Permian Brachiopods of West Texas, I G. ARTHUR COOPER and RICHARD E. GRANT SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY • NUMBER 14 SERIAL PUBLICATIONS OF THE SMITHSONIAN INSTITUTION The emphasis upon publications as a means of diffusing knowledge was expressed by the first Secretary of the Smithsonian Institution. In his formal plan for the Institution, Joseph Heniy articulated a program that included the following statement: "It is proposed to publish a series of reports, giving an account of the new discoveries in science, and of the changes made from year to year in all branches of knowledge." This keynote of basic research has been adhered to over the years in the issuance of thousands of titles in serial publications under the Smithsonian imprint, commencing with Smithsonian Contributions to Knowledge in 1848 and continuing with the following active series: Smithsonian Annals of Flight Smithsonian Contributions to Anthropology Smithsonian Contributions to Astrophysics Smithsonian Contributions to Botany Smithsonian Contributions to the Earth Sciences Smithsonian Contributions to Paleobiology Smithsonian Contributions to Zoology Smithsonian Studies in History and Technology In these series, the Institution publishes original articles and monographs dealing with the research and collections of its several museums and offices and of professional colleagues at otiier institutions of learning. These papers report newly acquired facts, synoptic interpretations of data, or original theory in specialized fields. These publications" are distributed by mailing lists to libraries, laboratories, and other interested institutions and specialists throughout the world. Individual copies may be obtained from the Smithsonian Institution Press as long as stocks are available. S. DILLON RIPLEY Secretary Smithsonian Institution INTERNATIONAL BOOK YEAR • 1972 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY • NUMBER 14 .i&- '••^sSSS^ri^i'^6^ :" *"^ i^ja/SIt* aa , i ' - • - • j . i - --- , — a..~*> IUM'V _ . ^ . « . , - . -*^_~*. -- - j « — •• r~M iL FIGURE 9.—South front of the Lenox Hills on Decie Ranch showing the formations and members (CM —Cathedral Mountain Formation with Third Limestone Member of Leonard Formation of P. B. King above W, DR=zDecie Ranch Member of Skinner Ranch Formation, LH = Lenox Hills Formation consisting of conglomerate below and shale above, lying under Decie Ranch Member ledge, PT = Poplar Tank Member of Skinner Ranch Formation, SP = Sullivan Peak Member of Skinner Ranch Formation, W = Wedin Member of Cathedral Mountain Formation with abundant Institella; see Plate 7: figure 3). tioned above, and it measures only 40 to 70 feet in places. Nevertheless, it maintains its conglomeratic characteristic throughout its extent in the Lenox Hills (Plate 5: figure 4; Plate 7: figure 3; Plate 12: figure 2). The Decie Ranch Member abounds in fossils. Several localities yielded extensive collections (USNM 707a, 714t). Large sessile types of brachiopods are the commonest fossils: Scacchinella, Derbyia, Geyerella, and Streptorhynchus. Fusulinids are abundant in places: Monodiexodina linearis, Schwagerina hessensis, and S. hawkinsi. Small faults have displaced the Decie Ranch member at both ends of the Lenox Hills. At the southwest end the member is thrown down to the base of the hill along the base of the escarpment facing U. S. Highway 90. Here the waterworn surface of the member displays many Scacchinella in cross-section (USNM 729g). At the northeast end of the Lenox Hills, on the west side of the Sullivan (Yates) Ranch road, a small fault has thrown the Decie Ranch member down to the base of the hill. It was here mistaken for the Wolfcamp (Lenox Hills) because of misidentification of the beds above it (P. B. King, 1931: map) (Plate 2: figure 3). In Hill 5021: Because of structural complications in this hill, the member is exposed along the base of the hill where it occurs as large float blocks. It also occurs in normal sequence above the sandstone and conglomerate of the Lenox Hills Formation. At the west end of the hill, the Decie Ranch Member is about 60 feet thick and forms the saddle between the two westernmost knobs. Here Scacchinella is abundant in several small bioherms. This is the easternmost locality at which the Decie Ranch is overlain by shale of the Poplar Tank Member (Plate 10: figures 1-3). Several large float blocks of considerable thickness occur along the base of hill 5021. One, at least 98 feet high, lies on the west side of the hill just north of the windmill, and another lies just south of the knob at hill 5021. Good collections of Scacchinella and other fossils of this member were taken from these blocks. The western block (USNM 707w) yielded the best silicified Scacchinella in the Glass Mountains (Figure 11; Plate 10: figure 2). East of hill 5021, the Decie Ranch Member cannot be readily distinguished; however, Scacchinella bioherms are common in the base of the Skinner Ranch Formation and for some distance above the base. These undoubtedly are at the level of the Decie Ranch Member even though some faunal differences can be detected. 46 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY '^- ' "' PT -*6*!** 3 f ^ a s FIGURE 10.—Spur at east end of Lenox Hills on the west side of the Sullivan (Yates) Ranch road; the Sullivan Peak Member of the Skinner Ranch Formation in the spur has been dropped into apparent continuity with the Decie Ranch Member of the Skinner Ranch Formation in the main body of the hill (CM —Cathedral Mountain Formation [orange-yellow] shale, DR = Decie Ranch Member, PT —Poplar Tank Member of Skinner Ranch Formation, SP = Sullivan Peak Member; see Plate 2: figure 3). POPLAR TANK MEMBER.—This formation takes its name from Poplar Tank located on the map by P. B. and R. E. King (in P. B. King, 1931) not far west of the Sullivan Ranch (now Yates Ranch) road about 1.5 miles south of the east end of the Lenox Hills. The type section is in hill 5300, where a nearly complete section is exposed. This member is predominantly shale, siltstone, or sandstone, but it is extremely variable with no two sections alike. At the type section, about 250 feet consists of thick beds of shale containing thin beds of conglomeratic limestone, pebble bands, and shell breccias. In the calcarenitic limestones there are 1- to 2-foot-thick layers of pinkish brown chert. Many beds in the upper part of the section contain thin skins of silica one or two inches thick, the surface of which is often covered with peculiar productoid brachiopods, especially the genus Spyridiophora. Occasional conglomerate bands contain cobbles of large size. The shale contrasts strongly with that of the Cathedral Mountain shale, which is usually platy. The Poplar Tank Member forms most of the lower slopes of the Lenox Hills between the two thick, bounding limestones, the Decie Ranch Member below and the Sullivan Peak Member above. An excellent locality, apparently produced in large part by the recent years of drought, appears on the south slope of hill 4801 at the south end of the Lenox Hills. This section, which measures 315 feet, has a thick shale forming the uppermost part, whereas at the type section the top is limestone (Plate 9: figure 2). The Poplar Tank Member is well exposed and 180 feet thick in Dugout Mountain. It is well displayed under the main high knob of the mountain, and also it is well established in the deep ravines on the north side of the crest. East of the type section the Poplar Tank member becomes thinner. In the fault block at the east end of the Lenox Hills on the west side of the Sullivan (Yates) Ranch road, the member is thinner than it is to the west. On the opposite side of the road, about a mile to the northeast in hill 5021, it is still thinner, measuring only 40-50 feet. Farther east in this hill, it wedges out and the Decie Ranch and Sullivan Peak Members come together (Plate 10: figures 1, 2). SULLIVAN PEAK MEMBER.