therefore, seems best that the boundary be drawn below the Brownville. A somewhat similar situation exists in the Texas sequence. The Belknap Limestone of the Brazos River section, heretofore considered to be of Pennsyl- vanian age, contains Leptotriticites extentus (Thomp- son) and L. eoextentus (Thompson) and Schubertella kingi Dunbar and Skinner (Kauffman and Roth, 1966). This fauna is also present in the "Waldrip No. 1" of the Colorado River section. The Belknap and "Waldrip No. 1" are judged to be equivalent, and Wolfcampian rather than Pennsylvanian in age. Strictly on the basis of fusulinid occurrences one is rather moved to carry the top of McCloud Zone A events in the Kansas section to the base of the Eskridge shale in the middle of the Council Grove Group. This has the immediate advantage of correlating their con- NUMBER 3 377 taining beds with widely occurring species of fusulinids. Leptotriticites americanus (Thompson) occurs in the Americus Limestone Member of the Foraker Lime- stone (Council Grove Group) of Kansas, and in the Laborcita Formation of New Mexico. Leptotriticites eoextentus (Thompson) occurs in the Americus, the Bursum Formation of New Mexico, and in "Waldrip No. 1" of the Pueblo Formation of Texas. Leptotri- ticites hughesensis (Thompson) is found in the Hughes Creek Shale Member of the Foraker of Kansas and in the Bursum Formation of New Mexico. Triticites meeki (Moeller) occurs in the Foraker of Kansas and Pueblo Formation of Texas. Triticites ventricosus (Meek and Hayden) occurs in the Laborcita of New Mexico, Foraker of Kansas, and Pueblo of Texas. Tri- ticites creekensis (Thompson) is found in the Laborcita and Bursum of New Mexico, and the Pueblo of Texas. Schwagerina longissimoidea (Beede) occurs in the lower half of the Council Grove of Kansas and the Pueblo of Texas. All of the above mentioned have been found in other localities, but only those which relate to the present discussion are given here. One of the immediate disadvantages in correlating McCloud Zone A with beds in Kansas all the way to the base of the Eskridge shale is that Paraschwagerina kansasensis (Beede and Kniker) occurs in the Neva Limestone just beneath the Eskridge. Occurring with P. kansasensis are three advanced species of Leptotri- ticites and Schwagerina longissimoidea (Beade). P. kansasensis occurs in King's Beds 4 and 5 of the type section of the Neal Ranch Formation of the Glass Mountains, Texas. Bed 4 of King's Section 24 at Wolf- camp (King, 1930, p. 55) lies only eight feet above the "Lower Gray Limestone," or Bed 2. So the correla- tion proposed here is probably not too far out of line with the true facts of the matter. Greater precision than this would not be expected on such a regional scale. Remainder of the Council Grove and overlying Chase Group in the Kansas sequence correlates easily with the north-central Texas Moran and Putnam for- mations, and by inference, with Zones B, C, and D of the McCloud of California. In fact, there seems to be a very close correlation between the Florence Lime- stone of Kansas and Gouldbusk of Texas. Both units contain Pseudoschwagerina texana Dunbar and Skin- ner and Stewartina moranensis (Thompson). McCloud Zones E, F (Lenox Hills), and G (Decie Ranch) apparently are missing from the Kansas sec- tion, although faunal evidence is lacking, so that it is difficult to be certain. It is possible that practically all of the overlying Sumner Group, now considered to be of Leonardian age, is actually Late Wolfcampian. Mer- riam (1963) reports a well-recognizable unconformity at the base of the Stone Corral Formation, the top unit of the Sumner Group. This unconformity could actually represent the Wolfcampian-Leonardian boun- dary. If most of the Sumner Group actually is Wolf- campian rather than Leonardian in age, this would have important implications for correlation of many Early Permian sequences throughout the mid-con- tinent and eastern Rocky Mountain regions. Literature Cited Bachman, G. O. 1968. Geology of the Mockingbird Gap Quadrangle, Lincoln and Socorro Counties, New Mexico. United States Geological Survey Professional Paper, 594- J:J1-J43. Blackwelder, R. E. 1964. Phyletic and Phenetic versus Omnispective Classifi- cation. In Phenetic and Phylogenetic Classification. Systematics Association Publication, 6:17—28. Cain, A. J., and G. A. Harrison 1960. Phyletic Weighting. Proceedings of the Zoological Society of London, 135:1—31. Cooper, G. A., and R. E. Grant 1964. New Permian Stratigraphic Units in Glass Moun- tains, West Texas. Bulletin of the American Associa- tion of Petroleum Geologists, 48:1581-1588. Douglass, R. C. 1962. In M. R. Mudge and E. L. Yochelson, Stratigraphy and Paleontology of the Uppermost Pennsylvanian and Lowermost Permian Rocks in Kansas. United States Geological Survey Professional Paper, 323 (part 2, Paleontology) : 55-64, plates 6-10. Dunbar, C. O., et al. 1960. Correlation of the Permian Formations of North America. Bulletin of the Geological Society of America, 71:1763-1806, chart 7. Dunbar, C. O., and J. W. Skinner 1937. Permian Fusulinidae of Texas. In The Geology of Texas, volume 3, part 2. University of Texas Bulle- tin, 3701:517-825, plates 42-81. Hoare, R. D. 1963. Permian Fusulinids from the Sunflower Reservoir Area of Northern Nevada. Journal of Paleontology, 37:1143-1149, plates 153-154. Kauffman, A. E., and R. I. Roth 1966. Upper Pennsylvanian and Lower Permian Fusu- linids from North-Central Texas: Cushman Foundation for Foraminiferal Research Special Publication, 8:1-49, 11 plates. 