SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY • NUMBER 90 Geology and Paleontology of the Lee Creek Mine, North Carolina, III Clayton E. Ray and David J. Bohaska EDITORS ISSUED MAY 112001 SMITHSONIAN INSTITUTION Smithsonian Institution Press Washington, D.C. 2001 ABSTRACT Ray, Clayton E., and David J. Bohaska, editors. Geology and Paleontology of the Lee Creek Mine, North Carolina, III. Smithsonian Contributions to Paleobiology, number 90, 365 pages, 127 figures, 45 plates, 32 tables, 2001.—This volume on the geology and paleontology of the Lee Creek Mine is the third of four to be dedicated to the late Remington Kellogg. It includes a prodromus and six papers on nonmammalian vertebrate paleontology. The prodromus con- tinues the historical theme of the introductions to volumes I and II, reviewing and resuscitat- ing additional early reports of Atlantic Coastal Plain fossils. Harry L. Fierstine identifies five species of the billfish family Istiophoridae from some 500 bones collected in the Yorktown Formation. These include the only record of Makairapurdyi Fierstine, the first fossil record of the genus Tetrapturus, specifically T. albidus Poey, the second fossil record of Istiophorus platypterus (Shaw and Nodder) and Makaira indica (Cuvier), and the first fossil record of/. platypterus, M. indica, M. nigricans Lacepede, and T. albidus from fossil deposits bordering the Atlantic Ocean. Robert W. Purdy and five coauthors identify 104 taxa from 52 families of cartilaginous and bony fishes from the Pungo River and Yorktown formations. The 10 teleosts and 44 selachians from the Pungo River Formation indicate correlation with the Burdigalian and Langhian stages. The 37 cartilaginous and 40 bony fishes, mostly from the Sunken Meadow member of the Yorktown Formation, are compatible with assignment to the early Pliocene planktonic foraminiferal zones N18 or N19. The Pungo River fish fauna is domi- nated by warm water taxa; the Yorktown fauna includes warm and cool water species. These changes are attributed to increased upwelling waters in Yorktown time. The abundant fossils provide the basis for several changes in selachian taxonomy and for two new species of bony fishes. George R. Zug records 11 taxa of turtles from the Yorktown Formation: a sideneck (Bothremys); six sea turtles (Caretta, IChelonia, Lepidochelys, Procolpochelys, Psephopho- rus, Syllomus); a softshell turtle (trionychid); two pond turtles (probably Pseudemys and Tra- chemys); and a giant tortoise (Geochelone). Albert C. Myrick, Jr., records the crocodylian Thecachampsa antiqua (Leidy) on the basis of fragmentary float material from the Pungo River or Yorktown Formation, or both. Robert W. Storer describes a new species of grebe of the genus Podiceps from the Yorktown Formation. Storrs L. Olson and Pamela C. Rasmussen record some 112 species of birds from the Pungo River and Yorktown formations. Apart from an undetermined number of shearwaters, only a few species are thought to come from the Pungo River Formation. The marine species from the Yorktown Formation include three loons, two grebes, five albatrosses, at least 16 shearwaters and petrels, one pelican, two pseudodontorns, three gannets, two cormorants, 9-11 auks and puffins, one skua, three jaegers, five gulls, two terns, and 20 ducks, geese, and swans. The less common land and shore birds are represented by 29 species, including three cranes, one rail, two oystercatchers, one plover, four scolopacids, one flamingo, one ibis, one heron, three storks, one condor, five accipitrids, one osprey, one phasianid, one turkey, one pigeon, and one crow. The fauna is dominated by a radiation of auks of the genus Alca. The early Pliocene fauna is very modem in aspect, suggesting that most modern lineages of birds were already in existence. Official PUBLICATION date is handstamped in a limited number of initial copies and is recorded in the Institution's annual report, Annals of the Smithsonian Institution. SERIES COVER DESIGN: The trilobite Phaecops rana Green. Library of Congress Cataloging-in-Publication Data Main entry under title: Geology and Paleontology of the Lee Creek Mine, North Carolina. (Smithsonian contributions to paleobiology ; no. 53- ) Includes bibliographies. 1. Geology, Stratigraphic—Tertiary—Addresses, essays, lectures. 2. Geology—Stratigraphic—Pleistocene— Addresses, essays, lectures. 3. Geology—North Carolina—Addresses, essays, lectures. 4. Paleontology—North Carolina—Addresses, essays, lectures. I. Ray, Clayton Edward. II. Series: Smithsonian contributions to paleobiology ; no. 53, etc. QE701.S56 no. 53, etc. 560s [551.7'8'09756] 82-600265 [QE691] © The paper used in this publication meets the minimum requirements of the American National Standard for Permanence of Paper for Printed Library Materials Z39.48—1984. Contents Page Prodromus, by Clayton E. Ray........................................ 1 Analysis and New Records of Billfish (Teleostei: Perciformes: Istio- phor1dae) from the yorktown formation, early pliocene of eastern North Carolina at Lee Creek Mine, by Harry L. Fierstine ............. 21 The Neogene Sharks, Rays, and Bony Fishes from Lee Creek Mine, Aurora, North Carolina, by Robert W. Purdy, Vincent P. Schneider, Shelton P. Applegate, Jack H. McLellan, Robert L. Meyer, and Bob H. Slaughter................................................ 71 Turtles of the Lee Creek Mine (Pliocene: North Carolina), by George R. Zug............................................ 203 Thecachampsa antiqua (Leidy, 1852) (Crocodylidae: Thoracosaurinae), from Fossil Marine Deposits at Lee Creek Mine, Aurora, North CAROLINA, USA, by Albert C. Myrick, Jr........................ 219 A New Pliocene Grebe from the Lee Creek Deposits, by Robert W. Storer ........................................................... 227 Miocene and Pliocene Birds from the Lee Creek Mine, North Carolina, by Storrs L. Olson and Pamela C. Rasmussen .......................... 233 ni Dedicated to Remington Kellogg 1892-1969 Geology and Paleontology of the Lee Creek Mine, North Carolina, III Prodromus Clayton E. Ray Prodromus... a preliminary publication or introductory work. Webster's Third New International Dictionary, 1964 The archaic title is intended to reflect the antiquarian nature of this paper and to emphasize my conviction that our work on the Lee Creek Mine project, a quarter century of effort by many people, is decidedly preliminary. Publication began with vol- ume I (Ray, 1983), which included papers on Remington Kellogg (to whom the series is dedicated), on the Lee Creek phosphate mine itself, and on stratigraphy and correlation, plants, and microfossils. The only paper specifically devoted to vertebrate fossils was that on otoliths of bony fish, included therein as "microfossils." That was primarily an unsuccessful effort to see the paper in print before the death of its senior au- thor, John Fitch, who was then terminally ill. Volume II (Ray, 1987) was devoted exclusively to mollusks, the most conspicu- ously abundant and well-preserved fossils in the mine. Initially, it was planned that all vertebrate fossils, other than otoliths, would be included in a third, concluding volume (Ray, 1983:3); however, subsequent productive collecting, especially that by able and devoted amateurs, has resulted in great accu- mulation of more and better fossils. These have been subjected to thorough research by the contributors and, combined, ex- pand the vertebrate papers beyond the reasonable confines of a single volume. The papers divide themselves conveniently into two sets, all groups other than mammals in this, volume III, to be followed by mammals, volume IV, which will include a tax- Clayton E. Ray, Department of Paleobiology, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560- 0121. onomic index to the publications of Remington Kellogg, pre- dominantly on mammals. This prefatory note continues the historical theme of those introducing volumes I and II, in which I attempted to review the early history of paleontological discovery and publication on the middle Atlantic Coastal Plain of British America. Hav- ing flattered myself that I had unearthed essentially everything, it is salutary to be reminded through several oversights that in antiquarian, as in paleontological, research one can never do too much digging. Returns in each are apt to be unpredictable and to be meager in relation to time invested (hardly "cost ef- fective"), but there will always be something new, and, to com- prehend it when found, one must be steeped in the subject. Thus, my primary objective is to rescue from obscurity or oblivion the additional early history that I have learned; not only to give credit to the pioneers, but to add to the foundation that may enable and inspire others to find out more, especially about American fossils surviving in European collections, and to dig further into the early literature. Thus, the present paper is an extension of those introducing volumes I and II and should be used in conjunction with them, as I have tried to avoid un- due repetition of text and literature cited. Although a full explication is beyond my scope herein and beyond my competence anywhere, I hope in reviewing these records to give some inkling of their importance, not only in the development of paleontology, but also in the broader intel- lectual concerns of the times. Fossils were more prominent in general scholarly discussions of the seventeenth and eighteenth centuries than at any time since. Although debate as to their na- SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY ture and significance has long since dropped from the forefront of general investigation, we have by no means won the day. In western culture many educated people, including scientists, ei- ther ignore fossils or reject them as evidence of organic evolu- tion, and humankind as a whole knows nothing of fossils (Mclver, 1992; Lemonick, 1996). No paleontologist can claim to be literate without thorough attention, not merely a perfunctory bow, to the role of fossils in western thought. Neglect of history is inexcusable in a histori- cal science such as paleontology, but its literature in English is very skimpy, and that written by practicing paleontologists is generally narrow and shallow (although not universally so; e.g., see Ward, 1990), much devoted to "correcting" past mis- takes and concepts in the light of present knowledge and fads. Of course we are obligated to correct objective errors in striv- ing to approach truth ever more closely, but this is less and less specifically useful as we delve deeper into the ontogeny of our science. Much more satisfying is to understand the thoughts on fossils in the context and constraints of the times and the rele- vance of those thoughts to subsequent developments. The best primers on this subject in English are Rudwick (1976), espe- cially the first 100 pages, and Drake (1996), the latter focused on Robert Hooke but with uncommon explication of context. Also very instructive are Challinor (1953), Morello (1979, 1981), and Young (1992). Davidson (2000:333) outlined the otherwise neglected role of Richard Verstegan in the early sev- enteenth century; however, she is mistaken in attributing the first published illustration, in 1605, of a shark's tooth to him. That distinction almost certainly belongs to Gesner (see Rud- wick, 1976:30, fig. 1.9), who in 1558 even included a modern comparative specimen alongside his fossil. Davidson (2000:343) cited Gesner's work as probably available to Ver- stegan but mentioned neither Gesner's figure nor the work of Kentmann of 1565 in Gesner (see Rudwick, 1976:11-17). These and other sources cited herein provide essential back- ground on the principal players in the founding of paleontolo- gy, including, among others, da Vinci, Colonna, Scilla, Steno, and Hooke, and those in the interrelated development of collec- tions, including Aldrovandi, Cospi, Giganti, Kircher, Mercati, Worm, and others. I refer to their work and its broader implica- tions only in the course of resurrecting the primary reports on American fossils. These allusions should be sufficient to show that these reports are not mere curiosities of antiquarian delight but were integral to cutting-edge (see Maienschein, 1994, re- garding this trendy term) intellectual concerns. Although there is no universal agreement as to what or when the Renaissance was, few would disagree that it was earlier, stronger, wider, and deeper in Italy than it was anywhere else. It is no accident that Italian names, notably those mentioned above, dominate the earliest stages in the history of paleontolo- gy and museums, and that Italian influence extended strongly into northern Europe and the British Isles. For example, Steno, or Niels Stensen, was a Danish cleric, but his scientific career was mostly Italian in locale, patronage, and material (Scherz, 1969, 1971); Olaus Worm, also Danish, probably was influenced by Aldrovandi in forming his museum (Schepelem, 1990:82); Aldrovandi's pioneering catalogs of his collection were emulated and cited frequently in much later catalogs in England (Grew, 1681; Sloane, see Thackray, 1994:125); and John Ray visited and was much impressed by Aldrovandi's collection (Torrens, 1985:206). Steno's work was immediately translated into English by Oldenburg, and it be- came the subject of great interest in the Royal Society (Eyles, 1958; Stokes, 1969:16). (Hooke accused Oldenburg and Steno of conspiring to plagiarize his ideas (Oldroyd, 1989:217); Drake (1996:116-117), especially, supported Hooke's claims, and, more importantly, documented his widely undersung con- tributions.) It has been suggested (e.g., Rudwick, 1976:39^11; Torrens, 1985:207) that recognition of fossils as remains of once-living organisms occurred in Italy before it did in northern Europe and England because the Italian fossils were "easy," being geo- logically young, little altered, and close to the sea and to living relatives, whereas those elsewhere were much older, in de- formed inland rocks, and the most conspicuous fossils were not closely related to living forms. Unfortunately, these factors can at best only partially explain away the Italian preeminence. Surely at least as important was the existence of an affluent so- ciety, with concomitant cultural sophistication, ready to under- write research and to accept truth through logical argument. Gould (1997) presented a convenient and timely analysis of Leonardo's brilliant and prescient insights on fossils, well fixed in the context of time and place. Both geologic and human his- tory preadapted Italy as the scene of these breakthroughs, and just as they were interwoven with a rich tapestry of culture, art, learning, and patronage, so also was the interrelated develop- ment of natural history collections. The literature in English re- veals little comprehension of the fact that natural history muse- ums developed (and survived in some cases) in continental Europe, especially in Italy, in some semblance of modem form, a century earlier than in the English-speaking world. It seems altogether too revealing that in 1995 I found the pages uncut in the Smithsonian Institution Library copy of MacGillivray's (1838) life of Aldrovandi. This neglect has been partially cor- rected in some excellent recent publications, including Impey and MacGregor (1985) and Findlen (1994). Ethnological and zoological objects from the Latin New World (then including Florida) have been well documented in these early collections (e.g., see Heikamp, 1976:458; Laurencich-Minelli, 1985), but to my knowledge no fossils have as yet been recognized. Nev- ertheless, the search for the beginnings of paleontology of the New World should begin in sixteenth century Italy, through di- rect examination of collections by appropriate specialists. The best hope might well be the collections of the great Ulisse (La- tinized as Ulyssis) Aldrovandi (1522-1605), who was known to have had a strong interest in the New World (Heikamp, 1976:458; Laurencich-Minelli and Serra, 1988). His catalogs, largely compiled during his lifetime but published posthu- mously (Ambrosinus, 1648; Figure 1), remained a powerful in- fluence long afterward in England (see above). NUMBER 90 onofonw F, FIGURE 1.—Title page (much reduced) of Aldrovandi's 1648 monumental catalog of his museum. It was com- piled and was widely known during his lifetime (1522-1605) but was published by Ambrosinus more than 40 years after Aldrovandi's death (Findlen, 1994:25). SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Some Early Records This brings me to the first instances to be added to early re- ports of fossils from the Atlantic coast of North America. As early as 1669 an allusion was made to natural history speci- mens from Virginia in a collection, long since forgotten and ir- retrievably lost, maintained by the East India Company at its headquarters in London (Hunter, 1985:162). The first explicit record to be added is that by Nehemiah Grew, who in 1681 published by subscription under auspices of the Royal Society a catalog of its collections, the title page of which is repro- duced herein (Figure 2; see Hunter, 1989, especially p. 142 et sqq., for genesis and nature of the catalog; see LeFanu, 1990, for Grew's life and contributions; see Clark, 1992, for an au- thoritative guide to histories of the society, its periodical publi- cations, and indices thereto). Included are two entries for fos- sils specifically stated to have come from the New World. A sort of MUSCUL1TES fill'd with Earth like Tobacco-Pipe Clay or Marie. Found amongst the earth of a Hill that was overturn'd at Kenebank in New England. (Grew, 1681:264) A great petrify'd SCALLOP. Figur'd by Ambrosinus {b) with the Name of Hippopectiniles. Given with several more of the same bigness, by Mr.______ Wicks. 'Tis half a foot over. Many of the same kind were taken out of a great Rock in Virginia, forty miles from Sea or River. (Grew, 1681:262) (b) Aldrov. Mus. Metall. (Grew, 1681:262, marginal citation) The first of these undoubtedly was a mussel shell, common in the late Pleistocene marine clays of the Presumpscot Forma- tion of coastal Maine, including the vicinity of Kennebunk (Stuiver and Borns, 1975; Thompson, 1982:212, 226). John Winthrop, Jr. (1606-1676), an original fellow and major con- tributor to the society's repository (Lyons, 1944:50, 64; Stearns, 1951:196,212,246, 1970:117-139), undoubtedly was the source of the specimen in question. In letters of 11 October 1670, printed in part in Birch (1756(2):473^t74) and quoted in part by Steams (1970:137), he alluded to "small shells" among the objects sent from a "hill near Kennebeck, Me, that turned over in summer last (June or July) into the River." The mysteri- ous "blowing-up" of the hill was reported also by John Josse- lyn (1674:210; see also White, 1956:180). The second entry is potentially of much greater interest. The marginal bibliographic citation is to Aldrovandi's monumental, classic illustrated catalog (Ambrosinus, 1648), which Grew cit- ed repeatedly, in this case alluding to a giant pecten illustrated on page 832 of volume 4. This raised the intriguing possibility that Aldrovandi's specimen might conceivably be a previously unsuspected and much earlier example from the New World. Unfortunately, my limited investigation to date has revealed no positive evidence that the giant pecten or any of Aldrovandi's fossils came from America; rather, Grew's allusion seems to be only an obsolete, broadly conceived synonymy, understandable for the time. The specimen has not been found among surviv- ing collections in Bologna, but it is thought to have come from the vicinity of the city (Sarti, in litt., 1993). Returning to Grew's specimens from Virginia, I had previ- ously been inclined to accept the argument that the specimen of giant pecten, Chesapecten jeffersonius, described and illustrat- ed by Lister, the first fossil so far known of any kind from the New World to be described and illustrated, probably had been collected by John Banister and sent directly to Lister, Petiver, or Sloane (see Ray, 1987:2), but now the Royal Society's Re- pository seems at least as likely. Not only was Lister's speci- men "half a foot over," but also Lister (1639-1712) and Grew (1641-1712) coincided in their activities in the Society (Hunt- er, 1994:188-189), and Lister is known to have used other specimens from the repository. The history of the repository is of great interest, not only in attempting to locate a potential historical treasure such as the giant pecten but also for its cautionary lessons to museologists in general. Early impetus to the establishment and support of the collection came from the need for a substantive rallying point for the struggling Royal Society and for a source of pub- lic prestige (Hunter, 1985, 1989:127, 128). Explicit and strik- ingly modem statements of the purposes of natural history col- lections were made by Grew (1681, preface), who advocated collections as an inventory of nature and as documentation of the ordinary, and by Hooke (1635-1703), who also took an ac- tive and at times official role in connection with the collections (see especially Hunter, 1989:125, 127, 139-141), and whose pioneering studies of fossil cephalopods stimulated his follow- ing statements (1705:338; also in Drake, 1996:236-237): And indeed it is not only in the description of this Species of Shells and Fishes, that a very great Defect or Imperfection may be found among Natural Histori- ans, but in the Description of most other things; so that without inspection of the things themselves, a Man is but a very little wiser.... It were therefore much to be wisht for and indeavoured that there might be made and kept in some Re- pository as full and compleat a Collection of all varieties of Natural Bodies as could be obtained, where an Inquirer might be able to have recourse, where he might peruse, and turn over, and spell, and read the Book of Nature, and ob- serve the Orlbographv. Etymologia, Synlaxis and Prosodia of Nature's Gram- mar, and by which, as with a Dictionary, he might readily turn to and find the true Figure, Composition, Derivation and Use of the Characters, Words, Phras- es and Sentences of Nature written with indelible, and most exact, and most ex- pressive Letters, without which Books it will be very difficult to be thoroughly a Literalus in the Language and Sense of Nature. The use of such a Collection is not for Divertisement, and Wonder, and Gazing, as 'tis for the most part thought and esteemed, and like Pictures for Children to admire and be pleased with, but for the most serious and diligent study of the most able Proficient in Natural Philosophy. And upon this occasion tho' it be a digression, I could heartily wish that a Collection were made in this Repository of as many variet- ies as could be produced of these kinds of Fossile-Shells and Petrifactions, which would be no very difficult matter to be done if anyone made it his care. Despite these and other resounding statements within the so- ciety, the reality (dictated largely by its dilettante membership) was that its collection continued to be much like that of a pri- vate cabinet of curiosities—devoted to the rare and bizarre rather than being a microcosm of what exists, ordinary as well as extraordinary (Hunter, 1989:150). This tension has yet to be resolved in museums, although the "inventory of nature" movement seems to be gaining ascendancy at last. Further, the society found that although establishing a museum is easy, maintaining it in the long term ("perpetuity") is almost impos- sible. From the beginning, much of the society's attention was NUMBER 90 MVSMVM REGALIS SOCIETATIS. O R A Catalogue &c Defcription Of the Natural and Artificial RARITIES Belonging to the ROYAL SOCIETY And preferved at Grefliam Colledge. MADE BytJS(ehemjab (jrewM. D. Fellow of the Royal Society, and of the Colledge of Thyfitiam. Whereunto is Subjoyned the Comparative Anatomy O F ^tomatljs ana (guts. "By the fame oJVTHOR. L0&CJD0N, Printed by W. prelim, for the Author, i68i. FIGURE 2.—Title page (reduced) of Grew's 1681 catalog of the repository of the Royal Society. SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY usurped by monetary problems, including difficulty in collect- ing dues and shortfall in funds to pay support staff. After years of vicissitudes in the care of its collections that entailed repeat- ed efforts at revitalization, and finally faced with a critical problem in space to house the collection, the society offered its collections to the British Museum in 1779 (Hunter, 1989:153-155), which accepted in 1781 (Lyons, 1944:211). Now, a mere two centuries later, the heir to that collection, The Natural History Museum (BMNH), London, the Gibralter of its kind, itself faces an uncertain future (e.g., see Nature. 1990), as do its counterparts elsewhere (Trescott, 1996; Butler, 1997). In any case, the "several" giant pecten(s) from Virginia should have gone in 1781 to the British Museum. Although it seems unlikely that such large, conspicuous shells would have been lost, sold, or destroyed, even in the vandalous "crema- tions" of curator Shaw (Steam, 1981:17), they have not as yet been recognized in the existing collections of BMNH (Nuttall, inlitt., 1993). The specimens in all probability represent Chesapecten jef- fersonius (see Ray, 1987), since 1993 the official fossil of the Commonwealth of Virginia (Anonymous, 1993) and thus of great historic and current interest if found. This leaves only the matter of the donor, "Mr._____Wicks," who turns out to be a subject of specific and general interest in spite of the paucity of information about him. The person in question undoubtedly is Michael Wicks, clerk of the Royal So- ciety for at least 20 years (Thomson, 1812:15, lists his years in office as 23), from the first meeting of the council on 13 May 1663 (Birch, 1756(1):236) at least until 27 November 1683, when it was resolved that "Mr. Cramer be clerk to the society in Wicks's place" and that "Mr. Wicks be told, that his atten- dance is of no farther use" (Birch, 1757(4):229). This resolu- tion seems however not to have had the finality that it implied, as Mr. Wicks was given orders at the meeting of 2 April 1684, and the treasurer was ordered on 14 January 1685 to pay him "fifteen pounds for a year and a half's salary" (Birch, 1757(4):277, 355). Robinson (1946:194-195) gave a summary of Wicks' employment by the society, indicating that the last mention of him is that of 13 November 1695, when a gratuity was voted him by the Council; however, Hunter (1994:235) noted a substantial payment to him as late as 1696. It should be noted that Robinson refers to Wicks as "Weeks," that he ap- pears as "John Weeks" in Weld (1848:562), secondarily as "Weekes" in Hunter (1994:235), and is omitted altogether by Lyons (1940:344). Apparently prior to Wicks' appointment with the society, Dr. Jonathan Goddard (1617-1675, professor at Gresham College) had employed "Mr. Mich. Weekes, who looked to his stills" (Aubrey, 1898(1 ):268). In this case, the stills were for produc- tion of ingredients to various secret medicinal nostra. It is thought that Wicks got the job as clerk through Goddard's in- fluence (Robinson, 1946:194). This seems plausible in view of Goddard's major role in the birth and early development of the society, from its unchartered gestation, beginning in 1645 (Copeman, 1960; McKie, 1960), through the turbulent period of the Civil War, Commonwealth, and Protectorate (inauspi- cious for the founding of anything "Royal"). In John Aubrey's (1626-1697) notes (dated 12 March 1689) for his brief life of Walter Raleigh, in connection with Ra- leigh's role in introducing tobacco to England, he states (Aubrey, 1898(2): 181-182), Mr. [Michael] Weekes, register" of the Royal Society and an officer of the cus- tome-house, does assure me that the customes of tobacco over all England is four hundred thousand pounds per annum. aSubst. for 'clerk.' There can be no doubt that Weekes and Michael Wicks are one and the same person. In response to my queries regarding Wicks, Mary Sampson (pers. comm., 1993), archivist to the Royal Society, found only one written communication by Wicks in the society's unpublished Classified Papers series (CL. P. XXIV.56), a brief undated note of some 13 lines, ad- dressed to Henry Oldenburg (his boss). I have been unable to decipher the handwritten note entirely, but the gist of it is that he put out some papers for Oldenburg stating, "I am sorry I could not wait upon you sooner, my business at Custome House being much more than ordinary." In 1993, Gillian Hughes, an independent researcher, under- took on my behalf a preliminary search in the Public Record Office for evidence of Michael Wicks in the Customs Estab- lishment. The earliest certain indication found by her lists Michael Wicks as Receiver for the Plantations among officers of his majesty's customes for 1673 and 1675 (PRO 30/32/15 and 17), and his name was last seen in those lists for 1693 (PRO, CUST 18/28). In the published Calendar of Treasury Books (Shaw, 1935:584), allusion is made under the date 17 April 1694 to "Mich. Wicks, late Receiver of the Plantation Duties and of the new impositions on tobacco and sugar... lately discharged from that service." The Calendar of Treasury Papers (Redington, 1868:338) indicates "confusion in the accounts of Mr. Wicks," and the Commissioners of Cus- toms "describe their perplexities about his accounts, and that to prevent further enlargement they had dispensed with his atten- dance at the Custom House__Dated 5 Jan. 1693 [now 1694]." Thus it seems clear that Michael Wicks (up to his dismissal under a cloud) was in an unusually favorable position for di- rect, frequent communication with merchant ships sailing to and from British America. At the meeting of the Royal Society on 13 June 1683, "Mr. WICKS was desired to procure from the East-India ships a quantity of the shining sand of St. Christo- pher's and James river in Virginia" (Birch, 1757(4):209). This request would hardly have been made had it not been anticipat- ed that Wicks could accommodate it. Interestingly, this is the only instance in the long employ- ment of Wicks by the society in which he was "desired" to do something, rather than "ordered" or "directed." This is proba- bly not accidental, but reflects a momentary deference to his position with the custom house. Otherwise, paid subordinates were addressed in the imperative, whereas the gentlemen Fel- NUMBER 90 lows were "requested" or "desired" to do something. Pumfrey (1991:12-16; Drake, 1996:17-18, 104-105, not withstanding) has made a persuasive case for this distinction in connection with his study of Robert Hooke's precarious position betwixt and between, which may have contributed to Hooke's appar- ently atypical egalitarian attitude toward subordinates (e.g., see Shapin, 1989:269), as well as to his prickly attitude toward the establishment. It should be recognized that the society's treatment of Wicks was not cruel and unusual but was in general correct for the so- cial system of the time and place. Even allowing for the free- wheeling attitude toward spelling in those days, it apparently was not important to get his name right or even to include it consistently in society records, nor perhaps was it important for Grew to remember or later insert Wicks's given name in the manuscript for the society's catalog. The next explicit report of specimens from the Atlantic Coastal Plain is that of Sloane (1697). He borrowed the speci- mens from his friend, Dr. Tancred Robinson, who had just re- ceived them from Maryland (most likely from the Rev. Hugh Jones, who arrived there in 1696 and was accused by Wood- ward of sending specimens to "rogues and rascalls," including Sloane, Petiver, Lister, and Robinson (Stearns, 1952:292, 306)). These specimens included at least three isolated tooth plates of the ray Aetobatis, illustrated in Sloane's figs. 7-12; it is unclear from his text whether the articulated partial tooth battery shown in his figs. 13 and 14 also is from Maryland (see Figure 3). At least the fragmental plate shown in his figs. 7 and 10 is among the very small number of the founder's specimens known to survive in BMNH, where it was featured in an exhi- bition on the history of paleontology (Edwards, 1931:61) and where it apparently is still to be found (Thackray, 1994:132). Obviously Robinson must have allowed Sloane to retain at least one of the fossils. Some or all of the others may be pre- served in Woodward's collection at Cambridge. In an appendix to his primary catalog of English "extraneous" fossils (catalog B of Price, 1989:94), however, Woodward (1728-1729) listed modem specimens preserved for comparison to his fossils, and on page 111, under his entry number 25, a modern ray denti- tion, he expressly stated that his ray tooth plates sent by Jones from Maryland "were digg'd up, together with those" reported by Sloane (see catalog B of Price, 1989). Sir Hans Sloane (1650-1753) is best known as a prodigious collector who provided the foundation for the collections of the British Museum and its offshoots. He also was a man of parts who was a successful doctor of medicine, an olympian letter writer, and a major force in the Royal Society, although he was not without his detractors, most notably John Woodward (e.g., see MacGregor, 1994:19). Most have made light of his abilities as a thinker and researcher. Nevertheless, his little paper on the fossil ray plates is an elegant example of modernity produced before any pattern was established. He placed the isolated un- knowns (considered by some to be bits of petrified mush- rooms) alongside the most appropriate specimens of known identity, articulated and disarticulated modem ray tooth batter- ies, found them to be similar in detail, illustrated them accu- rately in comparable orientations, and concluded that they de- rived from identical or closely related organisms. One's first impulse today might be to dismiss this approach as routine, but it was not such in the context of the time. Although spectacular examples of brilliant comparative methodology are known here and there from the sixteenth century onward (note that Grew used the approach and the term, "comparative anatomy," in 1681, see Figure 2), the techniques were not codified and uni- versally applied until the nineteenth century under the influ- ence of Cuvier, Owen, and Agassiz. This could not have oc- curred prior to the Age of Enlightenment/Reason, with the spread of the notion that all problems could be successfully solved through intensive inspection and that ordinary humans could rely on their own careful observations irrespective of au- thority. This approach was the cornerstone of the Royal Soci- ety. Until recently, this reliance was taken for granted, so much so that the sublime notion could be expressed profanely, if I may be permitted one homely example: Remington Kellogg, once asked by a colleague what criteria allowed him to con- clude that a certain fragmentary whale vertebra was in fact identifiable to a particular species, immediately replied re- soundingly, "because it looks like it, goddamit it!" He did not live to experience the postmodern entry of doubt introduced by phylogenetic systematics and social constructivism, in which we question the meaning of all our observations. Further, Sloane's (1697) note was written when the nature of fossils as vestiges of once-living organisms had by no means been universally accepted by serious scholars. This topic brings us conveniently to the next known report and illustration of fossils from the Atlantic Coastal Plain, that of a bone fragment and a shark tooth by Scheuchzer (1708), whose title page and figures are reproduced herein (Figures 4-6). Both specimens are preserved in the Paleontological Museum of Zurich (Leu, in litt., 1997; see also Furrer and Lev, 1998:33). Johann Jacob Scheuchzer (1672-1733), little familiar to the English-speaking scientific community, is known primarily as an object of derision for his Homo diluvii test is faux pas, based on a fossil salamander (see Jahn, 1969, for this and for a good thumbnail biography of Scheuchzer in English). Scheuchzer was actually a very substantial scientist who translated Wood- ward into Latin and promoted his ideas (notably organic origin of fossils) on the continent. Scheuchzer was in close contact with Sloane and other leading naturalists of the Royal Society. His historical significance would unquestionably be better ap- preciated had Jahn's (1975) promised bio-bibliography of Scheuchzer and translations of his major works materialized. The bone fragment from Maryland was attributed by Scheuchzer (1708:22) to the acetabular region of the innomi- nate bone of a mammal ("Animalis"), not a farfetched supposi- tion. This fragment, however, matches very well the portion of a small cetacean atlas vertebra that characteristically remains after the vertebra breaks at the weak points and rolls on the beach; it is illustrated (Figure 5) alongside a typical float speci- men and a well-preserved atlas from the Miocene of Chesa- SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY !Philojoz>h- (Transact- A~°. 2312 FIGURE 3.—Plate accompanying Sloane's 1697 report on fossil ray teeth from Maryland (xl). His figures 1-6 are of modern species, figures 7-12 are of fossil species from Maryland, and figures 13-14 are of fossils from an unknown locality, possibly Maryland. peake Bay in Maryland. This is probably the first cetacean (and mammalian) fossil from America to be illustrated. This Maryland specimen, especially if received by Scheuchzer from Petiver, probably was sent by Hugh Jones. Lhwyd (1660-1709) complained that Petiver and his pal Doo- dy got aboard ship and rifled collections from Jones intended for him (Gunther, 1945:343,462). Among specimens cataloged by Sloane that came to him in Petiver's collection were "shark NUMBER 90 PISCIUM QUERELAE ET VINDICIAE Expofit* Johanne Jacobo Scheuchzero Med. D> Acad. Lcopoldin. & So- cietatum Regg. Anglican, ac Pruk ficae Membro. ¦ TIGURI. SumtibusAuthoris, Typis Geffnerianis.. — M D feCr ¥llk.---- • \10S, FIGURE 4.—Title page of Scheuchzer's 1708 classic, Piscium Querelae et Vindiciae (x 1). teeth and other fossils sent from Maryland by the Revd Hugh his job as chaplain to the governor of Maryland through the ini- Jones" (Thackray, 1994:126). Jones communicated especially tial recommendation of Lhwyd, furthered by the Temple Cof- with Petiver and sent specimens from Maryland at least from fee House group that included Sloane, Petiver, Doody, Lister, 1696 to 1702, although he became ill in 1700. Jones had gotten and Robinson (Steams, 1952:292-294; Jessop, 1989). 10 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE 5.—Cetacean atlas vertebrae in cranial aspect (x0.82), a. Scheuchzer's figure reproduced; b. similar waterworn fragment, Lophocetus sp., probably collected on beach in Calvert County, Maryland, USNM 449525 (National Museum of Natural History, Smithsonian Institution, which houses collections of the former United States National Museum); c. complete atlas. Cove Point, Calvert County, Maryland, St. Marys Forma- tion, collected by Francis Markoe, Jr., 1841, part of the holotype oi Lophocetus calvertensis (Harlan), itself a historic specimen (Gilmore, 1941:311-312, 377; Simpson, 1942:162, 176). (Scale bar=l cm.) The second specimen, an incomplete tooth of Carcharodon megalodon Agassiz from the Carolinas (Figure 6a), is de- scribed by Scheuchzer (1708:20) as lacking serrations. The ab- sence of serrations is of no taxonomic significance because the tooth is clearly waterworn and is typical of the rolled speci- mens so abundant in the lower reaches of several rivers in South Carolina. Scheuchzer's comparison was to Luid number 1259, a simi- larly waterworn specimen from the British Crag (Figure 6c). The number refers to the collection of Edward Lhwyd, Lati- nized as Luid, among the many variations of the surname (see Gunther, 1945:vii) (see Roberts, 1989, for a succinct biogra- phy), whose specimen survives in the geological collections at the University Museum, Oxford (Powell, in litt., 1993). Scheuchzer's inferred outline of his incomplete specimen is a very early example of paleontological restoration, however modest. Jacob (or James) Petiver (71663-1718), identified as the do- nor, was a London pharmacist and perhaps second only to Sloane as a natural history collector and letter writer (see Stearns, 1952, for the fullest account of Petiver). Of course, Lhwyd and Woodward outdid Petiver in their geological col- lections (Torrens, 1985). Petiver's most productive correspondent in South Carolina was the Rev. Joseph Lord, who began sending him speci- mens in 1701 and continued at least until 1713 (Stearns, 1952:346, 362). Especially relevant may be Petiver's (1705:1960) account of two fossil shark teeth sent by Lord NUMBER 90 11 FIGURE 6.—Teeth of Carcharodon megalodon (x0.82), a, Scheuchzer's figure reproduced; b, well-preserved tooth, showing serrations, collected by P.J. Harmatuk, from Yorktown Formation spoil, Lee Creek Mine, North Carolina, USNM 350941; c, waterworn specimen, probably from the bone bed at the base of the Red Crag, Suf- folk, England (H.P. Powell, in litt., 1993), Lhwyd collection number 1259, Oxford University Museum, photo- graph courtesy of Oxford University Museum of Natural History. (Scale bar=I cm.) 12 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY the first of which could possibly be the very one illustrated by Scheuchzer. Here may be mentioned what is probably the earliest allusion to a fossil of a bony fish from the New World, other than Win- throp's 1636 allusion to "Fishes' bones" from the James River (Ray, 1983:4). In a letter to Petiver from his home at Dorches- ter (this would have been old Dorchester on the Ashley River), South Carolina, dated 1 September 1707, Lord writes: Herewith comes a small box with divers Fossils.... In a part of my land where some were digging after a sort of Marl... a stone was digged, somewhat flat & broad, but looking like ye marl among which it lay, on which was ye tail of a small fish & ye body near as far as ye Navel, of a brown colour, shewing fins & scales very apparently, but all Stone; & it seemed so distinct that I had a con- ceit I might separate it from ye rest of ye Stone, which I endeavoured to no pur- pose, but in trying much defaced ye Impress; & since, only lying in my Study made it more obscure: but however I have put it into ye Box. The letter was marked as received on 26 January 1708, as was presumably the accompanying small box. The original let- ter is preserved among the Sloane manuscripts (Sloane 4064, folio 150) in the British Library (permission to quote not re- quired; Taylor, in litt., 25 April 1994) as a result of Sloane's having purchased Petiver's papers and collections after the lat- ter's death (Steams, 1952:244; MacGregor, 1994:23). The fos- sil fish should have been among the Sloane specimens that ini- tiated the British Museum, but if it survives in the BMNH, it has not as yet been recognized (Thackray, 1994:132). In striking contrast to the loss or unknown fate of most North American fossils from the colonial era is the survival of those in the Scheuchzer collection and in the incomparable collection of his correspondent, John Woodward (1665-1728), preserved essentially intact, with data, at the University of Cambridge. The essential background to this collection can be learned from Woodward's own catalogs (Woodward, 1728-1729), Gunther (1937:424^133), and especially Price (1989). A measure of its volume and significance can be gained from the catalog entries. The North American fossils are contained in catalogs K and M of Price's notation (1989:93-94; table 1). Of 655 catalog en- tries for foreign fossils (only fossils in the modem sense, ex- cluding rocks and minerals), 74 are North American; these rep- resent a minimum of 127 of the total 1210 specimens. Thus, the North American material constitutes more than 10 percent of both total catalog numbers and specimens. Of the 127 speci- mens, 74 are invertebrates, mostly mollusks, and 53 are verte- brates, mostly sharks' teeth. Of the 74 catalog entries, 51 are from Maryland and 23 are from Virginia. Among the Maryland entries, at least 27 are attributed to William Vernon, 18 to Hugh Jones, and three to David Krieg, the three most important names in seventeenth-century natural history collecting in Maryland. Although focused primarily on their botanical col- lecting, the account of their activities in Maryland by Frick et al. (1987) is a convenient and authoritative source (see also Stearns, 1970:264-274). Jones, as previously noted, spanned the years 1696-1702 but was largely incapacitated for the last two. Vernon and Krieg overlapped almost exactly in their brief visits, during the spring and summer of 1698. There was keen interest and competition, in part unfriendly, among British nat- uralists for the specimens. Woodward, generally at odds with most of his contemporaries, boasted that "Mr. Doody had given him all or the greatest part of those fossils you [Jones] sent him" (Petiver to Jones, 10 March 1698; quoted in Frick et al., 1987:19; see also Steams, 1970:265). Of three catalog entries for Virginia specimens attributed to John Banister, two are explicitly stated to have been given by Doody, clearly a continuing benefactor of Woodward. Banis- ter's collecting could not have been later than 1692, the year of his death (see Ewan and Ewan, 1970, for a definitive account of Banister). All of the 20 North American entries (19 mollusks, 1 barna- cle) in Woodward's additional list (catalog M of Price, 1989) pertain to what was probably a single locality near the James River, 20 miles (-32 km) above its mouth. One specimen was found "by Lyons-Creek" (now Lawnes Creek, reverting to the place names of Christopher Lawne's Plantation, established in 1619), which empties into the James River just below Hog Is- land, opposite Williamsburg, some 20 miles (-32 km) up the James. All are attributed to a "Mr. Miller," who is probably the Mr. Miller described by Heame (Salter, 1915:148) as Wood- ward's "neighbor & particular Acquaintance for 30 years past, who often went abroad with him to gather Fossils, and assisted him often in packing up boxes, to be sent abroad to Professors & curious persons, & presented him himself with a Drawer or two from the West Indies." With the possible exception of the Miller specimens from the James River, all North American specimens in the Woodward collection were collected prior to 1700. Judging from the iden- tity of the collectors, their time, and Woodward's annotations, it seems highly probable that some of the specimens may have been studied or illustrated by Banister, Lister, Sloane, or con- temporaries. Price's valuable studies, cut short by his untimely death, were only just beginning to reveal the value of this unique resource, and it has not been feasible to examine the collection firsthand for the present project. Close study of the specimens with relevant literature at hand could scarcely fail to yield interesting results. Some of the shark teeth have been ex- amined recently by Shelton P. Applegate of the Universidad Nacional Autonoma de Mexico. The majority of the remaining entries for foreign fossils in Woodward's catalogs (catalogs K and M of Price, 1989) are from the extremely important collections of Scilla and Scheuchzer. Woodward appears to have been meticulous in cit- ing their specimens, but as yet none of his entries for them can be identified as pertaining to North American specimens. The collection should, of course, be searched for them. The next instance of early collecting that I wish to note is from a letter to Peter Collinson from John Custis of Williams- burg, Virginia, believed to have been written on 28 August 1737. In it Custis alludes to the extreme drought of that sum- mer, which necessitated his digging a deep well to water his garden. The letter is quoted in part from Swem (1957:47): As you are a very curious gentleman I send you some things which I took out of the bottom of A well 40 feet deep; The one seems to bee a cockle petre- NUMBER 90 13 fyd one a bone petrefyd; [this?] seems to have been the under beak of some large antediluvian fowl. Wish they may bee acceptable. In a letter of 5 December 1737, Collinson thanked Custis for "the Curious Fossils that you sent Mee last year" (Swem, 1957:60); again, in a letter of 5 March 1741 (Swem, 1957:71), Collinson alludes to fossils sent by Custis as "shells that was found so Deep when you was Makeing the Mill Dam." At least some of these fossils were on exhibit at Mill Hill School, on the site of Collinson's home, near London, in the early 1930s, but they have been lost sight of since (Swem, 1957:172; Hume, 1994:22). Interestingly, the 1964 archaeological reexcavation of Custis' 40-foot (-12 m) well in Colonial Williamsburg, Vir- ginia, yielded fossil shells and whale bones (Hume, 1994:20, 22). All of these specimens undoubtedly derive from the Pliocene Yorktown Formation. John Custis (1678-1749), educated in England, was a promi- nent citizen of Virginia and an avid horticulturist, which led to his association with Peter Collinson (Swem, 1957:11-20). Col- linson (1694-1768) was a successful business man with exten- sive interests in the American colonies, including a lifelong av- ocation to botany (Swem, 1957:1-9). He was singled out by Stearns (1951:194-195) as one of the most active fellows of the Royal Society in encouraging North American naturalists. He is perhaps best known in North America in connection with the vertebrate fossils of Big Bone Lick, Kentucky (Jillson, 1936; Simpson, 1943). Although Collinson was especially ac- tive in adding to Sloane's collection (Swem, 1957:3), no evi- dence has yet emerged to identify any fossils from Custis' dig- ging in Williamsburg in the surviving collections of BMNH. Lastly, although much later than the other reports cited here- in, I wish to supplement my earlier account (Ray, 1983:6-7) of Latrobe's 1799 report of vertebrate fossils from Richmond, Virginia, including sharks' teeth, fish vertebrae, a large bird fe- mur, and a partial porpoise flipper. Latrobe (1809:283-284) re- turned to this subject as follows: It was my intention then, to have offered to the [American Philosophical] Soci- ety, a series of geological papers, the materials of which I had collected, and of which this memoir [Latrobe, 1799] was the first. But my intention was delayed and partly defeated by the loss of a very large collection of all the principal fos- sils, necessary to elucidate my observations, in their passage by water, from Fredericksburg to Philadelphia.—This collection, intended for the American Philosophical Society, was made by the industry of my excellent friends, Mr. William Maclure now at Paris, of the late Dr. Scandella whose untimely death in 1798 science and friendship equally have to deplore, and of myself.—It con- sisted of specimens of loose and undecayed fossil shells, found on and near the surface, from the coast to the falls of the rivers of Virginia, of the shell rocks of York river, of the clays with impressions of shells in every fracture, but which shew no remaining evidence of any calcareous matter when subjected to chem- ical tests; of the exuviae of sea animals', bones of fishes, sharks' teeth, marsh mud, fossil wood and coral rock, dug from the deep wells about Richmond, of the marles of Pamunkey and Mattapony, of all the strata of the coal mines on James's river, of the varieties of the granite of Virginia, of the free stone of James's river and the Rappahannoc, with the vegetable petrefactions and coal belong to it; and of a variety of miscellaneous fossils. ...The loss of this collec- tion dispirited me, and the occupations of a most labourious profession de- prived me of time. "Drawings of some of the exuviae accompanied my memoir, to which refer.— The bones of the foot there represented, are probably those of a sea tortoise.... Had those collections survived and become available for re- search in Philadelphia, paleontology of the Atlantic Coastal Plain might have been advanced by some decades. In the same report Latrobe went on to discuss other geologic phenomena including delineation of the fall line and its significance in rela- tion to building stones. He was a practical man whose job at that time was "Surveyor of the Public buildings of the U. States," (Latrobe, 1809:293), which makes his closing observa- tion (Latrobe, 1809:292) regarding the geologic problems dis- cussed all the more revealing: It is fortunate that the solution of these aenigmas of nature are of no conse- quence whatever to our happiness, or of use to our enjoyments.—But the plea- sures of investigation, and of wonder, the offspring of ignorance, are not with- out a charm, which often entices the mere speculative philosopher into researches that produce results beneficial to mankind. We continue to vacillate in the unresolved and unresolvable stress between applied and pure research. In the most recent cy- cle, support for pure research probably reached a peak in the expansive mood of prosperity during the 1960s, when science could save us. We may hope that the retrenchment of the 1990s, with its demand for quick returns and the rise of pseu- doscience, is the nadir of the curve and not the precipitous slope of descent into continuing decay and rejection of science (see Sagan, 1995, especially chapters 14, 23, and 25, and Gross et al., 1996, for timely, accessible examinations of the problem; see Maull, 1997, for an example of the widespread and disas- trous confounding of science and scientists by social construe - tivists). Conclusions Review of these additional early publications on fossils from the Atlantic Coastal Plain leads to a few observations of seem- ingly wider relevance. These may be grouped conveniently for present purposes under the topics of "Firsts" for North America and for paleontology and of "Sharks' Teeth." Firsts.—Simpson (1942, 1943) was among the very few practicing vertebrate paleontologists in the modem era to have looked seriously into the early history of the subject in North America. Here, too, should be mentioned the historical re- search by Helen Ann Warren, under the aegis of Henry Fair- field Osborn (in Osbom, 1931 :ix, 1-33), which was similar in content and emphasis, if not in depth, to that of Simpson. Simpson was more than casually involved with preparation of the book (on Edward Drinker Cope) of which Warren's work was part, overlapped completely with her at the American Mu- seum of Natural History (Osborn, 1931:ix), and may have re- lied too heavily on her spadework. Be that as it may, he brought together a great deal of scattered information and quite correctly contrasted casual or inconsequent early finds (such as those by early Indians—interesting but not contributing to sci- ence) with those that were to become factors in the advance- ment of knowledge in western culture. In his words, "true dis- covery [is] that leading by a traceable route, however devious, to eventual elucidation of the problems concerned" (Simpson, 14 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY 1942:135), and again, "merely seeing a fossil bone or picking it up in idle curiosity is hardly discovery__scientific discovery was that which initiated continuous consciousness and record of the occurrence of fossil animals in America and had the first scientific studies as its sequel" (Simpson, 1943:26-27). Although these definitions are meaningful, Simpson was mistaken in every instance in applying them toward identifica- tion of firsts for North America, thus making his papers not the definitive work that he supposed (Simpson, 1943:26). Taking caution from his example, I do not propose that my candidates are in truth firsts, only that they are the earliest known to me (as indicated, I suspect, even hope, that there are still earlier ones, especially Italian, and thus I believe that the present ac- count is not the last word). It must be emphasized that Simpson provides a large target only because he had the rare insight to see the value of history and the ability to draw so much togeth- er from scratch. His well-earned stature and authority make doubly important the correction of his objective errors. Further, those errors reflect what I believe to be a pervasive lack of comprehension among American paleontologists of the sophis- ticated nature of natural history investigations by western Eu- ropeans in the late seventeenth and early eighteenth centuries. Simpson (1942:131) defined six periods in the history of ver- tebrate paleontology in America, the first two of which are of interest here. 1. Pre-scientific Period.—From the earliest times to about 1762. The first fossil discoveries were made. Toward the end of the period bones were collect- ed and sent to Europe. No truly scientific study of them had been made. 2. Proto-scienlific Period.—From about 1762 to about 1799. In 1762 Daubenton read a paper on American fossils treating them for the first time in what deserves to be called a scientific way. In reference to Lord's 1707 letter to Petiver about the fossil fish tail (see "Some Early Records," above) Simpson stated (1942:135), "The incident is...unique for its date, and for a long time there after, in involving a small fossil vertebrate. Most of the eighteenth century naturalists overlooked bones of animals smaller than the mastodon...." Simpson (1943:27) re- garded letters from Cotton Mather as the "first publication on American fossil vertebrates" (published in 1714 in the Royal Society's Philosophical Transactions), and Simpson thought they probably were based on mastodon remains. In allusion to Catesby's 1743 report of African slaves' recognition of fossil proboscidean teeth, Simpson (1942:134) credited them with the "first technical identification of an American fossil vertebrate," assuming the incident to have occurred prior to 1739. Based on the collection from Big Bone Lick, primarily of mastodon re- mains, Simpson stated (1942:135), "If Columbus discovered America in 1492, Charles Le Moyne, second Baron de Longueuil, discovered American fossil vertebrates in 1739." Simpson (1942:144-145) added that "Guettard (1756, read in 1752) published the first illustration of an American vertebrate fossil... [and] a decade later Daubenton (1764, read in 1762)... [provided]... an excellent example of the comparative method... one of the four most basic... principles in the rise of vertebrate paleontology and it may fairly be dated from Daubenton...." Both Guettard's and Daubenton's contribu- tions stemmed from the 1739 Longueuil collection of mast- odon remains. Both Sloane (1697) and Scheuchzer (1708) conspicuously antedate Guettard for the first description and illustration of North American fossil vertebrates. Sloane's paper in particular meets every possible criterion: the fossils reported were col- lected through a purposive scientific program (about which more beyond); Sloane was among the most prominent natural historians of his or any other era; he published in the premier scientific journal in English; his title alone reveals the signifi- cance of his subject; the specimens are small, and at least one survives today; and the paper is a model of comparative meth- odology. The larger point to be emphasized is the nature of the natural history enterprise in western Europe in the late seventeenth and early eighteenth centuries, for present purposes especially in England, and especially centered among fellows of the Royal Society. Their sustained, intensive, extensive interest in North America is well recognized and is woven into the modem liter- ature of zoology and especially of botany (Steams, 1970, and Frick et al., 1987, are superb examples) but is reflected hardly at all in that of paleontology (among notable exceptions is Ger- mon et al., 1987), especially of vertebrates. There was nothing in the least casual or chancy in the collec- tion of North American fossils; rather, they resulted from a pur- posive campaign. In fact, it is a little surprising that the results were so meager for fossils in light of the effort expended. Much of the voluminous correspondence of Sloane, Petiver, Woodward, and others was devoted to creating and maintain- ing a network of collectors, not least in the New World. A very good taste of the flavor of time and place can be had from Stearns' (1952:293-303) account of how the group cooked up a collector in cleric's clothing. The Bishop of Lon- don, in 1694, sought advice from Martin Lister in recommend- ing a candidate for chaplain to the governor of Maryland. This eventuated in Edward Lhwyd's putting forward his assistant, Hugh Jones, whose specific qualification was that he would be a worthy successor to John Banister. Jones was groomed in nat- ural history, run hastily through religious orders, and rushed off to Maryland. Besides Lister and Lhwyd, James Petiver, Samuel Doody, Jacob Bobart, and Tancred Robinson are known to have had specific roles in the care and feeding of Jones; Petiver quite literally—besides equipment, supplies, and literature, he sent Jones a Cheshire cheese and English beer, plus medicine and medical advice (Steams, 1952:297, 299, 303). John Woodward (1696) provided "brief instructions" to geo- logical collectors (see Eyles, 1971:403; Price, 1989:93, foot- note 7). Petiver also prepared instructions, which were sent out with travellers and to correspondents. These were highly so- phisticated, even to the point of recommending the stomach contents of sharks, and other great fish, as a source of "divers strange animals not easily to be met with elsewhere" (Steams NUMBER 90 15 1952:363). As to fossils (his "formed Stones), Petiver instruct- ed, "These must be got as intire as you can, the like to be ob- served in marbled Flints, Slates, or other Stones, that have the Impression of Plants, Fishes, Insects, or other Bodies in them; these are to be found in Quarries, Mines, Stone or Gravel Pitts, Caves, Cliffs, and Rocks, on the Sea shoar, or wherever the Earth is laid open" (Stearns, 1952:364). Thus, these natural historians knew exactly what they want- ed and devoted much thought, energy, and money toward get- ting it. Much of their massive correspondence concerns details of instructing, inducing, exhorting, even bribing others to col- lect (e.g., see MacGregor, 1995, on Sloane's correspondence and Steams, 1952, on Petiver's). SHARKS' TEETH.—Sharks' teeth are the quintessential enig- mas of nature, whose charm has inspired wonder, and finally researches, more widely and continuously than perhaps any other fossil. It would scarcely be possible to overemphasize their importance in cutting-edge debate on the meaning, nature, and definition of fossils in the sixteenth and seventeenth centu- ries. As indicated earlier, Rudwick (1976) has done a masterful job in laying out the major features of the story as it unfolded in the pioneering works of Colonna, Scilla, Steno, and Hooke; these need not be retold, but some essential points may be em- phasized. First, "fossil" continued for many years to encompass almost any, usually natural, object "dug-up" from the earth, notably mineral specimens. "Figured stones" was a common term for what we now understand as fossils. Until there was general ac- knowledgment that objects resembling living animals or plants actually were remains of once-living things, there was no logi- cal basis to require a distinction from other interesting things dug up. Sharks' teeth, as glossopetrae or tongue-stones, were widely and deeply embedded in European pre-scientific culture, ema- nating especially from Malta, where the fossils are abundant and are conveniently intertwined with the religious and magi- cal lore of St. Paul, serpents, and poison (for a sampling of this lore, see Zammit-Maempel, 1975, 1989, and Bassett, 1982). From our present god-like heights of sophistication we have tended to dismiss the seeming wrongheaded reluctance to rec- ognize sharks' teeth and other fossils for what they are as the ridiculous ignorance of benighted times; however, these gentle- men were no simpletons but rather the greatest minds of that or any other age. Even after presentation of the careful, logical ar- guments of Steno and Hooke, widely circulated in the Royal Society, that community of scholars did not rise as one in ac- ceptance. Instead, the subject was hotly contested for some 30 years before being laid to rest pretty much by the early 1700s. Grew, Hooke, Lhwyd, Woodward, Ray, Lister, Newton, and Scheuchzer all weighed in on the issue (e.g., see Stokes, 1969). Some, including Hooke, Woodward, and Scheuchzer, were de- cisive in their support of organic origin. In this group only List- er was adamant in his opposition. His views have been charac- terized as ridiculous in hindsight, but his problem, in part, may have been that he knew too much. Lister knew mollusks as per- haps none other of the time, and demanded, but did not find, exact correspondence between fossil and living forms. He was no fool—witness his coming close to "inventing" geologic mapping (Lyons, 1944:99; Steams, 1970:168). He might well be the Agassiz to Hooke's Darwin in this debate. Further, rec- ognition of fossils as such created serious problems in the frame of reference of the time. From it followed almost inevita- bly the problem of extinction of forms without modem coun- terpart, and this was unacceptable in a perfectly economical universe, whether divine or natural. It was in relation to this problem that fossil and modern natural history specimens from far off places, such as America, held special appeal. Locally extinct organisms might well survive elsewhere. With the possible exception of Lister, it might be observed that the practices of those who equivocated on the nature of fossils made sense only if they in fact accepted their organic or- igin. For example, Grew (1681:257) extrapolated (pretty suc- cessfully) on the size of shark (36 feet; -11m) from which large glossopetrae originated; Sloane's (1697) paper on ray teeth was based solidly on comparative methodology—his per- functory allusion to God's wisdom seems all too much like covering his flank. One is tempted to suspect persistence of a certain measure of accommodation to authority through lip ser- vice while proceeding operationally on the basis of persuasive new insights. Another fascinating aspect in which sharks' teeth illustrate how scientific discovery works is the fact that Steno, Scilla, Hooke, and Woodward were essentially coeval in their re- searches. Barring some more persuasive evidence of intellectu- al piracy than has thus far materialized, the interesting point is that this was an idea whose time had come. Hooke was a great and wide-ranging idea man, and there is no need to detract from his astounding originality. His geologic insights and pri- orities have at last been well presented (Drake, 1996). Never- theless, he clearly had a tendency toward jealousy of priority— whatever the topic, he thought of it first (which contributed strongly to his irreparable schism with Newton). Even if Steno was aware of Hooke's and/or Scilla's ideas, he has to be ac- corded primacy because he developed the idea fully with step- by-step logical procedure, which has been brought out best by Scherz (1969, 1971). Woodward clashed with almost everyone, was a thoroughly unsympathetic character, and was accused of pirating Scilla's ideas, but he probably was not a plagiarist (Jahn, 1972:210) (useful and accessible insights into Wood- ward's activities and character may be found in Eyles, 1971, and Levine, 1977). Another great truth illustrated by this history is that discover- ies do not stay discovered; they must be tended like a garden. Scilla (1670) illustrated what turned out to be the first known specimen of a sharktoothed porpoise, family Squalodontidae, a nice piece of a mandible with three teeth. This historic speci- men, preserved in the Woodward Collection at Cambridge, has since been the object of repeated attention in the paleontologi- 16 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY cal literature of the modern era. Although first formally de- scribed as a seal, and in one aberrant view regarded as a hippo- potamus (Owen, 1840-1845:564-565, pl. 142: fig. 3), it has long since become securely and correctly embedded in the lit- erature as a squalodont cetacean, the holotype of Squalodon melitensis (Blainville), where it has been alluded to and figured repeatedly (e.g., see McCoy, 1867:145; Kellogg, 1923:24; Gunther, 1937:433, unnumbered figure, p. 432; Fabiani 1949:26-29, figs. 9, 10; Rothausen, 1968:92). Then, in 1992, Gould (in Purcell and Gould, 1992:93-94, figs. 64, 65) misi- dentified the specimen as the jaw of a shark, invalid support for the valid interpretation of glossopetrae. Although as always we have a duty to correct objective mistakes, especially by con- temporary and influential authorities (I wrote to Gould imme- diately upon discovering the error, 8 March 1993), the signifi- cant point is hardly that even the greatest living spokesperson for paleontology to the world at large is fallible, but that ap- proach to truth is a fragile dynamic that requires continual vigi- lance. There may be some validity to Gould's (1996:110) claim that "persistent minor errors of pure ignorance are galling to perfectionistic professionals," but this has no bearing on the overriding requirement that each professional strive assiduous- ly to get things right and never knowingly let even "minor er- rors" persist. Finally, the history of sharks' teeth in relation to humans is a powerful cautionary tale against fashion in science. Fortunate- ly, people in general have maintained a seemingly innate curi- osity and interest in them throughout time. In professional pale- ontology, however, when I was a student some four decades ago at a prestigious university, only a naive beginner would risk being labelled childish, or worse, "amateurish," by betray- ing any interest in sharks' teeth (or dinosaurs). Now dinosaurs are the hottest topic in vertebrate paleontology, and even sharks are respectable subjects of investigation (Klimley and Ainley, 1996). Scientists are probably no more foolish as a group than the citizenry at large in lurching to extremes, but they may tend to appear so in retrospect because they put extreme views on record in emphatic terms. More reflective attention to the histo- ry of our science would undoubtedly tend to mitigate our most embarrassing emanations and perhaps damp down fadism. I hope that these few modest historical nuggets are enough to persuade readers that ancient specimens, many lost or mislaid, and the thinking and writing surrounding them are not mere quaint curiosities but are landmarks that can and should have meaning today. Secord (1996:459) has made a forceful case for the value of history not merely as entertainment or nostalgia but as an ac- tive force in research, concluding: Rather, a bold enquiry into the past can uncover the basic structures and large-scale patterns of change which lie behind our current dilemmas. We have inherited not just our institutions and practices, but our problems: and these can only be understood as products of history. A new culture of natural history will flourish only if it is effectively rooted in—and draws upon—a critical under- standing of the past. Acknowledgments This is perhaps the last seemly opportunity to thank David J. Bohaska, National Museum of Natural History (NMNH), Smithsonian Institution, Washington, D.C, for his steadfast and effective assistance not only in all aspects of my research but in preparation of volumes III and IV, as he now has the thankless status of coeditor in recognition of the reality of his indispensable contributions. In preparation and repreparation of illustrations, it has been our good fortune to have had the un- flagging, even enthusiastic, support and initiative of Irina Ko- retsky, Victor E. Krantz, and Mary Parrish (all of the NMNH). Long after our golden prose has tarnished, the pictures will re- main as new. For the substantive content of these volumes, we are of course indebted to two groups, in part overlapping. First, the collectors, mostly unpaid volunteers, without whom there could be no science of paleontology. They are named as ap- propriate in the individual papers. Second, the researchers who have contributed the papers that give meaning to the specimens. Many of them have waited, not totally with pa- tience, but they have waited, a quarter century, for my unreal- istic expectations of early conclusion to become reality. This delay has required the greatest forbearance from the most prompt authors, in that they have had to revise and update re- peatedly. Numerous individuals have aided generously in preparation of this small historical paper, with information and advice on specimens, records, and literature. These include, alphabetical- ly by surname, Vanessa J.A. Carr (Public Record Office, Chan- cery Lane, London), Mike Dorling (Department of Earth Sci- ences, Cambridge), Burkart Engesser (Naturhistorisches Museum, Basel), Paula Findlen (History Department, Universi- ty of California, Davis), Gerald M. Friedman (City University of New York), Ivor Noel Hume (Colonial Williamsburg, Vir- ginia), Michael Hunter (Department of History, Birkbeck Col- lege, University of London), Laura Laurencich-Minelli (Dipar- timento di Paleografia e Medievistica, Universita Degli studi di Bologna), Urs B. Leu (Zentralbibliothek, Zurich), Arthur MacGregor (Ashmolean Museum, Oxford), Patrick Nuttall (The Natural History Museum, London), H. Philip Powell (Geological Collections, University Museum, Oxford), Christa Riedl-Dorn (Archiv-Leitung, Naturhistorisches Museum, Wien), Mary Sampson (The Royal Society, London), Carlo Sarti (Dipartimento di Scienze Geologiche, Universita degli Studi di Bologna), Antony J. Sutcliffe (The Natural History Museum, London). G. Taylor (The British Library, London), K.J. Wallace (Central Archives, The British Museum, Lon- don), and George Zammit-Maempel (Birkirkara, Malta). For critical review and improvement of the manuscript I thank David J. Bohaska and I thank Ellen Compton-Gooding (now retired) and Lauck W. Ward (both Virginia Museum of Natural History, Martinsville). NUMBER 90 17 Literature Cited Ambrosinus, Bartholomaeus 1648. Ulyssis Aldrovandi, Patricii Bononiensis, Musaeum metallicum in libros III! distribulum. [vi]+979+[13] pages. Bologna: M.A. Bernia. [Edited and published by Ambrosinus after Aldrovandi's death. Many woodcuts, some full page, interspersed in text, not sequen- tially numbered. See Figure 1 for title page.] Anonymous 1993. Virginia's Official State Fossil. Virginia Explorer, 9(2):28, 2 un- numbered figures. Aubrey, John 1898. "Brief Lives," Chiefly of Contemporaries: Set Down by John Aubrey, between the Years 1669 & 1696. Volume I, xv+427 pages; volume 2, 370 pages, 5 plates. Oxford: Clarendon Press. [Edited from the author's manuscripts by Andrew Clark.] Bassett, Michael G. 1982. "Formed Stones," Folklore and Fossils. National Museum of Wales, Geological Series, 1:1-32,43 unnumbered figures. Birch, Thomas 1756-1757. The History of the Royal Society of London, for Improving of Natural Knowledge, from Its First Rise. Volume 1, [vi] + 511 pages (1756); volume 2, [ii]+501 pages (1756); volume 3, [ii]+520 pages (1757); volume 4, [ii]+558 pages (1757). London: A. Millar. Butler, Declan 1997. French Museum "In Decay" Fights for Its Life. Nature, 385(6615): 378. Challinor, J. 1953. The Early Progress of British Geology, I: From Leland to Wood- ward, 1538-1728. Annals of Science, 9(2): 124-153, 1 plate. Clark, Alan J. 1992. The History of the Royal Society: A Chronological Checklist of Books in English. Notes and Records of the Royal Society of Lon- don, 46(2):335-345. [Contains valuable guide to Society's periodi- cals and indexes.] Copeman, W.S.C. 1960. Dr. Jonathan Goddard, F.R.S. (1617-1675). Notes and Records of the Royal Society of London, 15:69-77, I unnumbered figure, plate 6. Davidson, Jane P. 2000. Fish Tales: Attributing the First Illustration of a Fossil Shark's Tooth to Richard Verstegan (1605) and Nicolas Steno (1667). Pro- ceedings of the Academy of Natural Sciences of Philadelphia. 150:329-344, 4 figures. Drake, Ellen Tan 1996. Restless Genius: Robert Hooke and His Earthly Thoughts, xiv+386 pages, 61 figures, 7 plates. New York and Oxford: Oxford Univer- sity Press. Edwards, W.N. 1931. Guide to an Exhibition Illustrating the Early History of Palaeontol- ogy. British Museum (Natural History), Special Guide, 8:i—vi, 7-68, 19 figures, 4 plates. [A revised edition, The Early History of Palae- ontology, was published by the museum in 1967, but the original is more valuable on account of its allusion to the exhibition and the ci- tation of specimens in it.] Ewan, Joseph, and Nesta Ewan 1970. John Banister and His Natural History of Virginia, 1678-1692. xxx+486+[2] pages, 9 "charts and exhibits," 70 figures. Urbana, Chicago, and London: University of Illinois Press. Eyles, VA. 1958. The Influence of Nicolaus Steno on'the Development of Geological Science in Britain. In Gustav Scherz, editor, Nicolaus Steno and His Indlce, pages 167-188. Copenhagen: Munksgaard. 1971. John Woodward, F.R.S., F.R.C.P., M.D. (1665-1728): A Bio-bib- liographical Account of His Life and Work. Journal of the Society for the Bibliography of Natural History, 5(6):399^t27, I plate. Fabiani, Ramiro 1949. Gli Odontoceti del Miocene inferiore della Sicilia. Memorie delt'ls- tituto Geologico dell'Universita di Padova, 16:1-32, 10 figures, 2 plates. Findlen, Paula 1994. Possessing Nature: Museums, Collecting, and Scientific Culture in Early Modern Italy, xviii+450 pages, frontispiece, 36 figures, 2 ta- bles. Berkeley: University of California Press. [Six of the figures are museum plans.] Frick, George F., James L. Reveal, C. Rose Broome, and Melvin L. Brown 1987. Botanical Explorations and Discoveries in Colonial Maryland, 1688 to 1753. Huntia, 7:5-59. Furrer, Heinz, and Urs B. Leu 1998. Der Landesplattenberg Engi: Forschungsgeschichte, Fossilien und Geologic 132 pages, 137 figures. Engi, Switzerland: Stiftung Landesplattenberg Engi. Germon, Raye N., Lauck W. Ward, and Clayton E. Ray 1987. Ecphora; Important Fossil from the Miocene Strata on the Chesa- peake Bay. Maryland Naturalist, 31(l):25-33, 11 figures. Gilmore, Charles W. 1941. A History of the Division of Vertebrate Paleontology in the United States National Museum. Proceedings of the United States National Museum. 90(3109):305-377, plates 49-53. Gould, Stephen Jay 1996. Afterword: The Truth of Fiction: An Exegesis of G.G. Simpson's Dinosaur Fantasy. In George Gaylord Simpson, The Dechronization of Sam Magruder, pages 105-126. New York: St. Martin's Press. 1997. Seeking to Explain Sea Fossils on Mountain Ridges, Leonardo De- veloped His Theory of The Living Earth. Natural History, 106(4): 18-21, 58-64, 1 unnumbered figure. [Listed in the table of contents under Gould's regular column as "This View of Life: Le- onardo's Living Earth."] Grew, Nehemiah 1681. Musaeum Regalis Societalis: Or a Catalogue & Description of the Natural and Artificial Rarities Belonging to the Royal Society and Preserved al Gresham College. [xii]+386+[2]+43 pages, 31 plates. London: W. Rawlins for the author. [See Figure 2 for title page; fi- nal 43 pages and plates 23-31 devoted to comparative anatomy.] Gross, Paul R., Norman Levitt, and Martin W. Lewis, editors 1996. The Flight from Science and Reason. Annals of the New York Acad- emy of Sciences, 775: xi+593 pages. Gunther, R.T. 1937. Early Science in Cambridge, xii + 513 pages, many illustrations. Ox- ford: The University Press. [Reprinted 1969, London: Dawsons of Pall Mall.] 1945. Early Science in Oxford, XIV: Life and Letters of Edward Lhwyd. xvi+576 pages, 14 plates. Oxford: The University Press. [Reprinted 1968, London: Dawsons of Pall Mall.] Heikamp, Detlef 1976. American Objects in Italian Collections of the Renaissance and Ba- roque: A Survey. In Fredi Chiappelli, editor, First Images of Amer- ica: The Impact of the New World on the Old, 1:455-^182, figures 31-53. Berkeley: University of California Press. [Michael J.B. Allen and Robert L. Benson, coeditors.] Hooke, Robert 1705. Lectures and Discourses of Earthquakes and Subterraneous Erup- tions; Explicating the Causes of the Rugged and Uneven Face of the Earth; and What Reasons May Be Given for the Frequent Finding of Shells and Other Sea and Land Petrified Substances, Scattered over 18 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY the Whole Terrestrial Superficies. In Richard Waller, The Posthu- mous Works of Robert Hooke.... pages 277^150, plates 1-7. [Re- printed. 1969, in The Sources of Science, 73: xxviii + [6] + 572 [actually 506] + [12] pages, 13 plates. New York and London: Johnson Reprint Corporation. A valuable new edition "transcribed, annotated, and with an introduction by Ellen Tan Drake" constitutes the major part of Drake, 1996.] Hume, Ivor Noel 1994. Custis Square: The Williamsburg Home and Garden of a Very Curi- ous Gentleman. Colonial Williamsburg. 16(4): 12-26, 14 unnum- bered figures. Hunter, Michael 1985. The Cabinet Institutionalized: The Royal Society's "Repository" and Its Background. In Oliver Impey and Arthur MacGregor, edi- tors, The Origins of Museums: The Cabinet of Curiosities in Six- teenth- and Seventeenth-Century Europe, pages 159-168. Oxford: Clarendon Press 1989. Establishing the New Science: The Experience of the Earlv Roval Society, xiv+382 pages. Woodbridge: Boydell Press. 1994. The Royal Society and Its Fellows, 1660-1700: The Morphology of an Early Scientific Institution. British Society for the History of Sci- ence, monograph 4: ix+291 pages, second edition. Impey, Oliver, and Arthur MacGregor, editors 1985. The Origins of Museums: The Cabinet of Curiosities in Sixleenlh- andSeventeenth-Century Europe, xiv + 335 pages, 108 figures. Ox- ford: Clarendon Press. Jahn, Melvin E. 1969. Some Notes on Dr. Scheuchzer and on Homo diluvii leslis. In Cecil J. Schneer, editor. Toward a History of Geology, pages 192-213, 3 figures. Cambridge, Massachusetts: M.I.T. Press. 1972. A Bibliographical History of John Woodward's An Essay toward a Natural History of the Earth. Journal of the Society for the Bibliog- raphy of Natural History. 6(3): 181 -213, 10 figures. 1975. A Note on Scheuchzer's Versification. Journal of the Society for the Bibliography of Natural History. 7(3):321-322. Jessop, L. 1989. The Club at the Temple Coffee House—Facts and Supposition. Ar- chives of Natural History. 16(3):263-274. Jillson, Willard Rouse 1936. Big Bone Lick: An Outline of Its History, Geology, and Paleontol- ogy, to Which Is Added an Annotated Bibliography of 207 Titles. Big Bone Lick Association Publications, 1: xvi +164 pages, frontis- piece, 4 unnumbered figures. Louisville, Kentucky: Standard Print- ing Company. Josselyn, John 1674. An Account of Two Voyages to New-England, Made during the Years 1638, 1663. xii+211 pages. Boston: William Veazie. [1865 edition, taken from second edition, 1675, and other sources.] Kellogg, Remington 1923. Description of Two Squalodonts Recently Discovered in the Calvert Cliffs, Maryland; and Notes on the Shark-Toothed Cetaceans. Pro- ceedings of the United Stales National Museum, 62( 16): 1-69, plates 1-20. Klimley, A. Peter, and David G. Ainley, editors 1996. Great While Sharks: The Biology of Carcharodon carcharias. xii+517 pages. San Diego: Academic Press. Latrobe, B.H. 1799. Memoir on the Sand-Hills of Cape Henry in Virginia. Transactions of the American Philosophical Society. 4:439^143, 1 plate. 1809. An Account of the Freestone Quarries on the Potomac and Rappah- annoc Rivers. Transactions of the American Philosophical Society, 6:283-293. Laurencich-Minelli, Laura 1985. Museography and Ethnographical Collections in Bologna during the Sixteenth and Seventeenth Centuries. In Oliver Impey and Arthur MacGregor, editors, The Origins of Museums: The Cabinet of Curi- osities in Sixteenth- and Seventeenth-Century Europe, pages 17-23. Oxford: Clarendon Press. Laurencich-Minelli, Laura, and Maurizio Serta 1988 ("1987"). Tra Museo e Biblioteca: Un esempio del metodo di lavoro di Ulisse Aldrovandi "Americanista." Museologia Scientiftca (Ve- rona), 4(3^t):221-228, 2 figures. [Date on title page is 1987; actu- ally published in 1988.] LeFanu, William 1990. Nehemiah Grew M.D.. F.R.S.: A Study and Bibliography of His Writings, xviii+182 pages, frontispiece, 7 plates. Winchester: St. Paul's Bibliographies. [Distributed in the Americas by Omnigraph- ics Inc., Detroit] Lemonick, Michael D. 1996. Dumping on Darwin. Time, March 18, 1996:81. Levine, Joseph M. 1977. Dr Woodward's Shield: History, Science and Satire in Augustan England, x+362 pages, frontispiece, 20 unnumbered figures. Berke- ley, Los Angeles, and London: University of California Press. Lyons, Henry 1940. The Record of the Royal Society of London for the Promotion of Natural Knowledge. Fourth edition, vii + 578 pages, 19 plates. Lon- don: Morrison and Gibb. for the Royal Society. 1944. The Royal Society, 1660-1940: A History of Its Administration Un- der Its Charters, x+354 pages. Cambridge: University Press. MacGillivray, William 1838. Memoir of Ulysses Aldrovandi. In William MacGillivray, British Quadrupeds. The Naturalist's Library, 22:17-58. Edinburgh. MacGregor, Arthur, editor 1994. Sir Hans Sloane. Collector. Scientist. Antiquary. Founding Father of the British Museum. 308 pages, 115 figures, 40 plates. London: British Museum Press, in association with Alistair McAlpine. 1995. The Natural History Correspondence of Sir Hans Sloane. Archives of Natural History, 22( I ):79-90. Maienschein, Jane 1994. Cutting Edges Cut Both Ways. Biology and Philosophy. 9( 1): 1-24. Maull, Nancy 1997. Science under Scrutiny. Han-ard Magazine, 99(4):22-25, 27. McCoy, Frederick 1867. On the Occurrence of the Genus Squalodon in the Tertiary Strata of Victoria. Geological Magazine. 4(34): 145, plate 8: figure I. Mel ver, Tom 1992. Anti-Evolution: A Reader's Guide to Writings Before and After Dar- win. xv+385 pages. Baltimore: Johns Hopkins University Press. McKie, Douglas 1960. The Origins and Foundation of the Royal Society of London. Notes and Records of the Royal Society of London, 15:1-38, plate 1. Morello, Nicoletta 1979. La Nascita del la Paleonlologia nel Seicenlo: Colonna. Slenone e Scilla. 265 pages. Milan: Franco Angeli. [In Italian.] 1981. De Glossopetris dissertatio: The Demonstration by Fabio Colonna of the True Nature of Fossils. Archives Internationales d'Histoire des Sciences, 31:63-71. Nature 1990. A Major Museum Goes "Populist." Nature, 345(6270): 1-2. Oldroyd, David R. 1989. Geological Controversy in the Seventeenth Century: "Hooke vs Wall is" and Its Aftermath. In Michael Hunter and Simon Schaffer, editors, Robert Hooke, New Studies, pages 207-233. Woodbridge: Boydell Press. Osborn, Henry Fairfield 1931. Cope: Master Naturalist, xvi+740 pages, 30 figures. Princeton, New Jersey: Princeton University Press. NUMBER 90 19 Owen, Richard 1840-1845. Odonlography; Or, a Treatise on the Comparative Anatomy of the Teeth; Their Physiological Relations, Mode of Development, and Microscopic Structure, in the Vertebrate Animals. Volume I, text, xix + lxxiv+655 pages, volume 2, atlas, 37 pages, 168 plates. London: Hippolyte Bailliere. Petiver, James 1705. An Account of Animals and Shells Sent from Carolina to Mr James Petiver, F.R.S. Philosophical Transactions [for May, 1705], 24(299): 1952-1960. [Pages 1953 and 1954 duplicated.] Price, David 1989. John Woodward and a Surviving British Geological Collection from the Early Eighteenth Century. Journal of the History of Collections, l(l):79-95, 13 figures. Pumfrey, Stephen 1991. Ideas above His Station: A Social Study of Hooke's Curatorship of Experiments. History of Science. 29(1)83:1-37. Purcell, Rosamond Wolff, and Stephen Jay Gould 1992. Finders, Keepers: Eight Collectors. 158 pages, 105 figures. New York: W.W. Norton. Ray, Clayton E. 1983. Prologue. In Clayton E. Ray, editor, Geology and Paleontology of the Lee Creek Mine, North Carolina, I. Smithsonian Contributions to Paleobiology, 53:1-14. 1987. Foreword. In Clayton E. Ray, editor, Geology and Paleontology of the Lee Creek Mine, North Carolina, II. Smithsonian Contributions to Paleobiology, 61:1-8, 3 figures. Redington, Joseph, editor 1868. Calendar of Treasury Papers, 1556/7-1696. 667 pages. London: Longman, Green, Render and Dyer, printed by Eyre and Spottis- woode for Her Majesty's Stationery Office. [Contains 42 "volumes" within one volume, paginated as one.] Roberts, Brynley F. 1989. In Search of Edward Lhuyd. Archives of Natural History, 16(I):49-57. Robinson, H.W. 1946. The Administrative Staff of the Royal Society, 1663-1861. Notes and Records of the Royal Society of London, 4(2): 193-205. Rothausen, Karlheinz 1968. Die Systematische Stellung der Europaischen Squalodontidae (Od- ontoceti, Mamm.). Palaontologische Zeitschrifl. 42(l/2):83-I04, figures 1-3, plates 11, 12. Rudwick, Martin J.S. 1976. The Meaning of Fossils: Episodes in the History of Palaeontology. Second edition, [xvi]+287 pages, 52 figures. New York: Science History Publications, USA. Sagan, Carl 1995. The Demon-Haunted World: Science as a Candle in the Dark. xviii+457+[4] pages. New York: Random House. Salter, H.E. 1915. Remarks and Collections of Thomas Hearne Volume 10 (March 27, 1728-December 8, 1731), xi+496 pages. Oxford: Oxford Historical Society, Clarendon Press. Schepelem, H.D. 1990. The Museum Wormianum Reconstructed; A Note on the Illustration of 1655. Journal of the History of Collections, 2( I ):81 -85,2 figures. Scherz, Gustav, editor 1969. Steno, Geological Papers. Ada Historico Scientiarum Naturalium et Medicinalium, 20: 370 pages, frontispiece, 29 figures (diagrams), 133 plates. Odense: Odense University Press. [Illustrations on pages 2, 216, 217, 246, 289-348; most "plates" two per page.] 1971. Dissertations on Steno as Geologist. Ada Historica Scientiarum Naturalium et Medicinalium, 23: 319 pages, frontispiece, many fig- ures. Odense: Odense University Press. [Most figures unnumbered, some separately numbered by chapter.] Scheuchzer, Johann Jakob 1708. Piscium Querelae et Vindiciae. 36 pages, 5 plates. Tiguri: Gessneri- anis. [See Figure 4 for title page.] Scilla, Agostino 1670. La Vana speculazione disingannata dal Senso; Lettera risponsiva circa i corpi marini, che petrificati si trovano in varii luoghi ter- restri. 168 pages, 28 plates. Naples. [Republished in Latin, 1747, as De corporibus marinis lapidescentibus... and reprinted New York: Amo Press, 1980.] Secord, James A. 1996. The Crisis of Nature. In N. Jardine, J.A. Secord, and E.C. Spary, ed- itors, Cultures of Natural History, pages AA7-A59. Cambridge: Uni- versity Press. Shapin, Steven 1989. Who was Robert Hooke? In Michael Hunter and Simon Schaffer, editors, Robert Hooke, New Studies, pages 253-286. Woodbridge: Boydell Press. Shaw, William A., editor 1935. Calendar of Treasury Books, January 1693 to March 1696, Pre- served in the Public Record Office. Volume 10, part 1, pages 1-592. London: His Majesty's Stationery Office. Simpson, George Gaylord 1942. The Beginnings of Vertebrate Paleontology in North America. Pro- ceedings of the American Philosophical Society, 86( 1): 130-188, 23 figures. 1943. The Discovery of Fossil Vertebrates in North America. Journal of Paleontology, !7(l):26-38, 2 figures. Sloane, Hans 1697. An Account of the Tongue of a Pastinaca Marina, Frequent in the Seas about Jamaica, and Lately Dug Up in Mary-Land, and Eng- land. Philosophical Transactions (for September, 1697), 19(232): 674-676, 1 plate. Steam, William T. 1981. The Natural History Museum at South Kensington: A History of the British Museum (Natural History) 1753-1980. xxiv+414 pages, 88 plates. London: Heinemann. [Published in association with the Brit- ish Museum (Natural History).] Steams, Raymond Phineas 1951. Colonial Fellows of the Royal Society of London, 1661-1788. Notes and Records of the Royal Society of London, 8(2): 178-246. 1952. James Petiver, Promoter of Natural Science, c. 1663-1718. Pro- ceedings of the American Antiquarian Society. 62(2):243-365. 1970. Science in the British Colonies of America. xx+760+[2] pages, 32 full-page illustrations. Urbana, Chicago, and London: University of Illinois Press. Stokes, Evelyn 1969. The Six Days and the Deluge: Some Ideas on Earth History in the Royal Society of London 1660-1775. Earth Science Journal. 3(l):l3-39. Stuiver, Minze, and Harold W. Boms, Jr. 1975. Late Quaternary Marine Invasion in Maine: Its Chronology and As- sociated Crustal Movement. Geological Society of America Bulletin, 86(l):99-103, 4 figures. Swem, E.G. 1957. Brothers of the Spade: Correspondence of Peter Collinson of Lon- don, and of John Custis, of Williamsburg, Virginia, 1734-1746. [viii]-*-196 pages, 3 figures. Barre, Massachusetts: Barre Gazette. [Reprinted from Proceedings of the American Antiquarian Society. 1949.1 20 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Thackray, John 1994. Mineral and Fossil Collections. In Arthur MacGregor, editor, Sir Hans Sloane, Collector, Scientist, Antiquary, Founding Father of the British Museum, pages 123-135, figures 30-35. London: British Museum Press, in association with Alistair McAlpine. Thompson, Woodrow B. 1982. Recession of the Late Wisconsinan Ice Sheet in Coastal Maine. In Grahame J. Larson and Byron D. Stone, editors, Late Wisconsinan Glaciation of New England, pages 211-242, 11 figures. Dubuque, Iowa: Kendall/Hunt Publishing Company. Thomson, Thomas 1812. History of the Royal Society, from Its Institution to the End of the Eighteenth Century, viii+552+xci pages. London: Robert Baldwin. Torrens, Hugh 1985. Early Collecting in the Field of Geology. In Oliver Impey and Arthur MacGregor, editors, 77i0.61) but within the range of other istiophorids. Both scapulas have a narrow, flat articular surface for the first pectoral-fin ray (Plate 5f,g). A nar- row width with a flat articular surface is characteristic of M. in- dica (Wapenaar and Talbot, 1964) (SNW/SL, Table 2). Makaira cf. M. indica (Cuvier, 1832) Plate 5/ Material.—1 scapula (USNM 488100). Remarks.—Scapula USNM 488100 (Plate 5/) has a narrow, curved surface for articulation with the first pectoral-fin ray. Makaira nigricans Lacepede, 1802 Plates da-m, 7a-m Material.—4 dentaries (NCSM 2124, 2125; USNM 475396, 475423); 8 parasphenoids (NCSM 5159, 11248; USNM 2855370, 421526, 481990, 488047, 488048, 488050); 13 predentaries (USNM 25741, 291066, 291114, 475415, 481935-481938, 481954, 481959, 481962, 481963, 488011); 38 rostra (NCSM 2129; USNM 286973, 286986, 286996, 290614, 297407, 475397, 475401, 475405, 475406, 475408, 475410,475419-475421, 476330, 476333, 481941—481943, 481947^181949, 481971, 481972, 481976, 481977, 488020, 488023, 488024, 8 USNM uncataloged); 5 scapulae (USNM 30 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Table 2.—Mean (i), observed range, and number of bones examined (n) tor variables (ratios) on 12 bones of eight extant species of the family Istiophoridae. Abbreviations for ratios are explained in the text and in the leg- ends to figures 1, 2, and 5. Ratio /. platypterus M. indica M. nigricans T. albidus T. angustirostri. T. audax T he/one T. pfluegeri (x) (range) n (x) (range) n (x) (range) n (x) (range) n (x) (range) « (x) (range) /; (x) (range) n (x) (range) n Articular AL/ASM .70(.59-.80)22 .90(.85-.94)2 .80(.66-.93)23 .71(.58-.91)15 - .78(.68-.88) 9 - -(.81)1 AW/AL .74(.53-.94)22 .70(.69-.70)2 ,84(.68-l.l)23 .91 (.65-1.3) 15 - .71(.58-.85) 9 - -(.89)1 AAL/AL .74(.59-.92)22 .59(.59-.60)2 .68(.47-.9l)23 .70(.54-.97)I5 - ,66(.55-.79) 9 - -(.29)1 ATW/AL 1.1 (.91-1.4)21 ,94(.94-.95)2 1.2(.98-1.6)22 1.3 (.96-1.9) 15 - 1.1 (.87-1.3) 9 - -(1.1)1 AW/ATW .67(.54-.78)26 .74(.74-.74)2 .71(.55-.78)22 .71(.57-.86)15 - .63(.58-.70)10 - -(.81)1 Dentary DAD/DJL .44(.33-.55)l5 -(.86)1 ,76(.55-.89)12 53(.47-.60)10 - .56(.39-.74) 7 - .26 (.22-30)2 First pectoral-fin ray (dorsal segment) FAW/FW .65(.55-.71)17 .27(.18-.37)8 .65(.5I-.72)20 ,66(.61-.72)14 -(.62)1 .62(.58-.67) 4 -(.72)1 -(.65)1 Maxilla MW/ML .39(.25-.57)14 - .64(.60-.70) 5 .47(.44-.5l) 8 - .47(.45-,49) 2 - .24(.23-.24)2 MH/ML .33(.26-.47)14 - ,43(.38-.49)14 .38(.34-.45)10 - .37(.36-.37) 2 - .33 (.31-.34) 2 MVW/MVH 1.4(.95-2.4) 9 - 1.2(1.1-1.3) 3 1.2(.80-1.8) 3 - 1.2(.96-1.5) 2 - - MVW/ML .22(.17-.36) 9 - .36(.35-.37) 3 .20(.18-.24) 3 - .28(.25-.3I) 2 - - MVW/MW .55(.48-.63) 9 - ,56(.53-.60) 3 .44(.40-.48) 3 - .60(.51-.69) 2 - - MVW/MH ,64(.59-.75) 9 - .90(.81-.96) 3 .54(.50-.58) 3 - .77(.68-.85) 2 - - Parasphenoid PAF/PAFW .58(.39-.67) 9 -(•32)1 ,41(.32-.52) 8 .51 (45-.60) 9 - .68(.63-.72) 2 - -(.63)1 PAF/PAD .55(.47-.67) 9 -(.54)1 .64(.57-.73) 8 .60(.47-.83) 9 - .62(.57-.68) 2 - -(.49)1 PAD/PAFW 1.1 (.74-1.4) 9 -(59)1 .63(.55-.79) 8 .86(.73-.98) 9 - 1.1(1.1-1.1) 2 - -(1.3)1 PAW/PAFW .70(.62-.86) 9 -(.85)1 .84(.67-1.0) 8 .53(.4l-.73) 9 - .58 (.47-68) 2 - -(.89)1 PAD/PAW 1.6(1.2-2.2) 9 -(.70)1 .77(.62-.97) 8 1.7(1.0-2.1) 9 - 2.0(1.6-2.4) 2 - -(15)1 Predentary PW/PL .50(.23-.63)21 ,54(.48-.62)5 .65(.45-l.I)23 .41(.34-.57)I5 .60(.54-.69)4 ,42(.33-.51)13 -(59)1 -(.65)1 PD/PL .27(.20-.37)21 ,38(.30-.41)5 .36(.26-.58)23 ,24(.20-.33)l5 .59(.52-.69)4 ,26(.22-.30)13 -(¦53)1 -(.60)1 PD/PW .56(48-1.0)21 .70(.63-.81)5 .56(.48-.61)23 ,59(.53-.65)15 .99(.97-1.0)4 .60(.53-.72)l3 -(.90)1 -(.92)1 Quadrate QAW/QH .17(.15-.20)11 .29(.28-.31)2 .23(.l9-.27)22 .21 (.18-24) 17 -(.16)1 .21(.20-.22) 4 - -(.17)1 QMW/QAW .68(.58-.8l)ll .85(.83-.88)2 .77(.57-.92)22 .73(.48-.88)18 -(.65)1 ,71(.67-.77) 4 - -(62)1 QAW/QHS .29(.24-.39) 11 .44(.44-.45)2 .36(.30-.43)2I .35(.30-.40)18 - .38(.37-.39) 2 - -(.30)1 QMW/QHS .I9(.17-.23)ll .38(.38-.38)2 .27(.22-.35)2I ,25(.19-.30)18 - .28(.27-.28) 2 - -(.18)1 QMW/QML .73(.62-.80)ll 1.1(1.1-1.1)2 .86(.63-1.0)22 .83 (.53-1.0)18 -(.56)1 .79(.65-1.0) 4 - -(.71)1 Scapula SGW/SL ,67(.55-.81)13 .44(.36-.53)8 .58(.51-.76)20 .64(.54-.72)l3 .70(.68-.71)2 ,63(.60-.64) 4 - 64(.64-.64)2 SNW/SL .44(.30-.53)l3 .16(.11-.19)8 .35(.22-.51)20 .42(.30-.55)13 .60(.53-.68)2 .56(.49-.63) 4 - ,56(.54-.57)2 SNW/SGW .65(.53-.78)l3 .36(.26-.47)8 .60(.37-1.0)20 .67(.50-.84)13 .87 (.74-1.0) 2 .89(.82-1.0) 4 - .87(.85-.90)2 Vertebra 1 AN/CL .61(.53-.69) 7 - .64(.51-.74) 6 .67(.57-.77) 4 - .64(.59-.70) 2 - -(.65)1 PN/CL .46(.40-.52)10 - ,41(.38-.47) 9 .37(.26-,42) 4 - .43(.39-.45) 5 - -(.41)1 ASW/VAD .82(.6I-1.0) 9 - 1.0(.79-1.6) 9 .89(.85-.96) 4 - ,90(.81-.97) 5 - -(.86)1 ASW/CL .62(.50-.78) 9 - ,95(.77-1.2) 9 .69(.64-.76) 4 - .8l(.76-.88) 5 - -(48)1 LAD/CL .68(.60-.77) 9 - ,92(.84-I.0) 9 .69(.62-.75) 4 - .78(.70-,87) 5 - -(.53)1 VAD/CL ,77(.62-.91)I0 - .94(.74-I.O) 9 ,77(.74-.79) 4 - .91(.82-.99) 5 - -(56)1 VAD/LAD 1.1 (.88—1.3) 9 - 1.0(.80-1.2) 9 1.1(1.1-1.2) 4 - 1.2(1.1-1.2) 5 - -(1.1)1 LPD/CL ,78(.72-.89)l0 - 1.0(.92-1.1) 8 .74(.69-.78) 4 - .86(.80-.90) 5 - - VPD/CL .76(.70-.83)I0 - .95(.86-l.0) 9 .73(.69-.76) 4 - ,85(.78-.91) 4 - - VPD/LPD .97(93-1.0)10 - .93 (.86-1.0) 8 ,98(.97-1.0) 4 - .98(93-1.0) 4 - - LAD/LPD ,89(.82-.97) 9 - ,9l(.86-.96) 8 .93(.9l-.96) 4 - .91(.88-.96) 5 - - VAD/VPD 1.0C.76—1.1)10 - .99(.76-1.1) 9 1.1(1.0-1.1) 4 - 1.1(1.1-1.2) 4 - - NW/CL .37(.22-.47)l0 - .63(.57-.74) 9 .41 (.39-.47) 4 - .45(.42-.52) 5 - -(•22)1 NW/LPD .47(.30-.56)10 - .62(.55-.68) 8 .56(.5l-.60) 4 - .52(.47-.58) 5 - - Vertebra 22 AN/CL .60(.57-.64) 11 -(.56)1 .60(.57-.62) 7 ,58(.55-.62) 5 - .60(.59-.60) 2 - -(.62)1 PN/CL .38(.32-.42)ll -(.38)1 .39(.36-.43) 7 .39(.36-.45) 5 - ,40(.40-.41) 2 - -(38)1 LAD/CL ,58(.45-.66)ll -(.68)1 .77(.58-.87) 7 .58(.54-.63) 5 - .61(.61-.61) 2 - -(48)1 VAD/CL .53(.47-.57)ll -(.50)1 .50(.42-.54) 7 .50(.46-.53) 5 - .52(.51-.52) 2 - -(.45)1 VAD/LAD .91 (.78-1.1) 11 -(.73)1 .66(.62-.73) 7 .87(.84-.92) 5 - ,85(.83-.87) 2 - -(.94)1 LPD/CL ,54(.48-.62)IO -(.74)1 ,74(.56-.84) 7 ,54(.5l-.57) 4 - .64(.63-.66) 2 - -(.46)1 NUMBER 90 31 TABLE 2.—Continued. Ratio /. platypterus (x) (range) « M. indica M. nigricans (x) (range) n T. albidus T. angustirostris (x) (range) n T. audax T. belone T. pfluegeri (x) (range) n (x) (range) n (x) (range) n (x) (range) n (x) (range) n VPD/CL .49(.45-.54)ll -(.50)1 ,48(.45-.51) 7 ,48(.46-.50) 4 - .54(.51-.57) 2 -(.40)1 VPD/LPD ,90(.80-.97)10 -(.68)1 .65(.60-.81) 7 .89(.83-.93) 4 - ,84(.77-.9l) 2 -(.87)1 LAD/LPD 1.1 (.92-1.2) 10 -(.92)1 1.0(99-1.1) 7 1.0(99-1.1) 4 - .95(.93-.97) 2 -(1.0)1 VAD/VPD 1.1(1.0-1.1)11 -(1.0)1 1.0(.94-l.l) 7 1.0(.98-1.1) 4 - .96(.93-1.0) 2 -(1.1)1 NW/CL .32(.26-.42)ll -(.41)1 .49(.34-.56) 7 ,32(.29-.34) 4 - .33(.33-.34) 2 -(.68)1 NW/LPD .58(.54-.62)10 -(.55)1 .66(.59-.72) 7 .60(.53-.67) 4 - .52(.51-.53) 2 -(1.5)1 Vertebra 23 AN/CL .69(.67-.75) 8 -(.69)1 .59(.50-.68) 7 .58(.56-.94) 6 -(.66)1 .61 (.61-.62) 2 -(.72)1 PN/CL .28(.20-.31)58 -(.24)1 ,36(.28-.47) 7 .32(.30-.35) 6 -(.29)1 .31(.30-.31) 2 -(.28)1 LAD/CL ,62(.56-,71) 9 -(.83)1 .88(.64-l.l) 7 .64(.57-.71) 6 -(.56)1 .72(.69-.75) 2 -(.52)1 VAD/CL .58(.48-.73)10 -(.56)1 .57(.52-,64) 7 ,57(.53-.62) 6 -(¦53)1 .62(.60-,64) 2 -(.48)1 VAD/LAD .90(.82-.98) 9 -(.68)1 ,65(.57-.82) 7 .89(.86-.94) 6 -(.94)1 ,86(.81-,92) 2 -(.92)1 LPD/CL .56(.48-.82)I0 -(.63)1 .68(.53-,83) 7 .57(.52-.66) 6 -(.48)1 .65(.63-.67) 2 -(.47)1 VPD/CL .50(.42-.63)10 -(•50)1 ,52(.44-.59) 6 .53(.48-.59) 6 -(.43)1 .54(.52-.56) 2 -(.42)1 VPD/LPD .90(.76-.99)IO -(.79)1 .77(.71-.88) 6 ,92(.88-.95) 6 -(.89)1 .83(.77-,89) 2 -(.91)1 LAD/LPD 1.2(1.1-1.2) 9 -(1.3)1 1.3(1.2-1.4) 7 1.1(1.1-1.1) 6 -(1.2)1 1.1(1.1-1.1) 2 -(1.1)1 VAD/VPD 1.2(1.1-1.2)10 -(1.1)1 1.1(1.0-1.2) 6 1.1(1.0-1.2) 6 -(1.2)1 1.2(1.1-1.2) 2 -(1.1)1 NW/CL .53(.38-.81)10 -(.53)1 .61(.44-.76) 7 .5I(.46-.59) 6 -(.41)1 .53 (.51-.54) 2 -(.45)1 NW/LPD .95(.76-l.l)10 -(.84)1 .90(.83-.92) 7 ,89(.84-.95) 6 -(.86)1 .81(.80-.82) 2 -(.97)1 Hypural HDD/HL .46(.38-.55)I0 -(.34)1 .38(.29-.43) 6 .47(.44-.49) 6 -(.47)1 .45(.41-.48) 3 -(•44)1 HDD/HH .24(.22-.26)10 -(.20)1 .21(.15-.25) 6 .26(.24-.28) 6 -(.24)1 .24(.23-.25) 3 -(.24)1 HDD/HW .41(.36-.47)10 -(•36)1 .39(.28-.46) 6 ,45(.44-.47) 5 -(.46)1 .45(.40-.52) 3 -(.44)1 HL/HH .53(.47-.58)10 -(¦58)1 .55(.52-.58) 6 .55(.52-.58) 6 -(.51)1 .53(.51-.56) 3 -(.55)1 HW/HL 1.1 (.94-1.4) 10 -(.96)1 .97 (.89-1.0) 6 1.0(1.0-1.1) 5 -(1.0)1 1.0(.94-1.1) 3 -(1.0)1 HW/HH .59(.51-.66)10 -(.55)1 ,54(.5O-.60) 6 .56(.54-.60) 5 -(.51)1 .54(.49-,57) 3 -(¦55)1 HNL/HL .37(.30-.47)10 -(.22)1 .28(.22-.31) 6 .34(.29-.40) 6 -(.41)1 .32(.31-.33) 3 -(.37)1 290197, 290211, 421527, 481929, 488110); 6 vertebra number 1 (USNM 481923,488059, 488071, 488078, 488088, 1 USNM uncataloged); 8 vertebra number 22 (USNM 286181, 488056, 488066, 488076, 488089, 488098, 2 USNM uncataloged); 16 vertebra number 23 (USNM 286179, 288000,488044, 488045, 488055, 488063, 488064, 488067, 488068, 488070, 488094-488096, 3 USNM uncataloged); 3 hypurals (NCSM 4938; USNM 488069, 488104). Remarks.—All specimens have one or more ratios only within the range of values measured for extant Makaira nigri- cans (Tables 1, 2). Seven of the eight parasphenoids have only one or two ratios out of a possible five used for identification. Most predentaries were assigned to this taxon on the basis of ratio PW/PL; however, four predentaries (USNM 475415, 481936 (Plate 6g,i), 481938, 488011) have two ratios (PW/PL, PD/PL) characteristic only of M. nigricans (Table 2). All scap- ulas have curved articular surfaces for the first pectoral-fin ray. Eight of the 38 rostra have only one ratio (either Dl/Wl or D2/W2) on which to base an identification. Two specimens have denticles on the prenasal bone, six lack denticles, and the presence or absence of denticles for the other 30 rostra is not known. Specimens range in size from moderate (USNM 475421) to huge (USNM 475405, 481941 (Plate 6fh,j)). Some specimens (e.g., USNM 297407, 475406, 481943, 481949, 481976,481977) have a naturally worn tip, sometimes with the nutrient canals exposed at the distal end (e.g., USNM 481976, Plate la-c). USNM 290614 has a small abnormal growth (keel) on the dorsal tip of the bill. All vertebrae 1, 22, and 23 have at least two or three ratios with values that are characteristic of M. nigricans. Makaira cf. M. nigricans Lacepede, 1802 Plates Za-j, 9a-f Material.—2 parasphenoids (USNM 488049, 488116); 3 quadrates (NCSM 6944; USNM 476372, 481903); 5 rostra (NCSM 7427; USNM 286958, 481944, 481974, 1 USNM un- cataloged); 5 vertebra number 1 (USNM 481897, 481910, 488057, 488072, 488073); 5 vertebra number 22 (USNM 286177, 488033, 488079, 488080, 488086); 7 vertebra number 23 (USNM 476371, 481909, 488065, 488097, 3 USNM uncat- aloged); 4 hypurals (USNM 283735,481979, 2 USNM uncata- loged). Remarks.—Specimens having one or more ratios that lie outside the range of values measured for extant istiophorids (Tables 1, 2) are placed here because the values for some of the other ratios fall within the range characteristic of only Makaira nigricans. Some of these unusual specimens include the fol- lowing: parasphenoid USNM 488116 (Plate %d,f) with a PAT/ PAD ratio of 0.94, indicating a shallower depth of the paras- phenoid bone; quadrate NCSM 6944 (Plate Sa,j) with three ra- tios (QAW/QH=0.37, QAW/QHS=0.52, QMW/QML=1.1), 32 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Table 3.—Mean (x), observed range, and number of specimens examined (n) for each of eight rostral variables (ratios), and presence (P) or absence (A) of denticles on prenasal bone for seven istiophorid taxa from Lee Creek Mine. Abbreviations for ratios are explained in the text and in the legends to figures 3 and 4. Istiophoridae, gen. and Istiophorus Istiophorus cf. /. Character sp. indeterminate platypterus (x) (range) n platypterus (x) (range) n Makaira nigricans (x) (range) n Makaira cf. M. nigricans (x) (range) n Makaira purdyi (x) (range) n Makaira sp. (x) (range) n (x) (range) n Ratio Dl/WI .79(.65-.89)12 .73(.72-.75)3 _ ,79(.77-.80) 3 -(.86)1 - Hl/Dl 17(.10-.24) 6 .22(.21-.23)3 - ,15(.09-.19)14 .14(.1I-.I9)3 - -(•20)1 DD1/D1 .48(.39-.64) 6 .37(.35-.42)3 - .52(.41-.59)I4 ,59(.47-.80)3 - -(.43)1 D2/W2 .77(.61-.91)27 -(.69)1 -(.75)1 ,76(.66-.83)36 .90(.86-1.0)4 (.95)1 .76(.73-.77)5 H2/D2 ,14(.09-.22)16 .24(.22-.25)2 -(.27)1 .15(.06-.22)I6 .09(.06-.12)2 (.16)1 ,17(.16-.17)2 DD2/D2 49(.32-.53)16 .44(.43-.45)2 -(.39)1 .52(.40-.62)I6 -(.57)1 (.57)1 .51(.47-.55)2 P/VSPM -(.39)1 - - - - (.38)1 - DZ/P .31(.15-.41) 4 - - .49(.38-.59) 3 -(.81)1 (1.4)1 -(.35)1 Denticles on (P/A/?) n (P/A/?) n (P/A/?) n (P/A/?) n (P/A/?) n (P/A/?) n (P/A/?) n prenasal bone (8/6/22) 36 (0/1/4) 5 (0/1/0) 1 (2/6/30) 38 (2/0/3) 5 (1/0/0) 1 (1/2/2) 5 quadrate USNM 476372 with two ratios (QAW/QH=0.37, QAW/QHS=0.54), and quadrate USNM 481903 with one ratio (QAW/QHS=0.51) indicating a wide surface for articulation with the articular; five rostra having either one or both ratios Dl/WI, D2/W2, with values greater than 0.83, indicating a rounder cross section; five vertebra number 1 each with a NW/ LPD ratio greater than 0.70, indicating a more hourglass- shaped centrum; five vertebra number 22 with a mixture of unique ratios, two with a NW/LPD ratio greater than 0.74, indi- cating a more hourglass-shaped centrum; seven vertebra num- ber 23 with an array of unique ratios, including three with VPD/LPD ratios less than 0.69, two with LAD/LPD ratios greater than 1.39, and one with VAD/CL and VAD/LAD ratios less than any other istiophorids; and four hypurals (USNM 283735, 481979 (Plate 9d,e), 2 USNM uncataloged), have HL/ HH ratios (0.72, 0.61, 0.63, and 0.62, respectively) indicating a shorter height from tip to tip. All vertebrae 1, 22, and 23 have a minimum of two ratios (except one USNM uncataloged vertebra 23 with one ratio) with values that are characteristic of M. nigricans. Hypurals USNM 283735 and two USNM uncataloged specimens have only three ratios out of a possible seven on which to base an identification, and they have just one ratio characteristic of M. nigricans. Makaira purdyi Fierstine, 1999a PLATE ia-c Material.—1 rostrum (holotype, USNM 481933). Remarks.—The rostrum is morphologically distinct from any extant istiophorid in the following combination of charac- ters: (1) the fused portion of the premaxillae is short and stout with denticles covering at least the distal one-half of its dorsal surface; (2) at 0.25L, the cross section is nearly round (D2/ W2=0.95) (Fierstine, 1999a). Makaira sp. Plates 9g-l, \0a-d.g-i MATERIAL.—2 dentaries (USNM 286997, 475418); 1 pre- dentary (USNM 481931); 2 quadrates (USNM 488006, 488075); 5 rostra (NCSM 11223; USNM 285384, 475390, 481969, 1 USNM uncataloged); 1 scapula (USNM 290204); 1 vertebra 23 (USNM 290542); 2 hypurals (USNM 481981, 488102). Remarks.—These specimens have one or more ratios that fall within the observed range of values measured for both Ma- kaira indica and M. nigricans (Tables 1, 2). Both dentaries are from large individuals. Predentary USNM 481931 (Plate \0a,b) is massive, much larger than any extant istiophorid ex- amined. Four of the five quadrates have four of five possible ratios on which to base an identification, and all are from large individuals. Rostrum USNM 285384 has an eroded tip with both nutrient canals exposed. All rostra are from small- to moderate-sized individuals with the exception of USNM 475390 (Plate 9g-i), which came from a very large fish. The scapula has a curved articular surface and is from a very large- sized individual. The hypurals (USNM 481981 (Plate \0h.i), 488102) came from large-sized individuals and have HNL/HL ratios within the range of both M. indica and M. nigricans. cf. Makaira sp. Plate \0e,fj-l Material.—2 dentaries (NCSM 2990; USNM 475395); 2 quadrates (USNM 481916, 481919); 1 hypural (USNM 488007). Remarks.—These specimens have at least one ratio near the observed ranges of values measured for Makaira indica and M. nigricans, but they have one or more ratios that lie out- side the ranges of values measured for extant istiophorids (Ta- bles 1, 2). Some of these unusual specimens include the fol- lowing: dentaries NCSM 2990 (Plate 10/) and USNM 475395 with DAD/DJL ratios (0.96 and 1.04, respectively) indicating a NUMBER 90 33 deep interdentary joint; quadrates USNM 481916 (Plate lOk.l) and 481919 with QAW/QHS ratios (0.47 and 0.53, respective- ly) indicating a wider joint surface for articulation with the ar- ticular; and hypural USNM 488007 (Plate lOef) with three ra- tios (HL/HH=0.71, HW/HL=0.85, HNL/HL=0.11) indicating both a longer centrum length and that it came from a small- sized specimen. Genus Tetrapturus Raflnesque, 1810 Tetrapturus albidus Poey, 1860 Plates \\a-j, \2a-d Material.—3 articulars (USNM 290193, 488029, 488043); 7 maxillae (USNM 290203, 475393, 475402, 475424, 488036, 488037, 488040); 2 parasphenoids (USNM 488027, 488046); 2 quadrates (USNM 481908, 488008). REMARKS.—I have reservations about recognizing Tetraptu- rus albidus at Lee Creek Mine because its identification is based on poorly preserved features of fragmentary material. Articulars USNM 290193 (Plate 1 la.b) and 488043, all seven maxillae, both parasphenoids, and quadrate USNM 488008 are from individuals of greater body length than any extant species of Tetrapturus measured. Five of the seven maxillae are poorly preserved, with only two ratios available for identification. Both parasphenoids are incomplete and have only ratio, PAD/ PAW, on which to base an identification. Quadrate USNM 481908 (Plate I2a,b) has a QAW/QH ratio (0.31) that falls only within the range of values for Makaira indica, but its QMW/QAW ratio (0.53) is only within the range of values for T albidus. The other three ratios encompass an extant species of Tetrapturus. Quadrate USNM 488008 (Plate 12c,d) has two ratios (QMW/QAW=0.53, QMW/QML=0.61) characteristic only of values measured for T. albidus. Tetrapturus cf. T. albidus Poey, 1860 Plate \2e,f Material.—2 maxillae (USNM 488035, 488085). Remarks.—Both maxillae have MVW/MH ratios (0.46 and 0.47, respectively) outside the range of values measured for ex- tant istiophorids (Tables 1, 2); however, other ratios are within the observed range of values for Tetrapturus albidus. Discussion Comparison of Lee Creek Specimens to Other Fossil AND RECENT SPECIES.