—The uppermost member of the Skinner Ranch Formation is the Sullivan Peak Member, the "First Limestone Member of the Leonard Formation" of P. B. King (1931:64, 66). It is well developed in Dugout Mountain, the Lenox Hills, and in hill 5021. The type section of the member is above that of the Poplar Tank Member in hill 5300 of the Lenox Hills. Here, about NUMBER 14 47 120 feet thick, it consists of detrital limestone, limestone conglomerate, limestone with scattered small pebbles, and ferruginous sandstone. The member is variable, and no two sections have much similarity (Plate 7: figure 3). In Dugout Mountain P. B. King's (1931:133) section 7 gives its thickness as 153 feet, somewhat thicker than it is at the south end of the Lenox Hills. It forms the prominent mass at the top of the mountain and forms bluffs and cliffs in several places. Strongly conglomeratic in the lower part, it is coarse calcarenite or calcirudite in the upper portion (Plate 13: figure 4). The Sullivan Peak Member has its best development in the Lenox Hills, where it varies from about 65 feet at the southwest end to 200 feet at the northeast end. The variation is due to intertonguing of beds and the appearance of bioherms in the section. This member contains the best bioherms to be seen in the Glass Mountains; the bioherms are discussed under a separate heading (Plate 18: figure 1). In the Lenox Hills the lower conglomeratic part of the member is conspicuous at the south end, where boulders up to five feet long have been seen. Furthermore, several bioherms here are surrounded by the coarse conglomerate. At the east end of the Lenox Hills, on the steep hill facing the Sullivan (Yates) Ranch road, the upper part of the member contains numerous bioherms. They are present also at the westernmost knob of hill 5021, where the easternmost exposure of the member is located. East of this point the member merges with the Decie Ranch Member by the wedging out of the Poplar Tank Member. Fossils of the Sullivan Peak Member are like those of the Decie Ranch Member with a liberal inheritance from the Poplar Tank. The bioherms contain Scacchinella, Coscinophora, Tropidelasma, Spyridiophora and Geyerella. Scacchinella becomes rare east of hill 5021. DUGOUT MOUNTAIN MEMBER.—This name was proposed for a sequence of limestones and intervening shales between the top of the Sullivan Peak Member of the Skinner Ranch Formation and the base of the Cathedral Mountain Formation. (Cooper and Grant, 1966). The sequence includes the Second, Third, and Fourth Limestone Members of the Leonard Formation on Dugout Mountain as described by P. B. King (1931:133, section 7). The Sullivan Peak Member represents the First Limestone of .the Leonard on Dugout Mountain, and this can be identified also on the south slopes of the Lenox Hills to the east, as explained above. In contrast, however, the Second, Third, and Fourth Limestones of the Leonard on Dugout Mountain are different lithologically and faunally from limestones similarly numbered by P. B. King in the Lenox Hills. These differences are indicated by the faunas of the Dugout Mountain limestones, which have fossils like those of the Sullivan Peak Member. King's Fifth Limestone of the Leonard Formation on Dugout Mountain has a fauna different from those below, but identical to that of the Second Limestone of the Leonard Formation in the Lenox Hills (see "Leonard Limestones" under "Cathedral Mountain Formation"). Leonard limestones 2—4 on Dugout Mountain stratigraphically underlie the Second Limestone of the Leonard of the Lenox Hills. grp:^t^f?g^*4s^i-_-^r^- -jsoo1 FIGURE 11.—Section through the west knob of hill 5021 (Decie Brothers Hill) showing thinned Poplar Tank Member, which is pinched out on the east side of hill 5021 by the Decie Ranch and Sullivan Peak Members of the Skinner Ranch Formation; note also detached block of Decie Ranch Member (CM —Cathedral Mountain Formation [shale and part of Third and Fourth Limestones of Leonard Formation of P. B. King here as one unit], D—Dimple Formation of Pennsylvanian System, LH —Lenox Hills Formation much thinned, DR —Decie Ranch Member of Skinner Ranch Formation, PT—Poplar Tank Member of Skinner Ranch Formation, SP — Sullivan Peak Member of Skinner Ranch Formation, SR=Skinner Ranch Formation, SR d b = detached block of Decie Ranch Member; see Plate 10: figures 1 and 3). 48 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE 12.—Section in the central part of the Lenox Hills through the hill capped by the closed 5250-foot contour, one-fourth mile west of hill 5300 (CM —Cathedral Mountain Formation with upper sandy and conglomerate limestone beds probably of Third Limestone of Leonard of P. B. King, DR — Decie Ranch Member of Skinner Ranch Formation, LH cgl = conglomerate of Lenox Hills Formation, LH sh —shale of Lenox Hills Formation, PT —Poplar Tank Member of Skinner Ranch Formation, SPzzSullivan Peak Member of Skinner Ranch Formation; sec Plate 7: figure 3) . The Dugout Mountain Member of the Skinner Ranch Formation has its type section in Dugout Mountain, on the north slope of the mountain, along the line of King's section 7 cited above. The sequence that is mostly yellow shale aggregates 514 feet in thickness. The intercalated limestones are mostly cherty calcarenite, dark brown in color, and commonly with siliceous coatings on the bedding surfaces. Fossils are distributed uniformly, but some small bioherms appear with their characteristic brachiopods and other fossils. The brown and somewhat bituminous limestone contains numerous ammonites, including small Perrinites, Propinacoceras, and others. Since some of the beds on the mountain have suffered from dolomitization, the fossils are not well preserved. The ammonites are exceptionally good, however, and these limestones are a prolific source of them. Two conspicuous knobs on the north slope of Dugout Mountain are excellent places to study these upper Skinner Ranch Beds (USNM 700r, 700s). Hill 4811 is an excellent place to study the Second and Third Limestones of P. B. King. The Sullivan Peak Member of the Skinner Ranch Formation occurs near the base of the hill. It is succeeded by 47 feet of chert, sandstone, and shale, on which rests 27 feet of the Second Limestone of P. B. King (USNM 700t). This is followed by 20 feet of yellow shale and chert, and the hill is capped by 15 feet of limestone, which is the Third Limestone of P. B. King (USNM 700s). Spyridiophora appears in the limestone and indicates relationship to the Skinner Ranch Formation. The total thickness of Dugout Mountain Member on hill 4811 is 109 feet (Plate 6: figure 2). The small knob capped by the 5000-foot contour just 0.5 mile northwest of the summit of Dugout Mountain is also capped by the Third Limestone Member of King. This