378 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY King, P. B. 1930. The Geology of the Glass Mountains, Part I: De- scriptive Geology. University of Texas Bulletin, 3038:1-149. King, P. B.; R. E. King; and J. B. Knight 1945. Geology of the Hueco Mountains, El Paso and Hudspeth Counties, Texas. United States Geologi- cal Survey Oil and Gas Investigations Preliminary Map, 36. Merriam, D. F. 1963. The Geologic History of Kansas. Kansas Geologi- cal Survey Bulletin, 162:1—317. Moore, R. C. 1936. Stratigraphic Classification of the Pennsylvanian Rocks of Kansas. Kansas Geological Survey Bul- letin, 22:1-256. Myers, D. A. 1968. Schwagerina crassitectoria, Dunbar and Skinner, 1937, a Fusulinid from the Upper Part of the Wichita Group, Lower Permian, Coleman County, Texas. United States Geological Survey Professional Paper, 600-B:B133-B139. Otte, C, Jr. 1959. Late Pennsylvanian and Early Permian Stratig- raphy of the Northern Sacramento Mountains, Otero County, New Mexico. New Mexico Bureau of Mines and Mineral Resources Bulletin, 50:1-111. Ozawa, T. 1967. Pseudofusulinella, a Genus of Fusulinacea. Trans- actions and Proceedings of the Paleontological So- ciety of Japan, new series, 68:149-173, plates 14-15. Ross, C. A. 1959. The Wolfcamp Series (Permian), New Species of Fusulinids, Glass Mountains, Texas, Journal of the Washington Academy of Sciences, 49(9) : 299-316, 4 plates. 1962. Fusulinids from the Leonard Formation (Permian), Western Glass Mountains, Texas. Cushman Foun- dation for Foraminiferal Research, Contributions, 13(1): 1-21, 6 plates. 1963. Standard Wolfcampian Series (Permian), Glass Mountains, Texas. Geological Society of America Memoir, 88:1-205, 29 plates. 1967. Eoparafusulina from the Neal Ranch Formation (Lower Permian), West Texas. Journal of Paleon- tology, 41:943-946,, 1 plate. Skinner, J. W., and G. L. Wilde 1965a. Lower Permian (Wolfcampian) Fusulinids from the Big Hatchet Mountains, Southwestern New Mexico. Cushman Foundation for Foraminiferal Research, Contributions, 16 (3) :95—104, 4 plates. 1965b. Permian Biostratigraphy and Fusulinid Faunas of the Shasta Lake Area, Northern California. Univer- sity of Kansas Paleontological Contributions, Pro- tozoa (6) : 1-98, 65 plates. 1966a. Permian Fusulinids from Pacific Northwest and and Alaska, Part 1. Permian Fusulinids from North- western Nevada. University of Kansas Paleontologi- cal Contributions, 4:1-10, plates 1-8. 1966b. Permian Fusulinids from Pacific Northwest and Alaska, Part 2. Permian Fusulinids from Suplee Area, East-Central Oregon. University of Kansas Paleontological Contributions, 4:11-16, plates 9-13. 1966c. Permian Fusulinids from Pacific Northwest and Alaska, Part 3. Permian Fusulinids from North- eastern Washington. University of Kansas Paleon- tological Contributions, 4:16-22, plates 14-17. 1966d. Type Species of Pseudofusulina Dunbar and Skin- ner. University of Kansas Paleontological Con- tributions, 13:1—7. 1967. Eowaeringella, New Generic Designation for Fusu- linids of the Group of Wedekindellina ultimata Newell and Keroher. Journal of Paleontology, 41 (Paleontology Notes): 1004-1005. Stark, J. T., and E. C. Dapples 1946. Geology of the Los Pinos Mountains, New Mexico. Bulletin of the Geological Society of America, 57:1121-1172. Steiner, M. B., and T. E. Williams 1968. Fusulinidae of the Laborcita Formation (Lower Permian), Sacramento Mountains, New Mexico. Journal of Paleontology, 42:51-60, plates 11-13. Stewart, W. J. 1963. The Fusulinid Genus Chusenella and Several New Species. Journal of Paleontology, 37:1150-1163. plates 155-158. Thompson, M. L. 1951. New Genera of Fusulinid Foraminifera. Cushman Foundation for Foraminiferal Research, Contribu- tions, 2(4) : 115-119, plates 13, 14. 1954. American Wolfcampian Fusulinids. University of Kansas Paleontological Contributions, Protozoa (5): 1-226, 52 plates. Verville, G. J., and D. H. Lokke 1956. Fusulinids of the Desmoinesian-Missourian Con- tact. Journal of Paleontology, 30:793-810, plates 89-93. Verville, G. J.; M. L. Thompson, and D. H. Lokke 1956. Pennsylvanian Fusulinids of Eastern Nevada. Jour- nal of Paleontology, 30:1277-1287, plates 133-136. Wheeler, H. E., and J. C. Hazzard 1946. Permian Fusulinids of California. Geological So- ciety of America Memoir, 17:1-77, 18 plates. Wilde, G. L. 1962. Lower Permian Biostratigraphic Relationships and Sedimentation. In Leonardian Facies of the Sierra Diablo Region, West Texas. Permian Basin Sec- tion Society of Economic Paleontologists and Min- eralogists Guidebook, pages 62—67, 68—90, plates 1-4. Williams, T. E. 1963. Fusulinidae of the Hueco Group (Lower Permian), Hueco Mountains, Texas. Peabody Museum of Nat- ural History Bulletin, 18:1-122, 25 plates. NUMBER 3 379 1966. Permian Fusulinidae of the Franklin Mountains, New Mexico-Texas. Journal of Paleontology, 40: 1142-1156, plates 147-150. Wilpolt, R. H.; A. J. MacAlpin; R. L. Bates; and G. Vorbe 1946. Geologic Map and Stratigraphic Sections of Pale- ozoic Rocks of Joyita Hills, Los Pinos Mountains, and Northern Chupadera Mesa, Valencia, Tor- ance, and Socorro Counties, New Mexico. United States Geological Survey Oil and Gas Investigations Preliminary Map, 61. Subject Index Abo Formation, 301 Admiral Formation, 375 Alacran Mt. Formation, 373 Deer Mt. Member, 374 Alaska, Permian rhynchonelloids, 313 Algae, as rock builders, 4 Antelope Valley Limestone, 102 Antietam Formation, 4 Appalachians, Paleozoic Stratigraphy, 3 Aqua Torres Formation, 373 Arctic, Permian linoproductid, 347 rhynchonelloids, Permian, 313 Victoria Island, Precambrian brachiopods are Cambrian, 74 Argentina trigonirhynchid, n. sp., Silurian, 139 Arkansas, triplesiid biostratigraphy, Silurian, 143 Aserian Limestone, 106 Athens Shale, 19 Australia, brachiopods, supposed Precambrian, 73 linoproductid, Permian, 347 Avis Limestone Member, 17 Bays Formation, 17 Becraft Limestone, 13, 15 Beekmantown Group, 12 Beirdneau Formation, 219 Lower Carbonate Member, 226 Sandstone Member, 226 Upper Carbonate Member, 226 Belknap Limstone, 376 Belle Plains Formation, 375 Biogeography, Appalachians, Northern, Ordovician