—Because fossil billfish have been re- viewed previously (Fierstine, 1974, 1978, 1990; Schultz, 1987), only specimens with direct relevance to Lee Creek and recent istiophorids are discussed herein. There are three widely recognized families of billfishes, each defined in part by its ros- trum. The Istiophoridae (marlin, sailfish, and spearfish) have an oval to round bill with paired nutrient foramina and have both fossil and recent representatives. The Xiphiidae (sword- fish) have a flattened bill with paired nutrient canals as well as a central chamber and have both fossil and recent representa- tives. The extinct Xiphiorhynchidae have an oval to round bill with one or more pairs of nutrient foramina and a central canal. Schultz (1987) recognized three questionable families of billfish, the extinct Blochiidae, extinct Paleorhynchidae, and extant Tetrapturidae. Too little is known about the first two families to determine if they are billfish (Fierstine, 1974), and the third should not be recognized (Fierstine and Voigt, 1996). Carroll (1988) placed many of these questionable billfish in the Xiphiidae, a decision that has no merit. Most early workers (see references in Fierstine, 1978, and Schultz, 1987) placed fossil specimens of Istiophoridae into new or existing fossil species of Istiophorus. Perhaps other genera and extant species were not considered because the sys- tematics of the recent Istiophoridae was poorly understood. Robins and de Sylva (1960, 1963), Nakamura et al. (1968), and Nakamura (1983, 1985) revised the extant Istiophoridae and recognized three genera, Istiophorus, Makaira, and Tetraptu- rus, although the number of species in Istiophorus and Makaira was equivocal. I follow Robins and de Sylva (1960, 1963) and Robins et al. (1991), and not Nakamura (1983, 1985), in treat- ing both the sailfish and blue marlin as single, world-wide spe- cies (/. platypterus and M. nigricans, respectively) and not as separate Atlantic Ocean (/. albicans and M. nigricans) and Indo-Pacific Ocean (/. platypterus and M. mazara) forms. Ge- netic studies support the status of a single, world-wide species each of sailfish and blue marlin (Graves and McDowell, 1995). Because most early workers placed fossil billfish of disparate morphologies in Istiophorus (sailfish genus), I have reclassi- fied specimens into other genera where warranted. If data to make an accurate identification were lacking, I left the speci- mens in Istiophorus but put quotes around the generic name. "Istiophorus" robustus (Leidy, 1860) (AMNH 5684, holo- type, ?Pleistocene, Ashley River, South Carolina) is a short (140 mm) distal rostral fragment that was refigured by Hussa- kof (1908). Based on published accounts, the specimen is oval in cross section (long axis is dorsoventral) for most of its length. Nutrient canals were not discussed or figured. Denticles are probably restricted to the ventral surface (dorsal surface in Leidy's figure, but the specimen is probably upside down) (Fierstine, 1974). Schultz (1987) placed the specimen in Aglyp- torhynchus Casier, 1967 (questionable billfish), and considered it to have been collected in the Eocene. Without reexamining the specimen I do not think it is relevant to Lee Creek istio- phorids. "Istiophorus" rotundus Woodward, 1901 (BMNH P8799, holotype, Tertiary phosphate beds of South Carolina), is a very stout and round rostral fragment that is extremely mas- sive for its length (313 mm). Due to poor preservation, I mea- sured the specimen's width and depth 95 mm distal from its proximal end (W=93.8 mm, D=78.5 mm, D/W=0.84). The rostrum has never been sectioned to determine the presence, number, position, and size of nutrient canals. No denticles or alveoli are visible, and I was unable to determine which sur- 34 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY i_ ra tt S § 2 § fN !>, -^ (J K (N CN fN p- — r- 3| U OU Os ¦rt •^ 3 5 c 1 1 1 1 1 1 1 1 1 CN 1 NO no l i 1 1 l 1 1 1 1 1 1 i i 1 1 1 1 1 1 1 1 g.« CO <^ ^ S3 r ¦ V. 1-. £ O OO NO ft Os m -rf =: rj m n cn n r» r- CN fN ^ rs - a oo ro r- oo fN tN — m 1/1 OJ p^ ' **o ; oo — —' On —' "^" /—» -—• >o "™" r? vq 70 C1 £" ,' ^r~ ,* •"3" o f I 1*1 L m -^ I t | ^ cn _L* oo o r | | I l 1 n —¦ ro - o ^> oo o t ir1 o _ *^ ^ ^_> NO K 1 1 O | O y-j ro ^ 1 2 no o vo »n p- oo 1 I 1 m p- *n r r f r- — — rO _J r i rO ^r 1 1 1 r f cn — _; on ! r NO NO '5 .& 2 O os no o oo m no w-i ro OO r-~ t o* OO OO "5 = Os Wi fN P~ NO fN VTi NO rr ¦* q> P- Os 1 — OS NO OS n vo i/i ro Os r- ^ a- ON OO T r~- r*i m m p NO — , l l [ l l i ' Illlll l no r- oo i ¦4- o cc 1 1 1 1 1 — ro •**¦ c vf O O £ £• r- Y T T ^ o -T V, «n cn ^r vq T 00 OO OO si <¦". oc m ¦o p^ fN OO OO s= c ll o> CJj ii^, Os O c 1 | Illlll 1 I 1 I I j 1 | 1 1 1 | | NO ---- CO | 1 1 1 | ,£ g 73 w ¦*Js ^ •£ 1 1 1 1 ~ ^r «tf T" rj fN fN q rO r- CN o o ,g u NO T vo <0 — wo CJ> |* l" r |" |" |' _c 73 1 l I l i 1 1 Illlll I 1 1 1 1 P- OC rr 1 1 1 1 1 P^ •^ r- 1 1 1 1 1 "O rn NO m « CN si - ri o 1C, -r NO OO NO -r vq f. — rO U 3 =: i- ^ 3 g: u OU c 1 l l l l l | Illlll 1 1 I 1 1 | | l 1 1 1 1 1 i I 1 | I 1 | | -^ c- g 5„S w '5 ^ »H -5 ^ ^ C4 fsl fN r* O 00 — U oil c C3 ^, X — JX, "f ^ "/-> r^ | VO |* f l i l l l 1 1 u*i -rj- r- O — TT — no -<3- fN m OC — O & C w ^ ^ pJ o; no p- Is- in j n tj- >o NO fN ON -p — p^ rsOO .*> CT —* r-- rr oo P- tj- CC 0O OO £ ci. 1 1 1 1 l ro 0s ^ pl 1 fN 1 1 Os OO 1 CN 1 o t 1 1 — vo O cn tJJ so no »/-> _; vo ro w, V, in VO (N rO t "T* T" T* rf CI TT T ^ T Y ^ •n fN TT NO ^T p- U-i rO Us .2 « ifl f) O* f) f| p- NO r- v» ^ O ^ vi no w~i os m -T y> CC m fN fN o NO — NO cn m CN OO p- os ^o __; r- W"i NO NO n nC P- o> ^q ro NO NO -f CC P^ NO c c .2 CO 2 , _i ^ £ ^ 2^ hJ j S S 2 S o 3[j£<£££<&g§ = |*>>>>£ ^<<<< 2 2 2 2 ;¦> e --111 s e o ^ d a d C Pu x < x -^ ^ aQ'o'O'5' 2 ^ > ^ * * «aa a a a a =5 J J O V) CA 00 — J « S ? ^ g y 3 o z z -s 5 00 > Cu Q ^ -J J £ y d ^ ^ 0 on 00 < < < J NUMBER 90 35 3:2 Illlll I I I I I I I I I I I I I I I I Illlll I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I rj tN n n n rj oi *3- rn v, ti- "* tr TT «n IT) VJ *n in VI m »n *f t v( ^r m vt v, CN ^r rr m rn cr> CN r-j © o r- ~ — — _ — — TT OO rM NO ^r NC, NO ^ vr | CC ' -t NO' NO ro r- <* ¦* ON Os Ov 00 rn ro OO OO Vi o P- rO in © Is* CC — _, r- r- vi m r- IT) NO P- ^r NO _J _' m r- NO m CO v> in p- ¦* -^r r- CN — Vl VI o -^r ro cn lo^NO^ttin^t^No vop-p-r-p^oot^f^r^vooooo ¦^ no _j __; no ^ CN CN v> vo vo ro — Orn — VI rO fN m rn -r © in rj ~ T rj tf V> _ Vi CN p- 1 C7-- r> TT OO fN v—' cn CN cn v, o V) oi [ p CN _ «* rr. *T _ ^r CN ** V, OS V, CN •* r*- l*» rO rn •* CN CN o CN ^O rsi CN r- O* T r T m r CN f r- r- c> Vi « ^t Os o cn r> in CN , NO cc CN n ._- NO CC i/ i ^J* 1 1 tr. o CN «c. m p» _ NO _J cn O] NO cr- in cn m (N m cm o o r^ r~ I I I I I I I I I NO NO P- TJ; O ON P- in cn HO cc •^t CN NO ~ — o vs. ro — NO NO TJ> CC v> CN IT, m r- CN O NO p*-oc. I I I I I I I I I I I I I I I I I I I I I I I I I o n n t o ^ ^i ¦^ (N ^ «-)__' V-» fS I I I I I I I I I I I I I I I I I I I r. id "O CA Ti C cn ra Illlll Illlll I I I I I I I I I I I I ro CN ro T T T OOOOOOP-P-.QOP-OSOO O O — NO vo I I I I o a \t o t n vi I I I I I I I I I I I I NOfN P- ro O OS O CN OO oq_^p^oo_;oq—"invq m CN rn r-j NO - o NO ro NO cn o CN CO rn ON NO v> fN Q Q -J < j -J a. q 5 q a S ? > 3 > > a 9 n Q Q Q Q -J J < j _i a. ou p- =d c o -j rs J o u _i rj ( ) _j -J > y D ° > > < < £ > i— o a DDq n q a a ^ 7" < < < a. a. o. < < _i > z z < D- _j > > _i -> > J > H »; " J j u u I I -S 5 2 < < z z e < a. -J > < _i 0 <- u Q _i U o 0. Q Q a. c a. _i Q < 0 > s a a c D D D * X Q Q X X J X J 7. -> _J ;> > -J > z ^ p. I X X X X X X 36 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Table 5.—Sample score sheet for species identification of selected predentary bones from Lee Creek Mine. Spe- cies score code: 0=outlier, 1= Istiophorus platypterus, 2-Makaira indica, 3=M. nigricans, 4=Tetrapturus albidus, 5=T. angustirostris, 6=T. audax, 7=T. belone. Catalog number PW/PL PD/PL PD/PW Value Score Value Score Value Score Species USNM 291066 .83 3 .50 3 .60 1,3,4,6 Makaira nigricans USNM 475399 .62 1,2,3,5 .40 2,3 .64 1,2,4,6 Makaira indica USNM 475412 .59 1,2,3,5,7 .35 1,2,3 .59 1,3,4,6 Istiophoridae USNM 481931 .58 1,2,3,5 45 3 .76 1,2 Makaira sp. USNM 481956 1.54 0 1.02 0 .67 1,2,6 Istiophoridae face is dorsal or ventral. Schultz (1987) placed /. rotundus in Xiphiorhynchus, but it probably belongs to the genus Ma- kaira. Until more is known about the morphology of /. rotun- dus, meaningful comparison with Lee Creek and recent istio- phorids is fruitless. "Istiophorus" solidus (Van Beneden, 1871) (IRSNB P643, holotype, late Eocene, Ghent, Belgium) is a poorly preserved rostral fragment with one pair of round nutrient canals that are placed more toward the lateral periphery than in other istio- phorids. The specimen is 17 mm long, 27.3 mm wide, and 17.3 mm deep at its proximal end (D/W=0.63). The exact position and size of the nutrient canals have not been recorded. Paired grooves run the length of the dorsal surface and indicate the presence of paired prenasal bones. Schultz (1987) placed /. solidus in Xiphiorhynchus, but because of the lack of a central canal and the presence of prenasal bones, I believe the speci- men is an istiophorid. Its morphology is unlike any species of recent billfish or billfish found at Lee Creek Mine, but the specimen is poorly preserved. Istiophorus calvertensis Berry, 1917 (USNM 9344, holo- type, late Miocene, Eastover Formation, Tar Bay, James River, Virginia), is a distal rostral fragment 310 mm long. Berry (1917) originally thought the specimen was collected in the Calvert Formation, but according to both R.E Weems (pers. comm., 1996) and L.W Ward (pers. comm., 1996) there is very little, if any, Calvert Formation at Tar Bay, and chances are very strong that it was collected in the late Miocene Easto- ver Formation of Ward and Blackwelder (1980). Both Weems and Ward admit, however, that the specimen could have been collected in the Yorktown Formation, although it is unlikely due to the Yorktown Formation's minor presence at Tar Bay. The specimen is 25.8 mm deep (DI) and 36.0 mm wide (Wl) at its proximal end, which approximates 0.5L. A pair of nutri- ent canals are exposed, each measuring 7.0 mm high (HI) and located 11.0 mm from the dorsal surface of the rostrum (DD1). The anterior extension of the prenasal groove (P) is 170 mm from the distal tip; 0.25L is estimated to be 134 mm from the distal tip. The rostrum is 21.2 mm deep (D2) and 30.8 mm high (W2) at 0.25L. On the dorsal surface of the rostrum, denticles extend posteriorly from the distal tip for 57.3 mm (DZ). Based on ratios computed from these measurements (Table 6) and on reexamination of the specimen (Fierstine, 1998), I identify this specimen as Istiophorus cf. /. platypterus. It has relatively large nutrient foramina (Hl/Dl) similar only to re- cent /. platypterus (Table 1), but the placement of the canals (DD1/D1) and area of the dorsal surface covered with denticles (DZ/P) are not sailfish-like. These latter two features are dis- counted because the dorsal denticular pattern may not have been completely preserved, and ratio DD1/D1 is nearly within the range of values for recent /. platypterus. Comparison of USNM 9344 to rostra from Lee Creek Mine (Table 3) shows that the Dl/WI ratio is similar to that in specimens identified as /. platypterus, the Hl/Dl ratio is unlike that in any specimen studied, and the DZ/P ratio is similar to that in specimens iden- tified as Makaira sp. On the basis of its large nutrient canals, Schultz (1987) clas- sified /. calvertensis in Pseudohistiophorus De Buen, 1950, a genus that Nakamura (1983) synonymized with Tetrapturus. Schultz made an erroneous decision because, as noted above, Table 6.—Eight rostral variables (ratios) for six fossil istiophorid taxa from localities other than Lee Creek Mine. The presence or absence of denticles on the prenasal bone is unknown for these taxa. Abbreviations for ratios are explained in the text and in the legends to figures 3 and 4. Istiophorus cf. /. platypterus. Makaira helgicus, Makaira courcelli, Makaira panamensis, Makaira teretirostris. type of /. calvertensis type specimen type specimen type specimen type specimen Makaira nigricans (USNM 9344) (IRSNB Pl 117) (MNHNP 250) (USNM 181710) (depository unknown) (LACM 17693) Dl/WI .72 - .67 - .87 - Hl/Dl .27 - - - .10 - DD1/D1 .43 - - - .58 - D2/W2 .69 .80 .66 .76 .84 .76 H2/D2 - .15 - .27 .11 .17 DD2/D2 - .34 - - .56 .56 P/VSPM - - - - - - DZ/P .34 - - - - - NUMBER 90 37 the size of the canals relative to the depth of the rostrum fits solely within the observed range of values for /. platypterus (Table 1). Fierstine (1990) thought critical review would syn- onymize /. calvertensis with M. nigricans; however, based on the information presented herein, the specimen belongs to the genus Istiophorus. Istiophorus platypterus (Shaw and Nodder, 1792) was iden- tified from a single, partial trunk vertebra (UCMP 125228) in upper Pliocene sediments, San Diego Formation, San Diego County, California, by Gottfried (1982). Until now, this verte- bra was the only fossil record of a sailfish in the literature since the revision of the extant Istiophoridae (Robins and de Sylva, 1960, 1963; Nakamura, 1983, 1985). Because the specimen's exact position in the vertebral column is unknown, and because I studied only vertebrae from Lee Creek Mine that could be ac- curately identified to position 1, 22, 23, or 24 (hypural), I was unable to compare it to Lee Creek material. Makaira belgicus (Leriche, 1926) (IRSNB Pl 117, holotype, middle Miocene, Anvers, Belgium) is a distal rostral fragment measuring 200 mm long, 32.3 mm wide (W2), and 25.9 mm deep (D2) at its proximal end. Prenasal bones are indicated by grooves, and paired nutrient canals are visible in cross section. Based on ratios in Table 6, the specimen falls within the range of values of recent M. nigricans (Table 1) and of M. nigricans from Lee Creek Mine (Table 3, except for the nutrient canals being closer to the dorsal surface of the rostrum (DD2/D2)). Makaira courcelli (Arambourg, 1927) (MNHNP 250, holo- type, early Pliocene, Algeria) consists of two rostra and several fragments. One well-preserved rostrum is 287 mm long, with the following widths and depths in mm: Dl=21.7, Wl=32.6, D2=16.4, W2=24.8. Ratios for this specimen are listed in Ta- ble 6. The other rostrum is crushed at its proximal end and measurements were not taken. Each rostrum contains one pair of nutrient canals, but their size and position have not been measured. Denticles are restricted to the ventral and lateral sur- faces of the well-preserved specimen, and paired prenasal grooves are present. Arambourg (1927) originally placed the specimens in Xiphiorhynchus; however, Schultz (1987) placed the specimens in Makaira and gave the age of the locality as late Miocene. Based on the ratios given in Table 6 and the dis- tribution of the denticles, I agree with Schultz's identification. The lack of other morphological information precludes a mean- ingful comparison between M. courcelli and other fossil and extant istiophorids. Makaira indica (Cuvier, 1832) was identified from a nearly complete head (including pectoral and pelvic girdles and fins) from the early Pleistocene, ?Cabatuan Formation, Luzon, Phil- ippines, by Fierstine and Welton (1983). The specimen was identified by its rigid pectoral fin, a diagnostic feature of the black marlin. Until the present study, this specimen was the only record of a black marlin in the paleontological literature. Makaira cf. M. nigricans Lacepede, 1802, was identified from an incomplete, disarticulated skull (USNM 375733) in the Eastover Formation, late Miocene, Virginia (Fierstine, 1998), and from a nearly complete rostrum (USNM 358534) in the Gatun Formation, late Miocene, Panama (Fierstine, 1999b). The Eastover specimen is similar to recent M. nigricans in 15 of 19 ratios, but it is dissimilar in four, three of which are out- side the observed range of all extant istiophorids. The Gatun specimen is similar to recent M. indica in 15 of 18 ratios, and it is similar to recent M. nigricans in 16 of 18 ratios. Two ratios that are similar to recent M. indica, but not M. nigricans, are discounted because they involve the denticular pattern on the dorsal surface of the rostrum, features that may have been in- completely presereved. Makaira panamensis Fierstine, 1978 (USNM 181710, holo- type, late Miocene or early Pliocene, Chagres Sandstone, At- lantic coast of Panama), was described from a large neurocrani- um with a poorly preserved rostrum attached. Except for its unique features (size of myodome, length of orbit, and relative size of nutrient canals), the specimen is most similar to recent blue marlin and black marlin; hence, the rationale for recogniz- ing it as a new species of Makaira. If the rostrum without the neurocranium had been among material collected at Lee Creek, it would have been listed under "cf. Makaira sp." Fierstine (1978) believed the Chagres Sandstone was late Miocene based on Woodring (1957, 1970, 1973), but its age is now considered to be late Miocene or early Pliocene (Woodring, 1982) or Pliocene (Coates et al., 1992). Makaira sp. was identified from several bones from late Mi- ocene localities in Southern California. Fierstine and Applegate (1968) studied a distal rostrum (LACM 17693) and predentary bone (LACM 16074) from separate localities in Orange Coun- ty, California, and Fierstine and Welton (1988) examined sev- eral associated bones (articular, dentary, preoperculum, ptery- giophores) of a single individual marlin (UCMP 118559) from the San Mateo Formation, San Diego County, California. The specimens were originally identified as Makaira sp. because of a lack of recent comparative material, but now that skeletal ma- terial is available, I have reexamined them with the following results. Extant Makaira nigricans (Table 1) is the common identifica- tion for all three ratios computed for the rostrum (Table 6). Be- cause the ratios of the predentary (PW/PL=0.62, PD/PL=0.37, PD/PW=0.60) fall within the observed range of values (Table 2) for both Istiophorus platypterus and M. nigricans, it is identi- fied as Istiophoridae, genus and species indeterminate. Analysis of the articular and dentary bones separately, as in the Lee Creek fossils, yields an identification of Istiophoridae, genus and species indeterminate, for both; however, when they are considered as skeletal elements from a single fish, then M. nigricans becomes the obvious choice. The ratios of the articu- lar are within the observed values for sailfish, blue marlin, and white marlin, whereas the ratio of the dentary is within the ob- served values for blue marlin and striped marlin; therefore, the common identification for the two bones is M. nigricans. Makaira teretirostris (Van Beneden, 1871) (?middle Mi- ocene, Belgium, exact locality unknown) is a large, distal ros- tral fragment (520 mm long) with paired nutrient canals and prenasal bones. Denticles and alveoli are neither mentioned nor 38 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY figured. The original description of the specimen was based on a cast and an artist's drawing, and disposition of the type is un- known. Schultz (1987) synonymized the specimen with M. bel- gicus and gave the type locality as southern France and the age as Pliocene. I made measurements from the drawing in Van Beneden (1871) and computed ratios (Table 6). Based on this analysis, M. teretirostris is slightly outside the observed range of ratios of recent M. nigricans (compare Tables 1, 6) but is within the range of values of M. cf. M. nigricans from Lee Creek Mine (Table 3). Without more information about the specimen, I recognize M. teretirostris and do not synonymize it with M. belgicus. Lawley (1876) described a well-worn, elongate, slender ros- trum from lower Pliocene rocks, Orciano, Italy, as Brachyrhyn- chus vanbenedensis. The specimen was redescribed, refigured, and placed into the living Histiophorus herschelii (Gray, 1838) by Barbolani (1910). Nakamura et al. (1968) synonymized H. herschelii with M. nigricans, and more recently Schultz (1987) placed B. vanbenedensis in synonymy with M. teretirostris. Until the specimen is studied further, I follow Schultz (1987). In summary, the family Istiophoridae has a fossil history from middle Miocene to recent, with the qualification that "Is- tiophorus" solidus (late Eocene, Ghent, Belgium) may be an is- tiophorid. Makaira belgicus (middle Miocene, Anvers, Bel- gium), Istiophorus cf. /. platypterus (late Miocene, Eastover Formation, Virginia, United States), and Tetrapturus albidus (early Pliocene, Yorktown Formation, North Carolina, United States) are the oldest known species within their respective genera. The temporal distribution of the Istiophoridae is given in Figure 7. Interspecific and Intraspecific Variation.—Any study of variation at Lee Creek Mine is dependent upon two factors: the number of fossil bones identified to species and the number of bones examined for each extant species. Only Makaira nig- ricans is sufficiently represented at Lee Creek to make a statis- tically meaningful comparison, and then only for a few bones. Significant differences exist only between the predentary, rostrum, scapula, and vertebrae 1 and 23 of extant M. nigricans and M. nigricans from Lee Creek Mine (Table 7). The preden- tary in the Lee Creek specimens tends to be wider (PW/PL), deeper (PD/PL), and rounder in cross section (PD/PW). The rostra from Lee Creek tend to have rounder cross sections throughout their length, and the nutrient canals are smaller dis- tally (H2/D2) and are more ventrally placed both proximally (DD1/D1) and distally (DD2/D2). In addition, the distal dorsal surface of the rostrum (DZ/P) is covered with more denticles than in the extant blue marlin. The scapulae from Lee Creek Mine have a narrower articular surface (SNW/SL, SNW/ SGW). First vertebrae from Lee Creek Mine exhibit a narrower anterior articular surface (LAD/CL) and a more constricted centrum (NW/LPD) with respect to their length. Twenty-third vertebrae of the blue marlin from Lee Creek have centra that are less depressed anteriorly (VAD/CL) and more constricted (NW/LPD) with respect to their posterior width. Blue marlin from Lee Creek Mine are more similar to extant black marlin than to extant blue marlin in two features (Table 8). The predentary bone in Lee Creek material is rounder (PD/ PW) than in extant blue marlin, and the dorsal surface of the distal rostrum has more denticles (DZ/P). So few vertebrae of extant black marlin were examined that no meaningful compar- ison with Lee Creek specimens was undertaken. Table 7.—Results of the unpaired /-test (Welch's modification) to determine significant differences between the means of variables (ratios) of the preden- tary, rostrum, scapula, and two vertebrae among Makaira nigricans from Lee Creek Mine and extant M. nigricans. Abbreviations for ratios are explained in the text and in the legends to figures 2-5. (*=P<0.05; **=P<0.01; »**=p<0.001.) Extant M. nigricans Ratio compared with Lee Creek M. nigricans Predentary PW/PL * PD/PL *«* PD/PW *** Rostrum Dl/WI «* DD1/D1 *** D2/W2 *** H2/D2 «* DD2/D2 ** DZ/P * Scapula SNW/SL •• SNW/SGW * Vertebra 1 LAD/CL *** NW/LPD * Vertebra 23 VAD/CL • NW/LPD * Table 8.—Results of the unpaired /-test (Welch's modification) to determine significant differences in the means of the same variables (ratios) listed in Ta- ble 7 for the predentary, rostrum, and scapula among Makaira nigricans from Lee Creek Mine and extant M. indica. Abbreviations for ratios are explained in the text and in the legends to figures 2-5. (n.s.=not significant (P>0.05); *=P<0.05; **=P<0.01; ***=P<0.001.) Extant M. indica Ratio compared with Lee Creek M. nigricans Predentary PW/PL *** PD/PL ** PD/PW n.s. Rostrum Dl/WI * DDI/DI • *• D2/W2 *** H2/D2 *»? DD2/D2 • * DZ/P n.s. Scapula SNW/SL ** SNW/SGW * NUMBER 90 39 Epoch MYA Species Pleistocene Pliocene M i o c e n e 1 a t e m i d e a r 1 y . 1.0 - 2.0 - 3.0 - 4.0 . 5.0 - 6.0 - 7.0 . 8.0 . 9.0 -10.0 -11.0 -12.0 .13.0 -14.0 -15.0 ¦ 16.0 ¦ 17.0 -18.0 ¦19.0 ¦20.0 ¦ 21.0 ¦22.0 to & to o 3 o to o -s: .| s o (J 5 ¦ — to 3 .Nj .§ a a s lO s •5! to K a 5 t>0 s to c g 3 a a, 5 a I I ¦f S to fc Ol e 5 ¦—i to a to k 3 I to 3 u to 3 3 I FIGURE 7.—Temporal distribution of the family Istiophoridae. Arrow indicates species that exist in recent time. Time scale is from Gibson (1983:38 (uncopyrighted)). Some of the rostra at Lee Creek Mine demonstrate abnormal- ities found in extant billfish (Gudger, 1940; Morrow, 1951; Fi- erstine and Voigt, 1996). For example, rostrum USNM 481976 (Plate la-c) has a foreshortened tip with nutrient canals ex- posed distally, and rostrum USNM 475409 (Plate 2h) has un- equal-sized nutrient canals. Rostrum USNM 481984 (Plate 2i,j) and a few others appear eroded, possibly by stomach acids after being consumed by a predator. Distribution and Life History of Recent Species.— The following information was taken liberally from Nakamura (1983, 1985) unless otherwise indicated. Emphasis is placed on species that inhabit the Atlantic Ocean and are found at Lee Creek Mine. In general, istiophorids are distributed throughout tropical and subtropical waters, some entering temperate cli- mates. All are oceanic and epipelagic species that usually favor waters greater than 20°C. Sexes are separate and are indistin- guishable externally. Each species has a distinct reproductive season and spawning ground, and mature individuals spawn several times each season by broadcasting gametes (Hopper, 1986). Adults are opportunistic predators, consuming cephalo- pods (squids) and many different species of pelagic fishes, in- cluding members of the Carangidae (jacks), Clupeidae (her- rings, pilchards, sardines), Coryphaenidae (dolphin-fish), Scombridae (mackerels, tunas, and allies), and Trichiuridae (snake mackerels, cutlass fishes). Although age estimates are given below for some species, aging of billfish is imprecise be- cause of the difficulty in establishing annual growth patterns in calcified structures (Hill et al., 1989). Frazier et al. (1995) con- cluded the function of the rostrum was unclear. It may serve in one or more capacities, possibly in hydrodynamics, food cap- ture (spearing/slashing), or defense/aggression, but the fish can get along without its bill because there are numerous records of apparently healthy billfishes with damaged, malformed, or missing rostra. Istiophorus platypterus (sailfish) inhabits the Atlantic, Indi- an, and Pacific oceans, possibly the Mediterranean Sea, and 40 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY sometimes enters the Red Sea through the Suez Canal. It is considered to be the least oceanic of the billfishes, often mi- grating to near-shore waters. Its distribution is influenced by wind and water temperature, favoring temperatures of 21°-28°C. In the western North Atlantic and western Pacific oceans, /. platypterus migrates northward in an extension of warm water during the summer and migrates southward with the onset of cold weather to form loose schools of up to 30 in- dividuals. The sailfish reaches a total length of around 3.2 m (tip of bill to tip of tail) and a weight of 58 kg in the Atlantic Ocean and reaches around 3.4 m total length and 100 kg in weight in the Indo-Pacific Ocean. Females are consistently larger than males (Jolley, 1974). Tagged sailfish are generally recaptured near their original release site; however, one speci- men caught off Isla Mujeres, Mexico, in the Caribbean was recaptured 2596 km distant off La Guiria, Venezuela (Nation- al Marine Fisheries Service, 1994). De Sylva (1957) reported that sailfish grow rapidly, attaining a weight of 9.1 kg within one year and having an estimated life span of two to three years. Jolley (1974), using a different aging method, found that sailfish in the Atlantic Ocean reached an age of seven years, with ages two to four being the most numerous. Makaira indica (black marlin) is restricted to tropical and subtropical waters of the Indian and Pacific oceans, except for occasional strays into the Atlantic Ocean via the Cape of Good Hope. It is found in waters with surface temperatures ranging from 15°-30°C and is often found as close to land masses as is Istiophorus platypterus. The distribution and abundance of M. indica off Natal are positively influenced by the presence of submarine canyons, some reaching a depth of 600 m within one km of shore (van der Elst, 1990). The black marlin reaches a total length of 4.48 m and a weight of 708 kg. Females grow more rapidly than males so that at any giv- en age males are much smaller than females (Hopper, 1986). Fish tagged off northern Queensland, Australia, generally are recaptured north or south of their point of release. One speci- men released off Baja California was recaptured north of New Zealand, thereby making a trans-Pacific and trans-equatorial migration of 5700 km (Pepperell, 1990). Makaira nigricans (blue marlin) inhabits the tropical and temperate waters of the Atlantic, Indian, and Pacific oceans. It is the most tropical of all billfishes, favoring blue water (depths greater than 100 m) at surface temperatures around 24°C (22°-31°C). The blue marlin in the Atlantic Ocean reaches a total length of approximately 4.0 m and a weight of 580 kg, whereas in the Pacific Ocean it reaches a total length of approximately 4.5 m and a weight of 906 kg. Although most tagged blue marlin are recaptured near their point of re- lease, one fish traveled at least 16,871 km from the North At- lantic Ocean off Delaware, across the equator to the Indian Ocean near Madagascar, presumably via the Cape of Good Hope (National Marine Fisheries Service, 1994). As in M. in- dica, females grow more rapidly than males, so that at the same age females are much larger than males (Hopper, 1986). Both sexes become sexually mature at around six to eight years of age. During the reproductive season (usually sum- mer), smaller males outnumber the larger females by as much as six to one (Hopper, 1990). Based on preliminary estimates from blue marlin caught off Kona, Hawaii, longevity is at least 18 years for males and at least 27 years for females (Hill etal., 1989). According to Browder and Prince (1990), blue marlin are most abundant off the mid-Atlantic coast in the summer. May and June are probably spawning months for blue marlin off Florida and the Bahamas. Adult fish off Cape Hatteras in June appeared to have already spawned. Tetrapturus albidus (white marlin) inhabits much of the At- lantic Ocean, including the Gulf of Mexico and Caribbean Sea, from 45°N to 45°S in the western South Atlantic and to 35°S in the eastern South Atlantic. A few individuals are known from the Mediterranean Sea and from off Brittany, France. The white marlin is usually found in blue water and favors surface temperatures greater than 22°C. Steep drop- offs, submarine canyons, and shoals are often scenes of im- portant feeding concentrations. They migrate to subtropical waters to spawn, with peak spawning in the summer. Females become larger than males, and during the spawning season males outnumber females on the spawning grounds. Most tagged white marlin were recaptured nearby, some after two or more years at large. Individuals, however, can travel long distances. One fish released in the United States Virgin Is- lands in 1991 was recaptured one year later off Mohammedia, Morocco, a transoceanic and transequatorial movement of 5840 km (National Marine Fisheries Service, 1994). White marlin attain a maximum size of 2.0 m (lower jaw to fork of caudal fin) and a weight of 79 kg. Tetrapturus audax (striped marlin) and T angustirostris (shortbill spearfish) inhabit mainly the tropical, subtropical, and temperate waters of the Pacific and Indian oceans, occa- sionally straying into the Atlantic Ocean via the Cape of Good Hope. Striped marlin prefer more temperate water than other billfishes, and shortbill spearfish are more oceanic, favoring waters greater than 900 m in depth. Tetrapturus belone (Medi- terranean spearfish) is limited in distribution to the Mediterra- nean Sea, and T. pfluegeri (longbill spearfish) is widely distrib- uted in the Atlantic Ocean from approximately 40°N to 35°S. Striped marlin attain the largest size and weight of any species of Tetrapturus, reaching a length (tip of bill to fork of caudal fin) of 3.5 m and a weight of 200 kg. Maxiumum body lengths (tip of lower jaw to fork of caudal fin) and weights, respective- ly, are approximately 2.4 m and 70 kg for the Mediterranean spearfish, 2.0 m and 52 kg for the shortbill spearfish, and 2.0 m and 45 kg for the longbill spearfish. A striped marlin tagged off Cabo San Lucas, Baja California Sur, Mexico, was recaptured near Norfolk Island in the South Pacific, a distance of 9600 km (Squire and Suzuki, 1990). Movement patterns of spearfishes are poorly understood. Implications of Distribution and Life History of Re- cent Species for Lee Creek Mine Fossils.—The presence of Makaira indica at Lee Creek Mine is unexpected based on NUMBER 90 41 its present distribution. During or prior to the early Pliocene, the black marlin may have had access to the North Atlantic Ocean via a more favorable environment than the Cape of Good Hope. According to Coates et al. (1992), the final clo- sure of the Isthmus of Panama occurred around 3.5 Ma BP, 1.0-1.5 Ma after Yorktown time at Lee Creek (Hazel, 1983). Coates et al. (1992) also believed that depths prior to closure ranged from shallow to shallow inner shelf (<200 m) and up- per slope (200-800 m) on the Caribbean side of Panama to a trench-slope environment on the Pacific side. Makaira indica is often found near land masses, islands, and coral reef areas (Nakamura, 1985). Therefore, during and prior to Yorktown time the Panama region was not a barrier but was a route for black marlin migrating between the Atlantic and Pacific oceans. Whitmore and Stewart (1965) believed the Canal Zone was the site of a narrow seaway throughout most of the Tertiary. There is other evidence that the Panama seaway was a migra- tion route for vertebrates. According to Purdy et al. (this vol- ume), the sharks Carcharhinus macloti and Triaenodon obesus are present at Lee Creek Mine. Both species inhabit only the tropical Indo-Pacific Ocean today (Compagno, 1984). Gillette (1984) concluded that the ichthyofauna of the Miocene Gatun Formation (Panama) resembles the marine faunas of North Carolina (Pungo Formation), Ecuador, and the Antilles. Final- ly, the presence of the sirenian Metaxytherium calvertense Kellogg in both the middle Miocene Calvert Formation (Mary- land, United States) and the correlative Montera Formation in Peru (de Muizon and Domning, 1985) and the presence of Met- axytherium arctodites Aranda-Manteca, Domning, and Barnes, 1994, in Mexico and California, further suggests there were faunistic exchanges between the Atlantic and Pacific oceans during the mid-Neogene. The presence of Istiophorus platypterus, Makaira nigri- cans, and Tetrapturus albidus at Lee Creek Mine fits the dis- tribution of extant species in the northwestern Atlantic Ocean today. The blue marlin is much more abundant at Lee Creek Mine than are the other two species, but this may be due to collection bias or to difficulty in identifying isolated and frag- mentary elements. Recent Makaira nigricans and most other billfish favor blue water (>100 m depth) and water temperatures of 22°-31°C, so it is likely that billfish during Lee Creek time preferred the same environments. Based on Gibson's (1983) conclusion that zone 1 of the Yorktown Formation was deposited at water depths of 80 m to 100 m and zone 2 was deposited at depths of less than 30 m, I believe that most, if not all, adult billfish spec- imens were collected in zone 1 (basal Yorktown Formation). If one assumes, based on my sample of 38 recent blue marlin where sex was known, that all fish with a rostrum width greater than 23.0 mm (measured at one-fourth bill length, or 0.25L) are female, then blue marlin at Lee Creek Mine have a sex ratio of approximately one male to two females. The ratio of males to females would be much larger in a spawning population (Hop- per, 1990); therefore, I hypothesize that the present pattern of blue marlin migrating northward in the western North Atlantic Ocean during the summer to feed after spawning was estab- lished during Yorktown time or earlier. Frazier et al. (1995) and Fierstine and Crimmen (1996) have reviewed the literature of extant istiophorids impaling inanimate and animate objects with their rostrum. They noted that healthy billfish have been captured with shortened bills and assumed damage occurred when the fish broke away from its impalement. Although the reasons for spearing behavior are unclear, foreshortened bills at Lee Creek Mine is evidence that the behavior was established during Yorktown times or earlier. Study Limitations.—Purdy et al. (this volume) have point- ed out many of the pitfalls that reduce understanding of the Lee Creek Mine fauna. They believed careful excavation of units 1-3 of the Yorktown Formation, rather than surface collecting from spoil piles, would have produced articulated skeletons and less fragmentary material. Strict stratigraphic control is critical for exploration of changes in species composition (seasonal or throughout Yorktown time) and for collecting paleoecological data. Partial or whole, articulated skeletons would have made identification more convincing because identification would have been based on measurements of several elements, not just those from one fragmentary bone. Also, more complete skeletal material from recent species, especially black marlin and most of the species of Tetrapturus, would have increased our knowl- edge of interspecific and intraspecific variation of the Istio- phoridae and would have made comparisons more certain. Conclusions The identification of Istiophorus platypterus, Makaira indi- ca, M. nigricans, M. purdyi, and Tetrapturus albidus at Lee Creek Mine is very significant. It is the only record of Makaira purdyi, the first fossil record of the genus Tetrapturus, specifi- cally T. albidus, the second fossil record of /. platypterus and M. indica, and the first record of /. platypterus, M. indica, M. nigricans, and T. albidus from fossil deposits bordering the At- lantic Ocean. The presence of M. indica at Lee Creek suggests that the Panama seaway may have been a migration route for billfish during the early Pliocene. The concentration of billfish at Lee Creek Mine supports the contention that the Yorktown Forma- tion represents a tropical to warm temperate (21°-28°C) ocean- ic environment that was deposited at depths greater than 100 m. Based on the size of isolated bones at Lee Creek Mine, /. platypterus and T. albidus are estimated to have reached much larger sizes and weights than their living representatives. Ma- kaira nigricans from Lee Creek are more similar to extant M. indica than to extant M. nigricans in two features, a more round predentary and a greater area of denticles on the dorsal surface of the distal rostrum. 42 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Literature Cited Arambourg, Camille 1927. Les poissons fossiles d'Oran. Materiaux pour la Carte Geologiques del'Algerie. seriesl (Paleontologie), 6:1-298. Aranda-Manteca, Francisco J., Daryl P. Domning, and Lawrence G, Barnes 1994. A New Middle Miocene Sirenian of the Genus Metaxylherium from Baja California and California: Relationships and Paleobiogeo- graphic Implications. In A. Berta and T.A. Demere, editors, Contri- butions in Marine Mammal Paleontology Honoring Frank C. Whitmore, Jr. Proceedings of the San Diego Society of Natural His- tory, 29:191-204. Barbolani, G. 1910. L'Histiophorus herschelii (Gray) nel terziario superiore. Paleon- tographia Italia. 16:1-22, 2 plates. Berry, E.W. 1917. A Sailfish from the Virginia Miocene. American Journal of Science, 43:461-464. Browder, Joan A., and Eric D. Prince 1990. Standarized Estimates of Recreational Fishing Success for Blue Marlin and White Marlin in the Western Atlantic Ocean, 1972-1986. In Richard H. Stroud, editor, Planning the Future of Billfishes, Part 2: Contributed Papers. Proceedings of the Second In- ternational Billfish Symposium, Kailua-Kona, Hawaii, 1-5 August 1988, pages 215-229. Savannah, Georgia: National Council for Ma- rine Conservation, Inc. Carpenter, Kent E., Bruce B. Collette, and Joseph L. Russo 1995. Unstable and Stable Classifications of Scombroid Fishes. Bulletin of Marine Science, 56:379-405. Carroll, Robert L. 1988. Vertebrate Paleontology and Evolution. 698 pages. New York, New York: W.H. Freeman and Company. Casier, Edgard 1967. Poissons de 1'Eocene inferieur de Katharinenhof-Fehmam (Schleswig-Holstein). Bulletin de I'lnstitut Royal des Sciences Na- turelles de Belgique, 43:1-23. Coates, Anthony G., Jeremy B.C. Jackson, Laurel S. Collins, Thomas M. Cro- nin, Harry J. Dowsett, Laurel M. Bybell, Peter Jung, and Jorge A. Obando 1992. Closure of the Isthmus of Panama: The Near-Shore Marine Record of Costa Rica and Western Panama. Bulletin of the Geological Soci- ety of America, 104:814-828. Compagno, Leonard J.V. 1984 Sharks of the World; An Annotated and Illustrated Catalogue of Shark Species Known to Date, Part 2: Carcharhiniformes. Food and Agriculture Organization of the United Nations (FAO) Fisheries Synopsis, I25(2):i-x+251-655. Cuvier, Georges 1832. Tetrapturus indicus. In G. Cuvier and A. Valcenciennes, Histoire Nalurelles Poissons, 8:554—626, Davie, Peter S. 1990. Pacific Marlins: Anatomy and Physiology. 88 pages. Palmerston North, New Zealand: Massey University Press, de Muizon, C, and Daryl P. Domning 1985. The First Records of Fossil Sirenians in the Southeastern Pacific Ocean. Bulletin du Museum National d'Histoire Naturelle. Paris, series 4, section C, 3:189-213. de Sylva, Donald P. 1957. Studies on the Age and Growth of the Atlantic Sailfish, Istiophorus americanus (Cuvier), Using Length-Frequency Curves. Bulletin of Marine Science of the Gulf and Caribbean, 7:1-20. Fierstine, Harry L. 1974. The Paleontology of Billfish—The State of the Art. In R.S. Sho- mura and F. Williams, editors, Proceedings of the International Bill- fish Symposium, Kailua-Kona, Hawaii, 9-12 August 1972, Part 2: Review and Contributed Papers. National Oceanic and Atmospheric Administration Technical Report, National Marine Fisheries Ser- vice, Special Scientific Report Fisheries Series, 675:34-44. 1978. A New Marlin, Makaira panamensis. from the Late Miocene of Panama. Copeia, 1978( 1): 1 -11. 1990. A Paleontological Review of Three Billfish Families (Istiophoridae, Xiphiidae, and Xiphiorhynchidae). In Richard H. Stroud, editor, Planning the Future of Billfishes, Part 2: Contributed Papers. Pro- ceedings of the Second International Billfish Symposium, Kailua- Kona, Hawaii, 1-5 August 1988, pages 11-19. Savannah, Georgia: National Council for Marine Conservation, Inc. 1998. Makaira sp., cf. M. nigricans Lacepede, 1802 (Teleostei: Perci- formes: Istiophoridae) from the Eastover Formation, Late Miocene, Virginia, and a Reexamination oi Istiophorus calvertensis Berry, 1917. Journal of Vertebrate Paleontology, 18( 1) :30^»2. 1999a. A New Shortbilled Marlin of the Genus Makaira, from the York- town Formation (Early Pliocene), Eastern North Carolina at Lee Creek Mine, USA. Teritary Research, 19(344):71-77. 1999b. Makaira sp., cf. M. nigricans Lacepede, 1802 (Teleostei: Perci- formes: Istiophoridae) from the Late Miocene, Panama, and Its Probable Use of the Panama Seaway. Journal of Vertebrate Paleon- tology. 19(3):430^t37. Fierstine, Harry L., and Shelton P. Applegate 1968. Billfish Remains from Southern California with Remarks on the Im- portance of the Predentary Bone. Bulletin of the Southern California Academy of Sciences. 67(1 ):29-39. Fierstine, Harry L., and Oliver Crimmen 1996. Two Erroneous, Commonly Cited Examples of "Swordfish" Pierc- ing Wooden Ships. Copeia, I996(2):472^i75. Fierstine, Harry L., and Nicole Voigt 1996. Use of Rostral Characters for Identifying Adult Billfishes (Teleo- stei: Perciformes: Istiophoridae and Xiphiidae). Copeia. 1996(1): 148-161. Fierstine, Harry L., and Bruce J. Welton 1983. A Black Marlin, Makaira indica. from the Early Pleistocene of the Philippines and the Zoogeography of Istiophorid Billfishes. Bulletin of Marine Science, 33(3):718-728. 1988. A Late Miocene Marlin, Makaira sp. (Perciformes, Osteichthyes) from San Diego County, California, U.S.A. Tertiary Research, 10(1): 13-19. Frazier, John G., Harry L. Fierstine, Sallie C. Beavers, Federico Achaval, Hiroyuki Suganuma, Robert L. Pitman, Yuichiro Yamaguchi, and Carlos Ma. Prigioni 1995. Impalement of Marine Turtles (Reptilia, Chelonia: Cheloniidae and Dermochelyidae) by Billfishes (Osteichthyes, Perciformes: Istio- phoridae and Xiphiidae). Environmental Biology of Fishes, 39:85-96. Gibson, Thomas G. 1983. Stratigraphy of Miocene through Lower Pleistocene Strata of the United States Central Atlantic Coastal Plain. In Clayton E. Ray, edi- tor. Geology and Paleontology of the Lee Creek Mine, North Caro- lina, I. Smithsonian Contributions to Paleobiology, 53:35-80, 37 figures. Gillette, David D. 1984. A Marine Ichthyofauna from the Miocene of Panama, and the Ter- tiary Caribbean Faunal Province. Journal of Vertebrate Paleontol- ogy, 4(2):172-186. Gottfried, Michael D. 1982. A Pliocene Sailfish Istiophorus platypterus (Shaw and Nodder, 1791) from Southern California. Journal of Vertebrate Paleontol- ogy, 2(2):151-153. NUMBER 90 43 GraphPad Software 1993. GraphPad InstafB. Macintosh Version 2.01. San Diego, California. Graves, John E., and Jan R. McDowell 1995. Inter-ocean Genetic Divergence of Istiophorid Billfishes. Marine Biology, 122:193-203. Gregory, William K. 1933. Fish Skulls, a Study of the Evolution of Natural Mechanisms. Trans- actions of the American Philosophical Society, 23(2):i-vii+75^I81. Gregory, William K., and G. Miles Conrad 1937. The Comparative Osteology of the Swordfish (Xiphias) and the Sailfish (Istiophorus). American Museum Novitates, 952:1-25. Gudger, E.W. 1940. The Alleged Pugnacity of the Swordfish and the Spearfishes as Shown by Their Attacks on Vessels. Memoirs of the Royal Asiatic- Society of Bengal. 12(2):215-315. Hazel, Joseph E. 1983. Age and Correlation of the Yorktown (Pliocene) and Croatian (Pliocene and Pleistocene) Formations at the Lee Creek Mine. In Clayton E. Ray, editor. Geology and Paleontology of the Lee Creek Mine, North Carolina, 1. Smithsonian Contributions to Paleobiol- ogy. 53:81-200, 4 figures, 38 plates. Hill, Kevin T., Gregor M. Cailliet, and Richard L. Radtke 1989. A Comparative Analysis of Growth Zones in Four Calcified Struc- tures of Pacific Blue Marlin, Makaira nigricans. Fishery Bulletin, United Slates, 87:829-843. Hopper, Carol N. 1986. The Majestic Marlin Solves Mysteries of the Sea. In World Record Game Fishes 1986, pages 58-61, 63, 65-66, 68-70. Fort Lauder- dale, Florida: International Game Fish Association. 1990. Patterns of Pacific Blue Marlin Reproduction in Hawaiian Waters. In Richard H. Stroud, editor, Planning the Future of Billfishes, Part 2: Contributed Papers. Proceedings of the Second International Bill- fish Symposium, Kailua-Kona, Hawaii, 1—5 August 1988, pages 29-39. Savannah, Georgia: National Council for Marine Conserva- tion, Inc. Hussakof, L. 1908. Catalogue of Types and Figured Specimens of Fossil Vertebrates in the American Museum of Natural History, Part I: Fishes. Bulletin of the American Museum of Natural History, 25:1-103. Johnson, G. David, and Colin Patterson 1993. Percomorph Phylogeny: A Survey of Acanthomorphs and a New Proposal. In G. David Johnson and William D. Anderson, Jr., edi- tors, Proceedings of the Symposium on Phylogeny of Percomorpha, June 15-17, Held in Charleston, South Carolina at the 70th Annual Meeting of the American Society of Ichthyologists and Herpetolo- gists. Bulletin of Marine Science, 52:554-626. Jolley, John W., Jr. 1974. On the Biology of Florida East Coast Atlantic Sailfish (Istiophorus platypterus). In R.S. Shomura and F. Williams, editors, Proceedings of the International Billfish Symposium, Kailua-Kona, Hawaii, 9-12 August 1972, Part 2: Review and Contributed Papers. National Oceanic and Atmospheric Administration Technical Report, Na- tional Marine Fisheries Service Special Scientific Report Fisheries Series, 675:81-88. Jollie, Malcolm 1986. A Primer of Bone Names for the Understanding of the Actinoptery- gian Head and Pectoral Girdle Skeletons. Canadian Journal of Zo- ology, 64:365-379. Lacepede, Bernard G.E. 1801. Histoire naturelle des poissons. Volume 3, 558 pages, 34 plates. 1802. Histoire naturelle des poissons. Volume 4, 728 pages, 16 plates. Lawley, R. 1876. Nuovi Studi sopra ai Pesci ed altri Vertebrati fossili delle Colline Toscane. 122 pages. Firenze. Leidy, Joseph 1860. Description of Fossil Vertebrates. In Francis S. Holmes, Post- Pleiocene Fossils of South Carolina, pages 99-122. Charleston, South Carolina: Russel and Jones. Leriche, Maurice 1926. Les poissons neogenes de la Belgique. Memoires du Muxee Royal d'Histoire Naturelle de Belgique, 32:365^172. Leviton, Alan E., Robert H. Gibbs, Jr., Elizabeth Heal, and C.E. Dawson 1985. Standards in Herpetology and Ichthyology, Part I: Standard Sym- bolic Codes for Institutional Resource Collections in Herpetology and Ichthyology. Copeia, 1985(3):802-832. Morrow, James E. 1951. A Striped Marlin (Makaira mitsukurii) Without a Spear. Copeia, 1951(4):303-304. Nakamura, Izumi 1983. Systematics of Billfishes (Xiphiidae and Istiophoridae). Publica- tions of the Seto Marine Biological Laboratory, 28:255-396. 1985. An Annotated and Illustrated Catalogue of Marlins, Sailfishes, Spearfishes and Swordfishes Known to Date. Food and Agricul- ture Organization of the United Nations (FAO) Fisheries Synopsis, 125(5):i-iv+l-65. Nakamura, Izumi, T. Iwai, and K. Matsubara 1968. A Review of the Sailfish, Spearfish, Marlin, and Swordfish of the World. Misaki Marine Biological Institute, Kyoto University Special Report, 4:1-95. National Marine Fisheries Service 1994. Cooperative Game Fish Tagging Program Annual Newsletter, 1992. National Oceanic and Atmospheric Administration Technical Mem- orandum, National Marine Fisheries Service—Southeast Fisheries Science Center, 346:1-23. Nelson, Joseph S. 1994. Fishes of the World. Third edition, 600 pages. New York: John Wiley and Sons, Inc. Pepperell, Julian G. 1990. Movements and Variations in Early Year-Class Strength of Black Marlin (Makaira indica) off Eastern Australia. In Richard H. Stroud, editor, Planning the Future of Billfishes, Part 2: Contributed Papers. Proceedings of the Second International Billfish Sympo- sium, Kailua-Kona, Hawaii. 1-5 August 1988, pages 51-66. Savan- nah, Georgia: National Council for Marine Conservation, Inc. Poey, Felipe 1860. Memorias sobre la historia natural de la Isla de Cuba, acompaha- das de sumarios latinos y extractos in frances etc., volume 2, pages 97-336. LaHabana. Rafinesque, Constantine S. 1810. Caratteri di alcuni nuovi generi e nuove specie de animali e piante delta Sicilia, con varie osservazioni sopri i medesimi. 105 pages. Palermo. Robins, C. Richard, Reeve M. Bailey, Carl. E. Bond, James R. Brooker, Ernest A. Lachner, Robert N. Lea, and W.B. Scott 1991. Common and Scientific Names of Fishes from the United States and Canada. American Fisheries Society Special Publication, 20:1-183. Robins, C. Richard, and Donald P. de Sylva 1960. Description and Relationships of the Longbill Spearfish, Tetraptu- rus belone. Based on Western North Atlantic Specimens. Bulletin of Marine Science, 10:383^113. 1963. A New Western Atlantic Spearfish, Tetrapturus pfluegeri. with a Redescription of the Mediterranean Spearfish, Tetrapturus belone. Bulletin of Marine Science, 13:84-122. Rojo, Alfonso L. 1991. Dictionary of Evolutionary Fish Osteology. 273 pages. Boca Raton: CRC Press. Schultz, Ortwin 1987. Taxonomische Neugruppierung der Uberfamilie Xiphioidea (Pisces, Osteichthyes). Annalen Naturhistorisches Museum Wien, series A 44 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY (Mineralogie und Petrographie, Geologie und Palaontologie, An- thropologie und Prahistories), 89:95-202. Shaw, George, and Frederick P. Nodder 1792. Xiphias platypterus: The Broadfinned Swordfish. The Naturalist's Miscellany, 28: pages unnumbered, plate 88. Squire, James L., and Ziro Suzuki 1990. Migration Trends of Striped Marlin (Tetrapturus audax) in the Pa- cific Ocean. In Richard H. Stroud, editor, Planning the Future of Billfishes, Part 2: Contributed Papers. Proceedings of the Second In- ternational Billfish Symposium, Kailua-Kona. Hawaii. 1-5 August 1988, pages 67-80. Savannah, Georgia: National Council for Ma- rine Conservation, Inc. Van Beneden, P.J. 1871. Recherches sur quelques poissons fossiles de Belgique. Bulletin de I'Academie Royale des Sciences, des Lettres et des Beaux-Arts de Belgique, series 3,31 (2):493-518. van der Elst, Rudy P. 1990. Aspects of the Biology and Sport Fishery for Billfishes in the SW. Indian Ocean. In Richard H. Stroud, editor, Planning the Future of Billfishes, Part 2: Contributed Papers. Proceedings of the Second In- ternational Billfish Symposium. Kailua-Kona, Hawaii. 1-5 August 1988. pages 147-158. Savannah, Georgia: National Council for Ma- rine Conservation, Inc. Wapenaar, Mary-Louise, and Frank H. Talbot 1964. Note on the Rigidity of the Pectoral Fin of Makaira indica (Cuvier). Annals of the South African Museum, Cape Town. 48(6): 167-180. Ward, Lauck W., and Blake W. Blackwelder 1980. Stratigraphic Revision of Upper Miocene and Lower Pliocene Beds of the Chesapeake Group—Middle Atlantic Coastal Plain. Bulletin of the United Stales Geological Survey, 1482-D:1-61. Whitmore, Frank C, Jr., and Robert H. Stewart 1965. Miocene Mammals and Central American Seaways. Science, 148:180-185. Woodring, W.P. 1957. Geology and Paleontology of Canal Zone and Adjoining Parts of Panama; Geology and Description of Tertiary Mollusks (Gastro- pods: Trochidae to Turritellidae). United States Geological Survey Professional Paper, 306-A: 1-145. 1970. Geology and Paleontology of Canal Zone and Adjoining Parts of Panama; Geology and Description of Tertiary Mollusks (Gastro- pods: Eulimidae, Marginellidae to Helminthoglyptidae). United States Geological Survey Professional Paper, 306-D:322^452. 1973. Geology and Paleontology of Canal Zone and Adjoining Parts of Panama; Geology and Description of Tertiary Mollusks (Additions to Gastropods, Scaphopods, Pelecypods: Nuculidae to Malleidae). United States Geological Survey Professional Paper, 306-E: 453-539. 1982. Geology and Paleontology of Canal Zone and Adjoining Parts of Panama; Description of Tertiary Mollusks (Pelecypods: Propea- mussiidae to Cuspidariidae; Additions to Families Covered in P306- E, Additions to Gastropods; Cephalopods). United States Geologi- cal Survey Professional Paper. 306-F:541-845. Woodward, Arthur S. 1901. Catalogue of the Fossil Fishes in the British Museum (Natural His- tory). Volume 4, 636 pages, 19 plates. London: British Museum (Natural History). Plates 46 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Plate 1 Istiophoridae, genus and species indeterminate Yorktown Formation, Lee Creek Mine a.b. NCSM 5576, right articular: a. lateral view; b, dorsal view of joint with quadrate. c. USNM 488084, left dentary, medial view of interdentary joint. d. USNM 290202, first pectoral-fin ray, view of articular facet for scapula. e. NCSM 5297, first pectoral-fin ray, view of articular facet for scapula. f,g. USNM 475425, left maxilla:/ lateral view; g, dorsal view. h,i. USNM 284815, parasphenoid: h, ventral view; /, left lateral view. j,k. USNM 481956, predentary: j, right lateral view; k, dorsal view. Each scale bar=2 cm NUMBER 90 48 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Plate 2 Istiophoridae, genus and species indeterminate Yorktown Formation, Lee Creek Mine a,b. USNM 481915, left quadrate: a. lateral view; b, view of articular condyle for articular. c-e. USNM 481939, rostrum: c, left lateral view; d. dorsal view; e, cross section. f-h. USNM 475409, rostrum:/ left lateral view; g, dorsal view; h, cross section. i.j. USNM 481984, rostrum: /, left lateral view;/ dorsal view. Each scale bar=2 cm NUMBER 90 50 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Plate 3 Makaira purdyi Fierstine, 1999a Yorktown Formation, Lee Creek Mine o-c USNM 481933 (holotype), rostrum: a. left lateral view; b, dorsal view; c, ventral view. Istiophoridae, genus and species indeterminate Yorktown Formation, Lee Creek Mine d. NCSM 7902, right scapula, view of articular surface for first pectoral-fin ray. e.f. NCSM 4914, vertebra 1: e. left lateral view;/ anterior view. g.h. USNM 481982, hypural: g, left lateral view; h, anterior view. Each scale bar=2 cm NUMBER 90 51 52 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Plate 4 Istiophorus platypterus (Shaw and Nodder, 1792) Yorktown Formation, Lee Creek Mine a,b. USNM 290198, left maxilla: a, dorsal view; b, lateral view. c—e. USNM 481973, rostrum: c, left lateral view; d, dorsal view; e, cross section. f-h. USNM 286949, rostrum:/ dorsal view; g, left lateral view; h, cross section. i-k. USNM 481967, rostrum: ;', left lateral view;/ cross section; k, dorsal view. Each scale bar =2 cm NUMBER 90 54 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Plate 5 Istiophorus cf. /. platypterus (Shaw and Nodder, 1792) Yorktown Formation, Lee Creek Mine a-c. USNM 286950, rostrum: a, left lateral view; b, dorsal view; c, cross section. d,e. USNM 488093, hypural: d, left lateral view; e, anterior view. Makaira indica (Cuvier, 1832) Yorktown Formation, Lee Creek Mine / USNM 481927, right scapula, view of articular surface for first pectoral-fin ray. g. USNM 488112, right scapula, view of articular surface for first pectoral-fin ray. h.i. USNM 475399, predentary: h. right lateral view;;', dorsal view. j.k. USNM 488009, predentary:/ right lateral view; k, dorsal view. Makaira ci. M. indica (Cuvier, 1832) Yorktown Formation, Lee Creek Mine /. USNM 488100, left scapula, view of articular surface for first pectoral-fin ray. Each scale bar=2 cm NUMBER 90 56 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Plate 6 Makaira nigricans Lacepede, 1802 Yorktown Formation, Lee Creek Mine a. NCSM 2125, left dentary, medial view of interdentary joint. b,d. NCSM 5159, parasphenoid: b, left lateral view; d, ventral view. c.e. USNM 291066, predentary: c, dorsal view; e, left lateral view. f.h.j. USNM 481941, rostrum:/ left lateral view; h, dorsal view;/ cross section. g.i. USNM 481936, predentary: g. dorsal view; I, left lateral view. k-m. USNM 481943, rostrum: k. cross section; /, left lateral view; m, dorsal view. Each scale bar =2 cm NUMBER 90 58 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Plate 7 Makaira nigricans Lacepede, 1802 Yorktown Formation, Lee Creek Mine a-c. USNM 481976, rostrum: a, left lateral view; b, dorsal view; c, view of nutrient canals exposed at distal tip. d. USNM 421527, right scapula, view of articular surface for first pectoral-fin ray. e. USNM 290197, right scapula, view of articular surface for first pectoral-fin ray. fih. USNM 286179, vertebra 23:/ left lateral view; h. anterior view. g.l. USNM 286181, vertebra 22: g. anterior view; /, left lateral view. i,k. USNM 481923, vertebra 1: ;', anterior view; k. left lateral view. j.m. NCSM 4938, hypural:/ anterior view; m, left lateral view. Each scale bar =2 cm NUMBER 90 60 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Plate 8 Makaira cf. M. nigricans Lacepede, 1802 Yorktown Formation, Lee Creek Mine aj. NCSM 6944, left quadrate: a, lateral view;/ view of articular condyle for the articular. b.c. NCSM 7427, rostrum: b, left lateral view; c, dorsal view. d,f. USNM 488116, parasphenoid: d, right lateral view;/ ventral view. e.g. USNM 481944, rostrum: e, left lateral view;g, dorsal view. h,i. USNM 481897, vertebra 1: h, left lateral view; /, anterior view. Each scale bar =2 cm NUMBER 90 62 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Plate 9 Makaira cf. M. nigricans Lacepede, 1802 Yorktown Formation, Lee Creek Mine a.b. USNM 286177, vertebra 22: a, left lateral view; b, anterior view. c,f USNM 481909, vertebra 23: c, left lateral view;/ anterior view. d,e. USNM 481979, hypural: d. right lateral view; e, anterior view. Makaira sp. Yorktown Formation, Lee Creek Mine g-i. USNM 475390, rostrum: g, dorsal view; h, cross section; f, left lateral view. //. USNM 290542, vertebra 23:/ left lateral view; /, anterior view. k. USNM 286997, right dentary, medial view of interdentary joint. Each scale bar=2 cm NUMBER 90 64 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Plate 10 Makaira sp. Yorktown Formation, Lee Creek Mine a,b. USNM 481931, predentary: a, left lateral view; b, dorsal view. c,d. USNM 488006, left quadrate: c, view of articular condyle for articular; d, lateral view. g. USNM 290204, left scapula, view of articular surface for first pectoral-fin ray. h.i. USNM 481981, hypural: h, right lateral view; /, anterior view. cf. Makaira indica sp. Yorktown Formation, Lee Creek Mine e.f USNM 488007, hypural: e, left lateral view;/ anterior view. / NCSM 2990, right dentary, medial view of interdentary joint. k.l. USNM 481916, left quadrate: k, view of articular condyle for the articular; /. lateral view. Each scale bar=2 cm NUMBER 90 66 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Plate 11 Tetrapturus albidus Poey, 1860 Yorktown Formation, Lee Creek Mine a,b. USNM 290193, left articular: a. lateral view; b, dorsal view of joint with quadrate. c,d. USNM 475393, right maxilla: c, lateral view; d, dorsal view. e.f. USNM 488046, parasphenoid: e. left lateral view;/ ventral view. g.h. USNM 290203, right maxilla: g, lateral view; h, dorsal view. i.j. USNM 488027, parasphenoid: /. left lateral view;/ ventral view. Each scale bar=2 cn NUMBER 90 68 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Plate 12 Tetrapturus albidus Poey, 1860 Yorktown Formation, Lee Creek Mine a.b. USNM 481908, right quadrate: a, lateral view; b, view of articular condyle for the articular. c.d. USNM 488008, left quadrate: c, lateral view; d, view of articular condyle for the articular. Tetrapturus cf. T. albidus Poey, 1860 Yorktown Formation, Lee Creek Mine e.f. USNM 488085, left maxilla: e, lateral view;/ dorsal view. Each scale bar=2 cm NUMBER 90 69 The Neogene Sharks, Rays, and Bony Fishes from Lee Creek Mine, Aurora, North Carolina Robert W. Purdy, Vincent P. Schneider, Shelton P. Applegate, JackH. McLellan, Robert L. Meyer, and Bob H. Slaughter ABSTRACT The fish remains, including 104 species from 52 families, col- lected at the Lee Creek Mine near Aurora, Beaufort County, North Carolina, constitute the largest fossil marine fish assemblages known from the Coastal Plain of the eastern United States. The fish faunas came principally from the Pungo River Formation (Burdigalian, planktonic foraminifera zones N6-7) and the York- town Formation (Zanclian, planktonic foraminifera zone N18 and younger). A few specimens were obtained from the James City Formation (early-middle Pleistocene). As an assemblage, the fishes found in the Pungo River Forma- tion, including 44 species of selachians and 10 species of teleosts, are most similar to those from the "Muschelsandstein" of the Swiss Molasse. The Yorktown Formation fish assemblage includes 37 species of selachians and 40 species of teleosts, derived mostly from the base of the Sunken Meadow Member. Although the Pungo River Formation fish fauna is dominated by warm-water (18°-25°C) taxa, the Yorktown Formation fossil fish fauna includes warm and cool water species. Both fish assem- blages occur with a cool-temperate invertebrate fauna. The abundant remains in both faunas permit us to make the fol- lowing interpretations concerning shark taxonomy. We reassign Megascyliorhinus to the family Parascyllidae and Parotodus bene- denii (Le Hon) to the Lamnidae. Among the mako sharks, we des- ignate the lectotype of Isurus desori (Agassiz) and synonymize it with /. oxyrinchus Rafinesque and separate Isurus xiphodon (Agassiz) from /. hastalis (Agassiz). Palaeocarcharodon, Procar- Robert W. Purdy, Department of Paleobiology, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560- 0121. Vincent P. Schneider, Paleontology Department, North Caro- lina Museum of Natural History, P. O. Box 27647, Raleigh, North Carolina 27611. Shelton P. Applegate, Instituto de Geologia, Apar- tado Postal 70-296, Cuidad Universitaria, 20 D.F., Mexico. Jack H. McLellan, 112 Shoveler Court, Georgetown, Kentucky 40324. Robert L. Meyer, Union Oil Company of California, 500 Executive Plaza East, 4615 Southwest Freeway, Houston, Texas 77027. Bob H. Slaughter (deceased). charodon, Megaselachus, and Carcharocles art synonymized with Carcharodon. Sphyrna laevissima (Cope) is synonymized with S. zygaena (Linnaeus), and Galeocerdo triqueter Cope is synony- mized with Alopias cf. A. vulpinus (Bonnaterre). This fauna produced four new records and two new species. Among the selachians, we note the first records of Megascyliorhi- nus, Rhincodon, Megachasma, and Isistius from the Atlantic Coastal Plain, and among the bony fishes, the first occurrences in the fossil record of Caulolatilus and Pomatomus. We also describe two new species of bony fishes, Lopholatilus rayus and Pagrus hyneus. Introduction The fossil faunas of both bony fishes and elasmobranchs at Lee Creek Mine are among the largest in the world, and the Pliocene remains represent one of the most abundant and di- verse fossil vertebrate faunas yet recorded in the scientific liter- ature (see Table 1). This fossil fish fauna consists of tens of thousands of selachian teeth (including six associated denti- tions) and bony fish remains (including teeth, cranial frag- ments, an articulated skull, several associated skeletons, verte- brae, fin spines, otoliths). This fauna extends the record of many extant fishes into the Neogene of the United States. We identify more than 35 species not previously recognized from this province (excluding the otolith record; see Fitch and Lavenberg, 1983) and provide the first thorough account of this faunal diversity and its paleoecological implications. PREVIOUS Work.—The literature concerning fossil fishes of the Atlantic Coast of North America is rather meager. Leriche (1942) provided one of the most comprehensive accounts of the fossil fishes of the area with a list of 50 Miocene and Pliocene species, and he also reviewed work previous to his own. The most useful of these works are illustrated papers by Gibbes (1848-1849) and Leidy (1877), concerning the phos- phate beds of South Carolina, and Eastman's (1904) summary of the fishes of the Chesapeake Group in Maryland. 71 72 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Fowler's (1911) lengthy work on the fossil fishes of New Jersey is useful as a catalog, but the specimen locality data are so poor that the informational value of the paper is seriously marred. Case's (1980) paper on the lower Miocene sharks of the Belgrade Formation is the only recent work about the Neo- gene fishes of the Atlantic Coast of North America. North Carolina fossil fishes are especially poorly known. Emmons (1858) included some species now known to occur prolifically at the Lee Creek locality. Cope (1869, 1871, 1875) made additions to this fauna, and Leriche's (1942) Duplin Marl occurrences are primarily from North Carolina. Project History.—A first manuscript describing the fossil fish fauna of Lee Creek Mine was completed in the mid-1970s. It reflected largely the work of Robert L. Meyer (sharks and rays) and Bob H. Slaughter (bony fishes). In the early 1980s Jack H. McLellan revised the manuscript and added to it addi- tional taxa that he and Robert W. Purdy had identified subse- quent to the first manuscript. Due to other commitments, Mey- er, Slaughter, and McLellan relinquished responsibility for final revisions to Purdy. Purdy, who revised the Chondrich- thyes and the geological portions of the manuscript, sought the help of Vincent P. Schneider to do the fossil bony fishes. Shelton P. Applegate reconstructed composite dentitions of the common Lee Creek shark taxa; he and Purdy subsequently refined these to reflect, as accurately as possible, the dentitions of these sharks. They are illustrated herein. We must state here that Applegate disagrees with Purdy's as- signment of Parotodus to the Lamnidae, Purdy's identification of the first upper anterior tooth of Parotodus and Carcharodon as the second anterior tooth, and Purdy's recognition of Galeo- cerdo contortus as a species separate from G aduncus Agassiz (= G. sp. herein). Since the completion of the McLellan manuscript in the early 1980s, much new information has been published about the pa- leoecology and paleooceanography of the Atlantic Coastal Plain, the ecology of extant fishes, and the taxonomy of fossil and extant fishes. Many additional specimens were added to the collections, including two associated dentitions of fossil sharks. Also, Gordon Hubbell made available to us his impor- tant collection of extant sharks' dentitions. This manuscript, therefore, represents a major revision of the earlier ones. ACKNOWLEDGMENTS We express special thanks to Peter J. Harmatuk and to Becky and Frank Hyne, who collected and generously donated several thousand specimens to the National Museum of Natural Histo- ry (NMNH), Smithsonian Institution, including many of the specimens cited below. We extend special thanks as well to Wayne and Aura Baker, who also donated many specimens to the Smithsonian. The Hynes and the Bakers also were instru- mental in encouraging collectors at the mine to donate impor- tant specimens to the Smithsonian. We also express gratitude to Kate and Gordon Hubbell, who graciously provided Bob Purdy with room and board during his several visits to their home to study shark dentitions in the Hubbell collection. R.W.P. extends sincere thanks to Jann Thompson and Clay- ton E. Ray for their encouragement and support of this research project. For the presentation of specimens to the Smithsonian Institu- tion we thank Wayne and Aura Baker, William Bean, Jim Bourdon, Tom and Pat Burns, Rann Carpenter, Gerard R. Case, Ralph Chamness, N.E. Cowell, Raymond Douglas, George C. Fonger, Harvey A. Franz, Jr., Tex Gilmore, David W. Grabda, Chris Harmatuk, Jack M. Hird, Jerry Hughes, Jeremy Jacobs, Geoffry Keel, Royal H. and Gene Mapes, the late Earl M. Ma- son, Vance McCollum, Gregory McKee, Arnold Powell, George Powell, George Prinsen, Norm Riker, Judy Schneider, Clyde Swindell, the late Charles Wells, the late Edgar A. Womble, and many members of the Calvert Marine Museum Fossil Club, the Myrtle Beach Fossil Club, the National Muse- um of Natural History field parties, and the North Carolina Fossil Club. Examination of collections or specimens of modern and fos- sil fishes was made possible with the assistance of Shelton P. Applegate and Camm Swift (Los Angeles County Museum of Natural History), Leonard Compagno (then at Stanford Uni- versity), M.N. Feinberg, (Department of Ichthyology, Ameri- can Museum of Natural History), W.I. Follett (Department of Ichthyology, California Academy of Science), Joseph Gregory (Department of Paleontology, University of California at Ber- keley), Gordon Hubbell (Key Biscayne, Florida), Alvaro Mon- es (Museo Nacional de Historia Natural, Montevideo, Uru- guay), Kazuhiro Nakaya (Hokkaido University), Dirk Nolf and Paul Sartenaer (Institut Royal des Sciences Naturelles de Belgique), Harold Wes Pratt, Jr. (National Marine Fisheries Service, National Oceanographic and Atmospheric Adminis- tration), Martin Sander (Palaontologisches Institut und Muse- um der Universitat Zurich), Earl Shapiro (Academy of Natural Sciences of Philadelphia), Scott Snyder (Department of Geol- ogy, East Carolina State University), Torn Taniuchi (Universi- ty of Tokyo), Giancarlo Tondi (Universita degli Studi di Roma "La Sapienza"), Stanley H. Weitzman and Victor Springer (Division of Fishes, NMNH), and Elizabeth Wing (Florida Museum of Natural History). Our research efforts were supported by the Smithsonian In- stitution through the Remington and Marguerite Kellogg Fund, the Research Opportunity Fund, and the Short-Term Visitors Fund. Leonard J.V. Compagno, David A. Ebert, Harry Fierstine, Bill Heim, Gordon Hubbell, James C. Tyler (NMNH), and David Ward aided us in the identification of several fossil fish- es. Leonard J.V. Compagno also reviewed portions of the manuscript; Michael Gottfried and Lance Grande reviewed the finished manuscript. We are grateful to the following individuals for translating into English papers in French, German, Italian, and Japanese: NUMBER 90 73 Steve Johnson, Marilyn Lebson, Albert Roland, Jeannette Saquet, and Hans Sues. National Science Foundation Grant GB 28598-1 supported the Southern Methodist University's field work at Lee Creek Mine. National Science Foundation Grant GB 36727 and a Smithsonian Institution research grant supported travel to mu- seums. Field work conducted by Smithsonian staff was sup- ported by the Smithsonian Institution through the Walcott Fund and the Remington and Marguerite Kellogg Fund. Methods The lack of adequate skeletal collections of extant western Atlantic fishes hampered our identification of the bony-fish re- mains. We identified the more tropical species of the fauna at the University of Florida, where Elizabeth Wing maintains an excellent synoptic skeletal collection of Florida fishes. One of us (V.P.S.), with the assistance of members of the North Caro- lina Division of Marine Fisheries, especially Fritz Rhode and Jim Franesconi, assembled a synoptic collection of the fishes presently living off the coast of North Carolina; this collection now resides in the North Carolina State Museum. Gordon Hub- bell's extensive private collection of extant sharks' dentitions assisted in our identification of the fossil teeth. We also exam- ined specimens from the private collections of Chris Harmatuk (Bridgeton, North Carolina) and Leonard Compagno (Cape Town, South Africa). These collections and the osteological specimens available in the Division of Fishes, NMNH, were used as the comparative basis for identifying the Lee Creek Mine material. To avoid the erection and perpetuation of unnecessary taxa, Leriche (1905, 1910, 1936b) advocated using reconstructed tooth sets for the study of fossil shark species. For lamnoid sharks, Applegate (1965, in prep.) supplemented Leriche's work by refining Leriche's tooth terminology and by identify- ing the morphological characters of the anterior teeth. These characters identify them to jaw position even when they occur as isolated teeth. For taxa with dentitions that are not easily dif- ferentiated into anteriors and laterals, Applegate recommended that the dentitions of closely related extant taxa be used as models for the reconstructions. Leriche's and Applegate's methods were employed in our study of the Lee Creek Mine sharks. The abundance and diversity of elasmobranch teeth in the Miocene and Pliocene sediments at Lee Creek Mine allowed us to reconstruct the dentitions of the common shark taxa. By comparing these reconstructions with the dentitions of extant sharks, we believe we have unraveled ambiguities that have plagued the study of fossil sharks. The unrecognized high degree of variation in dental mor- phology in extant sharks fostered this ambiguity. Compagno (1988:26) noted that "the general scarcity of comparative data for teeth of living sharks has led to many errors in the paleonto- logical literature. Unawareness of patterns of positional, devel- opmental, and sexual heterodonty have led both paleontologists and neontologists astray." In our attempts to establish the iden- tities of the Lee Creek Mine sharks, we examined as many den- titions of extant sharks as possible, both from different age groups and sexes and from different parts of the world. For most taxa, only a few specimens were available, and some of these were not sized or sexed and were from the same popula- tion. At this time, cladistic analyses of the taxa are not possible because of the fragmentary nature of the material and the un- availability of comparative osteological data for phylogenetic analysis. Where possible we have identified potential synapo- morphies, but we qualify them as tentative for the reasons giv- en above. The taxonomic classifications and nomenclature used in this paper follow Compagno (1977, 1984) and Carroll (1988). De- scriptions of taxa are from these sources unless noted other- wise. Dates for taxa published by Agassiz (1833-1843) were taken from Jeannet (1928, 1929), who published a schedule of publication dates for Agassis's work. Synonyms are provided only for fossil taxa whose identifica- tions are being changed. Due to the great number of fossil fish specimens recovered at Lee Creek Mine, only figured or measured specimens were cat- aloged. Thus, although the total number of specimens referred to each taxon is cited, with few exceptions, only the cataloged specimens are listed. Figured specimens are from Lee Creek Mine unless otherwise noted. All measurements were made in millimeters or centimeters. For shark teeth, tooth height was measured from the apex of the crown to a line tangential to the basal margin of the root, tooth width was measured at the greatest lateral extent of the tooth, and tooth thickness was measured in the area of the cen- tral foramen of the root. Total length is abbreviated as TL. Because the specimens were not collected in place, we made special efforts to collect from spoil areas where there was the least chance of mixing taxa from different horizons. We sifted large quantities of the Pungo River Formation, particularly the ore zone (unit 3) and units 4 and 5. Our principal source for ore-zone material was the coarse tailings found at the mill, and even in this material we occasionally found waifs from the Yorktown or James City formations. Despite these problems of mixing, we believe our collection techniques were careful enough to allow us to decide the stratigraphic occurrence of most of the taxa in this study. A field party from the Shuler Museum of Paleontology of Southern Methodist University undertook the initial bulk pro- cessing of sediments for fossils, especially microvertebrate re- mains. Subsequent bulk collecting was conducted by McLel- lan, Applegate, Schneider, and Purdy. We also collected on the spoil piles from areas that did not appear to be contaminated by mixed sediments and fossil invertebrates. Because we could not collect on the working face of the mine, we made special efforts to bulk sample uncontaminated piles of units 1 to 3 of 74 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY the Yorktown Formation. Some specimens were found in lumps of matrix or had matrix in cavities or grooves in the specimen; this associated material was submitted to Thomas G. Gibson (United States Geological Survey) for analysis of any foraminifera present. Gibson's reports helped us to stratigraph- ically place many of these specimens. The Lee Creek Mine collection has accumulated since the opening of the mine test pit in 1963. Field parties under the di- rection of Clayton E. Ray (NMNH) began actively prospecting the locality in August 1971. Bob Slaughter and a party from Southern Methodist University, Dallas, Texas, collected bulk samples for screen washing in July 1972. The cataloged specimens are housed in the collections of the North Carolina State Museum of Natural Sciences and the NMNH. ABBREVIATIONS.—Institutional abbreviations used in the text are as follows. AMNH American Museum of Natural History ANSP Academy of Natural Sciences of Philadelphia CAS California Academy of Science CMM Calvert Marine Museum ETHGI Palaontologisches Institut Universitat Zurich IG Brussels Museum of Natural History NCSM North Carolina State Museum of Natural Sciences TE-PLI Staatliches Museum fur Naturkunde Karlsruhe UNIG Universitat Neuchatel, Institut fur Geologie NMNH National Museum of Natural History collections (including col- lections of the former United States National Museum), Smithsonian Institution Stratigraphy The section exposed at Lee Creek Mine includes more than 30 m of fossiliferous marine sedimentary rocks assigned to the Pungo River and Yorktown formations. Gibson (1967, 1983) described these exposures in detail; unless otherwise noted, stratigraphic units herein are those of Gibson. The Pungo River Formation, a subsurface unit, is exposed in the lower part of the mine pit. The upper 3.7 m of this forma- tion consists of thin limestone layers interbedded with thin phosphate sand layers (units 4-7). Below this lies the ore bed, a phosphate sand layer about 12 m thick (units 1-3). Within the ore bed, a thin but persistent indurated sand layer containing diatomite occurs (unit 2). Below the main ore layer, about 3 m of low-grade indurated phosphatic sand occurs that is not mined. This layer rests unconformably on the Castle Hayne Formation of middle Eocene age. Gibson's units 1 to 3 corre- late with the lower Miocene Dunkirk beds of the Calvert For- mation of the Chesapeake Group to the north, which are Burdi- galian (Hoffman and Ward, 1989:55), and units 4 to 7 correlate with the Fairhaven Member of the Calvert Formation, which are Langhian (Hoffman and Ward, 1989:55). The Yorktown Formation, which at Lee Creek Mine uncon- formably overlies the Pungo River Formation, consists of blue clayey, fossiliferous sand. Ward and Blackwelder (1980) sub- divided the Yorktown Formation into four members; beginning with the oldest, they are the Sunken Meadow Member, the Rushmere Member, the Morgarts Beach Member, and the Moore House Member. Only the first three members occur at Lee Creek Mine. Hazel (1983) presented evidence that the Yorktown Formation at the Lee Creek locality is planktonic foraminifera zone N19 (early Pliocene). Riggs et al. (1982), Gibson (1983), Snyder et al. (1983), and others have confirmed this age assignment. At Lee Creek Mine, the Sunken Meadow Member is the source of most of the vertebrate fossils. Sources of the Fish Fossils Vertebrate fossils occur in both the Yorktown and the Pungo River formations. A few fish fossils were found in situ, in ex- posures along the pit walls. Most specimens, however, were obtained by searching piles of overburden in mined-out parts of the pit. Other sources of fossils were the piles of ore waiting to be slurried and pumped to the mill, the residue of coarse rubble left behind at the ore pumping sites, and the reject piles at the mill. The overburden consists of the Pleistocene sediments, York- town Formation, and the top 3.7 m or so of the Pungo River Formation. During mining the draglines cast the overburden aside into a previously mined cut, stacking this material in windrows or spoil piles. Rains wash finer sediments down the slopes of the spoil piles leaving the fossils behind. Fish teeth and bones and other fossils are common in this lag material. As the draglines cast the overburden aside, inevitable mixing of the overburden layers occurs. Because, however, the dra- gline buckets are very large and the sediments are cohesive, large masses of homogeneous material from identifiable hori- zons survive the drop from the bucket onto the spoil piles. Pungo River Formation.—Much of the Pungo River ma- terial studied herein was collected from the ore layers (units 1-3). We obtained this material at the active mining sites where the draglines stack thousands of tons of ore to await transport to the mill. By screening bulk samples of this unproc- essed ore through size 30 mesh screens, we collected many small bones and teeth. Smaller draglines move the ore to sumps; there it is mixed with water, screened to remove coarse particles, and pumped through pipelines to the mill. When all of the ore at a site has been pumped to the mill, large piles of particles too coarse for slurrying remain. These piles yielded many fossils. When the ore slurry reaches the washer section of the mill it is screened, and particles coarser than sand size are discarded. These piles of mill rejects, which we screened, also yield Pun- go River fossils; the pumping and screening, however, abrades the specimens. The upper 3.7 m of the Pungo River Formation (units 4-7) is cast upon the spoil piles along with the nonphosphatic York- town Formation. After weathering, these upper Pungo River sediments are easily recognized by their characteristic litholo- gies. Shark teeth and osteichthyan vertebrae occur embedded in NUMBER 90 75 the thin limestone layers, but the phosphatic sands that alter- nate with the limestones (unit 4) are a more prolific source of fossils. It was from these sources that we obtained the Pungo River fish assemblage. Although we did not collect them in place, we assign them with some confidence to the lower and middle parts of the formation (units 1-3) and to the upper part (units 4-7). Yorktown Formation.—The blue gray, silty sands con- taining some phosphate pellets, cetacean bone fragments, and Placopecten clintonius (Gibson, 1983), the invertebrate guide fossil for the basal part of the Yorktown Formation, identify sediments of the lowest 1.2 m of the Yorktown Formation. These sediments yielded many fish fossils. In the basal layer of the Yorktown Formation, fish fossils oc- cur in two states of preservation. They either are well preserved and unabraded or are blackened, broken, worn, and heavily etched. Intermediately worn specimens are uncommon. The dark, wom specimens are themselves divisible into two groups: (1) fossils that are common in the Pungo River Forma- tion and occur only in the basal layer of the Yorktown Forma- tion as redeposited specimens, and (2) fossil teeth known to oc- cur only in the basal layer of the Yorktown Formation. We believe they originated in post-Pungo River beds that were eroded completely before or during the Yorktown transgres- sion. The well-preserved component of the fossil fish assemblage accumulated after the initial high-energy transgression. We be- lieve that most of the identifiable bony-fish remains found on the spoil piles came from the bottom 4 m of the Yorktown For- mation (units 1-3). Sediments representing unit 3 occur only as small patches on the spoil piles, as the unit is only 0.6 m thick. These patches are readily identified by an abundance of fish otoliths, well-preserved small shark teeth, chalky crab chelae, and other characteristic fossils. Biostratigraphic Implications of the Fish Faunas Based on fish remains alone, the biostratigraphic position of this fish fauna is difficult to assess. Many of the Lee Creek Mine taxa cannot be distinguished from the extant species. Another problem is that many species of extant sharks segre- gate by size and/or by sex, and in some species the growth rates vary such that individuals of the same age in two different pop- ulations of the same species will be of measurably different siz- es. The effects of these distributions and size differences on the fossil record of sharks have not yet been assessed. Until they are, we believe that the usefulness of fossil shark teeth as guide fossils is questionable. The stratigraphic ranges of several large shark genera, such as Galeocerdo and Carcharodon, suggest the maximum ages of the deposits. Galeocerdo sp. (=G. aduncus of Agassiz, Cap- petta, Leriche, and others) and G. contortus, frequently used as guide fossils, have broad stratigraphic ranges: Rupelian to Serravalian for the former and Chattian to Tortonian for the latter (Cappetta, 1987:123). Based on specimens from a strati- graphically controlled collection he made from the Neogene of Belgium, Leriche (1926) extended the range of Galeocerdo aduncus into the Zanclian; these teeth, however, may be from juveniles of G. cf. G. cuvier. De Stefano (1909) questionably referred a tooth frm the Pliocene of Tuscany to G. aduncus; however, this is apparently referable to G. cf. G. cuvier and definitely is not G. aduncus. Galeocerdo contortus has a more restricted stratigraphic range than does G. sp., and along with the presence of Carcharodon subauriculatus, which ranges from the Chattian to the Burdigalian and possibly the Lang- hian, it suggests a maximum age of Chattian for the Pungo River fauna. In the Yorktown Formation the absence of Galeocerdo sp. and G. contortus and the presence of G. cf. G. cuvier corre- sponds with the occurrence of these sharks in Europe. The large Galeocerdo from the lower part of the Yorktown Forma- tion is quite modern in aspect. The only fossil teeth that to our knowledge are similar are those from the Ashley phosphate beds of South Carolina (now thought to be of Pliocene age, Sanders, pers. comm., 16 Jun 1990) and the Galeocerdo from the Orciano beds of Pliocene age in Italy (Lawley, 1876; De Stefano, 1909). Espinosa-Arrubarrena and Applegate (1981) also reported the occurrence of this shark as G. rosaliensis in the basal late Pliocene of Baja California, which may also be a synonym of G. capellini. The absence of Galeocerdo sp. in the basal Yorktown Formation and the presence of the more modern G. cf. G. cuvier supports an early Pliocene age for these beds. Fossil Shark Teeth From the beginnings of shark paleontology, most paleontolo- gists have identified the different morphotypes of shark teeth as belonging to different species, but within a species, tooth mor- phology varies considerably; the dentitions of one species may contain several different morphotypes. This practice for nam- ing fossil shark species has led to much taxonomic confusion. Agassiz (1833-1843) published the first extensive work on fossil shark teeth. Many of his species were based on incom- plete teeth or on teeth from different jaw positions of species he described earlier in his work; nevertheless, his classic work laid the foundation for the study of fossil shark taxonomy. Maurice Leriche, who studied fossil sharks for the first half of this century, revised many of Agassiz's species and those of other earlier workers. As the basis of his taxonomic studies, Leriche created artificial tooth sets for fossil species. He was the first fossil shark specialist to do this, and although many fossil shark specialists ignored Leriche's method of study, he made many significant contributions to the study of fossil sharks. Despite Leriche's work, the naming of new species seemed to be more important than scientific accuracy. As a result of this, common fossil shark taxa have appeared in the literature many times under many different names. Today, a morass of 76 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Table 1.—Faunal list of Lee Creek Mine fishes (l-6=stratigraphic unit, A=abundant, C=Common, U=uncommon, R=redeposited). Taxon Pungo River Formation Yorktown Formation 1 2 3 James City Formation Class CONDRICHTHYES HEXANCHIDAE Notorynchus cepedianus Hexanchus sp. ECH1NORHINIDAE Echinorhinus cf. E. blakei Squalidae Squalus sp. Isistius sp. Pristiophoridae Pristiophorus sp. RHINOBATIDAE Rhinobatos sp. Pristidae Pristis sp. P. cf. P. pectinatus Rajidae Raja sp. Dasyatidae Dasyatis say D. centroura D. cf. D. americana Myliobatidae Pteromylaeus sp. Aetobatus sp. Rhinopteridae Rhinoptera sp. Plinthicus stenodon Mobulidae Mobula sp. Mania sp. Squatinidae Squatina sp. Parascyllidae Megascyliorhinus miocaenicus Ginglymostomatidae Glnglymostoma sp. Rhincodontidae Rhincodon sp. Odontaspididae Carcharias taurus C. cuspidata C. sp. Odonlaspis ferox O. cf. O. aculissima Megachasmidae Megachasma sp. Alopiidae Alopias ci. A. superciliosus A. ci. A. vulpinus Cetorhinidae Cetorhinus sp. Lamnidae Parotodus benedenii Isurus oxyrinchus I. haslalis I. xiphodon Lamna sp. Carcharodon subauriculalus C. megalodon C. carcharias u u u c A c A c A C u u u C c c c c u u u u u u u c u c C c c c c c c c c c c A c U c A C u u u U7 U? U? R? U c A A A A A c c c c c c u u u u u u c c c c c u? c u u u u u C c A A A A U? C u u NUMBER 90 77 Table 1.—Continued. Taxon Pungo River Formation Yorktown Forn 1 2 lation 1 2 3 4 5 6 3 James City Formation SCYLIORHINIDAE Scyliorhinus sp. C C C C C R? Triakidae Galeorhinus cf. G. affinis U u U U U R? R? Hypogaleus sp. U Mustelus sp. u u u u U HEMIGALEIDAE Paragaleus sp. A A A A A R? Hemipristis serra A A A A A C C C U? CARCHARHIN1DAE Galeocerdo sp. A A A A A c G. contortus A A A A A C G. cf. G. cuvier A C U Carcharhinus brachyurus A A A A A C.falciformis u U U C. leucas U U U U u C C C C. macloti A A A A C R C. obscurus C C c C. perezi A A A C C. plumbeus U U U Rhizoprionodonl sp. U U U U U Negaprion brevirostris U u u U Triaenodon obesus U u U Sphyrnidae Sphyrna lewini S. cf. 5. media S. zygaena Class Osteichthyes Acipenseridae Acipenser ci.A. oxyrhynchus Lepisosteidae Lepisosteus osseus Elopidae Megalops cf. M. atlanticus CONGRIDAE Conger cf. C. oceanicus CLUPE1DAE Alosa cf. A. sapidissima ARIIDAE Bagre sp. Batrachoididae Opsanus lau LOPH1IDAE Lophius cf. L. americanus MERLUCCIDAE Merluccius bilinearis TR1GLIDAE Prionotus cf. P. evolans SERRANIDAE Epinephelus sp. Mycteroperca sp. BRANCHIOSTEG1DAE Caulolatilus cf. C. cyanops Lopholatilus rayus POMATOM1DAE Pomatomus saltatrix CARANG1DAE Seriola sp. Sparidae Archosargus cf. A. probatocephalus Lagodon cf. L. rhomboides Pagrus hyneus C u C u c u c u c u u u u u u u u u? c u c? A c c u u A c u c c? c A c c c u c? 78 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Table 1.—Continued. Taxon P.sp Stenotomus cf. S. chrysops SCIAENIDAE Sciaenops sp. 5. ocellatus Pogonias cf. P. cromis LABRIDAE Tautoga cf. T. on His URANOSCOP1DAE Astroscopus sp. SPHYRAEN1DAE Sphyraena cf. S. barracuda SCOMBRIDAE Sarda sp. aff. S. sarda Auxis sp. Thunnus sp. Acanthocybium solandri Xiphiidae Xiphias gladius Istiophoridae Istiophorus cf. /. platypterus Makaira cf. M. indica M. nigricans Tetrapturus cf. T albidus Hemirhabdorhynchus sp. BOTHIDAE Paralichthys sp. Monacanthidae Aluterus sp. TEtraodontidae Sphoeroides hyperostosus DlODONTlDAE Chilomyclerus schoepfi MOLIDAE Mola chelonopsis Indeterminate Emmon's fish tooth Pungo River Formation C C c C c U U u u C c u u Yorktown Formation 1 T~ 3 C c c A U li c A U A C U u A U H A A A U James City Formation C? c A u C A U u A U u A A A C? U? c? A A A scientific names exists in shark paleontology, and more new fossil taxa are described each year. Many of the type specimens for the species described in Eu- rope during the nineteenth century have yet to be rediscovered. Some were destroyed by the ravages of wars; others remained in private collections, many of which have been lost or de- stroyed. Even the type specimens that were not lost were rarely redescribed and refigured. These problems have further hin- dered fossil shark taxonomy. The lack of good museum collections of dentitions from ex- tant sharks and the paucity of scientific papers describing den- tal variation in extant sharks also have hampered the study of fossil sharks. In extant species, the extent of dental variation re- mains undocumented. Without knowing the range of dental variation, paleontologists cannot make sound judgements about the taxonomy of fossil sharks. In studying the Lee Creek Mine sharks, we attempted to make the most parsimonious interpretation possible of the fos- sil evidence. We compared the fossil teeth with those of as many related extant species as possible, looking for characters to separate or synonomize the fossil species. Where studies of extant species indicated that great variability in dental mor- phology existed but that variability was not defined, we identi- fied these taxa to genera only. In instances where no informa- tion exists about dental variability in the extant species, we identified the fossil teeth to the fossil species. For fossil teeth that we could not separate from those of the extant species, we give ecological information on the extant species and the prob- able total length of the fossil species. The latter we estimated from the measurements of the largest tooth in the upper jaw of extant sharks of known length. Tooth Terminology An extensive terminology, including many synonymous terms, has arisen for the description of shark teeth. Compagno (1988) called for a standardization of this terminology, and we NUMBER 90 79 cusplets distal apical mesial apical primary cusp mesial apex of crown basal lateral cusplet central foramen basal distal crown foot labial recurvature of tip labial \ lingual transverse groove cutting edge deep notch shoulder central foramen FIGURE 1.—Tooth terminology used in this paper. use the terms he suggested, which incorporate those of Apple- gate (1965a, 1967). See Compagno (1988:27-30) and Figure 1 for the definitions of these terms. In view of the importance of reconstructing tooth sets of fossil shark species, Applegate and Compagno's tooth terminology is reviewed below. Among the lamniform sharks, which include Alopias, Car- charias, Carcharodon, and Isurus, among others, several basic tooth types occur. Leriche (1905) and Applegate (1965a) iden- tified these tooth types as median (medial), symphysial, alter- nate, anterior, intermediate, lateral, and posterior (Figures 2, 5a; Compagno, 1988, fig. 3.4). These names indicate position in the shark's jaws. Compagno (1988:32-33) defined shark teeth that cannot be differentiated into anteriors and laterals as follows: When anteriors are not differentiated (as is often the case in the lower jaw) but posteriors are, the more mesial teeth are termed ANTEROLATERALS; when posteriors are not differentiated but anteriors are, the more distal teeth are LATEROPOSTERIORS; and when neither anteriors or posteriors are differen- tiated, the parasymphysial teeth are ANTEROPOSTERIORS. Small symmetrical and asymmetrical teeth occur in the sym- physial region of many species of sharks; Applegate recog- nized three types of teeth in this region: median (medial), sym- physial, and alternate (definitions from Compagno, 1988; occurrence of teeth in taxa from Applegate, 1965a). The lamni- form tooth types are described below. Upper and Lower Teeth.—The lower teeth are not as compressed as the upper teeth; their tips usually recurve toward the more convex side of the tooth or lingually. A straight-edged area parallel to the long axis of the flat or labial face of the crown and tangent to the base of the crown facilitates seeing this relationship (in upper teeth as well). (Exceptions to this are the upper anterior teeth. Except in Alopias, the tips of these teeth bend lingually, but this lingual bend is not as great as that of the lower anterior teeth). In the area of the central foramen, in the lower anterior teeth and the first two or three laterals, the root possesses a noticeable bulge or torus. In the lower lateral teeth, the angle formed by the root lobes is not as obtuse as that in the upper laterals. In the upper teeth, the tip of the crown in profile or lateral view is straight, or it may recurve labially, and the crown is more compressed or blade-like than are those of corresponding teeth of the lower jaw. The torus on the lingual face of the root is noticeably developed only in the anterior and intermediate teeth; the roots of the upper lateral teeth are flatter than those of the lowers. Exceptions, however, do occur. In the extant Lamna and in Isurus paucus we observed upper lateral teeth with slight lin- gual bends, and in two juvenile /. paucus dentitions from the same locality, the same upper lateral tooth in each dentition has a strong lingual bend. Medial Teeth.—Medial teeth are small, often symmetrical but may be asymmetrical, and occur at the juncture of the left and right jaws. These teeth are found in the Scyliorhinidae, Tri- 80 .a j= II - , 1 Cl 11 ^ o F BO « u> Q F LU II oi .; O 0> CO NUMBER 90 81 akidae, Pseudotriakidae, Carcharhinidae, Sphyrnidae, Hexan- chidae, Squalidae, and Heterodontidae. Symphysial Teeth.—These teeth, which usually look like miniatures of the first anterior teeth, have asymmetrical roots and occur on either side of the symphysis. They are usually in the lower jaw but also may occur in the upper jaw. Symphysi- als are found in the Odontaspididae, Cretolamnidae, Carchar- hinidae, Sphyrnidae, Hemigaleidae, and the Mitsukurinidae. Alternate Teeth.—These are "small teeth with asymmet- rical crowns that form two interdigitated rows on the symphys- is, with the cusps of each row hooked mesially towards the op- posite row" (Compagno, 1988:32). They are found in the Carcharhinidae and Hemigaleidae. We did not recover alter- nate teeth at Lee Creek Mine. ANTERIOR Teeth.—Two upper (three in Carcharias and Mitsukurina) and three lower anterior teeth occur in the denti- tions of lamnoid sharks. These teeth usually have erect, awl- like crowns; their tips may curve slightly toward the corner of the jaw (distally). The width of an anterior tooth is less than 80% of the tooth's height. In Odontaspis, Carcharias, Mitsuku- rina, Scapanorhynchus, and small individuals of Isurus, the an- terior teeth have a sigmoidal curvature in lateral view, but in upper teeth the sigmoidal curvature is not as great, and the root is not as thick in lateral view as it is in the lower anterior teeth (Figure 3). A shallow to deep hollowing of the labial face of the root occurs in these teeth to accommodate the torus of the next-forming tooth. This hollowing or concavity has its great- est development in the lower anterior teeth, which have the most prominent toruses, and in which the crown of the tooth overhangs the root. In upper anterior teeth, where the torus is less developed, the labial face of the root may be slightly recessed or flush with the labial face of the crown. Some exceptions to these characters do occur. ; In the upper teeth of Carcharodon, Isurus xiph- odon, and large individuals of I. paucus, the an- terior teeth lose their awl-like appearance, and in Carcharodon and /. xiphodon the angles of the root lobes are broader than in those of other lam- noid species. Because each anterior tooth exhibits a basic morphological pattern, the identification of the characters that define them is most important, I but one of these characters alone, such as the an- gle formed by the root lobes, is not sufficient for identifying this tooth type. One of us (R.W.P.) measured the angles of the root lobes (see Table 2) of the anterior teeth in dentitions from extant Figure 3.—Upper and lower anterior teeth in lateral view to show sigmoidal curvature of crown: a, upper; b, lower. Table 2.—Variation in the angle of the root lobes in the first two upper (Al, A2) and lower (al, a2) anterior teeth of extant lamnoid sharks. («=number of specimens.) Taxon n Al A2 al a2 Carcharias 2 38o^10° 44°-92° 37°-51° 48°-67° Lamna nasus 9 - 82°-143° 60°-120° 106°-141° L. ditropis 3 - 86°-125° 83°-135° 100°-!30° Isurus paucus 8 - 86°-137° 34°-60° 74°-lll° I. oxyrinchus 4 - 76°-91° 30°-50° 45°-75° Carcharodon 28 - 116°-156° 62°-108° 81°-115° lamnoid species; he found that these angles broaden as the shark increases in size and that the angles differ noticeably for the same tooth position in the left and right jaws. In identifying anterior teeth, then, the attitudes of the crowns and the propor- tional development of the root lobes are the most constant and the most important features (Figure 4). Even these features are variable, and they should be used with prudence. First Upper Anterior Tooth (Applegate's type A tooth): This is the most symmetrical tooth among the upper anteriors, and the root lobes are nearly equal in size and form an acute angle. The crown may appear to be symmetrical or slightly asymmetrical, but it remains erect. Second Upper Anterior Tooth (Applegate's type C tooth): The root lobes of this tooth form a wide acute to right angle, and the lobes are not equal in size (the longer root lobe is usually on the mesial side of the tooth). The crown has a slight distal slant, and the mesial cutting edge is slightly to very convex. Third Upper Anterior Tooth (Applegate's type E tooth): This is the shortest tooth in the upper anterior series. The mesial root lobe is onger than the distal lobe, and the crown leans distally; the distal cutting edge is slightly to very convex. First Lower Anterior Tooth (Applegate's type B tooth): The root lobes of this tooth are nearly equal or equal in length, form an acute angle, and are almost U-shaped in ap- pearance; one lobe may be somewhat flattened. The torus or swollen area surrounding the transverse groove and/or central foramen attains its greatest development in this tooth. In some species the roots of these teeth are elongated. The crown of this tooth, the most symmetrical of the lowers, has the least amount of distal curvature. Second Lower Anterior Tooth (Applegate's type D tooth): The root lobes of this tooth form an acute to small ob- tuse angle, and the mesial lobe is usually longer than the distal lobe. The curvature of the crown is similar to that of the first lower anterior tooth. Except in Carcharodon, this tooth is usu- ally the greatest in height. Third Lower Anterior Tooth (Applegate's type F tooth): This is the shortest tooth in the lower anterior series. The root lobes form a right to obtuse angle, and the mesial root lobe is noticeably longer than the distal one and may be pointed. The torus is more noticeable in this tooth than it is in its upper 82 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY counterpart. The tooth's crown has a strong distal lean to it; its mesial edge is almost straight or is concave. Intermediate Teeth.—Intermediate teeth occur in the up- per jaws between the anterior and lateral teeth. Although they are usually small, they may be almost as large as the neighbor- ing teeth. Two sharks of the same size and species can have in- termediate teeth of markedly differing sizes. These teeth, ex- cept in Carcharodon and Isurus xiphodon, have U-shaped roots, and their crowns, except in Carcharodon, slant toward the corner of the jaws. The mesial edge of the crown is slightly concave to convex, and the distal edge is concave. In Carchar- odon these teeth may be broad- or narrow-crowned, with root lobes that form right to obtuse angles. In Isurus xiphodon the intermediates have triangular crowns and have root lobes that form obtuse angles. LATERAL Teeth.—The root lobes in the lateral teeth form obtuse angles. These angles are usually smaller in the lower teeth than they are in the corresponding upper teeth, which are usually more blade-like than the lower teeth. In the upper teeth the crowns curve toward the angle of the jaws, whereas in the lower teeth they usually tend to be erect; however, we have seen strongly curved lower lateral teeth in Isurus oxyrinchus. Posterior Teeth.—In the posterior teeth, the crowns are small and are not well developed. Systematic Paleontology Class Chondrichthyes Order Hexanchiformes Family Hexanchidae (cowsharks, sixgill sharks, and sevengill sharks) Maisey and Wolfram (1984:172) identified three dental char- acters that, in addition to nondental characters, they considered to unite the living hexanchids: (1) "teeth compressed labiolin- gually; lateral teeth bladelike but with several cusps in a recti- linear series along the cutting edge;" (2) "upper and lower teeth distinctly different, the lowers generally being longer and hav- ing more cusps;" and (3) "posteriormost upper and lower teeth are small button-like, unserrated and lacking cusps." Maisey (1986:101) revised these characters and reduced the number of dental characters to one: "Labio-lingually flattened teeth, with the root and crown lying in the same plane and the basal surface enlarged to form the 'lingual' side of the root." His character, however, also may be applied to the Echi- norhinidae. Thies (1987:197) modified Maisey's dental character to read: "Lower lateral teeth flattened labiolingually and elongat- ed mesiodistally, producing a sawblade-like appearance to the tooth (Maisey and Wolfram 1984, in part)." His modification restricts the application of this character to the Hexanchidae. Thies (1987:197) also added the following two additional den- tal synapomorphies: "Main cusp of lower lateral teeth with a serration on the lower portion of its mesial edge or, alternative- ly, with mesial cusplets," and "tooth root of lateral teeth flat- tened labiolingually and rectangular in shape, with a straight basal edge." Compagno (1984:13) also identified elongate, comb-like, lower lateral teeth, which he called anterolaterals, as a charac- teristic of the family. These dental synapomorphies define the family Hexanchidae. Maisey and Wolfram (1984:173) stated that the lower medial tooth of Notorynchus has a vertical median cusp that is strongly inclined, whereas in Hexanchus it is almost vertical. The medi- an teeth in three dentitions, however, two ofH. griseus (USNM 176566, 188048) and one ofH. vitulus (USNM 110900), all with strongly inclined median cusps, contradict their observa- tions about this tooth; Thies (1987:195) also confirmed that their character is not taxonomically useful. Using the lower anterolateral teeth, Applegate (1965b: 124) identified characters for separating the genera of this family, stating, "Hexanchus possesses serrations on the front [mesial] edge of its most anterior crownlet. Notorhynchus [sic] has small unequal denticles [which increase in size apically] and the third genus in the family Hexanchidae, Heptranchus [sic] has one or two distinct narrow anterior denticles." Maisey and Wolfram (1984:173) and Compagno (1984:17, 19, 22) added that there are eight to 10 distal cusplets in the teeth of adult Hexanchus, five to six distal cusplets in the teeth of adult Noto- rynchus, and "an abruptly high cusp, and up to 7 or 8 distal cus- plets" in the teeth of adult Heptranchias. Applegate did not identify the taxonomic characters of the upper anterolateral teeth. Kemp (1978) was the first paleontologist to try to distin- guish the upper anterolateral teeth of Notorynchus from those of Hexanchus, with the following observations: "Teeth [of Notorynchus] in first rows a little higher than broad ranging through to a little broader than high in last rows. Teeth of Hexanchus are lower and broader in comparison. Primary cusp similar to H. griseus but with fewer crownlets distally, ranging from only 1 in first row to 4 to 5 or 6 in the last row." In the dentitions of the extant hexanchid genera available to us, Kemp's relationship of tooth height to breadth was vari- able in both genera; the Notorynchus condition was found in the upper anterolaterals of Hexanchus and the converse in No- torynchus. In addition to Kemp's character of the number of distal cusplets, the upper teeth of Notorynchus are distin- guished from those of Hexanchus by the presence of one or more cusplets on the mesial cutting edge of the anteriormost teeth and by the presence at the base of the mesial cutting edge of a small shoulder. Because only a small number of dentitions of the living species were available to us, we can- not ascertain the validity of these characters. Using these characters and those identified by Applegate (1965b), we NUMBER 90 83 identified only the genera Notorynchus and Hexanchus among the Lee Creek Mine fossil teeth. Herman et al. (1987:43-56), in their comparative morpholog- ical study of the posterior teeth of the Hexanchidae, found that these teeth possess characters that allow generic identification. These teeth, however, were not recovered at Lee Creek Mine. In the Hexanchidae, ontogenetic variation is known to occur in two genera, Hexanchus and Notorynchus. In juveniles of Hexanchus, Bigelow and Schroeder (1948:82) reported that in the lower lateral teeth "the inner [mesial] margins [are] smooth in newborn specimens, but finely serrate in large, with interme- diate sizes showing intermediate states." Concerning Notoryn- chus, Kemp (1978) noted, "As in Hexanchus the number of crownlets, especially in the lower laterals and the degree of denticulation of mesial margin of all teeth increases with age, and thus the size of the tooth." According to Ward and Thies (1987), the dental formula for each upper jaw is one to two symphysial, five to six ante- rolateral, and six to 13 posterior teeth; in the lower jaws it is one medial tooth; and in each jaw, it is five to six anterolater- al and four to 12 posterior teeth. They did not mention, how- ever, the upper medial tooth that is present in Notorynchus (Kemp, 1978). Kemp (1978) gave the dental formula for the upper jaw (one side) of Notorynchus cepedianus as two medi- al (one medial, one symphysial), six to seven lateral (antero- lateral), and 11 to 13 posterior teeth, and for the lower jaw (one side), one medial, six lateral (anterolateral), and eight to nine posterior teeth. He gave the dental formula for the upper jaw (one side) of Hexanchus griseus as two medial (both symphysial), nine lateral, and eight posterior teeth, and for the lower jaw (one side), one medial, six lateral, and eight to nine posterior teeth. Notorynchus cepedianus (Peron, 1807) Figure 4 Notidanusplectrodon Cope, 1867:141 [Miocene, Maryland]. Notidanus primigenius Agassiz, 1843:303, pl. 27: figs. 6-8, 13-17 [Miocene, Switzerland].