contains about 30 feet of biohermal limestone, some of it dolomitized and considerably altered, but nevertheless containing many fossils, some of them recoverable (USNM 700r) (Plate 10: figure 3). The Fourth Limestone of King proved to be poor in fossils, but we found Spyridiophora in it. The fact that it underlies the Fifth Limestone of P. B. King, which contains the basal fauna of the Cathedral Mountain Formation, and the presence of Spyridiophora lead us to place this limestone with the two below—in the Skinner Ranch Formation. At any rate, there is a considerable faunal change with the appearance of the Institella fauna in the Fifth Limestone (=Wedin Member), and it is here that we draw the boundary between the Skinner Ranch and Cathedral Mountain Formations in Dugout Mountain and the Lenox Hills. Between Dugout Mountain and the Lenox Hills is a 2-3 mile gap in the mountains, making it impossible to trace any beds from one mountain to the other. It is our belief that the thin limestones of the Dugout Mountain Member either pinch out east of Dugout Mountain or unite with the Skinner Ranch Formation in the same manner that the Decie Ranch and Sullivan Peak Members NUMBER 14 49 LENOX HILLS DUGOUT MOUNTAIN i . i . i FICURE 13.—Diagram shows probable correlation of Dugout Mountain Member of Skinner Ranch Formation within the Skinner Ranch (numbers 1-5 = P. B. King's Leonard Limestones in both sections, DRnDecie Ranch Member of Skinner Ranch Formation, PT = Poplar Tank Member of Skinner Ranch Formation, SP = Sullivan Peak Member of Skinner Ranch Formation [formerly- First Limestone Member of Leonard of P. B. King], W=Wedin Member of Cathedral Mountain Formation [Fifth Limestone Member of Leonard of P. B. King on north side of Dugout Mountain, but Second Limestone Member of Leonard of P. B. King in southwest end of Lenox Hills]; measurements and thicknesses of Dugout Mountain section are mainly from King's [1931:132] section 6 on west side of Dugout Mountain). 50 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY merge to the east. It is suggested that the shales intervening between the limestone beds become thin and are overlapped eastward. The possibility that the Dugout Mountain Member is eroded from the top of the Skinner Ranch Formation cannot be overlooked, but we have no evidence that this is the case. We do, however, have evidence from the sections to the east, that overlap of intervening shales and coalescence of beds does take place. This is true in the cited instance of the Skinner Ranch Formation, and it is known to take place in the Cathedral Mountain numbered limestones and the Word numbered limestones, the latter coalescing east of the Word Ranch near Split Tank. Another alternative is suggested by an occurrence of ammonites at the south end of the Lenox Hills (USNM 730k, 736d) identical to those in the Dugout Mountain Member. This ammonite bed is part of the Sullivan Peak Member just below the Cathedral Mountain Formation. The rock containing the ammonites is lithically like that of the limestone of the Dugout Member. Possibly this ammonite bed represents P. B. King's Fourth Limestone, or it may represent the united limestones 2 to 4 or either the Second or Third Limestones, with the others missing. At any rate it is a remnant of the Dugout Mountain Member. SKINNER RANCH FORMATION UNDIVIDED.—The type section of the Skinner Ranch Formation in the northwest knob of Leonard Mountain is 509 feet thick and consists mostly of calcarenite with very little shale but some fairly thick, pinkishbrown chert bands. The lower part of the Skinner Ranch Formation in Leonard Mountain has numerous bioherms containing Scacchinella just east of the type section. These can be followed along the dip up the mountain to the east for a considerable distance and furnish good specimens. Higher in the section, the Skinner Ranch Formation contains few bioherms, but it is mostly calcarenite in thin or thick beds, some of those near the top having a thin skin, an inch or two in thickness, of. dark brown siliceous material usually with abundant fossils, especially Spyridiophora and Glyptosteges. These peculiar and readily recognized productoids are characteristic of the Skinner Ranch Formation. Another fossil of importance is Torynechus (formerly Uncinuloides of R. E. King), the stenoscismatid with costellate exterior and flatly truncated anterior. This genus occurs in some abundance at the top of Leonard Mountain in the saddle just east of the top of the northwest knob. The crest of the knob in the type section is composed of 41 feet of extremely massive and coarse calcarenite with occasional small pebbles. This massive unit was not recognized on the southeast side of the mountain. It contains Institella and is now placed in the overlying Cathedral Mountain Formation (Plate 2: figure 1; Plate 6: figure 1). West of Leonard Mountain the undivided Skinner Ranch appears as bluffs and cliffs along the east face of hill 5280. The lower part of the secw E ^ J p f e W v ••'•••••• Zi*-L±l • ^ - ; - ^ V . ^ V - w . ; . v ^ : ^ - - . •,-.-,T,v, ••••••••.••••• •••••• .M:.,?. ;.•••-.:•- FIGURE 14.—South side of Leonard Mountain showing formations (G = Gaptank Formation, CM = Cathedral Mountain Formation, LH=Lenox Hills Formation [mostly shale], SR= Skinner Ranch Formation; see Plate 2: figure 1). NUMBER 14 51 /SCH i ***!-^ ^ Sr.jtytjftIL_j "* -"• - . \ .-^ «- ..' FIGURE 15.—East side of Leonard Mountain showing deep amphitheater and formations and distribution of two important fossils (CM —Cathedral Mountain Formation, a thin skin of shale and limestone on top of mountain, DOL = large dolomite mass composed of Lenox Hills and Skinner Ranch equivalents, G=zGaptank Formation [limestone and shale], LH=zLenox Hills Formation consisting of lower conglomerate and upper limestone on this side of mountain, SCz=Scacchinella beds of Skinner Ranch Formation that can be traced into dolomite, SCH ^Schwagerina crassitectoria Dunbar and Skinner, SRz= Skinner Ranch Formation; see Plate 6: figure 3) . tion is strongly conglomeratic, but the upper part is calcarenite. Scacchinella is rare in this hill, but it was found at the bottom and near the middle of the section. At the north end of this hill, most of the Skinner Ranch Formation has been dolomitized, and fossils or other details are difficult to see. At the top of the hill (USNM 723h) Spyridophora, Glyptosteges, and Torynechus occur in the uppermost layers as in the type section. The Skinner Ranch Formation is undivided at the east end of hill 5021. Good exposures of the formation also occur in the low hills between hill 5021 and hill 5280. At the west end of the former hill, the Skinner Ranch is differentiated into its three members, but the shale of the Poplar Tank Member wedges out in the middle part of the hill. At the top of the Skinner Ranch Formation in hill 5021, Spyridiophora and Torynechus were found in the upper beds as in Leonard Mountain and hill 5280. These fossils are therefore good guides to the upper part of the Skinner Ranch Formation (Plate 10: figure 1). The easternmost full section of the Skinner Ranch Formation is on the east side of Leonard Mountain, where a section of about 410 feet may be pieced together. Understanding of this section is marred by dolomitization of parts of the section. For example, the Scacchinella beds of the lower