brachiopods, 115 China, Devonian, rhynchonellid, 211 Circum-Pacific Permian linoproductid, 350 Eastern U.S., Devonian, meristellid, 181 Tethyan-Boreal Permian brachiopod faunas, 337 Texas, West, Permian sponges, 298 Biostratigraphy, see also Stratigraphy brachiopods, Mississippian, Iowa, 245 conodonts, significance of, 21 corals, Devonian, Quebec, 194 fusulinid, Early Permian, California, 363 meristellid, Devonian, eastern U.S., 181 rhynchonellid, Devonian, China, 211 sponge zonation, Permian, West Texas, 285 Blackford Formation, 12 Blackgum Formation, 145, 146 Bluefield Formation, 17 Bluestone Formation, 17 Bois Blanc Formation, 163 Bone Spring Formation, 285, 287, 292, 294 Brachiopoda biogeography, China, Devonian, rhynchonellid, 211 eastern U.S., Devonian, meristellid, 181 Northern Appalachia, Ordovician, 115 biostratigraphy, meristellid, Devonian, eastern U.S., 181 rhynchonellid, Devonian, China, 211 silicified fossils, Mississippian, Iowa, 245 branched spines, Lower Ordovician, 83 classification, articulate, 42 cyrtomatodont hinge-teeth, 35 deltidiodont hinge-teeth, 34 enigmatic n. gen., Silurian, 155 hinge evolution, 33 hinge structure, oldhaminoid, Pennsylvanian, Texas, 271 hinge-teeth terminology, 34 inarticulate, spine morphology, Poland, 83 inarticulate, Trematis, 93 linoproductinid, Permian, 347 as lithic components, 13 mantle retraction, 61, 64 meristellid, Devonian, eastern U.S., 181 morphology, Trematis, 93 muscle reconstruction, oldhaminoid, Pennsylvanian, 272 muscle terminology, inarticulates, 94 oldhaminoid, Pennsylvanian, Texas, 267 ontogeny, Trematis, 98 paleoecology, Trematis, 98 periostracum, growth features, 55 Precambrian ( ?), 71 productoid, Mississippian, western U.S., 257 rhipidomellid, Mississippian, Iowa, 246 rhynchonelloid, Devonian, Nevada, 175 rhynchonelloid, Permian, Arctic, 313 rhynchonelloid Yunnanella, Devonian, redescribed, 203 shell-growth, articulate, 47 periostracum, 55 primary layer, 56 secondary layer, 60 silicified, Mississippian, Iowa, 245 spiriferids, Mississippian, Iowa, 248, 250 stratigraphy, early Middle Ordovician, Northern Appala- chians, 113 syntrophopsid n. gen., Middle Ordovician, 125 Tethyan-Boreal Permian faunas, 337 trigonirhynchid, n. sp., from Argentina, Silurian, 139 triplesiid, biostratigraphy, Silurian, 143 morphology, Silurian, 146 381 382 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY Brallier Formation, 9, 15 Brassfield Formation, 143, 146 Brownville Limestone, 376 Burgess Shale, 78 Bursum Formation, 301, 373, 377 California, brachiopod, enigmatic, Silurian, 155 fusulinid biostratigraphy, Early Permian, 363 Camden Chert, 162 Capitan Formation, 285, 287, 297 Carboniferous, Mississippian brachiopods, Iowa, 245 Mississippian, Late, productid, western U.S., 257 Pennsylvanian oldhaminoid morphology, Texas, 267 Carlisle Center Formation, 161 Carlsbad Formation, 297 Cason Shale, 144, 145 Cathedral Mt. Formation, 286, 287, 294 Centinela Formation, 140 Cephalopods, goniatites, Permian, 301 goniatite wrinkle-layer in, 23 of Whiterock Stage, 101 Cerro Alto Formation, 373 Chambersburg Limestone, 19 Chapoghlu Shale Member, 73 Cherry Canyon Formation, 285 Chickamauga Formation, 18 China, rhynchonellid, Devonian, 203 Classification, Brachiopods, articulate, 42 fusulinid, Early Permian, 369 phenetic vs. phyletic, 369 Climate, Tethyan-Boreal Permian brachiopod faunas, 337 Clinch Sandstone, 14 Clyde Formation, 293 Cochrane Formation, 146 Coeymans Limestone, 13, 15 Conasauga Formation, 5 Conococheague Formation, 8 Conodonts, biostratigraphic significance of, 21 Silurian correlation, 150 Copper Ridge Dolomite, 7, 8 Corals, biostratigraphy, Devonian, Quebec, 193 Correlation, Permian fusulinids, U.S., 370 Permian linoproductinid, 352 Silurian conodont zonation, 150 Silurian Triplesia-bearing beds, central U.S., 145 Coyote Butte Formation, 367 Crown Point Limestone, 102 Day Point Limestone, 102 Denmar Limestone, 16 Devils Gate Limestone, 231 Devonian, clastic strata, Appalachians, 15 corals, Quebec, 193 Late, gastropods, Nevada, 231 Lower, ostracodes, New York, 161 meristellid, eastern U.S., 181 rhychonelloid, Nevada, 175 -Silurian limestones, Appalachians, 14 stratigraphy, Utah, 219 Echinoderms, in Appalachian carbonate rocks, 10 Edinburg Formation, 12, 19 Effna Limestone, 12, 13 Elbrook Formation, 7, 8, 12 Elgin Shaly Limestone Members, 132 Elway Limestone, 14 Erwin Formation, 4 Evolution, see also Phylogeny brachiopod, in Permian Tethyan-Boreal facies, 343 brachiopod hinge, 33 oldhaminoid, Pennsylvanian, 277 productoid brachial valve, Mississippian, 257 sponge, Permian, West Texas, 286 Famine Limestone, 193 Fig Tree Series, 71 Florence Limestone, 377 Foraker Limestone, Americus Limestone Member, 377 Hughes Creek Shale Member, 377 Functional Morphology, see also Morphology brachiopod hinge-teeth, 40 brachiopod shell growth, 47 brachiopod spines, 90 gastropod, Late Devonian, Nevada, 240 goniatite wrinkle-layer, 28 oldhaminoid, Pennsylvania, Texas, 267 rhynchonelloids, Permian, Arctic, 324 Fusulinids, phylogeny, biostratigraphy Early Permian, 363 Gaptank Formation, 371 Gastropods, in Appalachian dolomites, 12 morphology, Late Devonian, Nevada, 231 Getaway Limestone, 296 Goat Seep Limestone, 295 Graham Formation, 310 Grainger Formation, 16 "Gray Limestone" of King (1930), 371 Greenbrier limestones, 16 Gunflint Chert, 71 Halleck Formation, 327 Hampton Formation, Eagle City Limestone Member, 245 Healing Springs Sandstone, 15 Hess Formation, Taylor Ranch Member, 292 Hillsdale Limestone, 9, 18 Hinton Formation, 17 Honaker Dolomite, 5 Hueco Canyon Formation, 373 Powwow Member, 373 Hueco Formation, 285, 289, 292, 373 Hyrum Formation, 219 Lower Carbonate-Detritus Member, 223 Lower Dolomite Member, 223 Samaria Member, 222 Upper Carbonate-Detritus Member, 224 Upper Dolomite Member, 224 