—Eastman, 1904:77, pl. 29: fig. 6 [Miocene, Maryland].—Ler- iche, 1942:63-64, pl. 4: figs. 7-9 [Miocene, Maryland], Horizon.—Pungo River Formation (units 1-5); Yorktown Formation (units 1, 2). Referred Material.—Some 300 teeth, USNM 205296, 256290, 256312, 256315, 256316, 282771, 282776, 391921, 459874^459915, 474814-474857, 474871, 474872. Remarks.—Although these teeth are often assigned to the species Notorynchus primigenius, they are identical to those of the living N. cepedianus from the Pacific Coast of North America. Like the extant species, the attitudes of the primary cusps and cusplets range from almost erect to more recum- bent, with the latter attitude being more predominant. Figure 4a shows our reconstruction of this dentition. As noted above, the posterior teeth were not recovered at Lee Creek Mine. In the upper dentition, the first two mesial sym- physial teeth lack a shoulder at the basal extremity of the cut- ting edge, but in comparison to extant hexanchid dentitions, these teeth are most similar to those of Notorynchus. The first of these two teeth, the first symphysial, has a squarish root in labial view; in Hexanchus the root of this tooth appears to be triangular to rhomboidal. As in Notorynchus, in the second of these two teeth the mesial and apical edges of the root form a right angle. In Hexanchus this angle is absent, and the slanted contour of the cutting edge continues down to the base of the root. The remaining teeth in the upper jaw are characteristic of Notorynchus. {0 99 S© fiP ^P 9t { mm ^1P Wm 1.0 mm). NUMBER 90 Figure 5.—Hexanchus sp.: a, USNM 256313, lingual view of lower anterolateral tooth with gradually diminish- ing distal cusplets; b, USNM 474859, lingual view of lower anterolateral tooth with principal cusp much taller than distal cusplets; c, lingual view of seven upper anterior teeth; d. USNM 437771, lingual view of lower lateral tooth with enlarged serrations on its mesial edge; e, USNM 256289, lingual view of largest lower anterolateral tooth from Lee Creek Mine in USNM collections;/ USNM 474860, symmetrical lower medial tooth. (Scale bars: a.b=0.5> cm; c-f=\.0 cm.) Eight of the 22 teeth from Lee Creek Mine are from the an- terior portion of the upper jaw; these teeth possess one to two distal cusplets (Figure 5c). In the extant H. griseus the first two teeth usually do not have distal cusplets; in the teeth that follow these, the number of distal cusplets varies from denti- tion to dentition. In USNM 188048 (female, 433 cm TL), these teeth possess two to three distal cusplets. Ebert (pers. comm., 10 Dec 1990, 17 Apr 1992) has a male dentition from an individual of 333 cm TL having upper teeth with one to two cusplets, and he has examined another dentition from a female of 421 cm TL with two to three distal cusplets on the upper anterior teeth. As also noted by Kemp (1978), the num- ber of distal cusplets increases as the shark grows larger. This variation in the number of distal cusplets evidently is ontoge- netic and not sexually dimorphic. In four comparative dentitions and in dentitions illustrated in published accounts of both extant species, we noticed signifi- cant variation in the morphology of the teeth. The tips of the cusplets are either straight or recurved, and the cusplets of some teeth appear to be more erect than are those of others. In 86 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY dentitions where the mesial and distal ends of the teeth do not abut, the mesial edges of the roots are straight rather than con- cave, and in abutting teeth this concavity ranges from slight to deep. One fossil lower anterolateral tooth with eight distal cusplets (USNM 437771) exhibits a morphology that we have not ob- served in the teeth of the extant species; the mesial serrations, which are not recurved, are greatly enlarged, being as large as some of the distal cusplets (Figure 5d). Nothing about this tooth suggests that it is pathologic. Compagno (1984:20, 21) reported that the extant species in- habit continental and insular shelves in temperate and tropical seas in water depths exceeding 90 m; they feed on other sharks, bony fishes, carrion, seals, and crustaceans. Ebert (1994:216) noted that in the extant species, the young feed principally on cephalopods, and that as they mature, bony fishes and chon- drichthyans become an increasingly important part of their diet. The few individuals greater than 200 cm TL that he examined fed on cetaceans and on larger, more active teleosts, such as marlin and swordfish. Order Squaliformes Family Echinorhinidae (bramble sharks) Echinorhinus blakei Agassiz, 1856 Figure 6 Echinorhinus blakei Agassiz, 1856:272, pl. 1: figs. 7, 8, 17 [Miocene, Califor- nia].—[Not Echinorhinus blakei Jordan and Hannibal, 1923, pl. 4: figs. c,d.] HORIZON.—Pungo River Formation (units 4, 5). Referred Material.—13 complete teeth and fragments of teeth, USNM 207608, 207609, 280334, 281317, 281331, 287739, 287740, 412221, 457232^*57236. Remarks.—The upper and lower teeth of bramble sharks are alike (dignathic homodonty). This small to medium-sized shark has smooth-edged teeth, with the strongly oblique central cusp sloping toward the corners of the mouth. In published dentitions of Echinorhinus brucus (Bigelow and Schroeder, 1948, fig. 102; Bass etal., 1976, pl. 11; Herman et al., 1989, pi. 1) this obliquity varies noticeably. The greatest obliquity oc- curs in Bigelow and Schroeder's dentition of an unsexed indi- FlGURE 6 —Echinorhinus blakei: a. USNM 412221, posterolat- eral tooth, lingual view; b, USNM 207609, upper posterolateral tooth, lingual view; c. USNM 207608, posterolateral tooth. Echi- norhinus blakei. Miocene, California: d.e. views of holotype. Echinorhinus richiardi. Pliocene, Tuscany:/-A, syntypes. Echi- norhinus caspius, Oligocene, Armenia: i, lingual view. Echi- norhinus priscus. Eocene, Morocco: j,k. lingual views. Echi- norhinus brucus: l.m. outlines of partial dentitions of the extant Atlantic species, after Bigelow and Schroeder (1948); n. USNM 287739, lower posterolateral tooth, lingual view; o. same speci- men, labial view;/;, USNM 457235, upper posterolateral tooth, lingual view; q. same specimen, labial view; r. USNM 457236, upper posterolateral tooth, lingual view; s, same specimen, labial view. (Scale bars= I 0 cm.) m * n gfc J^jMin *VwS&gpB> ^^WJHf NUMBER 90 87 vidual, and the least occurs in Herman et al.'s dentition from a 228 cm female. In these dentitions the mesial edges of the teeth also exhibit noticeable variation; those of Herman et al.'s denti- tion are very convex, whereas in the other dentitions they are straight to slightly convex. In adults, one to three lateral cus- plets may be present on the mesial side of the central cusp, and one to two lateral cusplets may be present on the distal side; some of these cusplets may occur unpaired. In juveniles only the main cusp is present (Compagno, 1984:25). We refer 12 teeth from the Pungo River Formation to Echi- norhinus. They show considerable variation in the number of accessory cusplets and in the degree of curvature of the edges of the central cusp (Figure 6n-s). One tooth (Figure 6n,o; cf. Figure 6d,e) strongly resembles the holotype of E. blakei from the Miocene Temblor Formation of California. Of two extant species recognized by Compagno (1984), E. brucus (Bonna- terre, 1788) and E. cookei Pietschmann, 1928, the fossil teeth bear a closer resemblance to those of the latter species; how- ever, this comparison is based on the pubished dentitions of E. brucus cited above and on one of E. cookei (Garrick, 1960, fig. 3). The earliest known echinorhinid teeth, E. caspius Glikman (1964) from the early Oligocene of Armenia (Figure 6/) and E. priscus Arambourg (1952) from the early Eocene of Morocco (Figure 6j,k), are simple, lacking well-developed cusplets. In the extant species, the juvenile teeth (Figure 6m) and the adult posterior teeth also lack cusplets (Garrick, 1960; Compagno, 1984; Herman et al., 1989), and the type specimens of Glikman and Arambourg are very similar to the juvenile teeth of the ex- tant species. Without examining the types of the fossil species and many dentitions of the extant species, the affinities of these Paleogene species cannot be determined. Echinorhinus richiardi from the Pliocene of Tuscany (Figure (>f~h), with well-developed lateral cusplets, may belong to the extant species E. brucus. The teeth identified as Echinorhinus blakei by Jordan and Hannibal (1923) are fragmentary hexanchid teeth. The tooth in their pl. 4: fig. d is a median tooth of Hexanchus; the generic identity of the other tooth figured is indeterminate. We provisionally assign the teeth from Lee Creek Mine to E. blakei; without an extensive series of dentitions from the extant species we cannot determine if the teeth of the fossil and extant species are separable. The fossil teeth may be identical to those of E. cookei, and this latter species may be a junior synonym of E. blakei. The height of the teeth ranges from 9.4 to 11.8 mm, and the width ranges from 11.8 to 15.8 mm. The living species of this shark (Compagno, 1984:26, 27) are bottom dwellers in temperate to tropical seas, sometimes oc- curring in shallow water but primarily in deep water. Echi- norhinus brucus occurs in waters from 18 to 90 m deep, and E. cookei occurs in waters from 11 m to at least 424 m in depth. They feed on other sharks and on bony fishes, including cat- fish, hake, and flounder. Family Squalidae (dogfish sharks) Squalus sp. Figure 7a,b HORIZON.—Yorktown Formation (units 1, 2). Referred Material.—1 tooth, USNM 207546. REMARKS.—Unlike Squalus acanthias, the common North Atlantic dogfish, but like S. almeidae, this tooth is rather ro- bust. Its cutting edges are smooth but irregular. On its labial face (Figure la) the crown foot is slightly convex and extends basally to form a peg; this peg extends well below the basal margin of the root. Like S. almeidae, the lateral edges of the peg converge but become parallel near the peg's basal end. Lat- eral to the peg the boundary between the crown foot and the root is sinuous. On the lingual face of the tooth (Figure 7b), the crown foot extends basally to form a prominent process; this process has a deep central depression (also observable in S. acanthias), mak- ing it V-shaped. On either side of this process the root is exca- vated, and clearly marked foramina open into these depres- sions. The basal margin of the root forms a ridge, which is bisected by a transverse groove. In the extant species the teeth are alike in both jaws, being low crowned, blade-like, interlocked teeth with a single cusp and a distal enamel shoulder on a low root; the cutting edges are smooth in living species but are serrated in some fossil spe- cies (Compagno 1984:109). Bass et al. (1976:13) noted that in Squalus acanthias "slight sexual dimorphism is apparent, the male having teeth with more erect and pointed cusps." Although the teeth from Lee Creek Mine share characters with Squalus almeidae from the middle Miocene of Portugal, USNM 207546 differs from this species in the sinuosity of the mesial cutting edge, which is rectilinear in S. almeidae. Because at this time we cannot assess the taxonomic value of this charac- ter, we do not assign the Lee Creek Mine specimen to a species. This specimen measures 2.9 mm in height and 4.2 mm in width; its size falls within the size range of the extant Squalus acanthias. According to Compagno (1984), the latter species ranges in size from 22 cm at birth to 160 cm TL. The extant Squalus acanthias inhabits boreal to warm tem- perate waters from the intertidal zone to 900 m in depth. This shark feeds primarily on bony fishes (Compagno, 1984:112). Isistius sp. (cookiecutter sharks) Figure 7f,g HORIZON.—Yorktown Formation (units 1-3). Referred Material.—5 teeth, NCSM 11287, 11288, 11291, 11292, USNM 475362. Remarks.—While bulk sampling a Yorktown Formation spoil pile, one of us (V.P.S.) recovered five lower teeth of this 88 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY species. These teeth compare favorably with those illustrated by Bigelow and Schroeder (1948:510). All of the teeth have broad, flat, triangular crowns with smooth, very thin, transpar- ent cutting edges. Of the five teeth, NCSM 11287 (Figure If) and NCSM 11292 (Figure Ig) are nearly complete. NCSM 11287 is the largest, measuring 8.3 mm in height and 5.3 mm in width; a central foramen and shallow transverse groove are present on the lingual face of the root. In NCSM 11292 the root basal to the central foramen is missing; this tooth measures 3.7 mm in height and 2.4 mm in width. These teeth compare favorably with the types of Scymnus tri- angulus (Probst, 1879:175), which Cappetta (1970) rightly re- ferred to Isistius. The Lee Creek Mine teeth, however, differ from Probst's species and the living species in lacking the me- dian groove. With the lack of ample comparative extant and fossil material, we cannot judge the validity of this character or the validity of I. triangulus Probst. Of the two living species, /. brasiliensis, which has moderately large teeth, and /. plutodus, which has enormous teeth (Compagno 1984:93-95), the Lee Creek Mine species appears to be more closely related to /. brasiliensis. Although teeth of Isistius have been found in the Miocene and Pliocene sediments of Europe and South America (Ecua- dor) (Cappetta, 1987:64), the specimens from Lee Creek Mine represent the first occurrence of this taxon in North America. According to Compagno (1984:94), the extant I. brasiliensis is a tropical, oceanic shark, epipelagic to bathypelagic in distri- bution. In addition to feeding on squid, gonostomatids, and crustaceans, this shark "has highly specialized suctorial lips and a strongly modified pharynx that allow it to attach to the sides of large bony fishes such as marlin, tuna, albacore, wa- hoo, and dolphinfishes, as well as dolphins and other cetaceans and even the megamouth shark (Megachasma). The shark then drives its razor-sharp sawlike lower dentition into the skin and flesh of its victim, twists about to cut out a conical plug of flesh, then pulls free with the plug cradled by its scooplike low- er jaw and held by the hooklike upper teeth" (Compagno, 1984:94). Order PR1STIOPHOR1FORMES Family PRISTIOPHORIDAE (saw sharks) FIGURE 7.—Squalus sp.: a. USNM 207546, labial view; b. same specimen, lin- gual view. Pristiophorus sp.: c, USNM 207583, rostral spine, lateral view. Rhi- nobatos sp.: d, USNM 207544, lingual view; e. USNM 207545, lingual- occlusal view. Isistius sp.:/ NCSM 11287, lower tooth, lingual view; g. NCSM 11292, lower tooth, labial view. (Scale bars=0.25 cm.) Pristiophorus sp. Figure 7c HORIZON.—Pungo River Formation (units 1-3). Referred Material.—6 rostral teeth, USNM 207583, 299481, 467554-467556,467587. Remarks.—We assign these teeth to this genus rather than to Pliotrema because their dorsoventrally compressed and enameled crowns are not serrated on the distal edges. In Pliotrema (Fowler, 1941:283) these distal edges are serrated. The largest tooth in this sample measures 19.2 mm in length (Figure 7c). Because large samples of the dentitions of the ex- tant species were not available, we believe it is not prudent to identify these specimens beyond genus. According to Compagno (1984:137), the extant Atlantic spe- cies is "a little known, deep-water, tropical sawshark of the continental and insular slopes of the Bahamas region, occurring on or near the bottom at depths from 640 to 915 m." Its pres- ence in the shallower Pungo River seas may have been through the excretions or the regurgitated hard parts from a predator shark. NUMBER 90 89 Order Rhinobatiformes Family Rhinobatidae (guitarfishes) Rhinobatos sp. Figure ld,e HORIZON.—Pungo River Formation (units 4, 5). Referred Material.—9 teeth, USNM 207544, 207545. REMARKS.—In occlusal view, these teeth are oval to tetrago- nal (Figure 7e); a peg extends down from the lingual surface of the crown onto the surface of the root (Figure Id), which is lin- gually deflected. A transverse groove bisects the root. Because the range of dental variation in the extant species is not yet known and because we were unable to examine any specimens of the extant species, we believe it is unwise at this time to at- tempt a taxonomic evaluation of the fossil species or to attempt a more precise identification of the Lee Creek Mine specimens. The extant species, Rhinobatos lentiginosus, inhabits sub- tropical to tropical waters of the western Atlantic Ocean (Big- elow and Schroeder 1953:66-67); it occasionally reaches as far north as Cape Hatteras, North Carolina. Little is known about its range and habits. Order Pristi FORMES Family Pristidae (sawfishes) Pristis sp. Figure 8a, b Horizon.—Yorktown Formation (unit 1?). Referred Material.—3 rostral teeth, USNM 281389, 281390,412220. Remarks.—These teeth may be those of juveniles because they lack a groove along their posterior edges; this condition also is found in an extant species with rostral teeth of similar form (Pristis sp., USNM 232696). All three specimens are slightly abraded, and their surfaces exhibit bite marks. Teeth from Lee Creek Mine are relatively short (23.5-25.5 mm in length) and thin (2.6-2.9 mm in thickness; ratio of width to thickness 2.1-2.7). They show no trace of a subbasal barb. Bigelow and Schroeder (1953:28-29) reported that the ex- tant common sawfish (P pectinatus) inhabits warm temperate to tropical coastal waters and feeds on bony fishes and bottom- dwelling animals. Pristis cf. P. pectinatus Latham, 1794 Figure Sc.d HORIZON.—Yorktown Formation (unit 1?). Referred Material.—1 rostral tooth, USNM 412219. Remarks.—This rostral tooth does not differ appreciably from those of the extant Pristis pectinatus. The exserted part of FIGURE 8.—Pristis sp.: a, USNM 412220, dorsal view; b, same specimen, dis- tal view. Pristis cf. P. pectinatus: c, USNM 412219, dorsal view showing reg- ularly tapering exserted portion; d, same specimen, posterior view showing position of groove. (Scale bar= 1.0 cm.) the tooth is moderately short, being about equal to the inserted part. The apex of the crown is aligned with the mesial axis rath- er than being displaced distally. On the exserted portion of the tooth, the distal margin is straight (Figure 8c), with a noticeable groove (Figure Sd) that extends to the tip of the tooth, rather than a cutting edge. On the inserted portion of the tooth the dis- tal margin bends mesially. The outline of the mesial margin of the tooth is obliterated by wear and breakage, but the exserted portion was probably convex and rounded in cross section. The tooth is 52.7 mm in proximodistal length, 12 mm in anteropos- terior length, and 7.5 mm thick at the proximal end, tapering to 2.0 mm at the worn apex. Oral teeth, which also are important for the identification of fossil pristid remains, were not recovered at Lee Creek Mine. Order Rajiformes Family Rajidae (skates) Raja sp. Figure 9 HORIZON.—Yorktown Formation (unit 1). Referred Material.—2 teeth, USNM 476398,476399. Remarks.—Two teeth recovered from the spoil piles of principally Yorktown matrix compare favorably with those in a dentition of a male of the extant species Raja laevis (USNM 110962). USNM 476398 (Figure 9a,b), from the anterior por- tion of the jaw, has an awl-like crown that inclines strongly lingually; the crown foot extends laterally, creating a shelf that circumscribes the crown. Basal to the crown foot the root is constricted but flares basally. On its labial side the root is flattened, which gives it a semicircular cross section. A deep 90 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE 9.—Raja sp.: a, USNM 476398, labial view; b, same specimen, lateral view; c, USNM 476399, labial view; d, same specimen, lateral view. (Scale bar=0.l cm.) transverse groove bisects the root, giving it a bipedal appear- ance. This tooth measures 7.25 mm in height and 5.35 mm in width. In USNM 476399 (Figure 9c,d) the crown is stubby and rounded by wear; the root is similar to that of the above specimen. This tooth measures 5.20 mm in height and 5.40 mm in width. The extant Raja laevis feeds chiefly on larger crustaceans, but it also eats bony fishes, including tautog and hake (Bigelow and Schroeder, 1953:223). According to McEachran and Mu- sick (1975:119), this skate ranges from the Gulf of St. Lawrence to Cape Hatteras and occurs to depths of 375 m. Order Myliobatiformes Family Dasyatidae (stingrays) Dasyatis say (Lesueur, 1817) Figure 10a-r Horizon.—Pungo River Formation (units 1-5); Yorktown Formation (units 1, 2). Referred Material.—433 isolated teeth, USNM 207584-207588, 301684-302045. NUMBER 90 91 FIGURE 10.—Dasyatis say, male and female teeth: a, USNM 207584, Yorktown Formation, female tooth, occlusal view; b, same specimen, lateral view; c, same specimen, labial view; d, same specimen, lingual view; e, USNM 207585, Yorktown Formation, female lateral tooth, occlusal view (lingual side down);/ same specimen, lateral view; g, same specimen, labial view; h, same specimen, lingual view; /, USNM 207586, Yorktown Forma- tion, male tooth, occlusal view;/ same specimen, lateral view; k, same specimen, labial view; /, USNM 207587, Pungo River Formation, male tooth, occlusal view; m, same specimen, lateral view; n, same specimen, labial view; o, USNM 207588, Pungo River Formation, male medial? tooth, occlusal view; p, same specimen, lateral view; q, same specimen, labial view; r, same specimen, lingual view. Dasyatis cf. D. americana, USNM 207548, Pungo River Formation: s, lateral view; t, labial view; u, occlusal view. (Scale bars: a-r=0.5 cm; s-u=0.25 cm.) Remarks.—The Lee Creek Mine teeth are morphologically identical to those of the modern Dasyatis say. They both dis- play a punctate occlusal surface, a well-defined cutting edge di- viding the occlusal surface and lingual apron, and abbreviated striations that extend a short distance down the lingual apron from the cutting edge. Bigelow and Schroeder (1953:382) and Taniuchi and Shimi- zu (1993) reported that the teeth of D. say exhibit sexual di- morphism. The teeth of females and immature males are "qua- drangular with blunted corners, about as broad (transversely) as long (anteroposteriorly), the functional surface weakly rounded or more or less irregular from wear; those of mature males with low and broadly triangular cusps, largest in youngest rows" (Taniuchi and Shimizu, 1993:54). Teeth from both sexes occur at Lee Creek Mine; those of females are illustrated in Figure IQa-h, and those of males are illustrated in Figure 10/-r. The teeth range in height from 2.6 to 3.8 mm (mean=3.2 mm, «=5) and range in width from 2.8 to 3.9 mm (mean=3.3 mm, n=5). Although the extant species normally inhabits subtropical to tropical waters, it ventures into more temperate waters during the summer. Bigelow and Schroeder (1953:358-359) reported that D. say "has been reported as deep as 6-20 fathoms; and some may summer where the depth is as great as 20-30 fath- oms, if reports of their occurrence on Georges Bank are well 92 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY founded." The extant species probably feeds on crustaceans and mollusks. Dasyatis centroura (Mitchill, 1814) Figure Wa-d, g-i HORIZON.—Yorktown Formation (units 1, 2). Referred Material.—About 350 dermal denticles, USNM 182122, 182123, 182129-182135, 182137, 182139, 182141, 280071-280305, 353680, 353684, 445534,445544, 476402, 476403, 482224^182229. Remarks.—Numerous dermal denticles were recovered from the spoil piles, particularly in the Yorktown matrix, that are very similar to those found along the tail and back of the extant species Dasyatis centroura. These denticles are circular to oval in outline, with one or more centrally located conical points. Silas and Selvaraj (1985:252) reported that the sharp- ness of the conical points decreases as the base of the denticle Figure 11 .—Dasyatis centroura: a, USNM 353684, dorsal view of dermal denticle marked by radiating rugosi- ties with enameloid caps missing; b. USNM 280071, dorsal view of dermal denticle with three conical points with enameloid caps preserved; c, USNM 353680, dorsal view of dermal denticle with triangular enameloid cap Recent D. centroura: d, USNM 197504, dorsal view of caudal denticles. Ceratopterus unios, holotype, ANSP 8069, dermal denticle: e, dorsal view;/ lateral view. Dasyatis centroura: g, USNM 482224, dorsal view'of der- mal denticle; h. same specimen, lateral view; /, USNM 482225, dorsal view of dermal denticle with serrate enameloid cap. (Scale bars= 1.0 cm.) NUMBER 90 93 increases in size. Unlike D. centroura, which is known to have denticles that support only two or three conical points (Bigelow and Schroeder, 1953:355), the Lee Creek Mine specimens sup- port up to 10 conical points, which may be the result of the Lee Creek Mine denticles coming from specimens larger than those available to Bigelow and Schroeder. The external surfaces of these denticles may have fragile, tri- angular (narrow to wide) to conical enameloid caps (Figure 1 \b,c) (terminology of Reif, 1979). Basal to these caps, the ex- ternal surface of the denticle may be smooth to strongly striat- ed. According to Bigelow and Schroeder (1953:355), those with striated bases occur only on the tail. On the dermal denti- cles of a tail of a recent D. centroura (USNM 197504) (Figure 1 \d) and in a Lee Creek Mine specimen (USNM 280071, Fig- ure 1 lb), these striations extend under the enameloid cap. In Lee Creek Mine specimens that lack enameloid caps, these stri- ations extend almost to the apex of the denticle (Figure 1 \a). Paleontologists (Larrazet, 1886; Zittel, 1887-1890; Reif, 1979) have identified fossil specimens without their enameloid caps as Acanthobatis; these should be referred to Dasyatis. Figure 11/ shows a narrow denticle, USNM 482225, that we believe also belongs to Dasyatis centroura. The edges of the enameloid cap are serrated in the same manner as the caudal spines. According to Stehmann (1981), the extant species inhabits tropical latitudes, occurring in waters to depths of 300 m, and it feeds principally on bivalves, crustaceans, and worms. Dasyatis cf. D. americana Hildebrand and Schroeder, 1928 Figure \0s-u Horizon.—Pungo River Formation (units 1-3). Referred Material.—1 isolated tooth, USNM 207548. Remarks.—The root is dasyatid except that the specimen il- lustrated has two grooves rather than the usual one. The root constricts to one-third of the basal diameter at the crown-root boundary. Like most male dasyatid teeth, the crown is acumi- nate but asymmetrical, the occlusal surface is convex labially and divided into two ridges by a deep medial sulcus, and the basal margin is accentuated by a cingulum-like irregular ridge. The Lee Creek Mine specimen is 3.5 mm in maximum height and 1.9 mm in width (mesial-distal diameter). Except for the tripartite root, this tooth, which has a medial sulcus on the occlusal surface, resembles those of the male Dasyatis americana. According to Bigelow and Schroeder (1953:350-351), the extant species prefers warm-temperate to tropical coastal wa- ters. It is not known to occur at the depths represented by the Pungo River Formation; their presence there may be the result of regurgitation by sharks that fed on them. The extant Dasyat- is americana feeds on blue crabs, clams, shrimp, worms, and small bony fishes. Nota Bene Leidy (1876:86) described as a new species of manta ray (Ceratoptera unios) a large, thick, dermal denticle (length=8.2 cm, width=5.4 cm, thickness=3.6 cm) of Dasyatis. The holo- type (ANSP 8069), which Leidy mistook for a caudal spine, has an elongate, oval enameloid cap. Basal to this cap the den- ticle surface is striated. At one extremity of its long axis, a facet indicates that this denticle abutted another one. This specimen (Figure 1 le,/), which was found in the phosphate beds along the Ashley River, South Carolina, and may be from the Pliocene, is more massive than any found at Lee Creek Mine. Two fused denticles from Lee Creek Mine, USNM 482224, which together measure 4.8 cm in length (width=2.4 cm, thickness=1.4 cm), are similar in morphology (Figure \\g,h). These fused denticles support our assignment of Ceratoptera unios to Dasyatis. At this time we cannot ascertain the specific identity of Leidy's holotype. Family Myliobatidae (eagle rays) Pteromylaeus sp. Figure 12a,* Horizon.—Pungo River Formation (units 1-5). Referred Material.—7 partial dental pavements, USNM 24758, 25448,297760,464159. Remarks.—These specimens may be indistinguishable from the extant species, Pteromylaeus bovinus, which is pres- ently found in the eastern Atlantic Ocean and the Mediterra- nean Sea. With only one specimen of the extant species avail- able to us, we cannot confirm the identity of the fossil species; thus, we identify the teeth from Lee Creek Mine to genus only. The largest specimen of this species from Lee Creek Mine is USNM 464159 (Figure 12A:). We refer the Lee Creek Mine specimens to Pteromylaeus be- cause the length of the median teeth is seven times greater than their width, a character that McEachran and Capape (1984:206) and Capape and Quignard (1975:1335) used to separate the teeth of Pteromylaeus and Myliobatis. Garman (1913:438) characterized the teeth of Pteromylaeus as having three rows of very narrow lateral teeth; however, the width of the teeth ap- pears to be variable because Capape and Quignard (1975:1331) figured a dentition with lateral teeth that are identical to those of Myliobatis. Nishida (1990), who studied the phylogeny of the myliobatoids, stated that the size and shape of these teeth are highly variable. When large suites of dentitions of mylio- batids become available for study, our generic assignment for the fossil species may have to be revised. The extant species inhabits warm-temperate to tropical wa- ters (McEachran and Capape, 1984:205) and feeds on bottom- dwelling crustaceans and mollusks. 94 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE 12.—Pteromylaeus sp.: a. USNM 24758, lower dental battery, occlusal view. Aetobatus sp.: b, USNM 312263, lower dental battery, occlusal view; c, USNM 312264, upper dental battery, occlusal view. Rhinoptera sp.: d. USNM 24745, lower medial tooth, occlusal view; e, same specimen, anterior view. Plinthicus stenodon.fi USNM 207591, incomplete medial tooth showing narrow occlusal surface; g. same specimen, lingual view showing bifurcating grooves on crown and multilobed root; h. same specimen, labial view;y, same specimen, lat- eral view. (', USNM 207592, lateral tooth, occlusal view. Pteromylaeus sp.: k, USNM 464159, lower dental bat- tery, occlusal view. (Scale bars= 1.0 cm.) Aetobatus sp. Figures \2b.c, 13 Horizon.—Pungo River Formation (units 4, 5); Yorktown Formation (units 1, 2). Referred Material.— 3 uncataloged incomplete upper dental plates (Pungo River Formation phosphatic limestone); numerous uncataloged incomplete isolated teeth; 2 lower and upper dental batteries, USNM 312263, 312264; cast of 1 re- stored upper dental pavement, USNM 489120 (Yorktown For- mation). Remarks.—Most of the specimens from Lee Creek Mine are incomplete, isolated teeth. They include chevron-shaped lower teeth and gently arching upper teeth that are characteris- tic of Aetobatus. The largest and only specimen with complete teeth from Lee Creek Mine is in the collection of George W. Powell, Jr., who donated an excellent cast of this specimen to the Smithsonian Institution (USNM 489120, Figure 13). This specimen, which is 24.2 cm long and 12.5 cm wide, has the lin- gual portion of the dentition nearly complete; the labial portion of the dentition consists of teeth with large fragments missing. NUMBER 90 Figure 13.—Aetobatus sp., USNM 489120, Yorktown Formation, cast of upper dental battery collected and restored by George W. Powell, Jr., occlusal view. (Scale bar= 1 cm.) In form, the teeth are gently arcuate, with no lateral tooth rows present, which is characteristic of the genus (Bigelow and Schroeder, 1953:452). On the right lateral extremity of the den- tition, each tooth tapers to a rounded point. On the left side, the teeth are truncated and have rounded corners. Because denti- tions from large individuals of the extant species of Aetobatus were not available to us, we believe that at this time it is pru- dent not to identify this specimen to species. According Bigelow and Schroeder (1953:461), the extant species inhabits shallow to deep, warm temperate to tropical waters. It feeds on bivalve mollusks. 95 Family Rhinopteridae (cownose rays) Rhinoptera sp. Figure \2d,e HORIZON.—Pungo River Formation (units 1-5). Referred Material.—Approximately 200 fragments of teeth; 52 isolated teeth, USNM 24745, 284831, 312262. Garman (1913), Gudger (1933), and Nishida (1990) reported that the size, shape, and number of rows of these teeth (5-19) are highly variable. Because sufficient samples of the denti- tions of the extant species were unavailable to us, we believe it is premature to assign the Lee Creek Mine specimens to a fossil species or to evaluate the validity of the fossil species. Remarks (isolated teeth).—The crown height of unworn teeth is approximately equal to the width of the crown (crown height/width=0.9-1.3, mean=1.06, n=6). In medial teeth the crown heights are uniform from one side of the tooth to the other, but in the first rows lateral to the medial teeth, the crowns are higher on the mesial side than on the distal side. In unworn teeth, the occlusal surfaces either are smooth or are covered with faint and irregular depressions. The transverse edges of the crown show irregular vertical ridges; on the labi- al side of the tooth, the surface is smooth and shiny, but on the lingual side of the tooth, it usually has a finely pebbled tex- ture. On this same side of the tooth just below the crown, a prominent transverse ridge extends the full length of the tooth. The crown overhangs the root on all but the lingual side, where the root margin is even with or projects beyond the base of the crown. As the teeth grew, the number of grooves in the root decreased; for example, the number of grooves per cm (spacing of grooves) decreased from a maximum of 13.6 (small median tooth, transverse width=14 mm) to 7.5 (intermediate- sized upper medial tooth, transverse width=23 mm). The larg- est specimens have 7.5 to 8.5 grooves per cm. The extant cownose ray inhabits coastal to deep, warm-tem- perate to tropical waters, but it migrates into more temperate waters during the summer (Rogers et al., 1990). It feeds on bi- valve mollusks. Plinthicus stenodon Cope, 1869 Figure !2/-y HORIZON.—Pungo River Formation (units 1-6). Referred Material.—Several hundred isolated teeth, USNM 207591, 207592. Remarks.—This species, known only from isolated teeth, is easily recognized. The occlusal face is concave and is covered with labiolingually oriented rounded ridges and grooves. On all specimens the width of the occlusal surface is about 2.5 mm, or about one-third the height of the medial teeth and about one- half the height of the teeth of the presumed second lateral row. 96 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Labial and lingual articulating surfaces bear closely spaced (20 per cm) vertical ridges that often split and divide. When viewed laterally, the laminae on the root (which is the myliobatid type) are subcircular and are unusually thin (0.3 mm) and fairly widely spaced (6 per cm). Three types of teeth are present in the sample: (1) asymmet- rical teeth about 1 cm in length, higher mesially than distally; (2) symmetrical short teeth also about 1 cm in length; and (3) symmetrical long teeth. Teeth of the third type are almost al- ways broken, but the contacting pieces of a complete tooth sometimes can be recovered by careful collecting. Two such teeth, each 42 mm in length, were collected from the sandy layers of units 4 and 5 of the Pungo River Formation. When arranged in a rhinopterid pattern, the short teeth of types one and two would represent the second and first lateral rows, re- spectively, and the long teeth of type three would represent the medial row. This series formed a rather coarsely ridged pavement that, judging from the seemingly little wear that ap- pears on the occlusal surfaces, was not used for heavy-duty crushing. Cappetta (1970, 1987) assigned Plinthicus to the mobulids because the tooth is high and thin and little wear appears on the occlusal surface and because of the osteodentine histology; however, his sample was too small to reconstruct the dentition. When we reconstructed the dentition of this ray, the teeth best fit in a typical Rhinoptera pattern, with a transversely convex upper pavement and a nearly flat or only slightly transversely concave lower pavement. Both of the dental pavements were longitudinally convex, the upper more so than the lower. Each dental pavement appears to have consisted of seven rows of teeth: three distinct sets of transversely elongated teeth, a medi- al and two lateral rows, and four rows of lateralmost teeth with nearly equal transverse and longitudinal widths, the condition in the extant species. The specimens that Fowler (1911) attributed to Plinthicus belong in part to Myliobatis and in part to Aetobatus. A few teeth with lower crowns were found in the same lay- ers as the above specimens; these have a similar root pattern but a distinctly different occlusal surface. This surface is less rugose, and it inclines sharply downward toward the labial edge. It is uncertain if this represents a different form or is a result of wear. The occlusal surface does not appear to be abraded. Family MOBULIDAE (manta rays) Mobula sp. Figure Ha-p Horizon.—Pungo River Formation (units 1-5); ?Yorktown Formation (unit 1, possibly redeposited). Referred Material.—40 isolated teeth, USNM 207549, 207579-207582; 7 caudal spines, USNM 285372, 285381, 291226, 421695, 467584, 467585. Remarks.—Notarbartolo-di-Sciara (1987:9) reported that ontogenetic variation and sexual dimorphism occur in the teeth of Mobula, stating, "Heterodonty is one of the most salient mobulid characteristics. Sexual dental dimorphism, as well as ontogenetic, dignathic, and monognathic heterodonty, all occur in most Mobula species. One cause of such a high degree of variability may be that teeth often appear to grow in width, therefore increasing the number of cusps and of root lobes, pri- or to the branching of one row into two. As a consequence of the variety of tooth shapes which can be found within the same toothband, the use of tooth morphology as a systematic tool may be misleading in living forms, and quite problematical in palaeontological [sic] studies. An effort should be made of identifying tooth characters which remain constant within each species, if tooth morphology is to be used as a taxonomic aid in defining the systematics of the genus Mobula." In view of the above, we herein identify the three morpho- types that occur at Lee Creek Mine to genus only. The teeth of the first morphotype (USNM 207549, 207579, 207580; Figure 14o-i) have flat, triangular, ovoid, or rectangu- lar occlusal surfaces and one to four lingual cusplets. A row of tubercles may occur on the occlusal surface along the labial edge or along the labiolingually directed shallow striations. In sagittal section the root is wedge-shaped to trapezoidal and is basally divided into one to four grooves, the number of grooves increasing with an increase in the length of the tooth. Teeth of the second morphotype (USNM 207581; Figure \4m-o) are small (1.5 mm total height), with a prominent, me- dian, lingually directed cusplet bordered on either side by a pair of much shorter lateral cusplets. The occlusal surface is marked by relatively deep and labiolingually directed sulci that extend from the labial edge of the tooth, where they are best devel- oped, partly onto the median cusplet. The lateral edges of the crown are usually higher than the middle of the occlusal sur- face, making the latter concave. The root is either bipartite or tripartite. In the third morphotype (USNM 207582; Figure 14/'-/), the teeth have broad, labiolingually compressed crowns. The labial and lingual faces appear corrugated, with strong, vertical ridges separated by deep grooves; the ridges may bifurcate or end part way up the crown face. The occlusal face is flattened; this sur- face is triangular in occlusal view on narrow specimens but zig-zags on more elongate specimens, an effect caused by the jagged lingual edge. The root is small and is divided basally into two to five lobes. Seven caudal spines from the spoil piles at the Lee Creek Mine are referred to this genus. They compare favorably with the spine illustrated by Notarbartolo-di-Sciara (1987:58), ex- cept that just basal to where the exserted spine inserts into the bulbous base there are two thin, lobate structures (Figure \4p). NUMBER 90 97 The extant species (Bigelow and Schroeder, 1953:493) in- habits warm temperate to tropical coastal waters, feeding on small shrimp and minnows. Manta sp. Figure 14 yy *& ^ i& >x*f ^ V ¦¦&& FIGURE 22.—Alopias cf. A. vulpinus: a. composite dentition, with intermediate tooth, fourth to sixth and tenth to seventeenth upper lateral teeth, third lower anterior tooth, and sixth to seventeenth lower lateral teeth missing; b, USNM 464241, vertebra. Celorhinus sp.: c, USNM 312269, tooth, lateral view; d, same specimen, lingual view, e, USNM 467556, clasper spine, dorsal view. Holotype oi Alopias triqueter (=A. cf. A. vulpinus), ANSP 1214:/ lingual view; g, labial view. Alopias cf. A. superciliosus: h, USNM 475448. tooth, lingual view. (Scale bars: a,b,e,h=\.0 cm; c,rf=0.25 cm;/g=0.33 cm.) Leriche (1927:76) and Cappetta (1970:23) also identified incomplete cutting edges as a character to distinguish A. ex- igua from A latidens. In Probst's type suite for A exigua, the two non-Alopias teeth are the only ones with incomplete cut- ting edges; therefore, both species have teeth with complete cutting edges. The validity of A. latidens will have to await the restudy of Probst's types, if they still exist, study of Leriche's type spec- imens, and study of dental variation in a large population of A vulpinus and large, synchronous populations of the fossil spe- cies. In view of the above and the limited number of fossil and extant specimens, we cannot find characters to separate the Lee Creek Mine specimens from the extant species, and we believe the assignment of the Lee Creek Mine allopiid to a species is unwarranted at this time. The teeth from Lee Creek Mine (Figure 22a) have wide, stocky crowns with complete cutting edges, and the callosity on the labial face of the crown foot is rounded, sometimes NUMBER 90 109 forming a ridge, and is plicated. Of the three extant species of Alopias, these characters occur in the teeth of A. vulpinus. In the five dentitions available to us, the teeth of Alopias vulpinus exhibit considerable variation and characters not not- ed in the paleontological literature. The lower medial tooth var- ies in size and may be absent. The lower anterior teeth are 10% to 20% smaller than the upper anteriors. Among the upper ante- rior teeth, the third upper anterior (second from the symphysis) tooth is the largest. The tips of the intermediate teeth bend lin- gually rather than labially or being straight. In the lower lateral teeth, the attitudes of the crowns range from inclined with a straight distal cutting edge to hooked with a concave distal cut- ting edge. The tips of some of the upper lateral teeth, in lateral view, do not curve labially but rather curve slightly lingually. Mesial and/or distal cusplets may be present. Galeocerdo triqueter was established on the basis of an abraded lower anterior tooth. Eastman (1904:89) characterized the species as follows: "Teeth very robust, with elevated crowns, smaller and less twisted than those of G. contortus, and more faintly serrated along the coronal edges. Anterior margin only slightly arched, posterior notch inconspicuous." In examining the holotype of this species, ANSP 1214 (Figure 22f,g), we noticed that the faint serrations are confined to the basal portion of the distal cutting edge; the remainder of the cutting edge is smooth. We synonymize Galeocerdo triqueter with Alopias cf. A. vulpinus because the type specimen lacks the well-developed distal enamel shoulder characteristic of Ga- leocerdo and because there is a callosity on the labial face of the root that extends well onto the root. This tooth is almost identical to the first lower anterior tooth of Alopias vulpinus, and we believe it represents the same species as the specimens from Lee Creek Mine. The anterolateral teeth from Lee Creek Mine range in height from 0.8 to 1.5 cm (mean=1.2 cm, «=18). These sharks were probably 4.5 to 6 m TL. The vertebrae of this shark also are found at Lee Creek Mine. Their craniocaudal length is short (see Kozuch and Fitzgerald, 1989), and in side view they are septate, with the septa diverg- ing toward the rims of the vertebrae (Figure 22b). Their dorsal and ventral foramina vary in shape from oval to rectangular. These vertebrae are identical to those of a specimen of A. vulpi- nus (USNM 110242). Compagno (1984:233) reported that A. vulpinus inhabits coastal to oceanic temperate to tropical waters; the young are often found close inshore and in shallow bays. They feed main- ly on schooling fishes, such as mackerels and bluefish. Family CETORHINIDAE (basking sharks) Cetorhinus sp. FIGURE 22c-e Horizon.—Pungo River Formation (unit 3). REFERRED Material.—1 tooth, USNM 312269; 1 calcified clasper spine, USNM 467556. Remarks.—The single-cusped crown is nearly recumbent (Figure 22c,d); near the apex of the crown the labial face is al- most parallel to the basal face of the root. In cross section, the crown is triangular basally to oval apically. The labial face forms the short side of this triangular crown foot; the lingual faces are wider than the labial face and meet at an acute angle near the crown-root boundary. The root forms a pedestal for the crown. In basal view it has a triangular outline; a lingually placed central foramen occurs in a shallow transverse groove that bisects this face. The labial face protrudes somewhat, but it is flush with the crown; the la- bial end of the transverse groove divides this face. Both on the mesial face and the distal face of the root, three lateral canals open into a depression. The tooth measures 3.1 mm in total height, and the root is 3.3 mm in width, and 2.1 mm thick. In morphology, except for its strongly lingually directed crown, this tooth compares well with adult teeth of extant Ce- torhinus maximus, but it differs in its much smaller size, which is about half that of the extant species (compare with Herman et al., 1993, pl. 43). The strong lingual inclination of the crown is more characteristic of juvenile than of adult basking sharks (Herman et al., 1993:194-195), but unlike the teeth of the juve- niles, the crowns of the teeth from Lee Creek Mine are unoma- mented. Sufficient specimens of the extant species have not been examined to determine if ornamentation is a variable character in juveniles. In addition to the tooth, a calcified clasper spine (Figure 22e), was recovered from the spoil piles. This specimen is identical in morphology to those illustrated by Leriche (1926, fig. 195, pl. 37: figs. 6, 7). The living basking sharks are plankton feeders and inhabit boreal to warm-temperate coastal to pelagic waters (Compag- no, 1984:235). Along the Atlantic coast of North America they have been reported as far south as the Florida coast. Family Lamnidae (mackerel sharks, mako sharks, white sharks) There are three extant genera in this family, Carcharodon, Isurus, and Lamna. Of these, Lamna has the most primitive teeth1, which are very similar to those of Cretolamna appen- diculata. Lamna, however, has teeth with reduced, tapered lat- eral cusplets and root lobes that are not very lobate, and it has a first anterior tooth with root lobes that form an angle of 90° or 'in their phylogenetic analysis based on dental morphology, Long and Waggoner (1996) selected Mitsukurina as the outgroup. In comparison to Hy- bodus and Cretolamna, however, the teeth of Mitsukurina are very derived. Several of their synapomorphies, such as absence of diastem, short root lobes, multiple pairs of lateral cusplets, and moderate lateral cusplets, are plesio- morphic characters. 110 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY greater rather than the acute angle of those of Cretolamna. We identify these characters as autapomorphies for Lamna. For Carcharodon and Isurus, Compagno (1990a:372) identified as dental synapomorphies the characters "jaws and anterior teeth enlarged; lateral cusplets lost on teeth or present only in very young (?)," and as a synapomorphy of Isurus, he identified the flexed anterior teeth. This synapomorphy is absent in Isurus xi- phodon (see below). Only one of the synapomorphies Compag- no (1990a:372) identified for Lamna is found as a fossil: a cal- cified rostral node without lateral fenestra. In addition to these genera, we tentatively assign, as Kemp (1991) did, Parotodus to this family. The teeth of this genus exhibit characters that prompted Cappetta (1980) to assign it to the Otodontidae. Our comparisons lead us to assign Parotodus to the Lamnidae for the reasons discussed below. Cappetta (1980) included Parotodus in the Otodontidae be- cause of its very globular root, particularly in the lateral teeth, and because of the presence, in the Oligocene form, of lateral denticles. These characters, however, are not taxonomically useful. As in other lamnids, the globular roots are confined to the lower teeth where the toruses are well developed. Lateral cusplets, a primitive character, also occur in Lamna, the Odon- taspididae, the Alopiidae, Triaenodon (which have very broad lateral cusplets), and occasionally in the adult teeth of Isurus (Bass et al., 1975c:28) and Carcharodon. We did not find any derived characters to justify the assignment of Parotodus to the Otodontidae. At first we assigned Parotodus to the Alopiidae because al- though some individuals of living makos may have stocky teeth, their anterior teeth are more elongate than their lateral teeth. In our sample of P. benedenii, however, the heights of the largest lateral teeth (5.6-6.0 cm) and those of the anterior teeth (5.8-6.6 cm) overlap. This overlap in height, which does not occur in ex- tant Isurus or in Otodus, agrees with the observations on Alo- pias of Gruber and Compagno (1981:626): "Anterior teeth of threshers differ from lateral and posterior teeth in having nar- rower crowns relative to their height and more erect cusps, they are less well differentiated in Alopias than in lamnids, odontas- pidids, pseudocarchariids." Compagno (1990a:371) identified the reduced size of the anterior teeth as a synapomorphy of Alo- pias; this synapomorphy also pertains to Parotodus. Other characters, however, suggest that Parotodus is not an alopiid. While examining the Lee Creek Mine specimens, Compagno (pers. comm., 14 Apr 1993) noted that the morphol- ogies of the roots of Parotodus were like those of a mako rather than those of a thresher. We reexamined these teeth and those of Isurus and Alopias to verify his observation, and we found the following. (1) Unlike alopiids, in labial view, the lateral teeth of Parotodus do not have broad, shallow roots with enamel shoulders extending to the extremities of the root lobes. Like lamnids, in Parotodus the root lobes extend slightly be- yond the basal boundary of the crown foot, and the roots are deep. (2) Like lamnids and unlike alopiids, the lower anterior teeth are not 10% to 20% smaller in height than the upper oc- cluding teeth. (3) Like lamnids and unlike alopiids, the heights of the first three or four lateral teeth exceed their respective widths. These findings suggest to us that the less differentiated anterior teeth (see Compagno, 1990a:371) were derived inde- pendently of Alopias, and we assign Parotodus tentatively to the Lamnidae. Genus Parotodus Cappetta, 1980 Until 1980, paleontologists assigned the sole species of this genus, Parotodus benedenii, to the genus Isurus; in that year Cappetta erected a new genus for this species, Parotodus, which he characterized as having "very great thickness of the crown and a very globular root" (Cappetta, 1980:35). We re- vise his diagnosis to read as follows: upper and lower laterals with hooked, mako-like crowns; heights of first three or four lateral teeth exceed their respective tooth widths; anterior teeth not well differentiated from lateral teeth. Antunes and Jonet (1969-1970) and Antunes (1978) stated that the teeth identified as Isurus benedenii were the intermedi- ate teeth of Isurus hastalis. Cappetta (1980), however, correct- ly argued that the teeth of Parotodus benedenii represented dif- ferent jaw positions and that their thicknesses were proportionally greater than the intermediate teeth of I. hastalis. Despite their stocky crowns, the teeth of Parotodus bear some similarities to Isurus oxyrinchus. Like /. oxyrinchus, the lower anterior teeth have well-developed toruses and robust crowns, which in the Lamnidae are peculiar to Isurus oxyrin- chus. Also like /. oxyrinchus, the tips of the lower teeth, in lat- eral view, recurve labially. Unlike /. oxyrinchus, the sharply defined cutting edges on the anterior teeth extend to the crown foot. The roundness of the lingual crown foot, noticeable from the labial side in /. oxyrinchus, is not visible from this side. We do not believe, however, that these similarities warrant synony- mizing Parotodus with Isurus. Kuga (1985:14-16), apparently unaware of Cappetta's pa- per, erected the genus Uyenoa for this species, which he kept in the family Lamnidae; however, the senior name for this species is Parotodus. Parotodus benedenii (Le Hon, 1871) Figures 23,24 Isurus moniwaensis Hatai, Masuda, and Noda, 1974:19-20, pl. 2: figs. 20, 22 [Miocene, Japan]. Uyenoya benedenii Kuga, 1985:14-16 [Neogene, Japan]. Horizon.—Yorktown Formation (units 1, 2). Referred Material.—85 isolated teeth, USNM 24757, 279254, 279320, 281382, 282471, 283598, 289044, 289088, 289095, 289104, 289106, 293759, 293762,297466-297468, 302442, 312441, 312447, 324928, 421612, 421629, 421917-421921, 437885, 454539^154563, 457258-457285. NUMBER 90 111 Remarks.—Le Hon (1871:6) characterized this species as follows: "Species of enormous thickness that resembles the Cretaceous Oxyrhina crassidens of Dixon. The neck of the tooth is very wide and the gum imprint exceedingly pro- nounced as in all teeth with thick roots. The crown is incurved more or less, and some curved and hooked teeth having the same characters it seems to me to have belonged to the same animal" (translated from French by R.W.P.). The type speci- men that Le Hon illustrated is a second lower anterior tooth, which he noted was found in Pliocene sediments during the ex- cavation of the fortifications around Anvers, Belgium. Leriche (1910:281-283) reported the earliest occurrence of this species in the early Oligocene, Rupelian, of Belgium. These teeth are much smaller than those of the Pliocene form. In his suite of illustrated specimens, Leriche (1910) also in- cluded two teeth of Lamna rupeliensis (his pl. 16: figs. 5, 6), which he identified as Oxyrhina benedenii. Leriche (1910:282) noted that "the transformation of the heels of the crown into lateral cusplets is observed mainly in the lateral teeth of the up- per jaw (Fig. 5, 6, 8, 9)" (translated from French by R.W.P.). Although the teeth in his figs. 5 and 6 were not found in associ- ation with the others, his identification of them was not chal- lenged by subsequent workers. The sample of 85 teeth from Lee Creek Mine permitted us to reconstruct the dentition of this shark (Figure 23). As in lamni- form sharks, the upper and lower anterior teeth exhibit the same morphological characters that permit their differentiation: their crowns are more erect, and the angle formed by the root lobes is more acute than in the lateral teeth. In the lower anteri- or teeth there is a greater development of the torus. Only the second and third upper anterior teeth are present from Lee Creek Mine. The first anterior tooth may be absent in this genus; no teeth were recovered that were nearly symmetri- cal and with root lobes forming an acute angle. In the second anterior tooth, the crown inclines distally, the distal cutting edge is concave, the root lobes intersect at a right or slightly obtuse angle, and the distal root lobe has a greater mass than the mesial one. In the third upper anterior tooth, along the me- sial edge, the arc of the distal curvature is almost continuous from the tip of the crown to the base of the root. The distal cut- ting edge is nearly straight. The root lobes intersect at an ob- tuse angle, and as in the second anterior tooth, the distal root lobe has the greatest development. We identified three intermediate teeth: USNM 289088, 293762, and 312447 (Figure 24a-e). As in the anterior teeth, these teeth are much higher than they are wide, but they are smaller than the anterior and lateral teeth in our suite of speci- mens. Their root lobes form an acute to right angle, and as in the third upper anterior, their crowns have a distal curvature and, in lateral view, a very pronounced labial curvature. In the lower jaw, the anterior teeth possess toruses (lingual protusion of the root) (Figure 24/) with a much greater devel- opment than those of the lateral teeth. The first anterior tooth has an erect crown, with the root lobes forming an acute or right angle. In the second anterior tooth the erect crown has a slight distal curvature; the root lobes form a right or obtuse an- gle. In the third anterior tooth, which has a distally inclined crown, the mesial root lobe is longer than the distal one. In Parotodus, except for their size, the lateral teeth of both jaws have hooked crowns very similar to those of two denti- tions of an Isurus oxyrinchus in the Hubbell collection FIGURE 23.—Parotodus benedenii, composite dentition, lingual view. (Scale bar= 1.0 cm.) 112 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE 24.—Parotodus benedenii: a, USNM 289088, intermediate tooth, lingual view; b, USNM 293762, inter- mediate tooth, lingual view; c, USNM 312447, second lower intermediate tooth, lingual view; d, same specimen, lateral view; e, same specimen, labial view;/ USNM 457258, lower anterior tooth, lateral view showing devel- opment of torus. (Scale bars: a,b= 1.0 cm; c-e=0.9 cm;/=0.5 cm.) (D052188, MD62287). In both species the lower teeth are more erect than the upper teeth. Kemp (1991, pl. 32) noted and illustrated a specimen of this species consisting of 30 detached teeth in matrix from the ear- ly Miocene Batesford Limestone, Batesford, Australia, in- cluding the upper and lower anteriors and two symphysial teeth. The symphysial teeth appear to be from the upper and lower jaws; the tooth in his fig. El,2 is more compressed than the symphysial tooth in his fig. D; therefore, we concur with Kemp's identification of them as upper and lower symphysial teeth. We did not find any comparable teeth at Lee Creek Mine. The presence of symphysial teeth in the Batesford spec- NUMBER 90 113 imen may be the retention or reappearance of a primitive character. Davis (1888:26-27) described the species Oxyrhina von- haastii from the Oligocene of New Zealand and included in his type suite a mass of about 20 teeth in limestone. Although we have not seen his type suite, we believe that this species should be assigned to Parotodus, and it may be a junior synonym of P. benedenii. In the fall of 1992, Clyde Swindell and George Powell re- covered an associated, partial dentition of this shark from Lee Creek Mine (Kent and Powell, 1999). This dentition confirms the presence of two upper anterior teeth. Until this specimen can be compared with those from Australia and New Zealand, we believe it is premature to assess the relationship of this spe- cies to the Oligocene and early Miocene specimens and to oth- er lamnids. The anterior teeth of this shark range from 5.8 to 6.3 cm in height (mean=6.2 cm, «=6) and from 3.7 to 5.0 cm in width (mean=4.3 cm, «=6). Because their lateral and anterior teeth are not well differentiated by tooth height, there is no living lamnid that can be used as a model for estimating the total length of Parotodus benedenii, but we guess that large adults were between 6 and 7.5 m long. Hatai, Masuda, and Noda (1974:19-20, pl. 2: figs. 20, 22) described a new species, Isurus moniwaensis, from the Mi- ocene of Japan; their type specimen is identical to specimens identified herein as Parotodus benedenii. Roux and Geistodoerfer (1988) reported the occurrence of teeth of this species in the Pleistocene deposits in the Indian Ocean off New Caledonia (see "Carcharodon megalodon (Agassiz, 1835)," below, for further discussion). Compagno (pers. comm., 14 Apr 1993) suggested that this shark fed by grabbing prey, such as seabirds, porpoises, and seals, with its teeth and swallowing it whole. Genus Isurus Rafinesque, 1810 The teeth of Isurus have smooth cutting edges, have smooth labial and lingual coronal faces, usually lack lateral cusplets, and the central foramen usually does not open into a well-de- fined transverse groove. In both jaws the teeth are differentiat- ed into anteriors, intermediates (upper jaw only), laterals, and posteriors. Espinosa-Arrubarrena (1987:26) gave the formula for fossil and living makos as follows: (upper jaw) two anteri- ors, one intermediate, five to seven laterals, three to four pos- teriors; (lower jaw) three anteriors, five to seven laterals, three to four posteriors. Of these tooth types, the anteriors are the most important taxonomically; in the dentitions we have ex- amined so far, they exhibit the least amount of variation within a species. Ontogenetic heterodonty occurs in the extant species of Isurus. In individuals over 3 m TL, the teeth broaden and be- come thinner, even in the lower laterals, which normally have robust crowns, and the attitudes of the crowns may change considerably, becoming strongly arched distally. Espinosa- Arrubarrena (1987) noted that lateral cusplets occurred in the lateral and posterior teeth of juveniles, but in the available ju- venile dentitions of I. oxyrinchus (USNM 232652, TL=1901 mm?; USNM 232650, TL=1310 mm), Isurus paucus (USNM 196024, TL=1251 mm; USNM 196039, TL=1801 mm), and in numerous juvenile /. xiphodon teeth, we did not observe any lateral cusplets. Bass et al. (1975c:28), however, reported that in large individuals of I. oxyrinchus, minute lateral cus- plets occur on the distal shoulder of the more posterior lateral teeth. To date no sexual heterodonty has been noted. In the upper and lower jaws, differences in morphology ex- ist between teeth of the same jaw positions. The upper teeth have broader, thinner crowns, and their roots are not as well developed as are their counterparts in the lower jaw. In the lower teeth, in the area of the central foramen, the torus is well developed; the greatest development of the torus occurs in the lower first anterior tooth and diminishes in each succes- sive tooth until it is not discernible in the second or third later- al tooth. Espinosa-Arrubarrena (1987:110-117) recognized three groups of makos based on the elongate and cylindrical shape of the crowns of the anterior teeth and on the lengths of the root lobes of the upper anterior teeth. To facilitate discussion, we call his groups the oxyrinchus-paucus group, the desori group, and the hastalis group. He defined the oxyrinchus-paucus group as containing spe- cies with "small to medium-sized teeth, that have cylindrical to slightly flattened and very narrow crowns in the upper anterior teeth" (Espinosa-Arrubarrena, 1987:110-113). He subdivided this group into three lineages: the /. oxyrinchus lineage, the /. paucus lineage, and the /. sp. D-I. sp. E lineage. His /. oxyrin- chus lineage is characterized by having anterior teeth "with root branches that are unequally long [the mesial side is long- er]." His second subgroup, the /. paucus lineage, is character- ized by having "root branches that are equally long," and his third subgroup, the /. sp. D-I. sp. E lineage, is characterized by unequal root branches but with the distal branch longer than the mesial one (Espinosa-Arrubarrena, 1987:113). The desori group was defined as "isurid species with large and very robust teeth that represent an intermediate stage be- tween the narrow or cylindrical crowns of the uppers of group one [oxyrinchus-paucus group] and the wide and totally flat- tened upper teeth of the /. hastalis and /. planus lineages. The species included in this category are /. desori, I. retroflexus, and /. sp. F The upper anterior morphotypes in /. desori tend to be unequal (longer mesial side [root lobes]), and in /. retroflex- us and I. sp. F, the same elements tend to be equal in size" (Es- pinosa-Arrubarrena, 1987:113-114). Finally, his hastalis group was defined as "isurid species with wide crowns and totally labiolingually flattened upper teeth (triangular shaped crowns). The species groups of/. hastalis and /. planus (with /. sp. G and /. sp. H respectively) obviously represent the opposite end of the grasping-cutting se- 114 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY ries. These are teeth of very large size, the cylindrical (rounded cross section) shape of the crown has been completely lost. And the root branches of the upper anteriors are unequal in /. hastalis and very symmetrical (mesial and distal sides of the same size) in /. planus" (Espinosa-Arrubarrena, 1987:116-117). Espinosa-Arrubarrena and others have identi- fied as I. hastalis teeth that Agassiz named / xiphodon, which is not a junior synonym of I. hastalis (see below). When we applied these characters to recent and Lee Creek Mine specimens, we found the following exceptions to Espi- nosa-Arrubarrena 's groupings. (1) In a dentition from a 3.9 m TL female /. paucus (Hubbell collection, JG5379), the third up- per left anterior tooth has a mesial root lobe that is longer than the distal one. In another dentition from a 4 m TL female of the same species (Hubbell collection, NA91690), in all of the teeth except the upper intermediate and the third through fifth upper laterals, the mesial root lobes are longer than the distal root lobes. (2) In a dentition of a large /. oxyrinchus of over 3.7 m TL (USNM 309253), the second upper anterior tooth has a dis- tal root lobe that is longer than the mesial one. (3) In dentitions of/, paucus from individuals of over 3.7 m TL, the anterior teeth become broader and flatter, losing their cylindrical or rounded cross section. Compared to /. paucus, in /. oxyrinchus the compression of the anterior teeth occurs to a much lesser degree in some individuals (USNM 309253), and no compres- sion occurs in others (LJVC 901119). (4) The teeth in the type suite for I. desori (Agassiz, 1843:282), in size and roundness of cross section, can only be assigned to Espinosa-Arrubarrena's oxyrinchus-paucus group. (5) The broadness and flatness of the holotype of/, planus falls within the range of variation occur- ring in the extant /. oxyrinchus and /. paucus, particularly in in- dividuals of 3.7 m TL or more. In summary, the lengths of the root lobes and the roundness of the crowns of the anterior teeth in cross section are charac- ters that vary within a species. We believe, however, that there are three species of mako sharks represented at Lee Creek Mine. Those with more flex- uous, awl-like anterior teeth with a labial recurvature at the tips we identify as /. oxyrinchus; those with less flexuous an- terior teeth with more compressed crowns and with straight- tipped lower anterior teeth we identify as /. hastalis; and those with triangular, compressed upper anterior teeth with tips that may become labially recurved as they become larger and with straight-tipped lower anterior teeth we identify as /. xiphodon. Isurus oxyrinchus Rafinesque, 1810 Figures 25,26 Oxyrhina desori Agassiz, 1843:282, pl. 37: figs. 8, 9 [figs. 10-13 indetermi- nate; lectotype: ETHG1 P145, selected herein; Miocene, Switzerland],—Ler- iche, 1927:68, pl. 10: figs. 1-10 [Miocene, Switzerland]. Oxyrhina desori Gibbes, 1848-1849:203, pl. 27: figs. 169, 170 [Pliocene, South Carolina]. HORIZON.—Pungo River Formation (units 1-5); Yorktown Formation (units 1, 2). REFERRED MATERIAL.—115 teeth, USNM 207619, 207620, 207622, 207626, 207628-207630, 279119, 279136,279155, 293736, 312442, 312443, 312448, 312449, 312452, 336759, 339897, 421619, 421884, 421983, 425494, 425502, 425558, 425851, 425856, 425862-425872, 452469, 452481, 452483, 452487, 452488, 452491, 452494^52497, 452499^152501, 452505-452507, 452510, 452512, 452514, 452530, 452541-452543, 452550, 452563, 452565, 452566, 452572-452574, 452579, 452583, 452612, 452620, 453142, 453146, 453147, 453152, 453153, 453156, 453158, 453160^153162, 453164, 453170, 453174, 453175, 453184, 454280, 454281, 454283, 454286, 454302-454304, 454336, 454355, 454358, 454360, 454378, 454383, 454385, 454387^154389, 474966^*74975, 476296. REMARKS.—Agassiz (1843:282) characterized Oxyrhina desori as follows: "Relative to their height, the teeth of our Oxyrh. desori are much less broad than those of Oxyrh. hasta- lis; moreover they are thicker and nearly semicylindrical; and the cone of the tooth, instead of being straight, curves at first a little to the outside in order to turn back next to the inside; and when the tip, from its side, recurves in turn to the outside, the profile of the tooth takes on a very wavy appearance, which contrasts with the straight form and uniform bend to the outside of Oxyrh. hastalis" (translated from French by R.W.P.). Except for two, the syntypes are indeterminate, and Gibbes (1848-1849:203) related that Agassiz, subsequent to the pub- lication of his opus, felt that /. desori was identical to Lamna cuspidata. Gibbes, therefore, named a new species of fossil mako as O. desori to preserve Agassiz's name; the three teeth of his type suite are referable to Isurus oxyrinchus and /. xiph- odon (Gibbes, 1848-1849, pl. 27 figs. 169, 170, and fig. 171, respectively). Oxyrhina desori Gibbes, however, is a junior homonym of Oxyrhina desori Agassiz, and it is unavailable for use as a species name. Leriche (1910) also considered the syntypes of Agassiz's species to be indeterminate. Rather than abandon the name, Leriche (1910, 1927) referred his own specimens to "Oxyrhi- na desori" as identified by Sismonda (1849), which is pre- sumably, but not necessarily, also referable to Oxyrhina des- ori sensu stricto. Leriche's concept of this species, which is based on teeth that are similar to those of/, paucus, has pre- vailed until now even though, according to the rules of zoo- logical nomenclature, his redefinition of the species was in- valid. Had Agassiz and Leriche been correct about the type suite of O. desori, this species would be a nomen dubium and would be unavailable as a species name; however, one of Agassiz's syntypes is a diagnostic anterior tooth of Isurus. We obtained casts of three of Agassiz's (1843) syntypes, the spec- imens figured in his pl. 37: figs. 8-10. Of the three, the tooth NUMBER 90 115 »* " ^^^!?^| % «U*"-* ^^Sm iu?^*^^^| % ^ »^1 ¦ ^1 V 4 ^ \ * A 1 1 ^* v \ ~ ol A v ^ \ ' L i / fc J A Ij /-'-*,.. ^ * ^ • # i 1 l^JM 1 a — #? ^ r I k . A 1 '•-.« r p 1 1 Vt «f/i ¦ \b), and one of these, USNM 457247, has the appearance of buttocks, with a broad transverse groove. Another lower tooth, USNM 457251, the root of which is not swollen, has a deep transverse groove that is offset from the center of the lingual face of the root and which extends to the basal margin of the root (Figure 28/). Although the teeth of this species have been recovered from the Calvert Formation (middle Miocene), we did not find them in the Pungo River Formation. We suspect that this may be due to the absence or rarity of pinnipeds and cetotheres in the Pun- go River sea. In Figure 3lc-e, several examples of pathologic teeth are shown. One of them, USNM 457246, exhibits the same con- striction of the crown that Uyeno and Hasegawa (1974:258) used to diagnose a new species, Carcharodon akitaensis. Caus- es of tooth pathologies in mako sharks are not yet known. 122 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY v&r-"*-% FIGURE 31— Isurus xiphodon: a, USNM 278799, lingual view of right upper lateral tooth with swollen root; b, USNM 457247, lower lateral tooth with swollen root, lingual view; c, USNM 457246, pathologic upper lateral tooth, lingual view; d, USNM 474918, pathologic upper lateral tooth, lingual view; e, USNM 474919, pathologic upper lateral tooth, lingual view. (Scale bars= 1.0 cm.) NUMBER 90 123 The anterior teeth of this species range from 2.5 to 8.1 cm in height (mean=4.1 cm, «=46) and from 1.7 to 5.5 cm in width (mean=3.2 cm, «=46). We thought we could use the relationship between tooth size and total length in Carcharodon carcharias to estimate the size of/, xiphodon. In the extant species, however, we discovered that in individuals of nearly the same size, the anterior teeth of Isurus are greater in height than those of Carcharodon; the teeth of a 3.96 m TL /. oxyrinchus (Compagno collection, LJVC-901119) are equivalent in height to those of a 5.2 m TL Carcharodon carcharias (Hubbell collection, XI1384A). Therefore, on the basis of these data, we estimate that /. xiph- odon at Lee Creek attained 6 to 7.6 m TL. In one specimen of Callophoca obscura Van Beneden from the Yorktown Formation of Lee Creek Mine, USNM 467592, the distal portion of a right humerus exhibits long, straight, un- serrated bite marks (Figure 31 b), inflicted most likely by /. xi- phodon. Could this shark have had a dietary preference for warm-water phocid seals? Lamna sp. FIGURE 32 HORIZON.—Yorktown Formation (?unit 1). Referred Material.—6 rostral nodes, USNM 474994^174999. REMARKS.—Six rostral nodes collected from the spoil piles at Lee Creek Mine compare favorably with those of Lamna dit- ropis (Matsubara, 1955:115, fig. 15A-C; Compagno, 1988:69, fig. 7.1 A) and L. nasus (Parker, 1887, pl. 4: fig. 4, pl. 5: fig. 11; Garman, 1913, pl. 62). Concerning the Lee Creek Mine speci- mens, Compagno (pers. comm., 1982) stated that the calcifica- tion of these rostra is more characteristic of Lamna ditropis than of L. nasus, and that L. nasus has rostral cartilage less ex- tremely developed than that of L. ditropis. Figure 32a shows the hypercalcified development of the rostrum in L. ditropis (USNM 313874); comparable specimens of L. nasus were not available to us. Therefore, we are reluctant to make a specific identification of these specimens at this time. In more than 20 years of collecting at Lee Creek Mine, we have seen no teeth that can be identified definitely as Lamna, and in this time only six rostral nodes were found. Thus, this species is one of the rarest to occur here. In comparison to carcharhiniform calcified rostra (Figure 61a-a"), the only other group with tripodal rostra (Compagno, 1988:48), those of Lamna have dorsolateral rostral cartilages that, in dorsal view, attach to the anterior rostral node at a very acute angle rather than at a wide angle as in those of the car- charhiniform sharks, and on the dorsal surface of the rostral node, there is no median groove or fossa. These rostral nodes range in size (lateral width) from 1.9 to 3.6 cm. The extant species of Lamna inhabit shallow to epipelagic, cold waters and feed on schooling fishes (Compagno, 1984:247-249). Genus Carcharodon Smith in Miiller and Henle, 1838 Carcharocles Jordan and Hannibal, 1923:56. Procarcharodon Casier, 1960:13. Palaeocarcharodon Casier, 1960:13. Megaselachus Glikman, 1964:231. Until recently, the genus Carcharodon included C. carchar- ias, the extant great white shark; C. orientalis, a small Pale- ocene species; and the group of very large white sharks, the gi- ant-toothed species C. auriculatus, C. angustidens, and C. megalodon. The apparent absence of small-toothed species of Carcharodon from the end of the Paleocene to the middle Mi- ocene suggested to Casier (1960) that the giant-toothed species and the extant species had separate origins in the Tertiary, and the origin of the extant species from C. megalodon did not seem plausible. He concluded that the C. auriculatus-C. mega- lodon line was not related to C carcharias; therefore, he erect- ed the genus Procarcharodon (type species Carcharodon an- gustidens Agassiz, 1835, Oligocene of Europe) for these large sharks with finely serrated teeth and the genus Palaeocarchar- odon for the Paleocene species with coarsely serrated teeth. Casier (1960:13) characterized Procarcharodon as follows: "Teeth large and broad slightly compressed, with margins gen- erally regularly serrated, sometimes pectinate; with denticles present in the Eocene and the Oligocene, disappearing as a rule in later forms. Root well developed." His reason, however, for erecting these genera, the absence of small-toothed species in the middle Tertiary, was based on an inadequately sampled and biased fossil record. In his scenario for the evolution of these forms, Casier con- sidered the two most important characters to be the loss of lat- eral denticles and the appearance of marginal serrations. He stated (1960:15), "In their evolution one is assisted thus by the appearance, several times, of [these] two dental characters fol- lowing in succession__1. In the branch of Palaeocarcharo- don, where it [the appearance of marginal serrations] does not accompany the disappearance, nor even reduction of denticles; 2. in those of Procarcharodon, where it accompanies the loss of denticles, but only secondarily (the appearance of marginal serrations has preceded, in this case, the disappearance of later- al denticles); [and] 3. in the branch of Carcharodon s. str., where the appearance of marginal serrations has, on the con- trary, been preceded by the loss of lateral denticles (already lost even in the ancestral form: Oxyrhina hastalis)." Casier believed that two species, Otodus obliquus subserra- tus Agassiz and Isurus hastalis escheri (Agassiz), confirmed this hypothesis. The teeth of these species exhibit incipient marginal serrations, and both occur prior to the appearance in the fossil record of C. auriculatus and C. carcharias. Casier surmised, therefore, that these earlier species represented tran- sitional forms, O. o. subserratus to Carcharodon auriculatus and /. h. escheri to C. carcharias. Both of Casier's characters, however, occur widely in the ga- leomorph sharks. Applegate (1967:49) and Compagno (1988:28) reported that serrations have evolved several times in sharks (Paleozoic cladodonts, hexanchoids, squaloids, lam- 124 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE 32.—Lamna ditropis: a, USNM 313874, extant species with calcified rostral node. Lamna sp.: b, USNM 474998, rostral node, lateral view; c, same specimen, ventral view. (Scale bars: a=2.0 cm; b,c=\.0 cm.) NUMBER 90 125 noids, and carcharhinoids), and Compagno suggested that they may have evolved at least three times in the Carcharhiniformes. Likewise, among the lamnoid and carcharhinoid sharks, the loss of lateral cusplets represents a general evolutionary trend. Therefore, neither character can be used with confidence for developing phylogenetic scenarios. In 1969 Janvier and Welcomme reviewed the phylogeny of these sharks. They restricted the definition of Procarcharodon, with no documentation for this change, to include only the Eocene Carcharodon angustidens, but they placed the Oli- gocene form of this species back in the genus Carcharodon. Casier's concept of the genus Procarcharodon was chal- lenged by Glikman (1964), who found a greater resemblance between Otodus obliquus Agassiz and Carcharodon auricula- tus and C. angustidens (teeth with lateral cusplets) than be- tween the latter two species and C. megalodon. Glikman (1964:231) proposed to include C. auriculatus and C. angus- tidens in Otodus and characterized the teeth as "smooth mar- gined in Paleocene forms, sometimes serrate on the upper jaw of early Eocene forms and serrate on both jaws from the middle Eocene onwards. Neck well developed in the anterior and ante- rolateral teeth only. One pair of large accessory denticles." For teeth without lateral cusplets, he created the genus Megasela- chus, type species C. megalodon Agassiz, characterized by having "all teeth with serrate crowns; neck well developed; [and] most teeth without accessory denticles" (Glikman, 1964:231). Well-developed neck areas also occur in the teeth of Cretoxrhina mantelli, Parotodus benedenii, Carcharhinus leucas, C. longimanus, Carcharodon orientalis, C. auriculatus, C. angustidens, C subauriculatus, and, in a lesser degree of development, in the juvenile teeth of C. carcharias (see be- low). Accessory denticles occur in the juvenile teeth of C. meg- alodon and C. carcharias and may be retained in the teeth of adults. Glikman's characters are not useful for establishing the generic identity of these sharks. We believe Glikman's inclusion of Carcharodon auricula- tus (sensu lato) and C. angustidens in Otodus is unjustified. From a large suite of Otodus obliquus teeth from Morocco, one of us (R.W.P.) reconstructed a dentition of this shark and identified three upper anterior tooth positions, as opposed to two in Carcharodon. Also, unlike Carcharodon (including the giant-toothed species), the third lower anterior tooth points distally. A cast of a newly discovered associated denti- tion of this shark (Hubbell collection, uncataloged) from the Moroccan Eocene confirmed the presence of the first upper anterior tooth. In this dentition, one of us (R.W.P.) identified an intermediate tooth that points distally. This orientation is opposite to that found in Carcharodon, including the giant- toothed line. Rather than decreasing perceptibly in size, step fashion, as in Carcharodon, the second through sixth upper lateral teeth have similar heights, as they do in Lamna. Final- ly, the second lower anterior tooth is the largest tooth in the dentition. These characters of Otodus suggest that there is a greater morphological difference between teeth of Otodus and Carcharodon than has been assumed by previous workers, and that Otodus is not the ancestor of Carcharodon. Jordan and Hannibal (1923:56) created the genus Carcharo- cles with Carcharodon auriculatus (Blainville) as the type spe- cies, characterizing it as follows: "[Teeth] similar to Carcharo- don, but with a strong denticle on each side on the base of the tooth. Teeth narrower and more erect than in Carcharodon, their edges finely serrated." These characters also occur in an associated dentition that we identify as Carcharodon subauric- ulatus and in some juvenile teeth of C megalodon. Until re- cently, Jordan and Hannibal's genus was largely assumed to be a junior synonym of Carcharodon. Finally, Cappetta (1987) synonymized Procarcharodon Casier (1960) and Megaselachus Glikman (1964) with Car- charocles Jordan and Hannibal (1923), which of these three is the senior name. In two sets of associated teeth from the Rupelian of Bel- gium, Siverson (1989) observed two additional characters for separating Carcharocles from Carcharodon: the absence of dignathic heterodonty and the absence of an intermediate tooth. Dignathic heterodonty, however, is present in Carchar- odon auriculatus (see Dockery and Manning, 1986, pis. 2, 3), and it is very marked in the associated tooth sets from Lee Creek Mine (see below), which are from young individuals of Carcharodon subauriculatus. Like C. carcharias, dignathic heterodonty in the giant-toothed line becomes less marked in older individuals. In extant Carcharodon the intermediate teeth are highly variable in size and morphology (morphological variation is particularly evident in juveniles); they can range in height from about one-half to almost equal the height of the teeth in the adjacent tooth positions. Agassiz (1835, pl. F, 1843:91) and Leriche (1910:287, footnote 3, fig. 89) figured and de- scribed dentitions of C. carcharias with intermediate teeth the same size or almost the same size as the teeth of adjacent files. The form of the crown may range from being like that of a lower lateral tooth to like that of a broad, upper anterior tooth. Two characters, however, are consistent: the distal cut- ting edge is longer than the mesial one, causing the tip of the tooth to be erect or to point mesially rather than distally as in the lateral teeth, and the tooth has a strong labial curvature, stronger than any other tooth in the jaw. Uyeno et al.'s (1989, fig. 4) associated dentition shows the crown orientations of the teeth from the upper right jaw. Their tooth number three, which also is the position in the jaw of the intermediate tooth, has a tip that bends mesially. In their pl. 4, it also has a strong labial curvature. This tooth is undoubtedly an intermediate tooth (Gottfried et al., 1996, independently identified this character of the intermediate tooth as a synapomorphy for the genus Carcharodon). Also, in the associated tooth sets of C. subauriculatus, based on the above characters, we identified teeth that we are certain are intermediates. In detached tooth sets, these teeth may be mistaken for lower lateral teeth. In the Belgian specimens, the intermediate teeth may be absent or misidentified. 126 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Several researchers have expressed opinions contrary to those of the above workers, synonymizing the genera of Casier, Glikman, Jordan, and Hannibal with Carcharodon. Keyes (1972:239-240) disputed Casier's erection of Procarcharodon for the large-toothed species of these sharks and the derivation of C. carcharias from Isurus hastalis, but he did not document his position. Welton and Zinsmeister (1980:7) also questioned Casier's interpretations: "We do not find Casier's interpretations con- vincing, and we agree with Keyes (1972) that the similarities between / hastalis and Carcharodon (in the restricted sense of Casier 1960) might be convergence rather than phylogenetic relationship. The phylogeny of Carcharodon is best interpreted through detailed analysis of tooth morphology and dental for- mulae using specimens collected only under conditions of de- monstrable superpositional control. A detailed study of ontoge- netic heterodonty in the living C. carcharias and Isurus spp. would be extremely useful in resolving these problems." In 1983 Bendix-Almgreen stated that the structure of the coronoi'n fibrous architecture of Carcharodon carcharias and C. megalodon was too similar to be of divergent origins, and on the basis of this character, he synonymized Procarcharodon with Carcharodon. He did not, however, compare the coronoi'n of these sharks with that in the teeth of other lamnoids, and the usefulness of the coronoi'n fibrous architecture as a taxonomic tool for neoselachians has yet to be demonstrated. Uyeno and Sakamoto (1984:52) noted the similarity between the juvenile teeth, particularly the serrations, of Carcharodon megalodon from the middle Miocene sediments in the Chichi- bu Basin, Japan, and those of C. carcharias. Although they did not elaborate on them, they found that the characters used to identify Procarcharodon also occur in Carcharodon carchar- ias. They reiterated Welton and Zinsmeister's (1980) call for an extensive study of recent and fossil Carcharodon teeth to help to resolve the problem. In their report on Carcharodon auriculatus, Dockery and Manning (1986:16) stated that they did not use Procarcharo- don "because it appears to have been erected on the basis of the shared primitive characters of its component species," adding that "despite its huge size, considered a unique derived charac- ter here, C. megalodon shares with C. carcharias a reduction of the accessory cusps that are so prominent in C. auriculatus." But, the loss of lateral cusplets has occurred in other lamniform and carcharhiniform sharks earlier in their histories; therefore, Dockery and Manning's shared character for these two species, the loss of lateral cusplets, cannot be considered to be unique to them. Uyeno et al. (1989:83) reported on an associated dentition of Carcharodon megalodon, consisting principally of upper teeth, from the middle Miocene sediments of the Saitama Prefecture, Japan, and remarked, "In examining the teeth rows contained in the male great white shark (full length 3.9 meters), it can be seen that the teeth closely resemble C. megalodon." They did not illustrate, however, the teeth of the extant great white shark. The foregoing shows the diversity of opinion that exists about the taxonomy of the extinct great white sharks. Of these opinions, we agree with Welton and Zinsmeister (1980) about the need for an extensive study of the dentitions of the living and extinct species. The abundance of Carcharodon megal- odon at the Lee Creek Mine and Gordon Hubbell's excellent private collection of C. carcharias dentitions, including as well those of other lamnids, facilitated our study of this problem. Hubbell's collection and the collections of the CAS and the NMNH permitted us to study dental variation in 35 complete dentitions, ranging from neonates to very large adults, and in 28 partial dentitions consisting of anterior teeth. Although we would be more confident with larger samples of the living and fossil species, we believe that the specimens presently avail- able to us in these collections provide sufficient evidence to suggest that the genera of Casier, Glikman, and Jordan and Hannibal are indeed junior synonyms of Carcharodon. In the fossil and the extant species of Carcharodon, the juve- nile and adult teeth of both lines share similarities suggesting that they belong to a single genus. First, unlike other galeomor- ph sharks, in both lines of Carcharodon the cutting edges of ju- venile teeth are coarsely serrated, and they become finer in the larger replacement teeth. Second, unlike Isurus (including /. xi- phodon), in a juvenile C. carcharias (CAS 40905), as in C. megalodon, the upper anterior teeth have on their lingual crown feet a chevron-shaped neck area (Figure 33a). This character, which also was observed in juvenile dentitions of C. carcharias in the Hubbell collection, is lost in later generations of teeth. Finally, the juvenile teeth of C. carcharias, with their coarse serrations and lateral cusplets, resemble closely those of C. ori- entalis (Figure 33ft). These similarities are not found in Isurus and Otodus (see above). In addition to these similarities, these two lines of Carcharo- don share three derived characters that are not found in any other lamnids, including two associated dentitions of Isurus xi- phodon. (1) The mesial cutting edge of the first tooth from the symphysis in the upper jaw, a second upper anterior, rather than being convex for 50% or more of its length, is straight or nearly so from the lateral extremity of the crown foot to the apex of the crown or to within one-eighth of its total length of the apex (it may be convex in this area). This straightness is unique to Carcharodon. (2) This second upper anterior tooth, rather than the second lower anterior, as in other lamnids, is the largest tooth in the dentition. (3) The intermediate tooth has a distal cutting edge that is longer than the mesial cutting edge, causing the tip to be erect or to point mesially rather than point- ing distally. Thus, the tooth has a reversed appearance (Got- tfried et al., 1996) that is unlike all other lamnoid intermediate teeth, which have longer mesial cutting edges and distally pointing crowns. These three characters were present in the as- sociated fossil dentitions with complete sets of anterior teeth available to us, one of C. subauriculatus and two of C. megal- odon, and in Uyeno et al.'s (1989, fig. 4) associated dentition of C. megalodon. NUMBER 90 127 fs^jf »^,Jf W ^^^ \ 3 \^J ^y^ ^jap ^? %|r b — FIGURE 33.—Carcharodon carcharias: a, CAS 40905, dentition of juvenile. Carcharodon orientalis: b. USNM 412184, Aquia Formation, Prince Georges County, Maryland, upper anterior tooth, lingual view. (Scale bars=1.0 cm.) In view of the above, C. megalodon and C. carcharias and their respective Tertiary predecessors, including C. orientalis, share more morphologic characters in common with one anoth- er than with any other lamnoid shark. Consequently, we be- lieve that Palaeocarcharodon, Procarcharodon, Megasela- chus, and Carcharocles are junior synonyms of Carcharodon. The first upper anterior tooth, present in Carcharias taurus, is absent from the associated dentitions of Carcharodon subau- riculatus and C. megalodon from Lee Creek Mine, the associ- ated dentition of C. megalodon reported on by Uyeno et al. (1989), and another of this species, from the Bone Valley For- mation, in the Hubbell collection. In both the extant and the fossil species, a wide range of den- tal variation exists. In the anterior teeth, the first upper tooth (the second anterior) is predominantly symmetrical, but those of juveniles and sometimes those of adults are often slightly asymmetrical, and in the extant species, this asymmetry existed in 22 (35%) of the 63 dentitions that we examined2. These asymmetrical teeth are prevalent in white sharks with narrow 2Purdy (1996:69) identified the symmetry of this tooth as a synapomor- phy, but in view of its variability in the living species, we think this is a weak character. teeth. Because these asymmetrical teeth exhibited morphologi- cal characters found in the second upper anterior teeth of Car- charias taurus, we disagree with Applegate and Espinosa-Ar- rubarrena's (1996) assertion that this tooth is the first and not the second tooth position, which they stated is missing in Car- charodon. Our position that the first upper anterior tooth is ac- tually a second upper anterior agrees with Compagno's (1990a) conclusion. The intermediate teeth of both the giant-toothed and small- toothed Carcharodon may have slightly concave or straight distal cutting edges, with the mesial ones being convex or straight apically and concave or straight basally. In individuals of great size, the cutting edges may become nearly equal in length. In the extant species, tooth width varies widely. For exam- ple, six second upper anterior teeth measuring 5.8 cm in height had widths ranging from 3.8 to 5.0 cm, and a 6.1 cm tooth had a width of 3.6 cm. These variations in width occur in both males and females, and they are present in the teeth of Car- charodon megalodon. The number of central foramina in the teeth of C. subauricu- latus and C. megalodon varies from one large foramen to four smaller foramina. 128 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Compagno (1984:238) characterized the teeth of this genus as follows: "Flat, triangular, with broad, serrated, nearly straight cusps, and lateral cusplets only in juveniles below 2 m long (which may have at least some smooth-edged or partially smooth); intermediate teeth in upper jaw very large, over half height of upper anteriors." We would add that in early Tertiary species, lateral cusplets persist into adulthood. In the Lee Creek Mine fauna we recognize three species of Carcharodon: C. subauriculatus from the Pungo River Forma- tion, C. megalodon from the Yorktown Formation, and C car- charias from the Yorktown and James City formations. Carcharodon subauriculatus Agassiz, 1839 Figures 34-36 1 Carcharodon polygyrus Agassiz, 1838, pl. 30: figs. 9-12 [Miocene, Switzer- land]; 1843:253. Carcharodon auriculatus de Blainville.—Ameghino, 1906:181-182, fig. 48 [Miocene, Argentina]. ICarcharodon chubutensis Ameghino, 1906:183, fig. 49 [Miocene, Argentina]. Carcharodon megalodon (Agassiz).—Leriche, 1926:412-422, pis. 35, 36 Mi- ocene, Belgium]. Carcharodon megalodon var. chubutensis Ameghino.—Leriche, 1927: 80, pl. 12, pl. 13: figs. 1-3 [Miocene, Switzerland]. Horizon.—Pungo River Formation (units 1-5). Referred Material.—2 associated dentitions, USNM 299832, 411881; several hundred teeth, USNM 244350, 256331, 256333, 256334, 280557, 280564, 282356, 282457, 295331, 295339, 298362, 336370, 339920, 348132, 348185, 348186, 348201, 348206, 348210, 348228, 348230, 348237, 356965, 356968-356971, 356974. Remarks.—We distinguished this species from Carcharo- don megalodon by the presence of lateral cusplets on the ante- rior teeth of subadults and usually adults; these cusplets are not separated from the crown by a deep notch as they are in C. an- gustidens and C. auriculatus. Juvenile, anterior teeth of C. megalodon may possess well-defined lateral cusplets. Agassiz based Carcharodon subauriculatus on a suite of three anterior teeth, two of which (Agassiz, 1839, pl. 30a: figs. 11, 12) bear lateral cusplets. In the specimen in his fig. 11, the notches that separate the lateral cusplets from the crown have all but disappeared, and in the specimen in his fig. 12, a shallow notch separates the mesial lateral cusplet from the crown; the distal portion of the tooth is not shown. Agas- siz described three other species to which the Pungo River teeth might be referred: C. heterodon, C. polygrus, and C. megalotis. These species, however, are based on lateral teeth, which could belong to either C. subauriculatus or juvenile C. megalodon. Leriche (1926:420) synonymized the teeth of C. subauricu- latus with C. megalodon, and he assigned the early Miocene teeth with lateral cusplets to C. chubutensis Ameghino, which Ameghino (1906, fig. 49) based on a lateral tooth. Ameghino (1906, fig. 48) identified an anterior tooth as C. auriculatus, which Leriche (1926:420) synonymized with C. chubutensis. Both of these, however, are referable to C subauriculatus, the senior name. Peter J. Harmatuk recovered two sets of what appear to be associated teeth of this species, USNM 299832 (27 teeth) and USNM 411881 (106 teeth), concentrated in two separate small areas on the spoil piles. Both sets include the enameloid shells of partially formed teeth, and both sets possess morpho- logical peculiarities that suggest they represent associated dentitions. All of the teeth in USNM 299832 possess small "denticles," composed of osteodentine rather than the ortho- dentine and enameloid of cusplets, on the lateral extremities of the roots at the base of the crown. In USNM 411881, on the Figure 34.—Carcharodon subauriculatus. USNM 299832, Pungo River Formation, associated dentition from the right jaws, lingual view. (Scale bar= 1.0 cm.) NUMBER 90 129 FIGURE 35.—Carcharodon subauriculatus, USNM 4118 the left jaws, lingual view. (Scale bar= 1.0 cm.) II, Pungo River Formation, associated dentition from lingual face of the root lobes, a small slit or slits occur in ex- actly the same positions on each tooth. Some teeth possess a slit on one lobe only, which may be either the mesial or distal lobe. In both sets these teeth fit the sizes anticipated for the different tooth positions in a dentition of Carcharodon. These features suggest strongly to us that these tooth sets represent associated dentitions. Neither tooth set represents a complete dentition. In USNM 299832 (Figure 34), which consists of teeth from mainly the right side, the second upper anterior tooth, the upper interme- diate tooth, and the lower right anterior teeth are present. From the upper jaws, the right second through fifth lateral teeth are present; from the lower jaws, the first through third lateral teeth from both sides and the left fourth through fifth lateral teeth are present. In the second tooth set, USNM 411881 (Figure 35), the den- tition from the left jaw is nearly complete; only the second up- per anterior tooth and one upper and two lower posterior teeth are missing. From the right jaw, the dentition includes most of the upper lateral teeth, several lower lateral teeth, and an in- complete first anterior tooth. These tooth sets furnished us with information not available from isolated teeth. Like C. carcharias, the height of the sec- ond upper anterior tooth in USNM 299832 (8.9 cm) is greater than the height of the second lower anterior tooth (8.5 cm). In USNM 411881, which is from a slightly smaller individual (height of second lower anterior=6.8 cm), the second upper anterior tooth is missing; therefore, we were unable to make the same comparison for this specimen. Another similarity with C. carcharias is the nearly symmetrical second lower an- terior tooth in each tooth set (Figures 34, 35). In both tooth sets dignathic heterodonty is present. The up- per teeth are broader than the lowers, the lower anterior teeth and the anteriormost lateral teeth have better-developed torus- es than their upper-jaw counterparts, and the cutting edges of the lower teeth are more concave than those of the upper teeth. In both tooth sets, the third lower anterior tooth is neither reduced in size nor is it the shape of a lateral. In an associated dentition of C. megalodon from the Bone Valley Formation in the Hubbell collection, this condition also exists. Uyeno et al. (1989) noted, however, that the third lower anterior tooth in their associated dentition of C. megalodon from Japan was al- most the same size as the first lateral tooth, both teeth being only slightly smaller than the lower anterior teeth. At present, due to the very small number of associated tooth sets for these sharks, we cannot determine if the lack of lateralization of the third lower anterior tooth is a retention of a primitive charac- ter or a variable character in these sharks. The second upper anterior tooth (present only in USNM 299832), which is the first tooth from the symphysis, is slight- ly asymmetrical. Its mesial cutting edge is nearly straight, 130 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY whereas the distal cutting edge is slightly concave. In this specimen the second and third upper anteriors are very similar, having only minor differences in tooth height and width, crown width, shape of the cutting edges, and shape of the root. In USNM 411881 the crown is strongly inclined distally; its mesial cutting edge is strongly convex, and the distal one is concave. The intermediate teeth of both dentitions were easily identi- fied by their strong labial curvature. With concave cutting edges, both of these intermediate teeth resemble lower lateral teeth, but their crowns, which point mesially, are wider than those of the lower laterals of their respective dentitions. In both intermediate teeth the lateral cusplets are almost lost, and on the lingual face of the root, they have three central fo- ramina. The upper lateral teeth of USNM 299832 (five present) are broader than those of USNM 411881 (eight present) and have nearly straight cutting edges, whereas the upper lateral teeth of USNM 411881 have concave cutting edges. In USNM 299832 the crowns of the upper lateral teeth are more erect than are those of USNM 411881. The first upper lateral tooth has a stronger labial curvature than in the other upper lateral teeth. This tooth is distally inclined, and its mesial cutting edge is straight, whereas the distal one is strongly concave. Its mesial root lobe is usually more tapered than its distal one. Of the up- per lateral teeth, the second tooth usually has the greatest height in this series. Its crown is broader than that of the first lateral tooth, and the cutting edges are slightly concave to near- ly straight, giving the tooth a symmetrical appearance. The width of this tooth is usually 85% to 95% of its height. In the remaining lateral teeth, the crowns diminish perceptibly in size and have a slight distal inclination. The three lower anterior teeth have erect crowns. The mesial and distal cutting edges are slightly concave in the first and second teeth and are more so in the third, giving the crown of this tooth a narrower appearance than those of the first two. In USNM 299832 the distal root lobes of these anterior teeth are more rounded than the mesial lobes are, but in USNM 411881 the mesial and distal root lobes of the second tooth both are rounded, and in the third tooth, the mesial root lobe is tapered. In USNM 411881, the crown of this last tooth points straight up, but in USNM 299832, it has a slight mesial bend. On the lingual face, in the central portion of the root, the torus is prom- inent in USNM 299832 and forms a ridge concentric to the bas- al edge in USNM 411881. In both dentitions the width of the first lower lateral tooth slightly exceeds its height (Table 3), but in the Japanese (Uy- eno et al., 1989), Peruvian, and Bone Valley associated denti- tions, these teeth are higher than they are wide. Too few associ- ated dentitions of these sharks are available to show whether or not this difference between the Lee Creek Mine teeth and those in the Japanese, Peruvian, and Bone Valley dentitions is onto- genetic. In both sets the lateral cusplets are absent or are the least de- veloped in the anterior teeth and are the most developed in the more posterior lateral teeth. In USNM 411881 the three teeth in upper lateral file number three show the transition from well- developed lateral cusplets to almost none (Figure 36), which indicates that they are lost quickly. Among the various fossil faunas with Carcharodon megal- odon-iype teeth, teeth with lateral cusplets occur commonly in the Burdigalian, Aquitanian, and perhaps the Chattian stages. These teeth also are common in the Fairhaven Member of the Calvert Formation in Maryland and at the base of the Calvert Formation in Virginia. In the Chesapeake Group, the percent- age of such teeth drops sharply above Zone 4 (Shattuck, 1904) of the Calvert Formation. During the Chattian, Carcharodon subauriculatus may have had a nursery area in what is now South Carolina (Purdy, 1996, and in prep.). In a stratigraphically controlled sample recovered from the Chandler Bridge Formation (Sanders, 1980), juvenile teeth (m=98) of this shark are very common, with the anterior teeth ranging from 4.4 to 5.7 cm in height. Only five teeth from very large adults were recovered. Of these specimens, a first lower lateral tooth, the most anterior tooth in this size range found, measured 9.5 cm in height. This tooth is equivalent in size to those from the Yorktown Formation. These large teeth may be from females that came to the Chandler Bridge area, an area of subtropical waters during the Chattian, to bear their young. We have not seen any large teeth from the Pungo River For- mation. Anterior teeth from this area range in height from 1.8 Table 3.—Measurements (in cm) of associated dentitions of Carcharodon subauriculatus from Lee Creek Mine. (A2-A3=upper anterior teeth, I = intermediate tooth, Ll-L9=upper lateral teeth, al-a3 = lower anterior teeth, I l-l9=lower lateral teeth, B=incomplete tooth, M = tooth missing from dentition.) Specimen Upper lentition Lower dentition A2 A3 I LI L2 L3 L4 L5 L6 L7 L8 L9 al a2 a3 11 12 13 14 15 16 17 18 19 USNM 299832 height 8.9 8.8 7.1 8.5 8.3 7.3 6.2 5.2 M M M M B 8.5 8.2 6.1 6.6 M M M M M M M width 6.9 6.8 6,4 7.6 7.9 7.9 7.0 4.7 6.5 6.3 6.2 6.0 USNM 411881 height M 7.1 6.5 6.7 6.6 5.9 5.5 4.6 4.0 3.2 2.2 M 6.2 6.8 6.4 5.3 5.7 5.4 4.6 M M M M M width - 5.6 6.3 6.3 6.4 6.2 5.8 5.1 4.5 3.7 2.9 - 4.9 5.0 5.3 5.4 5.4 5.2 4.7 - - - - - NUMBER 90 131 FIGURE 36—Carcharodon subauriculatus, USNM 411881, lingual view of third upper lateral file showing loss of lateral cusplets. (Scale bar= 1.0 cm.) to 8.4 cm (mean=4.6 cm, «=56) and probably came from indi- viduals between 3 and 10 m in total length. In the extant Carcharodon carcharias, Klimley (1985) ob- served along the western coast of North America that adult fe- males give birth to pups south of Point Conception, California, and then move north to their feeding areas where there is an abundance of pinnipeds. The young white sharks, which feed principally on fish, move northward as they grow larger. Dur- ing the Neogene along the Atlantic coast of North America, Carcharodon subauriculatus may have pupped in the warm waters of the area of present-day Charleston, South Carolina, and as the juveniles increased in size, they may have moved northward to principal feeding areas where larger prey were probably more abundant. The distribution of Carcharodon sub- auriculatus in the Neogene of the Atlantic Coastal Plain, then, could have been governed by their size and the availability of prey. This might explain the absence of large C. subauriculatus in the Pungo River Formation. Carcharodon megalodon (Agassiz, 1835) Figures 37-42 Carcharodon akilaensis Uyeno and Hasegawa, 1974:257-260, figs. 1-3 [Mi- ocene, Japan]. Horizon.—Upper Pungo River Formation (units 4-6)?; Yorktown Formation (unit 1). Referred Material.— 1 associated dentition, NCSM 13073; several hundred isolated teeth, USNM 182108, 214947, 244350, 256331, 256333, 256334, 278515, 279338, 279353, 280509, 280557, 280564, 281392, 289087, 295331, 295339, 298362,298368, 299766, 336257, 336370, 339917, 348169, 348178, 348247, 348265, 348338, 348344, 348366, 348375, 350923, 350924, 355736, 355762, 355766, 355822, 355884, 355888,356965,356968-356975, 356981-357010, 445500, 474915, 474916, 475360, 476358. Remarks.—We distinguish these teeth from those of Car- charodon subauriculatus by the lack of lateral cusplets on sub- adult and adult teeth and often on the anterior teeth of juveniles. Agassiz (1835, pl. 29: figs. 1-6, 1843:247-248) founded this species on six teeth, three of which were from Malta; the prov- enance of the others is unknown. He characterized the species as follows: "Its general form is perceptibly equilateral....The marginal serrations are uniform on each edge of the tooth. The enamel overlaps scarcely the root at the limit of the latter; it is indented at an almost right angle on the internal face (fig. 2), while it is simply concave on the external face (fig. 3). The thickness of the tooth is not very considerable (fig. 2a); the in- ternal face is convex; the external face on the other hand is flat even a little concave. The root is very heavy; it forms alone more than a third of the total height of the tooth; in the other specimens these proportions are able to vary according to their position in the jaw" (Agassiz, 1843:247-248; translated from French by R.W.P.). Agassiz's (1835, pl. 29: figs. 2, 3) type specimen is a second upper anterior tooth preserved in the Staatliches Museum fur Naturkunde in Karlsruhe, Germany (TE-PLI 18) (Figure 39a"). Teeth satisfying this description occur in the upper layers of the Pungo River Formation, but they may be examples of adult C. subauriculatus that lack lateral cusplets. Some of the dark, worn teeth from the basal Yorktown Formation may be rede- posited from the Pungo River Formation or may be relicts from the hiatus. Other teeth from the lower Yorktown layer are beau- tifully preserved and are relatively light in color, and most cer- tainly they are contemporaneous. These large teeth are the most sought-after fossils in the Lee Creek Mine, and collecting expe- rience indicates that they are found only in the lower portion of the Yorktown Formation. The associated dentition (NCSM 13073) consists of the sec- ond and third anterior teeth from both upper jaws, a right inter- mediate tooth, the first through fourth lateral teeth from both upper jaws (Figure 37), and three incomplete lower lateral teeth. Another tooth under this number (13073.23) is a third an- terior tooth from a smaller individual. 132 The second anterior teeth (first tooth postion) are nearly symmetrical, the right more so than the left. Their cutting edges are convex, and there are four central foramina on the lingual faces of the roots. The third anterior teeth are asymmetrical; their mesial cut- ting edges are nearly straight but are convex apically, and their distal cutting edges are concave basally and convex apically. The crowns of these teeth have a labial curvature. There are three central foramina on the lingual faces of the roots. The mesial root lobes are rounded and are longer than the more lo- bate distal lobe. The intermediate tooth has the proportions (height/width) of an intermediate tooth rather than those of a lateral tooth; its crown bends labially, and its tip points mesially, but only slightly. The mesial cutting edge is nearly straight, whereas the distal one is concave. Four central foramina are present on the SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY lingual face of the root. The mesial root is rounded, whereas the distal root is somewhat lobate. The crowns of the lateral teeth are mainly asymmetrical with the exception of the tooth from the second position. On the lin- gual faces of the root lobes of these teeth, two to three central foramina are present. The mesial root lobes are usually round- ed, whereas the distal ones are lobate. The first lateral tooth has a nearly straight mesial cutting edge and a concave distal one. Its crown has a strong labial curvature. The second lateral teeth, which are the largest of the lateral series, have crowns that are nearly symmetrical, and the teeth as a whole are equilateral. The crown has a slight labial curvature. The third and fourth lateral teeth are similar to the first in general form. Only incomplete lower lateral teeth were present in this den- tition, and these had narrower crowns than did their upper FIGURE 37.