Skinner Ranch are well exposed in and on the side of the ravine that forms a broad amphitheater on the northeast side of the mountain. Here one can walk on the Scacchinella bed almost continuously until it merges with the lower part of the thick dolomite that forms the most conspicuous cliff on the southeast side of the mountain. On the southeast nose of Leonard Mountain, much of the dolomite belongs to the Skinner Ranch Formation. It is interesting to note discovery of the large gastropod Omphalotrochus in the lower part of the dolomite, a fossil abundant in the Scacchinella beds on the northeast side of the mountain and elsewhere to the east and north, especially in the Hess Ranch Horst. Part of the sequence that P. B. King (1931:62) described as Hess on this side of Leonard Mountain belongs to the Skinner Ranch Formation (beds 6 and 7, in section 17). This includes most of the thick dolomite and the overlying limestones exposed for about 0.25 mile from the edge of the mountain to the lowest of the shale of the Cathedral Mountain Formation, which in turn is exposed on the knob west of bench mark 5860. The Skinner Ranch Formation forms much of the steep bluff on the south side of Leonard Mountain, and the upper beds are well displayed below the knob just west of the bench mark on which the base of the Cathedral Mountain Formation is exposed. Skinner Ranch beds are also exposed on the north side of the mountain in the upper slopes. An isolated knob circumscribed by the 4750-foot contour 0.85 mile northwest of the Hess Ranch house is composed of biohermal beds and calcarenites with goniatites. Spyridiophora and Glyptosteges are present, indicating the upper part of the Skinner Ranch Formation. Three other exposures of the upper part of the Skinner Ranch appear in small fault blocks northeast of the Hess Ranch. The first of these is 1.55 miles northeast of the Hess house on the west spur of hill 5726. Here a small block of Skinner Ranch with Spyridiophora is overlain by Cathedral Moun52 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY tain Formation with Institella. A second, smiliar block is located 2.6 miles northeast of the Hess house. The third occurrence is in the elliptical hill on the east flank of the Hess Ranch Horst, exactly 3 miles northeast of the Hess house. This hill is composed of Skinner Ranch limestone, lying against the Lenox Hills Formation of the horst, and the igneous instrusive. Here, too, the Skinner Ranch is overlain by Cathedral Mountain limestone with Institella, which forms the east slope of the hill near the base. The lower part of the Skinner Ranch is well displayed just north of the Hess house and at the north base of hill 5816, and it forms the long north slope of hill 5305. In the exposure north of the Hess Ranch house, the section is greatly complicated by dolomitization and faulting. Overlying the Lenox Hills Formation is a veneer of the Skinner Ranch Formation perhaps 50-60 feet thick at the north end near the Hess pasture gate. Here the Skinner Ranch contains many bioherms abounding in Scacchinella, inter-biohermal detritals containing Omphalotrochus, and, in the higher part, the fusulinid Schwagerina crassitectoria. Dolomitization of the section has destroyed some of the fossil evidence. Fortunately, however, the dolomitization is not complete and patches of limestone within the dolomites contain good fossils of the lower Skinner Ranch. An undolomitized patch occurs at the very highest part of the hill (knob capped by the 5050- foot contour ( = USNM 71 Id). The dolomite with occasional limy patches occurs on the north side of the hill for about 0.75 mile, and all can be assigned to the lower Skinner Ranch Formation. An almost exact duplicate of the hill north of the Hess house occurs in hill 5305, the westernmost hill of the Hess Ranch Horst. Dolomitization is only a minor complication in this hill, which is an excellent place to collect the fossils of the lower part of the Skinner Ranch Formation. The south slopes of hill 5305 reveal the conglomerate and overlying limestone of the Lenox Hills Formation. Near the top of the hill appears an interval of dolomite with patches of limestone. This layer consists of dolomitized bioherms and undolomitized patches, which yield fossils of the lower part of the Skinner Ranch Formation. The north slope of the hill is completely made up of this formation; about 200 feet is exposed from the top of the hill to the base. Besides the characteristic Scacchinella, the lower and middle parts of the slope contain abundant Omphalotrochus and Schwagerina crassitectoria. This lower part of the Skinner Ranch Formation can be traced to the north base of hill 5816, highest and largest hill of the horst chain. Here the biohermal beds are not complicated by dolomitization and Scacchinella, and other bioherms are exhibited to perfection. About the same thickness appears here as on hill 5305. Unlike hill 5305, however, the Scacchinella beds cannot be traced to the south, up the long north slope of hill 5816, because they either pass into the facies of the Hess Formation or they are eroded off the lower and middle slopes of hill 5816. The Hesstype rocks at the top of the hill and knobs to the east are clearly Skinner Ranch equivalent because they contain Schwagerina crassitectoria. This relationship is discussed below. HESS FORMATION The Hess Formation is a great body of rock in the eastern part of the Glass Mountains, best exposed from the Hess Ranch house and eastward. It consists of a variety of lithic types that are mostly unfossiliferous and contain few fossils except fusilinids. The Hess Formation generally has been classified with the Leonard Series, but lately it has been discovered that it is wholly, or at least in part, pre-Leonardian, depending on one's point of view. Udden (1917:43) named the Hess Formation for exposures on the Hess Ranch, where it is well exposed. Udden explained that this formation overlies the Wolfcamp Formation unconformably and that it is separated from the underlying beds by a cobble conglomerate. He also believed that the formation was overlain unconformably by the Leonard Formation. Udden was uncertain as to the westward extent of the Hess Formation, and he suggested that it might be present in Leonard Mountain and perhaps in the ridge west of Iron Mountain. If present, Udden suggested that the formation would be considerably thinned in that direction. In spite of these doubts, the Hess Formation was indicated on his map across the entire mountain front. NUMBER 14 53 P. B. King and R. E. King (1929:126) pronounced the lower half of the beds on Leonard Mountain (referred to as the Leonard Formation by Udden) to be in reality Hess, and stated the Hess Formation was represented in the hills to the west of Leonard Mountain. They also recognized two facies of the Hess Formation: (1) the East Facies, constituting the fine-grained limestones and dolomites from Hess Ranch eastward, and (2) the West Facies, massive, usually non-dolomitic gray granular limestone. The two facies merged about 2 miles northeast of the Hess Ranch house. Much of the upper part of the section in the Hess Ranch Horst was also considered by P. B. King (1931:56, 141) to be Hess, and he described a mingling of the two facies on the horst. In 1932 P. B. King (page 347) indicated that the Leonard shale in the west end