India, brachiopods, supposed Precambrian, 72 Iowa, brachiopods, Mississippian, 245 trilobite, n. gen., Maquoketa Shale, 129 Iran, brachiopods, supposed Precambrian, 72 Jefferson Formation, 224 Birdbear Member, 224 Juniata Formation, 17 Kaibab Formation, 295 NUMBER 3 383 Kansas, Permian fusulinid correlations, 375 Keyser Limestone, 13, 14, 15 Knox Dolomite, 14, 17 Laborcita Formation, 301, 373 Leatham Formation, 225 Lee Formation, 17 Lenoir Limestone, 12, 14 Lenox Hills Formation, 289, 292, 370 Leonardian Clyde Formation, 375 Liberty Hall Limestone, 12 Licking Creek Limestone, 15 Lincolnshire Limestone, 15, 18 Little Valley Limestone, 18 Longview Limestone, 9, 18 Lower Suhuho Formation, 211 Maccrady (shale), 15 Maine, brachiopods, early Middle Ordovician, 113 Maquoketa Shale, 132 Martinsburg Formation, 3, 14, 19 Maruyama Limestone, 106 Maryville Limestone, 5, 6, 10 Maynardville Limestone, 7 McCloud Limestone, 363 Mecoyita Formation, 140 Millboro Formation, 3, 15 Mississippian, see Carboniferous Montana, brachiopods, supposed Precambrian, 74 Moran Formation, Gouldbusk Limestone Member, 375 Morphology, see also Functional Morphology brachiopod, branched spines in, 83 brachiopod hinge structures, 34 brachiopod Trematis, 93 gastropod coiling and septation, 231 goniatite wrinkle-layer, 23 linoproductinid spines, Permian, 353 oldhaminoid, Pennsylvanian, Texas, 268 triplesiid, Silurian, 146 Mosheim Limestone, 17 Neal Ranch Formation, 286, 287, 289, 292, 370 Needmore Formation, 15 Nevada, brachiopod, enigmatic, Silurian, 155 gastropods, late Devonian, 231 productid, Mississippian, 257 rhynchonelloid, Devonian, 175 syntrophopsid, n. gen., Ordovician, 125 Neva Limestone, 377 New Brunswick, brachiopods, early Middle Ordovician, 113 Newfoundland, brachiopods, early Middle Ordovician, 113 syntrophopsid n. gen., Ordovician, 125 Newland Formation, 74 New Market Limestone, 14 New Mexico, goniatites, Permian, 301 New York, ostracodes, Lower Devonian, 161 Pentagonia, Devonian, 181 New Zealand, linoproducted, Permian, 347 Nolichucky Formation, 6 Oklahoma, triplesiid biostratigraphy, Silurian, 143 Onondaga Limestone, 162, 181 Ontogeny, brachiopod Trematis, 98 dorsal valve in Poikilosakos, 270 septation in molluscs, 239 shell growth in articulate brachiopods, 47 Oranda Formation, 19 Ordovician, cephalopods, Whiterock Stage, 101 Lower, brachiopod spine morphology, 83 Lower, Poland, chalcedonites, 83 Middle, brachiopods, northern Appalachians, 113 Middle, new syntrophopsid gen., 125 Trematis, morphology and paleobiology, 93 trilobite, n. gen., from Maquoketa Shale, 129 Oriskany Formation, 15 Orthidiella zone, 126 Orthoceras shale, 102 Ostracodes, Lower Devonian, New York, 161 Ottosee Limestone, 18 Pacific-Northwest, Permian fusulinid correlations, 370 Paleoecology, brachiopod Trematis, 98 gastropods, Nevada, Late Devonian, 240 linoproductinid, circum-Pacific, Permian, 353 meristellid Pentagonia, Devonian, eastern U.S., 184 rhynchonelloids, Arctic Permian, 313 sponges, Permian, West Texas, 299 Paleozoic stratigraphy, Appalachian, 3 Park City Formation, Franson Member, 295 Pelecypods, recent family distribution data, 341 Pennsylvanian, see Carboniferous Permian, fusulinid phylogeny, 363 goniatites, New Mexico, 301 linoproductinid, 347 rhynchonelloids, Arctic, 313 sponge zonation, West Texas, 285 Tethyan-Boreal brachiopod faunas, 337 Phylogeny, fusulinid, Early Permian, 363 meristellid Pentagonia, Devonian, eastern U.S., 182 Mississippian productids, 257 oldhaminoid, Pennsylvanian, 277 sponges, Permian, West Texas, 285 Poland, brachiopod spine morphology, Lower Ordovician, 83 chalcedonites, Lower Ordovician, 83 Porifera, biostratigraphy, Permian, West Texas, 285 evolution, Permian, West Texas, 285 Portage lithofacies, 15, 16 Potsdam Formation, 4 Powwow Conglomerate, 310 Precambrian brachiopods, supposed, 71 Price Formation, 16 Princeton Formation, 9 Pueblo Formation, 375, 377 Putnam Formation, 375 Coleman Junction Limestone Member, 375 Quebec, corals, Devonian, 193 Recent clam family distribution data, 341 Red Eagle Limestone, 302, 311 Rickard Hill Member, 162 Road Canyon Formation, 287, 294 Roberts Mountain Formation, 156 384 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY Rocky Gap Formation, 15 Rogersville Shale, 5, 6 Rome Formation, 5 Row Park Limestone, 14 Rutledge Limestone, 5, 6 St. Clair Limestone, 144 St. Louis Limestone, 9, 18 Salona Formation, 19 San Andres Formation, 294, 296 Schoharie Formation, 161 Sedimentation, Devonian, Utah, 224, 228 Ordovician chalcedonites, Poland, 83 patterns in Appalachian Paleozoic, 3 Shady Formation, 4 Shaler Group, 74 Shin Brook Formation, 113 Silurian, brachiopod, affinities unknown, 155 Devonian limestones, Appalachians, 14 stratigraphic sequence, Argentina, 140 trigonirhynchid, H. sp., Argentina, 139 triplesiid biostratigraphy, Arkansas, Oklahoma, 143 Sinuites beds, 19 Skinner Ranch Formation, 286, 287, 292 Decie Ranch Member, 292, 372 Slaven Chert, 176, 178, 179 Soltanieh Dolomite, 73 Stone Corral Formation, 377 Stratigraphy Appalachian, Paleozoic, 3 Silurian-Devonian limestones, 14 Devonian and younger strata, 15 nonfossiliferous, 17 Devonian, Lower, New York, 161 Utah, 219, 223 and younger strata, Appalachia, 15 maverick fossils, value of, 18 Ordovician, early Middle, brachiopods, Northern Appalach- ians, 113 Paleozoic, Appalachian, 3 recurring faunules, 19 silicified fossils, value of, 20 Silurian-Devonian limestones, Appalachians, 14 Silurian sequence, Argentina, 140 Whiterock cephalopods, 101 Suket shales, 72 Systematics, articulate brachiopod, 42 fusulinid, Early Permian, 369 goniatite, 24 Ordovician cephalopod, 101 Table Head Formation, 126 Taxonomy {see also Systematics), ammonites, Permian, 301 "brachiopods", supposed Precambrian, 73 linoproductid Terrakea, Permian, 347 productid Carlinia, Mississippian, 