—Carcharodon megalodon. NCSM 13073, associated upper dentition, lingual view. (Scale bar=5.0 cm.) Aaaaa FIGURE 38.—Carcharodon megalodon, composite lower dentition, lingual view. (Scale bar=5.0 cm.) NUMBER 90 FIGURE 39.—Carcharodon megalodon: a, USNM 445500, heart-shaped fourth or fifth upper lateral tooth, lin- gual view; b, USNM 474915, same type of tooth, lingual view; c, USNM 355822, third upper anterior tooth with noticeable torus, basal view; d, USNM 476358, cast of holotype (Staatliches Museum TE-PLI 18), lingual view. (Scale bars= 1.0 cm.) 134 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY counterparts. Their cutting edges were concave, and their roots were incomplete. These teeth were not measured. As points of reference, we used some 40 dentitions of the liv- ing species, mostly uncataloged, in the USNM, CAS, and Hub- bell collections; the two associated dentitions of C subauricu- latus (USNM 299832, 411881); and the associated dentition of C. megalodon from the Bone Valley Formation (Hubbell col- lection) to identify the lower teeth (Figure 38) among the iso- lated Lee Creek Mine teeth. The crowns of all lower teeth bend lingually, and in most lower teeth the mesial root lobe is point- ed, and the distal one is lobate. The prominence of the root torus, which may be round or ridge-like, decreases progressive- ly in more distal tooth positions. The first anterior teeth are symmetrical, with concave to nearly straight cutting edges. The root lobes form a right or slightly obtuse angle. The second anterior teeth are slightly asymmetrical, pointing distally, with concave cutting edges. The angle of the root lobes is slightly broader than that of the first anterior tooth. In the third anterior tooth, the tip of the crown points mesial- ly. Near the apex, the mesial cutting edge is more convex than the distal one. Below this point the cutting edges are straight to slightly concave. The angle of the root lobes is broader than that of the second anterior teeth. Among the lateral teeth, the first or second is the largest; the remaining teeth decrease in size and become more asymmetri- cal toward the angle of the jaws. The distal cutting edges are more concave than are the mesial ones. The root lobes form ob- tuse angles ranging from 114° to 140°. We observed some tooth variations in C. megalodon that we have also observed in C. carcharias. The upper teeth are trian- gular (ranging from equilateral to isosceles) in outline. In some upper lateral teeth of C. megalodon, the mesial root lobe may be lobate, but this is not its usual form, which, unlike the distal lobe, is slightly tapered. The mesial cutting edge is convex (USNM 445500, 474915) (Figure 39a,b), which gives the teeth a heart-shaped appearance. The labial faces of the crowns of these teeth are convex, and the crown tips bend lingually slightly. Aside from their slightly pathologic appearance, these teeth agree in morphology with those of C. carcharias from the fourth and fifth upper lateral files. The torus is well developed (Figure 39c) on the lingual face of the root in a second upper anterior tooth (USNM 355822). Pathologies occur in many of the teeth; examples of these are illustrated in Figure 40a-c. The tooth germs of the two teeth in Figure 40a,b, USNM 474916 and 339917, respectively, were injured on their distal edges; these injuries caused the teeth to buckle lingually and distally. In USNM 339917, which re- ceived the greater injury, the distal cutting edge is distorted. The tooth germ of the tooth in Figure 40c, USNM 475360, was injured near its apex, which distorted its distal cutting edge. Cadenat (1962) noted that these deformities are caused by inju- a FIGURE 40.—Carcharodon megalodon, pathologic teeth: a, USNM 474916, lingual view; b, USNM 339917, lingual view; c, USNM 475360, third upper anterior tooth, lingual view. (Scale bar= 1.0 cm.) NUMBER 90 135 ries to the tooth germs on the insides of the jaws, usually caused by stingray spines or sea catfish spines. Once injured these tooth germs continually produce deformed teeth. Uyeno and Hasegawa (1974) described Carcharodon aki- taensis on the basis of a pathologic tooth of C. megalodon. This type of pathology also occurs in Isurus xiphodon. Another specimen worth noting is a tooth of this species with bite marks on it (USNM 336257, Figure 41a). When viewed under magnification (Figure 4\b), it is clear that these bites were made by a serrated tooth. Is it evidence of a shark eating a shark or of a loose tooth becoming lodged in the prey during feeding and then being bitten as the shark bit off a chunk of its prey? In the USNM collections, the largest tooth (USNM 214947), a second upper anterior tooth from Lee Creek Mine, measures 15.0 cm (5.9 in) in height and probably came from a 15 m TL shark. Larger teeth have been found in the Yorktown Forma- tion at New Bern (15.6 cm, Harmatuk collection, uncataloged) and from the Cooper River, South Carolina (16.2 cm, Hubbell collection, uncataloged). The teeth from the Yorktown Forma- FlGURE 41— Carcharodon megalodon, USNM 336257: a, lingual view of bitten tooth; b, close-up of bite mark showing serrations. (Scale bars: a=1.0 cm; 6=0.3 cm.) 136 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOG NUMBER 90 137 FIGURE 42 (opposite).—Carcharodon megalodon, composite dentition, USNM, reconstruction by John G. Maisey, AMNM. From left, top row: Robert J. Emry, Curator, Department of Paleobiology; Victor G. Springer, Curator, Department of Vertebrate Zoology; Peter J. Harmatuk, Collaborator, Depart- ment of Paleobiology; Ian G. Maclntyre, former Chairperson, Department of Paleobiology. Middle row: Sue Nell Voss, former Writer/Editor, Department of Exhibits; Clayton E. Ray, Curator, retired, Department of Paleobiology. Bottom row: Walter Erick Hock, Sr., former Modelmaker, Exhibits Central; Robert Purdy, Museum Specialist, Department of Paleobiology (all NMNH). (Photograph by Chip Clark, NMNH.) tion at Lee Creek Mine range in size from 5.6 to 15.0 cm (mean= 11.2 cm, n=25). A reconstruction of the jaws of C. megalodon (Figure 42) in- cludes as functional teeth specimens from Lee Creek Mine. The tooth in the left second lower anterior position is a left second upper anterior tooth. Several other teeth also are misplaced, in- cluding the right second lower anterior tooth, which is a right third upper anterior tooth; the right second lower and left third lateral teeth, which are upper intermediate teeth; and the right and left third lower anterior teeth, which are left third upper an- terior teeth from a smaller individual. Gottfried et al. (1996) published a corrected dental reconstruction of this shark. The earliest records of this species were reported from New Zealand, from the Oligocene sediments at Weka Pass (Davis, 1888:13) and from the early Oligocene sediments (Keyes, 1972:233). Keyes (1972:234) suggested that Davis' Weka Pass specimen was collected in the Pliocene "Greta Beds" that also occur at this locality. Many of the Oligocene specimens report- ed by Keyes were collected without precise stratigraphic data, and some of these are incomplete specimens, which in several instances lack the lateral cusplet area of the tooth; therefore, their stratigraphic position and in some instances their taxo- nomic identity cannot be ascertained. One specimen, however, with matrix adhering to it (Keyes, 1972: 235, figs. 10, 11), con- tains an imprint of a Venericardia, which Keyes matched with the sediments of the Chatton Marine Formation (middle Oli- gocene). This tooth, which is 12.0 cm in height, lacks the distal portion of the root and the mesial lateral cusplet, but we believe that it is from an adult C. subauriculatus and that other speci- mens from the Oligocene and early Miocene identified as C. megalodon also are C. subauriculatus. Roux and Geistodoerfer (1988:137) reported manganese-en- crusted specimens in Pleistocene sediments dating from 20,000 to 120,000 years ago dredged in the Indian Ocean off Madagas- car that may be the latest occurrence of C. megalodon. Based on the rates of accretion of manganese calculated by Burnett and Piper (1977), they estimated that the minimum age for these teeth is 10,000 years. Seret (1987) also reported dredged specimens from off New Caledonia in the southwestern Pacific Ocean; he estimated the age of these specimens at 1 Ma. Whether or not these specimens were continually exposed to conditions that permitted the accretion of manganese at a con- stant rate is not known, and they could be as old as Miocene or Pliocene. Several large cetacean bones with large, serrated bite marks, collected at Lee Creek Mine, confirm that this shark preyed and/or scavenged on large whales (see Purdy, 1996). Compagno (1990b:57) hypothesized that Carcharodon meg- alodon "may have been capable of preying on large baleen whales without the cooperative pack-hunting tactics that the smaller killer whale apparently needs to use to subdue difficult prey....Various reconstructions of the jaws of C. megal- odon... suggest that this shark had a predatory apparatus capa- ble of inflicting mortal injuries on even a fin whale or blue whale." Carcharodon carcharias (Linnaeus, 1758) Figures 33a, 43 HORIZON.—Yorktown Formation? (unit 1); James City For- mation. Referred Material.—22 teeth, USNM 214465, 214466, 256337,279304, 279311, 280579, 281051, 281055, 281059, 285623, 476340. Remarks.—These teeth are more compressed than are those of Carcharodon megalodon of equivalent size. The teeth of C. carcharias also have less massive roots and often have coarser serrations (Figure 43). Teeth of the extant species exhibit considerable variation in their morphology. The serrations of the cutting edges range from fine, as in Carcharodon megalodon, to coarse, with the finer serrations being more evident in the teeth of adults. In the upper anterior teeth, the tips of the second anterior teeth may be symmetrical, point distally, or (sometimes) point mesially. The third anterior tooth also may be symmetrical or nearly so. In the lower anterior teeth, the root lobes may arch labially. The intermediate teeth may have symmetrical or asymmetrical crowns, and in the same dentition, the morphology of the right and left intermediate teeth can be very dissimiliar. Like the in- termediate teeth, the upper first lateral tooth may have a strong labial curvature. The attitudes of the crowns of the upper later- al teeth may be erect or distally inclined. In the lower lateral teeth the extremities of the root lobes may be straight or rounded. Three juvenile dentitions of the extant species available to us for this study provided further information about variation in these teeth. In CAS 40905 (female, 1.6 m TL; Figure 33a) and in dentitions from a female (1.2 m TL) and a male (1.4 m TL) from the Hubbell collection, the upper teeth, rather than being erect, are inclined distally. The cutting edges of these teeth range from smooth near the tip to coarsely serrated basally. The cutting edges of the lateral cusplets range from smooth to coarsely serrated, particularly in the posterior lateral teeth. In all three dentitions the second upper anterior teeth are symmetrical; in the 1.2 m female, the distal root lobe tapers to a point and the mesial one is lobate, and in CAS 40905, the opposite condition exists. Unlike teeth of adults, the second lower anterior tooth is equivalent to or larger than the second upper anterior tooth 138 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Figure 43.—Carcharodon carcharias: a, USNM 285623, Yorktown Formation?, upper lateral tooth, lingual view; b, USNM 279311, Yorktown Formation?, upper lateral tooth, lingual view; c, USNM 281055, James City Formation, upper lateral tooth, lingual view; d, USNM 281059, upper anterior tooth, lingual view; e, USNM 476340, second upper anterior tooth, lingual view. (Scale bars: a= 1.0 cm; b-d= 1.25 cm; e=0.5 cm.) (CAS 40905 tooth heights: upper second anterior (A2)=1.56 cm, lower A2 = 1.61 cm; 1.2 m female, upper A2 = 1.48 cm, low- er A2= 1.54 cm; 1.4 m male, upper A2= 1.66 cm, lower A2= 1.66 cm). Among the intermediate teeth, the tip of the tooth points mesially in CAS 40905, is almost symmetrical in the 1.2 m female, and is erect but asymmetrical in the 1.4 m male. At Lee Creek Mine, most specimens are jet black, and the roots are very corroded. On the spoil piles, these teeth were found intermingled with masses of large molluscan shells of the James City Formation. A few teeth, however, compare in preservation to those from the basal Yorktown Formation (Fig- ure 43a,b). Because they have been found in early Pliocene beds elsewhere, this occurrence is not unusual (Leriche, 1936b:746; Muizon and De Vries, 1985:554). Their rarity in the basal Yorktown Formation may be related to the abundance of large teeth of C. megalodon; at Peruvian localities of equivalent age where C megalodon is uncommon, C. carcharias is abundant, and it is rare in Miocene sediments that yield the teeth of C. megalodon. Leriche (1927:82) also noted that C. carcharias "appears in the Miocene, where it is exceedingly rare. It is more widespread in the Pliocene, where it tends to replace C. megalodon." In the late Pliocene sedi- ments of Baja California, Espinosa-Arrubarrena and Applegate (1981) report that these two species do occur together. The stratigraphic resolution of the Baja California occurrence is not fine enough to determine if the two species occurred in the area at the same time of the year or at different times. With the ex- ception of this last, these occurrences suggests to us that these two species may have had allopatric distributions (For further discussions of the distribution and paleoecology of Carcharo- don, see Purdy, 1996). In their study of white sharks from along the coast of Califor- nia, Tricas and McCosker (1984:232-233) reported that "fish prey predominated in the diet of sharks [Carcharodon carchar- ias] approximately 3 m or less (TL), while pinnipeds and ceta- ceans predominated in those of larger sharks." Casey and Pratt (1985:10) made similar observations about white sharks occur- ring off the east coast of the United States. Compagno (1984:240; see also Cliff et al., 1989) noted the following: "Larger white sharks above 3 m long tend to prey heavily on marine mammals, while smaller sharks below 2 m long feed heavily on bony fishes and small sharks, but even large sharks are capable of eating smaller prey such as the 150 crabs, salmo- ns, hakes, and rockfishes found in a 4.4 m specimen from NUMBER 90 139 Washington State, USA. Pinnipeds may be especially impor- tant prey for white sharks where they occur together, but in tropical areas without these mammals the white shark is proba- bly capable of subsisting on other sharks, bony fishes, turtles and cetaceans." According to Compagno (1984:239-240), white sharks are primarily coastal and offshore inhabitants of continental and insular shelves. They are most commonly recorded in cold and warm-temperate seas. The Lee Creek great white sharks were between 4 and 5 m in total length. Order Carcharhiniformes Family Scyliorhinidae (cat sharks) Scyliorhinus sp. Figure 44a-h HORIZON.—Pungo River Formation (units 1-5); Yorktown Formation? (unit 1 ? probably redeposited from underlying sed- iments). Referred Material.—12 isolated teeth, 1 from Yorktown Formation, USNM 312266-312268. Remarks.—Compagno (1988:121) described the teeth of Scyliorhinus as follows: "[Teeth] similar in upper and lower jaws, cusps more oblique in upper jaw__Sexual heterodonty absent or poorly developed.... Anterolateral teeth have erect or semioblique, moderately high cusps and mesial and distal cus- plets, lower on posterior teeth; transverse ridges are confined to the basal ledges and do not extend onto the cusps." The Lee Creek Mine teeth match his description. The teeth from Lee Creek Mine are small (mean total height 3.1 mm (Figure 44a-h); mean maximum mesial-distal length 2.5 mm) and have a prominent, lingually inclined central cusp bordered on either side by a single lateral cusplet. The central cusp is conical apically, but its lower part is divided into labial and lingual faces by mesial and distal cutting edges. These ex- tend from one-third to one-half the height of the main cusp through a notch formed between the main cusp and the lateral cusplets, and then ascend onto the lateral cusplets. On the labial face, the main cusp and the lateral cusplets are moderately con- vex. This face of the crown is smooth apically, but it is some- times striated along the crown-root boundary. These striations are usually absent on the convex, lingual face of the central cusp, but they may be present on the lateral cusplets. The root has a prominent planar basilar face divided by an open trans- verse groove. The Lee Creek Mine specimens and those identified by An- tunes and Jonet (1969-1970) and Cappetta (1970) as Scyliorhi- nus distans (Probst, 1879) may represent the same species, but because little is known about the dental morphology of the ex- tant spotted cat sharks, of which there are 13 species, it is not prudent at this time to assign the fossil teeth to a species. Compagno (1984:366) reported that the extant chain cat shark inhabits waters of 73 to 550 m in depth from off southern New England to Florida and from the northern Gulf of Mexico to Nicaragua; its feeding habits are unknown. Family Triakidae (tope sharks) Galeorhinus cf. G. affinis (Probst, 1878) Figure 44i-m Horizon.—Pungo River Formation (units 1-5); Yorktown Formation (units 1, 2, possibly redeposited). Referred Material.—37 teeth, USNM 207450, 207451, 207453, 207454, 312270-312272. Remarks.—For the genus Galeorhinus, Compagno (1988:248) described the teeth as exhibiting weak dignathic heterodonty, and except for two upper and lower medials, the anterior and posterior teeth cannot be differentiated (anteropos- teriors). The medial teeth differ from the "anteroposteriors in having erect cusps and both mesial and distal cusplets, whereas anteroposteriors have oblique cusps and well-developed distal cusplets. Mesial cusplets are well developed in the fossil Eocene species Galeus rectoconus Winkler, 1873 [Cappetta as- signed this species to his genus Abdounia], and occasional adults of the living Galeorhinus galeus. Pegs absent from tooth crowns. Tooth roots with strong transverse notches" (Compag- no, 1988:248). In the seven dentitions of Galeorhinus galeus (Hubbell collection) available to us, the mesial edges are straight or dog-legged. In larger males, the teeth have broader crowns than do those of females, and the distal cusplets are saw-toothed. The teeth from Lee Creek Mine resemble those of females (Figure 44i-m). Probst's (1878:139, pl. 1: figs. 64-67) type suite for G. affi- nis contains four teeth of two different genera; only the two teeth in his figs. 66 and 67 belong to Galeorhinus; the other two teeth (figs. 64, 65) appear to be symphysials of Paragaleus. The specimens in his figs. 66 and 67 are very close in morphology to those illustrated by Compagno (1988, pis. 20: fig. J, 21: fig. J) and Herman et al. (1988:106, pl. 13). Because Probst's type specimens are lost and a large sample of dentitions from the extant species is not available to us, we cannot at this time assess the validity of his species. If it is valid, then the Lee Creek Mine teeth, which resemble those in Probst's figs. 66 and 67, should be assigned to Galeorhinus affinis. The Lee Creek Mine specimens are small; tooth height rang- es from 2.0 to 7.5 mm, width ranges from 2.1 to 9.9 mm, thick- ness ranges from 0.9 to 3.0 mm, and the number of distal cus- plets ranges from two to four. Compagno (1988:33) reported that the number of distal cusplets in Galeorhinus increases with the growth of the shark. This shark no longer inhabits the coastal waters off eastern North America. It is found in cold to warm-temperate continen- 140 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY tal seas in the eastern Pacific, the eastern North Atlantic, the South Atlantic, and off southern Australia, where it frequents depths from 2 to 471 m. It feeds on bony fishes and inverte- brates (Compagno 1984:387-388). Hypogaleus sp. Figure 44q HORIZON.—Pungo River Formation (unit 1). Referred Material.—1 tooth, USNM 207452. Remarks.—This tooth from the lower Pungo River Forma- tion compares favorably with the lower teeth figured by Compagno (1988, pl. 21: fig. I) and Herman et al. (1988, pl. 14). In the living species and in the Lee Creek specimen, the mesial edge of the crown is convex, and the distal cutting edge is concave on the apical half of the crown. This specimen mea- sures 4.1 mm in height and 5.5 mm in width. According to Compagno (1988:394), this shark inhabits deepish water in the tropical and subtropical western Indian and western North Pacific oceans; it eats bony fishes. Mustelus sp. Figure A4n-p Horizon.—Pungo River Formation (units 1-5). Referred Material.—5 isolated teeth, USNM 207589, 207590. Remarks.—These diminutive durophagous teeth are con- vergently similar to those of certain rhinobatids and pristids. The mesiodistal diameter (range= 1-1.3 mm) is twice that of the labiolingual diameter and is one and one-half times the maximum tooth height; the labiolingual diameter is approxi- mately equal to the medial height of the crown measured on the lingual side. In apical view, the occlusal surface is elliptical, more or less flattened, but rising slightly above the lingual peg and restricted from the lingual surface by a cutting edge. On the labial face (Figure 44c), the crown overhangs the root and is vertically striated. These striations, or more accurately, ver- miculating ridges, continue a short distance onto the occlusal surface of unworn teeth. These ridges also occur on the lingual surface of the crown and extend higher there than on the labial surface. On the lingual face (Figure 44p) the medial portion of the crown foot expands to form a relatively prominent, rhino- batid-like process or peg. The root is bipartite, with a flattened attachment face that is inclined at about a 30° angle to the plane of the occlusal surface. Compagno (1984:399), for purposes of identification, sepa- rated Mustelus into three groups, those with "cusps high on teeth," those with "no cusps on teeth," and those with "low blunt cusps present on teeth, crowns asymmetrical." The Lee Creek Mine specimens belong to the last category, and they are similar to those of M. canis, which inhabits the temperate and tropical continental waters of the western Atlantic and the Gulf of Mexico (Compagno 1984:405). It is found from the intertid- al zone to 200 m in depth and feeds mainly on crustaceans, al- though it also feeds on small bony fishes (Compagno 1984:406). Sffli ci bed e f g--------h ; nop Figure 44.—Scyliorhinus sp.: a, USNM 312266, incomplete, anteroposterior tooth, labial view; b, same specimen, lingual view; c, USNM 312267, anteroposterior tooth, lingual view; d, same specimen, lateral view; e, same specimen, labial view;/ USNM 312268, anteroposterior tooth, lateral view; g, same specimen, lingual view; h, same specimen, labial view. Galeorhinus cf. G. affinis: i, USNM 312270, lower anteroposte- rior tooth, lingual view;/ USNM 312271, upper anterior anteroposterior tooth, labial view; k, USNM 312272, upper anterior anteroposterior tooth, labial view; /, USNM 207450, upper medial tooth, labial view; m, USNM 207453, lower anteroposterior tooth, labial view. Mustelus sp.: n, USNM 207589, anteroposterior tooth, lateral view; o, same specimen, labial view; p, USNM 207590, anteroposterior tooth, lingual view. Hypogaleus sp.: q, USNM 207452, anterior tooth, lingual view. (Scale bars: a-e=0.s cm;f-h=0.25 cm; 1=0.5 cm;y-/=0.5 cm; m-^=0.1 cm.) NUMBER 90 141 Family HEMIGALEIDAE (weasel and snaggletooth sharks) Paragaleus sp. Figure 45 HORIZON.—Pungo River Formation (units 1-5); Yorktown Formation (unit 1, probably redeposited). Referred Material.—About 50 teeth, USNM 207455-207461, 475446, 475447, 476288-476292. Remarks.—Teeth from the Pungo River Formation (Figure 45) compare favorably with those of the extant species. Compagno (1988:258) described these teeth as follows: "Up- per anterolateral teeth with moderately long cusps that are no- ticeably longer than the distal cusplets. Lower anterolateral teeth slightly smaller than uppers. Cusps of lower anterolaterals short, stout, straight or slightly hooked, and differentiated from the crown foot by distal sometimes mesial notches. Some or al- most all cusps on lower anterolaterals oblique or semioblique. Distal and sometimes mesial cusplets on some lower anterolat- eral teeth.... Ventral edges of root lobes and crown feet straight or nearly so and horizontal on anterolaterals of lower jaw, giv- ing teeth an inverted T shape." In his description, Compagno does not mention the mesial cusplets that may occur on the up- per anterior tooth as illustrated by Bigelow and Schroeder (1948:277). Otherwise, the Lee Creek Mine teeth agree with Compagno's description. FIGURE 45.—Paragaleus sp., Pungo River Formation: a, USNM 207455, upper medial tooth, lingual view; b, USNM 207456, upper anterior tooth, lingual view; c, USNM 207457, lower anterior tooth, lingual view; d, USNM 207458, lower anterior tooth, lingual view; e, USNM 207459, lower anterolateral tooth, lingual view;/ USNM 207460, upper anterolateral tooth, lingual view; g, same specimen, labial view (reversed); h, USNM 207461, upper lateral tooth, labial view; i, same specimen, lingual view (reversed). (Scale bars=0.25 cm.) 142 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY The upper teeth from Lee Creek Mine have crowns that in- cline distally, with smooth cutting edges that become serrated basally on the distal enamel shoulder; the shoulder is separated from the crown by a deep notch. Their crowns are elongate, with convex distal and concave mesial cutting edges. As in the extant species, often the tips of the crowns bend lingually but not as much as those of the lower teeth. The upper teeth have more compressed roots than do those of the lower teeth. On the lingual face of the root there is a well-developed transverse groove that penetrates the basal margin of the root. In the lower jaw, the most anterior teeth lack the distal serra- tions, and their crowns may be slightly contorted. In the region of the transverse groove, their roots are thicker than those of occluding upper teeth. The more distal anterolateral teeth pos- sess serrated enamel shoulders. Although Jonet (1966:81-83; emended in Antunes and Jo- net, 1969-1970:169-171) described a new species, Paragaleus pulchellus, for his specimens, insufficient specimens of the four extant species are available to assess the validity of Jonet's species; therefore, we refrain from identifying the Lee Creek Mine specimens to a species. The anterolateral teeth from Lee Creek Mine are 6.4 to 8.5 mm in height (mean=7.4 mm, n=6) and 6.6 to 9.2 mm in width (mean=7.9 mm, n=6); they probably came from night sharks between 1.5 and 1.7 m in total length. According to Compagno (1984:442-444), Paragaleus inhab- its tropical and warm-temperate waters off western Africa, the northwestern India Ocean, and the western Pacific Ocean. Paragaleus pectoralis preferentially feeds on squids but will also take small bony fishes. Genus Hemipristis, Agassiz, 1843 Compagno (1988:269-270) characterized the teeth of this genus as follows: "The medials are very small, clawlike teeth with very narrow roots and strongly hooked cusps; symphys- ials are similar but larger, with broader symmetrical roots. Up- per anteriors are abruptly larger, broader, and flatter than the symphysials and differ from the laterals in having much nar- rower roots and crowns. The lower anteriors are very high, nar- row, deep-rooted and hook-cusped teeth having serrations or cusplets, if any, confined to the crown foot, and form a spike- studded impaling pad__The lower anteriors grade into the lat- erals by becoming lower-crowned, broader, shorter-cusped, and by having serrations or fine cusplets extending onto the distal crown feet. Upper laterals are very broad, wide, triangu- lar teeth, with compressed, horizontal-edged roots, broadly convex mesial edges, and arcuate distal edges having coarse serrations or cusplets. In small specimens the cusps of all the teeth are slightly more oblique than those of adults and sub- adults. In the young a few distal cusplets are present on upper laterals (five on fifth upper laterals of a 532 mm. specimen) but these become more numerous on adults and subadults (10 or more on fifth upper lateral) and turn into coarse serrations." Hemipristis serra Agassiz, 1835 Figures 46-48a Horizon.—Pungo River Formation (units 1-6); Yorktown Formation (units 1, 2, 3?). Referred Material.—About 700 teeth, USNM 278198, 435197, 444191, 451274-451327, 474917, 474920-474940; 6 vertebrae, USNM 467528^67532. Remarks.—The teeth of this species are among the most common large-selachian remains occurring at the Lee Creek Mine in both formations; their abundance permitted us to re- construct their dentitions (Figures 46, 47). The upper and lower symphysial teeth, the lower anterior teeth, and the first two laterals from the lower jaw are awl- like, with a few serrations or cusplets near the base of the crown. The more distal lower lateral teeth have coarsely ser- rated cutting edges and are much narrower than the upper lat- eral teeth. The upper anterior and lateral teeth are subtriangu- lar, labiolingually compressed, broad, and with coarsely serrated cutting edges. One of the more characteristic features of these teeth and of Hemipristis is the prominent lingual torus of the root. Hemipristis seems to increase in size through its evolution- ary history as do some of the other large sharks of the Neogene. At Lee Creek Mine, the largest teeth of Hemipristis, which are 30% larger than those from the Pungo River Formation, occur in the lower Yorktown Formation. These teeth are the largest yet reported in the literature. In the Tertiary sediments of Baja California, Applegate (1986) also noticed this increase in tooth size in Hemipristis; his tooth sizes correspond with those from Lee Creek Mine and those (measured by R.W.P.) from the ear- ly Miocene to early Pliocene of Florida. But is this size in- crease an evolutionary change? One exception to this increase in size through time does ex- ist: the teeth of Hemipristis, collected in place with late Oli- gocene whales and invertebrates (Sanders, 1980), range in height from 12.2 to 25.8 cm (mean=19.4 cm, «=21). These teeth compare in size with those from the middle and late Mi- ocene. We question, therefore, whether this size increase is re- lated to the evolution of Hemipristis; it may be due to ecologi- cal changes favoring increasingly larger sharks, which in earlier epochs may have inhabited areas not represented in the available fossil record. Are the Pungo River Hemipristis teeth those of juveniles? As noted above, Compagno (1988:269-270) reported the oc- currence of ontogenetic variation in the upper lateral teeth of the extant species, namely, that the number of cusplets on the distal cutting edge increases with age. In comparing the fifth upper laterals of the reconstructed Hemipristis dentitions from the Pungo River (tooth height=2.04 cm) and the Yorktown (tooth height = 3.41 cm) formations, both teeth possess 14 coarse, distal serrations. Therefore, the smaller, Pungo River Hemipristis teeth are from adults, possibly young adults. The apparent size difference between the Pungo River and York- NUMBER 90 143 FIGURE 46.—Hemipristis serra, Pungo River Formation, composite dentition, lingual view; second symphysial and last lateral tooth of upper series and last three lateral teeth of lower series missing. (Scale bar=2.0 cm.) FIGURE 47.—Hemipristis serra, Yorktown Formation, composite dentition, lingual view; two symphysials and two posterior lateral teeth of upper series and one symphysial and 10 lateral teeth of lower series missing. (Scale bar=1.0 cm.) town populations of this shark (see Branstetter et al., 1987) also may reflect different rates of growth. Compared to the extant species, the crowns of the upper teeth in both fossil dentitions are more erect than are those of the adult female ofH. elongatus illustrated by Compagno (1988, pl. 20N), whereas the attitudes of the crowns of the low- er teeth of the fossil species are similar to those of H. elongatus inhispl. 21N. The sizes of the Hemipristis teeth from the Pungo River and the Yorktown formations are contrasted below. Most of the Pungo River specimens are from the ore horizons, units 1-3, but some material from the upper Pungo River Formation is in- cluded as well. The measured Yorktown specimens are from units 1 and 2. Pungo River Formation Height Width Yorktown Formation Height Width Range (mm) 14.1-29.1 12.3-35.5 16.4-^1.0 14.0^13.5 Mean 20.3 21.4 29.1 25.9 Number 20 20 27 27 Six vertebrae (Figure 48a) from Lee Creek Mine are identi- cal with those associated with a dentition of Hemipristis serra from the Calvert Formation. These are short, aseptate centra 144 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY with rectangular to suboval dorsal and ventral foramina; diago- nal laminae bisect these foramina. In the vicinity of the forami- na and along the rims of the centra, pores are usually present. The extant species is a tropical coastal shark that inhabits waters from 1 to 30 m in depth in the Red Sea and Indian Ocean. It feeds on birds, sea catfish, mackerel, croakers, grey sharks, and other fish (Compagno, 1984:441). Family CARCHARHINIDAE (requiem sharks) Genus Galeocerdo Miiller and Henle, 1837 (tiger sharks) Compagno (1988:279-280) described these teeth as follows: "Dignathic heterodonty weak, but monognathic heterodonty well-developed, with strongly differentiated medials and an- teroposteriors in both jaws (anterolaterals grade into posteri- ors), and lower symphysial. Sexual heterodonty absent. Onto- genic heterodonty weak, with young having fewer cusplets, narrower, longer, and more oblique cusps, and more angular mesial edges on their anteroposterior teeth than adults. ... Anteroposterior teeth with characteristic cockscomb shape, with oblique primary cusps and strongly notched distal edges. Teeth secondarily anaulacorhizous, with no transverse groove and notch [in basal margin of root]." We would modify the above because the "strongly notched distal edges" are not predominant in some fossil species (e.g., Galeocerdo latidens). We tentatively identify three species of Galeocerdo from Lee Creek Mine: Galeocerdo sp. (identified by others as G. aduncus), G. contortus, and G. cf. G. cuvier. Our uncertainty regarding G. sp. and G. cuvier arises from the wide range of variability in tooth shape that exists in these sharks; our identi- fication of G. contortus is uncertain because its anterior teeth are unlike those of Galeocerdo. During the Tertiary history of Galeocerdo, two basic tooth forms reoccur and persist into the living species. In the earliest known form, Galeocerdo latidens, the distal notch of the cut- ting edge is not well developed. This tooth form, which is com- mon in the middle Eocene, recurs in abundance in the early Mi- ocene of Africa and in the late Miocene of Florida and occasionally in the teeth of the extant species (Hubbell, private collection; Alvaro Mones, pers. comm., 11 Jun 1991). The sec- ond tooth form, with a deeply notched distal cutting edge, oc- curs at widespread localities from the late Eocene (G clarken- sis White from Alabama) to the present; this tooth form is abundant in the late Oligocene of South Carolina, the middle Miocene of the Chesapeake Group, and most Pliocene marine deposits where Galeocerdo occurs. Rather than representing taxonomic differences, because both tooth types occur in the extant species, this patchwork distribution suggests to us that the teeth of the fossil species and possibly the extant species are highly variable in the development of the distal notch. Be- cause of this variability, we believe that the taxonomy of this genus needs revision and that the fossil species should be de- fined on the basis of the predominant tooth morphotype of sev- eral widely distributed, synchronous populations. None of these tooth morphotypes, however, matches exactly that of the teeth from the Pungo River Formation at Lee Creek Mine, which have a well-developed distal notch and the lateral extremities of the roots turned apically as in G. latidens. This last character is not present in the teeth of Galeocerdo from younger sediments. Concerning the Pliocene teeth that we identify as G. cf. cuvi- er, we believe that large samples of these teeth must be com- pared with large samples of teeth from the extant tiger shark, which are not presently available, to ascertain the identity of the Pliocene teeth; this also applies to ascertaining the validity of Applegate's G. rosaliensis. Our uncertainty with Galeocerdo contortus concerns its ge- neric identity. The teeth of G contortus differ markedly from all other species of Galeocerdo. By comparing the teeth of G. contortus to those of G. latidens and G. cuvier, we can contrast these differences. Like Galeocerdo cuvier, G. latidens has nearly homodont teeth, and on the lingual face of the root, the lower anterior teeth possess a slight but noticeable torus with a very shallow transverse groove; this character is harder to discern in the ex- tant species. In G. contortus, however, the lower anterior teeth have well-developed toruses with deep transverse grooves (Figure 48a"). Also, in the anterior anteroposterior teeth of G. cuvier and G. latidens, the width of the tooth exceeds its height, and in G contortus the opposite is true. In morphology, the teeth of G contortus do not fit the pattern for this genus. Both Leriche (1942) and Applegate (1978) proposed that the teeth of G contortus were the lowers of G. aduncus. If they are, for a carcharhinid, this shark had a very unusual dentition because the heights of the lower anteriors (mean=15.7 mm, «=136, range=12.0-19.4 mm) exceed those of the upper ante- riors (mean= 14.0 mm, «=65, range= 10.7-16.8 mm), which is the opposite of the condition in G. latidens and G. cuvier, where the height of the upper teeth exceeds that of their lower counterparts. Also, in screen-wash samples of shark teeth from the Eocene Castle Hayne Formation at Rose Hill, North Carolina, the teeth of G. latidens (w=100) exhibit the weak dignathic heterodonty seen in G. cuvier. Both uppers and lowers of the Eocene teeth can be identified, and no G. contortus-type teeth of comparable size (1.2-2.0 cm, anterolaterals) were found (G. contortus has not yet been found in the Eocene). Therefore, we cannot agree with the proposals of Leriche (1942) and Applegate (1978). Cappetta (1980) placed teeth from the Paleocene and Eocene similar to those of Galeocerdo contortus in his new genus Physogaleus and assigned them to P. secundus (Winkler, 1874), which he made the type species of the genus. Unlike Galeocerdo contortus, Cappetta's (1980, fig. 5A) specimens possess a deep distal notch. Also, Winkler's two syntypes, which are illustrated in lateral view only, are quite different in morphology. Winkler's specimens possess orectilobiform rath- NUMBER 90 145 er than carcharhiniform roots. Winkler described the root of the type specimen as "a wide and robust root, of a remarkable form: it presents a kind of a mound or a pyramid with three fac- es, and with a wide base, that forms below a triangle with rounded angles. In the middle of this lower face of the root one observes a circular depression, while the enamel of the crown forms a kind of ribbon, it terminates in a small tubercule on the internal face of the root." Winkler did not mention the presence of a transverse groove on the root, which is characteristic of carcharhiniform teeth, and his "enamel ribbon" may be the peg that is often found on the teeth of orectilobiform sharks. Until Winkler's syntypes are found, the identity of Physogaleus re- mains in doubt. The characters of G. contortus are not characteristic of Gale- ocerdo, but we feel that generic reassignment of these teeth should await the discovery of an associated dentition and a more thorough study of the Paleogene forms. Size is often used to separate fossil and extant species of Ga- leocerdo (Antunes, 1963; Antunes and Jonet, 1969-1970; Ap- plegate, 1978, 1986). A recent study by Branstetter et al. (1987) found that two populations of the extant species, one in the Gulf of Mexico and the other in the Atlantic Ocean off the coast of the southeastern United States, had different rates of growth and that tiger sharks of known age from the Gulf of Mexico were larger than individuals of equivalent age from the Atlantic population. Also, Hubbell (pers. comm., 1990) noted that the largest individuals of this species do not come from these two populations but from the Pacific Ocean off Panama, an area of upwelling and high food availability. Another study by Lowe et al. (1996:209) indicated that tiger sharks segregate by size—smaller ones live in shallow, nearshore waters, whereas larger ones are more pelagic. If these distributions also were true for fossil tiger sharks, then, except for nursery areas, large teeth would occur in sediments deposited in pelagic wa- ters, and small teeth would occur in sediments deposited in shallow, nearshore waters. If this hypothesis is correct, then the Neogene record of a size increase in tiger sharks may reflect an increase in water depth and in the availability of prey in the ba- sins of deposition rather than an evolutionary change. In identifying the Lee Creek Mine tiger shark teeth, we be- lieve that there is a wide range of variation in tooth morpholo- gy within a species and that two species are represented in the Miocene and another in the Pliocene. More extensive sampling of Miocene and Pliocene shark tooth horizons may prove us wrong. Galeocerdo sp. Figures 48b,e, 49 Horizon.—Pungo River Formation (units 1-6). Referred Material.—About 200 teeth, USNM 451233-451251,464118, 464119, 476376. Remarks.—Agassiz (1835, pl. 26: figs. 24-28, 1843:231) characterized Galeocerdo aduncus as having a greater con- FlGURE 48.—Hemipristis serra: a, USNM 467531, vertebra, dorsal view. Galeocerdo sp.: b. USNM 476376, pointed serrations on mesial edge of tooth; e, USNM 451240, basal view of root. Galeocerdo cuvier: c, USNM 476377, pointed, compound serrations on mesial edge of tooth. Galeocerdo contortus: d, USNM 451226, basal view of lower anterior tooth showing well-developed torus. (Scale bars: a=0.5 cm; b,c=0.5 mm; d,e=033 cm.) vexity in the distal enamel shoulder than exists in the living species; this character, however, is variable in the fossil and extant species. Cigala-Fulgosi and Mori (1979) described the only character that may separate middle Miocene teeth identi- fied traditionally as G. aduncus (=G. sp. herein) from G. cuvi- 146 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY er; they noticed that unlike the extant species, on the basal half of the mesial edge of the tooth the serrations are simple rather than compound (cf. Figures 48b, 48c). Incipient com- pound serrations, however, may be present. We do not know if these compound serrations are present or absent on Agas- siz's holotype, which was deposited at the Staatliches Muse- um fur Naturkunde, Karlsruhe, but is now lost. According to Agassiz, the holotype was found in the Schwabia region of southwestern Germany, where marine sediments were depos- ited from the Burdigalian into the Zanclian. Because the exact age of the type specimen cannot be determined, it may repre- sent a juvenile of G. cf. cuvier, or it may be the species identi- fied until now as G. aduncus by vertebrate paleontologists. Without examining the type specimen, the identity of G. aduncus cannot be ascertained. We believe, therefore, that Galeocerdo aduncus is a nomen dubium and is not available for use as a scientific name. At this time, however, we hesitate to identify the next avail- able junior synonym because the types of the subsequently de- scribed species are not now available to us, and we do not know their whereabouts. We believe that the resolution of this problem must await a major revision of the fossil species of this genus. Figure 49 illustrates a reconstructed dentition of this spe- cies. For illustration purposes, these teeth were separated by spaces; the lingual recurvature of the lateral extremities of the roots indicates that these teeth overlapped (Figure 48c), a con- dition termed alternate-imbricate overlap by Compagno (1988:31). The crowns of the upper teeth have a strong labial curvature, and as in the extant species, they are slightly more elongate than their lower-jaw counterparts. In our reconstruc- tion, the lower medial, an upper posterior, and a lower poste- rior tooth are missing; they were not found in the USNM col- lections. Well-preserved specimens of this species are found embed- ded in the mold and cast limestones in the upper part of the Pungo River Formation. The species also is common in ore- sump residues and from the reject gravels screened from the ore during processing. The condition of the few specimens found on spoil piles of lower Yorktown Formation sediments suggests that they were redeposited. We have not seen the specimens described by Leriche (1942) as Galeocerdo aduncus, but on the basis of his illustrations (pl. 7: figs. 33^12), the teeth appear to be the same as Galeocerdo contortus Gibbes (see discussion of G contortus). The anterior and anterolateral teeth range from 10.7 to 16.8 mm in height (mean=14.0 mm, «=65) and from 12.4 to 20.0 mm in width (mean= 16.2 mm, «=65). In the extant tiger sharks, teeth of these sizes are found in individuals of 1.8-3 m TL. Galeocerdo contortus Gibbes, 1848-1849 Figures 484 50 Galeocerdo acutus Storms, 1894:81-82, pl. 6: fig. 18 [Rupelian, Belgium]. Galeocerdo triqueter Eastman, 1904:89, pl. 32: fig. 12 [Calvert Formation, Maryland]. Physodon triqueter Leriche, 1942:79 [Calvert Formation, Maryland]. Galeocerdo aduncus Agassiz.—Leriche, 1927:88, pl. 14: figs. 1-3, 6 [middle Miocene, Switzerland]; 1942:87-88, pl. 7: figs. 33^12 [Calvert Formation, Maryland].—Caretta, 1972:54-57, pl. 11: figs. 1, 2, 4 [middle Miocene, Italy]. Horizon.—Pungo River Formation (units 1-6). Referred Material.—About 600 teeth, USNM 451212^51232, 476386-476393. REMARKS.—Galeocerdo contortus is distinguished from the other small tiger shark of the Pungo River Formation, G. sp. (see above), by the following criteria. Galeocerdo sp. Galeocerdo contortus crown relatively thin, flat; margins nearly in same plane mesial margin of crown uniformly curved in profile serrations on enamel shoulder com- pound, coarse apex angle of crown 35°-50° (mean=43°, n=16) root relatively thin; like G. cuvier, lacks prominent torus on lingual face width of anterior anteroposterior teeth exceeds height of tooth crown thick, twisted; mesial margin warped mesial margin of crown flexuous in profile serrations on enamel shoulder simple, fine apex angle of crown 22°—40° (mean=31°, n=40) root thick, with prominent torus on lingual face as in Hemipristis height of anterior anteroposterior teeth exceeds width of tooth %g§ *$E0& ^ ^0 %j^ \*& \f ** A> 3B* ^*^ J& J^ J^ J^ FIGURE 49.—Galeocerdo sp., composite dentition, lingual view. (Scale bar= 1.0 cm.) NUMBER 90 147 Galeocerdo contortus always occurs with G. sp. in Neogene localities on the East Coast of the United States. Counts of teeth collected by screen washing at Lee Creek Mine indicate it is twice as common as G. sp. This ratio has been constant in several separately collected samples. Figure 50 represents a reconstructed dentition of this species. In sorting the teeth, we identified both uppers and lowers. In the lower teeth, a prominent, lingual torus is present, particular- ly in the more anterior teeth (Figure 48a"), which becomes less prominent in the more posterior teeth. The tips of the crowns of these teeth exhibit the characteristic lingual bend. In the upper teeth, the root is more compressed, and the tips of the crowns have a noticeable labial curvature. Our reconstruction, based on G. cuvier, includes all tooth positions except the upper me- dial, if one is present, and one or two of the posterior teeth in each jaw. Cappetta (1987:123) stated that this species does not occur in Europe, but specimens identified by Storms (1894) as Galeo- cerdo acutus and by Leriche (1927) and Caretto (1972) as Ga- leocerdo aduncus are the teeth of G. contortus. In separating Galeocerdo contortus from G. sp. we asked ourselves how the Pungo River environment supported two such top predators. We compared the reconstructed dentition of G. contortus with those of other carcharhinids; the narrow up- per teeth are similar to those of Negaprion, which feeds on bony fishes and rays. Unlike G. sp., which has broad-bladed teeth for biting off chunks of flesh from larger animals, but like Negaprion, G contortus had teeth for feeding on smaller bony fishes and rays. Therefore, the dentitions of G. sp. and G. con- tortus suggest that these two sharks did not compete for the same prey. Tooth height of the anterior and anterolateral teeth ranges from 12.0 to 19.4 mm (mean=15.7 mm, «=136). Mesial-distal width of these teeth ranges from 12.0 to 19.5 mm (mean=15.5 mm, n=136). The average tooth size of G. contortus is larger than that of G sp. Galeocerdo cf. G. cuvier Peron and LeSueur, 1822 FIGURES 51, 52a,b Galeocerdo arcticus.—Leriche, 1942:88, pl. 7: figs. 1, 2 [Ashley phosphates (=Goose Creek Formation), middle Pliocene].—[Not G arcticus Faber, 1829=G. cuvier Peron and LeSueur, 1822.] Horizon.—Yorktown Formation (units 1-3). Referred Material.—649 teeth, USNM 451213^151218, 451225^151230, 456327, 457287, 457288, 474941^174960; 1 vertebra, USNM 467534. Remarks.—These teeth are as large as some of those of the largest individuals of the extant species and are twice as large as those of Galeocerdo aduncus (=G. sp. herein). They are identical in many respects to the teeth of G. cuvier, but unlike those of the extant species, the basal and apical portions of the mesial cutting edge are straight; this edge presents a distinct obtuse angle. In the extant species (Compagno, 1988:279) this obtuse angle is predominant in the juveniles rather than in the adults. Lawley (1876:16-17) erected Galeocerdo capellini on the basis of an isolated tooth from the Pliocene of Tuscany, Italy, but he did not illustrate the holotype. He first illustrated two teeth of this species from the type locality without referring to either of them as the holotype (Lawley, 1881:145-146, pl. 1: fig. 6, pl. 2: fig. 5). De Stefano (1909:578) synonymized this species with G. aduncus. The tooth, however, in Lawley's pl. 2: fig. 5, which is enlarged, has compound serrations on its mesial edge; therefore, it cannot be synonymized with G. aduncus. Landini (1976:115-116) and Cigala-Fulgosi and Mori (1979:125) noted that the Galeocerdo teeth from the Mediterranean Pliocene are indistinguishable from the extant species, and they synonymized the Pliocene species G. capel- lini with G. cuvier. A decision, however, about the synonymy of these two species must await the study of dental variation in a larger sample of G. cuvier from various parts of the world. *xy ytw^^ttt * * ^CiAiJC j*. Figure 50—Galeocerdo contortus, composite dentition, lingual view. (Scale bar-1.0 cm.) 148 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY a — « y f «/ *a Figure 51.—Galeocerdo cf. C. cuvier: o, composite dentition, lingual view; b, USNM 457287, pathologic tooth, lingual view; c, same specimen, labial view; d, same specimen, lateral view; e, USNM 457288, pathologic tooth, lingual view;/ same specimen, labial view. (Scale bars: a=I.O cm; fr-/=0.33 cm.) NUMBER 90 149 Applegate (1978) erected the species Galeocerdo rosaliaen- sis for Pliocene teeth from Baja California, Mexico, and stated that this species is distinguished "by having the combination of large size and shape similar to G. cuvier, but with the upper half of the anterior border flattened" (Applegate, 1978:59). These characters also occur in G. capellini and G. cuvier; therefore, we believe that Applegate's species is a junior syn- onym of G. capellini Lawley. A reconstruction of a dentition of this species (Figure 51a) shows that, as in the extant species, the tips of the crowns of the lower teeth have a slight lingual bend, whereas the tips of the upper teeth have a noticeable labial curvature; however, this la- bial curvature is obscured in teeth that have convex, labial, coronal surfaces. In the lower teeth, a slightly developed torus is present on the lingual surface of the root, and the arch formed by the root is often asymmetrical (skewed distally), whereas the arch of an upper-tooth root is usually more sym- metrical. Two pathologic teeth, USNM 457287 and 457288, were col- lected by Frank and Becky Hyne (Figure 5\b-f). Gudger (1937) and Cadenat (1962) noticed these pathologies in the ex- tant species, but they were unable to identify their cause. In other species of sharks, Cadenat (1962) found that similar de- formities were due to the tooth germs being damaged by sting- ray or sea catfish spines, and Compagno (1984:505) reported that tiger sharks prey on these animals. Tooth height ranges from 13.5 to 29.1 mm (mean=22.2 mm, «=38). Mesial-distal width ranges from 24.4 to 33.0 mm (mean=28.9 mm, n=38). Apical angle, the angle between the mesial and distal margins at the apex, ranges from 45° to 65° (mean=52°, «=33). One vertebra (Figure 52a, b) from Lee Creek Mine, probably from the Yorktown Formation, exhibits the characters identi- fied by Kozuch and Fitzgerald (1989) as belonging to Galeo- cerdo. The centrum is aseptate with scattered pores, and the dorsal and ventral foramina are oval and do not extend to the rims of the centrum. According to Compagno (1984:504), the extant tiger shark is a "wide-ranging coastal-pelagic tropical and warm-temperate shark. ...It often occurs in river estuaries, close inshore off wharves and jetties in harbours." He reported (1984:505) that it preys on a variety of fishes (including sea catfish, tarpon, mackerel, porcupine fishes, and puffers), marine reptiles, sea birds (including shearwaters, frigate birds, pelicans, and cor- morants), marine mammals (including seals, monk seals, and odontocete whales), carrion, and mollusks. Compagno (1984:505) reported that "this shark takes marine reptiles more than any species, and frequently preys on sea turtles (green, loggerhead, and ridley turtles) and is one of the most important predators on sea snakes." Genus Carcharhinus Blainville, 1816 Until recently, the taxonomic confusion that existed in the extant sharks of this genus hampered identification of the fossil teeth. Many paleontologists assigned the fossil teeth of Car- charhinus to different genera and assigned teeth now attribut- able to two or more species to one fossil species. Garrick (1982, 1985), however, revised the taxonomy of this genus and included descriptions and illustrations of the teeth of each spe- cies. Compagno's work (1984, 1988) also has added important information about the identity of these sharks on the basis of teeth. These works and the available dentitions of the extant species suggest that the differences in dental characters be- tween species is often subtle. Naylor and Marcus (1994) developed a morphometric meth- od for analyzing the variation in the upper teeth of Carcharhi- nus that segregates the teeth of the many extant species of this genus, and at present Naylor is developing this technique for the study of fossil species of this genus. When developed, his method may facilitate the identification of the Lee Creek Mine species. The tooth types present in this genus include upper and lower medials or alternates, usually lower symphysials, upper anteri- ors, upper lateroposteriors, lower anteroposteriors, and upper and lower posteriors. Compagno (1988:310) provided some ge- neric characteristics for the teeth, noting, [The anterior teeth of the lower jaw are] absent or poorly differentiated. Upper anteriors are usually abruptly narrower based than adjacent laterals. Sexual het- erodonty usually absent or weak,... ontogenic heterodonty very variable in dif- ferent species. Teeth often have thicker, less oblique cusps, and more numer- ous, less course [sic], and stronger serrations in adults than young. Cusplets, where present, are often better developed in young than adults, where they may disappear entirely or become coarse serrations. [In adults] crowns moderately high and cusps variably short to very long on teeth of both jaws. Cusps variably erect or oblique on the upper teeth, usually more or less erect on lower teeth but oblique in some species. Most species have well-developed serrations along the entire edges of the upper teeth. Some either lack them at all stages (C. ma- cloti) or have them poorly developed or absent in young (C. brevipinna, C. is- odon and C. leiodon) and poorly to well-developed in adults (C. brevipinna and C. isodon). Cusplets are entirely lacking from lower teeth but are variably de- veloped on upper teeth. Some species have distal cusplets on their upper teeth when young but lose them when adult (C. melanopterus); in others these per- sist throughout life (C. dussumieri, C. sealei, C. signatus and C. macloti). Me- sial cusplets are sometimes present on upper teeth (as in C. falciformis, C. sig- natus adults, and C. macloti). Roots of lower anterolateral teeth usually with nearly straight ventral edges but these are moderately to deeply arched in a few large-toothed species (such as C. leucas and C. ambionensis). Teeth usually holaulacorhizous, with transverse notches and grooves, but these are some- times absent. Basal ledges and grooves are vestigial or absent on tooth crowns. Hundreds of vertebrae assignable to this genus were found at the mine; five different morphotypes, which may be found in one or several species, are illustrated herein (Figure 52c-/). In cranial or caudal view, these aseptate vertebrae may be round or oval in outline; the ventral edge may be concave, and the dorsal edge may be slightly pointed (see Kozuch and Fitzgerald, 1989). The dorsal and ventral foramina also vary in shape, being square, rectangular, or oval. Too few speci- 150 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE 52.—Galeocerdo cf. G cuvier. USNM 467534, vertebra: a. dorsal view; b, ventral view. Carcharhinus spp., vertebrae: c. USNM 467600, ventral view; d. same specimen, view of articular surface; e, USNM 467360, dorsal view;/ same specimen, view of articular sur- face; g, USNM 467602, ventral view; h, same specimen, view of articular surface; i, USNM 467601, dorsal view;/ same specimen, view of articular surface; k, USNM 467599, dorsal view; /, same specimen, view of articular surface. (Scale bars=I.O cm.) mens of the extant species are available to identify them to species. At least seven species of Carcharhinus may be distinguished tentatively among the Lee Creek Mine teeth: C brachyurus, C. falciformis, C leucas, C macloti, C. obscurus, C. perezi, and C. plumbeus. Because tooth morphology varies greatly in the small number of dentitions now available from the extant spe- cies, the characters that we use to identify them may, after ex- amination of larger numbers of dentitions, fall within the range of intraspecific variation in two or more species. NUMBER 90 151 Carcharhinus brachyurus (Giinther, 1870) Figure 53a Corax egertoni Agassiz, 1843:228, pl. 36: fig. 6 [Miocene, Maryland]. Carcharhinus egertoni (Agassiz).—Antunes and Jonet, 1969-1970:189-190, pl. 15: figs. 110, 111 [middle Miocene, Portugal]. Carcharhinus priscus (Agassiz).—Cappetta, 1970:54-57, pl. 13: figs. 8-19 [middle Miocene, France]. Prionodon egertoni (Agassiz).—Leriche, 1942:80-82, pl. 7: fig. 4 [middle Mi- ocene, Maryland]. Sphyrna americana Leriche, 1942:86, pl. 6: figs. 7, 8 [fig. 6=C. cf. limbatus; early Pliocene, South Carolina]. Sphyrna prisca eastmani Leriche, 1942:85, fig. 7, pl. 7: figs. 28-32a [middle Miocene, Maryland]. HORIZON.—Pungo River Formation (units 1-5). Referred Material.—About 500 teeth, USNM 463978-463998,474888^174902. Remarks.—These teeth (Figure 53a) compare favorably with those of mature females of the extant species. Compagno (1984:465) characterized the upper teeth of the living species as having "narrow, strongly serrated, semierect to oblique [crowns], high bent cusps and transverse roots." All of these characters exist in teeth identified to this species from Lee Creek Mine. In our examination of dentitions (e.g., USNM 197665) from the recent species, we found that the upper teeth may be finely serrated rather than strongly serrated as indicated by Compagno. Although these teeth may be confused with those of Carcharhinus perezi, the convexity or angle of the me- sial cutting edge gives the tooth a truncated appearance, and the teeth are narrower than those of C. perezi. According to Bass et al. (1973:24), the upper teeth of "large males are distinctly hooked near the tips as compared to those of females." Garrick (1982:175) added that compared to fe- males, the upper teeth of males are proportionately longer and narrower, are more oblique to curved laterally, and have finer serrations. In females, hooked upper teeth do occur in the more posterior regions of the jaws but not in the anterior regions; it is in the posterior region of the jaw that this sexual dimorphism is more pronounced. None of the upper teeth of this species ex- amined from Lee Creek Mine exhibit the characteristics found in the teeth of males. Although the two syntypes of Corax egertoni were not avail- able to us for this study, one of them (Agassiz, 1843, pl. 36: fig. 6) is identical to those of the extant C. brachyurus. According to Agassiz (1843:228), these specimens are from the Tertiary of Maryland. The teeth of C. brachyurus are common in the Calvert Formation of Maryland and Virginia. At Lee Creek Mine, the anterolateral teeth range from 1.2 to 1.5 cm in height (mean=1.4 cm, n=2\) and from 1.3 to 1.8 cm in width (mean=1.6 cm, «=21); in the extant species, teeth of this size are found in individuals of 2 to 3 m TL. According to Compagno (1984:465), the extant species is poorly known; it inhabits inshore to offshore warm-temperate waters from the surfline to at least 100 m in depth. It eats a va- riety of bony fishes, including sea catfish, porgies, and hake, as well as spiny dogfish, rays, squid, and cuttlefish. Carcharhinus falciformis (Bibron, 1841, in Miiller and Henle, 1839-1841) Figure 5sb-f Prionodon egertoni (Agassiz).—Leriche, 1942:80-82, pl. 7: fig. 3 [Piocene, North Carolina, South Carolina], HORIZON.—Pungo River Formation (units 2-4). Referred Material.—5 teeth, USNM 476244-476248. Remarks.—Among the teeth from Lee Creek Mine (Figure 53b-f), we were able to identify only the uppers, which are identical with those of the extant species. On the lingual face of the root and penetrating its basal margin there is a well-devel- oped transverse groove. The distal cutting edges of these mod- erately broad-bladed teeth have a shallow, angular notch; api- cal to this notch the cutting edge is perpendicular to a line tangent to the basal margin of the root; basal to this notch, there is an enamel blade, which is finely serrated like the apical por- tion of the cutting edge. On their mesial cutting edges, at about the midpoint between the tip and the base of the cutting edge, there is a gap in the serrations. Basal to this gap the serrations are larger and therefore coarser. With the exception of this gap, the mesial cutting edge is fairly straight from the base to the tip of the crown. This gap in serrations and the straightness of the mesial cutting edge are characteristic of Carcharhinus falcifor- mis. In the extant species, the upper teeth vary somewhat in mor- phology. In USNM 232776, from a 2 m TL female, the tips of the crowns are slightly hooked, and in USNM 196026, from a 2.27 m TL female, the teeth are not hooked but are more distal- ly inclined. The mesial gap in the serrations is absent. In USNM 232780, from an unsized male, the mesial edge has a dog-leg turn as in C. brachyurus, and in USNM 112584, from 2.3 m TL male, the shallow, distal notch is rounded rather than angular. Among the Lee Creek Mine specimens, the only anterior an- terolateral tooth present measures 14.2 mm in height and 15.0 mm in width; it came from an individual of probably 3 m TL. The extant silky shark inhabits warm, nearshore waters from depths of 18 to 500 m. It feeds primarily on bony fishes, in- cluding sea catfish, yellowfin tuna, albacore, and porcupine fish (Compagno, 1984:471^472). Carcharhinus leucas (Valenciennes, 1839, in Miiller and Henle, 1839-1841) Figure 54a Corax egertoni Agassiz, 1843:228, pl. 36: fig. 7 [Miocene, Maryland]. Prionodon egertoni (Agassiz).—Leriche, 1942:80-82, pl. 7: figs. 11,12. Horizon.—Pungo River Formation (units 1-5); Yorktown Formation (units 1-3). Referred Material.—About 200 teeth, USNM 278376, 278379, 278381, 278386, 278398, 278411, 278428, 278436, 282953,457079, 457084,459778^159820,474961. REMARKS.—These teeth (Figure 54a), abundant in the York- town Formation and very rare in the Pungo River Formation at 152 FIGURE 53.—Carcharhinus brachyurus: a, compos- ite dentition. Carcharhinus falciformis: b, USNM 476244, upper anterolateral tooth, lingual view; c, USNM 476245, upper anterolateral tooth, lingual view; d, USNM 476246, upper anterolateral tooth, lingual view; e, USNM 476247, upper anterior tooth, lingual view;/ USNM 476248, upper sym- physial tooth, lingual view. (Scale bars: a= 1.0 cm; b-f= 1.0 cm.) SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY JEW J^ -^ As fisfl^- *"** ^T c d— e f Lee Creek Mine, are identical to those of the extant species. The lower teeth have thick crowns and roots, and the cutting edges are finely serrated. In some lower teeth the transverse groove is well developed and penetrates the basal margin of the root so as to be visible from the labial side of the tooth; in oth- ers, which appear to be functional teeth, the transverse groove is only partially developed or is absent. The basal margin of the root is arcuate. Punctae may be present on the labial face of the root and also on the roots of the upper teeth. The form of the upper teeth tends to be equilateral rather than elongate. Both cutting edges are finely serrated, with coarser serrations basally. The mesial cutting edge of the crown is straight, wavy, or slightly convex apically; the shal- low, angular notch is absent. The distal cutting edge is usually concave, but a shallow, angular notch may be present. The basal margin of the root is usually arcuate but may be angular. In one recent dentition (USNM uncataloged, vertebrate pale- ontology synoptic collection) the roots of the upper teeth are nearly lobate. The upper teeth of this species are very similar to those of Carcharhinus longimanus. Although the teeth of the latter spe- cies are usually more elongate than are those of C. leucas, one dentition (USNM 183804) has teeth that are equilateral in form, giving them a stocky appearance. These teeth, however, have broader crowns than do those of C. leucas, and in C. leu- cas the distal cutting edge in the area of the crown foot usually forms a noticeable blade. The lower teeth also are similar to those of C longimanus. In the latter species the crowns of the anteriors and the anterior laterals are asymmetrical; in C. leucas they are almost symmet- rical, but for both species, these observations were made on a small number of dentitions, and in larger populations of these species they may not be valid. One of Agassiz's two syntypes (1843, pl. 36: fig. 7) for Co- rax egertoni compares favorably with a lateral tooth of Car- charhinus leucas. The teeth identified and figured by Leriche (1942) as Prion- odon egertoni represent several species of Carcharhinus, but his figs. 11 and 12 of pl. 7 compare favorably with the teeth of the extant C. leucas. Cappetta (1987:125-126, fig. 106D) identified a tooth from the Yorktown Formation at Lee Creek Mine as Pterolamiops longimanus; Pterolamiops is a junior synonym of Carcharhi- nus (Compagno, 1988), and Cappetta's tooth also may belong to C. leucas. The anterolateral teeth from the Pungo River Formation range from 1.7 to 2.0 cm in height (mean height= 1.9 cm, n=8) and from 1.6 to 2.2 cm in width (mean width=1.9 cm, «=8), which falls in the range of teeth sizes in the extant species from individuals of 2 to 3 m TL. Those from the Yorktown Formation range from 1.9 to 2.4 cm in height (mean height=2.1 cm, «=16) and from 1.9 to 2.5 cm in width (mean width=2.3 cm, n= 16) and probably came from individuals of 2 to 4 m TL. Compagno (1984:479) stated that the extant species inhab- its shallow, tropical and subtropical waters "less than 30 m deep and occasionally less than a meter deep, but ranging into deeper water close to shore down to at least 152 m depth." The extant shark feeds on bony fishes, including tarpon, sea catfish, tuna, sea bass, and bluefish, and on sharks, rays, sea turtles, birds, whales, and invertebrates (Compagno, 1984:480). Carcharhinus macloti (Miiller and Henle, 1839) Figure 546 Hypoprion acanthodon (Le Hon).—Antunes and Jonet, 1969-1970:187, pl. 15: figs. 100-108 [Miocene, Portugal]. Horizon.—Extremely abundant in the ore layers of the Pun- go River Formation (units 1-3); less abundant but still the dominant carcharhinid in units 4 and 5; uncommon and proba- bly redeposited in the base of the Yorktown Formation (unit 1). Referred Material.—More than 400 isolated teeth, USNM 207464, 464164-464179, 464181, 464183-464185, 474903-474914. NUMBER 90 153 FIGURE 54.—a, Carcharhinus leucas, composite dentition; b, Carcharhinus macloti, composite upper dentition. (Scale bars= 1.0 cm.) Remarks.—Until recently this species was assigned to the genus Hypoprion, but Raschi et al. (1982) have shown that this genus is a junior synonym of Carcharhinus. Compagno (1984, 1988) and Garrick (1985) concurred with their decision. The upper teeth (Figure 54b) are small (3-7 mm in height, 3-8 mm in width), triangular, acutely pointed, and, except on the mesial and distal enamel blades, have smooth cutting edg- es. In two-thirds of the sample, the mesial enamel blade is ser- rated with one to five cusplets (mean=1.6; «=320). These ser- rations, which are unevenly developed, range from distinct, well-differentiated cusplets to indistinct, wavy interruptions of the mesial cutting edge. The distal enamel blade is rounded or rectilinear, with one to seven poorly to well-developed cusplets (mean=3.1; «=309). The anterior teeth tend to be erect and bi- laterally symmetrical, with short, convex, mesial and distal enamel blades. Toward the commissure of the jaws, the more posterior teeth become increasingly inclined, with a correlated deepening of the distal notch formed between the main cusp and its enamel blade. The root's rounded, lingual face is divid- ed by a rather broad transverse groove, which penetrates the basal margin of the root. According to Garrick (1985:17), the number of large serra- tions present on the upper teeth is variable; he reported that "in four juveniles and subadults (both sexes)... there were 1-2 ser- rae laterally and 0-1 medially, whereas in a mature female... there were up to 4 laterally and 3 medially; contrast- ing this with a mature male... [the male] had not more than 2 or 3 laterally (and these poorly defined) and 0 medially; the dif- ference between these two adults may reflect either sexual di- morphism or geographic variation." The lower teeth are smaller than the upper teeth and are sym- metrical and broad based, have narrow principal cusps, and lack cusplets and serrations. The anterolateral teeth of this species range from 5.0 to 7.3 mm in height (mean height=5.6 mm, «=21) and from 5.0 to 9.1 mm in width (mean width=6.6 mm, «=21). Although we did not have any dentitions from individuals of known size, these measurements fall within the size range for the extant species, 69-100 cm TL (Compagno 1984:487). According to Compagno (1984:487), this tropical, little- known, inshore shark inhabits continental and insular shelves. Carcharhinus obscurus (Le Sueur, 1818) Figure 55a Horizon.—Yorktown Formation (units 1-3). Referred Material.—About 110 teeth, USNM 456596^156635, 457077^157089. REMARKS.—These teeth (Figure 55a) compare very closely with those of the extant species. The upper teeth of this species are identified by their vertical to almost vertical distal cutting edges and by their apically convex mesial cutting edges; the tip of the tooth appears to be deflected distally. We could not iden- tify the lower teeth with certainty. In three dentitions of the extant species, we observed the fol- lowing morphological variations. The number of rows of sym- physial teeth varies from one to three in the lower jaws and from one to two in the upper jaws. In AMNH 89233 SD, in the anteriormost upper teeth near the apex, the mesial cutting edge is straight rather than convex. In both the upper and lower teeth of the same dentition, the root lobes of the teeth form obtuse rather than straight angles, and in the lower teeth the basal con- striction of the cutting edges is absent. A juvenile dentition (AMNH 89269 SD) has lower teeth with smooth to incipiently serrated cutting edges and has upper teeth with incipient serra- tions present. The upper anterolateral teeth from Lee Creek Mine range in height from 1.7 to 2.2 cm (mean=1.9 cm, n=32) and from 1.8 to 2.5 cm in width (mean=2.1, «=32). In the extant species, teeth of this size are found in sharks of 3 m TL. 154 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY T"y yp -^ -^ Figure 55.—a, Carcharhinus obscurus. composite denti- tion; b. Carcharhinus perezi. composite upper dentition. (Scale bars=1.0 cm.) According to Compagno (1984:490^191), the extant dusky shark is "a common, coastal-pelagic, inshore and offshore warm-temperate and tropical shark of the continental and insu- lar shelves and oceanic waters adjacent to them." This species feeds on bony fishes, such as tunas, barracuda, jacks, and blue- fish; sharks, as well as sawsharks, angelfish, and dogfish; and invertebrates. Carcharhinus perezi (Poey, 1876) Figure 556 Horizon.—Pungo River Formation (units 3-5); Yorktown Formation (unit 1). Referred Material.—50 teeth, USNM 463977, 464007-464010, 464012, 464014-464026, 464028-464046, 474874^174887. Remarks.—The teeth from Lee Creek Mine (Figure 55b) compare favorably with those of the extant species. In the ex- tant species, the crowns of the upper teeth are fairly narrow, and except for the teeth in the two anteriormost positions, they incline distally; the distal angular notch may be well de- veloped, or it may be rounded. The mesial cutting edge ranges from being slightly convex to being nearly straight. In the lower teeth, the erect crowns have complete, smooth cutting edges. These same characters are found in the Lee Creek Mine teeth. The Lee Creek Mine teeth range in height from 1.3 to 1.8 cm (mean=1.4 cm, «=40) and in width from 1.3 to 2.2 cm (mean=1.7 cm, «=37); in the extant species teeth of this size are found in individuals of about 2 m TL. Compagno (1984:493) reported that this tropical shark is found inshore on "continental and insular shelves, at depths down to at least 30 m." This shark eats bony fishes. Carcharhinus plumbeus (Nardo, 1827) Figure 56 Horizon.—Pungo River Formation (units 4, 5); Yorktown Formation (unit 1). Referred Material.—6 specimens, USNM 459788, 474962^174965, 476293. Remarks.—The upper teeth of this species are of moderate width; they are narrower and more elongate than those of Car- charhinus obscurus, and they lack the apical convexity of the cutting edge that characterizes the latter species. They may be confused with the elongate teeth of C. albimarginatus, but in this last species the tips of the teeth are hooked, and midway between the tip and the root on both cutting edges there is a no- ticeable, shallow notch. After the seventh tooth from the sym- physis, the mesial notch is lost. An amateur collector, Bill Heim, first identified a tooth from Lee Creek Mine that matches perfectly with the first upper an- terolateral tooth in the dentition (Figure 56a) of the extant spe- cies. Five other teeth (Figure 566-e) were recovered from the spoil piles; in one of these, which may not be a fully functional tooth, the transverse groove is absent. The three teeth that are from the anterior part of the jaw mea- sure 1.4 cm, 1.6 cm, and 1.9 cm in height. Extant sandbar sharks with teeth of this size are about 2 m in total length. Compagno (1984:494) stated that the extant species is a "coastal-pelagic shark, of temperate and tropical waters, found on continental insular shelves and in deep water adjacent to them." It feeds on "relatively small bottom fishes" (Compagno, 1984:495), including porgies and searobins, and on mollusks and crustaceans. NUMBER 90 155 FIGURE 56.—Carcharhinus plumbeus, upper antero- lateral teeth: a, USNM 459788, lingual view; b, USNM 474962, lingual view; c, USNM 474963, lin- gual view; d, USNM 474964, lingual view; e, USNM 476293, lingual view. (Scale bars: a-d= 1.0 cm; e= 1.0 cm.) Rhizoprionodon? sp. Figure 57a-g HORIZON.—Pungo River Formation (units 1-5). Referred Material.—About 50 isolated teeth, USNM 207535-207540. REMARKS.—In the extant Rhizoprionodon, Compagno (1988:295-296) noted the following: "Dignathic heterodonty weak and best developed in anteroposteriors closest to the sym- physis, these teeth having higher roots and crowns and more arched root edges in the upper jaw than in the lower. Upper lat- erals have slightly higher crowns, slightly stouter cusps, and less concave mesial edges than the lowers. Tooth row groups include small, weakly differentiated, and often double-bladed and erect-cusped medials, 2-3 rows of weakly differentiated upper anteriors, sometimes a lower symphyseal, and poorly differentiated posteriors. The anteriors differ from adjacent lat- erals in having narrower and higher crowns__Fine serrations present on teeth of adults of larger species__Tooth roots and crowns relatively low and deep. Teeth holaulacorhizous, with transverse notches and grooves. Basal ledges and grooves ab- sent or poorly defined on teeth." Springer (1964) revised and redivided the genus Scoliodon into three genera: Rhizoprionodon, Loxodon, and Scoliodon sensu stricto. Each has very similar teeth, all of which are con- vergent or parallel to those of some species of Sphyrna. Conse- quently, Springer was pessimistic about identifying the isolated teeth of these genera. Compagno (1984, 1988) included tooth characters in his di- agnoses for the species of Rhizoprionodon, but with one ex- ception, he also listed these as characters for Scoliodon and Loxodon. The one exception is the presence of serrations in the teeth of Rhizoprionodon. The Lee Creek Mine teeth, which have smooth or weakly serrated cutting edges, are iden- tical in form to those Compagno (1988, pl. 22G, 23G) illus- trated, but until more is known about the usefulness of these dental characters, we assign fossil teeth to this genus only ten- tatively. These fossil teeth (Figure 57a-g) are characterized by a com- bination of concave mesial cutting edges both on upper and lower teeth, smooth or indistinctly serrated cutting edges, deep- ly notched distal cutting edges, erect crowns, and distinct enamel shoulders. These characters, except for the notch on the distal enamel shoulder, are shared more or less by all species of Sphyrna with nonserrated teeth. Teeth from the lower jaw of these species of Sphyrna (e.g., 5". tiburo) are similar to those from the upper jaw of Rhizoprionodon, but differences exist in their proportions. Relative to their width, the crowns in sphyrnid teeth are higher than those of Rhizoprionodon, and the primary cusp is relatively wider. The Lee Creek Mine teeth closely resemble the more distinctive lower teeth of the extant Rhizoprionodon terraenovae. We found no certain criteria to eliminate Loxodon or Scoliodon. European Miocene representatives of the Scoliodon group have been referred to Rhizoprionodon taxandriae (Leriche, 1926; Antunes and Jonet, 1969-1970; Cappetta, 1970). Ler- iche's (1926, pl. 28: figs. 7-10) specimens of Carcharias (Scoliodon) taxandriae are poorly preserved, and certainly not all of them are Scoliodon sensu lato. The specimen in Leriche's fig. 7 is indeterminate; Cappetta (1970:62) thought it was Gale- orhinus affinis. Leriche's fig. 10 probably represents a small sphyrnid, and the specimens in his figs. 8 and 9 are either upper teeth of Rhizoprionodon or lower teeth from a smooth-toothed sphyrnid. Until these latter two specimens can be compared to the teeth of the extant species of Rhizoprionodon and Sphyrna, the validity of R. taxandriae remains in doubt. Three of the fossil teeth are from the more anterior portion of the jaw; they range in height from 3.2 to 5.2 mm (mean=4.1) and in width from 4.3 to 5.7 mm (mean=4.3). The extant species, Rhizoprionodon terraenovae, inhabits coastal warm-temperate to tropical seas from the intertidal zone to a depth of possibly 280 m (Compagno, 1984:533); it feeds on small bony fishes, including small jacks and tilefish. Genus Negaprion Whitley, 1940 (lemon sharks) Compagno (1988:342) characterized the teeth of this genus as follows: "Tooth row groups include upper and lower medials, lower symphysials [sic], upper anteriors, upper lateroposteriors, and lower anteroposteriors (or upper laterals, lower anterolater- als, and lower posteriors). Anteriors differ from adjacent later- als only in being smaller and narrower relative to their heights. Sexual heterodonty apparently little developed. Ontogenic het- 156 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY erodonty moderate: small specimens of 500-600 mm. total length have more oblique cusps, weaker mesial blades, and lack serrations on the blades of their upper lateroposterior teeth, but larger examples between 1-2 meters length have more erect cusps, stronger mesial blades and basal serrations.... Cusplets present on upper teeth of individuals below 650 mm. but lost in larger specimens. Roots of lower anterolateral teeth with nearly straight to slightly arched ventral edges. Teeth holaula- corhizous, with very weak transverse grooves and transverse notches weak or absent. Basal ledges very weak but faintly indi- cated even in teeth of adults, sometimes with nodular edge rep- resenting vestigial transverse ridges." Negaprion eurybathrodon (Blake, 1862) Figure 57/i-y Carcharias collata Eastman, 1904:85-86, pl. 32: figs. 3-5 [Miocene, Mary- land], Carcharias magna (Cope).—Eastman, 1904:86, pl. 32: figs. 6, 7 [Calvert For- mation, Maryland]. Sphyrna magna (Cope).—Leriche, 1942:85 [citation]. lAprionodon cf. collata (Eastman).—Leriche and Signeux, 1957:35, pl. 11: figs. 12, 13 [Miocene, France]. Negaprion cf. eurybathrodon (Blake).—Antunes and Jonet, 1969-1970: 175-177, pi. 13: figs. 80, 81 [middle Miocene]. Negaprion kraussi (Probst).—Cappetta, 1970:52-53, pl. 15: figs. 1-10, ?11, 12, 14-17 [early Miocene, France]. Horizon.—Pungo River Formation (unit 4 or 5; 1 tooth); Yorktown Formation (units 1, 2). Referred Material.—31 teeth, USNM 207490-207492, 451333-451340. Remarks.—Blake (1862) described a Miocene tooth of this species as that of Lamna eurybathrodon. White (1955) ob- served that the carcharhinid tooth recorded by Blake was ac- tually a Negaprion upper tooth and that teeth of the same gen- eral form were widely distributed in Miocene deposits. In the same paper he noted correctly that the teeth described by Eastman (1904) as Carcharias collata were the lower teeth of Negaprion, and that the teeth figured by Eastman (1904) and referred to C. magna (Cope) also should be referred to Negap- rion, but he hesitated to synonymize C. magna and N. eury- bathrodon. More recently, Antunes and Jonet (1969-1970) have rightly synonymized the two species. Cappetta (1970:52) referred his specimens of Negaprion to N. kraussi (Probst), but Antunes and Jonet (1969-1970:176) pointed out that the specimens that Probst described as Carcharias (Scoli- odon) kraussi were in poor condition and may be either Sphyrna or Negaprion. If they are Negaprion they are a junior synonym of N. eurybathrodon; thus TV. kraussi of Cappetta, but not necessarily C. (Scoliodon) kraussi Probst, is a junior synonym of N. eurybathrodon (Blake). If, however, these teeth are identical to those of TV. brevirostris (Poey), as they appear to be, N. brevirostris must be synonymized with Negaprion eurybathrodon. The anterolateral teeth from Lee Creek Mine range in size from 1.4 to 2.1 cm (mean=1.6 cm, «=9). In the extant lemon shark, teeth in this size range are found in sharks of 2.1-3 m TL, which is the size range for mature adults (Compagno 1984:520). According to Compagno (1984:519-520), this shark inhab- its inshore, tropical waters from the intertidal zone down to at least 92 m; it feeds on bony fishes, including sea catfishes, jacks, and porcupine fishes, on stingrays, occasionally on sea- birds, and on invertebrates. Genus Triaenodon Miiller and Henle, 1837 (whitetip reef sharks) In the extant species of this shark, the upper teeth can be dif- ferentiated into two medials, one symphysial, three anteriors, 15 to 18 laterals, and one or two posteriors; in the lower jaw, aside from two medials, one symphysial, and (sometimes) one or two posteriors, anterior and lateral teeth cannot be differenti- ated (Compagno 1988:352). When the posterior teeth can be identified, the remaining lower teeth are called anterolaterals, and when the posterior teeth cannot be differentiated, they are called anteroposteriors. Compagno (1988:352) characterized the teeth of this genus as follows: "Sexual heterodonty apparently weak or absent. Ontogenic heterodonty moderately developed, with adults hav- ing narrower and more differentiated medials than young. Young also have mesial cusplets absent from a few more rows of teeth adjacent to the ends of the dental bands than adults and have slightly more oblique cusps." Returning to the description of adult teeth, he continued, "Mesial edges differentiated into cusplets on most teeth except for a variable number of rows near the ends of the dental bands. Primary cusps present on all teeth, erect towards the symphysis but becoming oblique to- wards the ends of the dental bands. Primary cusps narrow- based and high. Distal edges strongly notched on all teeth, but serrations are absent. Tooth roots and crowns relatively high and moderately compressed. Teeth holaulacorhizous, with well-developed transverse grooves and notches. Basal ledges and grooves absent from teeth." Triaenodon obesus (Riippell, 1835) Figure 57A-n Otodus catticus Philippi, 1846:24, pl. 2: figs. 5-7 [early Miocene, Cassel], Lamna cattica (Philippi).—Leriche, 1926:395-397, pl. 28: figs. 50-52 ["Bold- erian," Belgium]; 1927:65-68, pl. 7: figs. 12-15, 17, 18 [Miocene, Switzer- land].—Cappetta, 1970:23-25, pl. 4: figs. 5-8 [Helvetian, France]. Horizon.—Pungo River Formation (units 1-3). Referred Material.—5 teeth, USNM 283497, 312434, 312436,459824,459825. Remarks.—The five teeth recovered from the Pungo River Formation (Figure 51k-n) are the typical upper lateral teeth of Triaenodon. These teeth are extremely labiolingually com- pressed and have tall, distally inclined and broad-based crowns that are bordered by lingually bent, large, lateral cusplets; at the NUMBER 90 157 FIGURE 57.—Rhizoprionodon? sp.: a, USNM 207535, lower lateral tooth, lingual view; b, same specimen, labial view; c, USNM 207536, lower anterior tooth, lingual view; d, USNM 207537, lower anterior tooth, labial view; e, USNM 207538, lower anterior tooth, labial view;/ USNM 207539, male? lower anterior tooth, lin- gual view; g, USNM 207540, lower lateral tooth, lingual view. Negaprion eurybathrodon: h, USNM 207490, upper anterior tooth, labial view; i, USNM 207491, upper lateral tooth, lingual view;/ USNM 451338, lower lateral tooth, lin- gual view. Triaenodon obesus: k, USNM 312436, upper lateral tooth, labial view; /, USNM 312434, upper lateral tooth, lingual view; m, USNM 459824, upper anterolateral tooth, lingual view; n, USNM 459825, lower lateral tooth, lingual view. (Scale bars: a-g,I = 0.5 cm; h-k,m = l.O cm; n=1.25 cm.) base of the crown there is a basal ledge. On the lingual face of the root, a transverse groove is usually present. Philippi's holotype (1846, pl. 2: figs. 5-7) of Otodus cattica is identical to an upper lateral of Triaenodon obesus, with its extremely labiolingually compressed, distally inclined crown, large lateral cusplets, and shallow root. A transverse groove, which Philippi did not mention, is not evident in his fig. 6, but if this tooth was not functional or if the groove was not well de- fined by its tooth germ, the groove would be absent. Antunes (1969-1970) used teeth from two genera to recon- struct a tooth set of Lamna cattica totuserrata from the Mi- ocene of Angola. The anterior teeth illustrated by Antunes (1969-1970, pl. 1: fig. 1, pl. 2: figs. 17-19, pl. 3: figs. 20-24) are identical to the upper and lower anterior teeth of Carchar- ias cuspidata, and the lower lateral teeth (his pl. 3: figs. 25, 26) also are identical to those of Carcharias cuspidata. Those that Antunes (1969-1970) identified as upper symphysial teeth, which are symphysial (his pl. 1: fig. 2) and intermediate (his pl. 1: fig. 3) teeth, also belong to C. cuspidata. (Some of these teeth bear on their cutting edges what appear to be in- cipient serrations, which are not as well developed as those on the other teeth in his reconstruction that are assignable to Triaenodon. The cutting edges of the functional teeth of the extant Carcharias taurus are often nicked, giving them the appearance of bearing incipient serrations.) The specimens Antunes (1969-1970, pl. 2: figs. 10-16) identified as upper lateral teeth, except for their finely serrated cutting edges, are identical to those of Triaenodon obesus. Antunes's recon- structed tooth set, therefore, includes teeth from two different genera. We do not think that serrations warrant separating the fossil teeth from Triaenodon because at least two genera in the Car- charhiniformes, Carcharhinus and Sphyrna, contain species with smooth and serrated teeth. We believe, therefore, that ser- rations are not a valid criterion for assigning these teeth to a ge- nus other than Triaenodon. On the basis of Antunes's reconstruction, Cappetta (1987:95) assigned Lamna cattica and L. totuserratus to the genus Carcharoides Ameghino, 1901. In view of the above, we believe this assignment is unwarranted. The referred teeth measure as follows (USNM 312434 is in- complete). Catalog number Height (mm) Width (mm) USNM 283497 10.2 10.1 USNM 312436 10.2 8.9 USNM 459824 13.4 13.2 USNM 459825 6.5 7.2 Compagno (1984:537) reported that the extant species is a bottom-oriented shark that inhabits tropical, shallow, clear wa- ter on or near coral reefs; it usually is found in waters of 8 to 40 m in depth, but it has been found occasionally as deep as 110 to 330 m. Family Sphyrnidae (hammerhead sharks) Genus Sphyrna Rafinesque, 1810 Compagno (1988:363) characterized these teeth as having weak ontogenic heterodonty, with the heights of the cusps and crowns relative to root width lower in juveniles than in 158 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY adults. Under the subgenus Sphyrna, which includes all of the Lee Creek Mine hammerheads, Compano (1988:366) de- scribed the teeth as follows: "Cusps of anterolateral teeth [lowers] slender and almost needlelike. Posterior teeth mostly cuspidate, not modified as molariform crushers. Basal ledges and grooves obsolete." Three species of hammerhead sharks occur at Lee Creek Mine, Sphyrna lewini, S. cf. S. media, and S. zygaena. ¦ H e" f^^ Sphyrna lewini (Griffith and Smith, 1834) Figure 58a-c Horizon.—Yorktown Formation (unit 3). Referred Material.—4 teeth, USNM 459728-459731. Remarks.—The teeth from Lee Creek Mine (Figure 58a-c) fit Compagno's (1984:545) description of teeth from the extant species: "Anterior teeth with moderately long, stout to slender cusps, smooth or weakly serrated, posterior teeth mostly cuspi- date and not keeled and [not] molariform." The anterior teeth of this shark range from 7.3 to 7.8 mm in height (mean=7.5 mm, n=4) and from 6.5 to 9.6 mm in width (mean=8.2 mm, «=4). Compagno (1984:546) reported that this species is "probably the most abundant hammerhead, a coastal-pelagic, semioceanic warm-temperate and tropical species occurring over continen- tal and insular shelves... [ranging] from the intertidal... to at least 275 m depth." He reported its principal prey as a wide va- riety offish, including sea catfish, bluefish, jacks, porgies, and other sharks. Sphyrna cf. S. media Springer, 1940 Figure 58<7-g -Pungo River Formation (units -5); Yorktown Horizon. Formation (units 1, 2). Referred Material.—111 teeth, USNM 207526-207529, 207533, 207534, 451341-451346. Remarks.—The upper teeth have triangular, distally in- clined crowns with smooth cutting edges (Figure 58o». The crowns are deeply notched distally, setting off a distinct, vaguely serrated, convex enamel blade. They are similar to those teeth referred to 1 Rhizoprionodon sp. but may be distin- guished by their wider-based crown, blunter point, and less concave mesial cutting edge. Of the sphyrnids, they resemble closely the teeth of Sphyrna media (Gilbert, 1967, fig. 14), but because dentitions of this species were not available to us for this study, we could not make direct comparisons between them and the fossil teeth. The teeth identified by Cappetta (1970:70-72, pl. 19: figs. 1-18) from the lower Miocene of France also closely resemble those of S. media. In size, the anterior teeth range from 4.8 to 7.8 mm in height (mean=6.3 mm, n=l) and from 5.6 to 8.0 mm in width (mean =7.8 mm, n=l). FIGURE 58.—Sphyrna lewini: a, USNM 459728, lower lateral tooth, lingual view; b, USNM 459729, upper lateral tooth, lingual view; c. USNM 459730, lower lateral tooth, lingual view. Sphyrna cf. S. media: d, USNM 207527, upper lateral tooth, lingual view; e, USNM 207534, upper lateral tooth, labial view;/ USNM 207529, lower anterior tooth, labial view; g, USNM 207533, lower lateral tooth, lingual view. (Scale bars: a-c= 1.0 cm; d-g=0.ss cm.) According to Compagno (1984:548), Sphyrna media, a little- known, inshore, tropical hammerhead, is found in the eastern Pacific Ocean, the southern Caribbean Sea, and the south At- lantic Ocean. The Lee Creek Mine occurrence, if it is Sphyrna media, indicates that this shark had a wider distribution during the Neogene. Sphyrna zygaena (Linnaeus, 1758) Figures 59, 60 Galeocerdo laevissimus Cope, 1867:141-142 [Calvert Formation, Maryland]. Carcharias laevissimus (Cope).—Eastman, 1904:84-85, pl. 32: fig. 2 [Calvert Formation, Maryland]. Carcharias (Scoliodon) kraussi Probst.—Leriche, 1927:83, pl. 14: fig. 16 [Mi- ocene, Switzerland] Sphyrna laevissima (Cope).—Leriche, 1942:84, pl. 7: figs. 23-27 [Calvert For- mation, Maryland].—Casier, 1958:40, pl. 1: fig. 23 [Miocene, Trinidad]. HORIZON.—Pungo River Formation (units 1-5); Yorktown Formation (units 1-3). REFERRED MATERIAL.—55 teeth, USNM 207520-207525, 459716-459719, 459721-459729. Remarks.—The Pungo River specimens (Figure 59a-e) are identical in size and form to teeth of juveniles and adults of the extant Sphyrna zygaena. In juveniles and young adults of the extant species, the teeth have smooth cutting edges, but in large individuals, they become weakly serrated (Gilbert, 1967:36). On the basal portions of their cutting edges, several teeth from the Pungo River Formation exhibit incipient serra- tions. These are similar in form to those found on the upper teeth from a 302 cm TL individual of the extant species (USNM 232633). Cope's (1867) type suite for S. laevissimus consists of 19 teeth (ANSP 1195-1213) (Figure 60), two of which are not Sphyrna. The remaining teeth compare favorably with those of 5. zygaena; therefore, we place S. laevissimus in synonymy NUMBER 90 159 FIGURE 59.—Sphyrna zygaena: a, USNM 207522, Pungo River Formation, lower anterolateral tooth, labial view; b, USNM 207524, Pungo River Formation, lower anterolateral tooth, labial view; c, USNM 207520, Pungo River Formation, upper lateral tooth, lingual view; d, USNM 207523, Pungo River Formation, incom- plete upper lateral tooth, labial view showing distinctive shape of mesial edge of larger teeth; e, USNM 207525, Pungo River Formation, upper lateral tooth, labial view;/ USNM 459716, Yorktown Formation, unit 3, upper lateral tooth, lingual view. (Scale bars: a-c,e=\.0 cm; <7=0.5 cm;/=0.33 cm.) with it. Although cataloged, the teeth were never numbered; in Figure 60 we give to each specimen a catalog number from the assigned series. The few specimens from the Yorktown Formation (Figure 59/) (mean height=1.3 cm, range= 1.2-1.4 cm, n=4), which are larger than the Pungo River Formation specimens (mean height=1.0 cm, range=0.8-l.l cm, «=10), are more serrate than those from the latter formation. FIGURE 60.—Type suite oi Sphyrna laevissimus: a, ANSP 1195, upper antero- lateral tooth, lingual view; b, ANSP 1196, lower anterolateral tooth, lingual view; c, ANSP 1197, lower anterolateral tooth, lingual view; d, ANSP 1198, lectotype, upper anterolateral tooth, lingual view; e, ANSP 1200, upper antero- lateral tooth, lingual view;/ ANSP 1201, upper anterolateral tooth, lingual view; g, ANSP 1202, upper anterolateral tooth, lingual view; h, ANSP 1203, upper anterolateral tooth, lingual view; i, ANSP 1204, lower anterolateral tooth, lingual view;/ ANSP 1205, upper anterolateral tooth, lingual view; k, ANSP 1206, upper anterolateral tooth, lingual view; /, ANSP 1207, upper ante- rolateral tooth, lingual view; m, ANSP 1208, upper anterolateral tooth, lingual view; n, ANSP 1209, upper anterolateral tooth, lingual view; o, ANSP 1210, lower anterolateral tooth, lingual view; p, ANSP 1211, lower anterolateral tooth, lingual view; q, ANSP 1213, lower anterolateral tooth, lingual view. (Scale bar= 1.0 cm.) 160 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY The extant species is a coastal-pelagic and semioceanic shark that inhabits temperate to tropical waters; it feeds on bony fish- es, including sea catfishes, sea bass, and porgies, and on sting- rays (Compagno, 1984:554). CARCHARHINIFORMES Figure 61 HORIZON.—Yorktown Formation (units 1?, 2?). Referred Material.—About 100 rostral nodes, USNM 476299^176302; about 20 postorbital and preorbital processes, USNM 476303-476306; 6 hypercalcified ?suprascapulae, USNM 476307, 476308. REMARKS.—The Lee Creek Mine carcharhiniform rostral nodes occur in three forms: broad dish-shaped (USNM 476299, Figure 61a), T-shaped (USNM 476301, Figure 61*), and bar- shaped (USNM 476302, Figure 6^,0"). In the first form, the two dorsal and one ventral rostral cartilages meet anteriorly to form a broad, unperforated, triangular dish. In the second form, in cranial or caudal view, the rostral cartilages form a "T," which may be perforated at the point of juncture of the rostral cartilages. In the last form, the two dorsal rostral cartilages meet anterior to the ventral one, forming a lateral bar in anteri- or view. Anterior to the juncture with the ventral rostral carti- lage, the dorsal cartilages are perforated. According to Compagno (1988:69), these types of rostral nodes occur in Mustelus mosis, Rhizoprionodon acutus, and Carcharhinus macloti, but without comparative material, we could not identify the Lee Creek Mine specimens, which may be too large to belong to these species. Hypercalcified postorbital (Figure 61c) and preorbital pro- cesses (Figure 61/g) similar to those found in the extant Rhizoprionodon and Mustelus also occur at Lee Creek Mine; however, these specimens are from individuals larger than the largest extant form, and the Lee Creek Mine teeth for these genera fall within the size range for the extant forms. Because of this size discrepancy and the lack of comparative material at this time, we cannot identify the Lee Creek Mine speci- mens. Finally, some bulbous specimens, which may represent su- prascapulae, also were recovered from the mine (Figure 6I/1). The largest, USNM 476307, has a maximum diameter of 5.3 cm. Again, due to the lack of comparative material, we could not identify these. FIGURE 61.—Carcharhiniform: a, USNM 476299, ros- tral node, ventral view; b, USNM 476301, rostral node, ventral view; c, USNM 476302, rostral node, ventral view; d, same specimen, ventral view; e, USNM 476306, postorbital process, dorsal view;/ same speci- men, ventral view; g, USNM 476303, preorbital pro- cess, dorsal view; h, USNM 476307, hypercalcified ?suprascapula, lateral view. (Scale bar= 1.0 cm.) NUMBER 90 161 Class Holocephali Order Chimaeriformes, family indeterminate Figure (>2a-d Horizon.—Pungo River Formation? REFERRED MATERIAL.—1 dentary fragment, USNM 282334; 1 tritor, USNM 476310; 1 partial dorsal spine, USNM 476309. REMARKS.—Only three specimens of chimaeroid have been recovered from the Lee Creek Mine; none of these possess any characters that would allow their identification to genus. The dentary fragment (Figure 62a), which is the anterior portion and is badly abraded, lacks any evidence for the positions of the tritors. The tritor (Figure 62b) is 7.3 cm long and 1.7 cm wide. In cross section the dorsal spine is triangular and has a dou- ble row of serrations caudally, which are separated by a groove; the cranial edge is smooth and sharp (Figure 62c,d). Its lateral surfaces are striated. This fragment is 3.9 cm in length and 1.5 cm in width. Chimaeroids are deepwater fishes that feed on invertebrates and small fishes. Class Osteichthyes Bone terminology herein follows Collette and Russo (1984) and Tyler (1980). Order ACIPENSERIFORMES Family Acipenseridae (sturgeons) Acipenser cf. A. oxyrhynchus Mitchell, 1814 Figures 62e-h, 63 HORIZON.—Yorktown Formation (units 1, 2). REFERRED Material.—Several thousand fragments, in- cluding bony scutes, USNM 207602, 207603, 207607,286998, 286999, 290533, 290645; pectoral spines, USNM 207604, 284823, 284908; neural spines, USNM 285375, 285384; skull bones, NCSM 9045, USNM 464059. Remarks.—The fossil sturgeon material from Lee Creek Mine consists of thick bony scutes or dermal plates with deeply ornamented outer surfaces (Figure 62c,/) and smooth inner surfaces. The pectoral spine (Figure 62g,h) is striated along its length and is ornamented at its proximal end. Among the skull bones we identified an intertemporal-supratemporal (Figure 63a) and a suborbital (Figure 63b). From the Atlantic Coastal Plain of the United States, fossil sturgeon remains have been reported from the Miocene Calvert Formations of Virginia (Leidy, 1873) and Maryland (Kimmel and Purdy, 1984) and from the Pleistocene of Florida (Swift and Wing, 1968). Based on a lateral scute from the Virginia Miocene, Leidy (1873) established a new species of fossil stur- geon, Acipenser ornatus. About this specimen, Leidy (1873:350) said, "Though exhibiting no positive distinctive character, it most probably pertained to a species now extinct." The type specimen was in a private collection and is now lost, but based on Leidy's description and illustration (Leidy, 1873, pl. 32: fig. 58), the scutes of Acipenser ornatus do not differ from those of Acipenser oxyrhynchus. In morphology, the Lee Creek Mine scutes are very similar to USNM specimens from the Miocene sediments of Virginia, Maryland, and Delaware and to those of the extant Atlantic sturgeon, Acipenser oxyrhynchus. They closely resemble those from a 183 kg specimen of the extant species (USNM 260347) except that the Lee Creek Mine specimens are more massive. This difference in thickness between the fossil and modern der- mal elements, however, may fall within the size range of Aci- penser oxyrhynchus, which reaches a length of 4.2 m TL and a weight exceeding 370 kg (Manooch, 1984). Acipenser oxyrhynchus is an anadromous species that occurs in the shallow waters of the Atlantic continental shelf from La- brador south to the northern coast of South America (Robins and Ray, 1986). Presently, it occurs off North Carolina in shal- low (to 20 m), nearshore waters from the late fall to the early spring (Holland and Yelverton, 1973). Ross et al. (1988) stated that in North Carolina in the spring, the American Atlantic sturgeon is common in the ocean near the mouth of the Cape Fear River. The abundance of sturgeon fossils in the Lee Creek Mine fauna suggests that this fauna may have existed near the mouth of a large river. The American Atlantic sturgeon is a bottom feeder, consum- ing worms, crabs, and small fishes (Manooch, 1984). Order LEPISOSTEIFORMES Family Lepisosteidae (gars) Lepisosteus osseus (Linnaeus, 1758) Figure (Aa-d Horizon.—James City Formation. Referred Material.—1 partial skeleton, USNM 279492. Remarks.—A partial skeleton consisting of eight vertebrae, a rostral fragment, and numerous scales are referred to Lepisos- teus osseus. The vertebrae are identical to those of the extant gar. Lepisosteus osseus, a freshwater and estuarine species, oc- curs on the Atlantic slope of the United States from New Jersey to Florida; it is also found in the Mississippi, Great Lakes, and Rio Grande drainages (Page and Burr, 1991). The gar feeds primarily on bony fishes, such as shiners, sun- fish, shads, and catfish (Manooch, 1984). 162 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE 62—Chimaeroid: a, USNM 282334, fragment of dentary, lateral view; b, USNM 476310, tritor, lateral view; c, USNM 476309, fragment of dorsal spine, lat- eral view; d, same specimen, caudal view. Acipenser cf. A. oxyrhynchus: e, USNM 207603, lateral dermal scute, external view;/ USNM 207607, ventrolateral dermal scute, external view; g, USNM 207604, incomplete pec- toral spine, dorsal view; h, same specimen, posterior view. (Scale bar= 1.0 cm.) Order Elopiformes Family Elopidae (tarpons) Megalops cf. M. atlanticus Valenciennes in Cuvier and Valenciennes, 1847 Figure (Aef HORIZON.—Pungo River Formation (units 5, 6). Referred Material.—9 large, worn vertebrae, USNM 290630, 291256, 459863-459867,476382, 476394. Remarks.—The Lee Creek Mine specimens, which range from 41.5 to 47.5 mm in length, closely resemble the vertebrae of the living tarpon, Megalops atlanticus. Like the extant spe- cies, the vertebral lateral surfaces are slightly concave (Figure 64/) and are striated. In axial view (Figure 64c) the vertebrae are oval, with the lateral axis being slightly greater than the dorsoventral one. Megalops atlanticus is a nearshore species that occurs from Virginia to Brazil (Robins and Ray, 1986). It feeds primarily on crabs and bony fishes, such as sardines, anchovies, mul- lets, silversides, hardhead catfish, and Atlantic Cutlassfish (Manooch, 1984). Order Anguilliformes Family Congridae (conger eels) Conger cf. C. oceanicus (Mitchill, 1818) Figure 65a,b HORIZON.—Yorktown Formation (unit 1). NUMBER 90 163 FIGURE 63.—Acipenser cf. A. oxyrhynchus: a, USNM 464059, partial left intertemporal-supratemporal, dorsal view; 6, NCSM 9045, partial right suborbital, dorsal view. (Scale bars= 1.0 cm.) f FIGURE 64.—Lepisosteus osseus, portions of partial associated skeleton, USNM 279492: a, partial premaxilla, labial view; b, vertebra, dorsal view; c, same specimen, articular view; d, scale. Megalops cf. M. atlanticus: e, USNM 290630, vertebra, articular view; / same specimen, dorsal view. (Scale bars: o=0.9 cm; b-d=0.75 cm; e,/=0.9 cm.) Referred Material.—1 partial left dentary without teeth, USNM 411948. Remarks.—This dentary compares favorably with that of the extant Conger oceanicus. Along its labial edge, the position of the main tooth row is indicated by oval alveoli, and lingual to the alveoli there is a row of alveoli for villiform teeth (Figure 65a). At its anterior end, the dentary widens labially to form a small tooth patch with 12 small alveoli in two rows. In this area 164 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY of the dentary, several extant specimens of Conger oceanicus (USNM 30710, 265105, 265109) exhibited variable numbers of tooth rows (one to four). Kanazana (1958) stated that the number of teeth in the jaw of the genus Conger increases di- rectly with age. The Lee Creek Mine specimen may be from a juvenile or a young adult. Conger oceanicus is a benthic species that occurs from Cape Cod south to the Gulf of Mexico (Robins and Ray, 1986). This species has been recorded in depths from 1 to 520 m (Kanaza- na, 1958). It feeds primarily on bony fishes (Hildebrand and Schroeder, 1928) but also has been reported feeding on shrimp and mollusks (Thomson et al., 1978). Order Clupeiformes Family Clupeidae (shads and herrings) Alosa cf. A. sapidissima (Wilson, 1811) Figure 65c,d Horizon.—Yorktown Formation (unit 1?). Referred Material.—Partial skull, USNM 411950. REMARKS.—This partial skull has a posteriorly extending, enlarged epiotic as seen in the genus Alosa. The posterior ex- tent of the epiotic is greater in the Lee Creek Mine specimen and in A. sapidissima than it is in the other species of Alosa. Also, the epiotics intersect at the midline of the skull to form a 90° V-shaped notch. In other species of Alosa (A. aetivalis, A. pseudoharengus, A. mediocris) these notches are greater than 90°. In this specimen, the preepiotic forms a depression at the side of the cranium, bounded by the epiotic, squamosal, and parietal bones, which also is characteristic of clupeids (Ridge- wood, 1904). Alosa sapidissima is an anadromous species that occurs from Newfoundland and the Gulf of St. Lawrence south to northern Florida (Robins and Ray, 1986). They are found off North Carolina in nearshore waters (0-20 m), but they have been cap- tured at depths of 160 to 250 m (Holland and Yelverton, 1973). It is a planktivore, feeding primarily on small crustaceans, in- sects, fish eggs, and algae (Manooch, 1984). Order Siluriformes Family Ariidae (sea catfish) Bagre sp. Figure 65e,/ Horizon.—Pungo River Formation (units 4, 5); Yorktown Formation (units 1, 2). Referred Material.—30 pectoral-fin spines, USNM 256274, 284106, 284107,284111-284113,284151, 284152 (Pungo River Formation); 6 dorsal-fin spines, USNM 284098-284100, 476375 (Pungo River Formation, unit 4 or 5), USNM 284114 (Yorktown Formation); about 100 vertebrae, USNM 285646, 286142, 286147, 286161, 290517, 290947 (Yorktown Formation). Remarks.—The fossil pectoral-fin spines from Lee Creek Mine are similar to those of Bagre marinus (Mitchill), the gaff- FlGURE 65.—Conger cf. C. oceanicus: a, USNM 411948, partial left dentary without teeth, occlusal view; b, same specimen, lingual view. Alosa cf. A. sapidis- sima: c, USNM 411950, skull, dorsal view; d, same specimen, ventral view. Bagre sp.: e, USNM 256274, partial pectoral-fin spine, posterior view. Opsanus tau: f, USNM 411949, partial right dentary, occlusal view; g, same specimen, labial view; h, NCSM 4285, partial right dentary, occlusal view. Bagre sp.: /', USNM 476375, dorsal-fin spine, anterior view. (Scale bars: a-g,i= 1.0 cm; A=0.25 cm.) NUMBER 90 165 topsail catfish, especially with regard to the shape of the proxi- mal articulating surface. This surface, which articulates with the coracoid, is useful for identifying genera of catfishes. In Bagre, the central hinge process is flattened ventrally and later- ally to form the outside part of the hinge, and medially it rises to form the hook-like hinge (Figure 65e). This differs from the pectoral attachment found in the genus Arius, which does not have its medial process ventrally flattened or hooked. Several dorsal-fin spines from Lee Creek Mine, which are referable to Bagre, have an outer surface that is notched with rows of cross ridges, each ridge consisting of one or two peaks (Figure 65/'). The serrae are very worn on both the dorsal- and pectoral-fin spines. Among the numerous fossil fish vertebrae from Lee Creek Mine, we were able to identify those of Bagre. The precaudal vertebrae are easily recognizable by the smooth lateral surfaces of their centra, which are perforated by a large and a smaller foramen, and by the strong, caudally curving haemal arches on the ventral surfaces. Other occurrences of sea catfish remains from sediments of the Chesapeake Group include a spine attributed to Arius (Ler- iche, 1942) and a skull with a left utricular otolith (Lynn and Melland, 1939), the holotype of Bagre stauroforus (Lynn and Melland), both from the Plum Point Member of the Calvert Formation. The spine assigned to Arius is incomplete and is too worn for generic identification. Because the skull and otolith of Bagre stauroforus were not associated with any pectoral-fin spines, comparison with the Lee Creek Bagre pectoral-fin spines is not possible. Bagre marinus occurs from Cape Cod to Venezuela (Robins and Ray, 1986) in shallow coastal and bay areas and seasonally in estuaries (Boschung et al., 1983). It feeds primarily on crabs, supplemented by shrimp and fishes (Gudger, 1918). Order Batrachoidiformes Family Batrachoididae (toadfishes) Opsanus tau (Linnaeus, 1766) FIGURE 65/-A HORIZON.