of Leonard Mountain passed into limestones of the Hess Formation (West Facies) and the latter into the East Facies in the vicinity of Hess Ranch (Plate 8: figure 1). P. B. King (1934: 730) next proposed that the Hess Formation be restricted to beds in the eastern part of the mountains, and relegated them to the rank of a member. In 1938 P. B. King (page 98) repeated his findings that the Leonard shale is separated from the Hess thin-bedded limestone by "limestone reefs" and that the Hess is best regarded as a member of the Leonard Formation. Jarvis (1957:6), as a result of studies of the Wolfcamp and Hess Formations, concluded that the conglomerates in the western part of the mountains, which were referred by P. B. King to the Wolfcamp, were, in fact, more like Hess conglomerates and should be referred to that formation. He also recommended that the Hess be returned to its original status as a formation. Ross (1959:299) redefined the Wolfcamp, establishing the Neal Ranch Formation for the shaly lower part and the Lenox Hills Formation for the upper conglomerate. The Lenox Hills conglomerates were traced by him eastward across the mountain front and were found to lie above the Neal Ranch Formation and to include the Hess conglomerate of the type Hess Formation. Later Ross (1960) described the fusulinids of the Hess Formation, but he continued to call it the Hess Member of the Leonard Formation. Cooper and Grant (1964) recommended that the Hess be regarded as a separate formation, and they indicated its lateral relationship to the Skinner Ranch Formation. The foregoing remarks on the history of this formation show that it originally was regarded as lying between the Wolfcamp and Leonard Formations, but later it was thought to be the equivalent of the Leonard shale but separated from it by limestone reefs. These views more recently have been shown to be erroneous; the Hess conglomerate is actually Wolfcampian, a part of the Wolfcamp Series. Ross, by his work on the fusulinids, indicated that the Hess for about 400 feet above the conglomerate actually contains fossils of the Wolfcamp Series. Our findings, detailed below, corroborate these views, which are further strengthened by extensive faunal evidence. In view of these discoveries, our discussion of the Hess will be in two parts: (1) Hess Formation (Lenox Hills equivalent), and (2) Hess Formation (Skinner Ranch equivalent). LENOX HILLS EQUIVALENT.—On the basis of the history just reviewed and by way of orienting the reader, it is necessary to offer some preliminary remarks before examining details of the Hess Formation. It is important to note that much of Udden and Bose's Hess in the western part of the mountains is actually the Lenox Hills Formation of Ross and is the Hess alluded to by Jarvis (1957:6). The band of Hess (western facies) appearing on P. B. and R. E. King's map (in P. B. King, 1931) along the base of the Lenox Hills overlies the Hess of Bose and Udden and is the Decie Ranch Member of the Skinner Ranch Formation. The Hess conglomerate on Leonard Mountain is part of the Lenox Hills Formation of Ross and is the equivalent of the Hess conglomerate of the type region. Therefore, most of the Leonard Mountain sequence is not Leonard, although it was part of the Leonard type section. This complication is explained under the discussion of the Leonard Formation. In discussing the part of the Hess Formation that is equivalent to the Lenox Hills Formation, it is best to describe the rocks in the type section. The interval in the western part of the mountains lies between the Gaptank Formation below and the Decie Ranch Member of the Skinner Ranch Formation above (the latter containing abundant Scacchinella or its cohort Schwagerina crassitectoria). In the type section of the Hess, therefore, 54 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY WOLF CAMP HILLS HESS RANCH HORST LIMESTONE FIGURE 16.—Correlation of the Lenox Hills Formation in the Hess Ranch Horst with that in the Wolf Camp Hills area (Gapt = Gaptank Formation, CM = Cathedral Mountain Formation) . NUMBER 14 55 the Lenox Hills Formation equivalent will be defined by the beds between the unconformity on the Gaptank Formation or the Neal Ranch Formation and the base of the Schwagerina crassitectoria zone. The conglomerate at the base of the Hess Formation can be traced east and west from the Wolf Camp Hills. West of these hills it can be found almost continuously at the base of the mountain front to the Hess Ranch, where the outcrops end. Just north of the Hess Ranch house it is well exposed at the base of the hill, where it consists of 187 feet of limestone cobble conglomerate. In Leonard Mountain, the next outcrop area to the west, this conglomerate appears above Gaptank limestone. It is the Hess conglomerate of King, identified by Ross as the Lenox Hills conglomerate. East of the Wolf Camp Hills the Hess conglomerate is traceable to the east side of the road (U. S. Highway 385) at Stockton Gap. It is not continuous to the east, and it thickens and thins, attaining a thickness of 200 feet in places in the eastern part of the belt. It is missing from the section in the somewhat conical foothill 1.5 miles northeast of hill 4852 at the east end of the Wolf Camp Hills. In this eastern range the conglomerate is composed of rounded limestone cobbles. Overlying the conglomerate is a varying sequence of red and green shale, thin limestone, and sandstone, with very few fossils. West of the Wolf Camp Hills the section contains progressively less shale, but east of the hills shale beds become thick and important. Shales and thin beds of dolomite attain a thickness of 300 feet east of the Wolf Camp Hills. The Lenox Hills equivalent part of the Hess Formation is limited at the top by thin-bedded limestone abounding in the fusulinids Schwagerina crassitectoria Dunbar and Skinner and S. guembeli Dunbar and Skinner, which characterize the base of the Skinner Ranch Formation on the Hess Ranch Horst and Leonard Mountain. The Lenox Hills equivalent of the Hess Formation thus proves to consist of variable conglomerates at the base followed by red and green shales, some thick sandstone, and all with thin limestone and dolomite beds intercalated. This part of the section ranges from 300 feet to about 460 feet thick. The boundary between Hess-Lenox Hills equivalent and Hess-Skinner Ranch equivalent is a paleontologie one, not a lithic break. SKINNER RANCH EQUIVALENT.