257 rhynchonellid Yunnanella, Devonian, 204, 208 rhynchonelloids, Permian, Arctic, 313 Texas, linoproductid, Australian, possible occurrence, 350 Permian fusulinid correlations, 370, 373, 375 West, sponge zonation, Permian, 285 Three Forks Formation, 228 Logan Gulch Member, 228 Trident Member, 228 Trace fossils, Appalachian, 9 Trilobite, calymenid, Ordovician, n. gen., 129 Tumbez Limestone, 12, 14 Tuscarora Formation, 14 United States, eastern, meristellid Devonian biostratigraphy, 181 Western, productid, Mississippian, 257 Upper Suhuho Formation, 211 USSR, linoproductid Terrakea occurrence, Permian, 352 Yunnanella, Nayunnella possible occurrence, 212 Utah, productid, Mississippian, 257 stratigraphy, Devonian, 219 Vaginatum Limestone, 106 Valcour Limestone, 102 Victoria Island, see Arctic Canada Victorio Peak Formation, 285, 294 Virgilian Holder Formation, 289 Wardell Limestone, 18 Wassum Limestone, 9 Water Canyon Formation, 219 Card Member, 224 Grassy Flat Member, 221 Weisner Formation, 4 Witten Limestone, 4, 9 Word Formation, 285, 287 China Tank Member, 287 Willis Ranch Member, 296 Wuting Limestone, 106 Wyoming, productid, Mississippian, 257 Wysoczki chalcedonites, 83 Zapla Formation, 139 Index to Genera and Species (New taxa and principal entries in italics) Acanthothiris, 90 Acinophyllum crassiseptatum, 196 rectiseptatum, 196 Acrothele, 72 Actinoconchus, 248 Aenigmastrophia, 155 cooperi, 156 (fig.), 158 greggi, 156 (fig.), 758 affine, Manticoceras, 26 Agoniatites oxynotus, 24, 28, 29 (fig.) Ahtiella, 115, 120 alata, Triplesia, 143, 146, 147 (fig.), 148 (fig.), 150 (fig.), 152 (fig.) Ambocoelia, 15 Amphigenia, 15 amsdeniana, Carlinia, 261, 262 (fig.) Anarcestes praecursor, 26 Anastrophia, 37 Ancillotoechia, 139 cooperensis, 139, 141 (fig.) Angusticardinia, 39 Anthracospirifer, 250 anticostiensis, Triplesia, 146 antiquatus, Cryptophragmus, 18 Antirhynchonella linguifera, 44 (fig.) Apatobolbina, 164 Arcestes (Proarcestes), 28 hauieli, 27 (fig.), 31 Archimedes, 17 arctica, Camerisma (Callaiapsida), 325 (fig.), 327, 328 (fig.), 332 (fig.) arctica, Terrakea, 354 (fig.), 357 (fig.) Arctomphalus grandis, 239 Arikloedenia, 165 Arthrophycus, 4, 9 Artinskia, 28 "aspera," Atrypa, 88 Athyris, 248 densa,249 Atrypa "aspera," 88 unisulcata, 181 Aulacophyllum sp., 198, 199 (fig.) Aulopora, 184, 190 (fig.) Bactroceras, 102 Bactrynium, 44 Bairdiacypris? sp., 166, 170 (fig.) Baltoceras, 102 bashkirica, Dzieduszyckia, 177 (fig.), 179 Bathmoceras, 101 Bathmoceras norvegicum, 102 Beekmanoceras, 103 belemnitiforme, Nanno, 103 Beloceras, 23, 26, 27 (fig.) sagittarium, 31 (fig.) "Beyrichia" kloedeni, 166 occidentalis, 166 Simplicibeyrichia (subgen.), 165 Beyrichoceras, 27 Beyrichoceratoides, 27 bicostata, Pentagonia, 185 bilobatum, Najaceras, 105, 108 (fig.) Bingeria, 165 biplicata, Pentagonia, 185 biseriatus, Pteridichnites, 16 bisulcata, Pentagonia, 185 brachythaerum, Terrakea, 353 (fig.) brachythaerus, Productus, 347 Bouchardia, 43 sp., 44 (fig.) Buthotrephis, 9 Buttsoceras, 108 Cacheoceras, 103 Cadomella, 35, 44 Calaurops, 232 Callaiapsida (subgenus), Camerisma, 323, 325 (fig.), 327, 328 (fig.), 330 (fig.) 332 (fig.) Calymene mammillata, 132 Camarocladia, 9 Camarophoria kutorgae, 315 Camarotoechia, 16 Camdenidea, 167 Camerisma, 323 Camerisma {Callaiapsida), 323 arctica,325 (fig.), 327, 328 (fig.), 332 (fig.) kekuensis, 330 (fig.), 332 (fig.) Cameroceras, 105 Cancrinella koninckiana, 350 capsella, Fermoria, 73 caractaci, Flexicalymene, 131 (fig.) cardilatus, Cuparius, 126, 127 (fig.) Carlinia, 258 amsdeniana, 261, 262 (fig.) diabolica, 262 (fig.) phillipsi, 259, 262 (fig.) Cartersoceras, 102 centrifuga, Straparollus (Serpulospira), 232 (fig.) Ceratopea, 12 cestriensis, Diaphragmus, 264 385 386 SMITHSO NIAN CONTRIB UTIONS TO PALEOBI OLOGY chaleurensis, Tubulibairdia, 170 Cheilcceras, 27 chevroniferum, Najaceras, 106 (fig.), 108 (fig.) Chonetes, 15, 16 Christiania, 115 lamellosa, 19 Chuaria circularis, 72 circularis, Chuaria, 72 Cladopora, 15 Cleiothyridina, 248 Clelandoceras, 103 Clinoceras, 108 Clistoceras, 304 Clitambonites squamatus, 40, 41 (fig.) Clymenia, 27 pseudogoniatites, 27 collilupana, Phanassymetria, 767, 170 (fig.) Composita, 17, 248 compressus, Merocanites, 27 corica, Lyttonia, 272 contrarius, Mimulus, 149 convolva, Gravicalymene, 131 (fig.) cooperensis, Ancillotoechia, 139, 141 (fig.) cooperi, Aenigmastrophia, 156 (fig.), 758 Nevadaspira, 232, 233 (fig.), 235 (fig.) Parabingeria, 165, 170 (fig.) Copiceras, 103 Coreanoceras, 103 corbuloides, Cythere, 170 cordatum, Manticoceras, 26 Costistricklandia, 36 lirata, 37 (fig.), 40 (fig.) crassicosta, Halorella, 178 Cryptophragmus antiquatus, 18 Cryptozoon, 8 Cumberloceras, 102 cummingi, Munhella, 117, 120 (fig.) Cuparius, 125 cardilatus, 126, 127 (fig.) Cupularostrum, 139 Cylomedusa plana, 71 Cylindrophyllum stummi, 195, 197 (fig.), 199 (fig.) Cymaclymenia, 27 Cyptendoceras, 102, 104 Cyrtoclymenia, 27 Cyrtospirifer disjunctus, 18 Cystiphylloides sp., 199 (fig.), 207 Cythere corbuloides, 170 davidsoni, Hesperorthis, 34 (fig.), 40 (fig.) decaturi, Tubulibairdia, 169 Delepinoceras, 27 Dendropoma, 241 densa, Athyris, 249 Planalvus, 249 diabolica, Carlinia, 262 (fig-) Diacalcymene, 130 diademata, 130, 131 (fig.) diademeta, Diacalymene, 130, 131, (fig.) Diaphragmus, 257 cestriensis, 264 Dicoelosia, 35, 36, 156 Dictyonina, 75 pannula, 78 sp., 75, 76 (fig.) Dideroceras wahlenbergi, 103 Diestoceras, 110 Dinorthis, 14 Discinisca, 93, 95, 98 laevis, 93 lamellosa, 100 Discoceras perornatus, 108 discoidale, Gonioloboceras, 301, 306 discoidalis, Mescalites, 304 (fig.), 305 (fig.), 306 (fig.), 308 disjunctus, Cyrtospirifer, 18 Spirifer, 16 Doleroides, 48 dubia, Perditocardinia, 247 Dzieduszyckia, 775 baschkirica, 177 (fig.), 179 kielcensis, 179 sp., 177 (fig.), 778 Ecculiomphalus, 232, 238 Ecdyceras, 101 Echinosphaerites, 