—Yorktown Formation (unit 1). Referred Material.—2 partial right dentaries, NCSM 4285, USNM 411949. Remarks.—Both of these partial right dentaries can be re- ferred to the extant oyster toadfish, Opsanus tau. The first specimen, USNM 411949 (Figure 65fg), which is 26 mm in length and broken at both ends, has a single row of alveoli stretching from the posterior end of the dentary to the anterior tooth patch, half of which is present. The second specimen, NCSM 4285 (Figure 65/z), is 28 mm in length and is broken only at the distal end. Unlike the first specimen, several super- numerary teeth occur on either side of the single tooth row at two locations. Several of the teeth are displaced from the line of this row. Other than these differences, the dentaries are iden- tical to those of the extant species. Ray et al. (1968) reported Opsanus sp. from the Pleistocene Kempsville Formation of Virginia; these specimens, the first fossil record for Opsanus, consist of a neurocranium and a right lower jaw and compare favorably with those from Lee Creek Mine. The Lee Creek specimens extend the fossil record for the genus Opsanus to the early Pliocene. Opsanus tau is a benthic species that occurs from Cape Cod south to Florida (Robins and Ray, 1986). The primary habitats for this species are areas that are rocky or have some form of reef habitat. It has been recorded in depths of 1 to 50 m (Ma- nooch, 1984). There is some evidence that this species migrates offshore during cold weather (Thomson et al., 1978). It feeds primarily on crabs, fishes, shrimp, amphipods, and worms (Manooch, 1984). Order Lophiiformes Family Lophiidae (goosefishes) Lophius cf. L. americanus Valenciennes, 1837 Figure 66a-d HORIZON.—Yorktown Formation (unit 1?). Referred Material.—21 dentaries, USNM 290214, 290625, 291101, 476362; 50 premaxillaries, USNM 291116, 295340; 3 palatines, USNM 285259, 291188, 476361; 4 verte- brae, USNM 290493. Remarks.—In both size and shape of the various elements, the Lee Creek Mine specimens are very similar to those of Lophius americanus. On the lingual ventral surface at the sym- physial end, the fossil premaxillaries (Figure 66a) bear several rugose, caniniform teeth. These teeth point lingually, and they are followed distally by up to 12 smaller, caniniform teeth with triangular bases that decrease in size distally. The labial ventral surface has four to five large alveoli; after a short gap, these are followed distally by a row of smaller, triangular teeth. At the symphysial end of the premaxillary, the articulating surface is square-shaped and slightly concave, whereas the articulating process is blade-like and nearly perpendicular to the ventral and articulating surfaces; this process is heavily striated on its lingual and labial surfaces. The fossil dentary (Figure 66b) has a central tooth row of ca- nine-like teeth with triangular bases. On the labial edge of the dorsal surface, a row of smaller teeth originates about one- quarter of the distance from the symphysial end and extends to the distal end of the dentary. On the dorsal surface, the numer- ous alveoli give a pitted appearance to the bone. At the sym- physial margin of the dentary, there is a ventral, rectangular process, which is striated on its distal surface. The fossil palatine (Figure 66c) has several rugose, canini- form teeth followed by a row of smaller caniniform teeth. The symphysial surface is squarish, and the ethmoid process has two 166 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY tooth-like projections that point lingually. The fossil vertebrae (Figure 66a") that we assign to Lophius are the anteriormost ver- tebrae in the vertebral column; they are dorsoventrally com- pressed, with striations on the external surface of the centra. The genus Lophius was recorded previously from the Pleis- tocene of Virginia (Ray et al., 1968) and the Oligocene of Bel- gium (Leriche, 1908, 1910). The Virginia specimens, a frag- mentary left dentary and a right scapulocoracoid, were identified to family only. The Belgian specimens were identi- fied as a new species of goosefish, Lophius dolloi, by Leriche (1908). Lophius americanus, a benthic species, occurs from Quebec south to northern Florida (Robins and Ray, 1986). It inhabits areas with sand, mud, and broken-shell substrates in waters ranging in temperature from 0°C to 24°C (Manooch, 1984). The goosefish has been recorded at depths of 1 to 200 m (Big- elow and Schroeder, 1953). It feeds on small sharks, skates, eels, herring, weakfish, tautog, butterfish, puffers, cod, had- dock, flounders, seabirds, lobsters, crabs, worms, shellfish, sand dollars, and starfish (Manooch, 1984). Order Gadiformes Family Merluccidae (hakes) Merluccius bilinearis (Mitchill, 1814) Figures 66e-/, 67 HORIZON.—Yorktown Formation (units 1, 2). Referred Material.—About 1000 dentaries, USNM 285254, 285273, 285291, 286875, 290546, 290553, 290561, 290584, 290603, 290613, 290617, 290635, 290640, 290660, 291102,298293,298302, 476311-476313; about 1000 pre- maxillae, USNM 290543, 290545, 290602, 290606, 290644, 290656, 291096, 291123, 291193, 291673, 476324-476326; 1 partial maxilla, USNM 476363; about 100 maxillae 476314^176320; 1 partial left angular, USNM 476321; 1 left angular, USNM 459851; 1 associated dentary, maxilla, and partial premaxilla, USNM 476322; 1 vomer, NCSM 10916; several hundred vertebrae, USNM 286131-286135, 286144, 286145, 286160, 286162, 290222-290229, 290231, 290238, 290240, 290242, 290243, 290251, 290253, 290255, 290256, 290259, 290264, 290272, 290275, 290326-290328, 290497, 290517, 290518, 290559, 291109, 291687, 291694, 476323; about 100 hyperostosed vertebrae, USNM 460120, 476351, 476352; 1 partial cleithrum, USNM 459849; about 100 hyper- ostosed cleithra, USNM 283825, 283869, 283905, 283960, 284012, 284043, 284045; 1 urohyal, USNM 476364; about 100 hyperostosed urohyals, USNM 421524, 421525. Remarks.—The Lee Creek Mine specimens are identical to those of Merluccius bilinearis. The presence of biserial teeth on both the dentary and premaxilla separates the genus Merluc- cius from all other gadiforms, which have single or multiple rows of teeth on the maxilla, dentary, or both. Although Mer- luccius differs from most gadoids in having a prognathous low- er jaw (Rosen and Patterson, 1969), this feature is difficult to prove in unarticulated fossils. In the Lee Creek dentaries, as in the extant Merluccius, the alveoli of the biserial tooth rows, which support medium-sized, triangular teeth, occupy approximately one-third of the depth of the dentary (Figure 66ef). The tooth-row surface slants lin- gually so that the alveoli of the uppermost labial tooth row are higher than those of the lingual row. In the fossil specimens, the labial row of teeth is more securely attached to the dentary than is the lingual row of teeth, which is absent. On the ventral, FIGURE 66.—Lophius cf. L. americanus: a, USNM 295340, premaxillary, labial view; b, USNM 476362, fragment of dentary, labial view; c, USNM 476361, fragment of palatine, ventral view; d, USNM 290493, vertebra, lateral view. Merluccius bilinearis: e, USNM 476313, left dentary, external view;/ same specimen, internal view; g, USNM 476326, right pre- maxilla, external view; h, same specimen, occlusal view; /', USNM 476363, partial maxilla, dorsal view. (Scale bar= 1.0 cm.) NUMBER 90 167 FIGURE 67.—Merluccius bilinearis: a, USNM 459851, left angular, external view; b, NCSM 10916, vomer, ven- tral view; c, USNM 459849, partial cleithrum, lateral view; d, USNM 290559, precaudal vertebra, articular view; e, USNM 291109, precaudal vertebra, articular view;/ USNM 460120, four hyperostosed caudal verte- brae, dorsal? view; g, same specimen, axial view; h, USNM 476351, five hyperostosed caudal vertebrae, dor- sal? view; /. USNM 476364, urohyal, lateral view. (Scale bars: a,c~g,h=1.0cm; 6=0.33 cm; /'=0.5 cm.) lingual surface of the dentary, a ridge extends from a notch in the ventral edge of the symphysial end down the entire length of the dentary. The ridge is thickest just posterior to the notch and then thins abruptly. On the ventral, labial surface of the dentary, a large groove extends from above the ventral notch to the posterior end of the dentary. The largest of these incom- plete dentaries measures 5.8 cm in length and 1.5 cm in depth at the symphysis. In the premaxillae (Figure 66g,h) the lingual row of teeth originates at a slightly more dorsal elevation than does the labi- al row. The much shallower dorsoventral depth of the premax- illae gives the teeth, which are in the same size range as those of the dentaries, a larger appearance. At the symphysial end, the premaxilla flares out to form the slightly concave articular surface; on its lingual edge is a small process that articulates with the maxilla. The dorsal surface of the premaxilla is smooth and rounded, whereas the lingual and symphysial sur- faces are sculptured. The maxilla is edentulous (Figure 66;). Its symphysial end, which articulates with the symphysial end of the premaxilla, has an elongated internal process dorsal to the shallow groove that accepts the articular process of the premaxilla. The shal- lowness of this groove is unlike the glove-shaped socket found in the maxillae of other teleosts at Lee Creek Mine. Mesial to this shallow groove there is a process that extends toward the symphysis. The distal end of the maxilla is flattened on its ven- tral surface but with a slight concavity; the more mesial distal surface is smooth and rounded and has a half-oval cross sec- tion. The vomer, angular, cleithrum, urohyal, and vertebrae are figured (Figure 67). With the exception of the urohyal and the precaudal vertebrae, these compare well with those of Merluc- cius bilinearis. The urohyal is hyperostosed (Figure 67/), giv- ing it an inflated appearance, but its morphology is close to that of the extant species, which is not hyperostosed. Although the precaudal vertebrae (Figure 61d,e) are very similar to those of the extant species, they are more strongly ossified, which may be hyperostosis, and on their ventral surfaces a central groove rather than a central ridge is present. Numerous hyperostosed vertebrae (Figure 61f-h) occurring in the basal Yorktown Formation are tentatively assigned to Merluccius. Although the hyperostoses give the vertebrae a large size, the centra they surround have a very small diameter (Figure 61 g). In adults of the extant Merluccius productus, the precaudal vertebrae have diameters (9 mm) three times larger than that of the second caudal (3.0 mm), and the diameters of 168 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY the third (3.2 mm) and fourth (3.5 mm) caudal vertebrae are slightly greater than the diameter of the second caudal vertebra. Of the taxa we have identified at Lee Creek Mine, only Meluc- cius has caudal vertebrae with very small diameters (mean=5.0 mm, range=3.6-5.9 mm, n=l). The Lee Creek Mine speci- mens represent three to five caudal vertebrae fused by hyperos- tosis (Figure 61 h); the greatest development (mean width= 15.1 mm, range= 11.6-25.5 mm, «=8) of this hyperostosis is along the lateral axis of the vertebrae, and as shown below, its devel- opment is independent of the diameter of the vertebral centra. Centrum diameter (mm) Width of hyperostosis (mm) USNM 460120 5.7 12.2 USNM 476351 4.8 18.6 USNM 476352 5.8 25.5 Because specimens of extant Merluccius bilinearis with hy- perostosis were not available to us, we cannot ascertain the identity of these vertebrae. The hake is a benthic species that occurs from the Gulf of St. Lawrence south to South Carolina (Robins and Ray, 1986). It has been recorded at depths of 1 to 400 m, with a preferred depth range of between 60 and 200 m (Almeida et al., 1984). The hake feeds on fishes and invertebrates (Thomson et al., 1978). Order PERCIFORMES Family Triglidae (searobins) Prionotus cf. P. evolans (Linnaeus, 1766) Figure 68 Horizon.—Yorktown Formation (units 1-3). Referred Material.—5 skulls, NCSM 2978, USNM 290627, 459826, 459828, 459862; about 100 partial skulls, USNM 289396, 289397, 290215, 290216, 290663, 291154, FIGURE 68.—Prionotus cf. P. evolans: a, USNM 290627, neurocranium, dorsal view of skull roof; b, same spec- imen, ventral view; c, same specimen, lateral view; d, USNM 291218, dorsal view of hyperostosed skull roof; e, same specimen, ventral view;/ USNM 459827, ?pathologic, hyperostosed skull roof, dorsal view; g, same spec- imen, ventral view; h, USNM 290663, dorsal view of hyperostosed skull roof; /', same specimen, ventral view;/ USNM 412167, lateral view of left operculum; k, NCSM 3106, lingual view of partial right dentary. (Scale bars: a-c,y=0.5 cm; d-i= 1.0 cm; k=0.5 cm.) NUMBER 90 169 291218, 291262, 459827; 4 preopercula, USNM 459830; 1 operculum, USNM 412167; 2 lacrimals, USNM 291192, 336359; 1 premaxilla, USNM 291156; 1 dentary, NCSM 3106. REMARKS.—Numerous fossil skulls have been found at the Lee Creek Mine site that are very similar to the extant northern searobin, Prionotus carolinus, and the striped searobin, Prion- otus evolans. They are easily recognized by the elaborate sculpturing of the external surfaces of the skull, which is char- acteristic of Prionotus. The posttemporals extend caudally well beyond the caudal margin of the supraoccipital. The skull roof is broad and somewhat flat and is slightly concave between the orbits, which penetrate the margin of the skull roof. The ros- trum tapers from the orbits to the ethmoid, and its dorsal sur- face is slightly convex. At Lee Creek Mine, three different skull types of Prionotus occur. In the first type (USNM 290627, Figure 68a-c), hyper- ostotic enlargement of the posttemporals is absent; only three skulls of this type are complete, and they measure 45.4 mm (USNM 290627), 35.8 mm (USNM 459862), and 45.6 mm (USNM 459826) in total length. In the second type (USNM 291218, Figure 680"^), the post- temporals are hyperostosed and fuse at the midline of the skull. None of these skulls are complete; the rostra are missing, and the largest, USNM 291218, measures 37.6 mm in total length. One specimen of this type, USNM 459827, may be pathologic (Figure 68/g). In the third type (USNM 290663), the hyperostosis is not a singular, bulbous mass as in the second type but develops in an elongate mass of a central and two lateral lobes, which are asymmetrical (Figure 68/?,/). Although in all of these speci- mens (n=43) the rostral and orbital regions are missing, the outline of the braincase is identical to that in the other two types of Prionotus mentioned above. The largest specimen of this third type, USNM 290663, measures 64.9 mm in total length. An isolated fossil operculum, USNM 412167 (Figure 68/'), was found at Lee Creek Mine, and it exhibits a sculpture and structure similar to those of the extant species. A partial right dentary also was recovered (Figure 68A:). Fossil searobins from the eastern coastal plain of the United States have been reported from the Pleistocene of Virginia (Ray et al., 1968), New Jersey (Selden, 1986), Maryland (Blake, 1953), and Florida (Swift and Wing, 1968). All of these specimens are partial neurocraniums that were identified as Prionotus sp., but Blake (1953) and Selden (1986) suggested their specimens had a closer affinity to P. evolans. Prionotus carolinus and P. evolans are benthic species that occur from Nova Scotia to Florida (Robins and Ray, 1986) at depths of 1 to 80 m (Thomson et al., 1978). Searobins feed on shrimp, amphipods, squids, and worms (Manooch, 1984). Family Serranidae (sea basses) Epinephelus sp. Figure 69a,b Horizon.—Yorktown Formation (units 1,2). Referred Material.—13 incomplete premaxillae, USNM 256272,285321, 286935, 290572, 290594, 290612, 290634, 291073, 291149, 291162, 291177, 291199, 291248; 5 incom- plete dentaries, NCSM 4797, USNM 256273, 286907, 286925, 290539. Remarks.—The generic assignment is based on the great similarity of the fossil premaxillae to the premaxillae of Epi- nephelus morio. Because genera of groupers, not to mention species, are rather difficult to distinguish based on jaw ele- ments alone, no attempt at specific identification is made. The premaxilla (Figure 69a) has a labial row of large teeth that decrease in size posteriorly; the alveolus for the largest and anteriormost tooth is 4 mm in diameter. Lingual to this row of large teeth is a triangular area of numerous small alve- oli supporting a cardiform tooth series. Alveoli in this area in- crease in size toward the lingual edge. Maximum height of the premaxilla (excluding teeth) is 20.5 mm, maximum width is 10.9 mm, and the width of the tooth-bearing area is 8.8 mm at the symphysial end. At the symphysial end, the ascending process is transversely compressed and is moderately low. The arrangement of the tooth series and the shape of the as- cending branch of the premaxilla closely resemble that of Epi- nephelus morio. The dentary contains teeth of three different sizes. On the la- bial side of the symphysial end of the dentary is a tooth patch (Figure 696) with large, conical teeth (usually two). Lingual to these teeth on this same tooth patch there are five to six rows of small teeth (about 1 mm in diameter at their bases), and lingual to these is a row of intermediate-sized teeth (about 1.5-2 mm in diameter at their bases); this row extends to the distal end of the dentary. On the main tooth surface of the dentary and labial to this row of intermediate-sized teeth there is a row of smaller teeth followed labially by a row of intermediate-sized teeth, three rows in all along most of the dentary. Dentary USNM 256273 is 8.9 mm wide and 16.9 mm deep (excluding teeth); the largest dentary, USNM 290539, which is incomplete, mea- sures 13.7 mm in width, 56.0 mm in length, and 22.6 mm in depth (excluding teeth). Groupers are offshore reef fishes that occur from Massachu- setts to Brazil (Robins and Ray, 1986) at depths of 24 to 210 m (Manooch, 1984). They feed on a variety of bony fishes, in- cluding snappers, scuds, porgies, and searobins, and on a vari- ety of invertebrates, including shrimp, crabs, squids, and worms (Manooch, 1984). 170 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE 69.—Epinephelus sp.: a. USNM 256272, right premaxilla, ventrolingual view; b, USNM 256273, ante- rior fragment of left dentary, labial view. Mycteroperca sp., USNM 290658, left dentary: c, lingual view; d, labial view. Seriola sp., NCSM 1857, partial left dentary: e, labial view;/ occlusal view. (Scale bars= 1.0 cm.) Mycteroperca sp. Figure 69c,d Horizon.—From spoil piles, probably lower Yorktown For- mation (unit 1). REFERRED Material.—1 left dentary, USNM 290658. REMARKS.—A single dentary (Figure 69c,d) with the alveo- lar ridge complete and with several conical teeth resembles those of modern Mycteroperca. At the symphysial end there are two large alveoli for conical teeth, about 6 mm in diameter; these are followed lingually by a row of very small teeth that extends distally a short distance, about 20% of the length of the dentary. This row of small teeth is between two rows of inter- mediate-sized teeth, with alveoli of 2 to 4 mm in diameter; these intermediate-sized teeth extend distally to the end of the dentary. The total length of the dentary is 14.6 cm, and the depth at the symphysis is 2.9 cm. NUMBER 90 171 Mycteroperca are offshore reef fishes that occur from Mas- sachusetts to Brazil (Robins and Ray, 1986) at depths between 10 and 130 m (Manooch, 1984). They feed on shrimp, scuds, porgies, snappers, grunts, squids, and sardines (Manooch, 1984). Family Branchiostegidae (tilefishes) Two genera of branchiostegids, Caulolatilus and Lopholatit- us, occur at Lee Creek Mine. This is the first occurance in the fossil record for Caulolatilus. Lopholatilus was previously re- corded from the Yorktown Formation as otoliths (Fitch and Lavenberg, 1983) and from the Calvert Formation of Maryland (Kimmel and Purdy, 1984). Based on a specimen from Miocene sediments near Oran, Algeria, Arambourg (1927) described Latilus mesogeus, but Dooley (1978) found Arambourg's illustration and description too vague to ascertain the identity of this species. Except for this specimen, the specimens from Lee Creek Mine are the only definite fossil branchiostegids yet known. Caulolatilus cf. C. cyanops Poey, 1866 Figure 70a, b Horizon.—Yorktown Formation (unit 1). Referred Material.—Associated right and left ceratohy- al, USNM 412151. REMARKS.—A pair of associated ceratohyals (Figure 10a, b) compare well with those of the extant blackline tilefish, Cau- lolatilus cyanops. They are similar in having an anterior projec- tion for articulation with the posteroventral notch of the hypo- hyal. The broad, flattened section of the ceratohyal is much shorter than in Lopholatilus and has only a short groove that extends from the upper one-third to about one-half of the length of the ceratohyal; in Lopholatilus this groove extends the entire length of the ceratohyal. Robins and Ray (1986) reported that Caulolatilus cyanops ranges from North Carolina south to the Gulf of Mexico and to northern South America. This species has been recorded at depths between 45 and 495 m, with a preferred depth range of 150 to 250 m (Dooley, 1978). Lopholatilus rayus, new species Figure 70c-f Holotype.—Partial skull including associated partial left dentary (30 mm in length) missing distal end, partial left pre- maxilla lacking distal end, partial left operculum, epiotic, frag- ment of center portion of preoperculum, and fragments of max- illa, hyomandibular, and palatine, NCSM 2900. Type Locality.—Lee Creek Mine, Aurora, North Carolina. Horizon.—Yorktown Formation (unit 1). Paratypes.—1 partial left dentary, 22 mm long, NCSM 1572; 1 partial left dentary, 22 mm long, NCSM 3233; 1 partial left dentary, 26 mm long, NCSM 3160; 1 partial skull, USNM 437559, including symphysial portions of both dentaries; 1 par- tial dentary, 37.3 mm long, USNM 336240; 1 vomer, 19.6 mm wide, USNM 437550. Etymology.—In recognition of Clayton E. Ray, whose de- termination has brought about the extensive study of the Lee Creek Mine fauna. Diagnosis.—A villiform tooth row extends from the anteri- or tooth posteriorly along the lingual margin of the coronoid process of the dentary. In all other branchiostegids, this villi- form tooth row is absent. Remarks.—This is one of the most common species of the nonpelagic fishes found at the Lee Creek Mine site. It also is one of the few fish species for which numerous partial skele- tons have been found. Although the otoliths from Lee Creek Mine have been described as Lopholatilus chamaeleonticeps, the extant species (Fitch and Lavenberg, 1983), we believe the difference between the dentaries of the fossil and extant species is sufficient to warrant the establishment of a new species. Numerous skeletal elements have been identified as belong- ing to L. rayus. These include premaxillaries, dentaries, hyo- mandibulars, vomers, parasphenoids, lateral ethmoids, angu- lars, ceratohyals, epihyals, vertebra, exoccipitals, and more. Most of these elements, including the premaxillaries, angulars, hyomandibulars, ceratohyals, epihyals, vertebra, and exoccipi- tals, agree very well in structure with L. chamaeleonticeps. The dentaries (Figure 70c,d) of L. rayus are very similar to those of L. chamaeleonticeps in having the same basic struc- ture, including a broad anterior tooth patch and a well-defined ventral exterior groove. In L. chamaeleonticeps the villiform teeth are confined to the anterior tooth patch, with the remain- der of the tooth row being made up of large, canine-like teeth. The vomer (Figure 10e,f) of L. rayus, although similar to that of L. chamaeleonticeps, differs in being broader and in having a flatter ventral base. The lateral processes of the vomer are prominent, thinning out to a sharp edge, and the anterior surfaces of the lateral processes are slightly striated and are flat or very slightly concave rather than noticeably concave as in L. chamaeleonticeps. Dooley (1978) reported that Lopholatilus chamaeleonticeps has a current range from Nova Scotia south to the Gulf of Mex- ico and to northern South America. This species has been re- corded at depths of 81 to 540 m, but it occurs generally in a rel- atively narrow zone along the continental slope and along the upper reaches of canyons at depths of 120 to 200 m (Dooley, 1978). There are two critical habitat requirements for tilefish, a suitable temperature, 9° to 14°C, and shelter, which can be pro- vided by rocks, boulders, or clay in which the tilefish can exca- vate vertical burrows (Grimes et al., 1986). Lopholatilus feeds on various crustaceans, snails, worms, fish, sea urchins, anemones, and sea cucumbers (Manooch, 1984). 172 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Family Pomatomidae (bluefishes) Pomatomus saltatrix (Linnaeus, 1766) Figure 70g-o HORIZON.—Yorktown Formation (units 1, 2). Referred Material.—Numerous premaxillae, USNM 286874, 290602, 290610, 290611, 290615, 290682, 291122, 291130; numerous dentaries, USNM 290600, 291063, 291088, 291100,291131,291172; 1 ethmoid, USNM 437564; 2 fron- tals, USNM 437562,437563; 1 ceratohyal, NCSM 1442; 1 ver- tebra, USNM 437566; 2 angulars, USNM 437565, 476366; 1 hyomandibular, NCSM 8363; 3 quadrates, NCSM 2908, 6685, USNM 476365; 1 maxilla, NCSM 3648; 1 vomer, USNM 437564. Remarks.—Numerous skeletal elements of this fish have been found at Lee Creek Mine, including several associated partial skeletons. The premaxillae compare well with those of modern bluefish, Pomatomus saltatrix, in having a labial row of laterally compressed caniform teeth and a row of smaller but otherwise similar teeth along the lingual side of the jaw (Figure lOg.h). The two rows converge posteriorly until they are side by side just forward of the postmaxillary process. Anteriorly, the labial and lingual rows are separated by a toothless ex- panse. The arrangement of the tooth rows and the shape and angle of the ascending and articular processes match those of the modem species. Other skeletal elements, including the dentary (Figure 70/), ceratohyal (Figure 70/), quadrate (Figure 10k), angular (Figure 70/), vertebrae (Figure 10m), frontal (Figure 70«), and vomer (Figure 70o), also agree well with Pomatomus saltatrix. This is Figure 70.—Caulolatilus cf. C. cyanops, USNM 4I2I5I, left and right ceratohyal: a, medial view; b, lateral view. Lopholatilus rayus, new species: c, NCSM 2900 (holotype), left dentary, occlusal view; d, same speci- men, labial view; e, USNM 437550 (paratype), vomer, ventral view;/ same specimen, labial view. Pomato- mus saltatrix: g, USNM 290615, right premaxilla, ventrolingual view; h, same specimen, occlusal view; i, USNM 291100, right dentary, labial view;/ NCSM 1442, ceratohyal, lateral view; k, USNM 476365, left quadrate, lateral view; /, USNM 476366, right angular, labial view; m, USNM 437566, vertebra, lateral view; n. USNM 437562, right frontal, dorsal view; o, USNM 437564, vomer, dorsal view. (Scale bars: a-d,g-j=\.Q cm; e,f,k-m=0.5 cm; n=0.7 cm; o= 1.0 cm.) NUMBER 90 173 the first occurrence of the genus Pomatomus in the fossil record. The extant bluefish is a pelagic species distributed world- wide except in the eastern Pacific Ocean (Robins and Ray, 1986). It feeds on various fishes, shrimps, lobsters, crabs, and worms (Manooch, 1984). Family Carangidae (jacks) Seriola sp. Figure 69e/ HORIZON.—Yorktown Formation (units 1, 2). Referred Material.—1 dentary, NCSM 1857. REMARKS.—The dentary (Figure 69c,/) is similar to that of Seriola dumerili in having the same villiform teeth, numerous foramina on the outer surface, and the same anterior profile, a noticeable curvature, from dorsal or ventral view. Seriola dumerili is an offshore reef fish with a world-wide distribution (Robins and Ray, 1986). It feeds on crabs, squid, round herring, round scad, filefish, little tunny, and other fishes (Manooch, 1984). Family Sparidae (porgies) Archosargus cf. A. probatocephalus (Walbaum, 1792) Figure 7la Horizon.—Yorktown Formation (unit 1). REFERRED Material.—Numerous frontals, USNM 287938, 476235-^76243. Remarks.—Numerous frontals (Figure 71a), which com- pare well with those of the extant sheepshead porgy, Archosar- gus probatocephalus, have been found at Lee Creek Mine. They are very thick and robust, and at their anterior ends, they have a V-shaped groove for accepting the ethmoid. On the pos- terior half of the dorsal surface, along the midline, there is an elevated crest. Based on frontals alone, we hesitate to make a specific identification. Archosargus probatocephalus is a benthic species that oc- curs from Nova Scotia south to Brazil, primarily in bays and estuaries (Robins and Ray, 1986). It feeds on shellfish and sea urchins (Manooch, 1984). Lagodon cf. L. rhomboides (Linnaeus, 1766) Figure 7 \b,c Horizon.—Pungo River Formation (units 2-6); Yorktown Formation (units 1,2). Referred Material.—Numerous isolated teeth, USNM 291137; 1 partial vomer-ethmoid, NCSM 9589. Remarks.—The isolated teeth (Figure lib) from Lee Creek Mine compare well with those of the extant Lagodon rhom- boides. In cross section, the teeth are round at the base and in- cisiform at the tip; a small notch divides the tip, and the result- ing lobes of the tip are rounded or pointed. The vomer-ethmoid fragment (Figure 71c) from the York- town Formation also compares well with that of the extant L. rhomboides; there are no characters in this specimen or in the isolated teeth that would warrant separating them from the ex- tant species, but the remains are too fragmentary to determine that they are the same species. We must remark on the abundance of these teeth at Lee Creek Mine. Because these teeth are small and phosphatic, they are found in the flotation concentrate produced by the mill. Along the base of the outdoor storage piles, the coarser grains of concentrate accumulate on the crests of wind ripples, and Lagodon teeth can be picked in great numbers from this coarse fraction. Because as many as two million tons of concentrate may be on the ground at a given time, the numbers of these fos- sil teeth are astronomical. There are only two other reported fossil occurrences of La- godon, one from the St. Marys Formation of Maryland (Berry, 1932) and one from a Pliocene deposit in Florida (Caldwell, 1957). Lagodon rhomboides is a benthic species that occurs from Massachusetts to the Yucatan Peninsula (Robins and Ray, 1986) at depths from several centimeters to 67 m; it feeds on worms, crustaceans, and mollusks (Manooch, 1984). Pagrus hyneus, new species Figure 71a"-/ Holotype.—One partial neurocranium from the frontals to the occipital area, NCSM 2916. Type Locality.—Lee Creek Mine, Aurora, North Carolina. Horizon.—Yorktown Formation (unit 1). Paratypes.—One partial neurocranium including the fron- tals and the occipital area and a nonswollen supraoccipital crest, USNM 479858; both frontals with supraoccipital crest, NCSM 10221. Etymology.—In recognition of Becky and Frank Hyne, whose collections at the Lee Creek Mine have increased our knowledge of the fossil fish faunas of North Carolina. Diagnosis.—This species differs from all other sparids by the narrow frontals, which thicken anteriorly; the articulation surface of the frontal with the ethmoid being, in profile, a deep- ly grooved "W"; the greatly hyperostosed supraoccipital, which in dorsal view is as wide as the frontals; and the apex of the supraoccipital crest being, in lateral view, above the poste- rior margin of the orbit. REMARKS.—In the holotype (Figure lld-f), the supraoccip- ital crest articulates along the midline ridge of the frontals, with its point of origin at the anterior end of the frontal. From the midline an oblique frontal ridge continues caudally onto the pterotic, where the ridge thickens and becomes grooved along 174 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY ¦ "•i...'!v;v^ FIGURE 71.—Archosargus cf. A. probatocephalus: a. USNM 476240, left frontal, dorsal view. Lagodon cf. L. rhomboides: b, USNM 291137, incisiform tooth; c, NCSM 9589, vomer-ethmoid, dorsal view. Pagrus hyneus, new species: d, NCSM 2916 (holotype), neurocranium, lateral view; e. same specimen, dorsal view;/ same specimen, ventral view; g, NCSM 10221 (paratype), both frontals with supraoccipital crest, lateral view; h, same specimen, dorsal view; i, same specimen, ventral view. Pagrus sp.:j. USNM 298296, paired frontals, dor- sal view; k, same specimen, ventral view. Stenotomus cf. S. chrysops: I, USNM 412154, neurocranium, dorsal view; m. USNM 336236, right premaxilla, lingual view. (Scale bars: a,b,d-f,j-l=\.0 cm; c,g=0.5 cm; h,i.m=0.75 cm.) NUMBER 90 175 its crest. Another ridge extending from the frontal-supraoccipi- tal suture intersects this ridge on the pterotic. Lateral to the frontal and anterior to the pterotic, there is a triangular sphenot- ic. Caudal to the pterotic, on the dorsal surface of the epiotic, is the epiotic condyle, which articulates with the posttemporal. Pagrus hyneus has characteristics that are similar to both Pagrus pagrus, the extant red porgy, and Stenotomus, the ex- tant Atlantic scup. Like the fossil species, P. pagrus possesses the same frontal-supraoccipital articulation, the same frontal- pterotic ridge morphology, and the same shapes of the epiotic condyle and the sphenotic. Pagrus hyneus shows greater similarity to the genus Pagrus than to Stenotomus in the following characters: (1) the supraoc- cipitals articulate along two-thirds of the length of the frontals rather than along one-half of their length as in Stenotomus; (2) the structure of the ridge that runs from the frontal-supraoccipi- tal articulation to the epiotic; and (3) the shape of the epiotic condyle. For these reasons we place the fossil species in the ge- nus Pagrus. We also think that the differences in the shape and placement of the supraoccipital crest, the thickness of the fron- tals anteriorly, and the profile of the articular surface of the frontals with the ethmoid warrant the establishment of a new species. Pagrus sp. Figure 71/* HORIZON.—Yorktown Formation (unit 1). Referred Material.—About 100 frontals, USNM 287925, 287927,287929, 289366, 290552, 290652, 298296. REMARKS.—Numerous frontals (Figure l\j,k), which com- pare well with the extant red porgy, Pagrus pagrus, occur at Lee Creek Mine. Like Archosargus, these frontals are thick- ened by an elevated crest, which is worn down in most speci- mens, along the midline of the posterior half of their dorsal sur- faces. Unlike Archosargus and P. hyneus, the anterior margins of the frontals are not grooved. Pagrus pagrus is a benthic species that occurs from New York to Argentina in the deeper parts of the continental shelf (Robins and Ray, 1986) at depths of 27 to 110 m (Manooch, 1984). The extant species feeds on worms, snails, crabs, and sea urchins (Manooch, 1984). Stenotomus cf. S. chrysops (Linnaeus, 1766) Figure 7 l/,m Horizon.—Yorktown Formation (units 1, 2?, 3?). Referred Material.—3 premaxillae, USNM 336236-336238; posterior portion of 1 neurocranium, USNM 412154. Remarks.—The Lee Creek Mine specimens agree well with those of Stenotomus chrysops, the extant Atlantic scup. The fossil neurocranium (Figure l\l,m) differs from Stenotomus chrysops in having a much broader and more massive anterior section of the parasphenoid, but it is still not as great as that in Pagrus. The premaxillae are similar to S. chrysops in having larger teeth distally, small villiform teeth toward the symphysial end and on the labial edge at the symphysial end, one row of large teeth, and a small descending process at the distal end of the premaxilla. Stenotomus chrysops is a benthic species that occurs from Nova Scotia to South Carolina, inhabiting the near-shore re- gion of the continental shelf; it feeds on shrimp, worms, clams, starfish, snails, crabs, and sea urchins and occasionally on small fishes (Manooch, 1984). Family Sciaenidae (drums) Sciaenops sp. Figure 72a-d HORIZON.—Yorktown Formation (units 1, 2). Referred Material.—14 incomplete premaxillae, USNM 244482, 285336, 285348, 285366, 286857, 290621, 290623, 290653, 291067, 291117, 291144, 291166, 291203, 291668; 6 dentaries, USNM 244470, 284899, 285341, 286902, 291086, 291098; 2 vertebrae, USNM 244481. Remarks.—This is a large sciaenid that closely resembles Sciaenops ocellatus. The premaxillae (Figure 12a,b) differ from those of the latter species in that the alveoli within the cardiform tooth area, which is lingual to the row of enlarged teeth, are not uniform in size. The bases of the lingualmost tooth row enlarge toward the symphysis, which suggests the presence of teeth that were about one-half the size of those of the enlarged labial row. Among the other sciaenids, the pres- ence of a labial row of large teeth distinguishes these premaxil- lae from those of Micropogon and Leiostomus. Lingual to this row, the presence of a cardiform tooth area distinguishes this form from Cynoscion and Otolithus. The closely spaced labial teeth are more like those of Sciaenops than like the widely spaced teeth of Menticirrus. The dentaries (Figure 72c) differ from those of S. ocellatus mainly in that the tooth-bearing area is convex dorsally rather than flattened. Also illustrated (Figure 72^) is an associated pair of anteri- ormost vertebrae. These vertebrae match those of S. ocellatus and are about the right size to be associated with the above mouth parts, but all of the sciaenids have very similar vertebrae in this region, and they cannot be identified with certainty. Sciaenops ocellatus (Linnaeus, 1766) Figure 72e Horizon.—Yorktown Formation (?unit 3, possibly much higher). Referred Material.—Fragmental skeletal parts from one individual, including pieces of both premaxillae, an opercular, 176 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY articular, quadrate, hyomandibular, atlas, and thoracic verte- brae lacking neural arches, USNM 475010. Remarks.—The stratigraphic position of this specimen is uncertain, but the preservation of the bone is similar to that of the material from unit 3 of the Yorktown Formation. The bones of this partial skeleton are so similar to the extant species that there is no question that the Lee Creek Mine specimen is this species. Consequently, the specimen may have come from a considerably higher position in the Lee Creek Mine section. The extant Sciaenops ocellatus is a benthic species that oc- curs from Massachusetts to Tuxpan, Mexico, in coastal and es- tuarine waters (Simmons and Breuer, 1962). It feeds on shrimp, crabs, sand dollars, and bony fishes, such as menhaden, mullet, pinfish, pigfish, searobins, lizardfish, spot, croaker, and floun- der (Manooch, 1984). Pogonias cf. P. cromis Cope, 1869 Figure 72/g HORIZON.—Pungo River Formation (units 4, 5); Yorktown Formation (units 1, 2, 3?). Referred Material.—6 complete upper pharyngeal plates, NCSM 3469, USNM 287978, 459832-459835; 1 com- plete lower pharyngeal plate, USNM 459836; numerous in- complete pharyngeal plates, USNM 256270, 256271, 290625, 290665, 290676, 291107, 291108, 291112, 291195, 291259, 291683; 2 premaxillae, questionably referred; isolated pharyn- geal teeth (Pungo River Formation, units 4, 5); 2 worn pharyn- geal plates (Yorktown Formation, unit 3). Remarks.—Compared to the extant black drum, Pogonias cromis, and the Miocene species, P. multidentatus Cope, the Lee Creek Mine pharyngeals are more similar to those of the extant species. Based on a nearly complete, upper right pharyngeal dental battery from the Calvert Formation of Westmoreland County, Virginia, Cope (1869) described P. multidentatus as having TABLE 4.—Comparison of the upper pharyngeal in fossil and extant Pogonias. (VZ=vertebrate zoology collection.) Specimen Pharyngeal length (mm) Number of teeth Age Pogonias cromis USNM VZ 26132 24.5 18 recent USNM VZ 110700 51 33 recent USNM VZ 110216 63 37 recent Pogonias multidentatus Holotype 45 45 Miocene Pogonias cf. P. cromis NCSM 3469 69 36 Pliocene USNM 287978 41 24 Pliocene USNM 459832 78 43 Pliocene USNM 459833 67 41 Pliocene USNM 459834 72 41 Pliocene USNM 459835 68 35 Pliocene more crushing teeth in the same relative area than does the modern species, P cromis. The number of teeth in the Lee Creek Mine specimens, all upper pharyngeals, falls within the range of the extant species (Table 4). In contrast to P. cromis, the Lee Creek Mine pharyngeals (Figure 12f,g), are more elongate. Perhaps correlated with this elongation, a consistent difference appears in the articulation of the pharyngeal to the epibranchial. On the dorsal surface of the upper pharyngeal of P. cromis, a ridge originates near the ante- rior border, slightly toward the symphysial side of the plate's midwidth. This ridge extends posteriorly and gives rise to a knob to which the second epibranchial attaches. This ridge originates relatively further back on the Lee Creek Mine speci- mens and is longer and narrower than that of the modem form. Therefore, we assign these specimens only tentatively to the extant species. The extant Pogonias cromis is a benthic species that occurs from New England to Argentina in coastal and estuarine waters (Choa, 1978). It feeds on clams, mussels, oysters, crabs, worms, and some fishes (Manooch, 1984). Family Labridae (wrasses) Tautoga cf. T. onitis (Linnaeus, 1758) Figure 72h-j Horizon.—Yorktown Formation (units 1,2). Referred Material.—4 incomplete premaxillae, USNM 207605, 459843^159845; 7 incomplete and 2 complete lower pharyngeals, USNM 290501, 297606, 281341, 459838-459841; 2 upper pharyngeals, 281341, 459842. Remarks.—Several premaxillaries (Figure 12h,i) collected at the Lee Creek Mine, except for size, compare favorably with those of the extant tautog, Tautog onitis. The fossil premaxil- laries are nearly twice the size of those of the extant species available to us. The extant tautog premaxillaries were from in- dividuals between 350 and 407 mm TL, which are less than half the maximum size (915 mm TL) for this species. This size difference, therefore, does seem not taxonomically significant. In dorsal or ventral view the lower pharyngeals (Figure 72/) form an isosceles triangle, with the top pointing anteriorly. The dorsal surface of the pharyngeal is covered with cylindri- cal teeth of varying sizes, with the largest toward the midline, and with blunt, rounded crowns. In the one complete speci- men with an associated upper left pharyngeal, the lower pha- ryngeal has an anteroposterior length of 16.7 mm and a width of 38.8 mm. In a nearly complete specimen, in which the ante- rior edge is missing, the anteroposterior length is about 21 mm, and it is 52.2 mm wide. These specimens, except for their sizes, are identical with those available of the extant spe- cies, Tautoga onitis. Based on an incomplete premaxilla from the Miocene of Vir- ginia, Leidy (1873) proposed the new genus and species Pro- tautoga conidens. He distinguished it from Tautoga onitis by NUMBER 90 177 FIGURE 72.—Sciaenops sp.: a, USNM 244482, anterior portion of left premaxilla, lingual view; b, same speci- men, occlusal view showing outer row of large alveoli and inner cardiform area; c, USNM 244470, anterior end of dentary, labial view; d. USNM 244481, associated atlas and second vertebrae. Sciaenops ocellatus: e, USNM 475010, lateral view of atlas. Pogonias cf. P cromis, upper right pharyngeal plates, occlusal view:/ USNM 256270; g, USNM 256271. Tautoga cf. T. onitis: h, USNM 207605, incomplete right premaxilla, lacking deli- cate posterior and elongate ascending processes, labial view; i, same specimen, lingual view;/ USNM 459838, lower pharyngeal, occlusal view. (Scale bars: o-g=1.10 cm; h-j=l.O cm.) its large size, greater number of teeth, and the wider spacing of its teeth. As in the extant species, however, these characteris- tics reflect different stages of growth, and Leidy's species Pro- tautoga conidens cannot be separated from the extant species, Tautoga onitis. The extant tautog is a benthic species that occurs from Nova Scotia south to South Carolina (Robins and Ray, 1986) at depths of 1 to 30 m (Bigelow and Schroeder, 1953). It feeds on mussels, barnacles, crabs, shrimp, and worms (Manooch, 1984). Family Uranoscopidae (stargazers) Astroscopus sp. FIGURE 73a-c HORIZON.—Yorktown Formation (unit 1). Referred Material.—5 opercula, USNM 412156, 412158-412160,459837; 1 preoperculum, USNM 412157. 178 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Remarks.—The Lee Creek Mine opercula (Figure 13a,b) are very similar to those of Astroscopus guttatus in the shape of the articular fossa of the opercula, which articulates with the opercular process of the hyomandibular, in the ridge line ex- tending from this facet posteriorly to approximately two-thirds of the mesial surface of the opercula, and in ornamentation. The preoperculum is a somewhat crescent-shaped bone with three radially divergent processes along the caudal edge (Fig- ure 73c). Like the extant species, the exterior surface of the bone is ornamented with circular depressions. The Lee Creek Mine specimens are insufficient for attempting a specific iden- tification. I \ FIGURE 7 s—Astroscopus sp.: a, USNM 412156, Yorktown Formation, operculum, medial view; b, same speci- men, lateral view; c, USNM 412157, preoperculum, lateral view. Sphyraena cf. 5. barracuda: d, USNM 295338, Yorktown Formation, anterior portion of left dentary, labial view; e, USNM 476367, Yorktown Formation, par- tial right dentary, lingual view;/ USNM 291168, Pungo River Formation, isolated tooth; g, USNM 291081, Yorktown Formation, vertebra, lateral view; h, USNM 476368, Yorktown Formation, premaxilla, occlusal view. (Scale bars=1.0cm.) NUMBER 90 179 Astroscopus has been previously reported from the Atlantic Coastal Plain. Ray et al. (1968) identified to this genus an otico- temporal portion of a neurocranium from the Pleistocene Kempsville Formation of Virginia, and Fitch and Lavenberg (1983) also identified otoliths from the Pliocene at Lee Creek Mine to this genus. Of the two extant species along the Atlantic coast of the United States, Astroscopus guttatus is a benthic species that oc- curs from New York to North Carolina, and A. y-graecum oc- curs from North Carolina to Mexico (Robins and Ray, 1986). Family Sphyraenidae (barracudas) Sphyraena cf. S. barracuda (Walbaum, 1792) Figure 7sd-h HORIZON.—Pungo River Formation (units 2-6); Yorktown Formation (units 1, 2). REFERRED MATERIAL.— 9 vertebrae, USNM 291076, 291081; 1 dentary lacking teeth, USNM 291664; 4 partial den- taries with teeth, USNM 291236, 295338, 437561, 476367; 2 partial premaxillae, USNM 476368; about 40 uncataloged frag- ments of jaws; about 50 isolated teeth, USNM 291168. REMARKS.—In both the Pungo River and Yorktown forma- tions, lanceolate and compressed barracuda teeth (Figure 73/) are among the more common osteichthyan remains. The teeth offer no certain characteristics that distinguish them from the modern form, Sphyraena barracuda. Fossil sphyraenid spe- cies founded on isolated teeth, such as S. major Leidy or 5. speciosa Leidy, probably cannot be distinguished from S. bar- racuda. The Lee Creek Mine dentaries (Figure 13d,e) are within the size range of S. barracuda. On their symphysial ends, they bear a large, lanceolate tooth that is followed by a single row of la- biolingually compressed to flat, carinate teeth. Some of these dentaries came from individuals that attained a very great size; the largest, USNM 295338, which is lacking its posterior half (Figure 73a"), has a depth at the symphysis of 37.5 mm. In the symphysial area of the premaxilla (Figure 73/z), two large lanceolate teeth are followed by teeth similar to those in the dentary, but they are much smaller than those in the oc- cluding dentary. Both the dentaries and the premaxillaries are indistinguishable from those of the extant Sphyraena barra- cuda. Vertebrae, which we believe are sphyraenid (Figure 73g), also were collected at Lee Creek Mine. Although the neural arch is broken away in the available specimens, the long, smooth, medially constricted centra with deep sulci dorsal to the midheight duplicate recent specimens. Sphyraena barracuda is an offshore reef species that occurs from Massachusetts to Brazil and also occurs worldwide in tropical and warm-temperate waters (Robins and Ray, 1986). It feeds on many species of bony fishes, including jacks, silver- sides, parrotfish, and filefish (Manooch, 1984). Family SCOMBRIDAE (mackerels and tunas) Sarda aff. S. sarda (Bloch, 1801) Figure 74 Horizon.—Pungo River Formation (units 4, 5); Yorktown Formation (units 1-3). Referred Material.—Numerous mesethmoids, USNM 290654, 291082, 291090, 291146, 291170, 291201; numerous portions of skulls, USNM 290544, 290595, 475012, 475013, 476225-476231; anterior end of 1 dentary, USNM 476250; nu- merous hypural elements, USNM 290681, 291092, 291095, 291104, 291140, 291152, 291158, 291160, 291167, 291169, 291197, 291233, 291252, 291677, 476396; numerous uncata- loged vertebrae. Remarks.—The incomplete dentary (Figure 74/) from the Yorktown Formation belongs to the tribe Sardini, which in- cludes the genera Cybiosarda, Orcynopsis, Gymnosarda, Al- lothunnus, and Sarda. Sarda sarda is the only member of the group now occurring in the western Atlantic Ocean, and it is with this species that the fossil compares favorably. A characteristic feature of Sardini dentaries, other than the large teeth, is the acute angle formed between the plane of the tooth row and the anterior margin of the dentary (see Collette and Chao, 1975). This angle is about 65° on the fossil, but it is usually less than 57° on dentaries of recent Sarda. In this as- pect the fossil dentary resembles Cybiosarda. In Sarda, Collette and Chao (1975) also noted the presence of a notch on the upper portion of the anterior margin of the dentary; in the Lee Creek Mine specimens, this notch is weakly developed or absent. The extant bonitos Cybiosarda, Orcynop- sis, and Gymnosarda lack this notch; however, unlike Cybio- sarda and Gymnosarda, the Lee Creek Mine dentaries possess a notch in the anterior, ventral margin that is similar to the notch in the dentaries of Orycnopsis and Sarda. Despite the noted differences between the fossil dentary and the dentaries of Sarda, the fossil dentary compares more favor- ably with those of this genus than with any of the others in the Sardini. It may prove to be a distinct species, but available fos- sil material is inadequate to establish its identity. Mesethmoids (Figure 14a,b) and incomplete skulls (Figure 14c,d) are common fossils in the lower Yorktown Formation. On the center of the ventral surface of the mesethmoid, the an- terior end of the vomer (or prevomer) is often present. Like most other bonitos except Orcynopsis, the anterior edge of the ethmoid is "concave... with an anteromedian projection and an anterolateral horn on each side" (Collette and Chao, 1975:539). These characters are found in the Lee Creek Mine meseth- moids, which are identical to those of the extant species. 180 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE 74.—Sarda aff. 5. sarda: a, USNM 290654, mesethmoid, dorsal view; b, same specimen, ventral view, with anterior end of prevomer; c. USNM 476225, incomplete skull, dorsal view; d, same specimen, ventral view; e, USNM 476369, incomplete hypural, lateral view;/ USNM 476250, anterior portion of dentary, labial view. (Scale bar= 1.0 cm.) The incomplete skulls are principally left or right portions of the skull roof from the anterior tips of the frontals to the su- praoccipital and epiotic. As in Sarda, the sphenotic is larger than it is in Thunnus atlanticus or T. obesus. The triangular hypural plate is composed of five fused hy- pural bones (Potthoff, 1975). According to Collette and Chao (1975), the dorsalmost hypural bone, hypural 5, does not completely fuse with the hypural plate in the bonitos or in higher tunas, Auxis to Thunnus; they did not present charac- ters for the hypural plates that could be used to separate Sar- da from other bonitos or from tunas. A vestige of the primi- tive hypural notch is present only in the bonito genus Gymno- sarda. While examining the skeletons of the extant Sarda sarda and comparing them to the Lee Creek Mine specimens, we no- ticed that the majority of ethmoid fossils were equal to or smaller than those from the extant species (550 mm TL size range). A hypural from a 550 mm TL Sarda sarda had a cranio-caudal length of 13 mm and a dorsoventral height of 16 mm. The Lee Creek Mine hypurals (Figure 74c), which we have assigned to Sarda sarda, are very small, 13-14 mm in cranio-caudal length and 15 mm in dorsoventral width, and lack a hypural notch; therefore, based on size, these hypurals are assigned to Sarda. Sarda sarda is a coastal pelagic species that occurs from the Gulf of St. Lawrence to Argentina (Robins and Ray, 1986); it feeds on squid, mackerels, anchovies, menhaden, silversides, and shrimp (Manooch, 1984). Auxis sp. Figure 75 Horizon.—Yorktown Formation (unit 1). Referred Material.— 1 nearly complete dentary, USNM 291139. REMARKS.—This dentary (Figure 75) is a flat, thin bone with a single row of minute teeth. The dorsal surface bearing the tooth row rises toward the distal end; at the mesial end it is bent slightly toward the symphysis. The teeth are broken off, but the preserved tooth bases are circular in cross section, uniform in size (approximately 0.5 mm in diameter), and evenly spaced (13 teeth per cm). On the lingual face the intermandibularis fossa is short and broad. It is bounded above by a shelf-like an- teroventrally directed ridge. The dentary is 11.6 mm high and 5.2 mm wide at the sym- physis. Height at the mesial termination of the notch dividing the dorsal and ventral rami is 14.8 mm. Width at the same place is 5 mm. The combination of a distally rising tooth row, small teeth, and the peculiar intermandibularis fossa allies this dentary with Auxis. The fossil dentary differs only in minute detail from a dentary of Auxis thazard (Collette collection, USNM; uncata- loged; 315 mm fork length). The fossil dentary is four times larger than that of the extant specimen; it has a shorter inter- mandibularis fossa, and at the distal end, the tooth row does not rise to as great an extent as it does in the extant species. The genus Auxis is currently represented by two species, Auxis rochei, the bullet tuna, and A. thazard, the frigate tuna. NUMBER 90 181 a FIGURE 75.—Auxis sp., USNM 291139, nearly complete dentary: a, labial view; b, lingual view showing abbre- viated intermandibular fossa. (Scale bar= 1.0 cm.) These are pelagic species that occur in the warm waters of the Indian, Pacific, and Atlantic oceans (Joseph et al., 1980). Thunnus sp. Figure 76 HORIZON.—Pungo River Formation (units 3, 4); Yorktown Formation (unit 1). Referred Material.—Numerous isolated vertebrae, USNM 291070, 475232-475234, 494371; 36 hypural bones, including 1 nearly complete caudal complex, USNM 291069, 291092, 291140, 291152, 291160, 291167, 291169, 291197, 291252, 291677; about 100 incomplete dentaries, USNM 290162, 290163, 290575, 290622, 290672, 291093, 291094, 291119, 291242; about 50 incomplete premaxillae, USNM 290551,475011; 1 articular, USNM 291147; 2 maxilla, USNM 291132, 291202; 5 quadrates, USNM 290648, 291111, 291129,291237,291247; 1 angular, USNM 319669. Remarks.—In the Yorktown Formation at Lee Creek Mine, the most abundant teleost remains are those belonging to Thun- nus. This material agrees well with most of the available skele- tons of the extant tunas. As Gibbs and Collette (1967) noted, the amount of variation that occurs in different skeletal ele- ments of the genus Thunnus often overlaps between different species; although some skeletal characters for identifying spe- cies do exist, they are not preserved in the material available to us. Species determination will have to await the availability of articulated skeletons. In the Lee Creek Mine collection, the numbers of tuna-like scombrid vertebrae (Figure 76a) exceed those of all other te- leostean taxa combined. Many of these are exceptionally large and duplicate in size and form those of the extant Thunnus thynnus. The precaudal vertebrae are slightly wider than tall and bear deep upper and lower fossa. At the anterior border of the ridge that separates these fossae, there is a small triangular depression for ligament attachment. In that the parhypural is not fused to the complex, the fused hypural plate does not bear a medial notch, and there is a deep fossa developed on the hypural just behind the terminal cen- trum. The hypural complex (Figure 16b) is typical of Thunnus. The large size of the complex suggests that it belongs to Thun- nus thynnus rather to some other species in the genus. The dentaries (Figure 76c) bear a single row of very small (1.5 mm in basal diameter), evenly spaced (4-6 per cm), coni- cal teeth. Due to a pronounced rounded crest on the lingual side situated above the intermandibularis fossa, the bone is unusual- ly thick for scombrid dentaries. The head for the attachment of the geniohyoideus muscle is at the head of the fossa below this ridge and is unusually prominent. Two incomplete dentaries from the Pungo River Formation, USNM 290162 and 290163, differ from those of T. thynnus in that the intermandibularis fossa does not reach up to the sym- physis and the mesial edge of the bone is only faintly notched rather than prominently notched (Figure 76o"). The available fragmental material is not sufficient to establish the specific identity of the Pungo River Thunnus, but it may be closely re- lated to T. thynnus. A small premaxilla (34 mm long; approximately 38 mm re- stored), USNM 290551 (Figure 76e), from unit 3 of the York- town Formation, has large, well-spaced teeth with labiolingual- ly compressed bases and conical, lingually curved tips. Tooth shape and spacing and the overall shape of the bone are very similar to a premaxilla of T obesus. The only apparent differ- ence is the much thinner dorsal process of the fossil premaxilla. Premaxillae, however, appear to be conservative elements in Thunnus skeletons so identifications based on them are neces- sarily tenuous. Tunas are pelagic species that have a world-wide distribution in tropical and temperate waters (Robins and Ray, 1986). They feed on larval crustaceans, squid, paper nautilus, filefish, trig- gerfish, jacks, mackerel, and many other bony-fish species (Manooch, 1984). Acanthocybium solandri (Cuvier in Cuvier and Valenciennes, 1831) Figure 77 Sphyraenodus bottii Capellini, 1878:250, pl. 3: figs. 1-6 [middle Miocene, Italy]. Scomberomorus both (Cappellini).—Caria, 1973:19, pl. 6: figs. 1, 2, pl. 7: figs. 1, 2, pl. 8: figs. 1-3 [Miocene, Sardinia]. HORIZON.—Yorktown Formation (unit 1). Referred Material.—1 dentary, USNM 319668; 1 in- complete dentary, USNM 291077; 3 incomplete premaxillae, USNM 25782; 4 hypurals, USNM 291084; 1 precaudal verte- bra, USNM 476370. Remarks.—In the fossil dentaries and premaxillae, each ridge, sulcus, and foramen as well as the shape and relative size of these jaw elements, and the replacement pattern of the teeth 182 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE 76.—Thunnus sp.: a, USNM 494371, Yorktown Formation, incomplete caudal vertebra, lateral view; b, USNM 291069, Yorktown Formation, associated hypural complex, lateral view; c, USNM 290575, Yorktown Formation, partial left dentary, labial view; d. USNM 290163, Pungo River Formation, partial right dentary, lin- gual view; e, USNM 290551, Yorktown Formation, left premaxilla with teeth, labial view. (Scale bars= 1.0 cm.) are duplicated in three western Atlantic specimens of Acantho- cybium solandri. The fossils are from larger individuals than the largest recent individual available to us (1780 mm TL; 36.8 kg), and they exceed the known size range of the species (max- imum observed size -83 kg). In the fossil form, the premaxillae (Figure 77a) have a great- er transverse diameter compared with jaw depth, particularly at the dorsal edge below the tooth row. The ratio of the transverse diameter (width) to depth at the tenth alveoli from the symphy- sis of the fossil premaxilla averages 0.54, whereas this ratio in the modern specimens averages 0.44. This difference is minor compared to the much more marked differences in proportions and tooth shape encountered when comparing corresponding bones from two allopatric species of Thunnus, for example. Furthermore, this is the sort of change one would expect from a normal allometric growth pattern. The most nearly complete and largest dentary (Figure lib) of this species from Lee Creek Mine, USNM 319668, exhibits the characters identified by Collette and Russo (1984:575, 578-579) as diagnostic of dentaries of Acanthocybium: the teeth are more tightly packed, the notch on the anteroventral margin is absent, and a prominent notch is present on the ante- rior margin of the dentary. The fossil hypural (Figure 77c) is typical of Acanthocybium solandri in that the parhypural, which carries the laterally ex- tending parhypurapophysis, is fused to the complex, and the posterior border of the plate carries a well-defined medial notch. The condition in Gymnosarda is similar (Collette and Chao, 1975) but differs in that hypural five and the parhypural are incompletely fused to the complex. Scomberomorus, the most closely related extant genus, also has a similar caudal construction, but the parhypural is not always completely fused. The only striking quality of the fossil caudal complex when compared to A. solandri is its large size. Dorsoventral di- ameters of the fossils range from 49 to 88 mm, but modern specimens that have been examined do not exceed 50 mm. Precaudal Acanthocybium vertebrae (Figure lid) are easily recognized by the development of three lateral sulci rather than the usual two. Capellini (1878) established Sphyraenodus bottii on the basis of an associated dentary and premaxilla, which also compares favorably with Acanthocybium solandri. NUMBER 90 183 FIGURE 77.—Acanthocybium solandri: a, USNM 25782, posterior end of premaxilla, labial view; b, USNM 319668, anterior end of dentary, labial view; c, USNM 291084, hypural, lateral view; d, USNM 476370, verte- bra, lateral view. (Scale bars: a=2.0 cm; b= 1.0 cm; c= 1.25 cm; rf=0.9 cm.) Acanthocybium solandri is a pelagic species that occurs from New Jersey to South America and is distributed world- wide in tropical and warm-temperate waters (Robins and Ray, 1986). It feeds on a variety of bony fishes, including frigate mackerel, butterfish, porcupinefish, and round herring (Ma- nooch, 1984). Family Xiphiidae (swordfishes) Xiphias gladius Linnaeus, 1758 FIGURE 78 Horizon.—Yorktown Formation (unit 1). Referred Material.—1 rostrum fragment, USNM 476396. Remarks.—In the course of studying the Lee Creek Mine marlins, Harry Fierstine (pers. comm., Nov 1993) discovered a FIGURE 78.—Xiphias gladius, USNM 476396, lateral view of rostral fragment showing rectangular chambers. (Scale bar=1.0 cm.) fragment of a swordfish rostrum (USNM 476396). He stated that the central chambers of the rostrum (Figure 78), which are rectangular in this specimen, are distinctive characters of Xiph- ias rostra (Poplin, 1975; Poplin et al, 1976). According to Manooch (1984:306), the extant swordfish is a warm-water species; in the summer, it ranges from Newfound- land to Argentina. It feeds on squid and bony fishes, including hake, mackerel, and barracuda. 184 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Family ISTIOPHORIDAE (marlins) For the genera Makaira, Istiophorus, and Tetrapturus, see Fierstine (this volume). Hemirhabdorhynchus sp. Figure 79 HORIZON.—Pungo River Formation (units 2-5). Referred Material.—7 broken rostra, USNM 24745, 421907^121909. REMARKS.—Leidy (1856a: 12) gave the name Cylindracan- thus ornatus to a fluted fish spine that was said to be from the Upper Cretaceous beds near Pemberton, Burlington County, New Jersey, but the type specimen was not illustrated. Later that year, in the same journal, Leidy (1856b:302) noted that Agassiz had previously described two species, Coelorhynchus rectus and C. sinuatus, based on the same type of spine, which Agassiz (1833-1843:92) said were from the London Clay (Ypresian) at Sheppey, England. In 1905 Leriche noted that Coelorhynchus Agassiz is a junior homonym of Coelorhynchus Giorna, and unaware of Leidy's genus Cylindracanthus, he erected the genus Glyptorhynchus. Four years later Leriche (1909:381-383) discovered the priority of Leidy's genus; at the same time he restricted the name Glyptorhynchus to fluted ros- tra with flat oral surfaces bearing two parallel bands of crowd- ed alveoli for very small acicular teeth. He did not, however, know of these rostra until after he erected Glyptorhynchus; this last genus, therefore, is a junior synonym of Cylindracanthus, and it is not available for Leriche's rostra with flat oral surfaces bearing two bands of tooth alveoli. Based on Leriche's holotype of G. costatus but excluding other species of Glyptorhynchus, Casier (1946:155) erected the genus Hemirhabdorhynchus, which he characterized as having "a more or less depressed cylindro-conical form, with longitu- dinal costae, limited to the dorsal half or nearly on this half, and by the existence of two ventral wide alveolar bands." These characters also may be applied to the species Casier ex- cluded. If Leriche's and Casier's separation of these rostra from Cylindracanthus is correct, then Hemirhabdorhynchus is the senior name for these rostra. The familial identity of these specimens is uncertain. On the basis of the two bands of small acicular teeth, Leriche (1910, 1936a, 1942) placed this genus in the Xiphiidae. As confirming evidence of this taxonomic assignment, Leriche (1910) men- tioned a rostrum associated with vertebrae and a hypural from the Oligocene of Belgium. Concerning the hypural of Leriche's specimen, as in the Istiophoridae but not the Xiphiidae, the par- hypural is fused to the ventral hypural plate; unlike the avail- able hypurals of the extant Xiphiidae and Istiophoridae, the hy- pural notch is extremely shallow. In other scombroids, however, both of these characters may vary within a family. Because the shape of the rostra and tooth bands are characteris- tic of the Istiophoridae rather than the Xiphiidae, we assign Hemirhabdorhynchus to the Istiophoridae. The rostra (Figure 79) are known only as detached elements, and they are very seldom found complete. They are long, straight, slender rods of dense bone that taper to a point, which may be rounded by wear. Except for the flattened tooth-bearing surface, they are covered completely by fine longitudinal grooves, which may be obscured by wear. The flattened surfac- es bear two parallel bands of tooth alveoli, which are similar to those found on istiophorid rostra (Figure 19b); between these parallel rows the surface of the spine is smooth to very finely grooved. Order Pleuronectiformes Family BOTHIDAE (lefteye flounders) Paralichthys sp. Figure 80o-c HORIZON.—Yorktown Formation (units 1, 2). Referred Material.—2 anterior portions of dentaries with teeth, USNM 412162, 412163; 1 left maxilla, USNM 412164; 1 partial right articular, USNM 412165. Remarks.—The maxilla (Figure 80a), which consists of the articular end, is identical to that of the extant Paralichthys den- tatus. The dentary (Figure 806) has long, stout, rugose teeth, which, as in P. dentatus, are almost one-half the height of the dentary. The lower one-third of the dentary is striated distally and has a sculptured texture. The partial right angular (Figure 80c) is lacking most of the anterior process, the coronoid pro- cess, and part of the postarticular process. Because the fossil FIGURE 79.—Hemirhabdorhynchus sp.: a, USNM 421907, rostral fragment, dorsal view; b, same specimen, ven- tral view; c, same specimen, cross-sectional view. (Scale bar= 1.0 cm.) NUMBER 90 185 material consists of isolated bone fragments, species determi- nation is not possible at this time. Paralichthys dentatus is a benthic species that occurs from Maine to northern Florida (Robins and Ray, 1986). It feeds pri- marily on menhaden, silversides, sand lances, herrings, ancho- vies, weakfish, squids, shrimp, and crabs (Manooch, 1984). Order Tetraodontiformes (Plectognathi) Family Monacanthidae (filefishes) AI uterus sp. Figure SOd-f HORIZON.—Pungo River Formation (units 4-6); Yorktown Formation (units 1-3). Referred Material.—About 200 hyperostosed vertebrae, USNM 286157, 286158, 286164, 290288, 290302, 290528, 291690, 291693, 291707-291729, 291731-291779,291801, 291803-291805, 291807-291823, 291825-291827, 291831, 291834, 291840-291845, 291847-291849, 291851-291856, 291858-291862,476354. Remarks.—The most common hyperostosed bones found in the lower Yorktown Formation are vertebrae, which are identi- cal in their morphology and hyperostosis to those of the extant Alutents shoepfi. Like the extant species, they are dorsoventral- ly compressed, with prominent dorsal and ventral grooves on the centra (Figure 80o». In articular view, the second precau- dal vertebra is hexagonal in outline. Two or three precaudal vertebrae are often found as a fused unit (Figure 80/"). The pre- caudal vertebrae from Lee Creek Mine have a more pro- nounced lateral ridge than do those of the extant species, and FIGURE SO.—Paralichthys sp.: a, USNM 412164, proximal portion of left maxilla, labial view; b, USNM 412162, symphseal portions of left dentary, labial view; c, USNM 412165, partial right articular, labial view. Aluterus sp.: d, USNM 290302 (left) and 290288 (right), vertebrae, dorsal view; e, same specimens, axial view; / USNM 476354, three fused caudal vertebrae, tlorsal view. (Scale bars: a=0.5 cm; 6=1.25 cm; c=1.5 cm; c/,e= 1.0 cm;/=1.25 cm.) 186 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY the centra in the more cranial vertebrae are more elongate than those of A. shoepfi. The average craniocaudal length is 2.1 cm (range =1.74-2.28 cm, m=11); the average dorsoventral height is 1.4 cm (range=1.15-1.63 cm, n=\ 1), and the average lateral width is 1.8 cm (range= 143-2.24 cm, «=11). The extant Aluterus shoepfi ranges from Nova Scotia to Bra- zil and also is found worldwide in temperate and tropical wa- ters (Robins and Ray, 1986). It prefers coral reefs and rocky bottoms. This species feeds on a variety of invertebrates, in- cluding crabs and shrimp (Thomson et al., 1978). Family Tetraodontidae (puffers) Sphoeroides hyperostosus Tyler, Purdy, and Oliver, 1992 Figures 81,82e HORIZON.—Yorktown Formation (units 1-3). Referred Material.—Skull and first 4 vertebrae, USNM 437601 (holotype); incomplete cranium, USNM 290643; skull roof and first 3 vertebrae, NCSM 11179; 8 dentaries, USNM 364354, 364356, 437601; 7 premaxillae, USNM 364349, 364350, 364351; about 3000 opercula, USNM 364328- FlGURE 81.—Sphoeroides hyperostosus: a, USNM 437601, holotype, dorsal view of skull; b, same specimen, lateral view of left side; c, USNM 364351, left premaxilla, labial view; d, same specimen, lingual view; e, USNM 364356, left dentary, labial view;/ same specimen, lingual view; g, USNM 457096, preoperculum, lat- eral view; h, USNM 364322, sightly hyperostosed suboperculum, lateral view; i, USNM 440874, extremely hyperostosed suboperculum, lateral view;y, USNM 364329, operculum, lateral view; k, USNM 283794, ventral postcleithrum, lateral view; /, USNM 437601, second to fourth abdominal vertebrae removed from holotype, lat- eral view. (Scale bars= 1.0 cm.) NUMBER 90 187 364339; about 500 preopercula, USNM 364340-364346, 457096; about 500 subopercula, USNM 364321-364327, 440814, 440859, 440864, 440867, 440874, 440885, 440893; about 500 ventral postsupracleithra, USNM 283794, 283839, 283891, 284070, 364361-364363; about 100 uncataloged ver- tebrae. REMARKS.—The remains of this fish are among the most abundant present in the lower units of the Yorktown Forma- tion. Tyler et al. (1992) identified these remains as a new spe- cies of pufferfish, Sphoeroides hyperostosus. The reader is re- ferred to this paper for a description of the Lee Creek Mine type material. Prior to Tyler et al., Weiler (1973:469-477) identified the suboperculum, preoperculum, and ventral postsu- pracleithra as hyperostosed parts of the fin skeleton of an inde- terminate fish. Subsequent to the publication of Tyler et al. (1992), one of us (V.P.S.) collected a third skull roof of this species, including the right operculum, cleithrum, dorsal postcleithrum, hyperos- tosed ventral postcleithrum, and three associated vertebrae from the Yorktown spoil piles (Figure 82e). From ethmoid to pterotics, the skull roof is nearly complete. On the left side, the frontal portion of the flange formed by the frontal and the sphe- notic is missing; on the right side it is incomplete. Like the ho- lotype, the lateral ethmoid is broad, with a flat, unornamented upper surface, and the posterior dorsal surfaces of the frontals have thick, curved crests that extend posteriorly as short pro- cesses over the epiotics. The cleithrum and matrix adhering to the ventral portion of the skull roof obscure the morphological details of this area. The closest living relative of Sphoeroides hyperostosus, S. maculatus, does not posess any hyperostosed bones (see Tyler et al., 1992). Sphoeroides maculatus is a benthic species that occurs from Newfoundland to northern Florida (Robins and Ray, 1986). It is found in sandy-bottom habitats in waters ranging from 1 to 54 m deep (Manooch, 1984). It feeds primarily on clams, mus- sels, shrimp, worms, sea urchins, sponges, sea anemones, sea squirts, and crabs (Manooch, 1984). Family Diodontidae (porcupineflshes) Chilomycterus schoepfi (Walbaum, 1792) FIGURE %2a-d HORIZON.—Pungo River Formation (units 4, 5); Yorktown Formation (units 1-3); James City Formation. Referred Material.—Partial skull, NCSM 8364; about 2000 mouthplates, USNM 291200, 291219 (Yorktown Forma- FlGURE 82.—Chilomycterus schoepfi: a, NCSM 8364, skull, dorsal view; b, USNM 291200, occlusal view of upper jaws showing trituration tooth plates; c, USNM 291219, occlusal view of lower jaws; d. USNM 476374, dermal spine, dorsal view. Sphoeroides hyperostosus: e, NCSM 11179, dorsal view of skull roof. (Scale bars: a=2.0 cm; b,c,e=\.0 cm; d=0.5 cm.) 188 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY tion), 476400, 476401 (James City Formation); numerous der- mal spines, USNM 476374. Remarks.—Although many jaws (Figure 82b,c) and a partial skull were available for study, these bones are identical in Di- odon and Chilomycterus; the skeletons can be separated only on the basis of dermal spines (Tyler, 1980). The dermal spines of Diodon are two-rooted and erectile, whereas those of Chilo- mycterus are three-rooted, short, and triangular. The spines found at Lee Creek Mine (Figure %2d) clearly resemble those of Chilomycterus. Because specimens of Diodon have not yet been found at the mine, we believe the other skeletal elements found at Lee Creek Mine also should be referred to Chilomycterus. The partial skull, NCSM 8364 (Figure 82a), includes well- preserved fused dentaries, ethmoids, and frontals. The fused premaxillae are badly crushed as is the posterior portion of the skull. The skull is 65 mm wide and 106 mm long. The most common remains of this fish at Lee Creek Mine are incomplete, fused dentaries and fused premaxillae that consist of the beak and tooth plate area; these are alike in both jaws and thus are difficult to distinguish from each other. In com- plete dentaries, on each side of the tooth plate area, there is a rounded, ventrally directed process that articulates with the an- gular. Remnants of these processes are often preserved in the fossils. A portion of the ventral surface of the dentary is often worn away, exposing a small area of the tooth-forming surface. On the dorsal surface of the premaxilla, the fragile processes that articulate with the vomer-ethmoid complex are missing, exposing a large area of the tooth-forming surface. In occlusal view, the beaks of the dentaries tend to be rounded; those of the upper jaw are usually pointed. Chilomycterus schoepfi is a benthic species that occurs pri- marily from North Carolina to Brazil. It is common in seagrass beds in bays and coastal lagoons (Robins and Ray, 1986). Por- cupinefishes feed primarily on shrimp and crabs (Manooch, 1984). Family Molidae (ocean sunfishes) Mola chelonopsis (Van Beneden, 1883) Figure %sa,b Horizon.—Yorktown Formation (unit 1). Referred Material.—5 premaxillae, USNM 265650, 291211, 457143,476341, 476395. Remarks.—Weems (1985:431-432) identified the Lee Creek Mine specimens to this species, and he characterized it as follows: "Premaxillary beak toothless, and lacking palatal tooth brace, toothless shelf anterior to location of the former tooth position much longer than in M. mola, such that the ante- ro-posterior beak length is greater than the lateral beak width at the level of the back shelf. Dentary beak comparable to M. mo- la." Due to the lack of comparative material of the extant spe- cies and the unavailability of Van Beneden's type specimen, we cannot confirm or refute Weems's identification. The Lee Creek premaxillae have attached to them portions of the edentulous palate (Figure 83a, b). In all five specimens the articular ends of the premaxillae are missing. Molas are pelagic fish that have a worldwide distribution in warm waters. In the Atlantic they range from Newfoundland to South America. They feed on jellyfishes, Portuguese man-of- war, ctenophores, and other soft-bodied pelagic invertebrates and on larval fishes (Robins and Ray, 1986). Teleost incertae sedis Emmons's "fish tooth" Figure 83c-e Horizon.—Yorktown Formation (unit 3). Referred Material.—About 12 specimens, mostly bro- ken, USNM 421518. Remarks.—Emmons (1858:244, figs. 99, 100) reported on a small fossil from the marl beds (probably the Yorktown For- mation) in Edgecombe County, North Carolina, which he con- sidered to be a "fish tooth," attached by ligaments in the throat. Emmons's fish tooth is not a tooth but a hyperostosed bone, the identity of which has eluded paleontologists and ichthyologists for over 140 years. At Lee Creek Mine only a few of these enigmatic forms have been found in the thin unit 3 of the Yorktown Formation. Unit 3 was described by Gibson (1967:646) as "2 feet of blue clayey fine sand," and the irregular contact and color contrast with the underlying unit 2 are shown very well in his pl. 1. Small, isolat- ed patches of unit 3 sediments can readily be spotted on the weathered spoil piles by their darker color and finer grain tex- ture. Close inspection of these patches has yielded distinctive small fossils (otoliths, porcupinefish spines, crab chelae, etc.), including the examples of Emmons's "tooth." Emmons's tooth is bilaterally symmetrical, consists of dense, brittle bone, and has a maximum height of 13 mm. It is bean- shaped but more attenuated at one end than the other, with a sulcus down the midline of the convex side, and some speci- mens split along this plane. On the concave side of the "bean," there is a longitudinal, deep groove at the midline. It is also on the concave side that the attenuated end curls inward, forming a "beak." In well-preserved specimens, the groove forms the back wall of a bifurcating tubular process that allowed the pas- sage of nerves and blood vessels. Paleoecology The 104 species of Lee Creek Mine fossil fishes represent the first fossil record of a marine vertebrate, high-use feeding area and the largest and most diverse fossil fish fauna known from the Atlantic Coastal Plain. Of these taxa, 55 are found in the Pungo River Formation, and 77 are found in the Yorktown Formation. Because we were unable to bulk-sample the York- town fauna, it may be even more diverse than available evi- dence indicates. NUMBER 90 189 FIGURE 83.—Mola chelonopsls, USNM 265650, fused premaxillae: a, occlusal view; b, lateral view of right side. Emmons's "fish tooth," USNM 421518, anatomical orientation unknown: c, front view; d, lateral view; e, rear view. (Scale bars= 1.0 cm.) These faunas, and their associated sedimentological and in- vertebrate paleontological data, permit us to interpret the paleo- ecology of the Pungo River and Yorktown seas. Temperature Lee Creek Mine is at the southwestern extremity of the Au- rora embayment, a deep Tertiary embayment (Figure 84) of the western Atlantic continental shelf (Popenoe, 1985). During the Miocene and early Pliocene, the 100 m depth contour was near the present mouth of the Pamlico River. During Pungo River and basal Yorktown time (Riggs, 1984; Snyder, 1988), this em- bayment allowed upwelling, colder waters to intrude over 100 km westward onto the continental shelf, an area covered by warm-temperate to subtropical surface waters. The Lee Creek fauna reflects this contrast in temperatures. Gibson (1967) concluded from his study of benthic foramin- ifera that the bottom temperatures during Pungo River and low- er Yorktown deposition were cool-temperate. In addition, Gib- son's (1987) study of the pectens demonstrated a greater similarity to those of the Calvert Formation to the north than to those of similar ages in Florida. In contrast, approximately two-thirds of the modern representatives of the fossil fish fauna of both formations now have tropical and/or warm-temperate distributions, whereas only one-third of the fauna suggests a cool-temperate influence. Pungo River Formation.—Taxa that can be broadly grouped as eurythermic tropical and warm-temperate domi- nate (70%) the Pungo River fauna. These are species that mi- grate to avoid temperature extremes (>25°-28°C, <15°C), coming north during the spring and summer and returning south during the winter. This fauna has a distinctly warm but not a tropical character. The following examples support this warm character and suggest temperatures no colder than 20°C during the warm season. Stingrays (Dasyatis, the dominant batoid) and other rays (Rhinoptera, Mobula) occur in the Carolinian province (North Florida to Cape Hatteras) only during the summer, when sur- face temperatures are between 20°C and 27°C (Bigelow and Schroeder, 1953:11). Except Rhinoptera, which seems to toler- ate cooler temperatures, they are not usually found north of the 21°C isotherm. The most common shark taxa at Lee Creek Mine are warm- water carcharhiniforms: twenty of the species (30%) and 11 of Middle Miocene coastline FIGURE 84.—Map showing location of Lee Creek Mine in relation to the Aurora Embayment (after Popenoe, 1985, published with permission of Kluwer Academic Publishers). 190 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY the genera (22%). This dominance also applies to the relative abundance of carcharhiniform remains. In samples obtained by screening substantial quantities of the phosphate ore, teeth of Carcharhinus macloti are as numerous as those of the other shark taxa combined. Following Carcharhinus macloti in these screen residues are (in decreasing order of abundance) Car- charhinus brachyurus, Galeocerdo, Hemipristis, and Sphyrna cf. S. media. The abundance of carcharhiniforms compared with other galeoid sharks parallels modern warm-temperate and tropical faunas. Also, the diversity of carcharhinid genera exceeds that now present on the middle Atlantic Seaboard and is similar to the diversity encountered in tropical seas. Temperature range during the Pungo River deposition was probably at least as great as that of the Carolina province today (15°-25°C). Warm-water taxa such as Sphyraena, Negaprion, Sphyrna cf. S. media, and Hemipristis suggest maximum tem- peratures as great as 27°C. None of the Pungo River taxa are found only in cool waters; therefore, we believe that warm- temperate to subtropical temperatures persisted year round. The warmth of the Pungo River sea may account for the lack of abundance of tuna and other pelagic fishes that are common in the Yorktown Formation. Horwood and Cushing (1978) ob- served that the largest stocks of pelagic fish are associated with thermal boundaries where the waters are high in nutrients. In his study on the phosphate-phosphorus and zooplankton vol- umes of the Pacific Ocean, Reid (1962:302) noted two condi- tions, which may also have affected the productivity of the Pungo River sea: (1) if the depth of the boundary between the warm surface water and the denser, cooler upwelling water is deeper than 75 m, the concentrations of phosphate-phosphorus, which are utilized by plankton, are low; and (2) "high tempera- tures are usually accompanied by lower plankton volumes." Popenoe (1985) noted also that in Pungo River time the Aurora embayment was isolated from both northward- and southward- flowing currents; the embayment was an area of quiet deposi- tion. These conditions may have reduced the productivity of the Pungo River sea, which would account for the scarcity of oceanic fish, such as tunas, and sperm whales (see below), which are common in the Yorktown Formation. Yorktown Formation.—Unlike the Pungo River fauna, the Yorktown fauna has a distinctly modem aspect, particularly among the teleosts, so paleoecologic inferences are more defi- nite. The dominant taxa are sturgeon (Acipenser), scombrids (Thunnus, Sarda, Acanthocybium), billfish (Makaira), filefish (Aluterus), hakes (Merluccius), bluefish (Pomatomus), tilefish (Lopholatilus), pufferfish (Sphoeroides), porcupinefish (Chilo- mycterus), and the sharks—Isurus, Hemipristis, and Galeocer- do. Except Hemipristis, this is a mix of tropical, warm-temper- ate, and cool-temperate forms with ranges that overlap near Cape Hatteras today, but their abundance suggests that a ther- mocline existed in the Lee Creek area of the Yorktown sea. The cool-temperate components of the assemblage (seven taxa, 15% of the fauna) include the genera Merluccius and Lop- holatilus, which are common at Lee Creek Mine, and the rarer Squalus, although its rarity may be due to collection bias. These taxa suggest temperatures in the 4°-18°C range (Grosslein and Azarovitz, 1982) (10°C is the optimum temper- ature for Squalus acanthias, according to the distributional data of Edwards et al., 1962). The abundance of Merluccius and Lo- pholatilus at Lee Creek Mine suggests that cool water was an important component of the Yorktown sea. The warm component of the fauna (14 taxa, 30% of the fau- na) includes the genera Sphyraena, Makaira, Acanthocybium, Epinephelus, Hemipristis, Isistius, Negaprion, and Pristis. Be- cause they occur exclusively in tropical areas, the temperature regime suggested by these fishes is 27°C (the temperature boundary between warm-temperate and tropical in the western Atlantic Ocean) or warmer. Abundant tuna, marlin, sea birds, sea turtles, and cetaceans suggest that the area was an important feeding ground. Off the west coast of Central America today, in waters underlain by a sharp thermocline, a similar abundance of marine vertebrates occurs: feeding tuna force schools of prey fish to the surface where cetaceans, sea turtles, and sea birds feed on them (Au and Perryman, 1985). Au (1991:346) elaborated on the make- up of these polyspecific associations with tuna schools, noting that "sharks clearly stood out among the fishes found associat- ed with tuna." The most common species was Carcharhinus falciformis; less common sharks were C. longimanus, C. leu- cas, and Rhincodon typus. The rays observed were mostly "me- dium- to large-sized manta rays (Mobulidae)" (Au, 1991:346). Among the bony fishes present, he reported billfishes, includ- ing Makaira, dolphinfish (Coryphaena), amberjack (Seriola sp.), wahoo (Acanthocybium solandri), and triggerfish (Balis- tidae). Cushing (1971), in his study of upwelling and the pro- duction of fish, mentioned that hake (Merluccius) also are abundant in upwelling areas. Au (1991) could identify only three of the sea turtles he observed: Lepidochelys olivacea, Dermochelys coriacea, and Chelonia mydas. Boobies (Sula spp.), shearwaters (Puffinus spp.), and frigatebirds Fregata spp.) were the most common birds in the area, and the porpois- es included Stenella spp. and Delphinus delphis (Au, 1991). With the exceptions of Carcharhinus longimanus and Dermo- chelys coriacea, all of these genera occur in the Yorktown fau- na. Their abundance at Lee Creek Mine suggests the presence of nutrient-rich, cold, upwelling waters meeting warm water from the Gulf Stream to provide an important feeding zone for both warm- and cold-water fishes and other marine vertebrates. The abundance of tuna and sperm whale remains in the Yorktown Formation suggests a divergence of the Gulf Stream into the Aurora embayment. Cushing (1971:315) noted, "The coastal upwellings are less important to the tuna than offshore divergences. ...Although tuna may be caught in the area of coastal upwelling, it is likely that they pass 100 km or more offshore." Concerning sperm whale catches, Cushing (1971:320) stated, "This suggests that the sperm whales do not in fact aggregate in the coastal upwellings themselves but in offshore divergences.... Thus the sperm whale must be a truly NUMBER 90 191 oceanic animal like the tuna and may be excluded to some ex- tent from the coastal upwelling themselves." At Lee Creek Mine, 100 km west of the edge of the continental shelf, these two oceanic vertebrates occur abundantly. Depth Three general bathytrophic groups of fishes occur at Lee Creek Mine (percentage constituted by the Pungo River assem- blage and the Yorktown assemblage, respectively, follows): predominantly coastal fishes of the inner shelf (0-100 m) (33%, 49%), outer-shelf benthic or mesopelagic fishes (100-1000 m) (12%, 9%), and epipelagic fishes (45%, 21%). About 9% of the Pungo River assemblage and 11% of the Yorktown assemblage are benthic groups that may occur over either the inner or outer shelf; these are excluded in the figures given above. In the Pungo River assemblage, the relative rarity of coastal fishes and the abundance of epipelagic, outer shelf, benthic, and mesopelagic fishes suggest a deeper depositional environ- ment than that prevailing during the Yorktown deposition. The abundance of "blue water" fishes (open-ocean epipelagic spe- cies, such as Thunnus thynnus and Acanthocybium solandri) and the rarity of strictly near-shore species suggest deposition of the Yorktown Formation at the Lee Creek locality occurred in the deeper part of the inner shelf but not as deep as that of the Pungo River Formation. In both formations, the same groups of coastal benthic fishes limit the maximum depth. Stingrays (Dasyatis), which are common in both formations, occur mostly in water shallower than 37 m and are unknown in water deeper than 110 m (Big- elow and Schroeder, 1953). Lagodon ranges from the shore down to 74 m. The largest individuals of L. rhomboides occur in the deepest part of the range (Caldwell, 1957); specimens from the Pungo River Formation are near the maximum report- ed size of the extant species. In the Yorktown assemblage, the only taxa present that pro- vide some indication of minimum depth are Hexanchus, Isis- tius, Scyliorhinus, a tilefish (Lopholatilus), and a hake (Merluc- cius). Both species of Hexanchus are mesopelagic or deep epipelagic in warm-temperate and tropical areas; they are usu- ally found at depths of 90 m and greater (Compagno, 1984). Similarly, the species of Isistius inhabit epipelagic to bathype- lagic waters (Compagno, 1984:93-96). The genus Scyliorhinus inhabits the outer shelf and continental slope, but there are ex- ceptions. The fossil scyliorhinid S.l distans, which occurs in the Yorktown Formation, cannot now be satisfactorily related to any extant member of the genus, so any inference as to its depth tolerance would be suspect. Merluccius and Lopholatilus are more limited in occurrence. If the habitats of Merluccius sp. and M. bilinearis are alike, fairly deep water is indicated. Merluccius bilinearis ranges from the shoreline to a depth of about 550 m off the New En- gland coast (Grosslein and Azarovitz, 1982:72). It is not known from the Carolinian coast, but it has been taken sporadically off the Virginia coast in waters between 160 and 350 m deep (Hildebrand and Schroeder, 1928). Winter ranges are some- what shallower along the middle Atlantic Coast. Edwards et al. (1962) recorded occurrences of M. bilinearis as shallow as 31 m and in relative abundance between 131 and 316 m off the mouth of Chesapeake Bay. According to Grosslein and Azaro- vitz (1982:83), Lopholatilus "ranges in depth from about 75 to 460 m," and in the Middle Atlantic Bight, "it is concentrated in depths of approximately 110 to 240 m." In view of the latitude of the deposit and the inferred temperature, it is unlikely that depths were much shallower than 50 to 60 m. In the Pungo River Formation, no species were found that would suggest a minimum depth. Based on foraminiferal as- semblages, Gibson (1983:63-64) reported that the lower Bel- haven Member "formed on the middle to outer shelf (approxi- mately 100- to 200-m water depth)" and the upper Bonnerton Member "formed on the middle to inner shelf (150 m to less than 70 m in the uppermost bed)." Paleoecological Comparisons Pungo River Fauna.—When we compared this fauna to other fossil fish faunas in the Atlantic Coastal Plain and Eu- rope, we discovered some important differences. The taxonom- ic composition of the two formations is very similar to that of the Calvert Formation of Maryland and Virginia, with three significant exceptions: the absence or scarcity of the mako shark Isurus xiphodon, the absence of tilefish, Lopholatilus, and the absence of sturgeon, Acipenser, in the Pungo River fau- na. The absence of the mako shark may be due to the absence or rarity of pinnipeds in the Pungo River Formation, which were present in the Calvert sea. Tilefish are burrowers (Grosslein and Azarovitz, 1982) that inhabit areas with clay and boulder substrates; the abundance of sand, which is unsuit- able for tilefish burrows, in the Pungo River Formation may explain their absence here. We found no clues to explain the absence of sturgeon. In the Calvert Formation near Smyrna, Delaware, which was deposited in shallow coastal or estuarine waters (Purdy, in prep.), the fish fauna differs from that of the Calvert Formation of Maryland and Virginia and the Pungo River Formation. Ex- cept for sea catfish at Smyrna, teleost remains are less com- mon. Unlike the Pungo River fauna, the teeth of Carcharodon megalodon, which represent very small individuals, and Isurus are uncommon. In contrast to the rarity of these species, the teeth of adult Negaprion, which are rare in the Calvert and Pungo River faunas, are very common. The shallower depths of the Delaware portion of the Calvert sea probably account for these faunal differences. A slightly older (Aquitanian) fauna from the Belgrade For- mation (Case, 1980), 47 km (30 mi) south of Lee Creek Mine at New Bern, is at the northern limit of the Atlantic Coastal Plain Miocene limestone beds and contains a subtropical fauna. 192 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY It differs from the Pungo River shark fauna in the presence of Heterodontus and Carcharhinus isodon (identified by Case (1980) as Aprtonodon acuarias). In the extant shark fauna, both taxa prefer water temperatures above 20°C (Compagno, 1984; Castro, 1983). The Lee Creek Pungo River sea may have been slightly cooler than these sharks preferred. Gillette (1984) noted a similarity between the Miocene fau- nas of Panama, Ecuador, and the Caribbean region and the Pungo River Formation, but some important differences exist. Odontaspids and species of Isurus are absent from the first three faunas and in them the number of carcharhiniform sharks (Panama, 8; Caribbean, 4; Ecuador, 7) is much less than the number present (19) in the Pungo River Formation. Although the taxa present in these faunas are similar to those in the Pun- go River Formation, they are cosmopolitan, warm-water taxa, and their presence sheds little light on the similarity of the en- vironments for these faunas. In comparison to the European Miocene, the Pungo River Formation has a greater abundance of carcharhiniform sharks (19 species) than either the Belgian Miocene (4 species) (Ler- iche, 1927; Nolf, 1988) or the Swiss Molasse (8 species) (Ler- iche, 1926). The Belgian Miocene sea appears to have been cooler than that of the Swiss Molasse or the Pungo River For- mation; Hemipristis, a subtropical shark common to the last two, was absent in the Belgian Miocene sea. Absent from both the Belgian and Swiss faunas is the subtropical lemon shark, Negaprion, which suggests that the Pungo River sea was warmer than those of Belgium or Switzerland. The differences between these faunas reflect the important influence of ecological factors on the distribution offish taxa, particularly sharks, an influence that must be considered in stratigraphic interpretations even in geographically close areas. Yorktown Fauna.—Unlike the Yorktown Formation, fish remains are uncommon in the Raysor Formation of South Carolina and the Duplin Formation of North and South Caroli- na, both of which were deposited in shallow, warm water dur- ing the early Pliocene (Ward et al., 1991). We know of no re- ports of fish remains from the Raysor Formation. From Duplin localities (south of the Neuse River), Leriche (1942) reported the presence of Carcharodon carcharias and Chilomycterus vetus, and from the Martin Marietta quarry at New Bern, ama- teur collectors Peter J. Harmatuk and Bob Johnson (pers. comm., 1991) reported the occurrence of Carcharodon megal- odon. These shallow, warm-water deposits were evidently not high-use areas for fishes and other vertebrates. The European Pliocene fish faunas offer some interesting contrast to the Yorktown fauna. They have fewer taxa, and they represent cooler environments. In the Belgian Pliocene fish fauna (Leriche, 1926), warm-water forms, such as Hemipristis, Negaprion, and Galeocerdo, are absent, and cold-water taxa, such as Gadus, Merluccius, Carcharodon carcharias, and Ce- torhinus, are present. In the Pliocene of Italy (Landini, 1976) more taxa are present than in Belgium, including warm-water species, but at least two of the shark taxa, Parotodus benedenii and Galeocerdo cf. G. cuvier, have smaller teeth than do those of the Yorktown Formation, and two subtropical taxa, Hemi- pristis and Negaprion, are rare or absent. In comparison with other Pliocene faunas of the Atlantic Coastal Plain and Europe, the Yorktown fauna at Lee Creek Mine has the greatest abundance and diversity of fossil fish and other marine vertebrates. This suggests to us that at Lee Creek in Yorktown time, the physiography of the environment of dep- osition, the presence at shallow depths of upwelling waters, and the presence of an eddy of the Gulf Stream were significant factors contributing to the congregation of many marine verte- brates in this area. Taxonomic Discussion The abundant remains at Lee Creek Mine allowed us to re- construct composite dentitions and to resolve some important problems in the taxonomy of fossil sharks, problems that can- not be resolved with a few isolated teeth. For example, by com- paring the anterior teeth of Carcharodon and Carcharocles, we were able to reassign the teeth of the latter genus to the genus Carcharodon. The absence of ontogenetic heterodonty be- tween juvenile and adult dentitions allowed us to separate Isu- rus xiphodon from /. hastalis. This fauna also provided suffi- cient specimens of Parotodus benedenii for us to reconstruct its dentition, and based on the mako-like nature of characters of the roots and the size relationships between the upper and low- er teeth we assigned this species to the Lamnidae. Composite dentitions, however, have their pitfalls. Shark teeth vary greatly within a species, even in the anterior teeth. Before we knew of the variation in mako teeth, we reconstruct- ed a composite dentition of the invalid species Isurus retroflex- us. The Miocene species of Galeocerdo are another example. Applegate (1978) believed that the teeth of Galeocerdo contor- tus were the lower teeth of G. aduncus Agassiz. Although his reconstruction seems very logical, we were able to identify characters that distinguish upper and lower teeth in the denti- tions of the extant Galeocerdo. When we applied these charac- ters to the teeth of fossil species, we were able to separate Ga- leocerdo contortus from G. aduncus (=G. sp. herein). Our experience with composite dentitions suggests that an exten- sive knowledge of tooth characters and variation in related ex- tant species is a prerequisite to reconstructing fossil shark den- titions accurately. Associated fossil dentitions are the most important tools of shark paleontology. At Lee Creek Mine, two associated denti- tions of Carcharodon subauriculatus, two of C megalodon, and one of Parotodus benedenii have been found, and these are more common than previously thought. The current commer- cial market for shark teeth, however, and the apathy of many amateur collectors toward preserving these scientifically im- portant specimens in museums will make these specimens harder to obtain for scientific research, and thus retard progress in understanding the systematics and evolution of fossil sharks. NUMBER 90 193 The rediscovery of some of Agassiz's fossil-fish type speci- mens allowed us to evaluate some of his species. Although Agassiz's syntypes of Isurus desori had not been lost, in over 150 years they were never examined and compared to the den- titions of extant makos. When we made these comparisons, be- sides designating a lectotype, we determined that one tooth of Agassiz's type series was definitely from a mako and that it is a junior synonym of /. oxyrinchus. In contrast, the unknown whereabouts of the holotype of Probst's Oxyrhina exigua (^Alopias exigua) leaves unresolved the generic identity of his species. Many of the Lee Creek Mine taxa cannot be separated from the living forms, for example, the Lee Creek specimens identi- fied as Notorynchus cepedianus, Isurus oxyrinchus, Carcharhi- nus brachyurus, Cfalciformis, C leucas, C macloti, C. obscu- rus, C. perezi, C. plumbeus, Negaprion brevirostris, Triaenodon obesus, Sphyrna lewini, and S. zygaena, which at- tests to the relative longevity of many shark species. Further study of the variability in the teeth of extant species may result in the synonymy of fossil species, such as Hexanchus gigas, with the extant species and extant species, such as Echinorhi- nus cookei and Isurus paucus, with the fossil species. Conclusions The Lee Creek fossil fish faunas are important because of their diversity and their abundance and because of the paleo- ecological information they yield about the Pungo River and Yorktown seas. Based on the taxa present and other geological data, we offer the following conclusions. First, the fish faunas suggest that the Pungo River sea was warm temperate to subtropical and that the Yorktown sea was warm temperate with the presence of upwelling cold water. Second, in Yorktown time this upwelling water supported a large vertebrate fauna of fish, sea turtles, sea birds, cetaceans, and pinnipeds. In Pungo River time, upwelling did not reach shallow depths and the fauna was not as diverse as in York- town time. Third, the Pungo River fish fauna suggests that it was depos- ited in an epipelagic, outer-shelf, benthic environment, whereas the Yorktown fish fauna suggests that it was deposited in an in- ner-shelf environment. Fourth, in comparing both Lee Creek faunas to those of the Atlantic Coastal Plain and Europe, the Pungo River fauna has the largest number of warm-water fishes but lacks the subtropi- cal forms found in the Belgrade Formation that outcrops 47 km to the south. In comparison to these other Atlantic Pliocene fish faunas, the Yorktown fauna has the greatest abundance and di- versity, and it has more warm-water forms than do the Europe- an faunas. Fifth, the Lee Creek sediments yielded new records of occur- rence for the Atlantic Coastal Plain for four fossil shark taxa: Rhincodon sp. from the Pungo River Formation, Megascylio- rhinus miocaenicus from the basal Pungo River Formation, Isistius sp. from the basal Yorktown Formation, and Mega- chasma sp. from the Pungo River and the Yorktown forma- tions. Sixth, the Yorktown Formation yielded two new species of bony fishes: Lopholatilus rayus and Pagrus hyneus. Seventh, the Lee Creek shark taxa suggest that shark species or tooth morphologies existed for several million years or more. Eighth, in the extant sharks, tooth form is highly variable; understanding this variation and knowing the tooth characters for each dental position is essential to reconstructing fossil shark dentitions accurately. Future Study As we stated earlier, our study is a beginning. Much work needs to be done in the study of the extant and fossil species. Recently, ichthyologists have made significant discoveries about the effects of segregation by size, water temperature, and ocean-floor topography upon the distribution of fishes (Munoz- Chapuli, 1984; Klimley, 1985; Galvan-Magana et al., 1989; Stevens, 1990; Smale, 1991; Klimley etal., 1992; Simpfendor- fer and Milward, 1993). Considering this information, the fos- sil record of Tertiary fishes needs to be reexamined in light of the ecology of modem fishes. In extant sharks, age/size differences can exist between two or more populations of a species (see Galeocerdo), and these differences may occur in other extant species of sharks and also in the fossil species. Shark vertebrae can be aged by ring counts, a technique that can be applied to fossil vertebrae, which are common at Lee Creek Mine. By applying this tech- nique to fossil faunas, the age and the mean tooth size of a fos- sil species could be compared with those from other areas. If, as seems possible (Applegate, 1967; Kozuch and Fitzgerald, 1989), shark vertebrae can be identified to species, the abun- dant shark vertebrae at Lee Creek Mine will be most useful for such a study, and by excavating the basal Yorktown Formation, some of these may be found associated with dentitions. A new area of investigation for paleontologists is the effect of segregation by size on the distribution of fossil sharks and rays. In the extant species, some sharks and rays segregate by size and/or sex. Females pup in areas that are safe from preda- tors, and as the pups increase in size, they migrate toward adult feeding areas. In many shark species, males and females—fe- males are usually larger than males—live and feed in different areas except during mating season. This segregation by size and by sex raises some questions about the fossil record of sharks: Do size increases in shark teeth reflect evolutionary or environmental changes? What types of paleoenvironments are fossil shark teeth found in, and in comparison to the extant spe- cies, would these paleoenvironments favor sharks of one size or sex? Do the fossil teeth contribute to the interpretation of segregation by size or sex? The answers to these questions re- quire careful stratigraphic sampling of shark-tooth-bearing 194 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY strata, the analysis of paleoecological data, and the comparison of the fossil species with, if they exist, extant species of the same genus. Water temperature and ocean-floor topography affect the distribution of extant fishes, but their effects upon the distribu- tion of fossil fishes have not been studied. In the Yorktown sea, large sharks (Carcharodon megalodon, Isurus xiphodon, Hemipristis serra, Galeocerdo cf. G. cuvier) and schooling fishes (Merluccius, Pomatomus, Thunnus) are more abundant in the area of the Aurora embayment than they are at other Yorktown localities in North Carolina and Virginia. The pres- ence of a submarine canyon or embayment and upwelling af- fected the distribution of these fishes. We can ask, What effects may water temperature and ocean floor topography have had on other fossil fish faunas? At Lee Creek Mine, do the differ- ences in the abundance and the diversity of taxa between units 1 and 3 reflect changes in water temperature? How great is the taxonomic difference between these two units? The fossil beds at Lee Creek Mine offer the potential for an- swering some of these questions and for increasing our knowl- edge about the taxa that lived there, their environment, and their relationships with other fossil organisms that lived with them. A careful excavation of the fossil beds in units 1 to 3 of the Yorktown Formation would greatly increase our knowl- edge of the fauna. It would allow us to observe any faunal changes that occur in these units that cannot be observed ade- quately by spoil-pile collecting. It would allow us to gather ad- ditional information about predator-prey relationships. It would allow us to obtain large samples of the fauna, including many articulated skeletons, which are so important to the studies of taxonomy and functional anatomy. Such an excavation would allow us to create a more detailed and accurate picture of life in the Yorktown sea. Lee Creek Mine is a hint of the fossil beds that may lie bur- ied deep beneath the sea. In comparison to the limited and geochronologically incomplete outcrops of Tertiary sedi- ments, the paleontological record of the outer continental shelf, where today marine vertebrates are abundant, is poten- tially much more extensive, and it may provide a more contin- uous record of the evolution of marine environments and ma- rine vertebrates. The Yorktown fauna at Lee Creek Mine may be an indication of the species diversity of marine vertebrate faunas that existed during the Neogene along the edge of the continental shelf of eastern North America. Today, Lee Creek Mine is the only exposure of this part of the fossil record available to us, a part of the fossil record that we have only begun to understand. NOTE ADDED IN PROOF After our paper was submitted for publication, we received three publications, one by Bourdon (1999) and two by Arnold Muller (1992, 1999), on the fossil fishes of the Atlantic Coastal Plain Tertiary sediments. Bourdon, on the basis of teeth, de- scribed a new species of manta ray, Manta hynei. Because little is known about dental variation in recent manta rays, we cannot assess the validity of this species, but we think the naming of it is premature. Miiller's publications are all but unavailable in the United States, and we were unable to comment on his findings in the present paper. His 1999 work is the publication of his thesis (1992), which is based on about 12,000 otoliths and some 1000 elasmobranch teeth collected in the Eocene-Pliocene sediments of the Atlantic Coastal Plain, including Lee Creek Mine. None of his new species has priority over the two de- scribed herein. Literature Cited Agassiz, L. 1833-1843. Recherches sur les poissons fossiles. Volume 3, i-viii +390 pages, atlas of 82 plates. Neuchatel. 1856. Notice of the Fossil Fishes Found in California by W.P Blake. American Journal of Science, 2(21 ):272-275. Almeida, F.P., TR. Azarovitz, L. O'Brien, and E.W. Pritchard 1984. The Distribution of Major Finfish and Shellfish Species Collected during NEFC Bottom Trawl Surveys, 1965-1978. Woods Hole Lab- oratory Reference Document, number 84-21: 101 pages, 95 figures. National Marine Fisheries Service, Northeast Fisheries Center. Ameghino, F. 1901. L'age des formations sedimentaires de Patagonie. Anales de la So- ciedad Cientifica Argentina. 51:65-91. 1906. Les formations sedimentaires du Cretace superieur ct du Tertiarc de Patagonie avec un parallele entre leurs faunes mammalogiques et celles de l'ancien continent. Anales del Museo Nacional de Buenos Aires, 3(15)8:1-508, 358 figures, 3 plates. Antunes, M. 1963. Sur quelques requins de la faune Neogene de Farol das Lagostas (Luanda, Angola). Memoires de I'Institut Frangais d'Afrique Noire, 68:47-63. 