—The Hess Formation above the Lenox Hills equivalent is a great mass of thin-bedded limestone and dolomite with some heavy, massive bands and occasional layers of conglomerate, shale, and sandstone. The limestone is described by Ross (1960:120) as silty biomicrosparite in the lower 200 feet and biosparite in the next 600 feet, while the uppermost 400 feet are mostly "limestone which is recrystallized biosparite or biomicrosparite." Tongues of shale, siltstone, and silty limestone appear in the eastern part of the mountains. Two important traceable beds are the "double ledge" and the "fossil bed" of P. B. King (1931:60). The formation is thickest in the eastern part of the hills, where it attains 1600 feet. The section thins to the west and is about 1290 feet thick about 2 miles east of the Hess Ranch house. Only two parts of this great mass of limestone concern this monograph. One is the lower fusulinid zone, and the other is the "fossil bed" of P. B. King (Plate 8: figure 1). Schwagerina crassitectoria Zone: This zone is generally about 200 feet thick and contains the two fusulinids Schwagerina crassitectoria Dunbar and Skinner and S. guembeli Dunbar and Skinner. Ross (1960:121) noted that, in the lower part of their range, the two species are morphologically similar and occur together. Higher in their range they become distinct, 5. guembeli increasing in size and rotundity and S. crassitectoria becoming more elongated and having less complicated axial deposits. Generally, at these levels, they are no longer associated in the same beds. Ross states that, in the upper part of its range, S. guembeli is commonest in biosparite and biomicrosparite whereas S. crassitectoria in the upper part of its range occurs in biomicrosparite. This suggests to Ross that the two gradually became adjusted to different environments. In the eastern Glass Mountains S. guembeli ranges a few feet into the succeeding zone of Parafusulina allisonensis Ross. The significance of this thick zone of S. crassitectoria and S. guembeli is the establishment in the Hess Formation of the level of the Scacchinella beds, approximately the level of the Decie Ranch Member of the western part of the Skinner Ranch Formation. As shown above, S. crassitectoria occurs 56 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY with Scacchinella on the northeast side of Leonard Mountain, on the north slope of the Hess Ranch Horst, and north of the Hess Ranch. The transition of biohermal basal Skinner Ranch into Hess facies occurs on the long north slope of the horst and on the east side of Leonard Mountain. Fossil Bed of P. B. King (— Taylor Ranch Member): This is an extremely important stratigraphic unit because of its abundance and variety of fossils and because of the clue the fossils give to the correlation and relationships of the Hess Formation. It can be traced from east of the Hess Ranch to the eastern part of the mountains. It is one of the key beds used by Ross (1960:120) in describing the stratigraphy of the Hess. In order better to understand this datum and to provide ease of designation, we have introduced the name "Taylor Ranch Member" of the Hess Formation (Cooper and Grant, 1966) and designated the type section on the Bill Neal Ranch on the southwest slope of the hill, capped by the 5750-foot contour, which is located between hills 5767 and 5821 (USNM 716o) (Plate 9: figure 1). Here, the following section was measured: feet Hess Formation, Taylor Ranch Member Fossiliferous light brown limestone with orangebrown chert 5 Cobbly limestone with some shale and abundance of fossils 33 Yellowish conglomerate with smooth matrix and pebbles of limestone up to 3 inches 2 Hess Limestone and Dolomite This member is not developed uniformly in all parts of the hill, because the thick fossil bed loses its abundance of fossils to the east, but the conglomerate is persistent and some fossils usually can be found by diligent search (Plate 9: figure 1). West of the type-locality, just under the nose of hill 5725, there occur thick sponge bioherms that mostly are composed of large Heliospongia and bead-like Girtycoelia. The sponges are accompanied by a variety of brachiopods, gastropods, and pelecypods to constitute one of the most prolific collecting places in the mountains (USNM 702d). Some important species for correlation occur in the sponge bioherms. Spyridiophora, which is rare but in excellent condition, constitutes one of the paleontological links of the Taylor Ranch Member to the upper part of the Skinner Ranch Formation (Plate 19: figure 2). The westernmost occurrence of the member is USNM 702e about 0.5 mile north of hill 5751. There the member contains a fairly prominent but thin shale bed and an abundance of Rhipidomella hessensis R. E. King that is not equalled elsewhere. This locality contains several important species. A cast of Scacchinella found in the vicinity of the sponge bioherms and a brachial valve found northwest of the Conoly Brooks Ranch are very important discoveries. P. B. King (1931:60), besides calling this the "fossil bed," also proposed the term "Perrinites compressus horizon" after the ammonite described by Bose (1917:166). The name is not apt, because the ammonite is excessively rare. Bose had only two specimens, and we have taken only two in all the collecting done in the mountains. Moreover, the species P. compressus Bose is now regarded as a synonym of P. hilli (Smith), the more widely distributed species of the genus. ABOVE THE TAYLOR RANCH MEMBER.—A considerable thickness of limestone and dolomite intervenes between the Taylor Ranch Member and the base of the Cathedral Mountain Formation, which is a conglomerate composed of small pebbles, many of them white quartz. P. B. King (1931) and Ross (1960) describe this interval, but they record considerable variation in the thickness. Both authors agree that the section thins to the west. The maximum thickness reported by P. B. King (1931: 145, section 27) is 410 feet, but Ross (1960:119, section 2) records over 700 feet in essentially the same section at the Conoly Brooks Ranch. Regardless of thickness, the interval lies between the Taylor Ranch Member and the base of the Cathedral Mountain, which is essentially the base of the Institella beds of the Cathedral Mountain Formation. These upper beds of the Hess are, like those below, thin limestone, dolomite, or massive dolomite with many unidentifiable fusulinids. In some places the fusulinids are coated by algal material. Ross (1960:119) records three species of fusulinids from this part of the section: Parafusulina spissisepta Ross, P. brooksensis Ross, and P vidriensis Ross. As explained in our chapter on correlation, these post-Taylor Ranch Member beds are included as a unit with the Hess Formation, as partial equivNUMBER 14 57 alents of the Skinner Ranch Formation. They are overlain by Institella throughout their extent, from near the Hess Ranch to a mile east of Split Tank, and therefore are regarded by us as belonging to the Hess Formation. Moreover, they are lithically like the Hess, and the few fossils found in them are Hess types (USNM 726n). In Leonard Mountain: King (P. B., 1931:62, 140, section 17) identified 746 feet (erroneously recorded as 646 feet) of Hess limestone in Leonard Mountain in 7 beds. This sequence now proves to be referable to the Gaptank, Lenox Hills, and Skinner Ranch Formations. Beds 3-5 in P. B. King's section (1931:62) belong to the Lenox Hills Formation of Ross, bed 3 being the Hess conglomerate (= Lenox Hills Formation) traceable in the hills east of the Hess Ranch, and the limestone forming bed 5 contains characteristic Wolfcamp Series fossils (= part of Lenox Hills Formation). Beds 6 and 7 of this section belong to the Skinner Ranch Formation, better exposed and more characteristic on the southwest side of the mountain. In King's section 17, bed 6 is composed largely of dolomite and is the great "reefy" mass so conspicuous on the southeast nose of the mountain. As explained elsewhere in more detail, dolomitization on Leonard Mountain has caused confusion, leading King to identify it as "east facies of Hess," when it is actually dolomitized west facies of Hess, which, in turn, is really partly Lenox Hills Formation and partly Skinner Ranch Formation. King's Hess