19 egrediens, Owenites, 30 (fig.), 31 elegantula, Resserella, 34 (fig.) eliasi, Gonioloboceras, 305 Elkanoceras, 110 elliptopora, Trematis, 93, 94 (fig.), 95 (fig.), 96 (fig.), 97 (fig.), 99 (fig.) Emmonsoceras, 103 Enantiosphen, 35, 44 Endoceras gladius, 103 Eoasianites subtilicostatus, 306 (fig-), 307 (fig.) Eobactrites, 102 Eochoristites, 250 Eodalmanella, 115 Eoparaphorhynchus, 206, 210 Eoplectodonta, 115 Eospirifer, 39 (fig.) radiatus, 35 (fig.) Eremotoechia, 115 Euomphalus (subgen.), Straparollus, 236 fayettensis, Flexicalymene, 134 fecunda, Tubulibairdia, 170 Fermoria capsella, 73 granulosa, 73 minima, 72 fidelis, Mimagoniatites, 24 Fistulogonites, 113, 776 novaterrensis, 117, 120 (fig.) Flexicalymene, 130 cf. caractaci, 131 (fig.) fayettensis, 134 gracilis, 134 fortunata, Mirifusella, 251 (fig.), 252 (fig.), 254 (fig.) Fusella, 250 galeata, Gypidula, 37 (fig.) gemmisulcata, Pentagonia, 181 NUMBER 3 387 genundewa, Ponticeras, 26 gibberosa, Planalvus,249, 250 (fig.), 252 (fig.), 254 (fig.) Girtyoceras, 27 glabra, Triplesia, 149 gladius, Endoceras, 103 Glossorthis, 117 cf. Munhella, 115 Glottidia, 98 goldringae, Pentagonia( ?), 194 Goniocoelia, 181 Gonioglyphioceras gracile, 308 (fig.) Gonioloboceras, 303, 305 discoidale, 301, 306 eliasi, 305 goniolobum, 304 (fig.), 305 (fig.), 308 (fig.) parrishi, 305 goniolobum, Gonioloboceras, 304 (fig.), 305 (fig.), 308 (fig. gracile, Gonioglyphioceras, 308 (fig.) gracilis, "Flexicalymene", 134 Gyroceratites, 28 Graciloceras, 110 Graftonoceras, 108 Grammysia, 15 Grandaurispina, 350 grandis, Arctomphalus, 239 grandis, Schohariella, 164, 170 (fig.) granulosa, Fermoria, 73 Gravicalymene, 129 convolva, 131 (fig.) hagani, 130 greggi, Aenigmastrophia, 156 (fig.), 758 Guadalupia, 287 Gurleyoceras, 305 Gypidulagaleata, 37 (fig.) Gyroceratites, 24 gracilis, 28 hagani, Gravicalymene, 130 halli, Heliophyllum, 196, 200 (fig.) Hallopora, 14 Halorella, 175 crassicosta, 178 intermedia, 178 hanburii, Yunnanella, 206, 209, 215 (fig.), 216 (fig.) hauieli, Arcestes (Proarcestes), 27 (fig.), 31 Hebertella, 48 Hebetoceras, 108 Heliophyllum halli, 796, 200 (fig.) sp. cf. H. proliferum, 198, 200 (fig.) Hemithyris psittacea, 35 (fig.), 44 (fig.), 63 Hesperorthis, 39 (fig.) davidsoni, 34 (fig.), 40 (fig.) Heterophrentis spp., 198, 199 (fig.), 200 (fig.) Huenella, 119 cf. Rugostrophia, 115, 119 Humeoceras, 101 "hunstvillae," Platycrinites, 16 husseyi, Papillicalymene, 131 Hyolithellus, 72 Hyolithes, 238 Hyrsynobolbina, 164 Ikesoceras, 110 Imbrexia, 250 incipiens, Trochoceras, 108 indianensis, Spirifer, 251 insulare, Orthoceras, 104 insularis, Triplesia, 147, 149 intermedia, Halorella, 178 iowensis, Perditocardinia, 246, 252 (fig-)> 254 (fig.) Isogramma, 156 jonesi, Protobolella, 73 josephinae, Vietor, 167, 170 (fig.) Juaboceras, 103 kekuensis, Camerisma (Callaiapsida), 330 (fig.), 332 keyserlingi, Timanites, 26 kielcensis, Dzieduszyckia, 179 Kiotoceras, 103 quadratum, 104, 106 (fig.) sp. 108 (fig.) kirkbyi, Moorea, 172 Kloedenia newbrunswickensis, 166 rectangularis, 166 koninckiana, Cancrinella, 350 Kosmoclymenia undulata, 27 kutorgae, Camarophoria, 315 Septacamera, 328 Lacazella, 43, 274 mediterranea, 44 (fig.) laevis, Discinisca, 93 lamellosa, Christiania, 19 Lecanospira, 12, 18 lenta, Meristella, 181 Pentagonia, 183 (fig.), 786, 188 (fig.), 190 (fig.) lentiformis, Meristella, 183, 788, 189 (fig.) Leonardoceras, 110 Leptobolbina, 164 Leptodesma, 15 Leptodus, 270 Leptostrophia, 15 Levisoceras, 102 linguifera, Antirhynchonella, 44 (fig.) Lingula, 14, 76, 98 Lingulella, 71, 73 montana, 74, 76 (fig.) Liomphalus, 236 lirata, Costistricklandia, 37 (fig.), 40 (fig.) Lithostrotionella, 9, 18 Litoceras, 108 Lituites pluto, 110 longula, Tubulibaiardia, 169 ludfordi, Pronorites, 28 lunulicosta, Pharciceras, 26 Lytospira, 232, 238 Lyttonia conica, 272 Macluritella, 237 Maeandrostia, 286 Maenioceras terebratum, 26, 28 (fig.) Magasella sanguinea, 58 (fig.), 63, 64 388 SMITHSON IAN CONTRIBU TIONS TO PALEOBIO LOGY mammillata, Calymene, 732 Thelecalymene, 131 (fig.), 732 (fig.), 134 (fig.) Manticoceras afnne, 26 cordatum, 26 sinuosum, 24, 26 (fig.) marshallensis, Spirifer, 225 Mastigospira, 238 mccalleyi, Pentremites, 17 mediterranea, Lacazella, 44 (fig.) Megathiris, 274 Megerlia truncata, 44 (fig.) mesastrialis, Spirifer, 15 Meikeloceras, 102 parvum, 103, 108 (fig.) Melocrinus, 16 Meniscoceras, 103, 105 Meristella, 788 lenta, 181 lentiformis, 183, 788, 189 (fig.) Meristina, 188 Merocanites compressus, 27 Mescalites, 304 discoidalis, 304 (fig.), 305 (fig.), 306 (fig.), 308 (fig.) mesleri, Neoaparchites, 766, 170 (fig.) "Paraparchites,'' 166 Michelinoceras, 106 primum, 106 Microcheilinella regularis, 170 seminalis, 170 millipunctata, Trematis, 95 Mimagoniatites fidelis, 24 Mimella, 48 Mimulus contrarius, 149 moera, 149 minima, Fermoria, 72 Protobolella, 73, 76 (fig.) Mirifusella, 250 fortunata,251 (fig.), 252 (fig.) 254 (fig.) moera, Mimulus, 149 montana, Lingulella, 74, 76 (fig.) Montyoceras, 108 Moorea kirkbyi, 172 Multispinula, 115 multitubulis, Tubulibairdia, 169 Munhella, 113, 777 cummingi, 117, 120 (fig.) Miinsteroceras, 27 Murrayoceras, 102 Najaceras, 103, 104 bilobatum, 105, 108 (fig.) chevroniferum, 106 (fig.), 108 (fig.) triangulatum, 105, 108 (fig.) Nanno belemnitiforme, 103 Nautilus, 23, 28,31 pompilius, 31 (fig.), 304 Nayunnella, 204, 205, 210, 211 Neoaganides, 27 Neoaparchites sp. aff. N. mesleri, 766, 170 (fig.) Neobolus, 72 Nevadaceras, 108 Nevadaspira, 232, 237, 240 cooperi, 232, 233 (fig.), 235 (fig.) newbrunswickensis, Kloedenia?, 166 nigricans, Notosaria, 49, 51 (fig.), 55 (fig.), 58 (fig.) northi, Straparollus (Euomphalus), 236 norvegicum, Bathmoceras, 102 Notosaria nigricans, 49, 51 (fig.), 55 (fig.), 58 (fig.) novaterrensis, Fistulogonites, 117, 120 (fig.) obesum, Tornoceras uniangulare, 27 obesus, Pentremites, 