1969-1970. Sur Lamna cattica ssp. lotuserrata; un cas de distribution an- tiequatorialc. Revista da Faculdade de Ciencias, Universidade de Lisboa, 16:37-62, plates 1-6. 1978. Faunes ichthyologiques du Neogene superieur d'Angola, Ieur age, remarques sur le Pliocene marin en Afrique Australe. Ciencias da Terra. 4:59-90, plates 1-3. Antunes, M., and S. Jonet 1969-1970. Requins de I'Helvetien superior et du Tortonien de Lisbonne. Revista da Faculdade de Ciencias, Universidade de Lisboa, 16: 119-280, plates 1-20. Applegate, SP 1965a. Tooth Terminology and Variation in Sharks with Special Reference to the Sand Shark, Carcharias taurus Rafinesque. Contributions in Science, Los Angeles County Museum, 86:1-18. NUMBER 90 195 1965b. A Confirmation of the Validity of Notorhynchus pectinatus; the Second Record of This Upper Cretaceous Cowshark. Bulletin of the Southern California Academy of Sciences, 64:122-126. 1967. A Survey of Shark Hard Parts. In P.W. Gilbert, R.F. Mathewson, and D.P. Rail, editors, Sharks, Skates, and Rays, pages 37-67, 7 fig- ures, 4 plates. Baltimore: Johns Hopkins Press. 1978. Phyletic Studies, Part 1: Tiger Sharks. Universidad Nacional Au- tonoma de Mexico, Instituto de Geologia, Revista, 2( 1 ):55-64, 9 fig- ures. 1986. The El Cien Formation, Strata of Oligocene and Early Miocene Age in Baja California Sur. Universidad Nacional Autonoma de Mexico, Instituto de Geologia, Revista, 6:145-162. Applegate, S.P., and L. Espinosa-Arrubarrena 1996. The Fossil History oi Carcharodon and Its Possible Ancestor, Cre- tolamna: A Study in Tooth Identification. In A.P. Klimley and D.G. Ainley, editors, Great White Sharks: The Biology of Carcharodon carcharias, pages 19-36. San Diego: Academic Press. Arambourg, C. 1927. Les poissons fossiles d'Oran. Materiaux pour la Carte Geologique del'Algerie, series 1, Palaeontologie, 6:1-298, 66 plates. 1935. Note preliminaire sur les vertebres fossiles des phosphates du Ma- roc. Bulletin de la Societe Geologique de France, 5(5):413-439, plates, 19,20. 1952. Les vertebres fossiles des gisements de phosphates (Maroc-Algerie- Tunisie). Notes et Memoires. Service des Mines et de la Carte Geologique (Rabat), 92:1-372, 62 figures, 46 plates, 7 tables. Au, D.W. 1991. Polyspecific Nature of Tuna Schools: Shark, Dolphin, and Seabird Associates. Fishery Bulletin, 89:343-354. Au, D.W.K., and W.L. Perryman 1985. Dolphin Habitats in the Eastern Tropical Pacific. Fishery Bulletin, 83(4):623-643, 12 figures. Bass, A.J., J.D. D'Aubrey, and N. Kistnasamy 1973. Sharks of the East Coast of Southern Africa: The Genus Carcharhi- nus (Carcharhinidae). Investigational Report, Oceanographical Re- search Institute, 33:1-68, 32 figures, 17 plates. 1975a. Sharks of the East Coast of Southern Africa: The Families Scyliorhinidae and Pseudotriakidae. Investigational Report, Ocean- ographical Research Institute, 37: 63 pages, 21 figures. 1975b. Sharks of the East Coast of Southern Africa: The Families Carchar- hinidae (Excluding Carcharhinus and Mustelus) and Sphyrnidae. Investigational Report, Oceanographical Research Institute, 38: 100 pages, 26 figures, 12 plates. 1975c. Sharks of the East Coast of Southern Africa: The Families Odon- taspidae, Scapanorhynchidae, Isuridae, Cetorhinidae, Alopiidae, Orectolobidae, and Rhiniodontidae. Investigational Report, Ocean- ographical Research Institute. 39: 102 pages, 24 figures, 15 plates. 1975d. The Families Hexanchidae, Chlamydoselachidae, Heterodontidae, Pristiophoridae, and Squatinidae. Investigational Report, Oceano- graphical Research Institute, 43: 50 pages, 12 figures, 9 plates. 1976. The Families Oxynotidae, Squalidae, Dalatiidae, and Echinorhin- idae. Investigational Report, Oceanographical Research Institute, 45: 103 pages, 36 figures, 11 plates. Bendix-Almgreen, S.E. 1983. Carcharodon megalodon from the Upper Miocene of Denmark, with Comments on Elasmobranch Tooth Enameloid: Coronoi'n. Bul- letin of the Geological Society of Denmark, 32:1-32, 10 figures. Berry, C.T. 1932. Some Miocene Teeth Belonging to the Genus Lagodon. American Journal of Science, 24:303-305, 1 figure. Bigelow, H.B., and W.C. Schroeder 1948. Sharks. Memoir of the Sears Foundation for Marine Research, 1(1): 59-546, figures 6-105. 1953. Sawfishes, Guitarfishes, Skates and Rays. Memoir of the Sears Foundation for Marine Research, 1(2): xv+588 pages, 127 figures. Blainville, H.M.D. de 1816. Prodrome d'une nouvelle distribution systematique de regne animal. Bulletin de Sciences de la Societe Philomatique de Paris, part 8:113-124. Blake, CC. 1862. Fossil Shark Teeth at Panama. The Geologist, 5:316. Blake, S.F. 1953. The Pleistocene Fauna of Wailes Bluff and Langleys Bluff, Mary- land. Smithsonian Miscellaneous Collections, 121(12): 32 pages. Bloch, ME. 1801. ME. Blochii...Systema Ichthyologiae iconibusCX illustratum; post obitum auctoris opus inchoatum absolvit, correxit, interpolavit J.G. Schneider. 2 volumes, lx+584 pages, 110 plates. Berolini [=Berlin]. Bonnaterre, J.P. 1788. Ichthyologie. In Tableau encyclopedique et methodique des trois regnes de la nature, lvi+215 pages, 102 plates. Paris. Boschung, H.T., Jr., C. Nehring, and J. Verner 1983. The Audubon Society Field Guide to North American Fishes, Whales, and Dolphins, viii+847 pages. New York: Knopf. Bourdon, J. 1999. A Fossil Manta from the Early Pliocene (Zanclean) of North Amer- ica. Tertiary Research, 19:3—4. Branstetter, S., J.A. Musick, and J.A. Colvocoresses 1987. A Comparison of the Age and Growth of the Tiger Shark, Galeo- cerdo cuvieri, from off Virginia and from the Northwestern Gulf of Mexico. Fishery Bulletin, 85(2):269-279, 6 figures. Burnett, W.C, and D.Z. Piper 1977. Rapidly-Formed Ferromanganese Deposit from the Eastern Pacific Hess Deep. Nature, 265:596-600, 4 figures. Cadenat, J. 1962. Notes d'icthyologie ouest-africaine, XXXVIII: Documents pour servir a la recherche des mecanismes de deplacement et remplace- ment des dents chez les Requins. Bulletin de IFAN, A(24)2: 551-579, 24 plates. Caldwell, D.K. 1957. The Biology and Systematics of the Pin Fish, Lagodon rhomboides (Linnaeus). Bulletin of the Florida State Museum, Biological Sci- ences. 2:77-173, 22 figures. Capape, C, and J. Quignard 1975. Contribution a la systematique et a la biologie de Pteromylaeus bo- vinus (Geoffroy Saint-Hilaire, 1817) (Pisces Myliobatidae) de cotes tunisiennes. Bulletin du Museum National d'Histoire Naturelle (Zo- ologie), 3(338):1329-1347, 8 figures. Cappellini, G. 1878. Delia pietra leccese e di alcuni suoi fossili. Memorie delta Reale Ac- cademia delle Scienze dell'Istituto di Bologna, 3(9):227-258, 3 plates. Cappetta, H. 1970. Les selachiens du Miocene de la region de Montpellier. Palaeover- tebrata. Memoir Extraordinaire, 139 pages, 22 figures, 27 plates, 6 tables. 1975. Selaciens et holocephale du Gargasien de la region de Gargas (Van- cluse). Universite de Provence, Annales, Geologie Mediter- raneenne, 2(3): 115-134. 1980. Modification du statut generique de quelques especes de selaciens cretaces et tertiaires. Palaeoverlebrata, 10(1):29—42, 6 figures. 1987. Handbook of Paleoichthyology; Chondrichthyes, II: Mesozoic and Cenozoic Elasmobranchii. 193 pages, 148 figures. Stuttgart and New York: Gustav Fischer Verlag. Cappetta, H., and D.J. Ward 1977. A New Eocene Shark from the London Clay of Essex. Palaeontol- ogy, 20(1): 195-202. Caretto, P.G. 1972. Osservazioni tassonomiche su alcuni Galeoidei del Miocene pie- montese. Bollettino delta Societa Paleontologica Italiana, 11(1): 14-85, 7 figures, plates 3-14. 196 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Caria, I.C 1973. / pesci del Miocene della Sardegna. 39 pages. Cagliari: Stabili- mento Tipografico Editoriale Fossataro. Carroll, R.L. 1988. Vertebrate Paleontology and Evolution. 698 pages. New York: W.H. Freeman and Co. Case, G.R. 1980. A Selachian Fauna from the Trent Formation, Lower Miocene (Aquitanian) of Eastern North Carolina. Palaeontographica, series A (Palaozoologie), 171(l-3):75-103, 6 figures, 10 plates. Casey, J.G., and H.L. Pratt, Jr. 1985. Distribution of the White Shark, Carcharodon carcharias, in the Western North Atlantic. Southern California Academy of Sciences Memoir, 9:2-14,4 figures. Casier, E. 1946. La faune ichthyologique de l'Ypresien de la Belgique. Memoires du Musee Royal d'Histoire Naturelle de Belgique, 104:1-267, 6 plates. 1960. Note sur la collection des poissons paleocenes et eocenes de I'En- clave de Cabinda (Congo). Annales du Musee Royal du Congo Beige, series A, 3(1)2:1^18, plates 1, 2. Castro, J.I. 1983. The Sharks of North American Waters. 180 pages, illustrated. Col- lege Station: Texas A & M University Press. Chariesworth, E. 1837. Notice of the Teeth of Carcharias megalodon Occurring in the Red Crag of Suffolk. The Magazine of Natural History, 1(5):225—227. Choa, L.N. 1978. A Basis for Classifying Western Atlantic Sciaenidae (Teleostei: Per- ciformes). National Oceanographic and Atmospheric Administra- tion Technical Report, National Marine Fisheries Service Circular, 415:1-64. Cigala-Fulgosi, F. 1983. First Record oi Alopias superciliosus (Lowe, 1840) in the Mediter- ranean, with Notes on Some Fossil Species of the Genus Alopias. Annali del Museo Civico di Storia Naturale di Genova, 84:211-229. 1988. Additions to the Eocene and Pliocene Fish Fauna of Italy; Evidence of Alopias cf. denticulatus. Cappetta, 1981 in the Bartonian-Pria- bonian of the Monte Piano Marl (Northern Apennines) and oiA. su- perciliosus (Lowe, 1840) in the Pliocene of Tuscany (Chondrichthyes, Alopiidae). Tertiary Research, IO(2):93-99, 1 fig- ure, 1 plate. Cigala-Fulgosi, F., and D. Mori 1979. Osservazioni tassonomiche sul genere Galeocerdo (Selachii, Car- charhinidae) (Peron & Lesueur) nel Pliocene del Mediterraneo. Bol- lettino della Societa Paleontologica Italiana, 18(1):117-132, 4 figures, 1 plate. Cione, A.L., and M.A. Reguero 1994. New Records of the Sharks Isurus and Hexanchus from the Eocene of Seymour Island, Antarctica. Proceedings of the Geologists' Asso- ciation, 105(1):1-14. Cliff, C, S.F.J. Dudley, and B. Davis 1989. Sharks Caught in the Protective Gill Nets off Natal, South Africa, 2: The Great White Shark Carcharodon carcharias (Linnaeus). South African Journal of Marine Science, 8:131 -144. Collette, B.B., and L.N. Chao 1975. Systematics and Morphology of the Bonitos (Sarda) and Their Rela- tives (Scombridae, Sardini). United Stales Fish and Wildlife Ser- vice. Fishery Bulletin, 73(3):516-625. Collette, B.B., and J.L. Russo 1984. Morphology, Systematics, and Biology of the Spanish Mackerels (Scomberomorus, Scombridae). Fishery Bulletin, 82(4):545-692. Compagno, L.J.V 1977. Phyletic Relationships of Living Sharks and Rays. American Zoolo- gist. 17:303-322, 15 figures. 1984. FAO Species Catalogue, Volume 4: Sharks of the World. 655 pages, illustrated. Rome: United Nations Development Programme. 1988. Sharks of the Order Carcharhiniformes. xxii+486 pages, 35 plates. Princeton, New Jersey: Princeton University Press. 1990a. Relationships of the Megamouth Shark, Megachasma pelagios (Lamniformes: Megachasmidae), with Comments on Its Feeding Habits. National Oceanographic and Atmospheric Administration Technical Report. NMFS, 90:357-379, 12 figures. 1990b. Alternative Life-History Styles of Cartilaginous Fishes in Time and Space. Environmental Biology of Fishes, 28:33-75. Compagno, L.J.V, and W.I. Follett 1986. Carcharias Rafinesque, 1810 (Chondrichthyes, Lamniformes): Pro- posed Conservation by the Use of the Relative Precedence Proce- dure; Z.N.(S) 2414. Bulletin of Zoological Nomenclature, 43(1): 89-92. Cope, E.D. 1867. An Addition to the Vertebrate Fauna of the Miocene Period, with a Synopsis of the Extinct Cetacea of the United States. Proceedings of the Academy of Natural Sciences of Philadelphia, 19:138-156. 1869. Descriptions of Some Extinct Fishes Previously Unknown. Pro- ceedings of the Boston Society of Natural History, 12:310-317. 1871. Observations on the Distribution of Certain Extinct Vertebrata in North Carolina. Proceedings of the American Philosophical Society, 12:210-216. 1875. Synopsis of the Vertebrata Whose Remains Have Been Preserved in the Formations of North Carolina. In W.C. Kerr, editor. Report of the Geological Survey of North Carolina, 1; appendix B:29-52, plates 5-8. Cushing, D.H. 1971. Upwelling and the Production of Fish. Advances in Marine Biology, 9:255-334, 20 figures. Cuvier, G., and A. Valenciennes 1831. Histoire naturelle des poissons. Volume 8, i-xix+5+509 pages, plates 209-245. Paris. 1847. Histoire naturelle des poissons. Volume 19, xix+544 pages, plates 554-590. Paris Davis, J.W. 1888. On Fossil-Fish Remains from the Tertiary and Cretaceo-Tertiary Formations of New Zealand. The Scientific Transactions of the Royal Dublin Society, 2(6): 1-60, 6 plates. De Stefano, G. 1909. Osservazioni sulla ittiofauna pliocenica di Orciano e San Quirico in Toscana. Bollettino della Societa Geologica Italiana, 28:539-648, plates 16-20. Dockery, D.T., and E.M. Manning 1986. Teeth of the Giant Shark Carcharodon auriculatus from the Eocene and Oligocene of Mississippi. Mississippi Geology, 7(1):7-19, plates 1-4. Dooley, J.K. 1978. Systematics and Biology of the Tilefishes (Perciformes: Bran- chiostegidae and Malacanthidae), with Descriptions of Two New Species. National Oceanographic and Atmospheric Administration Technical Report, National Marine Fisheries Service Circular, 411: 1-78. Eastman, C.R. 1904. Pisces. In Maryland Geological Survey—Miocene, i—clv+543 pages, 135 plates. Baltimore: Johns Hopkins University Press. Ebert, D.A. 1986. Aspects on the Biology of Hexanchid Sharks along the California Coast. In T. Uyeno, R. Arai, T. Taniuchi, and K. Matsuura, editors, Indo-Pacific Fish Biology: Proceedings of the Second Interna- tional Conference on Indo-Pacific Fishes, pages 437-449, 8 fig- ures. Tokyo: Ichthyological Society of Japan. 1991. Diet of the Sevengill Shark Notorynchus cepedianus in the Temper- ate Coastal Waters of Southern Africa. South African Journal of Marine Science, 11:565-572. 1994. Diet of the Sixgill Shark Hexanchus griseus off Southern Africa. South African Journal of Marine Science, 14:213-218. NUMBER 90 197 Edwards, R.L., R. Livingstone, and P.E. Hamer 1962. Winter Water Temperatures and an Annotated List of Fishes—Nan- tucket Shoals to Cape Hatteras; Albatross III Cruise No. 126. Spe- cial Scientific Report, United States Fish and Wildlife Service, Fisheries Series, 397: 31 pages. Emmons, E. 1858. Report of the North Carolina Geological Survey: Agriculture of the Eastern Counties Together with Description of the Fossils of the Marl Beds. 314 pages, 256 figures. Raleigh, North Carolina: H.D. Turner. Espinosa-Arrubarrena, L. 1987. Neogene Species of the Genus Isurus (Elasmobranchii, Lamnidae) in Southern California, U.S.A. and Baja California Sur, Mexico. 153 pages, 19 figures. Master of Science thesis, California State Univer- sity, Long Beach, California. Espinosa-Arrubarrena, L., and S.P. Applegate 1981. Selacifauna Pliocenica de Baja California Sur, Mexico y sus proble- mas de correlacion regionale. In Y.T. Sanguinetti, editor, Anais: 11 Congresso Latino-Americano de Paleontologia, Universidade Fed- eral do Rio Grande do Sul, Porto Alegre, Rio Grande do Sul, Brasil, 2: 667-681, 1 figure, 3 plates. Puerto Alegre, Rio Grande do Sul: Editora Pallotti. Faber, F. 1829. Naturgeschichte der Fische Islands, mit einem Anhange von den is- landischen Medusen und Strahlthieren 206 pages. Frankfurt au Main: Druck und Verlag von Heinrich Ludwig Bronner. Fitch, J.E., and R.J. Lavenberg 1983. Teleost Fish Otoliths from Lee Creek Mine, Aurora, North Carolina (Yorktown Formation: Pliocene). In Clayton E. Ray, editor. Geol- ogy and Paleontology of the Lee Creek Mine, North Carolina, I. Smithsonian Contributions to Paleobiology, 53:509-529, 4 figures. Fowler, H.W. 1911. A Description of the Fossil Fish Remains of the Cretaceous, Eocene, and Miocene Formations of New Jersey. Bulletin, Geological Sur- vey of New Jersey, 4:22-182, 108 figures. 1941. The Fishes of the Groups Elasmobranchii, Holocephali, Isospondyli, and Ostariophysi Obtained by the United States Bureau of Fisheries Steamer "Albatross" in 1907 to 1910, Chiefly in the Philippine Is- lands and Adjacent Seas. United States National Museum Bulletin, 100(13): 1-879. Galvan-Magana, F., H.J. Nienhuis, and A.P. Klimley 1989. Seasonal Abundance and Feeding Habits of Sharks of the Lower Gulf of California, Mexico. California Fish and Game, 75(2): 74-84. Garman, S. 1913. The Plagiostomia. Memoirs of the Museum of Comparative Zoology at Harvard College, 36: 528 pages, plates 1-75. Garrick, J.A.F. 1960. Studies on New Zealand Elasmobranchii, X: The Genus Echinorhi- nus, with an Account of a Second Species, E. cookei Pietschmann, 1928, from New Zealand Waters. Transactions of the Royal Society of New Zealand, 88(1): 105—117, 3 figures, plate 7. 1967. Revision of Sharks of Genus Isurus with a Description of a New Species (Galeoidea, Lamnidae). Proceedings of the United States National Museum, 118(3537):663-690, 9 figures, 4 plates. 1982. Sharks of the Genus Carcharhinus. National Oceanic and Atmo- spheric Administration, National Marine Fisheries Service Circu- lar, 445:1-194, 83 figures, 1 plate. 1985. Additions to a Revision of the Shark Genus Carcharhinus: Synon- ymy oiAprionodon and Hypoprion, and Description of a New Spe- cies of Carcharhinus (Carcharhinidae). National Oceanic and Atmospheric Administration, National Marine Fisheries Service Technical Report, 34:1-26, 14 figures, 4 tables. Gibbes, R.W. 1848-1849. Monograph of the Fossil Squalidae of the United States. Jour- nal of the Academy of Natural Sciences of Philadelphia, series 2, 1:139-147, 191-206, plates 18-21,25-27. Gibbs, R.H., Jr., and B.B. Collette 1967. Comparative Anatomy and Systematics of the Tunas, Genus Thun- nus. Fishery Bulletin, United Stales Fish and Wildlife Service, 66: 65-130. Gibson, T.G. 1967. Stratigraphy and Paleoenvironment of the Phosphatic Miocene Strata of North Carolina. Bulletin of the Geological Society of Amer- ica, 78:631-649, 4 figures, 2 plates. 1983. Stratigraphy of Miocene through Lower Pleistocene Strata of the United States Central Atlantic Coastal Plain. In Clayton E. Ray, edi- tor, Geology and Paleontology of the Lee Creek Mine, North Caro- lina, I. Smithsonian Contributions to Paleobiology, 53:35-80, 37 figures. 1987. Miocene and Pliocene Pectinidae (Bivalvia) from the Lee Creek Mine and Adjacent Areas. In Clayton E. Ray, editor, Geology and Paleontology of the Lee Creek Mine, North Carolina, II. Smithson- ian Contributions to Paleobiology, 61:31-112,25 figures, 31 plates, 19 tables. Gilbert, C.R. 1967. A Revision of the Hammerhead Sharks (Family Sphyrnidae). Pro- ceedings of the United States National Museum, 119(3539): 1-88, 22 figures, 10 plates. Gillette, D.D. 1984. A Marine Ichthyofauna from the Miocene of Panama, and the Ter- tiary Caribbean Faunal Province. Journal of Vertebrate Paleontol- ogy, 4(2): 172-186. Glikman, L.S. 1964. (Paleogene Sharks and Their Stratigraphic Significance). 266 pages. [In Russian; monographic collections of the USSR Academy of Sci- ences, Division of Earth Sciences, Moscow-Leningrad.] Gottfried, M.D., L.J.V. Compagno, and S.C Bowman 1996. Size and Skeletal Anatomy of the Giant "Megatooth" Shark Car- charodon megalodon. In A.P. Klimley and D.G. Ainley, editors, The Biology of the White Shark, Carcharodon carcharias, pages 55-66. San Diego: Academic Press. Griffith, E., and CH. Smith 1834. The Class Pisces, Arranged by the Baron Cuvier, with Supplemen- tary Additions by Edward Griffith, F.R.S., &c. and Lieut-Col. Charles Hamilton Smith, F.R., L.S.S. &c. In Baron Cuvier, The Ani- mal Kingdom...with Additional Descriptions of...Species...by E. Griffith...andOthers, 10: 680 pages. London. Grimes, C.B., K.W. Albe, and R.S. Jones 1986. Tilefish, Lopholatilus chamaeleonticeps. Habitat, Behavior, and Community Structure in Mid-Atlantic and Southern New England Waters. Environmental Biology of Fishes, 15(4):273-292. Grosslein, M.D., and T.R. Azarovitz 1982. Fish Distribution. MESA New York Bight Atlas Monograph, 15: 1-182. New York Sea Grant Institute. Gruber, S.H., and L.J.V. Compagno 1981. Taxonomic Status and Biology of the Big Eye Thresher, Alopias su- perciliosus. Fishery Bulletin, 79(4):617-640, 17 figures. Gudger, E.W. 1918. Oral Gestation in the Gafftopsail Catfish Felichthys felis. Papers from the Department of Marine Biology of the Carnegie Institution of Washington, 12(252):25-52. 1933. Abnormal Dentition in Rays, Batoidei. Journal of the Elisha Mitchell Scientific Society, 49(l):57-96, 21 figures, plate 2. 1937. Abnormal Dentition in Sharks, Selachii. Bulletin of the American Museum of Natural History, 73(2):249-280, 21 figures. Giinther, A. 1870. Catalogue of the Fishes of the British Museum. Volume 8, 543 pages. London: British Museum. Guzman, L., and I. Campodonico 1976. Notorynchus cepedianus (Peron, 1807) en la entrada oriental del Es- trecho de Magallanes (Elasmobranchii, Hexanchidae). Anales del Instituto de la Patagonia, 7:207-210, 1 figure, 1 table. 198 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Hatai, K, K. Masuda, and H. Noda 1974. Marine Fossils from the Moniwa Formation, Distributed along the Natori River, Sendai, Northeast Honshu, Japan, Part 3: Shark Teeth from the Moniwa Formation. Research Bulletin, Saito Ho-On Kai Museum, 43:9-26, plate 2. Hazel, J.E. 1983. Age and Correlation of the Yorktown (Pliocene) and Croatan (Pliocene and Pleistocene) Formations at the Lee Creek Mine. In Clayton E. Ray, editor, Geology and Paleobiology of the Lee Creek Mine, North Carolina, 1. Smithsonian Contributions to Paleobiol- ogy, 53:81-199, 38 plates, 4 figures. Herman, J., M. Hovestadt-Euler, and D.C Hovestadt 1987. Contribution to the Study of the Comparative Morphology of Teeth and Other Relevant Ichthyodorulites in Living Supraspecific Taxa of Chondrichthyan Fishes, Part A: Selachii, No. I: Order: Hexanchi- formes—Family: Hexanchidae, Commissural Teeth. Bulletin de I'lnstilut Royal des Sciences Naturelles de Belgique. Bio/ogie, 57: 43-56, 9 plates. 1988. Part A: Selachii, No. 2a: Order: Carcharhiniformes—Family: Triak- idae. Bulletin de I'lnstilut Royal des Sciences Naturelles de Bel- gique, Biologie. 58:99-126, 17 plates. 1989. Part A: Selachii, No. 3: Order Squaliformes—Families: Echinorhin- idae, Oxynotidae and Squalidae. Bulletin de I 'Institut Royal des Sci- ences Naturelles de Belgique. Biologie, 59:101-157, 19 figures, 23 plates. 1992. Order: Orectolobiformes, Families: Brachaeluridae, Ginglymosto- midae, Hemiscyllidae, Orectolobidae, Parascylliidae, Rhiniodon- tidae, Stegostomidae; Order Pristiophoriformes, Family: Pristiophoridae; Order: Squatiniformes, Family: Squatinidae. In M. Stehmann, editor. Contributions to the Comparative Morphology of Teeth and Other Relevant Ichthyodorulites in Living Supraspecific Taxa of Chondrichthyan Fishes, Part A: No. 4. Bulletin de I'Institut Royal des Sciences Naturelles de Belgique. Biologie, 62:193-254. 1993. Part A: Selachii, No lb: Order Hexanchiformes—Family: Chlamy- doselachidae; No 5: Order Heterodontiformes—Family Heterodon- tidae; No 6: Order Lamniformes—Families Cetorhinidae, Megachasmidae; Addendum 1 to No 3: Order Squaliformes; Adden- dum 1 to No 4: Order Orectolobiformes; General Glossary; Sum- mary Part A. Bulletin de I 'Institut Royal des Sciences Naturelles de Belgique, Biologie. 63:185-256, plates 1-54. Hildebrand, S.F., and W.C Schroeder 1928. Fishes of the Chesapeake Bay. Bulletin of the United Slates Bureau of Fisheries, 43(1): 388 pages. Hoffman, C.W., and L.W. Ward 1989. Upper Tertiary and Pleistocene Deposits of Northeastern North Carolina. In 28th International Geological Congress Fieldtrip Guidebook: Coastal and Marine Geology of the United States, pages 53-62, 2 figures. Washington, D.C: American Geophysical Union. Holland, BE, and CF. Yelverton 1973. Distribution and Biological Studies of Anadromous Fishes Offshore North Carolina. North Carolina Department of Natural and Eco- nomic Resources Special Science Report, 24: 132 pages. Holmes, F.S. 1859. Contributions to the Natural History of the American Devil Fish, with Descriptions of a New Genus from the Harbour of Charleston, South Carolina. Proceedings of the Elliott Society of Natural His- tory, 1:39^45, plate 3. Horwood, J.W., and Cushing, D.H. 1978. Spatial Distributions and Ecology of Pelagic Fish. In J.H. Steele, ed- itor, Spatial Pattern in Plankton Communities, pages 355-383. New York and London: Plenum Press. Hussakof, V 1908. Catalogue of Types and Figured Specimens of Fossil Vertebrates in the American Museum of Natural History, Part I: Fishes. Bulletin of the American Museum of Natural History. 25:1-103. International Commission on Zoological Nomenclature 1965. Opinion 723: Repeal of the Ruling Given in Opinion 47 Together with the Stabilisation of the Generic Names Carcharhinus Blain- ville, 1816, and Odontaspis Agassiz, 1838, in Their Accustomed Sense (Pisces). Bulletin of Zoological Nomenclature, 22(1 ):32—36. 1987. Opinion 1459: Carcharias Rafinesque, 1810 (Chondrichthyes, Lamniformes): Conserved. Bulletin of Zoological Nomenclature, 44 (3):216-217. Janvier, P., and J.-L. Welcomme 1969. Affmites et paleobiologie de 1'espece Carcharodon megalodon Ag. Squale geant des faluns de la Touraine et de 1'Anjou. Revue de la Federation Francaise des Societes des Sciences Naturelles, 8:1-6, 2 figures. Jeannet, A. 1928. Les poissons fossiles originaux conserves a ITnstitut de Geologie de 1'Universite de Neuchatel. Bulletin de la Societe Neuchateloise des Sciences Naturelles, 52:102-124. 1929. Additions et rectifications a la note intitulee: Les poissons fossiles orignaux conserves a l'lnstitut de Geologie de 1'Universite de Neu- chatel. Societe Neuchateloise des Sciences Naturelles, Bulletin, 53: 197-199. Jonet, S. 1966. Notes d'ichthyologie Miocene II: Les Carcharhinidae. Boletim do Museu e Laboratorio Mineralogico e Geologico da Faculdade de Ciencias. Universidade di Lisboa. 10:65-88. Jordan, D.S., and H. Hannibal 1923. Fossil Sharks and Rays of the Pacific Slope of North America. Bul- letin of the Southern California Academy of Sciences, 22:27-63, plates 1-9. Joseph, J., W. Klawe, and P. Murphy 1980. Tuna and Billfish—Fish Without a Country, vii+46 pages. La Jolla, California: Inter-American Tropical Tuna Commission. Kanazana, R.H. 1958. A Revision of the Eels of the Genus Conger with Descriptions of Four New Species. Proceedings of the United States National Mu- seum, 108(3400):219-267, plates I^t. Kemp, N.R. 1978. Detailed Comparisons of the Dentitions of Extant Hexanchid Sharks and Tertiary Hexanchid Teeth from South Australia (Selachii: Hex- anchidae). Memoirs of the National Museum of Victoria. 39:61-83. 1991. Chondrichthyans in the Cretaceous and Tertiary of Australia. In P. Vickers-Rich, J.M. Monaghan, R.F. Baird, and TH. Rich, Verte- brate Palaeontology of Australasia. 1:497-568, plates 1 —40. Lily- dale, Victoria: Monash University Publications. Kent, B.W. 1994. Fossil Sharks of the Chesapeake Bay Region. 146 pages. Columbia: Egan Rees and Boyer, Inc. Kent, B.W., and G.W. Powell 1999. Reconstructed Dentition of the Rare Lamnoid Shark Parotodus bcnedeni (le Hon| from the Yorktown Formation (Early Pliocene) at Lee Creek Mine, North Carolina. The Mosasaur, 6:1-10. Keyes, I.W. 1972. New Records of the Elasmobranch Carcharodon megalodon (Agas- siz) and a Review of the Genus Carcharodon in the New Zealand Fossil Record. New Zealand Journal of Geology and Geophysics, 15(2):228-242, 12 figures. 1984. New Records of Fossil Elasmobranch Genera Megascyliorhinos, Centrophorus, and Dalatius (Order Selachii) in New Zealand. New Zealand Journal of Geology and Geophysics. 27:203-216. Kimmel, P.O., and R.W. Purdy 1984. Fossil Fish of the Calvert and Eastover Formations. In L.W. Ward and K. Krafft, editors, Stratigraphy and Paleontology of the Out- cropping Tertiary Beds in the Pamunkey River Region, Central Vir- NUMBER 90 199 ginia Coastal Plain: Guidebook for the 1984 Field Trip Atlantic Coastal Plain Geological Association, October 6-7, 1984, pages 205-219. Norfolk, Virginia: Atlantic Coastal Plain Geological As- sociation. Klimley, A.P. 1985. The Areal Distribution and Autoecology of the White Shark, Car- charodon carcharias, off the West Coast of North America. Mem- oirs, Southern California Academy of Sciences, 9:15-40, 9 figures. Klimley, A.P., S.D. Anderson, P. Pyle, and R.P. Henderson 1992. Spatiotemporal Patterns of White Shark (Carcharodon carcharias) Predation at the South Farallon Islands, California. Copeia, 1992(3): 680-690, 7 figures. Kozuch, L., and C. Fitzgerald 1989. A Guide to Identifying Shark Centra from Southeastern Archaeo- logical Sites. Southeastern Archaeology, 8(2): 146-157. Kuga, N. 1985. Revision of Neogene Mackerel Shark of Genus Isurus from Japan. Memoirs of the Faculty of Science, Kyoto University, Geology and Mineralogy, 51< 1-2): 1-20, plates 1-11. Landini, W. 1976. Revisione degli «lttiodontoliti pliocenici» della collezione Lawley. Palaeontographla ltalica, 70, new series, 40:92-134, 5 plates. Larrazet, A. 1886. Des pieces de la peau de quelques selaciens fossiles. Bulletin de la Societe Geologique de France, (3)14:255-277, plates 13-16. Latham, J. 1794. An Essay on the Various Species of Sawfish. Transactions of the Linnean Society of London, 2:273-282. Lavenberg, R.J. 1991. Megamania: The Continuing Saga of Megamouth Sharks. Terra, 30(l):30-39. Lawley, R. 1876. Nuovi studi sopra ai pesci ed allri vertebrati fossili delle Colline Toscane. 122 pages, 5 plates. Florence. 1878. Confronto di denti fossili che si trovano nelle colline Toscane con la dentizione dell'Oxyrhina spallanzanii Bonp. vivente in Mediterra- neo. Atti della Societa Toscana di Scienze Natura/i Residente in Pisa, Memorie, 3:343-349. 1881. Studi comparativi sui pesci fossili coi vivenli dei generi Carcharo- don, Oxyrhina, Galeocerdo. 151 pages. Pisa. Le Hon, H. 1871. Preliminaires d 'un memoire sur les poissons tertiaires de Belgique. 15 pages. Brussels. Leidy,.J. 1856a. Description of Two Ichthyodorulites. Proceedings of the Philadel- phia Academy of Natural Sciences, 8:11-12, 1856b. Remarks on Certain Extinct Species of Fishes. Proceedings of the Philadelphia Academy of Natural Sciences, 8:301—302. 1873. Contributions to the Extinct Vertebrate Fauna of the Western Terri- tories. In Report of the United States Geological Survey of the Terri- tories, 1(1): 14—358, plates 1-37. Washington, D.C: Government Printing Office. 1876. Remarks on Fossils of the Ashley Phosphate. Proceedings of the Philadelphia Academy of Natural Sciences, 1876:86-87. 1877. Description of Vertebrate Remains, Chiefly from the Phosphate Beds of South Carolina. Journal of the Academy of Natural Sciences of Philadelphia, 2(8):209-261, plates 30-34. Leriche, M. 1905. Les poissons Eocenes de la Belgique. Memoires du Musee Royal d'Histoire Naturelle de Belgique, 3:51-228, figures 9-64, plates 4-12. 1908. Note preliminaire sur des poissons nouveaux de 1 'Oligocene beige. Bulletin de la Societe Beige de Paleonlologie et d'Hydrologie, 12: 378-384. 1909. Note preliminaire sur des poissons nouveaux de l'Oligocene beige. Bulletin de la Societe Beige de Geologie, de Paleonlologie el d'Hy- drologie, 22:378-384. 1910. Les poissons Oligocenes de la Belgique. Memoires du Musee Royal d'Histoire Naturelle de Belgique, 5:233-363, figures 65-156, plates 13-26. 1926. Les poissons Neogenes de la Belgique. Memoires du Musee Royal d'Histoire Naturelle de Belgique. 32:367-472, figures 161-228, plates 28^10. 1927. Les poissons de la Molasse suisse. Memoires de la Societe Paleon- lologique Suisse, 46-47: 120 pages, 12 figures, 14 plates. 1936a. Les poissons du Cretace et du Nummulitique de I'Aude. Bulletin de la Societe Geologique de France. 5(6)6-8:375-402, plates 25-27. 1936b. Sur I'importance des squales fossiles dans l'etablissement des syn- chronismes de formations a grandes distances et sur la repartition stratigraphique et geographique de quelques especes Tertiaires. Me- moires du Musee Royal d'Histoire Naturelle de Belgique, (2)3: 739-772, 10 maps. 1942. Contribution a I'etude des faunes ichthyologiques marines des ter- rains Tertiares de la Plaine Cotiere Atlantique et du centre des Etats Unis; le synchronisme des formations Tertiares des deux cotes de l'Atlantique. Memoires de la Societe Geologique de France, new series, 43:1-111, 8 figures, plates 1-8. Leriche, M., and J. Signeux 1957. Les poissons neogenes de la Bretagne, de 1 'Anjou et de la Touraine. Memoire de la Societe Geologique de France, 36(81): 1—64, 4 plates. Lesueur, CA. 1817. Description of Three New Species of the Genus Raja. Journal of the Philadelphia Academy of Natural Sciences. 1:41 —45. 1818. Description of Several New Species of North American Fishes. Journal of the Philadelphia Academy Natural Sciences, 1:222-235. Linnaeus, C. 1758. Systema Naturae. Tenth edition, volume I, ii + 824 pages. Holmiae (=Stockholm). 1766. Systema Naturae. Twelfth edition, volume 1, 532 pages. Holmiae (=Stockholm): Laurentii Salvii. Long, D.J., and B.M. Waggoner 1996. Evolutionary Relationships of the White Shark: A Phylogeny of Lamniform Sharks Based on Dental Morphology. In A.P. Klimley and D.C Ainley, Great White Sharki: The Biology o/Carcharodon carcharias, pages 37—47, 5 figures. San Diego: Academic Press. Lowe, C.G., B.M. Wetherbee, G.L. Crow, and A.L. Tester 1996. Ontogenetic Dietary Shifts and Feeding Behavior of the Tiger Shark, Galeocerdo cuvier, in Hawaiian Waters. Environmental Biol- ogy of Fishes, 47:203-211. Lowe, R.T. 1840. On New Species of Fishes from Madeira. Proceedings of the Zoo- logical Society of London, 8:36-39. Lynn, W.G., and A.M. Melland 1939. A Fossil Catfish (Felichthys stauroforus) from the Maryland Mi- ocene. Journal of the Washington Academy of Sciences, 29(1): 14-20. Maisey, J.G. 1986. The Upper Jurassic Hexanchoid Elasamobranch Notidanoides n.g. Neues Jahrbuch fur Geologie und Palaontologie, Abhandlungen (Stuttgart), 172(1):83-106, 9 figures. Maisey, J.G., and K.E. Wolfram 1984. "Notidanus." In N. Eldredge and S. Stanley, editors, Living Fossils, pages 170-180. New York: Springer Verlag. Manooch, C, III 1984. Fisherman !s Guide: Fishes of the Southeastern United States. 362 pages. Raleigh: North Carolina State Museum of Natural History. Martin, A.F. 1996. Systematics of the Lamnidae and the Origination Time of Carchar- odon carcharias Inferred from the Comparative Analysis of Mito- chondria! DNA Sequences. In A.P. Klimley and D.G. Ainley, Great 200 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY While Sharks: The Biology o/Carcharodon carcharias, pages 49-53. San Diego: Academic Press. Matsubara, K. 1955. Fish Morphology and Hierarchy. 1605 pages. Tokyo: Ishizaki- Shoten. McEachran, J.D., and C. Capape 1984. Myliobatidae. In P.J.P. Whitehead, M.-L. Bauchot, J.-C Hureau, J. Nielsen, and E. Tortonese, Fishes of the North-eastern Atlantic and the Mediterranean, pages 205-207. Paris: Unesco. McEachran, J.D., and J.A. Musick 1975. Distribution and Relative Abundance of Seven Species of Skates (Pisces: Rajidae) Which Occur between Nova Scotia and Cape Hat- teras. Fishery Bulletin, 73( 1): 110-136, 22 figures. Menesini, E. 1967. I pesci Miocenici delle "Arenarie di Ponsano." Atti della Societa Toscana di Scienze Natural! Residente in Pisa Memoire, series A, 74(l):l-22, plates 1-3. 1969. Ittiodontoliti Miocenici di terra d'Otranto. Palaeontographia Ital- ica, 65:1-61, plates 1-7. Mitchill, S.L. 1814. The Fishes of New York, Described and Arranged. Transactions of the Literary and Philosophical Society of New York, 1:355—492. 1818. Description of Three Species of Fish. Journal of the Philadelphia Academy of Natural Sciences, 1(2)407—412. Muizon, C, and T.J. De Vries 1985. Geology and Paleontology of the Late Cenozoic Marine Deposits in the Sacaco Area (Peru). Geologische Rundschau, 74(3):547-563, 4 figures, 2 plates. Miiller, A. 1992. Ichthyofaunen aus dem atlantischen Tertiar der USA: Systematik, palaookologie, biostratigraphie, evolution und palaobiologie. 393 pages, 17 plates. Doctoral thesis, Westfalischen Wilhelms-Univer- sitat, Munster. 1999. Ichthyofaunen aus dem atlantischen Tertiar der USA. Leipziger Ge- owissenschafien, 9/10:1-360, 17 plates. Miiller, J., and F.G.J. Henle 1837. Gatrungen der Haifische und Rochen nach einer von ihm mit Hm Henle untemomenen gemeinschaftlichen Arbeit iiber die Naturge- schichte der Knorpelfische. Bericht uber die zur Bekanntmachung geeignelen Verhandlungen der Akademie Wissenschaften Berlin, 1837:111-118. 1838. On the Generic Characters of Cartilaginous Fishes. Magazine of Natural History, 2:33-37, 88-91. 1839-1841. Systemalische Beschreibung der Plagiostomen. xxii+200 pages. Berlin. Munoz-Chapuli, R. 1984. Ethologie de la reproduction chez quelques requins de 1'atlantique nord-est. Cybium. 8(3): 1-14. Nardo, J.D. 1827. Prodromus observationum et disquisitionum ichthyologiae Adriati- cae. Oken 's Isis, 20(6)472-631. Naylor, G.J.P., and L.F. Marcus 1994. Identifying Isolated Shark Teeth of the Genus Carcharhinus to Spe- cies: Relevance for Tracking Phyletic Change Through the Fossil Record. American Museum Novitales, 3109:1-53. Nishida, K. 1990. Phylogeny of the Suborder Myliobatidoidei. Memoirs of the Faculty of Fisheries, Hokkaido University, 37(1-2): 108 pages, 73 figures. Nishimoto, H., and H. Karasawa 1991. Notes on Megascyliorhinus miocaenicus (Antunes and Jonet, 1970) (Orectolobiformes: Elasmobranchii) from the Miocene Bihoku Group, Southwest Japan. Bulletin of the Mizunami Fossil Museum, 18:107-110. Nolf, D. 1988. Dents de requins et de raies du Tertiare de la Belgique. Institut Royal des Sciences Naturelles de Belgique, pages 1-184, 59 plates. Notarbartolo-di-Sciara, G. 1987. A Revisionary Study of the Genus Mobula Rafinesque, 1810 (Chon- drichthyes: Mobulidae) with the Description of a New Species. Zoo- logical Journal of the Linnean Society, 91:1-91, 32 figures. Page, L.M., and B.M. Burr 1991. A Field Guide to the Freshwater Fishes: North America North of Mexico. 432 pages. Boston: Houghton Mifflin Co. [Peterson Field Guides.] Parker, T.J. 1887. Notes on Carcharodon rondeletii. Proceedings of the Zoological Society of London, 1887:27^10, plates 4-8. Peron, F. 1807. Voyage de decouvertes aux terres Australes: execute par ordre de Sa Majeste l'empereur et roi, sur les corvettes le Geographe, le Natural- iste, et la goelette le Casuarina, pendant les annees 1800, 1801, 1802, 1803 et 1804. Volume 1, 496 pages. Peron, F., and CA. Lesueur 1822. Description of a Squalus, of a Very Large Size, Which Was Taken on the Coast of New Jersey. Journal of the Philadelphia Academy of Natural Sciences, 2:343-352. Pfeil, F.H. 1984. Neoselachian Teeth Collected from Phosphorite-Bearing Greensand on Chatham Rise East of New Zealand. Geologisches Jahrbuch, se- ries D (Mineralogie, Petrologte, Geochemie), Lagerst Attenkunde, 65:107-115. Philippi, R.A. 1846. Tornatella abbreviata, Otodus milis, Otodus catlicus, und Mylio- bates testae. Palaeontographica, 1:23-25, plate 2. Pietschmann, V. 1928. Neuen fisch-arten aus dem Pacificischen Ozean. Anzeiger der Akad- emie der Wissenshaften in Wien. 65:297—298. Poey, F. 1866. Revista de los tipos Cuvierianos y Valenciennianos correspondi- entes a los peces de la isla de Cuba. Repertorio Fiscio-Natural de la Isla de Cuba, 1:308-328. 1876. Enumeracio piscium cubensium. Anales Soctedad Espahola Histo- ria Natural Madrid. 5:131-218. Popenoe, P. 1985. Cenozoic Depositional and Structural History of the North Carolina Margin from Seismic-Stratigraphic Analyses. In C.W. Poag, Geo- logic Evolution of the United Slates Atlantic Margin, pages 125-187, 43 figures. New York: Van Nostrand Reinhold Company. Poplin, C, F. Poplin, and A. de Ricqules 1976. Quelques particularites anatomiques et histologiques du rostre de 1'Espadon (Xiphias gladius L.). Compte Rendu Hebdomadaire des Seances de I'Academie des Sciences, series D (Sciences Naturelles), 282:1105-1108. Poplin, F. 1975. Restes de rostre d'Espadon trouves dans un gisement neolithique de I'etang de Leucate (Aude). Bulletin de la Societe Prehistorique Francaise, Etudes et Travaux, 72:69-70. Potthoff, T. 1975. Development and Structure of the Caudal Complex, the Vertebral Column, and the Pterygiophores in the Blackfin Tuna (Thunnus at- lanticus, Pisces, Scombridae). Bulletin of Marine Science, 25: 205-231. Probst, J. 1878. Beitrage zur Kenntniss der fossilen fische aus der Molasse von Bal- tringen. Jahreshefte des Vereinsfiir Vaterlandische Naturkunde in Wurttemberg, 34:69-103. 1879. Beitrage zur Kenntniss der fossilen fische aus der Molasse von Bal- tringen. Jahresheft des Vereinsfiir Vaterlandische Naturkunde in Wurttemberg. 35:127-191, plate 2. Purdy, R.W. 1996. Paleoecology of Fossil White Sharks. In A.P. Klimley and D.G. Ainley, editors, Great White Sharks: The Biology o/Carcharodon carcharias, pages 67-78. San Diego: Academic Press. NUMBER 90 201 In prep. The Late Oligocene Fish Fauna from Summerville, South Caro- lina. Rafinesque Schmaltz, CS. 1810. Caratteri di alcuni nuovi generi e nouve specie di Animali e Piante della Sicilia con varie osservazioni sopra I medesimi. 105 pages. Palermo. Raschi, W., J.A. Musick, and L.J.V. Compagno 1982. Hypoprion bigelowi, a Synonym of Carcharhinus signatus (Pisces: Carcharhinidae), with a Description of Ontogenetic Heterodonty in This Species and Notes on Its Natural History. Copeia, 1982(1): 102-109, 5 figures. Ray, C.E., A. Wetmore, D. Dunkle, and P. Drez 1968. Fossil Vertebrates from the Marine Pleistocene of Southern Vir- ginia. Smithsonian Miscellaneous Collections, 153(3): 25 pages. Reid, J.L., Jr. 1962. On Circulation, Phosphate-Phosphorus Content, and Zooplankton Volumes in the Upper Part of the Pacific Ocean. Limnology and Oceanography, 7:287-306, 6 figures. Reif, W.-E. 1979. Morphogenesis and Histology of Large Scales of Batoids (Elasmo- branchii). Paldontoiogische Zeilschrift, 53:26-37, 9 figures. Ridgewood, W.G. 1904. On the Cranial Osteology of the Clupeoid Fishes. Proceedings of the Zoological Society of London, 2:448-493. Riggs, S.R. 1984. Paleoceanographic Model of Neogene Phosphorite Deposition, U.S. Atlantic Continental Margin. Science, 223(4632): 123-131, 5 fig- ures. Riggs, S.R., D.W. Lewis, A.K. Scarborough, and S. Snyder 1982. Cyclic Deposition of the Upper Tertiary Phosphorites of the Aurora Area, North Carolina, and Their Possible Relationship to Global Sea Level Fluctuations. Southeastern Geology, 23:65-84, 4 figures. Risso, A. 1826. In Histoire naturelle des principales Productions de TEurope Meridionale et particulierement des celles des environs de Nice et des Alpes Maritimes, 3:97^180. Paris. Robins, C.R., and G.C Ray 1986. A Field Guide to Atlantic Coast Fishes of North America. 354 pages. Boston: Houghton Mifflin. Rogers, C, C. Roden, R. Lohoefener, K. Mullin, and W. Hoggard 1990. Behavior, Distribution and Relative Abundance of Cownose Ray Schools Rhinoptera bonasus in the Northern Gulf of Mexico. North- east Gulf Science. ll(l):69-76. Rosen, D.E., and C. Patterson 1969. The Structure and Relationships of the Paracanthopterygian Fishes. Bulletin of the American Museum of Natural History, 141(3): 361^174, 74 figures, 27 plates, 8 tables. Ross, S.W., F.C Rohde, and D.G. Linquest 1988. Endangered, Threatened, and Rare Fauna of North Carolina, Part II: A Re-evaluation of the Marine and Estuarine Fishes. Occasional Papers of the North Carolina Biological Survey, 1988-7:7-8. Roux, C, and P. Geistodoerfer 1988. Shark Teeth and Tympanic Bullae of Cetaceans: Nuclei or Manga- nese Nodules Collected in the Indian Ocean. Cybium, 12(2): 129-137, 6 figures. Riippell, E. 1835-1840. Neue Wirbellhiere, zu der Fauna von Abyssinien gehorig, entdeckt und beschrieben von Dr. Eduard Ruppel., 4 volumes. Frankfort am Main. Sadowsky, V. 1970. On the Dentition of the Sand Shark, Odontaspis taurus. from the Vi- cinity of Cananeia, Brazil. Boletim do Instituto Oceanografico da Universidade de Sao Paulo. 18( 1 ):37-44, 3 figures. Sanders, A.E. 1980. Excavation of Oligocene Marine Fossil Beds Near Charleston, South Carolina. National Geographic Society Research Reports, 12: 601-621, 8 figures. Selden, R.W. 1986. The Sea-Robin Prionotus (Teleostei: Scorpaeniformes) in the Pleis- tocene of New Jersey. Mosasaur, 3:161-167. Seret, B. 1987. Discovery of a Fauna with Procarcharodon megalodon (Agassiz, 1835) in New Caledonia (Pisces, Chondrichthyes, Lamnidae). Cy- bium, ll(4):389-394. Shattuck, G.B. 1904. Geological and Paleontological Relations, with a Review of Earlier Investigations. In Miocene; Maryland Geological Survey, civ + 543 pages, 9 plates, atlas of 126 plates. Baltimore: Johns Hopkins Uni- versity Press. Silas, E.G., and Selvaraj, G.S.D. 1985. On the Occurrence of the Rough-tail Sting-ray Dasyatis centroura (Mitchill) in Indian Waters. Indian Journal of Fisheries, 32(2): 248-255. Simmons, E.G., and J.P. Breuer 1962. A Study of Redfish, Sciaenops ocellata Linnaeus, and Black Drum, Pogonias cromis Linnaeus. Publications of the Institute of Marine Science of the University of Texas, 8:184-211 Simpfendorfer, C.A., and N.E. Milward 1993. Utilisation of a Tropical Bay as a Nursery Area by Sharks of the Families Carcharhinidae and Sphyrnidae. Environmental Biology of Fishes, 37:337-345, 5 figures, 3 tables. Sismonda, E. 1849. Descrizione dei pesci e dei Crostacei fossili nel Piemonte. Memorie della Reale Accademia della Scienze di Torino, series 2, 10:1-88. 1861. Appendice alia descrizione dei pesci e dei crostacei fossili nel Pie- monte. Memorie della Reale Accademia della Scienze di Torino, se- ries 2, 19:453-474, plate I. Siverson, M. 1989. The Dental Evolution of the Otodus-Carcharocles Lineage. Lund Publications in Geology, 74:23. Smale, M.J. 1991. Occurrence and Feeding of Three Shark Species, Carcharhinus brachyurus, C. obscurus, and Sphyrna zygaena, on the Eastern Cape Coast of South Africa. South African Journal of Marine Science, 11: 31—42, 7 figures, 5 tables. Snyder, S.W. 1988. Synthesis of Biostratigraphic and Paleoenvironmental Interpreta- tions of Miocene Sediments from the Shallow Subsurface of On- slow Bay, North Carolina Continental Shelf. Special Publication, Cushman Foundation for Foraminiferal Research, 25:179-189, 4 figures. Snyder, S.W., L.L. Mauger, and W.H. Akers 1983. Planktonic Foraminifera and Biostratigraphy of the Yorktown For- mation, Lee Creek Mine. In Clayton E. Ray, editor, Geology and Paleontology of the Lee Creek Mine, North Carolina, I. Smithsonian Contributions to Paleobiology, 53455-481, 6 figures, 4 plates. Springer, S. 1940. Three New Sharks of the Genus Sphyrna from the Pacific Coast of Tropical America. StanfordIchthyologicalBulletin, 1(5):161-I69. Springer, S., and R.A. Waller 1969. Hexanchus vitulus, a New Sixgill Shark from the Bahamas. Bulletin of Marine Science, 19(1): 159-174, 5 figures. Springer, VG. 1964. A Revision of the Carcharhinid Shark Genera Scoliodon, Loxodon, and Rhizoprionodon. Proceedings of the United States National Mu- seum. 115(3493):559-632, 14 figures. Stehmann, M. 1981. Dasyatidae. In W. Fischer, G. Bianchi, and W.B. Scott, editors, FAO Species Identification Sheets for Fishery Purposes: Eastern Central Atlantic, Fishing Areas 34, 47 (in part), 5:1-5. Stevens, J.D. 1990. Further Results from a Tagging Study of Pelagic Sharks in the North-east Atlantic. Journal of the Marine Biological Association of the United Kingdom, 70:707-720, 5 figures, 3 tables. 202 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Storms, R. 1894. Troisieme note sur les poissons du terrain rupelien. Bulletin de la Societe Beige de Geologie, de Paleonlologie et d'Hydrologie, 8: 67-82, plate 6. Swift, C, and E. Wing 1968. Fossil Bony Fishes from Florida. Plaster Jacket, 7:1-7. Taniuchi, T. 1970. Variation in the Teeth of the Sand Shark, Odontaspis taurus (Rafinesque) Taken from the East China Sea. Japanese Journal of Ichthyology. 17(l):37-44, 4 figures. Taniuchi, T., and M. Shimizu 1993. Dental Sexual Dimorphism and Food Habits in the Stingray Dasya- tis akajei from Tokyo Bay, Japan. Nippon Suisan Gakkaishi. 59(1): 53-60. Thies, D. 1987. Comments on Hexanchiform Phylogeny (Pisces Neoselachii). Zeitschrift fiur Zoologische Syslematik und Evolutionsforschung, 25: 188-204,2 figures, I table. Thomson, K.S., W.H. Weed III, A.G. Taruski, and D.E. Simanek 1978. Saltwater Fishes of Connecticut. Bulletin. Stale Geological and Nat- ural History Survey of Connecticut, 105:1-186. Tricas, T.C, and J.E. McCosker 1984. Predatory Behavior of the White Shark (Carcharodon carcharias), with Notes on Its Biology. Proceedings of the California Academy of Sciences. 43(14):221-238, 17 figures, 1 table. Tyler, J.C. 1980. Osteology, Phylogeny, and Higher Classification of the Fishes of the Order Plectognathi (Tetraodontiformes). National Oceanic and Atmospheric Administration Technical Report, National Marine Fisheries Service Circular. 434: 422 pages, 326 figures. Tyler, J.C, R.W. Purdy, and K.H. Oliver 1992. A New Species oi Sphoeroides Pufferfish (Teleostei: Tetraodon- tidae) with Extensive Hyperostosis from the Pliocene of North Carolina. Proceedings of the Biological Society of Washington, 105(3)462-482. Uyeno, T., and Y. Hasegawa 1974. A New Miocene Lamnoid Shark, Carcharodon akilaensis. from Central Japan. Bulletin of the National Science Museum. Tokyo, 17(3):257-260, 3 figures, plate I. Uyeno, T., and O. Sakamoto 1984. Lamnoid Shark Carcharodon from Miocene Beds Chichibu Basin, Saitama Prefecture, Japan. Bulletin of the Saitama Museum of Natu- ral History. 2:47-65, 10 plates. Uyeno, T., O. Sakamoto, and H. Sekine 1989. Description of an Almost Complete Tooth Set of Carcharodon meg- alodon from a Middle Miocene Bed in Saitama Prefecture, Japan. Bulletin of the Saitama Museum of Natural History. 7:73-85, 6 fig- ures, 18 plates. Valenciennes, A. 1837. In G. Cuvier, Le Regne Animal distribue d'apres son organisation, pour servir de base a I 'historie naturelle des animaux, el d 'intro- duction a I'Analomie Comparee, third edition; Les poissons. 2 vol- umes. Paris. Van Beneden, P.J. 1883. Sur quelques formes nouvelles des terrains tertiares du pays. Bulle- tin de I'Academie Royale des Sciences, des Lei Ires el des Beaux- Arts de Belgique, series 3, 6:132-133. Walbaum, J.J. 1792. In P. Artedi, P. Artedi renovati...Ichthyologica. part 3; Pelri Artedi Seuci Genera piscium: in quibus systema tolum ichthyologiae pro- ponilur cum classibus, ordinibus, generum characteribus, specierum differentiis, observationibus plurimis: redactis speciebus 242 ad Genera 52, 1-723, plates 1-3. Ward, D.J. 1979. Additions to the Fish Fauna of the English Palaeogene, 3: A Review of the Hexanchid Sharks with a Description of Four New Species. Tertiary Research. 2(3): 111-129, 2 figures, 3 plates. Ward, D.J., and D. Thies 1987. Hexanchid Shark Teeth (Neoselachii, Vertebrata) from the Lower Cretaceous of Germany and England. Mesozoic Research, 1(2): 89-106, 2 plates. Ward, L.W., R.H. Bailey, and J.G. Carter 1991. Pliocene and Early Pleistocene Stratigraphy, Depositional History, and Molluscan Paleobiogeography of the Coastal Plain. In J.W. Horton, Jr., and VA. Zullo, editors, The Geology of the Carolinas. pages 274—289. Knoxville: The University of Tennessee Press. Ward, L.W., and B.W. Blackwelder 1980. Stratigraphic Revision of Upper Miocene and Lower Pliocene Beds of the Chesapeake Group, Middle Atlantic Coastal Plain. Bulletin of the United States Geological Survey. 1482-D: 61 pages, 25 figures, 5 plates. Weems, R.E. 1985. Miocene and Pliocene Molidae (Ranzania, Mola) from Maryland, Virginia, and North Carolina (Pisces: Tetraodontiformes). Proceed- ings of the Biological Society of Washington. 98(2)422^138. Weiler, W. 1973. Durch hyperostose verdickte fischknoch aus dem oberen Sarmat von Nord-Carolina, U.S.A. Senckenbergiana Lethaea. 53(6): 469-477. Welton, B.J , and W.J. Zinsmeister 1980. Eocene Neoselachians from the La Mesata Formation, Seymour Is- land, Antarctica Peninsula. Contributions in Science, Los Angeles, 329:1-10, 4 figures. White, E.I. 1931. The Vertebrate Faunas of the English Eocene, Volume I: From the Thanet Sands to the Basement Bed of the London Clay. 121 pages. London: British Museum (Natural History). 1955. On Lamna eurybathrodon Blake. Annals and Magazine of Natural History. 12(8): 191-193, figures 1-14. Whitley, G.P. 1940. The Fishes of Australia. Part I: The Sharks. Rays. Devilfish, and Other Primitive Fishes of Australia and New Zealand. 280 pages. Sydney: Royal Zoological Society of New South Wales. Wilson, A. 1811. [Article on Clupea] The Cyclopedic: Or Universal Dictionary of Arts, Sciences, and Letters, 9: pages unnumbered. Winkler, T.C. 1874. Deuxieme memoire sur les Dents de Poissons fossiles du terrain Bruxellien. Extrait des Archives du Musee Teyler, 4( 1): 16-48. Zittel, KA. 1887-1890. Handbuch der Palaeontologie: Palaeozoologie, Band III, Vertebrata (Pisces, Amphibia. Reptilia. Aves). 900 pages, 717 fig- ures. Miinchen und Leipzig: Druck und Verlag Von R. Oldenbourg. Turtles of the Lee Creek Mine (Pliocene: North Carolina) George R. Zug ABSTRACT Eleven taxa of turtles have been recovered from the Lee Creek Mine: a sideneck turtle (Bothremys); six seaturtles (Caretta, IChe- lonia, Lepidochelys, Procolpochelys, Syllomus, and Psephopho- rus); two pond turtles (probably Pseudemys and Trachemys); a softshell turtle (trionychid); and a giant tortoise (Geochelone). The fossils are largely disassociated skeletal elements and fragments derived from spoil piles created by drag-line mining of phosphate. The mining removes and discards the Yorktown Formation (Pliocene) and processes much of the Pungo River Formation (middle Miocene), hence the Lee Creek Mine turtles are mainly from the lower Pliocene. The turtle fauna appears to be a natural assemblage of extant and extinct taxa. Caretta and Syllomus are the most abundant fossils; a few specimens of each had some adherent Yorktown matrix. Geochelone fossils are next in abun- dance, although an order of magnitude less than Caretta and Syllo- mus. The other genera are each represented by fewer than 10 fragments or elements. Cranial and carapacial differences indicate that the Lee Creek Caretta represents a new species, C. patriciae. The Geochelone also differs from its eastern North American Pliocene contemporaries by its larger size and unique plastral mor- phology. The fossils of the other taxa are too few and fragmentary to identify reliably to species or genus. Introduction The middle Miocene to early Pliocene faunas of the central Atlantic coast and coastal plain of North America included a variety of marine, freshwater, and terrestrial turtles. Marine species dominated the turtle fauna, and at least one species each of the sideneck turtle Bothremys; the three hard-shelled seaturtles Chelonia, Procolpochelys, and Syllomus; and the leatherback seaturtle Psephophorus have been reported. Other Miocene turtles from this region included a terrestrial tortoise, Geochelone, and a softshell turtle (freshwater trionychid). Ad- ditional hard-shelled seaturtles (Table 1) have been described George R. Zug, National Museum of Natural History, Smithsonian In- stitution, Washington, D.C. 20560-0162. from the central Atlantic and adjacent regions, but close exam- ination (Weems, 1974) of these fossils has shown these taxa to be synonyms of Syllomus aegyptiacus (Lydekker). Representa- tives of these seven genera of turtle occur in the Miocene ma- rine deposits of New Jersey, Maryland, and Virginia (Table 3). These turtles and the extant taxa of southeastern North Ameri- ca provided the comparative base for the identification and analysis of the temporally and geographically close fossil tur- tles of the Lee Creek Mine. The Lee Creek Mine turtles appear to derive primarily from the Yorktown Formation and, thus, are a more recent fauna than the turtles from the Calvert Formation of Virginia and Maryland. The mining operation, however, penetrates and dis- cards the top of the Pungo River Formation (temporally equiv- alent to the Calvert Formation), so there is a possibility that a few Calvert-aged (middle Miocene) turtles are mixed in with this predominantly (early to middle Pliocene) Yorktown fauna (see Gibson, 1983, for age and stratigraphy of mid-Atlantic coastal deposits). The mining operation scatters the fossils from the numerous beds of the Yorktown Formation, resulting in fewer associations of skeletal elements with one another or with their stratum of origin in the Lee Creek Mine fauna as compared to the Calvert fauna. This lack of positive association is unfortunate because the Lee Creek seaturtle fauna is diverse and straddles a faunal transition between a middle Tertiary and the Holocene fauna. My primary objective has been to identify the Lee Creek Mine turtles and briefly describe their fossil remains. This task has forced me to make taxonomic decisions on isolated bony elements, and in some instances the amount of comparative material has been limited. These necessarily tenuous decisions must be and can be confirmed only with less fragmented and better associated fossils from Yorktown deposits. Acknowledgments.—All fossil specimens described here- in are in the vertebrate paleontological collection of the Nation- al Museum of Natural History (NMNH, which houses collec- tions of the former United States National Museum (USNM)). Some of the USNM catalog numbers cited herein represent lots rather than individuals due to the quantity of disassociated ele- 203 204 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Table 1.—Miocene turtles described as new species from the central Atlantic coast of North America. Taxon Citation Type locality Current name Suborder Pleurodira Family PELOMEDUSIDAE Taphrosphys miocenica Collins and Lynn, 1936:155 Calvert County, Maryland (=Bothremys miocenica) Suborder Cryptodira Family CHELONIIDAE Chelonia grandaeva Leidy, 1851:329 Salem County, New Jersey (=Procolpochelys grandaeva) Chelonia marylandica Collins and Lynn, 1936:162 Calvert County, Maryland (=Syllomus aegyptiacus) Peritresius virginianus Berry and Lynn, 1936:176 Westmoreland County, Virginia (=Syllomus aegyptiacus) Syllomus crispalus Cope, 1896:139 Pamunky River, Virginia (=Syllomus aegyptiacus) Family DERM0CHELY1DAE Psephophorus calvertensis Palmer, 1909:370 Calvert County, Maryland Family TESTUDrNiDAE Testudo ducateli Collins and Lynn, 1936:166 Calvert County, Maryland (^Geochelone ducateli) Family Trionychidae Trionyx cellulosus Cope, 1868:142 Charles County, Maryland ments received from the mine. My use of this collection was aided by Robert Purdy and Clayton Ray (both NMNH). Glad- wyn Sullivan (NMNH) prepared and assembled the better-asso- ciated fragments. Victor E. Krantz (NMNH) photographed all of the specimens. A number of individuals assisted in this study: Robert Weems (U.S. Geological Survey), Eugene Gaffney (American Museum of Natural History), and Rainer Zangerl (Rockville, Indiana) broadened my narrow outlook on this fossil assem- blage, and I hope improved my interpretation of it, and they re- viewed early drafts of the manuscript; more recent versions benefitted from the comments of D. Bohaska (NMNH), C. Crumly (Academic Press), K. Dodd (U.S. Geological Survey), C. Ernst (George Mason University), R. Estes (deceased), W.R. Heyer (NMNH), A. Holman (Michigan State University), and R. Weems. I thank all of the above for their assistance. Seaturtle Identification The Lee Creek Mine turtle fossils are predominantly hard- shelled seaturtle fragments, and many of these cannot be identi- fied to species or even to genus. Having worked with these fos- sils intensely in the early 1970s and then only episodically until the final preparation of this manuscript in 1988, I discovered that my ability to assign taxonomic names with confidence was directly proportional to my current immersion in seaturtle oste- ology. To assist my memory, I developed diagnoses for the main fossil skeletal elements and include them herein to assist others in the identification of seaturtle elements and fragments. These diagnoses also document my criteria for the assignment of taxonomic names to the Lee Creek Mine fossils. The diag- noses are not complete; they emphasize the type and nature of the Lee Creek Mine fossils. For example, I describe only the tip of the dentary because only that part of the lower jaw has been recovered, and in Psephophorus, only osteoderms are known. The diagnoses also tend to state differences as absolutes when some of the differences are more subtle and subjective. The major features for distinguishing the various genera of extant cheloniid seaturtles are summarized in Table 2. A partial skull of Caretta (USNM 186731; Figures 3, 4) was reassem- bled from fragments. The two critical features for identifying this skull as Caretta are the exclusion of the frontal from the orbit and the absence of vomerine-premaxillary contact on the secondary palate. The frontal also enters the orbit in Pro- colpochelys and Syllomus. Dentaries can be differentiated by the nature of the ridges on the triturating surface. Syllomus has a complex pseudodont surface (Figure 7a) with a high, denticulate symphyseal ridge extending across the entire width of the dentary; a high, sharp-edged denticulate ridge on the lingual edge; and cone-shaped denticles along the labial border (Weems, 1980, fig. 2c). The dentary surface is nearly as complex in Chelonia; a high symphyseal ridge extends across the entire width of the dentary; a high, sharp-edged ridge is slightly inset along the entire lingual border; and the labial border is sharp-edged and occasionally faintly denticulate. In Eretmochelys a low sym- physeal ridge occurs on the posterior half of the dentary's trit- urating surface and enlarges near the labial border to form a large, pyramidal protuberance; a lingual ridge is often evident, although weakly developed. The triturating surface in Caretta and Lepidochelys is a smoothly concave surface curving gently to a sharp labial edge. The labial and lingual borders are sharp-edged but low. Some Caretta have a low, sharp-edged symphyseal ridge across the entire width of the dentary. Juve- nile Lepidochelys kempii (Garman) have a low pyramidal pro- tuberance (Figure 7b) at the posterior end of the symphysis and occasionally have a faint lingual ridge; the protuberance and lingual ridge are not evident in adult L. olivacea (Eschscholtz). The dentary of Procolpochelys is unknown. Of the many carapacial fragments, it is possible to distin- guish the linked osteoderm (=epithecal ossicle) shell of dermo- chelyids from the typical testudine shell of cheloniids. The os- teoderms of Psephophorus are large, thick, irregular polygons (Figure 7c), in contrast to the small, thin, irregular polygons of Dermochelys. The osteoderms forming the dorsal ridges of the NUMBER 90 205 Table 2.—Comparison of cranial characteristics of Holocene cheloniid seaturtles and the Lee Creek Mine sea- turtle skull. Abbreviations: Cc, Caretta caretta (Linnaeus); Cm, Chelonia mydas (Linnaeus); Ei, Eretmochelys imbricata (Linnaeus); Lk, Lepidochelys kempii; Lo, Lepidochelys olivacea; LCs, Lee Creek skull; +, structure present or as described; -, absent or not as described; ±, present or absent. Cranial characteristics Cc Cm Ei Lk Lo LCs Frontal in orbit - + + + + _ Strong temporal emargination + - - + + + Supraoccipital ridge blade-like dorsally + - - ± - - Premaxillary-vomer contact on secondary palate - + + + + - Trochlear process of pterygoid elongate and thin + - - - - +? Articular surface of quadrate broad + - _ _ _ + Triturating surface of dentary strongly r dged - + + - ± - carapace in Dermochelys are as large in circumference as those of Psephophorus but are thinner. The reticulated external sur- face of the Syllomus carapace is unlike the surface texture of any other cheloniid, although the surface texture might be con- fused with that of trionychids. The shell elements are distinctly thinner (absolutely and relatively) in Syllomus than they are in any of the other Lee Creek Mine cheloniids. Neurals are found frequently. In most cheloniids, the neu- rals are elongate hexagons with the posterior segment two to three times longer than the anterior segment (i.e., cas- ket-shaped). Only in Procolpochelys are the neurals regular hexagons; the neurals also are proportionately thicker in Pro- colpochelys than they are in any of the other seaturtles except Psephophorus. The neural series in Caretta and Lepidochelys appears to be evolutionarily undergoing fragmentation and size reduction. Regular polygonal neurals lie between elongate ones in these two taxa. In addition to their unique surface tex- ture, Syllomus neurals often bear a longitudinal ridge along the entire length of each neural; the ridge ranges from a faint indi- cation to a distinctly elevated (-5 mm), sharp-edged keel. Young Caretta and Lepidochelys (of extant species) also have keeled neurals; in the small juveniles, the middorsal ridge is continuous only in the youngest individuals. The ridge has five spines or knobs extending well above the keel. These spines are most evident in small (carapace length (CL) <25 cm) Lepidochelys; in larger juveniles (CL >40 cm) only the second spine may persist, and none remains in adults. In L. kempii the spines occur at the posterior edge of each vertebral scute, hence on neurals 1, 4, 7, and 10 (neural number may differ slightly because of tendency for neural fragmentation in carettine seaturtles) and on the posterior suprapygal. The re- duction or loss of spines appears to occur from posterior to an- terior, with the second spine being the last to disappear, and the external surface of all neurals flattens with increasing cara- pace length. This external surface is planar in all size classes of Chelonia and Eretmochelys. Costal fragments are unidentifiable for most genera, al- though the surface texture of the Syllomus carapace is unique and readily identifies even small fragments. Peripherals also are difficult to assign to genus, other than those of Syllomus. In general, the larger ninth, tenth, and eleventh peripherals with distinct, serrate borders were identified as Caretta peripherals. The pygals of carettine turtles characteristically show a wide, deep, medial V-shaped notch posteriorly (this notch is small or absent in Procolpochelys) and have medially slanted peripher- al-pygal articular surfaces. The cheloniine pygal has a narrow, shallow notch posteriorly and nearly parallel peripheral-pygal articular surfaces. These differences emphasize the extremes, and pygal morphology in cheloniids forms a continuum. The most numerous limb bones are humeri, and their mor- phology appears generically diagnostic in most cases. The Syl- lomus humerus (Figure lC,E; Weems, 1974, pl. 3: figs. 1-3) is the most distinctive; however, rather than describe the entire humerus of each genus sequentially, a comparative description of each part of the humerus is offered. Humeral morphology terminology follows that proposed by Zug et al. (1986); the major difference from previous use is the recognition that the cheloniid humerus possesses both a radial (lateral) process and a deltopectoral ridge. The ulnar (medial) process of seaturtles is elongate and ex- tends proximally. This process is greatly elongated and pointed in Syllomus (Figure 1C,E) and extends proximally well beyond the humeral head (roughly the width of the head beyond); this elongation produces an attenuated appearance, although the Syl- lomus humerus is proportionately of the same width as that of the other genera. In Caretta, Lepidochelys, and Procolpochelys the ulnar process is rounded and extends only slightly beyond the head proximally. This process is intermediate in length and is somewhat acute in Chelonia and Eretmochelys. The radial (lateral) process is low and lies distal to the level of the humeral head (Figure 1). In Chelonia, Eretmochelys, and Syllomus the process forms a narrow ridge extending nearly two-thirds across the ventral surface of the shaft. This ridge is broader in Caretta, Lepidochelys, and Procolpochelys and ex- tends across no more than one-half of the shaft. The articular (cartilage-supporting) surface of the humeral head is ellipsoidal in all cheloniids (Figure 1). This surface in Syllomus is narrower and more elongate than in the other five genera and is moderately pointed at its pre- and postaxial ends. This surface is usually continuous with the radial process in Caretta, Lepidochelys, and Procolpochelys, continuous or sep- arate in Eretmochelys, usually separate in Chelonia, and al- ways separate in Syllomus. The head appears to extend farther off the diaphysis in Chelonia and Syllomus. 206 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Figure I.—Morphology of cheloniid right humeri. A, ventral view of a Holocene Chelonia mydas humerus (USNM, uncataloged); B,D, ventral and dorsal views, respectively, of a Holocene Caretta caretta humerus (USNM 235590); c,E, ventral and dorsal views, respectively, of a Syllomus aegyptiacus humerus (USNM 433179). Abbreviations: dc, deltopectoral crest or ridge; ef, ectepicondylar foramen; if, intertubercular fossa; re, radial epicondyle; rp, radial process; ue, ulnar condyle; ur, ulnar process. (Scale bar= 1 cm.) NUMBER 90 207 In Syllomus the deltopectoral crest is a cone-like tubercle projecting strongly beyond the preaxial surface. In the other genera the crest extends longitudinally along the preaxial sur- face, and it is continuous (or nearly so) with the lesser trochant- er in all cheloniids except Syllomus, where it is separated from the radial process by a deep, U-shaped groove. In Caretta and Lepidochelys the crest is truncate and moderately projecting, and it is flattened and slightly projecting in Chelonia and Eret- mochelys. The Syllomus humerus differs strikingly from that of the oth- er genera (Figure 1). Using the deltopectoral crest to divide the humerus into proximal and distal segments, these segments are subequal in length in Syllomus, and the proximal segment is one-third to three-eighths the length of the distal segment in the other five genera. The condylar surface has distinct trochlear and capitellar ridges in Syllomus; these articular ridges are low and rounded or are absent in the other genera. Ulnae are fairly abundant. The ulna of Syllomus differs from the ulnae of other cheloniids by its robustness and shape. It is proportionately shorter; has a deep, concave, proximal articular surface facing postaxially; a raised, sharp-edged postaxial crest; and a broad, radial articular surface. The cheloniid femur is similar in all of the Lee Creek Mine genera, with subtle differences that allow some differentiation of the taxa. The head is round (nearly circular in outline) in Chelonia, Eretmochelys, and Syllomus and is ellipsoidal in Caretta, Lepidochelys, and Procolpochelys. The greater (posterior) trochanter is large and angular, forming a broad sur- face anterior to the head in Chelonia, Eretmochelys, Caretta, and Lepidochelys. In Syllomus this trochanter is equally large, but the anterior border is curved and extends proximally be- yond the head; it also is hook-shaped above the intertrochanter- ic fossa (Weems, 1980, pl. 1: figs. 7, 8). The greater trochanter of Procolpochelys is narrow and straight-edged relative to the diaphysis. The lesser (anterior) trochanter is large and protrud- ing in Chelonia, Eretmochelys, Caretta, and Lepidochelys. It is about the same size in Syllomus, but the preaxial border is en- larged and rugose. It is only moderately protruding in Pro- colpochelys. The condylar surface bears distinct articular ridg- es only in Syllomus. Turtle Fauna Family PELOMEDUSIDAE (sideneck turtles) Bothremys Figure 2 Collins and Lynn (1936) described the sideneck turtle Taph- rosphys miocenica from an anterior lobe of a plastron. Later, Gaffney and Zangerl (1968) reassigned this fossil to Bothre- mys; however, they were reluctant to confirm its specific iden- tification owing to the incompleteness of the fossil. They did emphasize that this piece of plastron represented the only un- questionable sideneck turtle from the Tertiary of North Ameri- ca. Later, Gaffney (1975) noted that the type material of T. mi- ocenica was too incomplete to provide a reliable diagonsis, hence this species is a nomen dubium. Several pieces of carapace and plastron match the Bothremys material. A single hexagonal nuchal (USNM 186773; Figure 2a,b) is 59 mm long at its midline, 47 mm wide anteriorly, and 84 mm wide posteriorly. The nuchal is thin (11 mm at thickest region) and possesses smooth dorsal and ventral surfaces. The scute sutures are lightly etched on the surface. A cervical scute is absent. Sutures of the left and right first marginals, left and right first pleurals, and first vertebral scutes are present dorsal- ly on the nuchal. No scute sutures are visible ventrally. The ventral scute surface occupies the anterior third of the nuchal. The nuchal's shape and the absence of a cervical scute identify it as a pelomedusid element. Four fragments are from the plastron. Three of these (USNM 358462A (Figure 2c), 358747, 358784) are xiphiplastral frag- ments with pubic or ischial articular scars (fusion of pelvic gir- dle to plastron is characteristic of pleurodires), and the remain- ing fragment (USNM 358462B) is unidentifiable to plastron location. A small fragment of a costal (USNM 425594) has the texture of the other Lee Creek pleurodiran fragments. A complete left humerus (USNM 358316; Figure 2d,E) is as- signed to Bothremys. It is a short, robust humerus with a widely flaring greater trochanter, a squat, rugose lesser trochanter, and an ectepicondylar canal on the anterodorsal edge of the diaphy- sis (canal does not intersect the condylar articular surface). It shares some of the features of the humerus of Taphrosphys sul- catus (Leidy) (Gaffney, 1975, fig. 12C,D). Family Cheloniidae (hard-shelled seaturtles) I recognize five species of cheloniid and one species of der- mochelyid seaturtles in the Lee Creek Mine fauna. Caretta and Syllomus are represented by hundreds of elements, the other seaturtles are represented by many fewer elements. A partial skull and mandibular fragments match the mor- phology of these elements in Caretta. Numerous carapacial fragments (particularly posterior peripherals) and humeri pos- sess the carettine morphology and also are assigned to Caret- ta. These fossil elements show sufficient differences to indi- cate that they represent a species distinct from extant Caretta caretta. Caretta patriciae, new species Figures 3-6 Holotype.—USNM 186731, a partial skull lacking basioc- cipital, basisphenoid, and left quadrate-squamosal complex through and including left jugal. Collected by J.H. McLellan, 17-20 Jul 1972. TYPE Locality.—North Carolina, Beaufort County, Lee Creek Mine (35°23TSI, 76°47'30"W; United States Geological 208 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY B < '^ - Figure 2.—Fossil remains of the sideneck turtle Bothremys from Lee Creek Mine. a,b, dorsal and ventral views, respectively, of a nuchal (USNM 186773); c, dorsal view of a fragmented xiphiplastron (USNM 358462A), showing a pelvic girdle attachment scar. D,E, dorsal and ventral views, respectively, of a humerus (USNM 358316). (Scale bar=l cm.) Survey quadrangle map, 7.5-minute series, Bath, North Caro- lina, quadrangle), south side of Pamlico River, near Aurora; from a spoil pile. Horizon and Age.—Presumably from the Yorktown For- mation, lower Pliocene. Etymology.—The specific epithet is a patronym in honor of my wife, Patricia, for her years of support and love. It is pro- posed as a noun in the genitive case. Definition.—A cheloniid seaturtle with frontals excluded from orbits by prefrontal-postorbital contact, maxillary contact on secondary palate separating premaxillae from vomer, slight temporal emargination, and deep pterygoid grooves. Triturat- ing surface of dentary smoothly concave, with or without a low symphyseal ridge. Carapace morphology carettine, with strongly serrate posterior border, pygal widely and deeply notched posteriorly; neural series in adults bearing large, pro- jecting spines or knobs on neurals 1, 4, 7, 10, and posterior su- prapygal; suprapygal spine very large. Description of Holotype.—Most of the dorsal surface of the skull is present (Figures 3, 4a). The skull roof has a slight transverse arch and an equally slight longitudinal arch. Togeth- er, the parietals are trapezoidal and are 91 mm long, 50 mm NUMBER 90 209 FIGURE 3.—Cranial skeleton of Caretta patriciae, new species. A,B, dorsal and ventral views, respectively, of a skull (USNM 186731, holotype). (Scale bar=l cm.) 210 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY B /*£%=¦. , '} y ¦JfP c FIGURE 4.—Cranial skeleton oi Caretta patriciae. new species. A, lateral view of a skull (USNM 186731, holo- type); B-D, dorsal, lateral, and ventral views, respectively, of a dentary (USNM 186730). (Scale bars=l cm.) wide anteriorly, and 98 mm wide posteriorly. The frontals have a pentagonal outline, are 29 mm long medially, and are exclud- ed from the orbits by the prefrontals and the postorbitals. The nasals are truncated anteriorly and are 20 mm long medially and 42 mm long laterally. The anterior palatal region of the skull lacks only the right maxilla and left premaxilla. The triturating surface and adja- cent secondary palatal surface are flat and smooth. There is a slight depression in the premaxilla for the tip of the mandible. Posteriorly, the left side of the skull is represented by the artic- ular process of the quadrate and the adjacent part of the ptery- goid. The articular surface is ellipsoidal with no distinct medial constrictions, and the surface is inclined only slightly anterior- ly. The pterygoid groove is deep and is bordered laterally and medially by well-developed ridges. Additional Specimens.—Numerous fossil elements pos- sess characteristics of Caretta or carettine seaturtles and are re- ferred to C patriciae. The morphology of the carapace is based entirely on these isolated elements (Figures 5, 6), even though their association with the skull is uncertain. A few elements are described below. Parts of a supraoccipital and right opisthotic are present. In addition to unidentified skull fragments, pieces of a right jugal and right quadratojugal are recognizable. The skull fragments NUMBER 90 211 rfi 1S/9 FIGURE 7.—Fossil remains of Lee Creek Mine seaturtles. A,B, dorsal views oi Syllomus aegyptiacus (USNM 427790) and Lepidochelys (USNM 425612) dentaries, respectively; c, dorsal view of three osteoderms (USNM 214649) from the carapace of a Psephophorus; D,E, pygals (USNM 358461, 358457, respectively) oi Syllomus aegyptiacus. (Scale bar= 1 cm.) Procolpochelys A piece of a left hyoplastron (USNM 214648; principally the medial and posterior portion) possesses a strong xiphiplastral notch, which is covered ventrally by a bony shelf and projects from the midline at about a 30° angle. The shape and depth of this notch matches well plastral fragments of juvenile Pro- colpochelys from the Calvert fauna. Several neurals match Cal- vert Procolpochelys neurals in shape and thickness. Syllomus aegyptiacus Figure 7a,d,e Syllomus is represented by numerous fragments. The dis- tinctive surface texture of its carapace allows even the most fragmentary carapacial elements to be recognized. Neurals are numerous; most have distinct longitudinal keels. Eight pygals (two figured; Figure 7d,e) have been found, and all but the largest one bear an attenuate tip with a distal bifurcation. Hu- meri are extremely abundant, with 44 left and 56 right humeri recognized from either entire elements or proximal halves. One humerus (USNM 187122) contained a small amount of matrix, and sedimentological analysis of this matrix indicates that the humerus was derived from the Yorktown Formation. Thus, Syllomus survived into the Pliocene and probably was a contemporary of Caretta patriciae. Only six femora have been found. Family Dermochelyidae (leatherback seaturtles) Psephophorus Figure 7c Three articulated carapacial osteoderms (USNM 214649; Figure 7c) represent this genus. They are 10 mm thick, and the largest plate has a maximum length of 39 mm. Each of the three plates is of a different size and shape. There are eight other isolated osteoderms, and most of these are derived from a keeled area of the carapace. The three articulated osteo- derms are darker and more mineralized than are most Lee Creek Mine fossils and may derive from the Pungo River For- mation. Family Emydidae (pond turtles) Chrysemys complex Figure 8 A few emydid shell fragments represent the Chrysemys com- plex. Over the past three decades, the contents of this complex have been variously considered to be members of one genus or of two or three genera; the number of species has remained es- sentially static. Herein, I follow the three-genera concept 214 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Figure 8.—Carapacial and plastral elements from tur- tles of the Chrysemys complex. A, ventral view of a Pseudemys left hyoplastron (USNM 187104); B,C, dor- sal views of a Trachemys posterior peripheral (USNM 358314) and pygal (USNM 186774), respectively. (Scale bar=l cm.) (Ward, 1984; Seidel and Smith, 1986): Chrysemys, monotypic with picta; Pseudemys, containing thefloridana and rubriven- tris species groups; and Trachemys, containing the scripta spe- cies group. The Lee Creek Mine emydine fragments appear to derive from both Pseudemys and Trachemys. An emydine plastron is represented by a left hyoplastron (USNM 187104; Figure 8a). This hyoplastron lacks most of the bridge buttress, but the hy- poplastral and hypoxiphiplastral sutures and the position of the abdominofemoral sulcus are distinct. The element is 39 mm long and 33 mm wide, approximately the size of a hyoplastron of an adult Chrysemys picta; however, its morphology is more similar to that of a juvenile Pseudemys. A piece of left hyo- plastron (USNM 358315A) and a smooth-edged ninth or tenth right peripheral (USNM 358315B) also appear to be derived from a Pseudemys. Two pygals (USNM 186774 (Figure 8c), 359009) and a tenth or eleventh left peripheral (USNM 358314; Figure 8b) bear deep notches that have the serrated border of a Trachemys carapace. Family Testudinidae (tortoises) Geochelone A giant tortoise is represented by a complete shell (USNM 336458; Figure 9) and miscellaneous shell fragments, princi- pally peripherals. The complete shell has an estimated carapace length (CL) of 88 cm and an estimated plastron length (PL) of 70 cm. The costals and neurals have collapsed into the body cavity but retain their alignment. The shell was high-domed, with a smooth surface and distinct but lightly incised scute bor- ders. In outline, the shell is slightly obovate, wider posteriorly than anteriorly; the peripherals possess a slight lateral flare. The plastron is smaller than the shell opening and bears a well- developed epiplastral lip, which extends beyond the anterior margin of the carapace. The plastral surface is very lightly etched with scute outlines. Three lineages of tortoises are known from the late Tertiary of eastern North America: Geochelone (Caudochelys), Geo- chelone (Hesperotestudo), and Gopherus. The Lee Creek Mine tortoise is large, has a narrow nuchal scute, parallel-sided cos- tals, and a plastron smaller than the carapace opening, traits that ally it to Geochelone. Two species of Geochelone, G. ducateli (Collins and Lynn) (Calvert Formation, Maryland) and G. tedwhitei (Williams) (Hawthorne Formation, Florida), occur in Miocene faunas (Auffenberg, 1974). Both of these species are moderate-sized tortoises of less than 40 cm PL and are considered to be members of the subgenus Caudochelys. Larger tortoises of the subgenus Hesperotestudo have mem- bers in midcontinental Miocene faunas but do not appear in At- lantic coast faunas until the Pliocene and then only in the Southeast. The Lee Creek Mine tortoise fossils do not closely match either of these two species groups. Although some of the isolated peripherals fall within the size range of G. ducate- li, the Lee Creek Mine peripherals are proportionately thinner. The epiplastral lip of the shell also is more angular and project- ing than that of G. ducateli. As in many Hesperotestudo, the fossil's humeropectoral scute border lies on the hyoplastron immediately posteriorad to the entoplastron; however, the pec- toroabdominal border is widely separated from the humeropec- toral border (pectoral/abdominal midline lengths, 0.30%). This feature distinguishes the Lee Creek Mine tortoise from the Hesperotestudo lineage, and this separation also is greater than in Caudochelys. Of the eastern Pliocene Geochelone, only G. (Caudochelys) hayi (Sellards) is a large species, encompassing the size of the Lee Creek Mine tortoise. Both G. (Hesperotestudo) alleni Auffenberg and G (H) turgida (Cope) have plastron lengths of less than 25 cm and large plastra filling their shell openings. The Lee Creek Mine tortoise may be G. hayi; however, without additional comparative material, such an identification is tenta- tive. The type of G. hayi (USNM 8815) has a carapace of equivalent size but has a proportionately larger plastron with broader epiplastra and a deeper xiphiplastral notch. The type NUMBER 90 215 Figure 9.—Ventral view of the complete shell (USNM 336458) of the Lee Creek Mine Geochelone. A, shell resting in a plaster jacket (ruler on right= 15 cm); B, reconstruction of the plastron proportionately matching the lengths and widths of the fossil elements. (Scale bar=15 cm.) also has strongly flaring posterior peripherals, which flared only slightly in the Lee Creek Mine specimen. Family Trionychidae (softshell turtles) Genus undetermined Three carapacial fragments are referable to trionychid turtles. All possess the strongly pitted and ridged surface texture of the trionychid shell. A proximal end of a costal (USNM 186677) is extremely thick, with the thickness about 25% of the width. A neural (USNM 508057) is of equal thickness. These elements are darker and more mineralized than are most other Lee Creek Mine fossils and perhaps are derived from the Pungo River Formation. The extant North American softshell turtles represent a monophyletic group (Apalone; Meylan, 1987) of three species. Although the Lee Creek Mine fragments likely represent Apal- one, shell fragments, indeed entire carapaces or plastrons, are insufficient for the differentiation of Apalone from its Asian relatives. Family incertae sedis A nearly complete right ilium, USNM 187103, lacking the distal sacral border, is indeterminate. It is a small (-40 mm long), stout element with a straight anterior edge and a fan-shaped posterior edge. It possesses the stoutness of a Che- lydra ilium and the shape of an Emydoidea one. A heavily mineralized right parietal (USNM 187100), -45 mm long and unquestionably turtle, cannot be reliably assigned to genus. Its manner of fossilization suggests that it came from the Pungo River Formation. Discussion The Lee Creek Mine fauna has 11 recognizable turtles: a sideneck turtle, six seaturtles, two pond turtles, a softshell tur- tle, and a giant tortoise. This fauna derives principally from the Yorktown Formation, although the mining operation may have introduced some elements from the Pungo River Formation. Both of these formations are marine, yet the fauna has repre- sentatives of marine, freshwater, and terrestrial turtles. None- theless, it seems likely that both faunal components derive 216 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY from the waters and land immediately adjacent to the deposi- tional site. None, in my opinion, requires a long-distance trans- portation hypothesis. The turtles occurring in the two Miocene and one Pliocene marine deposits of the mid-Atlantic coast are summarized in Table 3. Only three species have been recovered from the New Jersey Miocene, and the "Chelonia" is almost certainly from an Eocene deposit. This small number probably reflects a lack of good collecting localities rather than a depauperate fauna. The Calvert and the Lee Creek faunas are similar in size and con- tent, sharing six taxa: Bothremys, Procolpochelys, Syllomus, Psephophorus, Geochelone, and a trionychid. The freshwater and terrestrial taxa are all extant taxa, al- though only the pond turtles occur in the Lee Creek Mine area today. The fragmentary nature of the fossil pond turtles allows only a statement of their presence in the fauna, not their specif- ic identity or ecological significance. Their presence is not un- usual because both are known from other late Tertiary and Quaternary faunas. Trachemys is both common and widespread in Cenozoic faunas east of the Rocky Mountains and has a tem- poral distribution from the early Miocene (Williams, 1953, pl. 4; Jackson, 1988) to the present in eastern North America. Wil- liams (1953) also pictured a Pseudemys floridana-\ike turtle from an early Miocene deposit in Florida. Specific identifica- tion of the Lee Creek Pseudemys and Trachemys will require more complete specimens from Lee Creek Mine and a more comprehensive examination of the fossil history of the Chryse- mys complex. The Lee Creek Mine specimens extend the geo- graphic occurrence of these taxa in the Pliocene into the mid-Atlantic region. Trionychids occurred at the Lee Creek site. Today, they are not present in that area or adjacent to the other two mid-Atlan- tic fossil sites, yet fragmentary fossils of the trionychids dem- onstrate their Pliocene or Miocene occurrence (Table 3) in the rivers of the mid-Atlantic coastal plain. Geochelone also is a common member of late Tertiary and Quaternary faunas of North America. Four of the previously known species (G. ducateli, G tedwhitei, G. alleni, G. turgida) from eastern Miocene-Pliocene faunas are much smaller tor- toises. Some Lee Creek tortoise fossils match the size of these species, but they are too fragmentary to discern whether they represent a second, smaller species in the Lee Creek Mine fau- na or represent juveniles of the giant tortoise. The complete shell and many fragments show the Lee Creek Geochelone to be a giant tortoise, the first from the mid-Atlantic Tertiary. It may be G. hayi, but comparative material is inadequate for confirmation. The pelomedusid Bothremys is considered to be a marine sideneck turtle. The presence of the xiphiplastra with girdle scars confirms its presence in the Lee Creek Mine fauna. Its oc- currence is important because it may extend the temporal range of this genus from the Calvert Formation through the York- town Formation. Without precise stratigraphic data, however, the Lee Creek sideneck turtle must be assigned questionably to the Pliocene. Whether Miocene or Pliocene, the Lee Creek Mine occurrence confirms the presence of sidenecks in the North American Tertiary. Procolpochelys and Psephophorus are very rare in the Lee Creek Mine assemblage, perhaps because they are from the Pungo River Formation; however, Psephophorus was recently discovered (Dodd and Morgan, 1992) in a Pliocene deposit in central Florida. They are assumed to be highly pelagic species. Although this pelagic behavior may account for their relative rarity, Dermochelys, the modern day counterpart of Psepho- phorus, seasonally migrates along the Atlantic coast (Shoop, Table 3.—Occurrence of Miocene and Pliocene turtles in marine deposits of the central Atlantic coastal plains of North America. Symbols: +, species occurs in fauna; -, species absent from fauna; ?, occurrence doubtful. Taxon Miocene fauna of New Jersey Calvert fauna of Maryland Lee Creek Mine fauna of and Virginia North Carolina Family Pelomedusidae Bothremys Family Cheloniidae Caretta Chelonia Lepidochelys Procolpochelys Syllomus Family Dermochelyidae Psephophorus Family Emydidae Chrysemys complex Family Testudinidae Geochelone Family Trionychidae cf. Apalone •?i 'Cope (1868) considered two fragments to represent Chelonia; Weems (1974) believed them to be from Syllo- mus and Procolpochelys, respectively. NUMBER 90 217 1987), regularly enters the larger estuaries (e.g., Chesapeake Bay; Musick, 1988), although briefly and in small numbers, and strands regularly on Atlantic beaches (Prescott, 1988). I suspect that the rarity of Psephophorus is not because they are pelagic and their carcasses were lost at sea, but because the Lee Creek depositional environment was estuarine, equivalent to today's Albemarle and Pamlico Sounds. Stranding on high en- ergy ocean-front beaches destroys carcasses and provides little opportunity for fossilization. This destruction occurs to all sea- turtles, whether they are near-shore or pelagic species. Extant Caretta caretta and Lepidochelys kempii use the estu- aries, bays, and sounds of North America (south of Cape Cod) as summer feeding grounds for juveniles and often occur in high densities in these areas. The abundance of Syllomus and Caretta patriciae suggests that the Lee Creek area was similar- ly used by these extinct species. This suggestion is further strengthened by the numerous limb bones of juvenile Caretta and Syllomus. To extend this suggestion into speculation, I note that juvenile Caretta caretta, Chelonia mydas, Eretmochelys imbricata, and Lepidochelys kempii are year-around residents in some Florida bays and sounds (Ehrhart, 1983). During the winter, they burrow into the bottom of these bays and possibly hibernate (Ogren and McVea, 1982). It seems likely that the Lee Creek Caretta and Syllomus also were year-around resi- dents of the Pliocene Lee Creek estuary. The abundant fossils of the latter two seaturtles might be attributed to cold-stunning (K. Dodd, pers. comm., 1991), a regular event in some estuar- ies (Meylan, 1986; Witherington and Ehrhart, 1989) that kills many resident seaturtles. Without stratigraphic control, suggestions on the origin of the Lee Creek vertebrate fauna are speculative. The common- ness of tortoises and juvenile seaturtles and the types of seatur- tles present argue for a shallower, near-shore deposition. The teleost fish data is less precise, indicating a "deposition at 60 to 100 m, but could in fact have been much shallower or a great deal deeper" (Fitch and Lavenberg, 1983:527). The similarity of the Caretta patriciae skull to that of C caretta indicates a similar diet, dominated by mollusks and crustaceans (Mortimer, 1982; Plotkin, 1989). The skull of Syl- lomus is more elongate (Weems, 1980) and generally resem- bles that of Eretmochelys, so it may have shared a preference for sponges (Meylan, 1988) as well. One of the more striking features of Syllomus, however, is its humeral morphology, which is unlike that of any modern seaturtle. Syllomus un- doubtedly swam with the aquatic flight locomotor pattern, but the proportional and shape differences of the humerus suggest a modification of the typical pattern, perhaps a more rapid or powerful stroke. Rather than eating sponges, was it capable of chasing and capturing fish or squid? Another peculiarity of Syl- lomus is the surface texture of the carapace, suggesting a differ- ent type of epidermal covering. Scutes were present, but they may have been softer, less keratinized, perhaps similar to the scutes of Natator depressa (Garman). The relationships of these two taxa require closer examination. The abundance of Caretta fossils in the Lee Creek Yorktown deposits and their absence from the Calvert Formation indicate a Pliocene arrival to the mid-Atlantic coast. Caretta has been reported from faunas as early as the Eocene and questionably the late Cretaceous (Mlynarski, 1976). These early fossils (Cre- taceous and Eocene) are suspect, and their identities must be confirmed. Zangerl and Turnbull (1955) placed the Miocene Pro- colpochelys grandaeva Leidy in the cheloniid tribe Carettini. They considered Procolpochelys to be a pelagic divergent and not ancestral to the extant carettines, Caretta and Lepidochelys. The presence of Caretta in the Lee Creek Mine assemblage is additional evidence against Procolpochelys as an ancestor of extant carettines. Carapace structure of these two is similar in two characteristics. Both lack surface sculpturing and possess costoperipheral fontanelles; however, the extent of fontanelle development can not be determined from the present Lee Creek Mine fragments. The fontanelles probably never closed in Pro- colpochelys. In extant Caretta, closure does occur but appar- ently only after sexual maturity. The extent of closure, its tim- ing, and intra- and interpopulational variation remain undocumented. Some of the peripherals from Lee Creek Caret- ta are equivalent in size to those of extant, reproductively ac- tive Caretta, and these peripherals lack costoperipheral sutures. Further, the shape of the largest (and clearly adult) posterior su- prapygal indicates the presence of large costoperipheral fon- tanelles in the posterior aspect of the carapace. Conclusions Examination of the turtle fossils from the Lee Creek Mine re- veals the following: (1) The Pliocene turtle fauna of the mid-Atlantic coast and coastal plain contained extinct and modern genera. The marine or estuarine taxa were Bothremys, Caretta patriciae, IChelonia, Lepidochelys, Procolpochelys, Syllomus, and Psephophorus. The freshwater taxa were two pond turtles (probably Pseudemys and Trachemys) and a tri- onychid, and there was a single terrestrial taxon, Geochelone. (2) The estuarine/near-shore nature of the Lee Creek Mine de- posit and the abundance of Caretta and Syllomus indicate that these two taxa were regular residents of the estuaries and coast of the Albemarle Embayment. Juveniles and adults are repre- sented, so the area likely included both feeding grounds and nesting beaches, just as it does for Caretta caretta today. (3) The Pliocene Caretta is morphologically distinct from the modern species and is recognized as a new species, C. patrici- ae. (4) The Lee Creek Geochelone is the earliest Cenozoic record of a giant tortoise from the mid-Atlantic coast of North America. This tortoise also appears to be morphologically dis- tinct from previously known Miocene-Pliocene Geochelone. (5) The presence of xiphiplastral fragments with pelvic girdle articular scars confirms the presence and likely extends the temporal range of pleurodiran turtles in North America. 218 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Literature Cited Auffenberg, Walter 1974. Checklist of Fossil Land Tortoises (Testudinidae). Bulletin of the Florida Stale Museum. Biological Sciences, 18(3): 121 -251. Berry, Charles T., and W. Gardner Lynn 1936. A New Turtle, Peritresius virginianus, from the Miocene of Vir- ginia. Proceedings of the American Philosophical Society, 76(2): 175-190. Collins, R. Lee, and W. Gardner Lynn 1936. Fossil Turtles from Maryland. Proceedings of the American Philo- sophical Society, 76(2): 155-173. Cope, ED. 1868. An Addition to the Vertebrate Fauna of the Miocene Period, with a Synopsis of the Extinct Cetacea of the United States. Proceedings of the Academy of Natural Sciences of Philadelphia. 19:138-156. 1896. Sixth Contribution to the Marine Fauna of North America. Proceed- ings of the American Philosophical Society, 35:139-146. Dodd, C. Kenneth, Jr., and Gary S. Morgan 1992. Fossil Sea Turtles from the Early Pliocene Bone Valley Formation, Central Florida. Journal of Herpetology. 26(1): 1-8. Ehrhart, Llewellyn M. 1983. Marine Turtles of the Indian River Lagoon System. Florida Scien- tist. 46(3/4):337-346. Fitch, John E., and Robert J. Lavenberg 1983. Teleost Fish Otoliths from Lee Creek Mine, Aurora, North Carolina (Yorktown Formation: Pliocene). In Clayton E. Ray, editor, Geol- ogy and Paleontology of the Lee Creek Mine, North Carolina, I. Smithsonian Contributions lo Paleobiology, 53:509-529. Gaffney, Eugene S. 1975. A Revision of the Side-necked Turtle Taphrosphys sulcatus (Leidy) from the Cretaceous of New Jersey. American Museum Novilates, 2571:1-24. Gaffney, Eugene S., and Rainer Zangerl 1968. A Revision of the Chelonian Genus Bothremys (Pleurodira: Pelome- dusidae). Fieldiana: Geology. 16(7): 193-239. Gibson, Thomas G. 1983. Stratigraphy of Miocene through Lower Pleistocene Strata of the United States Central Atlantic Coastal Plain. In Clayton E. Ray, edi- tor. Geology and Paleontology of the Lee Creek Mine, North Caro- lina, I. Smithsonian Contributions to Paleobiology, 53:35-80. Jackson, Dale R. 1988. A Re-examination of Fossil Turtles of the Genus Trachemys (Testudines: Emydidae). Herpetologica, 44(3):317-325. Leidy, J. 1851. Description of Fossils from the Greensand of New Jersey. Proceed- ings of the Academy of Natural Sciences of Philadelphia, 5: 329-330. Meylan, Anne Barkau 1986. Riddle of the Ridleys; Frozen and Stranded, Were These Rarest of Sea Turtles Giving a Clue to Their Secret Migrations? Natural His- tory. 95(l):90-96. 1988. Spongivory in Hawksbill Turtles: A Diet of Glass. Science, 239: 393-395. Meylan, Peter Andre 1987. The Phylogenetic Relationships of Softshelled Turtles (Family Tri- onychidae). Bulletin of the American Museum of Natural History. 186(1):1-101. Mlynarksi, Marian 1976. Testudines. Handbuch der Paldoherpelologie. 7: vi+130 pages. Stuttgart: Fischer. Mortimer, Jeanne A. 1982. Feeding Ecology of Sea Turtles. In Karen A. Bjorndal, editor, Biol- ogy and Conservation of Sea Turtles, pages 103-109. Washington, D.C: Smithsonian Institution Press. Musick, J.A. 1988. The Sea Turtles of Virginia with Notes on Identification and Natural History. Second revised edition, iv+22 pages. Glouster Point: Vir- ginia Sea Grant College Program, Virginia Institute of Marine Sci- ence. [Virginia Institute of Marine Science Educational Series, 24.] Ogren, Larry, and Charles McVea, Jr. 1982. Apparent Hibernation by Sea Turtles in North American Waters. In Karen A. Bjorndal, editor, Biology and Conservation of Sea Turtles, pages 127-132. Washington, D.C: Smithsonian Institution Press. Palmer, William 1909. Description of a New Species of Leatherback Turtle from the Mi- ocene of Maryland. Proceedings of the United States National Mu- seum. 36(1669):369-373. Plotkin, Pamela 1989. Feeding Ecology of the Loggerhead Sea Turtle in the Northwestern Gulf of Mexico. NOAA Technical Memorandum, NMFS-SEFC- 232:139-141. Prescott, Robert L. 1988. Leatherbacks in Cape Cod Bay, Massachusetts, 1977-1987. NOAA Technical Memorandum, NMFS-SEFC-214:83-84. Seidel, Michael E., and Hobart M. Smith 1986. Chrysemys, Pseudemys, Trachemys (Testudines: Emydidae): Did Agassiz Have It Right? Herpetologica. 42(2):242-248. Shoop, C. Robert 1987. Sea Turtles. In R.H. Bachus and H.H. Bourne, editors, Georges Bank, pages 357-358. Cambridge, Massachusetts: MIT Press. Ward, Joseph P. 1984. Relationships of Chrysemyd Turtles of North America. Special Pub- lications, The Museum. Texas Tech University, 21:1-50. Weems, Robert E. 1974. Middle Miocene Sea Turtles (Syllomus. Procolpochelys, Psepho- phorus) from the Calvert Formation. Journal of Paleontology, 48(2):278-303. 1980. Syllomus aegyptiacus. a Miocene Pseudodont Sea Turtle. Copeia, 1980(4):621-625. Williams, Ernest E. 1953. A New Fossil Tortoise from the Thomas Farm Miocene of Florida. Bulletin of the Museum of Comparative Zoology, 107( 11 ):537-554. Witherington, Blair E., and Llewellyn M. Ehrhart 1989. Hypothermic Stunning and Mortality of Marine Turtles in the Indian River Lagoon System, Florida. Copeia, l989(3):696-703. Zangerl, Rainer, and William D. Tumbull 1955. Procolpochelys grandaeva (Leidy), an Early Carettine Sea Turtle. Fieldiana: Zoology, 37:345-384. Zug, George R., Addison H. Wynn, and Carol Ruckdeschel 1986. Age Determination of Loggerhead Sea Turtles, Caretta caretta, by Incremental Growth Marks in the Skeleton. Smithsonian Contribu- tions to Zoology, 427: 34 pages. Thecachampsa antiqua (Leidy, 1852) (Crocodylidae: Thoracosaurinae) from Fossil Marine Deposits at Lee Creek Mine, Aurora, North Carolina, USA Albert C. Myrick, Jr. ABSTRACT Fossil remains of crocodilians at Lee Creek Mine have been few and fragmentary. Nevertheless, isolated teeth and dermal plates from the spoil piles are identifiable as Thecachampsa antiqua (Leidy). This species was established by Leidy on characters of teeth presumably from the Calvert Formation of Virginia (late early to early middle Miocene), but the species is now validated by skulls and numerous elements from the middle Miocene forma- tions of the Chesapeake Group. The same form is common in the Florida early Pliocene, where it is represented by several fairly complete skeletons known as Gavialosuchus americanus (Sell- ards). It also is known from numerous skulls from early and mid- dle Miocene formations near Lisbon, Portugal, as Tomistoma lusitanica (Vianna and Moraes). Teeth and a dermal scute indicat- ing the same species have been collected from late early Miocene deposits. The resemblance among teeth, scutes, and skulls from the Chesapeake Group, Florida, and from Portugal indicates that these forms are conspecific. The earliest available name, Theca- champsa antiqua, has priority. Introduction A small collection of isolated and fragmentary remains of fossil crocodilians from Lee Creek Mine has accumulated in the collections of the National Museum of Natural History (which includes collections of the former United States Na- tional Museum (USNM)), Smithsonian Institution, over the past 25 years. The sample is a result of collecting from mixed tailings from the Pungo River Formation (late early to early middle Miocene; Gibson, 1983) and the overlying Yorktown Albert C. Myrick, Jr., Research Associate, Vertebrate Paleontology Division, Los Angeles County Museum of Natural History, 900 Expo- sition Boulevard, Los Angeles, California 90007. Formation (early Pliocene; Gibson, 1983; Hazel, 1983). It provides a first opportunity to study the Lee Creek crocodilian materials in relation to other similar forms of comparable geo- logic age. The purpose of this paper is to describe the Lee Creek Mine crocodilian specimens and to comment on the taxonomic status and paleozoogeographic implications related to the Lee Creek Mine occurrence and to occurrences of closely related forms elsewhere. ACKNOWLEDGMENTS.—The Lee Creek Mine specimens cit- ed in this report were collected and donated by P.J. Harmatuk, F. Hyne, and B. Hyne. I thank L.G. Barnes (Los Angeles County Museum of Natural History) for providing refresher information on the California specimens. S.D. Webb's (Uni- versity of Florida) age estimations of the Florida localities were of great help. C. Repenning (United States Geological Survey) loaned crocodilian teeth from California. A. Sanders (The Charleston Museum) and B. Erickson (Science Museum of Minnesota) gave helpful advice. I am especially grateful to T. Demere, at the San Diego Museum of Natural History, who permitted me to study the Museum's holdings of recent verte- brate paleontology literature. Materials and Provenience The fossil crocodilian remains from Lee Creek Mine consist of approximately 24 isolated teeth, 15 unassociated vertebrae, two incomplete dermal scutes, three skull fragments, a man- dibular fragment, a right maxilla fragment, and several bones and bone fragments from appendicular skeletons (Figures 1-5). These have been assigned USNM catalog numbers. Field data indicate that none of the specimens were collected in association or in situ. Adherent matrix indicates at least one specimen (fragment of right maxilla, USNM 307548) came from the Pungo River Formation, but, because the spoil piles 219 220 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY consist of disturbed mixtures of sediments from the Pungo River Formation and the Yorktown Formation, others could equally well have come from the Yorktown Formation. The most that can be said at present is that the Lee Creek Mine crocodilian fossils are of late early Miocene and possibly early Pliocene age. Morphological Description The appendicular elements, vertebrae, and skull fragments in hand are of little immediate diagnostic use below the family level. If, however, one may judge from the several robust verte- brae (USNM 412243, 412254) and the large size of the posteri- or fragment of a right maxilla (USNM 307548; Figure 2), which contains complete teeth, some of these animals probably attained greater sizes than most modern crocodilians. Most of the crocodilian teeth collected at Lee Creek Mine, including those in the maxillary (USNM 307548) and mandib- ular (USNM 437930) fragments, are in a state of excellent preservation and are of diagnostic importance (Figures 1-3). They are heavily built simple cones, elliptical to subcircular in cross section, with bluntly pointed apices. A carina of variable prominence occurs on anterior and posterior surfaces of the crowns. It typically extends basally from near the apex to about one-half the length of the crown, although carinal distance also is highly variable. The carinae divide the tooth subequally, such that the greater half of the crown lies toward the labial side. The longer teeth are slightly recurved, the shorter, stubbi- er ones are barely so. Labiolingual compression is exhibited es- pecially in the shorter, presumably more posterior, teeth. The enamel is thin, variably transparent, and finely rugose, with a silky fibrous surface texture. Within the enamel can be seen various light and dark colored bands encircling the crowns hor- izontally at irregular intervals. The two specimens of dermal plates or scutes (USNM 244391, 412252) also have diagnostically important charac- ters (Figure 4). They are rectangular with one corner more or less rounded and with two edges more strongly beveled than the others. The external surface of the scute contains very large, irregularly rounded pits with no discernable pattern. In addition, the scutes have no keels (i.e., plate-like ridges that project at right angles from the pitted external surface), which commonly are found in the dermal scutes of most modern crocodilians. Comparison with Other Materials I confine my remarks to the North American Miocene and Pliocene and the Portuguese Miocene forms. In their diagnostic characters, the crocodilian teeth and scutes are identical to numerous isolated teeth and scutes from the Miocene formations of the Chesapeake Group of Mary- land and Virginia. They also closely resemble elements asso- FlGURE 1.—Teeth of Thecachampa antiqua from Lee Creek Mine, North Caro- lina (a.c, USNM 412246; b.d. USNM 299794): a.b, apical view; c.d. lingual view. (Scale bar= 1 cm.) ciated with one well-preserved skull (USNM 25243; Figure 5b) from the Calvert Formation in northern Virginia (late ear- ly Miocene) and with at least several skulls from the Bone Valley Gravel and the Alachua formations (early Pliocene; S.D. Webb, pers. comm., 1967) of Florida. In addition, they do not differ in any consistent feature from teeth and dermal scutes associated with skulls from Miocene deposits near Lis- bon, Portugal. Finally, the only crocodilian material known from late Tertiary marine deposits in southwestern California and in Baja California, one partial scute and more than a doz- en teeth, compares favorably in detail with the Lee Creek Mine materials. In cases where skulls and other skeletal elements have been closely compared and described (i.e., from the Virginia Cal- vert Formation, Maryland St. Marys Formation, Florida Pliocene formations, and Miocene formations near Lisbon, NUMBER 90 221 FIGURE 2.—Posterior fragment of right maxilla containing teeth of Thecachampsa antiqua from Lee Creek Mine, North Carolina, USNM 307548: a, dorsal view; b. lateral view; c, medial view. (Scale bar=5 cm.) 222 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE 3.