section on Leonard Mountain is Hess equivalent, but it does not represent all of the section. A similar situation exists on the Hess Ranch Horst, but there only the lower part of the Hess is involved. In the Hess Ranch Horst: The great mass of the Hess Ranch Horst in hill 5816 rests on the Neal Ranch Formation (all shale at this place) and is capped by beds with Schwagerina crassitectoria. The rocks are thus blocked in by the same boundaries that define the Lenox Hills Formation. P. B. King (1931:56, 141, section 20) recognized 305 feet of Wolfcamp limestone and shale above the "Uddenites shale" ( = Neal Ranch Formation of present usage) and recognized 429 feet of Hess Formation composed of 7 units that included some conglomerate (bed 1), 214 feet of limestone said to be eastern Hess facies (beds 5 and 6), the latter topped by 116 feet of light gray limestone in massive beds (bed 7) resembling "the cliffs on Leonard Mountain, made up of the western facies of the Hess.'' All of these beds now are known to contain Wolfcampian fusulinids and to belong to the Lenox Hills Formation. The horst sequence, assigned to the Wolfcamp and Hess by King, is thus proved to belong to that part of the Hess (of its type section), which includes the interval from the basal conglomerate to the Schwagerina crassitectoria Zone, the part that in the eastern Glass Mountains is mainly red and green shale and sandstone. The passage of these facies of western and eastern expression is difficult to demonstrate lithologically because of unfortunate gaps in continuity, but it can be demonstrated amply and adequately by paleontology. FACIES RELATIONSHIPS.—The facies relationships of the Hess Formation as outlined by King and others are considered almost classic examples in the geological and stratigraphical literature, but they are only partly true. The relationships depicted for Leonard Mountain and the western part of the Hess Ranch area—in which it is indicated that shaly Leonard on the northwestern side of Leonard Mountain passes into reefs on the southeastern side and then into the Hess thin-bedded Old Word Ranch CM FIGURE 17.—Section through the Wolf Camp Hills to Old Word Ranch showing great development of the Hess Formation (GK—Gaptank Formation, NR = Neal Ranch Formation, H = LH z= lower part of Hess equals Lenox Hills Formation of Ross, Hz=Hess Formation, fb = TR = fossil bed of P.B. King equals Taylor Ranch Member, CM = Cathedral Mountain Formation, RC=Road Canyon Formation). 58 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY L E O N A R D MOU N T A IN SKINNER RANCH A=DETACHED BLOCK of LENOX HILLS LIMESTONE CM'CATHEDRAL MOUNTAIN YELLOW SHALE G- GAPTANK ° *° ° ' " 2 - ' / 4 M I L ES LH = LENOX HILLS SR= SKINNER RANCH GAPTANK B=DOLOMITE: part LENOX HILLS and port SKINNER RANCH _:_ :-. : ^ = = E 1 1 / / 1 1 1 / / 1 - - I - 1 V | - K N - 1-1 K c o | » . | o i==l»" = . . l . o l . «..!» = •".•yn< TESNUS FIGURE 18.—Section through Leonard Mountain showing dolomite block (B) and detached block of Lenox Hills Formation (A) (see Plate 6: figure 3; Plate 8: figure 2) limestone—is in part true, but the picture must be carried farther west than Leonard Mountain to get the full perspective. We have shown that the Poplar Tank shaly member is pinched out rather than interfingered into the Skinner Ranch Formation, but, although a few yellow shale beds can be seen in the Skinner Ranch Formation, we found no such relationship as indicated for Leonard Mountain (P. B. King, 1932:347). The Skinner Ranch Formation underlies the main mass of the "Leonard shales" ( = Cathedral Mountain Formation) on Leonard Mountain and to the west. The Skinner Ranch Formation constitutes a great mass of calcarenite, and it is the massive formation of Leonard Mountain that can be demonstrated by paleontological correlation to be the lateral equivalent of the main part of the Hess thin-bedded limestone. The relations of the Lenox Hills Formation to the lower part of the Hess Formation are shown on Leonard Mountain, on the mountain mass east of Hess Ranch, and in the Hess Ranch Horst. The lower 300-500 feet of the Hess east of Hess Ranch and all but the main knobs of the Hess Ranch Horst are equivalent to the Lenox Hills Formation. Demonstrating the equivalency of the Skinner Ranch with the bulk of the Hess Formation is more difficult because a great gap exists between Leonard Mountain and the hills east of Hess Ranch and the north slope of the Hess Ranch Horst. In spite of this gap, concrete evidence of the presence of beds equivalent to those higher in the Skinner Ranch Formation than are found on the horst appears in the small hill 0.85 mile northwest of the Hess house (USNM 709a, 727f, 727m, 727n). There, numerous fossils of the upper part of the Skinner Ranch were taken, including not only brachiopods of significance, but also ammonites. The 100 or 200 feet of lower Skinner Ranch Formation on the north slope of the west end of the Hess Ranch Horst (hill 5305) contains abundant Schwagerina crassitectoria, and the same fusulinid occurs on the summit of the horst (USNM 714h) but in characteristic Hess thinbedded lithology. It is evident that a lithological change takes place up the long north slope of the horst, but also that the mingled lithologies largely have been eroded away (see Figure 26). Leonard Series Like the Wolfcamp Series below it, the Leonard was elevated in rank from the status of a formation to that of a series (Adams et al., 1939). Furthermore, the content of the series has suffered some change from its original definition to its recent reconstitution. TYPE SECTION OF THE LEONARD FORMATION Udden, Baker, and Bose (1916:51) only indirectly designated the type section of the Leonard, stating that the "section characterised here has been measNUMBER 14 59 ured in Leonard Mountain and north of it." They did not state what parts of Leonard Mountain were measured nor how far north they carried the measurements. Udden (1917:46) was more specific in stating that the "formation which makes the greater part of the south face of Leonard Mountain has been given the name of this prominent feature in the landscape north of Marathon." Udden's section started at the west end of the mountain and extended northward, parallel to Gilliland Caynon (Plate 8: figure 2). In 1929 P. B. and R. E. King (page 131) pointed out that the lower half of Udden's type section was Hess and part Gaptank (= Lenox Hills and Skinner Ranch Formations of this report). P. B. King (1931:63) reiterated this view and commented upon "the massive limestones which outcrop on the south face of Leonard Mountain, and which have been shown by the writer to correspond to the upper part of the Hess at its type locality." King, therefore, placed the contact of the two formations (Hess and Leonard) at the natural line of subdivision between the shales above and the massive limestones below. So defined, the only Leonard Formation on Leonard Mountain is that forming the knob 0.25 mile northwest of bench mark 5860 and that appearing on the northwest slope of the mountain. The Leonard thus occupies the north slope of the mountain and the valley and lower slopes of the hills on the north side of the valley. In 1932 P. B. King suggested that the Leonard Formation consisted of three interlocking facies: a shaly sequence in the west, a thin-bedded limestone in the east, and the