17 Obollella, 72 Oderoceras, 103 Odontomaria, 238 Oelandoceras, 103 Oldhamina, 270 Oligorhynchia, 12 Onnicalymene, 130 opitula, Septacamera, 321, 330 (fig.) ordinatum, Sudeticeras, 24 (fig.) Orthambonites, 115 Orthoceras insulare, 104 regulare, 108 Orthonychia, 237 Orthorhynchula, 14 ortoni, Triplesia, 145, 147, 152 Oslogonites cf. Fistulogonites, 115 Owenites, 28 egrediens, 27 (fig.), 31 (fig.) Oxoplecia, 12 oxynotus, Agoniatites, 24, 28 (fig.) Pachydomella tumida, 168 Paleoneilo, 15 pannula, Dictyonina, 78 pannulus, Trematis, 75 papillata, Papillicalymene, 131 (fig.) Papillicalymene, 131 husseyi, 131 papillata, 131 (fig.) Parabingeria, 165 cooperi, 165, 170 (fig.) Parabufina, 172 Parabufina? sp. 770 (fig.) "Paraparchites" mesleri, 166 Parodiceras, 27 parrishi, Gonioloboceras, 305 parvum, Meikeloceras, 103, 108 (fig.) Paterula, 72 paucitubulis, Tubulibairdia, 169 peersi, Pentagonia, 181, 183 (fig.), 785, 186 (fig.), 189 (fig.), 190 (fig.) Pentagonia, 787, 182, 184 bicostata, 185 biplicata, 185 bisulcata, 185 gemmisulcata, 181 ?goldringae, 184 lenta, 183 (fig.), 786, 188 (fig.), 189 (fig.), 190 (fig.) peersi, 181, 183 (fig.), 785, 186 (fig.), 189 (fig.), 190 (fig.) ?sp. 182 unisulcata, 183 (fig.), 787, 188 (fig.) NUMBER 3 389 Pentremites mccalleyi, 17 obesus, 17 Perditocardinia, 35, 44 dubia, 247 iowensis, 246, 252 (fig.), 254 (fig.) "Pericyclus", 27 perornatus, Discoceras, 108 petaloides, Poikilosakos, 278 (fig.), 281 (fig.) Phanassymetria, 168 collilupana, 167, 170 (fig.) Pharciceras lunulicosta, 26 phillipsi, Carlinia, 259, 262 (fig.) Productus, 259 Piloceras, 103 Pionodema, 48 Placotriplesia, 146, 149 praecipta, 149 (fig.) 152 (fig.) plana, Cyclomedusa, 71 Planalvus, 248 densa, 249 gibberosa, 249, 250 (fig.), 252 (fig.), 254 (fig.) Platyceras (Orthonychia), 237 Platycrinites "huntsvillae", 16 Platygoniatites, 27 Platystrophia, 115 pluto, Lituites, 110 Poikilosakos, 267, 269(fig.) petaloides, 278 (fig.), 281 (fig.) pollex (subsp), Terrakea, 357 (fig.) pompilius, Nautilus, 31 (fig.), 304 Ponticeras genundewa, 26 Ponticeras (?) uchtense, 26 Porambonites, 35, 36, 115, 125 potteri, Schedophyla, 122 (fig.) praecipta, Placotriplesia, 149 (fig.), 152 (fig.) praecursor, Anarcestes, 26 Prasopora, 14 prima, Syringospira, 58 (fig.) Primapis cf. crosotus, 132 primum, Michelinoceras, 106 Proarcestes (subgen.), Arcestes, 28 Probeloceras, 26, 27 (fig.) aff. lutheri, 27 (fig.), 28 (fig.) Probillingsites, 108 Productella, 15, 16 Productorthis, 115 Productus brachythaerus, 347 phillipsi, 259 solidus, 348 Pronorites, ludfordi, 28 Prorichthofenia, 269 Protobolella jonesi, 73 minima, 73, 76 (fig.) Pseudoclymenia, 27 Pseudofusulinella, 363, 366 pseudogoniatites, Clymenia, 27 Pseudosyrinx, 16 psittacea, Hemithyris, 35 (fig.), 44 (fig.), 63 Pteridichnites biseriatus, 9 Pterotocrinus, 16 punctulata, Tubulibairdia, 768, 170 (fig.) pybensis, Septacamera, 320 (fig.), 321 (fig.), 330 (fig.) quadratum, Kiotoceras, 104, 106 (fig.) radiatus, Eospirifer, 35 (fig.) rectangularis, Kloedenia, 166 rectiseptatum, Acinophyllum, 196 Redpathoceras, 108 regulare, Orthoceras, 108 regularis, Microcheilinella, 170 Rennselandia, 223 Resserella elegantula, 34 (fig.) reticulosa, Schohariella, 164, 170, (fig.) Treposella, 164 retusa, Terebratulina, 44 (fig.), 55 (fig.), 63 Rhabdiferoceras, 102 Rhinidictya, 14 Rhipidomella, 15 thiemei, 247 Richtofenia, 65 Robsonoceras, 103 roemeri, Sieberella, 44, (fig.) Rossoceras, 103 Rostricellula, 12, 14, 139 Rotaia, 315 Ruedemannoceras, 110 Rugostrophia, 113, 778 silvestris, 779 (fig.), 122 (fig.) sagittarium, Beloceras, 31 (fig.) sanguinea, Magasella, 58 (fig.), 63, 64 Scacchinella, 65 Schedophyla, 113, 720 potteri, 122 (fig.) Schizocrania, 100 Schizophoria, 48 Schnurella, 205 Schohariella, 164 grandis, 164, 170 (fig.) reticulosa, 164, 170 (fig.) Scolithus, 4, 5 (fig.), 6 (fig.), 7 (fig.) seminalis, Microcheilinella, 170 Septacamera, 314, 375, 316 kutorgae, 328 (fig.) opitula, 321, 330 (fig.) pybensis,320 (fig.), 321 (fig.), 330 (fig.) sp. 322, 328 (fig.) stupenda, 318 (fig.), 319 (fig.), 328 (fig.), 330 (fig.) Serpulosira (subgen.), Straparollus, 234, 237 Sieberella, 37 roemeri, 44 (fig.) Silus sp., 167 silvestris, Rugostrophia, 119 (fig.), 122 (fig.) simplex, Tubulibairdia, 169 Simplicibeyrichia (subgen.), Beyrichia, 165 Sinoceras, 108 Sinuites, 19 sinuosum, Manticoceras, 24, 26 (fig.) Siphonophrentis sp., 2, 200 (fig.), 207 sp. cf. S. yandelli, 798,199 (fig.) 390 Siphonotreta(?) sp., 83, 84, 85 (fig.), 86 (fig.), 87 (fig.), 88 (fig), 89 (fig.) Sobolewia, 27 (fig.) virginiana, 26, 27 (fig.), 28 (fig.) Solenopora, 9, 12 solidus, Productus, 348 Sowerbyella, 14 Spinolyttonia, 278 Spirifer, 16 disjunctus, 16 indianensis, 251 marshallensis, 255 mesastrialis, 15 Spirigerella, 248 squamatus, Clitambonites, 40 (fig.) Straparollus (Euomphalus) northi, 236 (Serpulospira), 234, 237 centrifuga, 232 (fig.) Stricklandia, 36 Stringocephalus, 223 stummi, Cylindrophyllum, 195, 197 (fig.), 199 (fig.) stupenda, Septacamera, 318 (fig.), 319 (fig.), 328(fig.), 330 (fig.) Stylopegma, 286 subtilicostatus, Eoasianites, 306 (fig.), 307 (fig.) Sudeticeras, 26, 27 ordinatum, 24 (fig.) Syntrophopsis, 125 Syringospira, 65, 66 (fig.) prima, 58, (fig.) Syringothyris, 16 Talarocrinus, 16 Telemarkites, 74 Tentaculites, 15 Terebratulina retusa, 44 (fig.), 55 (fig.), 63 terebratum, Maenioceras, 26, 29 (fig.) terminalis, Trematis, 95 Terrakea, 347, 350, 353, 354 arctica, 354 (fig.), 357 (fig.) brachythaerum, 353 (fig.) pollex (n. subsp.), 357 (fig.) sp., 357 (fig.) Tetracamera, 16 Tetradium, 18 Thecidellina, 43, 274 Thecospira, 35, 44 Thelecalymene, 130 mammillata, 131 (fig.), 732 (fig.), 134 (fig.) thiemei, Rhipidomella, 247 Timanites keyserlingi, 26 Tornoceras, 27 (fig.), 31 arcuatum, 27 uniangulare obesum, 27 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Trematis