—Distal ends of mandibles of Thecachampsa antiqua from Lee Creek Mine, North Carolina, USNM 437930, dorsal view. (Scale bar=5 cm.) Portugal), I have not been able to identify consistent morpho- logical differences. They all have long robust snouts that are gradually tapered from strongly built skulls (Figure 56). Among other unifying features, they share the same or similar sutural patterns, bone proportions, tooth counts, and serial tooth diameter differentiation patterns. Taxonomy Although this is not the place for a revision of the European and North American thoracosaurine crocodiles, the fossil skulls are distinctly different in generic features from the modern false gavial, Tomistoma schlegelii Miiller, of south- east Asia. Several studies (Toula and Kail, 1885; Mook, 1921, 1924; Vianna and Moraes, 1945; Auffenberg, 1954) have carefully pointed out that the European and North American "tomistomines" of Miocene-Pliocene age represent a genus separate from Tomistoma. Modern workers who have accept- ed evidence for a separate genus have used the name Gavialo- suchus to designate the more robust Miocene and Pliocene forms. The name Gavialosuchus dates from Toula and Kail (1885). Mook (1921) referred the Florida fossils to Gavialo- suchus, but Antunes (1994) continues to use the more conser- vative name Tomistoma lusitanica (Vianna and Moraes) for the Portuguese fossils. Until a partial skeleton with skull, teeth, and dermal scutes (USNM 25243; Figure 5) was collected from the Calvert For- mation in Virginia, in the late 1960s, it was not possible to show that Gavialosuchus americanus and Tomistoma lusitani- ca were junior synonyms of Leidy's species, based on the tooth characteristics, and that Gavialosuchus was a junior synonym of Cope's genus Thecachampsa, in which taxon Cope included Leidy's species antiqua in 1869. From my own study of USNM 25243 and other Chesapeake Group specimens, however, it now seems apparent that the four species of Thecachampsa (antiqua, sericodon, sicaria, and contusor) recognized by Cope on the basis of differences in tooth shape should be naturally combined under one name by virtue of the variation in teeth exhibited along the tooth rows of Thecachampsa antiqua skulls. In addition, consider- ing that there seem to be no important morphological differ- ences between skulls of different nominal species, in my opin- ion there is no useful purpose in maintaining taxonomic distinctions between specimens from the Chesapeake Group, Florida, and Lisbon, Portugal (and presumably now from Lee Creek Mine and possibly from southern and Baja California). I therefore consider these taxa to be conspecific and propose the following taxonomic scheme, which follows Steel (1973) in part. NUMBER 90 223 crocodylidae Thoracosaurinae Thecachampsa Cope, 1867 Thecachampsa antiqua (Leidy, 1852) Crocodylus antiquus Leidy, 1852 Thecachampsa sericodon Cope, 1867 Thecachampsa contusor Cope, 1867 Thecachampsa antiqua (Leidy).—Cope, 1869 Thecachampsa sicaria Cope, 1869 Tomistoma americana Sellards, 1915 Gavialosuchus americana (Sellards).—Mook, 1921 Gavialosuchus americanus (Sellards).—Auffenberg, 1954 Gavialosuchus americanus (Sellards) var. lusitanica Vianna and Moraes, 1945 Tomistoma lusitanica (Vianna and Moraes).—Antunes, 1961 Paleozoogeography of Thecachampsa antiqua The earliest known occurrences of this species in North America are from the Kirkwood Formation of New Jersey and the Calvert Formation of Maryland and Virginia (both late ear- ly to early middle Miocene, Gibson, 1983). This is probably a little later than the early Burdigalian appearance of the species in Portugal (Antunes, 1961; Benson, 1998). A fragment of a right maxilla (USNM 307548) from the Pun- go River Formation at Lee Creek Mine establishes a new southern record for eastern North America for Thecachampsa antiqua in the middle Miocene. Further, if some of the other specimens are from the Yorktown Formation at Lee Creek Mine, then these would establish the most northern extension in the early Pliocene. Thecachampsa antiqua remains have been collected from the St. Marys Formation (Chesapeake Group), but none have yet been reported in the Yorktown For- mation (early Pliocene) in Virginia. Although the species is well represented in the Bone Valley Gravel and Alachua formations in Florida (early Pliocene), it is surprising that there is no record of the species from the Haw- thorn Formation (middle Miocene, Webb, pers. comm., 1967). These results may indicate that the T. antiqua populations were vacating Maryland-Virginia waters by the beginning of York- town deposition and may not have come to inhabit coastal ar- eas of Florida in force until late in the Miocene. The sparse remains that seem to place this or a related spe- cies on the southwestern coast of North America have been collected from Barstovian Age (Calvert equivalent?) marine deposits in California and deposits in Baja California of possi- ble early Pliocene age (Barnes, pers. comm., 1988). The lack of crocodilians in otherwise abundantly rich fossiliferous marine and coastal deposits of late Miocene age is enigmatic. The problem is still under investigation. FIGURE 4.—Partial dorsal dermal scute of Thecachampsa antiqua from Lee Creek Mine, North Carolina, USNM 412252: a, dorsal view; b, ventral view. (Scale bar=l cm.) 224 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Figure 5.—Partial skeleton of Thecachampsa antiqua from the Calvert Formation of Virginia, USNM 25243: a. teeth (4) (scale bar=l cm); b, skull, dorsal view (scale bar=l m); c, dorsal dermal scutes (8), dorsal view (scale bar=3 cm). NUMBER 90 225 Literature Cited Antunes, M.T. 1961. Tomistoma lusitanica, crocodilien du Miocene du Portugal. Revista de Faculdade de Ciencias, Universidade de Lisboa, series 2, 9:5-88, 13 figures, plates 1-12, 3 tables. 1994. On Western Europe Miocene Gavials (Crocodylia); Their Paleo- geography, Migrations and Climatic Significance. Comunicacoes Instituto Geologico e Mineiro, 80:57-69. Auffenberg, W. 1954. Additional Specimens oi Gavialosuchus americanus (Sellards) from a New Locality in Florida. Quarterly Journal of the Florida Acad- emy of Sciences, 17(4): 185-209, 15 figures, 3 tables. Benson, R.N. 1998. Radiolarians and Diatoms from the Pollack Farm Site, Delaware: Marine-Terrestrial Correlation of Miocene Vertebrate Assemblages of the Middle Atlantic Coastal Plain. In R.N. Benson, editor, Geol- ogy and Paleontology of the Lower Miocene Pollack Farm Fossil Site, Delaware. Delaware Geological Survey Publication, 21:5-19. Cope, E.D. 1867. An Addition to the Vertebrate Fauna of the Miocene Period with a Synopsis of the Extinct Cetacea of the United States. Proceedings of the Academy of Natural Sciences of Philadelphia, 1867:136-156. 1869. Third Contribution to the Fauna of the Miocene Period of the United States. Proceedings of the Academy of Natural Sciences of Philadel- phia, 1869:6-12. Gibson, T.C 1983. Stratigraphy of Miocene through Lower Pleistocene Strata of the United States Central Atlantic Coastal Plain. In CE. Ray, editor, Geology and Paleontology of the Lee Creek Mine, North Carolina, I. Smithsonian Contributions to Paleobiology, 53:35-80, 37 figures, 19 tables. Hazel, J.E. 1983. Age and Correlation of the Yorktown (Pliocene) and Croatan (Pliocene and Pleistocene) Formations at the Lee Creek Mine. In CE. Ray, editor, Geology and Paleontology of the Lee Creek Mine, North Carolina, I. Smithsonian Contributions to Paleobiology, 53:81-200, 4 figures, 38 plates. Leidy, J. 1852. Description of a New Species of Crocodile from the Miocene of Virginia. Journal of the Academy of Natural Sciences ofPhiladel- phia, 2(2): 135-138, 1 plate. Mook, CC. 1921. Skull Characters and Affinities of the Extinct Florida Gavial Gavia- losuchus americanus (Sellards). Bulletin of the American Museum of Natural History, 44:39^16, 5 plates. 1924. Further Notes on the Skull Characters of Gavialosuchus americanus (Sellards). American Museum Novitates, 155:1-2, 1 figure. Sellards, E.H. 1915. A New Gavial from the Late Tertiary of Florida. American Journal of Science, 4(11):138-138, 2 figures. Steel, R. 1973. Crocodylia. In O. Kuhn, editor, Handbuch der Palaoherpetologie, 16: vii +116 pages. Stuttgart: Gustav Fischer. Toula, F., and J.A. Kail 1885. Uebereinen Krokodilschadel aus den Tertiarablagerungen von Egg- enberg, in Nieder Oesterreich. Denkschriften der Koniglichen Akad- emie der Wissenschaften, Mathematisch-naturwissenschaftlichen Klasse (Vienna), 50:299-355, 3 figures, 3 plates. Vianna, A., and A. Moraes 1945. Sur un crane de crocodile fossile decouvert dans le Miocene de Lis- bonne. Boletim da la Sociedade Geologica de Portugal, 4(3): 161-170. A New Pliocene Grebe from the Lee Creek Deposits Robert W. Storer ABSTRACT A new species of Podiceps (Aves: Podicipedidae) is described from the early Pliocene Lee Creek marine deposits in North Caro- lina. The holotype is a femur. Referred material includes entire or partial femora (7), tarsometatarsi (5), coracoid (1), humeri (7), and ulna(l). Introduction Among the thousands of bird bones found in the Neogene marine deposits at Lee Creek Mine, near Aurora, Beaufort County, North Carolina, are 22 bones or parts of bones belong- ing to a small species of grebe. Different bones representing the same element of the skeleton vary considerably in overall size and in the positions of the muscle scars on them; however, comparable differences may be found within series of a single recent species (e.g., in a series of skeletons of the Horned Grebe, Podiceps auritus (Linnaeus)), and there is no reason to believe that they represent more than a single species. Accord- ing to the characteristics listed by Murray (1967:278) for the appropriate elements, the Lee Creek Mine grebe is referable to the recent genus Podiceps. Comparisons with skeletons of liv- ing grebes in the University of Michigan Museum of Zoology (UMMZ) confirm this placement. The fossil form is about the size of P. auritus but differs in several respects that warrant de- scribing it as a new species. Acknowledgments.—I am grateful to the curators of the National Museum of Natural History (which includes collec- tions of the former United States National Museum (USNM)), Smithsonian Institution, and the University of Kansas Museum of Natural History (KUVP) for permission to borrow the fossils described herein, to L. Delle Cave (Museo Geologia e Paleon- tologia, Universita di Firenze) for providing a cast of the holo- type of Podiceps pisanus (Portis), and to the curators of the University of Michigan Museum of Paleontology (UMMP) for Robert W. Storer, Museum of Zoology and Department of Biology, University of Michigan, Ann Arbor, Michigan 48109-1079. permission to study comparative material. Clayton E. Ray, Storrs L. Olson, and T.J. Cohn offered valuable comments on the manuscript; Kama Steelquist and Jennifer Emry prepared the figure. Tom and Pat Burns, Raymond Douglas, Frank and Becky Hyne, Peter J. Harmatuk, and Clyde Swindell collected many of the Lee Creek Mine specimens used in this study. Podiceps howardae, new species Figure 1 Holotype.—Complete right femur, vertebrate paleontologi- cal collections of the National Museum of Natural History, Smithsonian Institution, USNM 252314. Paratypes.—Seven other femora (KUVP 21240, USNM 177918, 178151, 206413, 215453, 215649, 460785) probably represent this species. Their measurements are shown in Table 1. Type Locality.—Lee Creek Mine, near Aurora, Beaufort County, North Carolina (35° 18TST, 76°48'W), collected in 1977 by Peter J. Harmatuk. Horizon and Age.—Yorktown Formation, early Pliocene. Etymology.—Named in honor of Hildegarde Howard in recognition of her many important contributions to the study of fossil birds. DIAGNOSIS.—Podiceps howardae was a small grebe, ap- proximately the size of the recent P auritus, and had similar leg proportions. It differed from the recent species in confor- mation of the known skeletal elements as noted below. It was smaller than the fossil species Podiceps oligoceanus (Shufeldt), P. subparvus (L. Miller and Bowman), P. parvus (Shufeldt), and P dixi Brodkorb and was larger than P. pisa- nus, P. discors Murray, and Pliolymbus baryosteus Murray. Measurements of Holotype.—Overall length 32.5 mm, width at head 8.6 mm, width at distal end 8.7 mm, least width of shaft 3.6 mm. Description of Holotype.—The specimen is similar in size to femora of Podiceps auritus, but it is considerably nar- rower across the distal end and has a much narrower external condyle. The latter is difficult to measure, but the differences are readily seen when the bones are viewed from the anterior or distal aspects. 227 228 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE I.—Specimens oi Podiceps howardae: a. holotype femur USNM 252314; b, paratype femur USNM 177918; c, tarsometatarsus KUVP 21239; d, coracoid USNM 177927; e. humerus USNM 243764;/ distal por- tion of humerus USNM 215034. (Scale bar= 17.5 mm.) Additional Specimens.—A nearly complete tarsometatar- sus (KUVP 21239) is approximately the length of that of the largest available male of Podiceps auritus and, in general, it is similarly shaped. The foramen between the trochleae for digits three and four is longer in the fossil, and the trochlea for digit four is less offset and does not extend distally as far as that of digit three (in 27 out of 31 P auritus, the trochlea for digit four is the longer). The proximal half of a tarsometatarsus (USNM 250773) is as wide as that of a large example of P. auritus, whereas the distal portion of another (USNM 210531) is no- ticeably more slender, differing to the same degree as tar- sometatarsi of males and females of recent P auritus. Two oth- er fragmentary tarsometatarsi, a proximal portion (USNM 193175) and a distal portion (USNM 206326), are more similar in size to the smaller one. Unfortunately, the trochlea for digit four has been lost in each of the distal pieces. The nearly complete tarsometatarsus measures 48.8 mm (if it were complete, it would measure approximately 50 mm.) As- suming that this bone and the largest femur (USNM 177918, paratype) belong to the same sex of the same species, the ratio of femoral length to tarsometatarsal length would be about 0.72. The comparable ratios of four specimens each off auri- tus and P grisegena (Boddaert) are 0.73 and 0.76, respectively, whereas those of Aechmophorus occidentalis (Lawrence) and Podilymbus podiceps (Linnaeus) are 0.60 and 1.02, respective- ly. Thus, the new bird presumably had similar hind-limb pro- portions to those of P. auritus. The proximal two-thirds of a tarsometatarsus (USNM 250773) has a nearly complete articular portion and resembles that of a large male P. auritus in size and in width of the proxi- mal end. This is in contrast with the relatively narrow distal end of the femur. A nearly complete coracoid (USNM 177927) resembles co- racoids of P. auritus in size and form but has a relatively deep- er sternal facet. The bone is 30.5 mm long, 12.4 mm wide at the base, 3.0 mm in least width of shaft, and 4.0 mm in maximum depth of the external facet. In the shape of the head it differs markedly from that illustrated for the holotype of Pliodytes lanquisti Brodkorb (1953:954) of the Bone Valley Formation. A nearly complete humerus (USNM 243764) measures 75.4 mm in length and thus is within the range of measurements of females of P. auritus. Five partial humeri, consisting of a proximal portion (USNM 183430) and four distal portions (USNM 193242, 215034, 368557, 430524), also fall within the range of P. auritus. A fifth distal portion (USNM 407798) is somewhat larger and heavier (Table 2) than the extreme of P. auritus and probably is from a large male of P howardae. The shaft of the humerus is somewhat wider and flatter in the fossil than it is in the living NUMBER 90 229 form, but not enough to suggest an adaptation for using the wings under water. The distal portion of an ulna (KUVP 21292) is slightly heavier than it is in males of P. auritus and measures 5.8 mm in maximum width at the distal end, which is near the maximum for males of P. auritus. Comparisons.—According to Brodkorb (1963b:227), the earliest fossil species of Podiceps, and the earliest record of the Podicipedidae, is P. oligoceanus from the early Miocene of Or- egon. The holotype femur, as figured by Wetmore (1937:197), is considerably heavier and somewhat longer than that of P. howardae. Storrs Olson (in litt., 1986) reported that according to Jane Gray (in litt.) the provenance, and likewise the age, of the type of P. oligoceanus are in doubt. He added that "it is a typical modern grebe and could as well be Pleistocene as early Miocene." Podiceps pisanus, from the middle Pliocene of Italy, is known from the distal portion of a humerus. According to Re- galia (1902:233-234, pl. 27[I]: figs. 21, 22), P. pisanus was somewhat larger than P "auritus," and the holotype is charac- terized by the shape of the scar for the attachment of M. brachi- alis anticus near the distal end of the bone. A cast of the holo- type was compared with four humeri of P. howardae (USNM 193242, 215034, 368557, 407798) and with a series of skele- tons of recent P. auritus in the UMMZ (Table 2). In size it is at or near the lower limits of females of P. auritus (from this, it appears that Regalia's comparisons were made with the smaller species, P. nigricollis Brehm, which for many years was called "auritus"). In its shorter, wider, more transverse scar for the at- tachment of M. brachialis anticus, P. pisanus differs from P auritus (and also from P. howardae) as described by Regalia. Podiceps subparvus, described by Miller and Bowman (1958:6-7) from the middle Pliocene of San Diego, California, was somewhat larger than P. howardae and was wider across the distal end of the femur. Podiceps discors, described by Murray (1967:279-282) from the late Pliocene Rexroad Formation of Kansas, appears to have been a slightly smaller species than P. howardae. The type, a well-preserved tarsometatarsus (UMMP 52465), is smaller and more slender than the tarsometatarsi referred to P. howardae. The latter specimens also differ from P. discors and resemble P. auritus and P nigricollis in having the support of the internal condyle more flared internally. Murray (1967:281-282) referred several specimens from the Hagerman local fauna of Idaho to P. discors. One of these, a femur (UMMP 52423), is within the range of P. howardae and is proportionally too narrow at the distal end for P. auritus. It differs from the holotype of P. howardae in having facets for the insertion of M. obturator internus and M. ischiofemoralis TABLE 1.—Measurements (in mm) of femora of Podiceps howardae and P. auritus. Data for P. auritus are the ranges of 12 individuals, six of each sex, in the collection of the UMMZ. Specimen Total length Width at head Least width of shaft Width at distal end Width at distal end/total length Podiceps howardae Holotype USNM 252314 32.5 8.6 3.6 8.7 0.27 Paratypes USNM 177918 35.8 9.5± 3.5 9.5± 0.27± USNM 178151 - 9.9 3.8 - - USNM 206413 - 10.0 4.0 - - USNM 215453 32.5 8.7± 3.6 - - USNM 215649 35.0 9.0 3.3 9.0± 0.26± USNM 460785 32.4 9.9 3.8 9.3 0.29 KUVP 21240 Podiceps auritus 31.3-35.7 8.7-10.3 3.7 3.4-3.9 9.3-10.9 0.27-0.31 TABLE 2.—Measurements (in mm) of the distal portions of humeri oi Podiceps auritus (UMMZ), P. howardae, and P. pisanus (UMMP). Data for P. auritus are range and mean standard deviation for 10 individuals of each sex. Specimen Width at distal end Least width of shaft Height of shaft Podiceps auritus 7.2- -8.1 7.69±0.26 3.3- -3.95 3.62±0.19 2.75- -3.5 3.08±0.I6 Podiceps howardae USNM 193242 7.5 4.0 3.4 USNM 215034 7.5 3.8 3.15 USNM 368557 7.9 4.1 3.35 USNM 407798 8.3 4.4 3.75 USNM 243764 7.3 4.0 3.25 USNM 430524 8.0 - - Podiceps pisanus (cast of holotype) 7.2 3.75 2.95 230 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY lying more on the lateral plane of the bone than in P. howard- ae, P. auritus, or P. nigricollis, and in having a larger, deeper depression for the insertion of M. obturator externus. Two of the three coracoids assigned to P. discors by Murray (UMMP 52277', 49590) have considerably shallower external sternal facets than in P. howardae. Pliolymbus baryosteus, also described by Murray (1967:278-279) from late Pliocene deposits in Kansas, was placed in a new genus on the basis of characters in the sternum, an element that so far is unknown in Podiceps howardae. The other skeletal elements of Pliolymbus are much smaller than the corresponding ones of P. howardae. The holotype of Podiceps parvus has been discussed and fig- ured by Wetmore (1937:195-197, 200-201) and reviewed by Miller and Bowman (1958:4-5). This was a larger species than P. howardae and has been stated to range from the early mid- dle Pliocene to the middle Pleistocene. Podiceps dixi, from middle Pleistocene beds in Florida, is known from the proximal part of the carpometacarpus, an ele- ment as yet unknown for P. howardae. Measurements given in the original description (Brodkorb, 1963a:54) indicate that it was a larger bird than P auritus, and hence, than P. howardae. Steadman (1984:49), after reviewing the literature on fossil grebes, pointed out the "unsatisfactory nature" of this species and preferred to regard it "as a synonym of P auritus." Podiceps gadowi, described from Quaternary deposits on Mauritius (Hachisuka, 1953:124-125), is known from a single right ulna (205k) in the Cambridge University Museum of Zo- ology. This was examined by S.L. Olson (pers. comm.) in Au- gust 1985, who found that it had been annotated by Graham Cowles of the British Museum (Natural History) (now The Natural History Museum, London) as being from a whimbrel (Numenius phaeopus, Scolopacidae). According to Olson, the specimen measures 81.8 mm in length and is definitely not a grebe. The fossil Thiornis sociata Navas from the middle Miocene of Spain has been recognized as a grebe and was redescribed by Olson (1995:131-140), who placed it tentatively in the ge- nus Podiceps, although "in its general morphology, particularly the pelvis and hind limb, Thiornis sociata is decidedly more similar to Tachybaptus than to modern species of Podiceps." Because comparisons with skeletons of recent grebes show that P howardae clearly belongs in the genus Podiceps, compari- son with Thiornis was not attempted. Remarks.—Podiceps howardae was a small grebe; it aver- aged slightly larger in size than the recent Horned Grebe, Podi- ceps auritus, but it was similar in proportions. The known specimens all come from offshore marine deposits; however, because the floating nests of all living grebes are subject to damage or loss by wave action and fluctuations in water level, these birds do not nest near large expanses of open water or in tidal situations. Assuming that P. howardae had similar nesting requirements, it probably nested inland on fresh water and win- tered on salt water. Thus, specimens of this Pliocene species may well be expected in inland localities. Such a pattern of dis- tribution is already known for the fossil species Podiceps par- vus (Miller and Bowman, 1958:5). In spite of the similarities in size and proportions between P. howardae and P auritus, whether the former was ancestral to the latter is unclear. Fjeldsa (1983) has provided convincing evidence for character displacement in the bill length of grebes, and character displacement also is evident in the overall size of the Red-necked Grebe (Podiceps grisegena), which is consid- erably smaller in Europe, where it is sympatric with the larger Great Crested Grebe (P. cristatus (Linnaeus)), than it is in North America, where no larger congener occurs (Palmer, 1962:63-87). Even greater geographic variation in size is found in the White-tufted Grebe (Rollandia rolland (Quoy and Gaimard)), in which tarsometatarsal length of study skins var- ies from 51.7 mm in a male of the race on the Falkland Islands (Islas Malvinas) to 31.0 mm in a small female from northern Argentina (Storer, unpublished data). Smaller, but significant, differences in size in this species occur between lakes Junin and Titicaca, where R. rolland is found with different assem- blages of grebes. There is no reason to doubt that geographic variation occurred in the past as it does today, and this should be taken into consideration in any analysis of fossils (Storer, 1992:419^22). It is thus evident that size in grebes is not necessarily an indi- cation of close relationships. It is therefore to be expected that as faunas change with the disappearance of some forms and the appearance of others, shifts in the size of at least some of the species can be expected. This being the case, other characteris- tics, especially the conformation of bones, should be more use- ful in assessing relationships among closely related species of grebes. Differences in proportions can arise from more than one source. In birds using similar types of locomotion, wings must increase more rapidly than overall size in order for a bird to maintain the ability to fly. This explains most, if not all, of the differences in the ratios of humeral length to tarsometatarsal length mentioned above. Other proportional differences, such as those in the toes, are more likely to reflect differences in the way the foot is used and are presumably more significant in phylogenetic studies. Compared with other pre-Pleistocene grebes, Podiceps howardae is represented by a fair number of specimens. Until other species become known from more adequate material, little can be shown about the relationships among them and recent species. NUMBER 90 231 Literature Cited Brodkorb, P. 1953. A Pliocene Grebe from Florida. Annals and Magazine of Natural History, series 12, 6:953-954. 1963a. A New Pleistocene Grebe from Florida. Quarterly Journal of the Florida Academy of Sciences, 26( 1): 53-55. 1963b. Catalogue of Fossil Birds. Bulletin of the Florida Slate Museum, Bi- ological Sciences, 7(4): 179-293. Fjeldsa, J. 1983. Ecological Character Displacement and Character Release in Grebes Podicipedidae. Ibis, 125(4):463-481. Hachisuka, M. 1953. The Dodo and Kindred Birds, xvi + 250 pages. London: H.F. and G. Witherby. Miller, L., and R.I. Bowman 1958. Further Bird Remains from the San Diego Pliocene. Contributions in Science of the Los Angeles County Museum. 20:1-15. Murray, B.G., Jr. 1967. Grebes from the Late Pliocene of North America. Condor, 69(3):277-288. Olson, S.L. 1995. Thiornis sociata Navas, a Nearly Complete Miocene Grebe (Aves: Podicipedidae). Courier Forschungsinsul Senckenberg, 181: 131-140, 4 figures. Palmer, R.S., editor 1962. Handbook of North American Birds. Volume 1, x+567 pages. New Haven and London: Yale University Press. Regalia, E. 1902. Serte Uccelli Pliocenici. Palaeontographia Ilalica, 8(1902): 219-238, plate 27[I]. Steadman, D.W. 1984. A Middle Pleistocene (Late Irvingtonian) Avifauna from Payne Creek, Central Florida. Carnegie Museum of Natural History Spe- cial Publication, 8:47-52. Storer, R.W. 1992. Intraspecific Variation and the Identification of Pliocene and Pleis- tocene Grebes. In K.E. Campbell, Jr., editor, Papers in Avian Pale- ontology Honoring Pierce Brodkorb. Science Series, Natural History Museum of Los Angeles County, 36:419-422. Wetmore, A. 1937. A Record of the Fossil Grebe, Colymbus parvus, from the Pliocene of California, with Remarks on Other American Fossils of This Family. Proceedings of the California Academy of Sciences, series 4,23:195-201. Miocene and Pliocene Birds from the Lee Creek Mine, North Carolina Storrs L. Olson and Pamela C. Rasmussen ABSTRACT An account is given of a collection of over 10,000 fossils repre- senting at least 112 species of birds from middle Miocene (Pungo River Formation) and early Pliocene (Yorktown Formation) deposits exposed during phosphate mining in the Lee Creek Mine, near Aurora, Beaufort County, North Carolina. Relatively few of these species are derived from the Pungo River Formation, as determined partly by similarity to contempo- raneous species from the Calvert Formation of Maryland and Vir- ginia. The tremendous avifauna now known from the Yorktown Formation consists of nearly 100 species, including three species of loons, two grebes, five albatrosses, at least 16 shearwaters and petrels, one pelican, two pseudodontorns (horizon less certain), three gannets, two cormorants, at least nine auks and puffins (probably 11 or more), one skua, three jaegers, five gulls, two terns, and 20 species of ducks, geese, and swans. Incidental land and shore birds, a few of which likely originated in the Pungo River Formation, are represented by 29 species, including three cranes, one rail, two oystercatchers, one plover, four scolopacids, one flamingo, one ibis, one heron, three storks, one condor, five accipitrids, one osprey, one chachalaca, one phasianid, one turkey, one pigeon, and one crow. Three new species are described, one in each of the genera Gavia, Phoebastria, and Calonectris. In num- bers of individuals, the avifauna is dominated by a radiation of auks of the genus Alca. The avifauna of the Yorktown Formation indicates conditions of much greater marine productivity, with accompanying greater diversity of marine birds, than existed in the middle Miocene. Loss of diversity since the early Pliocene involved extinction of most of the radiation of Alca, two species of albatrosses, two species of gannets (Moms), a pelican, and possi- bly a cormorant. Diversity in the western North Atlantic Ocean also was reduced by the withdrawal of three species of albatrosses, numerous petrels, and a puffin (Fratercula) to the Pacific Ocean. Other species, mostly of land and shore birds, appear to have with- drawn from North America but have persisted in Europe and Asia. Apart from the relatively few extinctions and several range retrac- tions, the avifauna in the early Pliocene of North Carolina was very modem in aspect, and many modern species lineages of birds may already have been in existence at least five million years ago. Storrs L. Olson, National Museum of Natural History, Smithsonian In- stitution, Washington, D.C. 20560-0131. Pamela C. Rasmussen, Mich- igan State University Museum, East Lansing, Michigan 48824-1045. Introduction This paper originated as a survey of an extensive fossil avi- fauna from Miocene and Pliocene marine deposits obtained in the late 1960s and early 1970s at what was then called the Texasgulf Lee Creek Phosphate Mine in North Carolina (now owned by the Potash Corporation of Saskatchewan). Bird fos- sils from this site were originally turned over to Alexander Wetmore for study, and he was to have prepared a paper on these specimens for inclusion in a projected volume on the ge- ology and paleontology of the mine. Because of Wetmore's commitment to his magnum opus on the birds of Panama, Olson later volunteered to take on the bulk of the work associ- ated with preparing a manuscript on the Lee Creek Mine birds. Olson and Wetmore (Figure 1) produced a manuscript entitled "Preliminary Survey of an Extensive Miocene and Pliocene Marine Avifauna from Lee Creek, North Carolina," which was submitted to the volume editor on 28 September 1973 and sub- sequently was optimistically cited as "in press" (Olson, 1975, 1977; Alvarez and Olson, 1978). Years went by and the single volume that was originally en- visioned was expanded to two, then three, and ultimately four volumes. The first of these appeared a decade after the original manuscript on birds had been submitted (Ray, 1983) and was followed by a second volume four years later (Ray, 1987). Meanwhile, the quantity of avian fossils from Lee Creek Mine increased. At the same time, Wetmore's health declined, and he was no longer able to participate in the several attempts Olson made to keep the Lee Creek manuscript up to date. By the time Wetmore died on 7 December 1978, the manuscript bore little resemblance to the first draft to which he had contributed. Some of the ideas first developed by Olson in the early drafts appeared subsequently in an overall summary of the fossil record of birds (Olson, 1985d). Because so many new fossils were acquired after Olson's last lone revision of the manuscript, Jonathan Becker was en- listed to identify new specimens and worked to produce an up- dated version of the manuscript. This, too, eventually became almost completely obsolete, both apart from, and because of, 233 234 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE 1.—Olson (left) and Alexander Wetmore (right) in the early 1970s, examining the available fossil bird bones from Lee Creek Mine. Since then, the number of specimens has increased by probably more than an order of magnitude. acquisition of new material. By this time Becker had changed careers. Olson and Rasmussen began collaboration on a revi- sion of the Lee Creek Mine avifaunas in 1991, greatly expand- ing the material included and gaining entirely new insights into systematics, biogeography, and evolution. The present manu- script thus bears little resemblance to its predecessors. We must therefore acknowledge with gratitude the contributions of our previous collaborators, Wetmore and Becker, to the early stag- es of study of this massive avifauna, while relieving them of any responsibility for our conclusions. To our knowledge, the Lee Creek Mine collection constitutes what is probably the largest Tertiary marine avifauna now known, with at least 112 species of birds, most being pelagic forms, represented by over 10,000 specimens, not all of which, however, have been included in the present study. The fossils from Lee Creek Mine are from two distinct faunas, one of mid- dle Miocene age (-14 Ma) derived from the Pungo River For- mation, and a much larger early Pliocene (3.7-4.8 Ma) avifau- na derived from the Yorktown Formation. Because of the nature of the mining operations, it is often uncertain which ho- rizon produced a given fossil, and this mixing contributes to the complexity of analyzing certain groups. The Pungo River Formation is in part equivalent to the Cal- vert Formation of the Chesapeake Group of Maryland and Vir- ginia. The avifauna of the Calvert Formation is now reasonably well known from fossils, many still undescribed, found in pre- cise stratigraphic contexts. Thus, the horizon of certain Lee Creek Mine fossils may be inferred from their identity with NUMBER 90 235 species known from the Calvert deposits. We also have a few additional undescribed specimens of fossil birds from the Choptank, St. Marys, and Eastover formations in Maryland and Virginia that we propose to study by systematic groups in con- junction with fossils from the Calvert Formation before at- tempting to synthesize the East Coast marine avifaunas that preceded those of the Pliocene. For this reason, we have not dwelt at length on the avifauna of the Pungo River Formation in the present report, the species of which will be treated in more detail later in conjunction with those of the Calvert Formation. Instead, we emphasize herein the much more extensive and diverse avifauna of the Yorktown Formation, fossils of which are extremely rare in surface exposures. This very important marine avifauna would be all but unknown were it not for mining operations. The present study is preliminary, and no one is more aware of its shortcomings than the authors. We have intended this mainly as a faunal survey, but of necessity we must deal with many issues of systematics as well. Needless to say, a great deal of revisionary work is still urgently needed, particularly among such difficult groups as the shearwaters (Procellari- idae). By making these avifaunas better known, we hope to stimulate taxonomists to undertake revisions of various taxa, and we invite them to include the Lee Creek Mine material, especially that from the Yorktown Formation, in their studies. ACKNOWLEDGMENTS.—We thank the officials of Texas- gulf, Inc., and their successors, for their continuing consider- ation in permitting paleontological investigations at the Lee Creek Mine. Field parties from the Smithsonian Institution were funded by the Smithsonian through the Remington and Marguerite Kellogg Fund and the Walcott Fund. We owe our greatest debt to the many collectors whose sharp eyes spied thousands of small bird bones on the mine spoil piles and whose generosity in donating the specimens to the Smithsonian has permitted this study. In this connection the massive contributions of Peter J. Harmatuk and Frank and Becky Hyne must be singled out. The following additional collectors also donated specimens: Calvin Allison, David R. Amos, Elizabeth and Wallace L. Ashby, Donnie Bailey, Aura L. and J. Wayne Baker, Peter Ballmann, William C. Bean, N. Bikan, David and Paula Bohaska, Captain and Mrs. Michael Boroff, John Boyd, Eldon Branch, Pat and Tom Burns, Kerry Button, Ken Carpenter, John H. Carson, Richard Carter, Gerard R. Case, Ralph Chamness, Richard Chandler, Mike Cohen, Phillip Cox, Larry Decina, Daryl Domning, Raymond C Douglass, Duke University Marine Lab (through Dan Rittschof), Ralph Eshelman, John Everette, Alan Feduccia, James Firebaugh, Mark Florence, George C. Fonger, Frank Garcia, Steve Gotte, Michael D. Gottfried, Fred Grady, Richard W. Grier, Jr., Robert W. Grier, Todd Grimsley, Leslie Hale, Christopher J. Harmatuk, Eugene F. Hartstein, Barbara Harvey, Bill Heim, Anne Henderson, Wayne Henschel, Linda Heritage, Carolynne and Scott Hertenstein, Pam Hester, Mr. and Mrs. L. Hodges, Candace Holiday, Sid Hostetter, Jane Hubbard, Fran Hueber, Aileen and Stanley Hyne, Kelly Irwin, Ronald M.A. Ison, Jeremy Jacobs, Helen James, Walter Johns, Linda Johnson, Ralph Johnson, James Kaltenbach, William G. Keel, John Keshishian, Anne Leightner Kienlen, Jim Knight, Trish Kohler, Arnold Lewis, Lloyd Logan, Andreas Luettge, Royal H. Mapes, Earl Mason, Vance McCollum, Peter A. Mc- Crery, Rita McDaniel, Tom Mclntyre, Gregory S. McKee, Jack H. McLellan, Robert Metzger, Robert L. Meyer, Jon Moore, Wayne Morgan, A.C. Myrick, Jr., John and Myrna Nay, Randy Nunnery, Gene C. Oliver, John A. Onderdonk, Jr., Tom Parks, Franklin Pearce, Michael A. Pehachek, James and Susan Pendergraft, Diane Pitassy, John L. Pittman, Pamela Platt, George Potter, George W. Powell, Jr., Robert W. Purdy, Clayton E. Ray, Charles A. Repenning, William J. Rieger, Norman L. Riker, Albert J. Robb III, Betty L. Roberts, Rocky Mount Children's Museum, Annette Salensky, Mr. and Mrs. Vincent P. Schneider, Randy Scott, William Shelton, David Oscar Siegert, Edward S. Slagle, Wilton Smith, Albert and Cheryl Snelson, Scott W. Snyder, Gladwyn Sullivan, Clyde Swindell, Calvin E. Taylor, Bill and Juanda Taylor, Reginald Titmas, Jerry D. Tracy, John Waldrop, David Wells, James Westgate, Alexander Wetmore, Frank C, Jr., and Martha Whitmore, Gaye Williams, Druid Wilson, E.A. Womble, Roger Wood, Wendell P. Woodring, W.W. Zack, and George Zug. David J. Bohaska, Mark Florence, Robert W. Purdy, and Clayton E. Ray, of the Department of Paleobiology, National Museum of Natural History (NMNH), Smithsonian Institu- tion, have assisted repeatedly in many aspects of this study. We are indebted to Frank B. Gill and Earl A. Shapiro (Acade- my of Natural Sciences of Philadelphia), William N. Orr (Condon Museum of Fossils, University of Oregon), Michael Gottfried (then of the Calvert Marine Museum), and Larry D. Martin (University of Kansas, Lawrence (KU)) for making fossil material available for study. Janet Hinshaw lent recent comparative material from the University of Michigan Muse- um of Zoology, Ann Arbor (UMMZ). S. David Webb, Bruce J. MacFadden, Gary S. Morgan, and Pierce Brodkorb made material from the Bone Valley Formation, now in the Florida Museum of Natural History (UF), available for study. Tom Demere kindly lent specimens, including holotypes, from the San Diego Natural History Museum (SDSNH). Over the years we have benefited from study of both fossil and modern mate- rial from the Natural History Museum of Los Angeles County (LACM), and we thank Kenneth E. Campbell for lending fos- sils from Rancho La Brea for comparison in this study. We also had access to specimens from the Big Sandy Formation, Arizona, lent from the Frick Collection at the American Muse- um of Natural History (F:AM), New York. We are particularly grateful to Thomas G. Gibson (United States Geological Sur- vey) for analyzing matrix samples from some of the bird fos- sils. Raphael Alvarez (formerly UMMZ) provided initial iden- tifications of some of the duck fossils that were available in the 1970s. Kenneth I. Warheit (Washington Department of 236 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Fish and Wildlife, Seattle) and Helen F James (NMNH) aided in a variety of ways. Camm C. Swift (formerly of the Natural History Museum of Los Angeles County) identified fish prey of an auk. Douglas Siegel-Causey and Jiri Mlfkovsky translat- ed portions of articles in Russian. David W. Steadman provid- ed measurements of some albatross fossils at UF. The many photographs are the patient work of our dedicated associate of many years, Victor E. Krantz, of the Smithson- ian's Office of Photographic Services. Brian K. Schmidt (NMNH) assisted with revisions of some of the plates. Pierce Brodkorb, Robert W. Storer, Miklos D.F. Udvardy, John Far- rand, Jr., and Albert C. Myrick commented long ago on por- tions of an early draft of this paper. We are especially grateful to Steve Emslie (University of North Carolina, Wilmington) and Cecile Mourer-Chauvire (Universite Claude Bernard, Ly- on) for supplying numerous references and for their comments on the manuscript. Craig Ludwig (NMNH) checked the accura- cy of catalog numbers of modern comparative material in the manuscript with the specimen database in the Division of Birds, National Museum of Natural History, Smithsonian Insti- tution. Finally, with dedication and tireless attention to detail, Mark Florence (NMNH) checked each of the catalog numbers of fossils in the manuscript against the specimens themselves, correcting a number of errors. Synopsis of the Geology of the Lee Creek Mine The Lee Creek fossil locality is an open-pit mining operation for obtaining commercial quantities of phosphate from the middle Miocene marine sediments of the Pungo River Forma- tion. The mine is located on the south bank of the Pamlico Riv- er, north of the town of Aurora, Beaufort County, in central eastern North Carolina (35°23'22"N, 76°47'06"W). Mining was started at what was then the debouchment of Lee Creek into the Pamlico River (McLellan, 1983), the lower reaches of the lesser stream having since been obliterated by the mine to which it gave its name (see Ward and Blackwelder, 1987, figs. 1,2). This sacrifice of a bit of coastal topography has been more than compensated by the tremendous increase in knowledge of geology and paleontology that was made possible by the min- ing operations at Lee Creek. There are no naturally occurring outcrops of the Pungo River Formation anywhere, so the only fossil birds known from it have come from Lee Creek Mine. Practically the same may be said of the overlying Yorktown Formation, surface outcrops of which are very limited and which have yielded at most a handful of bird bones. This is in marked contrast with the thousands upon thousands of bird fos- sils of Yorktown age from Lee Creek Mine, an assemblage that constitutes the most extensive marine paleoavifauna in the world in terms of both species and numbers of specimens. Without these fossils we would scarcely have an inkling of the dramatic turnover that took place in the pelagic avifaunas of the North Atlantic Ocean between the moderately well-known avifauna of the Calvert Formation, some 14 million years ago, and the present. Mainly for the benefit of those who do not have access to the other volumes in this series, we have included herein a very brief synopsis of the geology of Lee Creek Mine, which, unless otherwise indicated, has been extracted from the lucid and in- formative papers of Gibson (1983a, 1983b). Neogene sediments at Lee Creek Mine accumulated in a neg- ative feature, the Albemarle embayment, flanked by two east-west trending positive geological structures, the Norfolk arch, running across southern Virginia, and the New Bern arch, running across central eastern North Carolina, just south of Pamlico Sound. Although the sediment column at the mine is capped unconformably by late Pliocene and Pleistocene sedi- ments (see also Ward and Blackwelder, 1987), we are unaware of any fossil birds from these strata. If these deposits ever were exposed on the surface of spoil piles at the mine, they likely would not attract much attention from the amateur collectors responsible for collecting the majority of the fossil birds at the mine, who search mainly in sediments in which large sharks' teeth are likely to be found. Thus, attention is focused herein on the other strata exposed at the mine—those of the Pungo River Formation and the over- lying Yorktown Formation. The Pungo River Formation was laid down during an extensive marine transgression responsible for the mainly contemporaneous Calvert Formation to the north, around Chesapeake Bay, and the Kirkwood Formation in New Jersey (Figure 2). Commercially profitable phosphate de- posits occur in the Pungo River Formation, so the mine, of course, does not extend below this level, although it is known from well cores that the formation rests unconformably on Pa- leogene sediments, as does the Calvert Formation. The Pungo River Formation is more than 122 m thick at its maximum, but at Aurora, near Lee Creek Mine, it is about 28 m thick. Sedi- ments at the mine are characterized by the absence of terrige- nous elastics and include diatomaceous clay, carbonates, and phosphatic sands. The lower part of the Pungo River Formation (Belhaven Member) formed on the middle to outer shelf in wa- ter depths of 100 to 200 m, whereas the upper part (Bonnerton Member) formed on the inner to middle shelf in water depths of 150 m to less than 70 m. The top of the formation is an ero- sional surface. In the middle to late Miocene, three lesser marine transgres- sions resulted in the Choptank, St. Marys, and Eastover forma- tions, but these all lie to the north and are not represented at Lee Creek Mine. The next great transgression gave rise to the Yorktown Formation and its lateral equivalents (the Cohansey Sand and the Duplin Formation), extending from New Jersey to South Carolina (Figure 3). The depth and composition of the sediments varies greatly geographically due to environmental differences. At Lee Creek, the 15 m thick Yorktown Formation rests unconformably on the Pungo River Formation. Its basal unit consists of muddy, gravelly phosphatic sand containing re- worked phosphate nodules from the underlying Pungo River. NUMBER 90 237 !£¦*— S ^\n 40" PENNSYLVANIA s^ J •NEW'^KHl 1 ,¦¦''"" ~'..^ i MARYI,ANl)W^;<^'-.-i(7 VIRGINIA \Vj /Y W N north fC,«*r r 1 M CAROLINA V^jF 3 / - Ik ^^ aPc New BcrnKicP^/^ llatteras \ \ }—kf \ > iO jiooj 1/ /H '¦--------Y, ' '^J /"^ Cape / 0 50 I (10 Miles Fcar / 0 80 / \ 1 160 Kilometers 80° 75" FIGURE 2.—Isopachous map of middle Miocene strata showing the extent of the Calvert/Pungo River embayment (shaded). l=Kirkwood Formation, 2=Calvert Formation, 3=Pungo River Formation; contours in feet. Based on Gibson (1983a:39, fig. 3). 160 Kilometers FIGURE 3.—Isopachous map of early Pliocene strata showing the extent of the embayment (shaded) of the Yorktown Formation and equivalent deposits. l=Cohansey Sand, 2=Yorktown Formation, 3=Duplin Formation; contours in feet. Based on Gibson (1983a:70, fig. 32). This is succeeded by several fossiliferous units of muddy sand and sandy mud (see section in Snyder et al., 1983, fig. 2). The top of the formation is an erosional unconformity. There are lithic and faunal changes within the Yorktown For- mation, as well as scour surfaces, although it is not known whether these formed below water or subaerially. The York- town Formation in the Albemarle embayment, which contains Lee Creek Mine, contains older strata deposited in deeper wa- ter than was in the Salisbury embayment to the north. The basal unit at Lee Creek Mine is thought to have formed under 80 to 100 m of water. The majority of vertebrate fossils from the mine are considered to have been derived from this basal part of the Yorktown Formation. This accords very well with the highly pelagic nature of almost the entire avifauna and the great scarcity of shore and land birds in the collections, indicat- ing deposition on the high seas in relatively deep water. Foraminifera could be obtained only from the upper 3.7 m of the Pungo River Formation. These correlated with foraminifer- al zones N8 and N9, indicating a late early and early middle Miocene age, equivalent to the European Langhian stage, the middle of which is about 14 Ma, the same age as the upper part of the Calvert Formation. Foraminifera from the basal part of the Yorktown Formation indicate an age of N19/20, which is early Pliocene, falling within the European Zanclian stage. The Yorktown Formation at Lee Creek Mine is considered to range in age from 3.7 to 4.8 Ma (Hazel, 1983:97). The majority of bird fossils from Lee Creek Mine come from the basal part of the Yorktown Formation, hence their age would be closer to 4.8 Ma. Paleoenvironment Marine conditions at Lee Creek during the period of deposi- tion of the Pungo River and Yorktown formations have been ably summarized by Purdy et al. (this volume), and the avifau- na strongly supports their reconstructions. Fossils were depos- ited well offshore in water of about 100 m depth at the south- western end of the Aurora Embayment, a deep depression that allowed cold waters to upwell 100 km west of the margin of the continental shelf (Popenoe, 1985). Bottom temperatures in both time periods were cool temperate (Gibson, 1967). The ichthyofauna of the Pungo River Formation indicates a warm but not tropical environment with few pelagic elements, suggesting an absence of cold upwelling. Fishes of the York- town Formation, on the other hand, are a mixture of tropical, warm-temperate, and cool-temperate taxa, with abundant tuna and other pelagic fishes, indicating a sharp temperature gradi- ent in the region with nutrient-rich cold upwelling. In the vicin- ity of Lee Creek Mine, this created what Purdy et al. (p. 188, this volume) call, rather understatedly, a "marine vertebrate, high-use feeding area," in which tuna drove prey fish to the surface where they were fed upon by abundant cetaceans, sea- birds, and sea turtles in company with an extremely abundant and diverse shark fauna. 238 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY This must have been one of the most spectacular feeding as- semblages of marine vertebrates the world has ever known. To put it in perspective, the total number of species of birds known from Lee Creek Mine is greater than the total number of spe- cies of fishes collected there. The figure is perhaps somewhat inflated for bird species by the relatively high proportion of very rare incidental shore and land birds, but nevertheless it is an extremely diverse avifauna that in terms of numbers of indi- viduals is overwhelmingly dominated by pelagic, piscivorous species, particularly auks, along with abundant shearwaters, al- batrosses, gannets, and loons. Taphonomy Although we have not attempted a taphonomic analysis, we discuss herein some impressions that may help to explain why there were so many fossils of birds in the area of Yorktown-age deposits that have been excavated at Lee Creek Mine. When the size of the mine is taken into consideration, however, the concentration of bird bones seems less impressive than that conveyed by the numbers of museum drawers filled with fos- sils. As of January 1998, the area excavated at Lee Creek Mine was 2408 hectares (5949 acres), or 24.1 km2 (9.3 mi2). Using a conservative figure of 10,000 avian fossils collected so far from the mine, this averages out to only 0.7 bird specimens per hectare. Very few fossil birds have been found in the Yorktown For- mation as associated partial skeletons. These have been found in nodules that apparently come from higher levels in the stra- tum, as opposed to the basal part of the formation, from which most of the bird bones are believed to have been derived. In the basal part of the Yorktown Formation, we suspect that much of the accumulation of bird bone ultimately came from regurgita of predators. The Yorktown seas off present-day North Carolina must have supported one of the greatest levels of marine productivi- ty in the history of the earth. As a consequence, these waters seethed with a diversity of sharks, seals, and carnivorous ceta- ceans that has no parallel on the planet today. One may envi- sion vast schools offish, among which a multitude of auks, loons, and diving shearwaters swam in pursuit of their prey. These invaders of a piscine world, despite being well adapted to their environment, would have been highly susceptible to predation by larger carnivores. A large shark wreaking havoc on a school of hake would doubtless much prefer a large, fat auk, with a beakful offish, to any smaller prey. Not only diving birds but any others that may occasionally venture below the surface, such as albatrosses and gulls, also would have been susceptible. Sharks, particularly tiger sharks (Galeocerdo), which were abundant at Lee Creek, will at times even take birds from the surface (e.g., Moseley, 1892:49; Gudger, 1949; Dodrill and Gilmore, 1978). Most bird fossils from Lee Creek Mine are found singly and are typically broken. For example, at one point in our study we had examined over 2700 alcid humeri, of which only 83 specimens (3%) were complete or nearly so. The most abun- dant bird fossils at Lee Creek Mine are the humeri, coracoids, and ulnae of the larger auks, particularly Alca antiqua (Marsh). These are among the densest and heaviest bones of any of the birds in the fauna and thus would have been the least subject to complete digestion. Furthermore, many appear etched as well as being broken. Obvious tooth marks of either predators or smaller scavengers are preserved on some (Fig- ure 4). Sharks cannot pass solid objects such as these through their spiral intestines, so "indigestible bodies must come out where they went in—through the mouth" (Gudger, 1949:46). Thus, much of the accumulation of bird bones at Lee Creek Mine may be the result of shark regurgita, which would intro- duce considerable bias into the fossil record, not only in the relative abundance of species represented, but in the propor- tion of different elements of the body as well, with bones of the wing and pectoral girdle predominating over hindlimb ele- ments, and the latter over cranial elements. Figure 4.—Parallel scratches on a tarsometatarsus of an albatross (Phoebast- ria anglica, USNM 430533), presumably made by the teeth of a predator or scavenger. (Magnification x-4.3.) NUMBER 90 239 The Species Question In dealing with an avifauna such as that of the Yorktown Formation, we have been forced to deal with difficult ques- tions, both practical and philosophical, regarding the limits of species, particularly through time, that are inescapable in a work of this nature. As might be expected, our own views evolved during the course of our investigations, and so some discussion of the approach used herein is warranted. During the third quarter of the twentieth century, avian pale- ontology in North America developed under the influence of Pierce Brodkorb, who presumed that most species of birds did not cross epoch boundaries (Brodkorb, 1960). On the face of it, there is little reason why this should be so, and it is counter- intuitive to presume that speciation occurred simultaneously in all lineages at the transition from the Pliocene to the Pleis- tocene, for example. Such may have been the wish of alpha- level systematists for whom the description of new species was an end in itself, but this is not helpful in documenting and comprehending evolution. There are also workers, whose re- search is more faunally oriented, who subscribe to nomencla- tural recognition of any morphological variation that is tempo- rally removed from modern taxa, advocating that "in the case of a smallest constant morphological difference a new system- atical name is reasonable" (Janossy, 1987:190). Another matter that compounds the difficulties of referring to species in a fauna such as that at Lee Creek Mine is the cladistic viewpoint that ancestors cannot or should not be identified in the fossil record (e.g., Englemann and Wiley, 1977; Eldredge and Cracraft, 1980; Norell, 1996), although this has been vigorously contested by others (e.g., Wagner, 1995, 1996; Foote, 1996). The proposition that, within a rela- tively small and circumscribed basin such as the North Atlan- tic, none of the multitude of species that existed four to five million years ago in Yorktown times was ancestral to any ex- tant species is hardly tenable. The main problem arises when a given lineage may have split into two species. How, then, does one recognize (and deal nomenclaturally with) the common ancestor of the two? The Pliocene loons at Lee Creek Mine may reflect just such a problem, with Gavia concinna Wet- more possibly being ancestral to the living Pacific Loon, G. pacifica (Lawrence), and the Arctic Loon, G. arctica (Lin- neaus), and with a new large species (described below) con- ceivably being ancestral to the modern Common Loon, G. im- mer (Briinnich), and the Yellow-billed Loon, G. adamsii (Gray). This is rather a different matter from saying that the harle- quin duck (Histrionicus) at Lee Creek Mine, although differ- ing somewhat from the single living species, is still probably on a direct line with that species and need not be named as a distinct taxon. Compounding the difficulties of communica- tion are the semantic problems that arise from the fact that the same terminology—species—is applied to two completely dif- ferent phenomena. As Haffer (1995) has discussed at length, extant species, regardless of which concept one employs to de- fine them, are geographical entities that occupy areas with more or less defined borders that may or may not overlap with congeneric species. In contrast, temporal "species" are arbi- trary, morphologically defined units of a continuous lineage through time. Haffer rightly concluded that Linnean species nomenclature should not be used for temporal entities, but no consensus on a useful alternative terminology has yet emerged. Mammalian paleontologists have had much longer to con- tend with such issues, whereas the fossil record in birds has only recently become sufficiently extensive, particularly among seabirds, to provide investigators with any meaningful possibility of detecting changes within a lineage. Differentiat- ing between extant lineages and those that are extinct, with no modern descendants, should be one of the most important ac- tivities of paleontologists dealing with late Cenozoic faunas. We can hardly have any sensible discussion of faunal changes and "turnover" without having at least some grasp of whether the "disappearance" of individual units is due to extinction or to the evolution of new morphologies. In the present analysis, when the material from the York- town Formation is reasonably extensive and we can still find little or no differences from existing species, we have simply referred the fossils directly to living taxa, as, for example, among some of the albatrosses. We attempted to identify extinct versus extant lineages, but there are various sources of potential error in this undertaking. Not the least of these is that different groups of birds have changed morphologically (but not necessarily speciated) at very different rates. Shearwaters and albatrosses (Procellarii- formes), for example, have evolved very slowly. Some procel- lariiform lineages appear to date back at least to the late Oli- gocene with little change in morphology. Puffinus conradi Marsh from the Calvert Formation was very similar to the liv- ing Greater Shearwater, P. gravis (O'Reilly), so it would be very difficult to separate the Yorktown-age representative of this lineage from earlier and later manifestations. On the other hand, loons (Gaviidae) evolved very rapidly in the Neogene. Since the middle Miocene, they have increased greatly not only in size but also in degree of specialization of both the leg and wing elements for underwater propulsion. Be- cause there are three species of loons in the Yorktown Forma- tion, it is tempting to suggest that these are the Pliocene repre- sentatives of the three major extant lineages, which may well be the case. Two of the three modern lineages of loons now consist of pairs of sibling species, of unknown times of diver- gence. Two of the early Pliocene forms therefore may repre- sent the common ancestors of the sibling-species pairs rather than direct ancestors of any of the sibling species. Through the collection of more fossils from different time intervals and geographic areas, we can hope to resolve this problem satis- factorily. Sometimes the problem is too many fossils. We can easily recognize a flightless early Pliocene relative of the modern 240 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Great Auk, Pinguinus impennis (Linnaeus), lineage in the Yorktown deposits. In addition to the rare but unmistakable Pinguinus bones, however, the fauna contains a vast array of fragmentary and usually unassociated fossils of closely related species of the genus Alca that constitutes a continuum varying so much in size that we calculate it to have encompassed four species at minimum. Nevertheless, there is only one modern descendent of this assemblage, the Razorbill (Alca torda Lin- naeus). Although the ancestor of Alca torda clearly is repre- sented among all these broken bones, assigning individual fragmentary limb bones to this lineage is not always possible. The other side of the coin, too few fossils, is the situation most frequently encountered. Many late Neogene species of birds have been founded on material too inadequate even to distinguish extant from truly extinct lineages. We have designated lineages that are apparently ancestral to modern species with the prefix "aff." (Latin affinis, related to, neighboring), even if minor differences in structure can be dis- cerned (for example, in Histrionicus). We have listed as syn- onyms previously named fossil taxa that evidently are continu- ous with modern lineages. We have used the designation "cf." (Latin confer, compare) to denote only a general morphological similarity, not necessarily a close relationship. Even the ambiguous designation "aff." may be too precise when confronted with taxa such as the dabbling ducks of the genus Anas. Among modern species of this genus it is very dif- ficult or impossible to identify fragmentary material to species because many species do not differ postcranially other than in size, and among species there may be considerable variation and overlap. In such such osteologically difficult groups, we used the designation "magn." (from Latin, magnitudino) to in- dicate the approximate size range of the fossils in comparison with modern species, without implying possible relationship. The waterfowl, i.e., ducks and geese (Anatidae), provide par- ticularly illustrative examples of some of the problems we have attempted to address. The most abundantly represented duck at Lee Creek Mine turns out, unexpectedly, to be a harlequin duck, Histrionicus, the material of which is sufficient to show that the Yorktown bird was very similar to the extant species, although the wing was slightly less specialized. By previous standards, this would have been a nameable species, yet only the quality and abundance of the material would allow it to be distinguished from the modern species, and then only in some of the elements. Although it is interesting to find that the wing has changed slightly in this lineage in the past few million years, does this mean the Pliocene bird is a new species? When we found that this apparently new species was bracketed by names given to extremely scanty, generically misidentified fos- sils that were both older and younger than the Yorktown For- mation, it reinforced our opinion that the use of Linnean no- menclature for such fossils that appear to belong to existing lineages is untenable. If we have been overly enthusiastic in our species attributions, we hope nevertheless to have fostered a healthy new trend in viewing late Cenozoic avifaunas. Methods In a few instances, bird fossils were found on spoil piles of known stratigraphic position. In most cases, however, speci- mens were brought in as mixed lots gleaned from the spoil piles with no stratigraphic information. Some differences in preservation are apparent; bones from the Pungo River Forma- tion are usually better preserved, black in color, less water- worn, and have a more polished appearance than do those from the Yorktown Formation, but such criteria cannot always be re- lied on. The stratigraphic position of a particular fossil can of- ten be determined by examination of the adherent matrix and its contained microfauna. Thomas G. Gibson examined the mi- crofossils in matrix samples from 43 fossil bird bones selected at random from the early collections from Lee Creek Mine. His determinations indicate that all of these probably came from the basal part of the Yorktown Formation. The great quantity of fossil material, in particular that repre- senting the Alcidae, and the unresolved systematics of some taxa make it impractical to list all specimens thus far identified. We have, however, listed by catalog number referred material for all but the most common species; for the latter we list spec- imens that are unusually complete or diagnostic or that belong to uncommonly represented skeletal elements. Unless designat- ed by some other museum acronym (see "Acknowledgments"), all listed material is in the Vertebrate Paleontology collections of the National Museum of Natural History (NMNH, which in- cludes collections of the former United States National Muse- um (USNM)), Smithsonian Institution. To attempt to determine whether specimens from Lee Creek Mine were from the Pungo River Formation or the Yorktown Formation, we made comparisons with fossil birds from the Calvert Formation in Maryland and from the Upper Bone Val- ley Formation in Florida. Most of the Maryland specimens also are housed in the Vertebrate Paleontology collections of the National Museum of Natural History; most of those from Bone Valley are housed in the Florida Museum of Natural History (UF), which now includes the collection of Pierce Brodkorb (UF PB). Modern comparative material was mostly from the collections of the Division of Birds, Department of Vertebrate Zoology, National Museum of Natural History, Smithsonian Institution, and also have the USNM acronym, although these collections have a separate numbering system. Measurements of bones were taken according to the illustra- tions in von den Driesch (1976), unless otherwise stated, and are given in millimeters. Osteological nomenclature is modi- fied from Howard (1929). Literature citations are given for au- thorities of scientific names of fossil taxa only. Authorities for extant North American taxa are generally those cited in the seventh edition of the A.O.U. Check-list (American Ornitholo- gists' Union, 1998); those for taxa extralimital to that work are from Sibley and Monroe (1990). Elements are listed in the following order: skull, vertebrae (except synsacral vertebrae), sternum, furcula, coracoid, scapu- la, humerus, ulna, radius, carpometacarpus, alar phalanges in NUMBER 90 241 numerical order, pelvis, femur, tibiotarsus, tarsometatarsus, and pedal phalanges in numerical order. Right elements are listed before left elements, and complete elements are listed be- fore partial elements, with proximal portions listed before dis- tal portions (scapular portions before sternal portions in the case of coracoids). Systematic Paleontology Pelagic Birds The species account are divided into two sections. The sec- ond section deals with shore and land birds that are incidental to the fauna, usually being known by a single bone or fragment or by at most a few specimens. The present section deals with the pelagic avifauna, which for the most part consists of species that lived and fed in the vicinity of deposition. Some of the spe- cies of waterfowl and gulls would probably not have met this criterion, although most are more commonly represented than the incidental shore and land birds. On the other hand, most of the species represented at Lee Creek Mine in these families would certainly have been members of the pelagic avifauna, and we did not consider it practical to split the accounts of these families along speculative ecological lines. Order Gaviiformes (loons) Family Gaviidae (loons) Genus Colymboides Milne-Edwards, 1867 Colymboides? sp. Material.—Distal end of right humerus, USNM 302286. Left carpometacarpus missing distal end, minor metacarpal, and proximal parts of carpal trochlea, KU 21251. Horizon.—Pungo River Formation inferred from known temporal distribution of genus. Measurements (mm).—Humerus: Distal width, 8.8. Carpometacarpus: Length of alular metacarpal, 9.0. Remarks.—These specimens appear to be referable to Co- lymboides rather than to Gavia based on the following charac- ters. In the humerus, the attachment of the anterior articular lig- ament extends proximally only to the level of the ectepicondyle, and the surface is nearly plane, rather than sloping medially; in distal view it is much flatter, and the condyles are much less bulbous; the olecranal fossa does not extend proximo-intemally between the internal condyle and the entepicondylar process; and the ridge on the internal border of the impression of the bra- chialis anticus is more diagonally oriented with respect to the shaft. In the carpometacarpus, the alular metacarpal is very short; the shaft is relatively unflattened; the proximal symphysis between major and minor metacarpals is very short; and the lig- amental attachment of the pisiform process is not flattened or produced anteriorly. These specimens are from a species smaller than any of the other loons from Lee Creek Mine or the Chesapeake Group. Their size suggests that they pertain to an undescribed species, the humerus being smaller than that of Colymboides anglicus Lydekker but larger than that of C minutus Milne-Edwards. If correctly referred to genus, they would provide the first record of the genus Colymboides outside of Europe. The latest occur- rence of Colymboides in Europe is in the early Miocene of the Cheb Basin of the Czech Republic, where the tiny species C. minutus, otherwise known only from the early Miocene of France, has been found (Svec, 1980, 1982; Mlikovsky, 1996). This species is known only from freshwater deposits, so it is quite possible that the genus Colymboides may have persisted somewhat later in marine environments (see discussion of Ga- via egeriana, below). Genus Gavia Forster Gavia egeriana Svec, 1982 Material.—Right coracoid, USNM 430523. Proximal two- thirds of right ulna, USNM 241423. Horizon.—Pungo River Formation inferred from similarity to fossils from Calvert Formation. Measurements (mm).—Coracoid: Length with sternal facet flat on calipers, -35; width and depth of shaft at midpoint, 4.3x3.8. Ulna: Proximal width and depth, 7.3 x 8.3. Remarks.—Gavia egeriana is a very small species of loon described from two distal ends of humeri from early Miocene deposits at Dolnice, in the Cheb Basin in the Czech Republic (Svec, 1982). The micromammal zone of these deposits is MN 4b (Mlikovsky, 1996); thus, these deposits are probably slight- ly older than the Calvert or Pungo River formations, and fossils of this species form the earliest occurrence of the genus Gavia. Several specimens of Gavia have been recovered from the Calvert Formation in Maryland, Virginia, and Delaware (Ras- mussen, 1998) and appear to be referable to two species, differ- ing in size. We were able to compare the holotype of Gavia egeriana directly with these specimens, and we assign the larg- er of the two Calvert loons to that species. The smaller Calvert loon is undescribed and is not known from Lee Creek Mine. The coracoid from Lee Creek Mine listed above is black and phosphatized. It compares well with a coracoid from the Cal- vert Formation that is associated with a sternum, scapula, and furcula (USNM 23717) that we refer to G. egeriana based on size. Likewise, the ulna from Lee Creek Mine is close to one from the Calvert Formation that is associated with a radius and the proximal end of a humerus of G. egeriana (USNM 237204). We therefore assume that the specimens from Lee Creek Mine are from the Pungo River Formation, which is fur- ther supported by the preservation of the coracoid. The Calvert material will be described in detail elsewhere. 242 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Gavia howardae Brodkorb, 1953c PLATE ia,c,d,fhJ,l,m,o,q MATERIAL.—Right coracoids, USNM 244209, 321271; complete left coracoids, USNM 192845, 366413; right cora- coid lacking part of both ends, USNM 446482; left coracoid lacking part of both ends, USNM 430515; scapular end of right coracoid, USNM 257461; medial portion of left coracoid, USNM 302363. Anterior end of right scapula, USNM 430472. Left humerus, USNM 206448; proximal end of left humerus, USNM 446473; distal ends of right humeri, USNM 206347, 242363, 257480, 430508, 430510-430512, 446474,446475; distal ends of left humeri, USNM 215850, 257465, 366664, 446478. Proximal end of right ulna, USNM 215436; proximal ends of left ulnae, USNM 252348, 446471; distal ends of right ulnae, USNM 215473, 215629, 430500; distal ends of left ul- nae, USNM 193204,430494. Distal end of right radius, USNM 177820. Right carpometacarpus missing alular and minor metacarpals, USNM 460766; proximal portions of left car- pometacarpi, USNM 177783, 366702, 460767; distal end of left carpometacarpus, USNM 192449. Right femora, USNM 215426, 367161, 446480; right femur lacking distal end, USNM 248590; distal end of left femur, USNM 250681. Distal end of right tibiotarsus, USNM 192676. Left tarsometatarsus lacking most of both ends, USNM 430520; proximal end of right tarsometatarsus, USNM 446481; proximal ends of left tarsometatarsi, USNM 193274, 430522; distal ends of right tar- sometatarsi, USNM 183423, 183431, 366565, 460768; distal ends of left tarsometatarsi, USNM 177879, 193059. Horizon.—Yorktown Formation. Additional Material Examined.—San Diego Forma- tion, California: Right humerus, SDSNH 42762; right hu- merus lacking most of proximal end, SDSNH 42776; left hu- merus, SDSNH 42763; fragmented left humerus, SDSNH 42781. Distal half of right ulna, SDSNH 35252. Left radius, SDSNH 42778. Left carpometacarpus, SDSNH 42774; proxi- mal end of right carpometacarpus, SDSNH 42768. Right femo- ra, SDSNH 42764, 42775; left femur, SDSNH 42779. Right ti- biotarsus, SDSNH 42765; left tibiotarsus, SDSNH 42772; distal end of left tibiotarsus, SDSNH 42773. Proximal half of right tarsometatarsus, SDSNH 35251. Measurements.—See Table 1. Remarks.—Gavia howardae was first described from the late Pliocene (Blancan) San Diego Formation of California (Brodkorb, 1953c). Chandler (1990a) referred a subsequent specimen from the same deposits to this species, and numerous additional specimens, which we examined (see list above), have since been found there. As in the Red-throated Loon, G. stellata (Pontoppidan), the bones of G. howardae are slim rela- tive to those of the Arctic Loon, G. arctica, or the Pacific Loon, G. pacifica, and many of them are shorter than those of G. stel- lata (Table 1). Characters given for G howardae by Chandler (1990a) were used to identify the Lee Creek Mine specimens, and other elements were referred to this species mainly on their small size and gracile proportions. The small size of G. howardae suggests a relationship to G. stellata, the smallest modern loon, and this also is supported by small fossil cora- coids from Lee Creek Mine. These agree with G. stellata in having only an incisura on the medial edge of the coracoid, whereas in most individuals of all other modern and fossil loons (in which the coracoid is known), there is a distinct, closed procoracoid foramen. Gavia moldavica Kessler (1984) was described from por- tions of all the major wing elements from the early late Mi- ocene (Middle Sarmatian) of Kishinev (Chisinau), Moldavia. The measurement given for the distal end of one of the paratyp- ical humeri is 12.0 mm, which is within the range of G. howardae (Table 1). The length of a paratypical radius of G. moldavica was 87.6 mm, compared with 86.5 mm for a radius of G howardae (SDSNH 42778). There is, however, a consid- erable period of time between the probable age of G moldavica (MN ?9 (=planktonic foraminifera zone N15) according to Mlikovsky, 1996), the Yorktown Formation at Lee Creek Mine (N19), and the even younger San Diego Formation (N21). Giv- en that loons appear to have been increasing rapidly in size dur- ing the last half of the Neogene, it is possible that G. moldavica could have been ancestral to the larger Pliocene species G con- cinna. Gavia schultzi Mlikovsky (1998), from the middle Miocene (upper Badenian) of Austria, is somewhat younger than the Calvert/Pungo River formations and apparently falls within the lower size range of G. howardae, being considerably larger than any middle Miocene loon yet known from North America. Gavia brodkorbi Howard (1978) is known from a complete ulna from the early late Miocene (Clarendonian; N14-16) at Laguna Niguel, California. This is shorter and more robust than in G. howardae. The holotype of Gavia paradoxa Umanskaja (1981) is the proximal portion of an ulna from the late Miocene (MN 11-13 = ~N 16-17) of the Ukraine. The published dimen- sions of G paradoxa suggest that it is very similar in size to G. brodkorbi, from which it appears to differ in its long attach- ment for the anterior articular ligament. The relationship of these late Miocene loons to the tiny spe- cies of the middle Miocene and the larger ones of the Pliocene can only be determined with more and better material and di- rect comparisons with the types. Gavia howardae is very similar to the modern Red-throated Loon, Gavia stellata, and is probably on a direct line with that species. Although there is overlap in size (Table 1), the fossil form on average is smaller, and some individuals were smaller than any of the individuals in the modern sample. The main qualitative difference noted was that the pectoral crest of the humerus in G. howardae is not as long and low as it is in the modern form. NUMBER 90 243 Gavia concinna Wetmore, 1940 PLATE 2a,c,d,fg,i,k-n,p,r-t,v,x,y Gavia concinna Wetmore, 1940:25. Gaviapalaeodytes Wetmore, 1943a:64. Gavia sp., Howard, 1982:3. Material.—Because of the abundance of material of this species from Lee Creek Mine, only the best-preserved speci- mens and rarer elements are listed. Right coracoid, USNM 430477; left coracoids, USNM 366592, 430476; scapular ends of left coracoids, USNM 192029, 192033, 192772. Anterior end of left scapula, USNM 192083. Proximal end of left humerus, USNM 430501; distal ends of right humeri, USNM 367020, 430470, 446489, 460771; distal ends of left humeri, USNM 181037, 252360, 366642, 366894, 430460, 430503. Proximal ends of right ul- nae, USNM 430457, 430458, 446470, 446485, 446491, 460770; proximal ends of left ulnae, USNM 430453, 446472; distal ends of right ulnae, USNM 252340, 252377; distal ends of left ulnae, USNM 430443, 430456, 446488, 460769. Proxi- mal end of left radius, USNM 446484. Right carpometacarpus missing minor metacarpal, USNM 430451; proximal ends of right carpometacarpi, USNM 215873, 446483; proximal ends of left carpometacarpi, USNM 367132, 430441. Proximal half of fused synsacral vertebrae, USNM 460774. Right femora lacking part of distal ends, USNM 183477, 430518, 446479; left femora lacking part of distal ends, USNM 366000,460783. Proximal end of left tibiotarsus, USNM 430519; distal ends of right tibiotarsi, USNM 241388, 446490; distal ends of left ti- biotarsi, USNM 215646, 430481, 430482. Right tarsometatar- si, USNM 366714, 430485, 430486; left tarsometatarsus, USNM 193359; proximal end of left tarsometatarsus, USNM 430446; distal ends of right tarsometatarsi, USNM 430490, 460773; distal end of left tarsometatarsus, USNM 460772. Pedal phalanx, USNM 192543. Horizon.—Yorktown Formation. Additional Material Examined.—San Diego Forma- tion, California: Left humeri, SDSNH 42761, 42763; distal half of left humerus, SDSNH 42767. Right carpometacarpus missing minor metacarpal, SDSNH 42769. Right major alar digit phalanx 2, SDSNH 42771. Left femur, SDSNH 42777. Right tarsometatarsi, SDSNH 22916, 42766; left tarsometatar- si, SDSNH 42770,42780. Bone Valley Formation, Florida: Right coracoids, UF PB 132, USNM 256375. Proximal half of right humerus, UF PB 306; distal thirds of right humeri, UF PB 88, 524, USNM 256376; distal halves of left humeri, UF PB 297, USNM 256395. Distal third of right ulna, UF PB 89. Right femora, UF PB 133, USNM 256378; left femur, UF PB 298. Proximal half of right tarsometatarsus, USNM 256374. Horizon Uncertain: Distal end of left radius, USNM 460775, from Renny Creek, New Bern, Craven County, North Carolina. Yorktown or equivalent deposits are exposed there, but the specimen might be Pleistocene and is included herein for purposes of illustration (Plate 2/). Measurements.—See Table 1. REMARKS.—Gavia concinna Wetmore (1940) was described from the proximal end of an ulna from the Etchegoin Forma- tion of Monterey County, California, which is late Hemphillian in age (Becker, 1987) and thus is approximately contemporane- ous with the Yorktown and Bone Valley formations. This spe- cies was said to be intermediate in size between the modern species G. stellata and the Common Loon, G. immer. It was not stated how the fossil species could be discriminated from mod- ern G. arctica or G. pacifica, however, which are in this inter- mediate size range. Wetmore (1943a) then described Gavia palaeodytes from an imperfect coracoid from the Bone Valley Formation of Florida. He considered the coracoid to be smaller than that of any then-known species of loon, living or fossil; however, it is stout and has a heavily rimmed procoracoid fora- men, unlike G. stellata or the Lee Creek material referred to G. howardae. Brodkorb (1953c) referred additional material from California to G. concinna, which he regarded as being most similar to G. pacifica, although approaching G. immer in size. He also referred material from Bone Valley to G. concinna, which he considered to be a larger species than G. palaeodytes, which in turn was said to be the size of G. stellata. Delle Cave et al. (1984) described the skull of a loon from the Pliocene of Italy and identified it as Gavia cf. concinna. Chandler (1990a) dismissed all published records of the oc- currence of Gavia concinna in the San Diego Formation as misidentifications. At least seven specimens collected there in 1989 and 1990 (listed above), however, are best referred to this species, being far too large to be G. howardae; none is so large as definitely to pertain to the new species of Gavia described herein (see below). A worn tarsometatarsus from the San Mateo Formation (ear- ly Hemphillian), in San Diego County, California, was referred to by Howard (1982) only as "Gavia sp.," but its size and age are compatible with G. concinna as defined herein, to which we tentatively refer the specimen. The material from Bone Valley, including part of that re- ferred by Brodkorb (1953c) to G. concinna, appears to be a composite, as it contains at least two specimens (proximal end of humerus UF PB 593; distal end of humerus UF PB 90) that are too large for that species and that we refer to the new spe- cies of Gavia described below. We regard all of Brodkorb's (1953c) material of G. palaeodytes as belonging to G. concin- na. Although there is considerable size variation shown in this series, with some specimens being quite small, none of the ma- terial we have yet seen from Bone Valley can convincingly be referred to G. howardae. Therefore, we regard the great major- ity of specimens of loons from Bone Valley, including the ho- lotype of G palaeodytes Wetmore (1943a), to be referable to G. concinna Wetmore (1940). Emslie (1998) referred to G. concinna the distal end of a tar- sometatarsus from the earliest Pleistocene of Florida, where it was contemporaneous with G. pacifica. Otherwise, G. concin- na would presumably be part of the lineage that includes G. arctica and G. pacifica, which have often in the past been con- 244 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY sidered conspecific. The fossils differ very little from the mod- ern forms. A much more detailed analysis would be required to determine whether the split between G. arctica and G. pacifica took place before or after the early Pliocene, if indeed such a determination can be made osteologically. Gavia fortis, new species FIGURES 5, 6; PLATES 3a.c,e,g-i,k,l,n, 4a-c,e,fh.j.l,n,p,q,s.u,v,x,z Holotype.—Associated partial skeleton consisting of a por- tion of shaft of right humerus, distal end of left humerus, prox- imal ends of right and left radii and ulnae, proximal end of right carpometacarpus, synsacrum, fragment of left innominate with most of the acetabulum, distal end of right tibiotarsus, complete right tarsometatarsus, pedal phalanx, and various small fragments of bone, USNM 252432. Type Locality.—Texasgulf Inc. Lee Creek Mine, south side of Pamlico River, near Aurora, Beaufort County, North Carolina (35°23'22"N, 76°47'06"W). Horizon and Age.—As determined from foraminifera and sedimentary characters of matrix, the holotype is from the low- er to middle part of the Yorktown Formation, lower Pliocene. Paratype USNM 179222 is from the basal Yorktown, and paratype USNM 178148 is from the lower to middle part of the Yorktown Formation. Other paratypes are assumed to be of the same age. Distribution.—Known from the type locality and from the Bone Valley Formation in central Florida. Measurements of Holotype (mm).—Humerus. Distal width, 17.5. Ulna: Proximal width, 13.0; proximal diagonal, 13.1. Radius: Proximal width, 7.4 Carpometacarpus: Proximal depth, 16.0. Tibiotarsus: Distal width, 13.9; distal diagonal, 13.1. Tarsometatarsus: Length, 77.5; proximal width, 14.1; dis- tal width, 11.0. TOPOTYPICAL PARATYPES.—Associated Specimen: Partial skeleton consisting of vertebral fragments, proximal end of left radius and ulna, left femur, proximal and distal ends of left ti- biotarsus, left tarsometatarsus, and pedal phalanx, USNM 302392. Individual Elements: Cervical vertebra, USNM 460782. Left coracoid, USNM 215562; scapular ends of right cora- coids, USNM 215463, 215502; scapular ends of left coracoids, USNM 206368, 206432, 210452. Anterior end of right scapula, USNM 206587. Left humerus lacking most of proximal end, USNM 206625; proximal ends of right humeri, USNM 244212, 460778, 460779; proximal ends of left humeri, USNM 215840, 366588; distal ends of right humeri, USNM 178149, 192848, 192981, 193009, 252370, 252372,460780; distal ends of left humeri, USNM 177742, 192450, 192771, 193230, 242173, 252353. Right ulna, USNM 250778; proximal end of right ulna, USNM 367044; proximal ends of left ulnae, USNM 206450, 215753; distal end of right ulna, USNM 366414; distal ends of left ulnae, USNM 178148,460777. Left radius, USNM 192060/192065 (two pieces fitting together); proximal end of right radius, USNM 460795; distal end of right radius, USNM 215749; distal end of left radius, USNM 430444. Right car- pometacarpus lacking proximal end, USNM 460776; proximal end of right carpometacarpus, USNM 430442; proximal ends of left carpometacarpi, USNM 244210, 430440. Right femora, USNM 183459, 206348, 275779, 367062,482593; left femora, USNM 177912, 179222, 250713, 257491, 275842, 460781; right femur lacking proximal end, USNM 308212; left femur lacking distal end, USNM 366681. Proximal end of right tibio- tarsus, USNM 256254; distal ends of right tibiotarsi, USNM 178057, 430483; distal end of left tibiotarsus, USNM 206593. Right tarsometatarsus, USNM 206629; proximal ends of right tarsometatarsi, USNM 183492, 206533, 308193; proximal ends of left tarsometatarsi, USNM 215570, 242340, 308204, 366559, 446492, 482592; distal ends of right tarsometatarsi, USNM 206302, 257501, 430449, 430450; distal ends of left tarsometatarsi, USNM 192880, 215772, 250753, 430447, 430448, 446493. Pedal phalanx, USNM 236812. Additional Paratypes.—Bone Valley Formation, Florida: Proximal third of right humerus, UF PB 593; distal third of left humerus, UF PB 90. Measurements of Paratypes.—See Table 1. Etymology.—Latin fortis, strong, powerful; from the ro- bustness of the bones in comparison to the Common Loon, Gavia immer. Diagnosis.—Larger than any living or fossil species of Gavia except G. immer and the Yellow-billed Loon, G. adam- sii. Smaller than all but the smallest individuals of G. immer, but all skeletal elements markedly more robust. Differs from G. immer in having the sacrum in lateral view more curved ventrally. Remarks.—This common species at Lee Creek Mine is larger than any of the known fossil forms of Gavia but is simi- lar in overall size to the modern G. immer-G. adamsii super- species. Gavia fortis is similar to G. immer and differs from G. concinna and G. arctica in that the ectepicondylar prominence of the humerus is more laterally produced, and the attachment for the anterior articular ligament is longer; the distal end of the ulna is much expanded, especially the palmar edge of the shaft immediately proximal to the articular surface, the base of the internal cotyla is rounded and heavy, the internal-palmar edge of the shaft is less ridge-like, and the internal cotyla and inter- nal condyle are more produced palmarly; the distal end of the radius is more expanded; there is a larger spur on the posterior edges of the external and internal condyles of the tibiotarsus, the distal intercondylar sulcus is wider, and the entire distal end is more expanded; and the distal foramen of the tarsometatar- sus is proportionately larger in posterior view, and the inter- trochlear notch between trochleae III and IV is narrower. Regalia (1902) described a large species of loon, Gavia por- tisi, based on a broken tenth or eleventh cervical vertebra from the Pliocene of Italy (middle Pliocene at Orciano Pisano near 3 u <-. ^ CO II u •—' 2 !* a £ I OJ BO C d & tq to c OJ M U S a to * C) bo c ¦ £ R to »0 CN II e CO C 61 Q 2 .5 £J u ^ GO a b CO p2 fN VO un —^ — . "** t oo"r-f rn on h * m> 2 _. c o « J j £ Q 5 —•mm o m o vb no' on un U-( fN un vO 0 "t — ¦"* — T O NO vo rn m o V-J —¦ — OO CO On vd OC fs r-' t—" ^f" -3-" rN ON ^r r- ^t fN 00 fN un un vO O O O _ NO — un """ rN rN rn pm NO fN m »i r-* q ^^ «:^ r*i T T 7 7 T T t*; ^ on ^ p n JN 2 00 J2 ^ — »n cn fN CTv — ON ^4 un m on f*! —* "* —' CN ^_: P ^ rs OC OC NO rN ON m ON NO m NO un — O O 0 O 0 O fN r- NO — CN ^i- NO OO m rn ^t ON ¦^r - ~ r* un un un _ OO rN 2 Vi NO ON un 5 rs rs un rn un r- un m rN m 00 m O 0 «> n ^ «, un'O fN <* vo r- — 00 --'1 ¦^ vo r- no Mh O r^ vO VO — m on ©' O —. p 00 vo vo" m' r-; -^ t^ m' «S rn tt no un rn un rn rn — r^ rN 00 T fN ON 1 un m 5 00 m J_ CO E ~ r^ 0 m ¦nr m -0 0 £ E * Et o .2 DCS J o, Q 3 1 O CO g -a "5 E >< 0 c 0h c 0 00 •a CO ¦3 c0 =3 c M ¦a CO F ¦S & U TO Efl c 0 00 CO TO E TO D. U TO CO 2 4J J= S M 3 1-0 t) 3 CO a; X CO CO CO X 2 to 00 c £¦ ° 5 _> < Id O £ a D X> £ a Q 6 -J u 0- Uh H H -S -c 5= S m ¦0 J= s J= ¦0 0. TJ ^ — ca S .t^ fj fc CD C a* 5^ ca T3 0 vo u -1 0. (J 246 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Val di Fine). From the illustrations it appears that the specimen was correctly identified as a loon. Although Brodkorb (1953c) stated that this came from a species the size of modern G. im- mer, the measurements seem to indicate a smaller species (Delle Cave et al., 1984), but the holotype can no longer be found (Delle Cave et al., 1984; Delle Cave, 1996). Although Colymbus portisi Regalia is the earliest name applied to any Pliocene loon and almost certainly pertains to one of the younger epithets we have used in this paper, the application of the name can seemingly no longer be determined. Thus, we re- gard Colymbus portisi Regalia as a nomen dubium and we have not used it. There appear to be no records of Gavia immer earli- er than the late Pleistocene (Emslie, 1998). Differences in proportions between the modern and fossil loons were confirmed by a bootstrapped principal components (PC) analysis of 12 variables (listed in Table 2) for Gavia im- mer (n= 13), G. pacifica (n= 10), and G. fortis (n=2 associated partial skeletons) using a covariance matrix of logi0-trans- formed data (Figure 5). By far the greatest amount of variation (89%) was explained by PC-I, on which all variables loaded highly and positively, as is typical for a general-size axis (Ta- ble 2). All T statistics (resampled eigenvectors divided by their standard errors) for PC-I were greater than T value (|T|)=5, which can be considered the level above which T sta- tistics are significant (Marcus, 1990), although tarsometatar- sus-shaft width had a much smaller |T| than did any other vari- able (Table 2). Gavia fortis is similar on factor I, and thus in overall size, to small G. immer and large G. pacifica. The only length measure available for both of the associated skeletons of G. fortis, however, was tarsometatarsus length, which was shorter than that of G. immer; elements from other specimens of G. fortis are shorter than similarly stout elements of G im- mer (Table 1). On PC-II, tarsometatarsus-shaft width was by far the most important variable (Figure 6), with a |T| >2 (Table 2), contrasted with the less important ulna-shaft width. On PC- III, which remained the third axis after 500 bootstrap itera- tions, tibiotarsus distal depth and tarsometatarsus length (Fig- ure 6) were the most important variables, based on their T val- ues (Table 2). On factors II and III, G. immer and G. pacifica overlap widely (Figure 5), whereas G. fortis differs from both extant species in having a combination of a relatively short, heavy tarsometatarsus (scores low on PC-III and high on PC- II), and a relatively thin ulnar shaft immediately distal to the external condyle (scores high on PC-II). In proportions, the u q- 0.75- 0.50- 0.25- 0.00- -0.25- -0.50- -0.75- 64 C 0 heavy ulna heavy tarsometatarsus i i i i i--------1--------r -0.75 -0.50 -0.25 0.00 0.25 0.50 0.75 PC-II Figure 5.—Scatter plot of individual factor scores from principal components analysis of loons Gavia immer (I), G. pacifica (P), and associated specimens of G fortis, new species (F). NUMBER 90 247 two associated specimens available of G. fortis differ more from these two recent species than they differ from one anoth- er. In fact, the principal components analysis underrepresents the distinctness of G. fortis because of the scarcity of usable length measures for this species. PC-II 1.0 Tarsometatarsus _ _ shaft width PC-I -l.ol- a PC-I -11.0 _ J Ulna shaft width 1 -1.0 PC-II PC-III 1.0 Tarsometatarsus length I Tibiotarsus 1 distal depth PC-II -1.0|- +—T I Ulna shaft width PC-II -11.0 Tarsometatarsus distal breadth Tarsometatarsus shaft width -1.0 PC-III Figure 6.—Component loadings for principal components analyses of loons Gavia immer, G. pacifica, and G. fortis, new species: a, factors I and II; b, fac- tors II and III. TABLE 2.—Summary of results (T statistics, eigenvalues, and percent of vari- ance explained) for factors I—III of a bootstrapped principal components analy- sis of 12 skeletal variables for loons Gavia immer, G pacifica, and G. fortis, new species. Variable T statistics PC-I PC-II PC-III Ulna proximal width 19.76 -1.34 -0.15 Ulna shaft width (immediately distal to base of external condyle) 10.41 -1.60 -0.51 Ulna proximal diagonal 21.87 -0.79 -0.19 Tibiotarsus distal width 19.68 0.71 0.95 Tibiotarsus distal depth 19.37 0.06 1.76 Tibiotarsus condyle height 14.28 -0.33 -0.36 Radius proximal width 24.90 0.16 0.52 Tarsometatarsus length 14.90 0.03 1.46 Tarsometatarsus proximal width 21.84 0.41 0.89 Tarsometatarsus distal width 13.80 0.25 -0.67 Tarsometatarsus distal depth 20.74 0.85 -0.27 Tarsometatarsus shaft width 7.82 2.04 -0.58 Eigenvalues 0.0296 0.0009 0.0007 Percent of variance explained 89.26 2.9 2.2 Discussion of Gaviidae Although modern loons breed in the boreal zone and are en- tirely confined to the Northern Hemisphere, even in winter, the fossil record shows that the family once occurred, and perhaps even originated, in the Southern Hemisphere. Once thought to belong to the Mesozoic toothed divers of the order Hesperomi- thiformes, Neogaeornis wetzeli Lambrecht (1929), from the Late Cretaceous of Chile, has been shown to belong to the Gaviidae (Olson, 1992). Another Late Cretaceous fossil, from Seymour Island, Antarctica, also is referable to the Gaviidae (Chatterjee, 1989) and possibly to Neogaeornis. The published fossil record of loons resumes with Colym- boides anglicus Lydekker (1891a), from the late Eocene of En- gland (Lydekker, 1891a; Harrison and Walker, 1976). The ge- nus Colymboides comprises two species that differ considerably from Gavia in many aspects of their osteology (Storer, 1956; Cheneval, 1984), and they are believed to repre- sent a separate lineage (Storer, 1956). The type of the genus is C. minutus Milne-Edwards (1867), a tiny species described from the rich early Miocene (Aquitanian) deposits at St.- Gerand-le-Puy, France. Colymboides minutus also has been identified from the early Miocene of the Dolnice basin in the Czech Republic, where it occurs together with a small species of Gavia (Svec, 1980, 1982), and from the early Miocene Faluns de Saucats in France (Cheneval, 1984). The two speci- mens from Lee Creek Mine referred to Colymboides sp. and presumed to be from the Pungo River Formation thus probably represent the latest occurrence of the genus. The genus Gavia, as represented by G. egeriana Svec (1982), was first known from the early Miocene of the Czech Republic. This was a very small species, which we also have identified from the Calvert and Pungo River formations. The Calvert material differs considerably, in characters that are 248 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY doubtless primitive, from modern species of Gavia. Contempo- raneous with G. egeriana in the Calvert Formation is a second, somewhat smaller, undescribed species. Both of these middle Miocene species are much smaller than any known later spe- cies of Gavia. Loons intermediate in age between the early middle Miocene and the early Pliocene are known from several localities in Eu- rope and are briefly discussed above. Abundant material of Ga- via from the early Pliocene Yorktown Formation clearly en- compasses at least three species, which differ in size. These are all much larger than any loons of the Calvert Formation or de- posits of equivalent or earlier age, and they may represent the three modern lineages, consisting of G. stellata, G. arctica/ pacifica, and G. immer/adamsii. The three extant species-groups of loons are essentially Hol- arctic in distribution, breeding on fresh water at much higher latitudes than North Carolina but wintering almost entirely at sea. All forms except G. stellata have such high wing loadings that their flight would presumably be impaired by any loss of remiges during molt. Because of the necessity of leaving their breeding grounds before the waters freeze, adults do not molt until they reach their wintering grounds at sea, where they molt the remiges simultaneously and undergo a flightless period (Woolfenden, 1967). Gavia fortis, certainly, and G. concinna, probably, were al- ready sufficiently large by the early Pliocene as to have neces- sitated this pattern of molt of the flight feathers, and it seems reasonable to assume that all loons from the Yorktown Forma- tion at Lee Creek Mine were individuals on their wintering grounds. Thus, the basic pattern of loons wintering at sea had evidently already evolved by the early Pliocene. If the three species identified at Lee Creek Mine are really the predecessors of the three modern species groups, then there have been some interesting changes in the wintering dis- tributions of the lineages. Gavia stellata is "a common mi- grant and regular winter resident" off North Carolina (Lee, 1995:119), although it is much scarcer farther south in Geor- gia and Florida. Gavia immer is the common wintering loon in eastern North America today, being abundant in Florida, yet what we have identified as G. fortis is very rare in the Bone Valley deposits. Gavia concinna is by far the most com- mon loon in the Bone Valley deposits and is at least as abun- dantly represented at Lee Creek Mine as is either of the other two species. If this represents the G. arctica/pacifica lineage, then its status has changed dramatically since the Pliocene be- cause G. pacifica is only a casual visitor to the western North Atlantic, where it does not regularly winter. For North Caroli- na there are only a few sight reports and one specimen found just north of the North Carolina/Virginia border (Lee, 1995:119). Order PODICIPEDIFORMES (grebes) Family Podicipedidae (grebes) Genus Podiceps Latham Pliodytes Brodkorb, 1953d:953. Podiceps aff. auritus (Linnaeus) PLATE 5a-g "Fulica sp. (Pisana Nob.)" Portis, 1888:195. "Fulica sp. (pisana Portis)" Portis, 1891:13. Podicepespisanus (Portis).—Regalia, 1902:233. Podiceps pisanus (Portis).—Lambrecht, 1933:262. Pliodytes lanquisti Brodkorb, 1953d:953. Podiceps howardae Storer, p. 227, this volume. Material.—Right coracoid, USNM 177927. Right humer- us, USNM 243764; proximal three-fourths of right humerus, USNM 183430; distal ends of right humeri, USNM 193242, 215034, 407798; distal ends of left humeri, USNM 368557, 430524. Right femora, USNM 215453, 215649, 252314, 460785; left femur, USNM 177918; proximal ends of left fem- ora, USNM 178151, 206413. Proximal ends of right tarsometa- tarsi, USNM 193175, 250773; distal end of right tarsometatar- sus, USNM 206326, distal end of left tarsometatarsus, USNM 210531. Horizon.—Yorktown Formation inferred from identity with specimens in Bone Valley Formation and similarity to modern species. Additional Material Examined.—Bone Valley Forma- tion, Florida: Proximal half of right humerus, USNM 447059. Measurements.—See Storer (p. 227, this volume). Remarks.—Grebes are uncommon among the Lee Creek Mine fossils, almost all of the few specimens appearing to be from a single species that was very similar in size and other characters to the modern Horned Grebe, Podiceps auritus, which is the common species of grebe wintering at sea in North Carolina today. We compared the appropriate specimens from Lee Creek Mine with a cast of the holotype of Podiceps pisanus (Portis, 1888, original in the Museo di Geologia e Paleontologia dell'Universita di Firenze, Italy), which was described from the distal end of a humerus from the Pliocene of Italy, and we could detect no meaningful differences. The species originally was described by Portis (1888) as a coot (Fulica, Rallidae) and at that point was almost a nomen nudum. It was well described and figured by Portis, still as a coot, in a subsequent publica- tion (Portis, 1891). Brodkorb (1963:227) cited the latter as the original description, with the date 1889, but the Zoological Record (1892, volume 28:21) lists the publication for 1891. The supposed extinct genus and species Pliodytes lanquisti Brodkorb (1953d), based on a coracoid from the Bone Valley NUMBER 90 249 Formation in Florida, is in this size range. The length of a cora- coid from Lee Creek Mine (USNM 177927) is exactly the same as that given for the holotype of Pliodytes lanquisti and falls within the range of Podiceps auritus, from which it shows no significant differences. The proximal end of a humerus from Bone Valley (USNM 447059), which presumably is from the same species of grebe as represented by the holotype of Pliodytes lanquisti, is indistinguishable from comparable ele- ments from Lee Creek Mine, and these in turn are identical to Podiceps auritus. Storer (this volume), although recognizing the affinities of the Lee Creek grebe with P. auritus, according to long-accepted practice has emphasized slight differences and has named it as a new species, Podiceps howardae. In ac- cordance with the philosophy outlined in our introduction, we emphasize its similarities to P. auritus. Podiceps sociatus (Navas, 1922), a fossil species of grebe also the size of P. auritus, is known from as far back as the middle Miocene of Spain, although it was more primitive in some respects than the modern species (Olson, 1995). The Horned Grebe is circumpolar in distribution and is one of the commoner grebes in the Northern Hemisphere, especial- ly in marine environments, so it would not be surprising if its antecedents occurred in the same situations in the Pliocene. Podicipedidae, genus and species indeterminate Material.—Left tarsometatarsus lacking distal end and with proximal end badly damaged, USNM 501509. HORIZON.—Uncertain, probably Yorktown Formation. Measurements (mm).—Estimated length, 36.8. Remarks.—This poorly preserved specimen comes from a grebe considerably smaller than Podiceps aff. auritus (above). The tarsometatarsus is relatively shorter and more robust than it is in the Eared Grebe, Podiceps nigricollis Brehm, and per- haps comes from a grebe with less specialized tarsal morpholo- gy, such as the species of Tachybaptus Reichenbach (see Olson, 1995). Order PROCELLARHFORMES (tubenoses) Two of the four families of this order, Diomedeidae and Pro- cellariidae, are abundantly represented at Lee Creek Mine. The diving-petrels (Pelecanoididae) are known only from the Southern Hemisphere and would not be expected, whereas the absence of storm-petrels (Oceanitidae=Hydrobatidae auct.) is almost certainly an artifact of collection and taphonomy. Al- though storm-petrels are common in the same area today and would doubtless have been so in the Pliocene, they are the smallest members of the order and feed entirely from the sur- face of the water; thus, they would be less likely to fall prey to submarine predators. This is probably the main factor contrib- uting to their absence at Lee Creek Mine, although the small size of their bones also would make them less likely to be spot- ted by collectors. The only Tertiary locality where fossils of storm-petrels have been found in numbers is a coastal site in South Africa thought to be in the immediate vicinity of insular breeding colonies (Olson, 1985b), whereas in nearby pelagic deposits storm-petrels were all but absent (Olson, 1985c). Family DIOMEDEIDAE (albatrosses) Albatrosses now occur largely in the southern oceans, with only three modern species found in the North Pacific and none found in the North Atlantic, except as very rare vagrants. Thus, it is of considerable interest that over 500 specimens represent- ing five species of albatross have been recovered from Lee Creek Mine. The only albatross known from the Calvert For- mation is a rare, very small species, even smaller than the smallest one known from Lee Creek Mine. Because of this and the similarity of the Lee Creek birds to modern species, and as indicated by microfossil analysis of the matrix associated with some of the fossils, all of the Lee Creek albatross material is re- garded as being from the Yorktown Formation. In attempting to identify the Lee Creek fossils and in making comparisons with modern taxa, we found that all the North Pa- cific albatrosses (Short-tailed Albatross, Phoebastria albatrus (Pallas); Black-footed Albatross, P. nigripes (Audubon); Lay- san Albatross, P. immutabilis (Rothschild)) differ from all the smaller Southern Hemisphere albatrosses (mollymawks, Thalassarche Reichenbach; sooty albatrosses, Phoebastria Re- ichenbach) in having the tarsometatarsus proportionately long- er and more slender (Plate 8). In this respect the North Pacific albatrosses were more similar to the "great" albatross group, which consists of the Wandering Albatross, Diomedea exulans Linnaeus; Amsterdam Albatross, D. amsterdamensis Roux et al.; and Royal Albatross, D. epomophora Lesson, than to the mollymawks or the sooty albatrosses. These osteological observations have been corroborated by DNA sequences in which four major groups of albatrosses were recognized (Nunn et al., 1996): the North Pacific species just mentioned (plus the Waved Albatross, D. leptorhyncha Coues (=D. irrorata Salvin auct.)), which take the name Phoe- bastria; the great albatrosses, genus Diomedea Linnaeus, which forms the sister group of Phoebastria; the mollymawks, genus Thalassarche; and the sister-group of the latter genus, Phoebastria. We have followed this classification herein. Genus Phoebastria Reichenbach Phoebastria anglica (Lydekker, 1891a), new combination Plates da, 7e,g,i,l Diomedea anglica Lydekker, 1891 a: 189. IDiomedea californica L. Miller, 1962:471. Diomedea sp. A, Chandler, 1990a: 100. Material.—Fragment of ramus of furcula, USNM 206434. Proximal ends of right scapulae, USNM 192688, 430529, 250 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY 464253. Proximal end of right humerus, USNM 242324; distal end of right humerus, USNM 366454; distal end of left humer- us, USNM 183511. Proximal end of left ulna, USNM 430602; distal ends of right ulnae, USNM 193160, 252321; distal end of left ulna, USNM 430528. Associated(?) distal ends of left ulna and radius, USNM 215748, 215751. Proximal ends of right carpometacarpi, USNM 193170, 321242, 366628; proximal end of left carpometacarpus, USNM 430590; distal ends of right carpometacarpi, USNM 302354, 366922, 464254. Major alar digit phalanx 1, USNM 430526. Left femur, USNM 210458; proximal end of right femur, USNM 368546; distal ends of right femora, USNM 181085, 308201, 366689, 366906, 430530; distal ends of left femora, USNM 464255, 464256. Distal ends of right tibiotarsi, USNM 206486, 206638, 241429, 248495, 464257; distal ends of left tibiotarsi, USNM 181045, 248533, 321230, 430531, 430532, 464258. Right tar- sometatarsi, USNM 250732, 430538; left tarsometatarsi, USNM 250777, 430533; proximal ends of left tarsometatarsi USNM 250739, 366627, 368543, 430622, 464259, 464260; shafts of right tarsometatarsi, USNM 192809, 215478, 241431; shaft of left tarsometatarsus, USNM 256257; distal ends of right tarsometatarsi, USNM 193142, 257457, 366716, 430539, 464261-464263; distal ends of left tarsometatarsi, USNM 215884, 250743, 250758, 252303, 256250, 275850, 321236, 430534, 430535, 430537, 430619, 464264, 464265. Pedal pha- langes, USNM 192469, 257482, 366380, 464266-464274. Horizon.—Yorktown Formation (USNM 181085 from bas- al Yorktown as determined from foraminifera in matrix). Additional Material Examined.—Red Crag Formation, England: Right tarsometatarsus, USNM 215038 (cast of ho- lotype). Coralline Crag Formation, England: Proximal two-thirds of right ulna, USNM 215040 (cast of paratype). Bone Valley Formation, Florida: Distal end of left ulna, UF 123829. Right carpometacarpus, UF 65765. Distal end of right tibiotarsus (cast), USNM 16751. Left tarsometatarsus, UF 53942; distal end of left tarsometatarsus, UF 57309. San Diego Formation, California: Right tarsometatarsus, SDSNH 27872. Measurements.—See Table 3. REMARKS.