elliptopora, 93, 94 (fig.), 95 (fig.), 96 (fig.), 97 (fig.), 99 (fig.) millipunctata, 95 pannulus, 75 terminalis, 95 Treposella reticulosa, 164 triangulatum, Najaceras, 105, 108 (fig.) Triplesia, 146 alata, 143, 146, 147 (fig.), 148 (fig.), 150 (fig.), 152 (fig.) anticostiensis, 146 glabra, 149 insularis, 147, 149 ortoni, 145, 147 (fig.), 152 (fig.) wenlockensis, 149 woodlandensis, 149 Tritoechia, 115 Trochoceras incipiens, 108 Trocholites, 108 Tropidoleptus, 35, 44 truncata, Megerlia, 44 (fig.) Tubulibairdia, 168 chaleurensis, 170 decaturi, 169 fecunda, 170 longula, 169 multitubulis, 169 paucitubulis, 169 punctulata, 768, 170 (fig.) simplex, 169 tubulifera, 169 windomensis, 169 tubulifera, Tubulibairdia, 169 tumida, Pachydomella, 168 uchtense, Ponticeras (?), 26 undulata, Kosmoclymenia, 27 Unispirifer, 250 unisulcata, Atrypa, 181 Pentagonia, 181, 183 (fig.), 787, 188 (fig.) Valcourea, 115 Vermetus, 236 Vermicularia, 237 Vestinautilus sp., 28 (fig.) Vietor, 167 josephinae, 167, 170 (fig.) virginiana, Sobolewia, 26, 27 (fig.), 28 (fig.) wahlenbergi, Dideroceras, 103 wenlockensis, Triplesia, 149 Wewokites, 303 Wiedeyoceras, 303 Williamsoceras, 103 windomensis, Tubulibairdia, 169 woodlandensis, Triplesia, 149 Yunnanella, 203, 204, 208, 209, 211 hanburii, 206, 209, 215 (fig.), 216 (fig.) Zygospira, 14 o Publication in Smithsonian Contributions to Paleobiology Manuscripts for serial publications are accepted by the Smithsonian Institution Press, subject to substantive review, only through departments of the various Smithsonian museums. Non- Smithsonian authors should address inquiries to the appropriate department. If submission is invited, the following format requirements of the Press will govern the preparation of copy. (An instruction sheet for the preparation of illustrations is available from the Press on request.) Copy must be typewritten, double-spaced, on one side of standard white bond paper, with V/z" top and left margins, submitted in ribbon copy with a carbon or duplicate, and accom- panied by the original artwork. Duplicate copies of all material, including illustrations, should be retained by the author. There may be several paragraphs to a page, but each page should begin with a new paragraph. Number consecutively all pages, including title page, abstract, text, literature cited, legends, and tables. The minimum length is 30 pages of typescript and illus- trations. The title should be complete and clear for easy indexing by abstracting services. Taxonomic titles will carry a final line indicating the higher categories to which the taxon is referable: "(Ammonoidea: Goniatitidae)." Include an abstract as an introductory part of the text. Identify the author on the first page of text with an unnumbered footnote that includes his professional mailing address. A table of contents is optional. An index, if required, may be supplied by the author when he returns page proof. Two headings are used: (1) text heads (boldface in print) for major sections and chapters and (2) paragraph sideheads (caps and small caps in print) for subdivisions. Further headings may be worked out with the editor. In taxonomic keys, number only the first item of each couplet; if there is only one couplet, omit the number. For easy reference, number also the taxa and their corresponding headings throughout the text; do not incorporate page references in the key. In synonymy, use the short form (taxon, author, date, page) with a full reference at the end of the paper under "Literature Cited." Begin each taxon at the left margin with subsequent lines indented about three spaces. Within a taxon, use a period-dash (.—) to separate each reference. Enclose with square brackets any annotation in or at the end of the taxon. For references within the text, use the author-date system: "(Jones, 1910)" or "Jones (1910)." If the reference is expanded, abbreviate the data: "Jones (1910,p. 122,pi. 20: fig. 1)." Simple tabulations in the text (e.g., columns of data) may carry headings or not, but they should not contain rules. Formal tables must be submitted as pages separate from the text, and each table, no matter how large, should be pasted up as a single sheet of copy. Illustrations (line drawings, maps, photographs, shaded drawings) can be intermixed through- out the printed text. They will be termed Figures and should be numbered consecutively; how- ever, if a group of figures is treated as a single figure, the individual components should be indicated by lowercase italic letters on the illustration, in the legend, and in text references: "Figure 9b." If illustrations (usually tone photographs) are printed separately from the text as full pages on a different stock of paper, they will be termed Plates, and individual components should be lettered (Plate 9b) but may be numbered (Plate 9: figure 2). Never combine the numbering system of text illustrations with that of plate illustrations. Submit all legends on pages separate from the text and not attached to the artwork. In the bibliography (usually called "Literature Cited"), spell out book, journal, and article titles, using initial caps with all words except minor terms such as "and, of, the." (For capitali- zation of titles in foreign languages, follow the national practice of each language.) Underscore (for italics) book and journal titles. Use the colon-parentheses system for volume, number, and page citations: "10(2) :5—9." Spell out such words as "figures" and "plates" (or "pages" when used alone). For free copies of his own paper, a Smithsonian author should indicate his requirements on "Form 36" (submitted to the Press with the manuscript). A non-Smithsonian author will receive 50 free copies; order forms for quantities above this amount with instructions for payment will be supplied when page proof is forwarded.