—Diomedea anglica was originally described by Lydekker (1891a) from an associated tarsometatarsus and pha- lanx 1, supposedly of pedal digit IV, from the late Pliocene Red Crag at Foxhall, Suffolk, England. The phalanx, however, is actually that of digit II, the phalanx for digit IV in albatrosses being proportionately much longer and more slender. The proximal end of an ulna from the underlying Coralline Crag in the same vicinity (Lydekker, 189lb:395) was soon after con- sidered to belong to the same species. Both these records are now considered to be late Pliocene (MN 16-17) in age (Mlik- ovsky, 1996:766). Later, Wetmore (1943a) referred the distal end of a tibiotarsus from Bone Valley, Florida, to D. anglica. Lydekker (1891a) characterized the tarsometatarsus of Di- omedea anglica as being somewhat smaller and proportionate- ly more slender than that of D. exulans. Harrison and Walker (1978) concluded that D. anglica was a valid species that seemed most similar to D. albatrus but was larger. An albatross the size of D. anglica is one of the two most common albatross species at Lee Creek Mine. We examined additional material from the Bone Valley Formation in Florida that falls in the same size class, as well as a specimen from the San Diego For- mation in California first reported by Chandler (1990a: 100) as an unidentified species of Diomedea. We refer all of this mate- rial to Lydekker's species Diomedea anglica, but under the ge- nus Phoebastria, as explained below. The species known as Diomedea californica L. Miller, 1962, may be the Miocene representative of this same lineage. It was originally described from the distal end of a tarsometatarsus from the middle Miocene of Sharktooth Hill (Miller, 1962), with the distal end of a humerus and another tarsometatarsus from the same locality being referred later (Howard, 1966, 1978). A tibiotarsus from the late Miocene at Laguna Niguel, Orange County, also was referred, with a query, to D. californi- ca (Howard, 1978). The distal width of the holotypical tar- sometatarsus was 20.6 mm, and that of the referred specimen was about 21.5 mm (as extrapolated from the percentages giv- en by Howard, 1978), which is within or very near the range for the series of tarsometatarsi from Lee Creek Mine assigned to Phoebastria anglica (18.7-21.2 mm, «=21). The humerus referred to D. californica had a distal width of 27.5 mm, which compares very well with three assigned to P anglica from Lee Creek Mine (27.3, 28.5, 30.2 mm). There are few living albatrosses the size of Phoebastria an- glica, which was larger than all known species except the great albatrosses. The largest living species are the Wandering and Royal albatrosses (Diomedea exulans and D. epomophora). Based on specimens available to us, these have larger and much more robust tarsometatarsi than does P. anglica (Table 3). Our series of both modern species, however, was quite inad- equate because there are several recognized subspecies in this complex, some of which differ in size, so there may be more overlap than was apparent in our comparisons. Another enigmatic member of the great albatross group is the Amsterdam Albatross, Diomedea amsterdamensis (Plates 6b, Ifh.m), known from a small remnant population on Amster- dam Island in the southern Indian Ocean. This was named as recently as 1983 (Roux et al., 1983), and although photographs of living specimens were published with the original "descrip- tion," the so-called "holotype" was evidently a composite as- sortment of subfossil bones, perhaps belonging to several dif- ferent individuals. It was not stated what elements of the skeleton were included in the "holotype," nor were any mea- surements or comparisons made of these bones, the character- ization of the species thus being utterly inadequate. The taxon is sometimes considered to be a subspecies of D. exulans (e.g., Warham, 1990:424). We were able to examine a small composite assortment of bones of Diomedea amsterdamensis (USNM 560597), which NUMBER 90 251 shows this species to be smaller than any of the available spec- imens of D. exulans, although it is about the size of the fossils we have referred to Phoebastria anglica (Table 3). Although P. anglica possibly falls within the lower size ranges of the great albatrosses of the restricted genus Diomedea, there is no other indication of now-exclusively Southern Hemisphere albatross- es in the Northern Hemisphere. Thus, it appears much more likely that anglica is a very large member of the Northern Hemisphere albatrosses of the genus Phoebastria, to which we refer it. As such, it may be regarded as an extinct lineage with no living descendents. Phoebastria aff. albatrus (Pallas) Plates 6f,k, 7a,j, Sa Diomedea howardae Chandler, 1990a:96. Material.—Portion of ramus of furcula, USNM 430609. Fragment of shaft of right coracoid, USNM 206548. Anterior end of right scapula, USNM 193298. Proximal end of left hu- merus, USNM 460858; distal ends of right humeri, USNM 242347, 256249, 430607; distal end of left humerus, USNM 430606. Proximal ends of left ulnae, USNM 179230, 181039; distal ends of right ulnae, USNM 430598, 430599, 430601; distal ends of left ulnae, USNM 206623, 430603, 430605. Proximal ends of left radii, USNM 181068, 250825; distal ends of right radii, USNM 193149, 241368; distal ends of left radii, USNM 215563, 244300, 464290. Left carpometacarpus lack- ing minor metacarpal, USNM 430588; proximal ends of right carpometacarpi, USNM 192023, 206499, 256211, 275854, 430593; proximal end of left carpometacarpus, USNM 430589; distal end of right carpometacarpus, USNM 193319; distal end of left carpometacarpus, USNM 430591. Major alar digit pha- lanx 1, USNM 193378. Left femur, USNM 192945; distal ends of left femora, USNM 177921, 366925, 430610. Proximal end of right tibiotarsus, USNM 308215; distal ends of right tibio- tarsi, USNM 241361, 308249; distal ends of left tibiotarsi, USNM 275849, 430614, 430616. Right tarsometatarsi, USNM 193223, 275847, 430629; left tarsometatarsi, USNM 181095, 430618; proximal end of right tarsometatarsus, USNM 368545; proximal end of left tarsometatarsus, USNM 178191; distal end of right tarsometatarsus, USNM 464245; distal ends of left tar- sometatarsi, USNM 192858, 192876,430626. Pedal phalanges, USNM 464275^164280. Horizon.—Yorktown Formation. Additional Material Examined.—Bone Valley Forma- tion, Florida: Distal end of right ulna, UF 53915. Left tar- sometatarsus, UF 94549. San Diego Formation, California: Distal half of left car- pometacarpus, SDSNH 25244. Right major alar digit phalanx 1, SDSNH 25243. Right tarsometatarsus, SDSNH 25245 (holo- type of Diomedea howardae). Measurements.—See Table 3. Remarks.—The Short-tailed Albatross, Phoebastria al- batrus (Plates 6l-o, lb), was probably the common inshore al- Table 3.—Length (mm) of the tarsometatarsus in large living and fossil alba- trosses Diomedea and Phoebastria. (n=number of specimens.) Species n Range Mean D. epomophora D. exulans subsp. D. amsterdamensis 2 5 2 125.0-128.1 113.2-125.8 106.9-111.6 126.5 121.3 109.2 P. anglica (holotype) P. anglica, Lee Creek Mine 1 4 110.3—117+ 110.9 113.5 P. anglica, Bone Valley P. anglica, San Diego Formation P. albatrus I 1 25 89.5-104.0 118.5 112.8 98.4 P. aff. albatrus, Lee Creek Mine P. aff. albatrus, Bone Valley P. albatrus, San Diego formation (holotype of D. howardae) 5 1 1 97.7-103+ 100.9 105.2 101.7 batross in the North Pacific in pre-human times. Hunting and fishing cultures around the northern Pacific rim took a heavy toll on the species, at least on the wintering grounds. In the nineteenth and twentieth centuries the combination of Japanese feather hunters and volcanic eruptions nearly exterminated the species from its remaining breeding grounds in the Volcano Is- lands, south of Japan, and the population of this species is still perilously low but has been increasing. The commonest albatross at Lee Creek Mine is similar to P. albatrus in size, being larger than any of the Northern Hemi- sphere species except P. anglica. We have identified speci- mens from Bone Valley, Florida, as belonging to the same spe- cies. Further evidence of the presence of this lineage in the North Atlantic comes from a mid-Pleistocene deposit on Ber- muda, where there was clearly a breeding colony of P. al- batrus, as numerous bones were found of non-volant juveniles, and in some cases even of embryos (Olson, unpublished data). A rapid and very high sea-level stand about 450,000 years ago (Hearty et al., 1999) probably caused the extinction of this spe- cies on Bermuda and in the entire North Atlantic. The Lee Creek Mine material shows some differences from P. albatrus, such as in the shape and orientation of the internal tuberosity of the humerus and the less proximally oriented head of the femur. These may be only temporal differences in the same species lineage. The species described as Diomedea howardae from a tar- sometatarsus from the late Pliocene San Diego Formation in California (Chandler, 1990a) is the size of Phoebastria al- batrus (Table 3), but it was not compared with that species in the original description. It was diagnosed on supposed differ- ences in the shape of the distal foramen that do not hold up in a series of modern P. albatrus. We consider Diomedea howard- ae Chandler, 1990a, to be synonymous with Phoebastria al- batrus (Pallas, 1769). That albatrosses of this size have occurred in the Pacific well before the early Pliocene is shown by a tarsometatarsus from the middle Miocene Astoria Formation in Oregon (USNM 424081, Plate 6g) that has a decidedly modem aspect. 252 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Phoebastria aff. nigripes (Audubon) Plate %d Material.—Distal ends of right humeri, USNM 430608, 464243; distal end of left humerus, USNM 366898. Distal ends of right tibiotarsi, USNM 183457, 248572, 275798; distal ends of left tibiotarsi, USNM 177900, 192950,430613, 430615. HORIZON.—Yorktown Formation. Measurements.—See Tables 4, 5. Remarks.—The Black-footed Albatross, Phoebastria ni- gripes (Plates lk, 86,/), and the Laysan Albatross, P. immuta- bilis (Plates 6d, 8c,g), are medium-sized species confined today to the North Pacific, where their main breeding grounds are in the northwestern Hawaiian Islands. Despite their considerable differences in plumage, the two species hybridize with some regularity. Although P. nigripes is definitely the larger species, there is overlap in the measurements of some of the elements (e.g., see Table 4) so that many of the fossils in this general size range from Lee Creek Mine probably cannot be identified except as belonging to one or the other of these two species. We have not undertaken a comprehensive analysis of this material but have attempted only sufficient comparisons to establish that both species are indeed present in the fauna. Among eight distal ends of humeri in this size class recov- ered at Lee Creek Mine (Table 4) are three that are larger than the largest available specimen of P. immutabilis and hence are assigned to P. aff. nigripes. Three others are in the area of overlap between the two species and could belong to either. The situation with distal ends of tibiotarsi was more com- plex. In our comparative series, there was essentially no over- lap in this measurement between the two species, P. nigripes being larger (Table 5). All of the fossil specimens that had originally been assigned to this size class, however, fell within the range of P immutabilis. We then found that some of the specimens that had been assigned to the P. albatrus size class were too small for that species. The smallest of six modern specimens of P. albatrus, a juvenile, had the distal width of the tibiotarsus 16.9 mm, which is barely larger than the largest P. nigripes, but the shaft in this specimen is markedly more ro- bust. Based on measurements of distal width and visual com- parison of the robustness of the shaft, we assign the tibiotarsi listed above to P. aff. nigripes. Although more material than we have identified doubtless belongs to this species, P. aff. nigripes is apparently less com- mon in the Lee Creek Mine deposits than is P. aff. immutabi- lis. We also examined the distal end of a left ulna from the Bone Valley Formation in Florida (UF 95654) that belongs in the P. nigripes/P. immutabilis size range and which appears to be the first record of an albatross other than P. anglica in those deposits. Phoebastria aff. immutabilis (Rothschild) Plates 6c. 7c Material.—Distal end of right humerus, USNM 367115; distal end of left humerus, USNM 430677. Right carpometac- arpus lacking minor metacarpal, USNM 460841. Distal ends of right tibiotarsi, USNM 193387, 302293, 321274, 366669, 430686-430688, 460864, 460865; distal ends of left tibiotarsi, USNM 183440, 430680, 430681, 430683, 430684, 460847, 464247, 464248. Right tarsometatarsus, USNM 464250; left tarsometatarsus, USNM 430689. Horizon.—Yorktown Formation. Measurements (mm).—Carpometacarpus: Length, 96.9 (within range of Phoebastria immutabilis and smaller than any of four females of P. nigripes (100.0-103.6); see also Tables 4, 5). REMARKS.—As discussed under the preceding species, sev- eral distal ends of tibiotarsi and two distal ends of humeri from Lee Creek Mine are too small to belong to Phoebastria ni- gripes and fall within the range of variation of P. immutabilis. In addition, there are two complete tarsometatarsi with lengths of 87.5 and 87.8 mm, values well within the range of females of P. immutabilis (Table 6), from which they are inseparable. Phoebastria aff. nigripes or P. aff. immutabilis The following specimens from Lee Creek belong to one or the other of these two species but have not been further identi- fied. Proximal, shaft, and distal fragments of left humerus, USNM 193231; distal ends of right humeri, USNM 193250, 430679; distal ends of left humeri, USNM 430678. Proximal ends of right ulnae, USNM 366663, 366717; proximal ends of left ul- nae, USNM 210448, 242315, 430656, 430657; distal ends of right ulnae, USNM 181057, 192487, 206350, 215743, 366417, 366648, 430658-430663, 430666, 460840, 460859; distal ends of left ulnae, USNM 181075, 193388, 206475, 366803, 366939, 367000, 368466, 368542, 430664, 430665, 430667^130669, 430671, 430672. Proximal end of left radius, USNM 366435; distal ends of left radii, USNM 464288, 464289. Proximal portions of right carpometacarpi, USNM 430654, 460854; proximal halves of left carpometacarpi, USNM 430647^130649, 460855, 460856; distal ends of right carpometacarpi, USNM 206566, 275799, 368554, 430592, 460857; distal ends of left carpometacarpi, USNM 178090, 256237, 430655. First phalanges of major alar digit USNM 206557, 460846, 460862, 460863, 464282, 464283. Right fe- mur, USNM 460860; proximal ends of right femora, USNM 321302, 430674; proximal end of left femur, USNM 215479; distal ends of right femora, USNM 308223, 430675, 430676, 460861; distal ends of left femora, USNM 321312, 430673. Proximal end of right tibiotarsus, USNM 430685. Proximal and distal ends of right tarsometatarsus, USNM 464251; proxi- mal and distal ends of left tarsometatarsus, USNM 430691; proximal ends of right tarsometatarsi, USNM 275816, 366346, NUMBER 90 253 430694,430695, 430707; proximal ends of left tarsometatarsi, USNM 192654, 241372, 252306, 366347, 430690, 464252; distal ends of right tarsometatarsi, USNM 181110, 192705, 275823, 302317, 430696, 430697, 460851, 460852, 464249; distal ends of left tarsometatarsi, USNM 248542, 257500, 302292, 430693. Pedal phalanges, USNM 177819, 248577, 321229,464284^164287. Phoebastria rexsularum, new species Plates 6h, i, 8e HOLOTYPE.—Right tarsometatarsus, USNM 302313. Type Locality.—Texasgulf Inc. Lee Creek Mine, south side of Pamlico River, near Aurora, Beaufort County, North Carolina (35°23'22"N, 76°47'06"W). Horizon.—Yorktown Formation, lower Pliocene. Distribution.—Known so far only from the type locality and from the Rappahannock River in Middlesex County, Vir- ginia. Measurements of Holotype (mm).—Length, 73.9; proxi- mal width, 13.7; width and depth of shaft at midpoint, 5.5 x 6.0; distal width, 13.8; depth of middle trochlea, 8.4; width through outer and middle trochleae, 9.7. TOPOTYPICAL Paratypes.—Distal end of left humerus, USNM 302414. Distal ends of right tibiotarsi, USNM 250848, 275795, 308185; distal ends of left tibiotarsi, USNM 178050, 430682,430701, 430702,464291. Right tarsometatarsi, USNM 430704, 430705; proximal ends of right tarsometatarsi, USNM 242334, 366607, 430708, 460848, 460849; proximal end of left tarsometatarsus, USNM 460850; distal ends of right tarsometa- tarsi, USNM 366611,460838; distal ends of left tarsometatarsi, USNM 430703, 460853. Pedal phalanx, USNM 464281. The following specimens are tentatively referred to this spe- cies, most being very fragmentary or undiagnostic, although they appear to be too small for Phoebastria aff. immutabilis. Shaft of right coracoid, USNM 430700. Three fragments of proximal end of left humerus, USNM 430699. Distal end of right ulna, USNM 430670; distal end of left ulna, USNM 460839. Proximal ends of right carpometacarpi, USNM 178066,430653; proximal ends of left carpometacarpi, USNM 460842, 460843; distal ends of right carpometacarpi, USNM 250702, 460844, 460845; distal end of left carpometacarpus, USNM 257494. Additional Paratype.—Left tarsometatarsus lacking proximal end, USNM 256620, south side of Rappahannock River, 3 mi (5 km) upstream from Stingray Point, near Deltaville, Middlesex County, Virginia; collected by Eldon Branch, received in 1979. The specimen was found as a "float" and is presumed to be from the Yorktown Formation, which crops out there. Measurements of Paratypes (mm).—Tarsometatarsus (USNM 256620): Distal width, 14.1. For measurements of other paratypes see Table 6. Etymology.—"King of the gannets," from Latin rex, king, and the genitive plural of Sula, now used as the generic name Table 4.—Comparison of distal width (mm) of humerus of modern and fossil medium-sized albatrosses, Phoebastria. Measurements are in list form for fos- sil species. («=number of specimens.) Species P. immutabilis females P. immutabilis males P. aff. immutabilis, Lee Creek Mine P. aff. immutabilis/nigripes, Lee Creek Mine P. aff. nigripes, Lee Creek Mine P. nigripes females P. nigripes males Range Mean 9 20.9-22.2 21.5 9 21.9-23.5 22.9 2 21.6,21.8 - 3 22.6, 22.9, 23.0 - 3 23.7,23.8,24.0 - 4 22.6-24.0 23.4 5 23.1-24.3 23.8 Table 5.—Comparison of distal width (mm) of tibiotarsus of modern and fos- sil medium-sized albatrosses, Phoebastria. (n=number of specimens.) Species n Range Mean P. immutabilis females 6 12.1-13.2 12.6 P. immutabilis males 8 14.4-15.0 14.7 P. aff. immutabilis, Lee Creek Mine 17 13.0-14.8 14.0 P. aff. nigripes, Lee Creek Mine 5 14.9+-I6.5 15.8 P. nigripes females 4 14.9-15.4 15.2 P. nigripes males 5 15.1-16.7 15.6 Table 6.—Measurements (mm) of hindlimb elements of Phoebastria rexsu- larum, new species, compared with females (the smaller sex) of modern P. im- mutabilis. Element P. rexsularum, n sp. P. immutabilis, females n Range Mean n Range Mean Tibiotarsus Distal width 6 12.1-13.2 12.6 9 13.3-14.1 13.9 Tarsometatarsus Length 3 73.8-82+ 78.5 9 83.7-88.5 86.1 Proximal width 9 12.5-14.2 13.5 9 14.8-15.8 15.4 Distal width 6 12.9-14.2 13.6 9 14.4-15.8 15.0 of boobies but derived from the Scandinavian name applied to the Northern Gannet, Morus bassanus. The name comes from the extraordinary example of a single female of the Southern Hemisphere Black-browed Albatross, Thalassarche mel- anophris (Temminck), that appeared in a gannetry on Myggen- aes Holm in the Faeroe Islands in 1860 and returned each sum- mer for 34 years until shot in 1894 (Andersen, 1894, 1895; Murphy, 1936:511). During this time it was known among the Faeroese who visited the gannetry as "Sulekongen," or "Gan- net King," as the gannets would make way for the albatross when it moved about the colony. DIAGNOSIS.—This species is referable to Phoebastria by the slender configuration of the tarsometatarsus; it is without ex- panded articular surfaces seen in Thalassarche, but it is smaller than any other member of the genus and is probably smaller than any existing species of albatross. Remarks.—In sorting the albatross material from Lee Creek Mine, it became evident that some of the specimens at the small end of the observed size variation were too small to be encompassed by the range of variation seen in the size class in- cluding P. nigripes and P. immutabilis. Although in life there may have been overlap in size between this smallest species 254 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY and P. aff. immutabilis, we have assigned fossil leg elements to the new species only when they are smaller than observed in a series of females of modern P. immutabilis. The holotype of P. rexsularum is from a particularly small individual, being a full centimeter shorter than the shortest available tarsometatarsus of P. immutabilis. The distal width of the humerus assigned to P. rexsularum is 20.7 mm, which is smaller than that in any available specimen of P. immutabilis or P. nigripes (Table 4). The pedal phalanx referred to P. rexsularum is the diagnostic long, slender proxi- mal phalanx of the fourth toe, which measures 43.9 mm in length. The same element in a small female of P. immutabilis was 51.8 mm long, again illustrating how small some individu- als of P. rexsularum must have been. Howard (1966) described a new species of rather small alba- tross, Diomedea milleri, from the middle Miocene deposits of Sharktooth Hill, California. The holotype is the most proximal portion of a left ulna and was compared only with a single modern specimen of D. (=Phoebastria) nigripes. The proximal end of a tarsometatarsus from the same locality was referred to D. milleri in the same publication. Howard and Barnes (1987) later tentatively referred another proximal end of an ulna with associated fragments of radii from middle Miocene deposits at Oso Creek, Orange County, California, to D. milleri. The prox- imal widths of the holotype and of the Oso Creek specimen were 14.4 and 15.8 mm, respectively, whereas in a small fe- male of Phoebastria immutabilis (USNM 488178) the same measurement is 12.7 mm. No measurements were given for the referred tarsometatarsus, but if the illustration (Howard, 1966, fig. le) is at natural size, as stated, the proximal width would be about 18 mm, which is considerably larger than it is in any female of P. immutabilis measured (Table 6). The material of D. milleri is probably inadequate for characterization of a new species of albatross, but the available evidence suggests that it was not smaller than P. immutabilis. Therefore there is no rea- son to identify it with P. rexsularum. Chandler (1990a: 100) mentioned the distal end of a tibiotar- sus from the San Diego Formation as belonging to a small alba- tross ("Diomedea sp. B") that he said was "86 percent that of D. nigripes in distal width." Because no measurements were given for either species, there is no way to assess the actual size of the specimen from his publication. Phoebastria rexsularum represents the least-common size class of albatross at Lee Creek Mine. Curiously, this smallest of the Northern Hemisphere albatrosses, and D. anglica, the largest, were the ones that became extinct, whereas the three species of intermediate size have persisted elsewhere up to the present. Family Procellariidae (shearwaters and petrels) By far the majority of fossils of this family from Lee Creek Mine are referable to the genus Puffinus Brisson and the close- ly related genus Calonectris Mathews and Iredale. The identifi- cation of these fossils was extremely difficult for reasons given below. To begin with, many fossil species of Puffinus have been described and named from deposits of various ages on both sides of the Atlantic and from the eastern Pacific. A prop- er treatment of the Lee Creek fossils would have to take these fossil taxa into account, necessitating a great deal of revision- ary work. Another problem with the Lee Creek Mine material in partic- ular is that species lineages in the Procellariidae appear to have changed very slowly, if at all, through time. Species of Puffinus in the Calvert Formation, for example, may be extremely simi- lar to living species. Thus, when precise stratigraphic informa- tion is absent it is impossible to determine whether a given form of Puffinus from Lee Creek Mine is middle Miocene or early Pliocene in age. Shearwaters of the genera Puffinus and Calonectris were treated in an admirable monograph by Kuroda (1954), who ex- amined osteology in addition to external characters. He was careful to attempt to distinguish between primitive and special- ized characters well before the advent of cladistics. In summary, the species of shearwaters are marked by a pro- gression from a primitive, aerially adapted condition (Calonec- tris) to increasing use of both the wings and feet for underwater propulsion—what Kuroda referred to as "aquatic" adapta- tions—in which the humerus becomes flattened, the forewing shortened, the pelvis laterally compressed and lengthened, the femur stouter and more curved, the cnemial crest of the tibio- tarsus lengthened, and the tarsometatarsus more laterally com- pressed—all of these being typical diving adaptations also found in other groups of birds. The groups consisting of the Wedge-tailed Shearwater (Puffinus pacificus (Gmelin)) and Buller's Shearwater (P. bulleri Salvin, subgenus Thyellodroma Stejneger), and the Pink-footed Shearwater (P. creatopus Coues) and Flesh-footed Shearwater (P. carneipes Gould, sub- genus Hemipuffinus Iredale), are only slightly more specialized along these lines than is Calonectris and are hardly separable from one another except on size. The Greater Shearwater, Puffinus gravis (O'Reilly) (subgenus Ardenna Reichenbach), occupies an intermediate position between those species and the most specialized members of the family, which are of the subgenus Puffinus. Among the last, the Short-tailed Shearwa- ter, P. tenuirostris (Temminck), appears to be the least derived. Despite the distinct osteological differences between Ca- lonectris and the different subgroups of Puffinus, the nature of the fossils from Lee Creek Mine still renders them difficult to identify. There is practically no associated material. Bones of the wing are much more frequently represented than are ele- ments of the hindlimb, and these are almost always broken. Whereas complete humeri of Puffinus would be relatively easy to assign to one subgroup or another, the differences become greatly blurred when hundreds of fragmentary distal ends with varying degrees of wear are compared. Therefore, we have not attempted to identify many of the specimens of Procellariidae NUMBER 90 255 to species. Instead we have selected only a few specimens of some of the more diagnostic elements in order to attempt to as- sess the minimum number of species that are present in the fau- nas. Further refinements will have to await much patient revi- sionary analysis. The great majority of the procellariid fauna at Lee Creek Mine is apparently made up of the five medium- large species Calonectris aff. diomedea, C. aff. borealis, Puffi- nus aff. gravis, Puffinus (Ardenna) sp., and P. aff. pacificoides. Below, we identify 16 species of Procellariidae from the Lee Creek deposits. This is undoubtedly a minimum and others will almost certainly be distinguished among the specimens already at hand. Genus Pterodromoides Segui et al. (in press) Pterodromoides minoricensis Segui et al. (in press) Plate 9o Material.—Distal end of left humerus lacking ectepi- condylar spur, USNM 464315. HORIZON.—Yorktown Formation (see below). Measurements (mm).—Distal width, 8.9; distal depth through ulnar condyle, 5.8; width and depth of shaft 20 mm above distal extremity, 4.4 x 3.1. REMARKS.—This bone seemingly cannot be assigned to any living genus in the family. There is practically no expansion of shaft above the entepicondyle, even to the rather limited extent seen in Pterodroma Bonaparte and quite unlike Puffinus, and the brachial depression is large and deep. The specimen ap- pears to be closest to that of the so-called "fulmarine" petrels, exemplified by Fulmarus Stephens, but it is smaller than any of those birds. It is of a rather unusual size class for the family, being intermediate in size between the largest of the smaller species of Pterodroma and the smallest of the larger species of that genus. We had originally listed this specimen only as "Procellari- idae, genus and species indeterminate," the preceding para- graph having been written before we thought to compare the specimen with Pterodromoides minoricensis Segui et al. (in press). This is a new genus and species described from Mio/ Pliocene deposits on Menorca in the Balearic Islands of the Mediterranean, which was shown to have similarities to the Fulmarinae. Olson examined and compared some of the type material and agreed that it could not be referred to any known genus of Procellariidae. The distal width of the humerus of P. minoricensis was given as 9.2, 9.3, and 9.6 mm, and the slight- ly smaller size of the Lee Creek fossil (8.9 mm) would proba- bly fall within the range of a larger sample. The fact that, as re- marked above, this is an unusual size class within the Procellariidae makes the assignment of the Lee Creek fossil all the more likely. The deposits on Menorca from which P. minoricensis was obtained could not be dated directly but were reasoned from faunal evidence to be Mio/Pliocene, but younger than Langhian (middle Miocene), and they were inferred to be late Miocene. The shared presence of this species in Menorca and North Carolina suggests that the Lee Creek Mine specimen came from the Yorktown Formation. The temporal range of such a seabird could surely encompass both the late Miocene and the early Pliocene. This trans-Atlantic correlation suggests that the wintering grounds of P. minoricensis included the western North Atlan- tic. That it is so rare at Lee Creek Mine may be due to tapho- nomic processes discussed above relating to its having been a surface feeder. Genus Procellaria Linnaeus This genus, including Adamastor Bonaparte, is now used for the four largest species of the family, apart from the albatross- sized giant petrels Macronectes Richmond. One of the species of petrels at Lee Creek Mine is of this size class, but it appears to be more closely related to Calonectris and is listed under that genus. The living species of Procellaria are practically confined to southern oceans, only two of them being found in the South Atlantic. For our comparisons we had only a single skeleton of each of the species except the White-chinned Pe- trel, P. aequinoctialis, of which we had 13. Procellaria cf. aequinoctialis Linnaeus Plate 9y Material.—Distal end of right tibiotarsus, USNM 464312. HORIZON.—Uncertain, probably Yorktown Formation. Measurements (mm).—Distal width, 9.7; depth through inner condyle, 9.7; width of shaft 30 mm above distal extremi- ty, 5.7. Remarks.—This bone is from a very large procellariid the size of males of Procellaria aequinoctialis, there apparently being no other species of the family that falls in this size range or any other procellariid fossils from Lee Creek Mine of such great size. At present, the breeding stations of P aequinoctialis in the Atlantic are at Inaccessible Island in the Tristan da Cun- ha group, the Falklands, South Georgia, and possibly Gough Is- land. The normal modern range at sea does not extend north much beyond 15°S, so the Lee Creek bird may have been only a vagrant. Procellaria cf. parkinsoni Gray Plate 9q.r Material.—Proximal half of right humerus lacking most of pectoral crest and internal tuberosity, USNM 430845; distal third of right humerus, USNM 430726. HORIZON.—Uncertain, probably Yorktown Formation. Measurements (mm).—Greatest depth through head, 5.8; width and depth of shaft below pectoral crest, 7.6 x 6.1; distal width, 14.4; depth through radial condyle, 8.2 mm. 256 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY REMARKS.—The proximal portion of the humerus appears to be nearly identical with that of Parkinson's Petrel, Procellaria parkinsoni, not only in size and details of attachments, but also in the shape of the shaft, which is noticeably flattened on the anconal surface. The distal portion is an equally good match. The smallest species in its genus, P. parkinsoni breeds only in New Zealand but crosses the Pacific to winter off western America as far north as the latitude of Guatemala. Evidently it or a related form occurred in the Atlantic in the early Pliocene, possibly having been a vagrant entering through the Panamani- an seaway. Genus Pterodroma Bonaparte Gadfly petrels of the genus Pterodroma are all but absent in the fossil record except at their nesting islands. Their rarity in marine deposits may be due to their feeding entirely from the surface, thus making them much less susceptible to predation by aquatic carnivores. Measurements (mm).—Width and depth of shaft at ap- proximate midpoint, 3.7 x 3.0, 3.9 x 3.2. Remarks.—The relatively terete shafts of these fragmentary specimens remove them from association with species of Puffi- nus. They indicate a species larger than Bulwer's Petrel, Bulw- eria bulwerii (Jardine and Selby), and about the size of the smallest of the forms of Pterodroma or of Jouanin's Petrel, Bulweria fallax Jouanin. The only species in this size range re- ported from the Atlantic Ocean is B. bifax Olson (1975), known only from Quaternary deposits on St. Helena Island (Olson, 1975), although the Indian Ocean species B. fallax has been recorded as a vagrant in Italy (Olson, 1985a). Genus Pachyptila Illiger The prions of the genus Pachyptila are small, filter-feeding petrels that are now entirely confined to the southern oceans. Thus, the following record was quite unexpected. Pterodroma magn. lessonii (Garnot) Plate 9/7" Material.—Left tarsometatarsus lacking only the inner tro- chlea, USNM 430854. HORIZON.—Uncertain, probably Yorktown Formation. Measurements (mm).—Length, 45.0; proximal width, 7.9; width through outer and middle trochleae, 6.0; width and depth of shaft at midpoint, 3.9 x 3.7. Remarks.—The lack of either lateral compression or asym- metry of the shaft distinguishes this bone from any of the spe- cies of Puffinus. It is too small for any of the species of Procel- laria or Fulmarus, and it is compatible with the morphology in Pterodroma, within which, however, it is very large, being the size of the White-headed Petrel, P. lessonii. We did not have appropriate comparative skeletal material for this species, but the fossil matched very well in size with the tarsus as visible in skin specimens. Pterodroma lessonii is very much a southern species today, being found circumpolarly from Antarctica to about 33°S latitude. Gadfly petrels the size of those resident in the western North Atlantic, the Black-capped Petrel (P. hasita- ta (Kuhl)) and the Cahow (P. cahow Nichols and Mowbray), have not been found at Lee Creek Mine. Genus Bulweria Bonaparte Bulweria? sp. Material.—Most of shaft of left humerus with distalmost portion of pectoral crest, USNM 464298; proximal two-thirds of shaft of left humerus with portions of pectoral and bicipital crests, USNM 501505. HORIZON.—Uncertain, either Pungo River Formation or Yorktown Formation. Pachyptila sp. Plate 9a,c,e Material.—Distal end of right humerus, USNM 464313. Distal end of left tarsometatarsus, USNM 496162. Horizon.—Uncertain, probably Yorktown Formation. MEASUREMENTS (mm).—Humerus: Distal width, 7.6; dis- tal depth through ulnar condyle, 6.0. Tarsometatarsus: Distal width, 4.7; depth of middle tro- chlea, 2.7; width and depth of shaft 10 mm above distal ex- tremity, 2.3 x 1.8. REMARKS.—These specimens are indistinguishable from medium-sized modern species of Pachyptila, the systematics of which is complex. The only previously known fossils of the ge- nus are those reported from deposits contemporaneous with the Yorktown Formation in South Africa, where fossils in the size range of modern species occurred with those of a much larger extinct species, P. salax Olson (1985b). Genus Calonectris Mathews and Iredale Calonectris krantzi, new species Plate 9m. u Holotype.—Distal end of left humerus, USNM 430724. Type Locality.—Texasgulf Inc. Lee Creek Mine, south side of Pamlico River, near Aurora, Beaufort County, North Carolina (35°23'22"N, 76°47'06"W). Horizon and Age.—Yorktown Formation, early Pliocene, inferred from preservation and lack of any known species of Procellariidae of such large size in deposits of middle Miocene age. Distribution.—Known so far only from the type locality. NUMBER 90 257 Measurements of Holotype (mm).—Distal width, 15.8; depth through radial condyle, 9.5; shaft width and depth at proximal margin of brachial depression, 9.9 x 5.5. PARATYPES.—Right coracoid lacking part of head and exter- nal distal angle, USNM 250805. Proximal end of left humerus, USNM 464308; distal third of right humerus lacking ectepi- condylar spur, USNM 430728; distal fourth of left humerus lacking ectepicondylar spur, USNM 464307. Distal two-thirds of right carpometacarpus, USNM 430718. Measurements of Paratypes (mm).—Coracoid: Head to tip of internal distal angle, 37.4. Humerus: Proximal width, 21.2; distance from head to dis- tal extent of pectoralis scar, 28.6; distal width, 15.9, 15.0. Etymology.—To Smithsonian photographer Victor E. Krantz, who, for nearly 30 years, has photographed thousands of fossil bird bones to illustrate dozens of publications, includ- ing this one, in recognition of his long service to avian paleon- tology. DIAGNOSIS.—Referable to Calonectris by the following combination of characters: ectepicondylar spur relatively short and somewhat triangular (unlike the more elongate process in Procellaria); attachment of anterior articular ligament relative- ly short and wide (longer and narrow in Procellaria); brachial depression large and deep, extending proximally well past proximal margin of ectepicondylar spur, and shaft showing no signs of flattening or expansion of entepicondylar area (in the last two respects differing from Puffinus). Larger than any known species of Calonectris, falling within the size range of smaller species of Procellaria. Remarks.—The coracoid and carpometacarpus are referred to this species almost entirely on size and could belong with one of the other large species, although the coracoid differs in shape from that of Procellaria. The largest form of Calonectris is Cory's Shearwater, C. bo- realis Cory, in which the distal width of the humerus ranges from 13.0 to 14.6 mm («=12, all unsexed, average 13.9 mm). These figures do not adequately convey the more massive na- ture of the bones of the new species. Calonectris aff. borealis (Cory) Plate 9t Material.—Distal two-thirds of right humerus lacking most of ectepicondylar spur, USNM 501506. Horizon.—Uncertain, probably Yorktown Formation. Measurements (mm).—Distal width, 14.2. Remarks.—This specimen is clearly within the size range of Cory's Shearwater, Calonectris borealis, which is the larg- est of the living taxa of the genus. The Atlantic forms of Ca- lonectris, all often considered to be subspecies of the Mediter- ranean Shearwater, C. diomedea (Scopoli), differ considerably in size. The material from Lee Creek Mine suggests that at least two of these lineages have been separate for some 5 Ma, for which reason subspecific rank seems inappropriate. The Cape Verde Shearwater, C. edwardsi (Oustalet), of the Cape Verde Islands, is the smallest form and is supposed to winter in the vicinity of its natal islands, although this appears to be little more than an assumption. The largest species, C. borealis, nests on islands in the North Atlantic (Azores, Canaries, Ma- deira, Desertas, Porto Santo, Salvages, and the Berlengas off Portugal). Calonectris diomedea is of intermediate size and nests on islands of the Mediterranean. The nonbreeding ranges of C. diomedea and C. borealis have been confounded because both taxa have usually been considered together. Both, howev- er, certainly occur in the North Atlantic Ocean off North Caro- lina today and evidently did so in the past as well (see follow- ing species). We depart from any of the modern lists we have cited in recognizing these taxa at the specific level. Calonectris aff. diomedea (Scopoli) Plate 9j Material.—Distal ends of right humeri, USNM 215433, 366013; distal end of left humerus, USNM 430745. Left femur lacking internal distal condyle, USNM 430950. Horizon.—Uncertain, probably Yorktown Formation. Measurements (mm).—Humerus: Distal width, 12.5, 12.0, 12.7. Femur: Length, 38.1; proximal width, 8.7. REMARKS.—These specimens agree in size and general mor- phology with Calonectris diomedea, the Mediterreanean mem- ber of the genus. These birds leave the Mediterranean in win- ter, and large numbers occur in the western North Atlantic Ocean off North Carolina (Lee, 1995:126). This also is the most abundant bird in a beach deposit on Bermuda dating to about 450,000 BP (Olson, unpublished data), which, with the Lee Creek Mine specimens, suggests that this pattern of winter distribution has been established for a very long time. Genus Puffinus Brisson Puffinus all. pucificoides Olson, 1975 Plate 9v,ee Material.—Associated Specimen: Partial skeleton con- sisting of right and left coracoids, right humerus lacking head, proximal two-thirds of shaft of left humerus, and left car- pometacarpus lacking minor metacarpal, USNM 464335. Individual Elements: Left humerus lacking portions of pec- toral crest, USNM 193130; right humeri lacking most of proxi- mal end, USNM 430751,430840. Left femora, USNM 242227, 501507; left femur lacking proximal end, USNM 501508. Right tarsometatarsus, USNM 430852; left tarsometatarsus, USNM 464297. Horizon.—Yorktown Formation. MEASUREMENTS (mm).—Coracoid: Length with sternal facet flat on calipers, 25.8. 258 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Humerus: Length, 99.4; length from distal extent of pecto- ral crest to ectepicondyle, 76.9, 74.5, 77.2, 77.7; distal width, 11.9,11.8,11.7,12.4. Femur: Length, 35.2, 37.8; proximal width, 8.2, 8.7; distal width, 8.1,8.4, 7.9. Tarsometatarsus: Length, 47.6, 47.3; proximal width, 7.6, 8.0; distal width, 7.7, 6.9. Remarks.—These specimens indicate a shearwater refer- able to the subgenus Thyellodroma, the existing members of which are Puffinus pacificus, of wide distribution in the Indian and Pacific oceans, and P. bulleri of New Zealand, wintering to the eastern Pacific. That this group formerly extended outside the Indo-Pacific was established by Olson (1975), who de- scribed a new species, P pacificoides, from Pleistocene depos- its on St. Helena Island in the South Atlantic Ocean. Most of the Lee Creek Mine fossils are from a species more robust than either P pacificus or P bulleri but quite similar to P. pacifi- coides. One femur (USNM 242227) is slightly smaller and de- cidedly more gracile than the other Lee Creek femora or those of P. pacificoides and closely resembles that of P. bulleri, pos- sibly indicating a third species in the Thyellodroma group (see following species). Regardless, this group of shearwaters had a long history in the Atlantic, from which it vanished at some time in the Pleistocene. Puffinus (Thyellodroma) sp. Plate 9gg Material.—Distal end of right humerus, USNM 464303; distal end of left humerus, USNM 464301. Complete left tar- sometatarsus, USNM 250810. Horizon.—Uncertain, but Pungo River Formation inferred from similarity to fossils of an unnamed species from Calvert Formation and differences from any modern species lineage. MEASUREMENTS (mm).—Humerus: Distal width, 9.7, 9.0. Tarsometatarsus: Length, 42.1; proximal width, 6.0+; dis- tal width, 6.3. Remarks.—The humerus and tarsometatarsus in this spe- cies have a morphology consistent with that in the Thyellodro- ma group of shearwaters but come from a species much smaller than any modem taxon. Puffinus (Ardenna) sp. Plate 9bb,cc Material.—Distal end of right humerus, USNM 179269; distal ends of left humeri, USNM 177786, 179249, 242329, 366613. Left tarsometatarsus lacking inner and outer trochleae, USNM 366625; proximal ends of left tarsometatarsi, USNM 177806, 366977; distal ends of left tarsometatarsi, USNM 181072,215620. HORIZON.—Uncertain, either Pungo River Formation or Yorktown Formation. Measurements (mm).—Humerus: Distal width, 14.3, 14.6, 15.2, 14.3, 14.0. Tarsometatarsus: Length, 65.1; proximal width, 9.5+, 9.7, 9.2; distal width, 8.3. Remarks.—This shearwater is a fairly common species in the Lee Creek Mine fauna. The humerus and tarsometatarsus are similar in morphology to those of Puffinus gravis (subge- nus Ardenna Reichenbach) but are larger. Puffinus aff. gravis (O'Reilly) Plate 9p Material.—Distal ends of right humeri, USNM 177890, 178187; distal end of left humerus, USNM 430709. Horizon.—Uncertain, either Pungo River Formation or Yorktown Formation. Measurements (mm).—Distal width, 13.7, 12.8, 12.4. Remarks.—The modern Greater Shearwater breeds at Tristan da Cunha and Gough islands and breeds in limited numbers in the Falklands, spending the boreal summer in the North Atlantic, a pattern of distribution that the fossils suggest may have been in effect for millions of years. Puffinus conradi Marsh (1870), from the Calvert Formation, is quite similar to the modem P. gravis and would be expected in the Pungo Riv- er Formation, as would the Pliocene representative of this sub- genus. Puffinus aff. tenuirostris (Temminck) Plate 9k,z Material.—Distal portions of right humeri, USNM 430760, 430761; distal two-thirds of left humerus, USNM 464333. Complete left ulna, USNM 366014. HORIZON.—Uncertain, probably Yorktown Formation. MEASUREMENTS (mm).—Humerus: Distal width, 10.8+, 11.9, 11.7. Ulna: Length, 88.0. Remarks.—These remains indicate a shearwater with a very compressed shaft of the humerus. This species is smaller than the Sooty Shearwater, Puffinus griseus (Gmelin), but is larger than the Christmas Island Shearwater, P. nativitatis Streets, or any of the Manx Shearwater, P. puffinus (Briinnich), assem- blage. No existing shearwater with these qualities is found in the Atlantic today, the only other one being P. tenuirostris, which breeds on islands around Australia and ranges widely over most of the Pacific. The well-preserved left humerus listed above (Plate 9k) is virtually identical to that of P. tenuirostris except for the shorter scar for the anterior articular ligament. The ulna is at the low end of the size range for P. tenuirostris or perhaps is smaller, but it has no closer match among other modern species of Puffinus. The species described as P. holei (Walker et al., 1990; emended to P. holeae by Micheaux et al., 1991:806, note 11), from the Pleistocene of the Canary Islands, may bear scrutiny in connection with this material. NUMBER 90 259 Puffinus cf. puffinus (Briinnich) sensu lato Plate 9/ Material.—Right humerus lacking head, internal tuberosi- ty, and ectepicondylar spur, USNM 430765. HORIZON.—Uncertain, probably Yorktown Formation. Measurements (mm).—Length from distal extent of pecto- ralis scar to entepicondyle, 64.5; distal width, 10.8. REMARKS.—Among the Lee Creek Mine collections are sev- eral specimens of humeri with compressed shafts that fall in the general size range of the various taxa often grouped as subspe- cies of the Manx Shearwater, Puffinus puffinus. The best pre- served of these, listed above, has the size and proportions of the Levantine Shearwater, P. mauretanicus Lowe, a Mediterranean breeder that winters in the eastern North Atlantic. Nominate P. puffinus breeds on both sides of the North Atlantic and is high- ly migratory. The fossil taxa P. nestori (Alcover, 1989) and P. olsoni (McMinn et al., 1990), from the late Pliocene of Eivissa (Ibiza) and the Holocene of the Canary Islands, respectively, also belong in this group of small shearwaters. Puffinus magn. Iherminieri Lesson Plate 9g, h h, a Material.—Proximal end of right humerus, USNM 464332. Proximal end of left tarsometatarsus, USNM 464311; distal end of right tarsometatarsus, USNM 178074. HORIZON.—Uncertain, probably Yorktown Formation. Measurements (mm).—Humerus: Proximal width, 11.3. Tarsometatarsus: Distal width, 5.8. REMARKS.—Remains of Puffinus in this smallest size class of the genus (which also includes the Little Shearwater, P. assi- milis Gould) are extremely rare at Lee Creek Mine, despite the fact that Audubon's Shearwater, P. Iherminieri, now occurs abundantly off North Carolina. The distal end of a tarsometa- tarsus listed above is only tentatively included because, al- though it is of an appropriate size, it is not referable with cer- tainty to Puffinus. Order Pelecaniformes (pelicans, cormorants, and allies) Family Pelecanidae (pelicans) Genus Pelecanus Linnaeus Pelecanus schreiberi Olson (1999) Plate\0a-d Material.—Distal end of right femur, USNM 192077 (ho- lotype); distal end of left femur, USNM 263567 (paratype). Pedal phalanx 1 of digit III, USNM 446506 (paratype); pedal phalanx 2 of digit III, USNM 421948 (paratype). Horizon.—Basal Yorktown Formation from matrix analysis of paratypical femur (Olson, 1999); also tentatively identified from early Pliocene Bone Valley Formation in Florida. Additional Material Examined (paratypes).—Bone Val- ley Formation, Florida: Right quadrate lacking orbital pro- cess, UF 125031. Axis vertebra lacking dorsal spine, UF 65677. Measurements.—See Olson (1999). Remarks.—This large species exceeds in size either of the living North American pelicans (American White Pelican, Pelecanus erythrorhynchos Gmelin; Brown Pelican, P. occi- dentalis Linnaeus) and would have equalled the two largest modern pelicans in size (Dalmatian Pelican, P. crispus Bruch; Great White Pelican, P. onocrotalus Linnaeus). The fossil fem- ora differ from all modem pelicans in having the rotular groove markedly deeper and narrower, the internal condyle more ele- vated anteriorly, and the intercondylar fossa narrower and deeper. The pedal phalanges are much more robust than are the comparable elements of modern pelicans. The shaft of the femur of the holotype is filled with dense medullary bone. This serves as a calcium reserve in females, forming 10 to 14 days prior to egg-laying and being quickly re- sorbed afterwards (see references cited in Ballmann, 1979, and Mourer-Chauvire et al., 1999). Thus, this species must have been breeding in the vicinity of deposition. Family Pelagornithidae (pseudodontorns, false-toothed birds) These gigantic, soaring, pseudotoothed birds are rather well represented in the Lee Creek Mine collections, but, as is usual in this group, only by scraps because the extremely thin walls of the bones are easily broken. Olson (1985d) reviewed some aspects of the systematics of these birds, concluding that all those known are referable to a single family, Pelagornithidae, which has several synonyms. The first species to be named was Pelagornis miocaenus Lartet (1857), from the Miocene of France. So far, all of the considerable material of these birds found in late Oligocene and Neogene deposits around the North Atlantic appear to be referable to a single genus—Pelagornis— although this includes a variety of species. Evidence from other localities suggests that as many as three species of pseudodon- torns may have coexisted. We have not attempted to describe the Lee Creek Mine material in detail, pending planned revi- sionary studies of the group by K.I. Warheit and Olson. Adding to the problems of identifying the Lee Creek pseudo- dontorns is the fact that we do not yet know from which hori- zon any of the material comes. The preservation of much of the material resembles that of Yorktown-age specimens. The most frequently represented species of pseudodontorn in the Calvert Formation, including the Pollack Farm site in Delaware (Ras- mussen, 1998), is smaller than either of the two species known so far from Lee Creek Mine, which also is indirectly suggestive of their derivation from the Yorktown Formation. This line of reasoning has its limitations, however, because there was an 260 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY extremely large pseudodontom in the late Oligocene of South Carolina (Warheit and Olson, unpublished data) and presum- ably there would have been large species in the middle Mi- ocene as well. Although mainly fragmentary, the material from Lee Creek Mine is important, as certain elements represented are other- wise unknown, or nearly so, for the entire family. At least two species can be distinguished, mainly on the very different sizes of comparable elements: in this case, the heads of humeri, frag- ments of coracoid, and distal ends of femora. Genus Pelagornis Lartet, 1857 Pelagornis sp. 1 Plate 1 \p.q,t,u Material.—Fragment of left coracoid with glenoid and scapular facets, USNM 464329. Heads of right humeri, USNM 215442, 308155. Proximal ends of left femora, USNM 321289, 446499; distal end of right femur, USNM 242202; distal ends of left femora, USNM 205473, 242327, 446498. Inner and middle trochleae of left tarsometatarsus, USNM 446500. HORIZON.—The stratigraphy is uncertain; preservation and abundance suggest that most specimens are probably from the Yorktown Formation, but the Pungo River Formation cannot be ruled out because pseudodontorns were already diverse in the late Oligocene and early Miocene in the North Atlantic. Measurements (mm).—Femur: Proximal width, 29, 30; distal width, 29.6, 30.9, 32.7, 30.6. Tarsometatarsus: Width and depth of middle trochlea, 11.7x 17.1. REMARKS.—The very long but light humeri of these birds did not preserve well at Lee Creek Mine, and they are repre- sented only by the heavy, rounded heads. These and the scap- ulae articulate well with the single fragment of coracoid and associate these elements with the smaller species present in the deposits. There are two size classes of femora, the larger species (see below) being represented only by a very worn distal end. All of the remaining femoral fragments are referred to the smaller species, along with the fragment of tarsometa- tarsus. The shafts of all of the femora are filled with medul- lary bone, as are those of a specimen from the Calvert Forma- tion, so either all of these were from laying females or the presence of medullary bone in these birds has some other sig- nificance. Pelagornis sp. 2 PLATE ll/.M Material.—Fragment of sternal articulation of right cora- coid, USNM 425109; fragment of left coracoid with glenoid and scapular facets, USNM 446501. Heads of right humeri, USNM 425111,446497; head of left humerus, USNM 464331. Distal end of left femur, USNM 252307. Horizon.—The stratigraphy is uncertain; preservation and abundance suggest that most specimens are probably from the Yorktown Formation, but the Pungo River Formation cannot be ruled out because pseudodontorns were already diverse in the early Miocene and late Oligocene in the North Atlantic. Measurements (mm).—Femur: Distal width, >33. Remarks.—The coracoidal fragment is much larger than that referred to the preceding species, and the humeral heads articulate well with it, except for USNM 446497, which ap- pears almost intermediate in size. The sternal portion of the co- racoid is so large and massive as to suggest its placement here rather than with the smaller species. Pelagornis sp. 1 or Pelagornis sp. 2 Plate Wa-d.fih.j Material.—The following material cannot at this point be identified certainly with one or the other of the above species and is referred only to the genus Pelagornis. Rostral "tooth" (pseudotooth), USNM 464325; fragment of mandible with "tooth," USNM 182106; distal end of right mandible, USNM 425108; associated distal ends of right and left mandibles, USNM 446494; right quadrate, USNM 446495 (height 45.6 mm); right pterygoid, USNM 425110 (length 40.1 mm, greatest anterior width 17.2 mm). Cervical vertebrae: ax- es, USNM 275777, 446502, 464326; vertebral centra, USNM 250715, 425102, 425103; basal cervical vertebra, USNM 425101. Anterior portion of carina of sternum, USNM 464328 (depth of furcular facet 44.0 mm). Proximal end of left radius, USNM 183512 (greatest diameter 22.4 mm). Left radiale, USNM 446496 (greatest diameter 27.0 mm). Distal end of right tibiotarsus lacking external condyle, USNM 446507; dis- tal end of left tibiotarsus lacking internal and part of external condyle, USNM 448913. Pedal phalanx, USNM 183506. Also of questionable placement are the following scapulae: anterior end of right scapula lacking much of acromion, USNM 464330 (greatest width 35+ mm), anterior end of left scapula lacking much of acromion, USNM 425104 (greatest width 38.7 mm), and anterior portion of shaft of left scapula with part of glenoid facet, USNM 425105. These are all of roughly the same size, which is approximately the size of the largest Oli- gocene pseudodontom known from South Carolina (K.I. War- heit, pers. comm., 1998). This suggests that they should be re- ferred to the larger species at Lee Creek Mine, but the two scapulae that preserve the coracoidal heads do not articulate well with the largest coracoidal fragment having a scapular fac- et (USNM 446501, Pelagornis sp. 2), yet they fit nearly per- fectly with the smaller one (USNM 464329, Pelagornis sp. 1). Family Sulidae (boobies and gannets) The phylogeny and paleontology of the Sulidae have been studied extensively by K.I. Warheit, who included fossils from NUMBER 90 261 Lee Creek Mine in his investigations. Some discussion of fos- sils may be found in his dissertation and elsewhere (Warheit, 1990, 1992), but most of his paleontological studies remain to be published. Pending his revisionary study of this material, we have set forth herein only a skeletal outline (with measure- ments omitted) concerning the sulid faunas from Lee Creek Mine based partly on Warheit's unpublished data and partly on our own observations. Specimens listed follow Warheit's identifications. Genus Morus Vieillot The modern Sulidae consists of the tropical boobies (Sula Brisson and Papasula Olson and Warheit) and the temperate gannets (Morus). The three living species of gannets, often considered to form a superspecies, occur in the North Atlantic, in southern Africa, and in New Zealand and along the southern coast of Australia. Fossil taxa also are known from the North Pacific. Although neontologists have frequently included Morus in the genus Sula, the two are quite distinct osteologi- cally, with Morus being the more derived genus (Warheit, 1990). The distinguishing characters of Morus are evident at least as far back as the late early Miocene, which is argument enough for recognizing the genus. All fossil Sulidae from Lee Creek Mine and the Calvert For- mation are referable to the genus Morus, which was much more diverse in the North Atlantic in the past, with at least three con- temporaneous species occurring sympatrically in both the mid- dle Miocene and the early Pliocene. Gannets are the most abundant fossil birds in the Calvert Formation. For some time it has been known that at least three species are represented in the fauna, and that despite their hav- ing been originally referred to Sula, all are more closely relat- ed to gannets (Olson, 1984). Thus, the three named taxa are herein referred to Morus. All are smaller than any modem spe- cies of Morus, and one, M. avitus (Wetmore), is smaller than any living member of the family. These same three species also occur at Lee Creek Mine, where they are known or as- sumed to come from the Pungo River Formation. All are gen- erally less pneumatic than the larger Pliocene species. Of the three species of gannets from the early Pliocene, the smallest, which is about the size of or slightly smaller than the living Northern Gannet, Morus bassanus Linnaeus, is referred to the species Morus peninsularis Brodkorb (1955), previously known from the Bone Valley Formation in Florida. The other two are new, as yet undescribed species, both larger than living gannets. It is not known whether the three Pliocene species are larger descendents of the three Miocene lineages or which of the Pliocene species, if any, gave rise to modern gannets. The Pliocene species all have pneumaticity well developed, in con- trast to those from the Miocene. Morus avitus (Wetmore), new combination Sula (subgenus Microsula) avita Wetmore, 1938:22. Material.—Left coracoid, USNM 177798; left coracoid lacking sternal end, USNM 501510. Distal half of right humer- us, USNM 178033. Proximal end of right ulna, USNM 501511. Distal end of radius, USNM 210530. Left carpometacarpus missing minor metacarpal, USNM 215722; proximal ends of right carpometacarpi, USNM 183481, 215638; distal end of right carpometacarpus, USNM 178186. Left femur, USNM 215761. Left tarsometatarsi, USNM 181029, 426057; distal end of left tarsometatarsus, USNM 426063. HORIZON.—Pungo River Formation inferred from similarity to fossils from Calvert Formation. Remarks.—A small amount of material exists that is insep- arable from the type of Sula (Microsula) avita, from the Cal- vert Formation of Maryland. Several elements (ulna, coracoid) were collected by J.H. McLellan from spoil derived from units 4 and 5 of the Pungo River Formation. This was a tiny gannet, much smaller than the smallest living member of the family. Morus atlanticus (Shufeldt, 1915), new combination Sula atlantica Shufeldt, 1915:62. Material.—Distal end of left humerus, USNM 411963. Proximal end of right ulna, USNM 426031; proximal ends of left ulnae, USNM 302327, 368551, 412039; distal ends of right ulnae, USNM 192574, 206512, 275806, 367103; distal ends of left ulnae, USNM 177943, 215686, 241374. Left carpometac- arpus missing minor metacarpal, USNM 193364; proximal end of left carpometacarpus, USNM 257493; distal ends of left car- pometacarpi, USNM 210439, 412056. Left femur, USNM 252310. Distal end of right tibiotarsus, USNM 368562; distal end of left tibiotarsus, USNM 426068. HORIZON.—Pungo River Formation inferred from similarity to fossils from Calvert Formation. Remarks.—The holotype of this species came from the Kirkwood Formation in New Jersey, which is temporally equivalent to the Calvert Formation, whence specimens insepa- rable from the holotype have been collected. This species was intermediate in size between M. avitus and M. loxostylus (Cope), but there is probably some overlap in size with the lat- ter (K.I. Warheit in litt., 1997). Morus loxostylus (Cope, 1870) Sula loxoslyla Cope, 1870:236. Morus loxostyla [sic].—Brodkorb, 1963:259. Material.—Right coracoid lacking sternal end, USNM 241426; right coracoid lacking most of both ends, USNM 367004. Distal ends of right humeri, USNM 366796, 411973; distal ends of left humeri, USNM 181116, 215456. Proximal end of right ulna, USNM 177935; proximal end of left ulna, USNM 366936; distal end of right ulna, USNM 236836; distal 262 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY end of left ulna, USNM 367088. Proximal ends of radii, USNM 412008, 426089; distal ends of radii, USNM 178055, 192075. Right tarsometatarsus, USNM 215476; left tarsometa- tarsus, USNM 412010; distal end of right tarsometatarsus, USNM 366715; distal ends of left tarsometatarsi, USNM 193002,412089,412094. Horizon.—Pungo River Formation inferred from similarity to fossils from Calvert Formation. Remarks.—This fairly common form is about the size of the modern Blue-footed Booby, Sula nebouxii Milne-Ed- wards. The fossil material compares well with material from the Calvert Formation of Maryland assigned by Wetmore (1926) to Morus loxostylus, the holotype of which has been lost. Study of the original illustration and description (K.I. Warheit, pers. comm., 1997) indicates that it is reasonable to assign the largest Calvert gannet to this species, as has been done for material from the Calvert Formation of Delaware (Rasmussen, 1998). Morus peninsular is Brodkorb, 1955 PLATES \2h.l, \Sa,e,g,i,m,n Material.—Right coracoids, USNM 412054, 426052; left coracoids, USNM 367093, 412016, 412021, 426053, 501512, 501514; right coracoid lacking part of sternal end, USNM 308220, left coracoid lacking much of both ends, USNM 412042; sternal end of left coracoid, USNM 426054. Proximal portions of left humeri, USNM 464323, 501514; distal end of right humerus, USNM 215691; distal ends of left humeri, USNM 215716, 411971, 411981. Proximal two-thirds of right ulna, USNM 412017; proximal end of left ulna, USNM 412023; distal half of right ulna, USNM 412006. Left car- pometacarpi missing minor metacarpals, USNM 412066, 426023; proximal end of left carpometacarpus, USNM 366806. Left femora, USNM 177910, 412048. Right tarsometatarsus, USNM 426061; left tarsometatarsus lacking much of proximal end, USNM 275838; distal ends of left tarsometatarsi, USNM 179261,248510. HORIZON.—Yorktown Formation (USNM 179261 probably from the middle Yorktown as determined from foraminifera and sedimentary characteristics of matrix). Remarks.—The most abundant sulid at Lee Creek Mine is a large gannet that appears to be identical to material from the Bone Valley Formation described by Brodkorb (1955) as Morus peninsularis, which was characterized as being some- what smaller than the modern Northern Gannet, M. bassanus, similar in size to or slightly smaller than either the Australian Gannet, M. serrator (Gray) or the Cape Gannet, M. capensis (Lichtenstein), but slightly larger than the Masked Booby, Sula dactylatra Lesson. Morus, undescribed species 1 Plates \2a-g,i,n, [3b,k,o MATERIAL.—Associated Specimen: Partial skeleton con- sisting of mandible (lacking symphysis), 3 cervical vertebrae, proximal portions of right and left humeri, distal ends of left ulna and radius, left ulnare, and proximal end of left car- pometacarpus, USNM 181052. Individual Elements: Left coracoid, USNM 464322; right coracoid lacking much of sternal end, USNM 193086,412053. Proximal half of right humerus, USNM 368548; distal ends of left humeri, USNM 411966,411982,426011. Proximal ends of right ulnae, USNM 321227, 412077, 426030; proximal ends of left ulnae, USNM 366880, 412087, 426029; distal end of right ulna, USNM 366883; distal end of left ulna, 412078. Proximal end of radius, USNM 193019; distal ends of radii, USNM 411983, 412083. Proximal ends of right carpometacarpi, USNM 412065, 412067; proximal ends of left carpometacarpi, USNM 308197, 321241. Distal two-thirds of right tibiotarsus, USNM 426069. Left tarsometatarsus, USNM 366904. Horizon.—Yorktown Formation. Remarks.—This is a large gannet, being intermediate in size between Morus peninsularis and the following extremely large species; it is larger than M. bassanus by roughly the same amount as the latter is larger than M. capensis. Morus, undescribed species 2 Plates \2j.m. 13c Material.—Scapular ends of left coracoids, USNM 183499, 321253; sternal ends of left coracoids, USNM 366666, 412047. Proximal end of left humerus, USNM 426009; distal end of right humerus, USNM 242374; distal ends of left hu- meri, USNM 411951, 411987 Proximal end of left ulna, USNM 215558; distal end of left ulna, USNM 183508. Distal end of radius, USNM 177881. Right carpometacarpus missing minor metacarpal and most of distal end, USNM 206346; prox- imal end of right carpometacarpus, USNM 412062; distal end of right carpometacarpus, USNM 256212. Horizon.—Yorktown Formation. Remarks.—This is a huge species, at least half again larger than the largest living gannets (Morus). Family Phalacrocoracidae (cormorants) Genus Phalacrocorax Brisson Because cormorants have never been found in the Calvert Formation, we assume that it is unlikely that any of the cormo- rant fossils from Lee Creek Mine come from the Pungo River Formation. Furthermore, the morphology of the Lee Creek cor- morants does not differ greatly from that of modem species, which would not be expected of taxa as old as middle Miocene. NUMBER 90 263 Phalacrocorax wetmorei Brodkorb, 1955 Figure 7; Plate \4a,b,d,e,g,h,j,n,o,q,s,t MATERIAL.—Associated Specimen: Distal end of left ulna, carpometacarpus, and radiale, USNM 179307. Individual Elements: Mandibular articulation, USNM 215700. Anterior fragment of sternum, USNM 446468. Right coracoids, USNM 242346, 446444-446450; scapular ends of right coracoids, USNM 192013, 192922, 192944, 206303, 206482, 206586, 248565, 302344, 366352, 366411, 446451, 446452; scapular ends of left coracoids, USNM 177761, 192868, 193279, 193306, 215467, 215752, 215842, 241407, 308186, 446454^446464; sternal ends of right coracoids, USNM 256244, 446453; sternal ends of left coracoids, USNM 178179,250762. Anterior end of right scapula, USNM 242361. Proximal end of right humerus, USNM 501515; proximal ends of left humeri, USNM 446409, 446410, 446469; distal ends of right humeri, USNM 177745, 215818, 366340, 366446, 446407, 446408; distal ends of left humeri, USNM 446411^446413. Proximal ends of right ulnae, USNM 215745, 252320, 256265, 302380, 446397-446399; proximal ends of left ulnae, USNM 177846, 181088, 250701, 367019; distal ends of right ulnae, USNM 177859, 192491, 206436, 210482, 236825, 250776, 250821, 252428, 321250, 446400-446404; distal ends of left ulnae, USNM 177856, 177863, 178099, 192982, 206613, 215636, 215672, 241428, 242352, 308194, 366919, 446405, 446406. Proximal end of radius, USNM 446467. Right carpometacarpus, USNM 446391; left car- pometacarpus, USNM 446394; proximal ends of right car- pometacarpi, USNM 306343, 366445, 367139, 446392; proxi- mal ends of left carpometacarpi, USNM 192942, 206472, 215650, 242336, 242370, 446395, 446396; distal ends of right carpometacarpi, USNM 178204, 446393; distal end of left car- pometacarpus, USNM 206534. Alar phalanges, USNM 193068, 206542. Synsacral fragments, USNM 210440, 308191, 446465. Right femora, USNM 242205, 302304, 446414,446415; left femora, USNM 179250, 367159,446419, 446420; proximal ends of right femora, USNM 366692, 446416; proximal ends of left femora, USNM 179291, 446421, 446422; distal ends of right femora, USNM 215542, 248562, 368478, 446417, 446418; distal ends of left femora, USNM 177938, 192935, 248571, 250748, 366950, 446423, 446424. Right tibiotarsus, USNM 446425; distal ends of right tibiotarsi, USNM 177882, 181115, 446426, 446427; distal end of left ti- biotarsus, USNM 446428. Right tarsometatarsi, USNM 193098, 275785, 366903, 446429; left tarsometatarsi, USNM 210425, 215429, 446435^146437; proximal ends of right tar- sometatarsi, USNM 193129, 215815, 252302, 257499, 367030, 368556, 446430^146432; proximal ends of left tar- sometatarsi, USNM 242364, 446438-446443; distal ends of right tarsometatarsi, USNM 181065, 446433, 446434; distal ends of left tarsometatarsi, USNM 178184, 179229, 179262, 206354,215765, 248504. Pedal phalanges, USNM 192458. Horizon.—Yorktown Formation. Additional Material Examined.—Bone Valley, Florida: Right coracoid, UF 95488. Right femur, UF 101946. Distal end of right tibiotarsus, UF 65772. Right tar- sometatarsus, UF 94550. Measurements.—See Table 7. The common species of cormorant at Lee Creek Mine overlaps broadly in size with Phalacrocorax wetmorei Brodkorb (1955) from the upper Bone Valley Formation but tends to average smaller. When skeletal measurements of the two modem subspecies of Dou- ble-crested Cormorant (P. auritus auritus Lesson, P. a. florida- nus Audubon) from the Atlantic coast of North America are plotted with those of the fossil cormorants from Lee Creek Mine and Bone Valley, all populations show broad overlap in most measurements. Two measurements of the humerus (width of the proximal end and depth of the head), however, separate the fossil populations from the modem ones (Figure 7). Remarks.—Phalacrocorax wetmorei was described by Brodkorb (1955) from the Bone Valley Formation as a new species of cormorant similar in size to P. auritus. Comparisons of humeri and tarsometatarsi of cormorants of this size class from Lee Creek Mine with those of Bone Valley revealed no osteological differences. Detailed comparisons were made between the Pliocene fos- sils and the European Shag, P. aristotelis Linnaeus, a species of somewhat similar size that occurs along the coasts of the eastern Atlantic and Mediterranean. Both fossil populations, however, are decidedly more similar to P. auritus than to P. ar- istotelis in most qualitative characters, as well as in size. Most Lee Creek and Bone Valley specimens of P wetmorei differ from P. auritus in having the posterior rim of the medial condyle of the tarsometatarsus forming an angle with the shaft, rather than being confluent with it; the accessory distal fora- men of the tarsometatarsus positioned nearer the primary distal foramen; the bicipital crest of the humerus less laterally ex- panded (in some P. wetmorei); and the ventral tubercle of the humerus less laterally produced. In these four characters the fossils seem more like P. aristotelis, whereas in all other re- spects they are more similar to P auritus. Although it seems likely that P. wetmorei is only a temporal manifestation of P. auritus, which latter may date back at least to the earliest Pleis- tocene of Florida (Emslie, 1998), we have used the former name for the Lee Creek Mine fossils pending future revisionary work. Phalacrocorax, large species PLATE \4c,fii,k-m,p,r,u,v Material.—Nearly complete right coracoid, USNM 177880; scapular end of right coracoid, USNM 242335. Distal end of right humerus, USNM 215774. Proximal end of right ra- dius, USNM 179256. Right femur, USNM 177791. Proximal ends of left tibiotarsi, USNM 177901,215597; distal end of left tibiotarsus, USNM 177787. Distal end of left tarsometatarsus, USNM 430872. 264 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Additional Material Examined.—Bone Valley, Florida: Right tibiotarsus, UF 67935. HORIZON.—Yorktown Formation. Measurements.—See Table 7. Remarks.—These bones are from one or more species of cormorant considerably larger than the preceding. The only large cormorant in the North Atlantic today is the Great Cor- morant, Phalacrocorax carbo Linnaeus. Compared with a large male of that species from the Netherlands (USNM 555669), the fossil femur is larger, but the shaft is slightly less robust and curved. One of the tibiotarsi proximal ends is not quite as large as that of P. carbo, but the fibular crest is longer, extending farther proximally. The other is poorly preserved and is considerably smaller. The distal ends of the tibiotarsi and tar- sometatarsi are larger and have much heavier shafts than they do in the modern species. The fossil pectoral elements are all either slightly or much smaller than in P carbo. The fossil cormorants of North America are much in need of revision, and many of the described taxa, which have been summarized by Emslie (1995b), are known from inadequate material. On the other hand, Emslie (1995b) described a new species, Phalacrocorax filyawi, from abundant material from a late Pliocene death assemblage in central Florida. This was re- lated to the large Pacific cormorants (Brandt's Cormorant, P. penicillatus (Brandt); the extinct Pallas's Cormorant, P. perspi- cillatus Pallas; and the Flightless Cormorant, P. harrisi Roths- child) and represents a lineage that must have invaded the At- lantic prior to the closing of the Panamanian seaway. It would have existed in a region of cold upwelling that was disrupted after the emergence of the Panamanian isthmus, causing the ex- tinction of the cormorant. Another large cormorant, Phalacrocorax idahensis (Marsh, 1870), has been reported from late Pliocene and Pleistocene de- posits in Idaho and Florida, although there is some doubt that all the material ascribed to it belongs to the same species (Emslie, 1995b, 1998). Emslie (pers. comm., 1997) made com- parisons of the large cormorant bones from Lee Creek Mine, noting that the femur compared well with P. idahensis, the co- racoids with P. filyawi, and the remaining bones with neither of those species. Given the amount of intraspecific variation in some modern cormorants, however, it is highly unlikely that the large cormorant material from Lee Creek Mine would be referable to more than two species. s s 13 E •a o h a 3 U B 3 m 25 24 23 22 21 20- 19 Bone Valley P. wetmorei Lee Creek P. cf. wetmorei I ~ 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5 Humerus head depth (mm) Figure 7.—Bivariate scatter plot of depth versus width of humerus head for cormorants: Lee Creek Mine Phala- crocorax wetmorei. Bone Valley P. wetmorei, and two Atlantic coast subspecies of Double-crested Cormorant (P. auritus auritus and P. a. floridanus). NUMBER 90 265 TABLE 7.—Measurements (mm) for recent and fossil cormorants (Phalacrocorax): P. wetmorei and P. large spe- cies from the Lee Creek Mine (LC); P. wetmorei from the upper Bone Valley Formation (BV), and modem spec- imens of P. auritus auritus (7 males, 7 females). Measurements are in list form when number of specimens (n)<3 for fossil species. (s=standard deviation.) Element P. wetmore (LC) P. large sp. (LC) P. wetmorei (BV) Pa. auritus (n= 14) n Range Mean s n Range Mean s Range Mean s Coracoid Length - 61.3,63.5 - - - 4 59.3-67.9 64.4 2.6 - 64.3 2.8 Glenoid facet Length - 11.3,11.7 - - 13.0 4 12.0-12.6 12.4 0.3 11.5-13.3 12.4 0.5 Carpometacarpus Length 3 67.8-73.5 69.9 3.1 - - - 69.6 - 63.8-73.3 69.0 2.5 Proximal depth 10 13.4-15.3 13.9 0.7 - 24 13.1-14.2 13.7 0.4 12.7-14.1 13.3 0.4 Distal depth 5 4.4-4.8 4.7 0.2 - 8 4.4-5.4 4.9 0.3 4.6-5.4 5.0 0.2 Distal width 5 7.1-7.5 7.3 0.2 - 9 7.2-7.8 7.4 0.2 7.1-7.7 7.4 0.2 Humerus Proximal width - 22.0 - - - 16 21.5-24.8 23.2 0.9 21.0-24.6 22.9 1.0 Head depth - 6.3, 6.7 - - - 16 6.3-7.3 7.0 0.3 10.5-12.3 11.1 0.5 Shaft depth - 6.0 - - - 18 5.4-7.0 6.2 0.3 6.0-7.7 6.8 0.4 Distal width 7 13.2-15.0 14.2 0.7 - 35 14.9-16.8 15.6 0.4 15.4-16.9 15.9 0.5 Distal depth 8 9.8-11.0 10.4 0.4 - 34 9.4-11.1 10.3 0.4 10.0-11.9 10.9 0.6 Ulna Proximal width 6 10.8-12.4 11.4 0.7 - 26 10.6-12.8 11.5 0.4 11.4-12.7 11.9 0.4 Proximal depth 6 10.0-11.2 10.7 0.5 - 26 10.5-12.3 11.6 0.4 11.5-13.5 12.5 0.6 Femur Length 4 50.0-56.6 53.9 3.2 62.0 9 53.3-59.2 55.5 1.6 49.5-58.7 54.9 2.4 Head depth 7 5.9-7.0 6.6 0.4 8.0 16 6.3-7.2 6.8 0.3 6.4-7.6 7.0 0.3 Proximal width 6 14.2-17.5 15.3 1.4 17.9 19 14.5-16.7 15.4 0.6 15.4-17.4 16.3 0.6 Shaft width 12 5.7-7.0 6.2 0.3 7.9 17 5.9-7.0 6.5 0.3 5.6-7.1 6.5 0.4 Distal width 10 13.2-16.4 15.0 1.1 - 17 14.4-16.0 15.1 0.5 14.9-16.6 15.7 0.6 Tibiotarsus Shaft width _ 6.0 - - - 5 6.6-7.2 7.0 0.3 6.6-7.4 6.9 0.2 Shaft depth - 4.8 - - - 5 5.0-5.2 5.1 0.1 4.8-6.5 5.4 0.5 Tarsometatarsus Length 5 59.8-66.8 63.3 2.9 - 5 60.3-69.2 64.6 3.7 55.2-65.0 62.0 2.4 Proximal width 10 11.5-13.4 12.6 0.6 - 22 12.1-14.1 13.0 0.5 12.0-13.6 12.8 0.4 Order Charadriiformes (shorebirds, gulls, and auks) Family Stercorarhdae (skuas and jaegers) The skuas and jaegers constitute a small group of predaceous relatives of gulls, the humerus of which appears to be primitive (Olson, 1985d) within the suborder Lari (as is also that of the noddies (Anous) and skimmers (Rynchops)). The jaegers, Ster- corarius, comprise three species that breed entirely in the Arc- tic, and the skuas, Catharacta, consist of a complex of taxa that breed at high latitudes in the Southern Hemisphere, with the exception of the Great Skua, C. skua Briinnich, which breeds on islands in the eastern North Atlantic, having recently ex- panded into the Barents Sea (Cohen et al., 1997:188). We have retained the two genera mainly to promote comprehension as we can find no osteological differences between the two groups, Catharacta merely being an enlarged version of Ster- corarius as far as the skeleton is concerned. Various authors, with the perspectives offered by both morphology and behav- ior, have previously suggested merging these two genera (see references cited in Braun and Brumfield, 1998:998). Recent molecular and other evidence has made the tradition- al view of relationships within this family the object of consid- erable contention (Cohen et al, 1997; Braun and Brumfield, 1998). The most unexpected result was the apparent closer re- lationship of the Pomarine Jaeger (Stercorarius pomarinus (Temminck)) to Catharacta in general, and to the Great Skua (C. skua) in particular (Cohen et al., 1997). This, combined with the proposed very recent origin of C. skua in the North At- lantic (Andersson, 1973), led Cohen et al. (1997) to suggest three different possible hypotheses of origin of S. pomarinus, including hybridization between Catharacta and one of the other species of Stercorarius. We do not intend to dwell on the specifics of these recent studies but wish instead to elucidate what bearing the fossil record may have on the issues. At Lee Creek Mine there are scarce fossils assignable to all three size classes of modern jaegers, including S. pomarinus. Although there are some slight differences from modem species in some of the bones, the similarities are greater, and all of the fossils are assumed to have come from the Yorktown Formation. There also are a 266 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY few bones, including a diagnostic distal end of a tarsometatar- sus, that clearly belong to a much larger species the size of Catharacta. The last is not necessarily an indication of a breeding popu- lation of skuas in the North Atlantic at that time. Given the propensity of the modem Southern Hemisphere forms of Ca- tharacta to wander across the equator and the presence at Lee Creek Mine of other Southern Hemisphere vagrants, especial- ly among the Procellariidae, it is just as likely that bones of Catharacta at Lee Creek Mine represent strays from the south, particularly as skuas and jaegers commonly follow flocking shearwaters. Nevertheless, the fossil record indicates that the diversity of Stercorariidae in the North Atlantic was just as great nearly five million years ago as it is today. The size range of Sterco- rarius that existed then appears to encompass three species, and we assume that these are the same three lineages represent- ed by the modem species. Although it is still quite possible that the North Atlantic Great Skua (Catharacta skua sensu stricto) may be of much more recent origin, the fossil record suggests that a much greater period of time may need to be allowed for the point of origination of S. pomarinus. Genus Catharacta Briinnich Catharacta sp. Plate \5a,c,d Material.—Distal end of left ulna, USNM 366602. Proxi- mal end and shaft of left carpometacarpus, USNM 430889. Left tarsometatarsus lacking most of both ends, USNM 430892; distal end of left tarsometatarsus, USNM 183482. HORIZON.—Uncertain, probably Yorktown Formation. Measurements.—See Table 8. Remarks.—The above elements are from a large stercorari- id the size of the Great Skua, Catharacta s. skua Briinnich, or the South Polar Skua, C maccormicki (Saunders). The distal portion of the tarsometatarsus from Lee Creek Mine differs from that of Catharacta skua and C. maccormicki in having a smaller, rounder distal foramen, especially notable in posterior view. The metatarsal facet is more deeply excavat- ed in the fossil and extends farther distally. In posterior view, the external edge of trochlea III extends farther proximally in the fossil. In anterior view, there is no distinct furrow proximal to the internal intertrochlear notch. The carpometacarpus is as large and heavy as that in large female specimens of C. mac- cormicki, and the preserved portion is similar to modern spe- cies examined. The ulna is from a very large skua, larger than C. maccormicki, with an especially heavy shaft, but the speci- men is poorly preserved. These cold-water birds occur mainly in the Southern Hemi- sphere, with the isolated Great Skua being the only form to breed in the Northern Hemisphere. Its breeding is restricted to the subarctic of the eastern North Atlantic, but it winters regu- larly in the western North Atlantic, virtually never being found close to land. The South Polar Skua (C. maccormicki) occurs rarely but regularly in the western North Atlantic in the warmer months only. The material from Lee Creek Mine provides the first fossil record of Catharacta from the North Atlantic, but its scarcity makes it equivocal as to whether this constitutes evidence for birds of this group having established breeding colonies north of the equator by that time. As discussed above, the possibility of the fossils having originated in vagrants from the south is perhaps just as likely. Genus Stercorarius Brisson Stercorarius aff'. pomarinus (Temminck) Plate \5f,h.j.k,m,o Material.—Left coracoid lacking sterno-coracoidal pro- cess, USNM 430891. Partial distal end of right humerus, USNM 192818; distal end of left humerus, USNM 366015. Distal third of left ulna, USNM 430890. Horizon.—Uncertain, probably Yorktown Formation. Measurements.—See Table 8. Remarks.—The shaft of the coracoid is longer than it is in the Pomarine Jaeger, Stercorarius pomarinus. In other respects the fossil material agrees with the modern species. Emslie (1995b:324) reported a coracoid of "Stercorarius sp." from the late Pliocene of Florida that was considered likely to be an un- described species larger than S. pomarinus but smaller than any species of Catharacta. He also noted differences between that specimen and USNM 430891. Stercorarius aff. parasiticus (Linnaeus) Plate \5q,r,t,u,z Material.—Distal ends of right humeri, USNM 178064, 460819. Proximal end of right carpometacarpus, USNM 193269. HORIZON.—Uncertain, probably Yorktown Formation. Measurements.—See Table 8. Remarks.—The distal end of the humerus has the shaft much more flattened in medial view and has a deeper pit in the brachial depression than in 15 Parasitic Jaegers (Sterco- rarius parasiticus) examined. Larine ulnae (USNM 237149, 237176) that resemble Stercorarius and are the size of S. par- asiticus (Plate 15w,jc) also have been recovered in the Calvert Formation. Stercorarius aff. longicaudus Vieillot Plate \5bb,cc,eeff,hh Material.—Distal ends of left humeri, USNM 430893, 448914. Proximal three-fourths of left carpometacarpus, USNM 460818. NUMBER 90 267 TABLE 8.—Measurements (mm) for recent and fossil skuas (Catharacta) and jaegers (Stercorarius) from Lee Creek Mine. Measurements are in list form when number of specimens («) <5 and for fossil samples. (r=standard deviation.) Element C. skua («=4) C. sp. C. maccormicki («=17) S. pomarinus («=7) S. aff. pomarinus 5. parasiticus («=16) 5. aff. parasili cus (n=2) Range S. longicaudus (n=9) S aff Range Mean s Range Mean s Range Mean s Range Mean s longicaudus Coracoid Median length 51.0,53.4, 56.5, 57.2 - 47.7-53.5 51.2 1.7 38.7^43.0 40.8 1.5 44.7 32.8-36.5 34.4 1.2 - 28.8-32.3 30.2 1.0 - Humerus Distal width 18.2, 19.3, 19.7,20.5 — 17.0-19.7 18.2 0.6 13.6-15.3 14.5 0.6 15.3 12.4-13.6 13.0 0.4 12.4, 13.0 10.4-11.8 11.0 0.4 11.4 Ulna Distal diagonal 11.6,12.9, 12.9, 14.0 - 11.3-13.3 12.4 0.6 9.3-10.6 10.0 0.5 10.7 8.4-9.2 8.9 0.3 - 7.1-8.2 7.8 0.3 - Carpometacarpus Proximal width 15.6, 16.8, 16.9, 17.3 16.0 14.5-16.8 15.6 0.6 12.1-13.6 13.0 0.6 - 11.1-12.2 11.5 0.3 11.7 9.5-10.9 10.1 0.4 - Tarsometatarsus Distal width 12.3, 12.9, 12.9, 14.9 11.2+ 11.3-13.0 12.2 0.5 8.2-9.0 8.7 0.3 - 6.7-7.8 7.2 0.3 - 6.1-6.8 6.5 0.2 - HORIZON.—Uncertain, probably Yorktown Formation. MEASUREMENTS.—See Table 8. REMARKS.—These humeri are indistinguishable from the modem Long-tailed Jaeger, Stercorarius longicaudus, a regular migrant in the western Atlantic. Family Laridae (gulls and terns) Genus Larus Linnaeus Larus aff. argentatus Pontoppidan PLATE \6a,c,d,f,h Material.—Portion of right mandibular ramus, USNM 430931. Proximal end of left humerus, USNM 366923; distal half of left humerus, USNM 366896. Proximal ends of right ul- nae, USNM 430896, 430897, 430903; distal end of right ulna, USNM 430905. Proximal end of right carpometacarpus, USNM 430895. MEASUREMENTS (mm).—Humerus: Proximal width, 20.3; distal width, 16.1. Ulna: Proximal width, 12.7,13.7, 14.9+; distal width, 10.5. Carpometacarpus: Proximal width, 15.0, 15.7. HORIZON.—Yorktown Formation. Remarks.—The above material represents a gull the size of the modern Herring Gull, Larus argentatus. Emslie (1995b:324) referred the two humeral fragments listed above to his new species Larus perpetuus, from the late Pliocene of Florida. The holotype of that species is a complete humerus with a length of 107.0 mm and a distal width of 14.2 mm, whereas the material from Lee Creek Mine indicates a larger species, within the lower size range of Larus argentatus, and from which we can detect no significant differences. Larus aff. delawarensis Ord PLATE \7a,c,e,h,j,m,p,q Larus elmorei Brodkorb, I953b:94. Material.—Associated left coracoid lacking sternal end and fragment of shaft of left humerus, USNM 215592. Scapu- lar halves of right coracoids, USNM 430914, 460820; shaft of right coracoid, USNM 430915. Right humerus lacking most of both ends, USNM 181025; proximal portions of right humeri lacking much of proximal ends, USNM 430911, 460821; par- tial proximal end of left humerus, USNM 460822; distal ends of right humeri, USNM 192869, 242326, 250678, 321294, 366791, 367136, 430912, 430913, 460826, 460827; distal ends of left humeri, USNM 206408, 242184, 275819, 366791, 430907^130910,460823^160825. Proximal third of right ulna, USNM 430918; proximal ends of left ulnae, USNM 430900, 430917; distal ends of right ulnae, USNM 308228, 430904, 430906, 460830, 460831; distal ends of left ulnae, USNM 366295, 366651, 430901, 430902, 460832. Right carpometac- arpus lacking minor metacarpal, USNM 460833; proximal half of right carpometacarpus, USNM 256349. Distal end of right tibiotarsus, USNM 464293; distal end of left tibiotarsus, USNM 430848. Distal end of right tarsometatarsus lacking tro- chleae II and IV, USNM 178070; distal end of left tarsometa- tarsus, USNM 236892. Horizon.—Yorktown Formation. Additional Material Examined.— Cobham Wharf James River, Surrey County, Virginia: Distal third of left hu- merus, USNM 237211. Lower Yorktown Formation, 25 ft (7.6 m) above beach. Collected by Warren Blow, received 1965. Bone Valley Formation, Florida: Distal ends of right hu- meri, USNM 447048, 447049. Proximal half of left ulna, UF 61951. Proximal half of left carpometacarpus, USNM 447047. 268 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Measurements (mm).—Humerus: Distal width, mean=13.6, standard deviation (s)=0.63, range= 12.9-14.8, n=9. Ulna: Proximal width, 10.5, 11.2, 11.9; distal width, mean=9.1,i=0.48,range=8.4-9.7, «=8. Carpometacarpus: Length, 57.4; proximal width, 12.0; distal diagonal, 7.4. Tibiotarsus: Distal width, 7.5, 8.2. Remarks.—These elements are from a gull of the same size and general morphology as Larus elmorei Brodkorb (1953b) from Bone Valley (Plate llfg.k.o), thus approximating the modem Ring-billed Gull, L. delawarensis. Larus aff. atricilla Linnaeus Plate 16/,/.« Material.—Shaft of left coracoid, USNM 192102. Distal ends of right humeri, USNM 430921, 430922, 464292, 464299; distal ends of left humeri, USNM 215622, 367112, 430920. Distal two-thirds of right ulna, USNM 366451. Right carpometacarpus missing minor metacarpal, USNM 460828. Horizon.—Yorktown Formation. MEASUREMENTS (mm).—Humerus: Distal width, 10.4, 11.0, 11.0, 11.3, 11.5, 11.6, 11.7. Ulna: Distal width, 7.8. Carpometacarpus: Length, 54.1; proximal width, 11.0. Remarks.—These specimens, indistinguishable from the Laughing Gull, Larus atricilla, indicate a species larger than the Franklin's Gull, L. pipixcan Wagler; Black-headed Gull, L. ridibundus Linnaeus; or Sabine's Gull, L. (Xema) sabini Sab- ine, and smaller than L. delawarensis or the Royal Tern, Ster- na maxima Boddaert. The carpometacarpus is similar in size to that of 5. maxima but differs from it in having the alular metacarpal with the extensor attachment more rugose and less slanted. Larus aff. minutus Pallas Plate ]6v,w,y,z,bb Material.—Left humerus lacking most of proximal end, USNM 430916; partial distal end of right humerus, USNM 430924; distal ends of left humeri, USNM 430923, 464337. Right ulna, USNM 460817; distal end of left ulna, USNM 464338. HORIZON.—Yorktown Formation. Measurements (mm).—Humerus: Distal width, 7.7, 8.2, 8.6. Ulna: Length, 61.8+; proximal width, 6.9; distal diagonal, 5.4. Remarks.—The virtually complete ulna is almost indistin- guishable from that of the Little Gull, Larus minutus. The fos- sil has the distal end less rotated internally than it is in modem L. minutus. The Lee Creek ulna and those of modem L. minut- us differ from North American species of Sterna of similar size as follows: in the former, the prominence for the anterior articular ligament juts farther internally; the distal edge of the external cotyla is smoothly continuous with the shaft, not squared; the shaft is not swollen just proximal to the distal end so that the carpal tuberosity is more distinct from the shaft; the anconal edge of the external tuberosity is less flared; and there is no distinct pit proximal to the internal condyle, but instead a ridge extends from this area to the carpal tuberosity. The fossil ulna and that of L. minutus differ from the Black Noddy, Anous minutus Boie, in having a straighter shaft from internal view; having the external cotyla more flared externally; and lacking a strongly shelf-like internal edge of the proximal ra- dial depression. The Little Gull is essentially an Old World species, although it has recently begun nesting in North America. As a result of this recent colonization, the species now regularly occurs dur- ing the nonbreeding season in North Carolina, where the first record was in 1971 (Lee, 1995:147), and elsewhere in eastern North America. Larus magn. ridibundus Linnaeus Plate 16/? MATERIAL.—Proximal half of left humerus lacking internal tuberosity, USNM 210461. HORIZON.—Yorktown Formation. MEASUREMENTS.—No standard measurements possible. Remarks.—The lack of extension to the median crest distal- ward along the shaft shows this specimen to be a gull rather than a tern. It differs from the slightly smaller Larus sabini in having a less rounded bicipital crest. This specimen is similar in size to both L. ridibundus and L. pipixcan; we cannot distin- guish between these species on the basis of the proximal end of the humerus. The ulna and radius of a gull of about this size was reported from the late Miocene Big Sandy Formation of Arizona (Bickart, 1990). Larus sp. Plate 16r,/ Material.—Scapular half of right coracoid, USNM 430834. Horizon.—Pungo River Formation. Collected from Pungo River spoil pile (fide R. Purdy, Smithsonian Institution). Measurements (mm).—Length from head to distal rim of scapular facet, 11.0; depth through head, 8.9; length and width of glenoid facet, 7.6 x 4.7. Remarks.—This specimen lacks the pneumaticity under the furcular facet found in other gulls and thus resembles the Ivory Gull, Larus (Pagophila) eburnea Phipps. Perhaps this is a primitive feature, retained among living gulls only in L. ebur- nea, that may have been characteristic of most or all gulls in the early middle Miocene. NUMBER 90 269 Genus Sterna Linnaeus Sterna aff. maxima Boddaert Plate\6dd MATERIAL.—Proximal end of right carpometacarpus, USNM 215643; proximal end of left carpometacarpus, USNM 430898. HORIZON.—Uncertain, probably Yorktown Formation. Measurements (mm).—Proximal width, 10.8, 11.9. REMARKS.—These specimens differ from Larus atricilla and are similar to the Royal Tern, Sterna maxima, in having the ex- tensor attachment of the alular metacarpal strongly slanted and in lacking the pronounced rugosity of L. atricilla. The Royal Tem is a common species in coastal environments in eastern North America and breeds in North Carolina today. Sterna aff. nilotica Gmelin Plate \6ff Material.—Right carpometacarpus lacking minor metacar- pal, USNM 495588. HORIZON.—Uncertain, probably Yorktown Formation. Measurements (mm).—Length, 37.8. Remarks.—This specimen is recognizable as a tem by the longer and more slender alular metacarpal and by having the carpal trochlea more squared in outline than it is in Larus. It is too small for the Sandwich Tem, Sterna sandvicensis Latham, or for any of the tem species larger than that, and it is too large for any North Atlantic species except the Gull-billed Tern, Sterna nilotica, with which it agrees very closely. The fossil is slightly smaller than the comparative modem material avail- able to us, but this series of three modem skeletons is insuffi- cient to determine the amount of size variation in S. nilotica. The Gull-billed Tem is a nearly cosmopolitan species that oc- curs during the breeding season along most of the eastern coast of North America, including North Carolina. Family Alcidae (auks and puffins) Auks are by far the most abundantly represented birds at Lee Creek Mine both in numbers of specimens and in individuals. Thousands of fossil bones, representing no fewer than 11 spe- cies, have been collected. Three skeletal elements are the most commonly found: the humerus, ulna, and coracoid. We have examined thousands of specimens just of these elements. By contrast, elements of the pelvic limb are scarce, and portions of the skull and mandible are decidedly rare. Greatly compound- ing the difficulties of sorting and identifying the alcid material is its fragmentary nature and the fact that most of the taxa present are the product of a radiation from a single basic stock, so that there are very few qualitative differences between spe- cies in the wing elements. A further complication is the apparently great intraspecific variation in size. We know from the modern Common Murre, Uria aalge (Pontoppidan), and Thick-billed Murre, U. lomvia (Linnaeus), that morphological differences between sibling species may be very subtle (Spring, 1971), and if such pairs of sibling species occur in the fossil record, as seems probable, they would be very difficult to distinguish. At the same time, they would increase the amount of variation observed within a size class. Intraspecific size variation in modem alcids may be correlat- ed with latitude or with oceanic environment, with larger indi- viduals occurring in colder regions, as in the Atlantic Puffin, Fratercula arctica (Linnaeus), and the Great Auk, Pinguinus impennis (Moen, 1991; Bumess and Montevecchi, 1992). If the Lee Creek Mine sample consists in part of wintering birds from different parts of a species' range, this would increase the vari- ability within species and contribute further to the difficulty of separating one species from another. In spite of the difficulties, it has been possible to identify a minimum of 11 species of auks from Lee Creek Mine, nine of which are thought to be from the Yorktown Formation and presumably would have been contemporaneous. Our estimate of four species in the great radiation of Alca is almost certain- ly too low, and if there were sibling species of the same gen- eral size and if more taxa are included at the high and low ends of the size variation, as we suspect, then the number of species of Alca might well double. The fossil record of alcids at Lee Creek Mine is so extensive that the absence of certain taxa, such as Uria Brisson and Cepphus Pallas, must be con- sidered as reflecting reality rather than being a product of sampling bias. The compatibility analysis of the family by Strauch (1985) and the phylogenetic analysis by Chandler (1990b) were con- sulted, but many of the conclusions in the latter unpublished work seem untenable, and we have not considered them fur- ther. Both the American Ornithologists' Union (1998) and Strauch (1985) divide the Alcidae into tribes while omitting any division at the subfamilial level. Because there are no crite- ria for determining whether characters are of subfamilial or tribal value, the next level for grouping genera below the fami- ly level is the subfamily, so we list the subdivisions of the Al- cidae at that level. Subfamily ALCINAE (auks) This subfamily has usually been considered an Atlantic radi- ation and includes the Dovekie (Alle), the murres (Uria), and the auks (Alca, Pinguinus, and fossil relatives). Rather unex- pectedly, however, the genus Uria, at least in its modem as- pect, appears to have originated in the Pacific, as it is entirely absent at Lee Creek Mine and in all other Atlantic deposits that have yielded alcids older than Pleistocene. The only fossil that is truly referable to Uria that has been discovered so far in the 270 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Atlantic is the type of U. affinis (Marsh, 1870), which is only about 12,000 years old (see Ray and Spiess, 1981), whereas two species of Uria have been named from the late Miocene of California (Howard, 1981, 1982). Thus, despite its similarities in plumage to Alca, the genus Uria appears to have had a long, separate history in the Pacific and is probably only distantly re- lated to Alca. The genus Miocepphus Wetmore, 1940, known from the Chesapeake Group in Maryland and Virginia, as well as from Lee Creek Mine, has nothing to do with the genus Cepphus (see Howard, 1978; Olson, 1985d:184), but it is closely related to and possibly ancestral to the Alca radiation. Genus Miocepphus Wetmore, 1940 Miocepphus mcclungi Wetmore, 1940 Material.—Right coracoids, USNM 460810, 495602. Left humeri, USNM 495597, 495599; proximal ends of right hu- meri, USNM 177884, 192467; proximal ends of left humeri, USNM 192775, 495598; distal ends of right humeri, USNM 210424, 210526; distal ends of left humeri, USNM 178052, 250674. HORIZON.—Pungo River Formation inferred from similarity to fossils from Calvert Formation. MEASUREMENTS.—Omitted pending revision of genus. Remarks.—Wetmore (1940) described a new genus and species of alcid, Miocepphus mcclungi, from a right humerus collected in the Calvert Formation of Maryland. Another hu- merus from the same locality was later referred to this species (Wetmore, 1943b), and we now have on hand several addi- tional specimens of various elements from the Calvert Forma- tion in both Maryland and Virginia. The specimens listed above from Lee Creek Mine are the size of Miocepphus and compare very favorably with the specimens of M. mcclungi from the Calvert Formation, although the stratigraphic posi- tion of the Lee Creek specimens was not otherwise deter- mined. Miocepphus, undescribed species Material.—Left humerus, USNM 302324; right humeri missing heads, USNM 430948, 495584, 495600; proximal end of right humerus, USNM 192879; proximal ends of left hu- meri, USNM 178015, 192691, 192698; distal ends of right hu- meri, USNM 215499, 242316; distal end of left humerus, USNM 242178. Distal end of radius, USNM 366359. HORIZON.—Pungo River Formation inferred from similarity to fossils from Calvert Formation. Measurements.—Omitted pending revision of genus. REMARKS.—We have recognized a second species of Mio- cepphus in the Calvert Formation, although it is as yet unde- scribed. This was somewhat larger than M. mcclungi, as are the above specimens from Lee Creek Mine. Both species of Miocepphus were rather small alcids, about the size of small individuals of the Black Guillemot, Cepphus grylle (Linnaeus). Miocepphus mcclungi was likened by Wet- more (1940) both to Cepphus and to Brachyramphus. The broad, flattened humeral shaft in Miocepphus, however, is quite unlike the distinctively terete, and probably primitive, shaft of Cepphus. The less-expanded distal end, more distinct pectoralis scar, and higher, more pointed ectepicondylar pro- cess of Miocepphus are other points that set it apart from Cep- phus. Miocepphus is actually very similar to the Atlantic Alca group of auks, although it is smaller, and it may well have been ancestral to some of the Pliocene forms of that assemblage. In any case, it now appears to be quite unrelated to Cepphus, the pre-Pleistocene history of the latter probably having taken place entirely in the Pacific. Genus Alca Linnaeus Figures 8,9 Australca Brodkorb, 1955:25. The great masses of alcid material from Lee Creek Mine come mainly from several species belonging to the same genus as the modern Razorbill, Alca torda Linnaeus, which is con- fined to the North Atlantic. The humeri of Alca are most easily distinguished from those of Uria by having the internal and ex- ternal tricipital grooves of equal width, whereas in Uria the in- ternal groove is noticeably wider. Also, the ectepicondylar prominence extends higher on the shaft and projects farther in a distinct point, whereas in Uria this prominence is lower and more rounded. The premaxilla is definitely known for only one of the extinct fossil species, but this is high and laterally com- pressed, as in Alca, and is unlike the terete, pointed bill of Uria. To attempt to determine how many species might be encom- passed by the range of size variation shown in bones identified as Alca, a principal components (PC) analysis was done using the following four measurements of the distal ends of 566 fos- sil humeri from Lee Creek Mine: greatest distal width (with calipers parallel to axis of shaft); width through the condyles; mediolateral width of shaft just proximal to ectepicondyle; and shaft thickness (palmar-anconal) at the same point. This sam- ple included all specimens then available in which all four measurements were unaffected by wear or breakage. A correla- tion matrix was used, and axes were unrotated. PC-I scores were plotted as a polygon in a density graph, where proportion per standard unit is the number of cases in each interval divided by the standard deviation (s). Factor I was a general-size axis and was the only factor with eigenvalues above 1.0; this factor explained 93% of the varia- tion. Because the eigenvalues for factors II and III were far be- low 1.0, these axes were not interpreted, and instead, PC-I was taken as a measure of size and was plotted as a density polygon (Figure 8). This density graph of PC-I shows at least a trimodal pattern, with each peak (at PC-I=—3.0, -1.5, 0, 2.5) possibly representing the mode of a different species of Alca. NUMBER 90 271 0.7- P ¦d 0.6- 0.5- 0.4- i 0.2- 0.1- FIGURE 8.—Density polygon of factor I scores from a principal components analysis of four measures of the distal end of the humerus of auks, Alca spp., from Lee Creek Mine. Factor I is a general-size axis explaining 93% of the variation. To determine the minimum number of species of Alca that must have been present at Lee Creek, we estimated the maxi- mum expected variance (s2) within species of Strauch's (1985) tribe Alcini, following the rationale developed by Warheit (1992). However, rather than using a parsimony procedure to determine an ancestor's expected variance where the variance of the two recent sister taxa differed (Warheit, 1992), to avoid oversplitting we simply used the maximum variance of species within the clade. The Dovekie (A lie alle (Linnaeus)) was omitted from this analysis because its relationships with the subfamily Alcinae are equivocal. Greatest width of the distal end of the humerus was measured for 56 Common Murres, Uria aalge; 18 Thick- billed Murres, U. lomvia; 27 Razorbills, Alca torda; and 203 Great Auks, Pinguinus impennis. Sexes were pooled for each species because size dimorphism between the sexes of extant alcids is minimal (Storer, 1952; Livezey, 1988), and the sexes could not be determined for the specimens of Pinguinus. For distal width of the humerus, variances of taxa thus deter- mined in the subfamily Alcinae were s2=Q.33 (Uria aalge), i2=0.38 (U. lomvia), s2 = 0.5\ (Pinguinus impennis), and s2=0.51 (Alca torda). The maximum variance expected for any fossil species within that clade (assuming no evolution in that trait between ancestors and descendents; Warheit, 1992) is therefore 0.51, that shown by Pinguinus impennis and Alca tor- da; Alca torda also is the only living representative of the en- tire radiation of Alca. To determine how best to assign individual specimens to each of four species, a K-means cluster analysis of the four measurements of the distal end of the humerus used in the pre- viously mentioned principal components analysis was done, using K=4 clusters. In addition, K=3 and K=5 analyses were done to determine if these would provide better clustering. All analyses resulted in highly significant F-ratios. With K>3, cluster 3 includes a group with the variance of s2=0.96 for the distal width of the humerus, which is considerably larger than that for any extant species in the Alcinae, and the within- groups sum-of-squares (ss) is large (245.18). With K=5, on the other hand, cluster 5 has a very small variance (s2-0.\ 1), and clusters 1, 2, and 5 have largely coincident ranges and similar means, the range of one of the clusters falling entirely within that of another. With K=4, however, the standard deviations of the distal widths of the humeri for clusters 1 through 4 were 0.32, 0.20, 0.21, and 0.45, respectively, thus better fitting the predicted level of variation for any given species in this clade. In addition, there was less overlap between ranges of each group than if five clusters were used, and the within-groups sum-of-squares is not much larger than for K=5 (s.sK=4=162.19; ss^=5= 112.6). Also, four species each with a variance of 0.51 would account for most of the total variance shown by the distribution of distal humerus width. Therefore, based on the four measurements of the distal end of the humer- us, fossils of Alca from Lee Creek Mine are statistically best considered as belonging to four species. Species intervals as defined by the cluster analysis with K=4 are shown on a histogram of distal humerus width of 621 fossil specimens of Alca (Figure 9). For the entire distribution the mean=13.5, range=9.0-17.2, s=l.46, and the distribution is strongly skewed to the left (skewness=-0.58) but is not kurtot- ic (kurtosis=0.07). The skewness simply reflects the smaller number of specimens of the smallest species of Alca at Lee Creek Mine. Alca aff. torda Linnaeus Figure 1 Oc.d MATERIAL.—Associated Specimens: Left humerus and ul- na, USNM 321314; proximal two-thirds of scapula, anterior end of left coracoid, left humerus lacking most of both ends, and alar digit, USNM 252414. Individual Elements: Anterior portion of sterna, USNM 215778, 495678. Right coracoids, USNM 177758, 179258, 181070, 306263, 367014, 495625, 495626; left coracoids, USNM 308158, 321264, 495596, 495628, 495629. Right hu- meri, USNM 257519, 446653, 446657, 446658, 446686, 446691, 495589, 495668; right humeri lacking part of proxi- mal ends, USNM 495591, 495592; proximal half of left hu- merus, USNM 193252; distal ends of right humeri, USNM 193259, 250772; distal half of left humerus, USNM 178032. Right ulnae, USNM 193184, 215809, 448898, 448904, 448908, 448909; left ulnae, USNM 368515, 446644; left ulna lacking part of proximal end, USNM 495594; proximal half of left ulna, USNM 495595. Right carpometacarpi, USNM 178072, 368532; left carpometacarpus, USNM 193292; right carpometacarpus missing minor metacarpal and part of distal end, USNM 495593; proximal ends of right carpometacarpi, USNM 183421, 215586, 215836; proximal ends of left car- pometacarpi, USNM 192731, 367107. Right femur, USNM 272 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY 460797; left femur 430949; proximal half of left femur, USNM 460798; distal end of right femur, USNM 460805; dis- tal end of left femur, USNM 459393. Proximal ends of left ti- biotarsi, USNM 179243, 495641; distal ends of right tibiotar- si, USNM 366660, 366712; distal ends of left tibiotarsi, USNM 215862, 495638-495640. Left tarsometatarsus, USNM 495633. HORIZON.—Yorktown Formation. Measurements.—See Figure 9. Remarks.—The smallest size-class of Alca at Lee Creek Mine is similar to and probably contains the ancestral form of the modem Razorbill, Alca torda. Some of the fossil specimens grouped here seem too small and gracile to be included in the same species, and further analysis may make it possible to dis- cern a smallest species that could be described as new. This will need to be carefully distinguished from the much earlier species of Miocepphus that are mixed in among the Yorktown alcids, so we have postponed the attempt until undertaking a revision of the Miocene auks. Alca ausonia (Portis, 1891), new combination Uria ausonia Portis, 11 tration]. Figure \0e; Plate 18a,/ : 195 [nomen nudum]; 1891:15 [description and illus- MATERIAL.—Because of the abundance of this species at Lee Creek Mine, only the most important and best preserved elements are listed. Associated Specimen: Left coracoid and distal end of left humerus, USNM 495623. Individual Elements: Partial cranium, USNM 430925. Right mandibular articulation, USNM 206627; left mandibular articulations, USNM 177804, 460790. Anterior portion of ster- na, USNM 178152, 496140^96146. Symphyseal portion of furcula, USNM 193317. Right coracoids, USNM 177759, 181077, 192718, 192841, 193075, 210498, 215779, 250666, 252300, 308159, 366741; left coracoids, USNM 177751, 179242, 192452, 193145, 206464, 215512, 215852, 242332, 248515, 248575, 366363, 367007, 367155. Proximal ends of scapulae, USNM 192608, 193353, 215480, 495674, 495675. Right humeri, USNM 181038, 215443, 366571, 366584, 446661, 446670; left humeri, USNM 179220, 183425, 275870, 368479, 368480, 446685, 446698, 495670-495672. Right ul- nae, USNM 192706, 192983, 193411, 206353, 210459, 306258, 446545-446547, 446553, 446556, 446568, 446578, 446599, 448902, 448903; left ulnae, USNM 178173, 183432, 183433, 183490, 215723, 215905, 242223, 242300, 252331, 302357, 302364,366567,446622,496160; pathological left ul- na, USNM 446647. Right carpometacarpi, USNM 178109, 178182, 183463, 192866, 215673, 215855, 250720, 306313; left carpometacarpi, USNM 177784, 178053, 181102, 192679, 192948, 193399, 236873, 241379, 446536. Synsacra, USNM 275868, 496122-496128. Right femora, USNM 275789, 446702,446703; left femora, USNM 446721-446723, 460799. Proximal ends of right tibiotarsi, USNM 366561, 495645, 495646; proximal ends of left tibiotarsi, USNM 206394, 206461, 495642^195644; distal ends of right tibiotarsi, USNM 178137, 206569, 366401, 368477; distal ends of left tibiotarsi, USNM 177902, 206440, 210496, 241396, 252311, 256251. Right tarsometatarsi, USNM 183468, 321272, 446728, 446729, 446731, 495635; left tarsometatarsi, USNM 177812, 179232, 183491; incompletely ossified right tarsometatarsus, USNM 206329; incompletely ossified left tarsometatarsus, USNM 206393. HORIZON.—Yorktown Formation (USNM 178137, 179220, and 183425 from basal Yorktown Formation as determined from foraminifera in matrix). Measurements.—See Figure 9. Remarks.—This species was hitherto known only from the holotype, the distal two-thirds of a left humerus from the Pliocene at Orciano Pisano, about 25 km SSE of Pisa in the province of Pisa, Italy. Although Portis's 1888 publication is usually cited as the original description (e.g., Lambrecht, 1933; Brodkorb, 1967), the species is stated only to accord well in size and characters with Uria rhingvia (sic) Briinnich (= U. aal- ge), which does not diagnose it in any way. Th