two separated by reef limestone. King (1934:730) proposed the Hess Formation as a member of the Leonard Formation, a nomenclature that existed until Jarvis (1957) recommended restoration of the Hess as an independent formation (see Figure 18). Ross (1960) did not follow Jarvis' suggestion to reinstate the Hess as a formation, but he continued to regard it as a member of the Leonard. In 1962, however, Ross (1962b) recommended splitting the Leonard Formation into three members, which he designated in ascending order A, B, and C. Member A includes the basal conglomerate limestone (Decie Ranch Member of the Skinner Ranch Formation) and the succeeding shale (Poplar Tank Member) to the base of the first Leonard limestone. Member B includes all of the numbered limestones from the first to the fifth. Member C extends from the base of an extensive sandstone to the base of the Word Formation (Road Canyon Formation) and includes the great mass of Leonard soft shale in the western part of the mountains. CATHEDRAL MOUNTAIN FORMATION Cooper and Grant (1964) revised the Leonard Formation, recommended that the name be restricted to the series designation, and proposed Cathedral Mountain for the siliceous sequence from the top of the Skinner Ranch Formation (= Member A of Ross plus the first Leonard limestone of P. B. King) to the base of the Road Canyon Formation ( = First Limestone of the Word Formation). The Cathedral Mountain Formation thus includes Ross Member B (minus the first limestone) and all of Member C. The Ross members do not form natural bio- or lithostratigraphic units in our opinion; therefore, we have not adopted them. Furthermore, in our treatment of the Leonard Series, we expand it to include the Road Canyon Member of the Word Formation, which now has become the Road Canyon Formation of the Leonard Series. As explained in the chapter "Faunas and Correlations of Glass Mountains Formations," this assignment has a paleontological as well as stratigraphical basis. We also individualize the Fifth Limestone of the Leonard on Dugout Mountain and the Second Limestone of the Leonard in the Lenox Hills as the Wedin Member of the Cathedral Mountain Formation. IN THE TYPE AREA.—The type section of the Cathedral Mountain Formation is on the line of P. B. King's (1931:66) section 12 and includes his beds 19-38, aggregating 1245 feet in thickness. Beds 1—18 of this section belong to the Poplar Tank and Sullivan Peak Members of the Skinner Ranch Formation. The type section, predominantly shale, chert, and sandstone, contains some conspicuous beds of sandstone and limestone. The shale of the Cathedral Mountain, especially the yellow, platy kind is well indurated, silty, or often sandy, cherty, and blocky, very hard and usually a beautiful orange and red, the staining often paralleling the blocky joints and creating interest60 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY L E 0 N A R D < o u. -J O 5 SKINNER RANCH COOPER-GRANT CLASSIFICATION WEST EAST ROAD CANYON CATHEDRAL MOUNTAIN FORMATION WEDIN SULLIVAN PEAK POPLAR TANK DECIE RANCH LENOX HILLS FORMATION CATHEDRAL MOUNTAIN much thinned, faunolly like sequence In west TAYLOR RANCH HESS FORMATION MEMBERS ot ROSS WEST EAST BASE of WORD FORMATION MEMBER C upper CM only MEMBER B MEM BER A MEMBER C Includes all of CM HE SS B FOSSIL BED MEMBER A LENOX HILLS FORMATION L E O N A R D a FIGURE 19.—Comparison of members of the Leonard "Formation'' of Ross with the Skinner Ranch and Cathedral Mountain Formations. ing patterns. The limestones are commonly conglomeratic, the pebbles usually small and composed of quartz or chert, occasionally red or white. In the conglomeratic, calcarenitic limestones, the ammonite Perrinites commonly occurs. The lower orange and pink, hard, indurated, silty shales and cherts form the knobs on the north slopes of the Lenox Hills, and the upper part of the formation lies on the lower slopes of Cathedral Mountain beneath the Road Canyon and Word Formations, which form most of the upper part of the mountain. The valley between the Lenox Hills and Cathedral Mountain is underlain by the Cathedral Mountain Formation. Good exposures appear in low hills and knobs in the valley. One hill composed mostly of sandstone occurs 1.35 miles south of Sullivan Peak. It is unusual because the nearly solid mass of sandstone surrounds a characteristic smooth-limestone bioherm that contains the brachiopod Institella, among other fossils. This is a possible erratic block of the Wedin Member. The contact with the Sullivan Peak Member of the Skinner Ranch Formation is sharp and clear because it consists of yellow Cathedral Mountain shale and chert on light-gray Sullivan Peak limestone in most places in the Lenox Hills. It also occurs in the valley on the south side of hill 4920. The upper contact is more difficult to find because it is covered by slide from the Word Formation. On the south side of Cathedral Mountain, where rock is exposed, the shale of the Cathedral Mountain Formation is in contact with bioherms of the lower Road Canyon Formation. East of the type section at the Clay Slide an excellent exposure of the upper shale can be studied (Plate 7: figure 2). In this vicinity the base of Cathedral Mountain is in contact with the Sullivan Peak Member of the Skinner Ranch Formation at the constriction of the 4750-foot contour on hill 5021. Here the yellow siliceous shale lies on the Skinner Ranch Formation, which forms the saddle. The limestone or dolomite above the Skinner Ranch in the lower Cathedral Mountain King (P. B., 1931: map) has mapped as the First Leonard Limestone. This is actually not the First Limestone, but it is the same as the one marked "Limestone 3-4," which makes hill 4920. The "First Limestone of the Leonard" forms the top of hill 5021 and is the upper part of our Skinner Ranch Formation. The other limestone is a dolomite rather than a limestone and contains abundant Perrinites. The valley between hill 5021 and the Clay Slide is underlain by the Cathedral Mountain Formation. The low hill about 0.35 mile east of Clay Slide, containing interesting exposures of limestone and sandstone, is a good place to collect the fauna of the Cathedral Mountain Formation. Perrinites is common and Institella is present but difficult to find. Fossils also are abundant in a ravine just east of Clay Slide (USNM 707q). The upper beds of the Cathedral Mountain Formation appear in several ravines entering Gilliland Canyon, such as the one just southeast of elevation 4869 on the Gilliland Canyon Road. LEONARD LIMESTONES OF P. B. KING IN THE LENOX H I L L S . - P . B. King's (1931) sections 10-12 in the Lenox Hills all indicate the presence of limestone beds in the Leonard shale, which he designated by numbers 1-4. The First Limestone was separated by Cooper and Grant (1964) as the Sullivan Peak Member of the Skinner Ranch Formation. Inasmuch as this First Limestone is correctable across the Lenox Hills to the west onto Dugout Mountain, it has been inferred hitherto that the other numbered limestones of the two hill masses are the same. Certainly the consecutive numbering in both places encouraged such a view. NUMBER 14 61 4500' SULLIVAN PEAK ROAD CANYON VERTICAL a HORIZONTAL SCALE' Imm. SO)««l FIGURE 20.—Section through the hill capped by the closed 5250-foot contour, just west of hill 5300, to Sullivan Peak in Cathedral Mountain (DR = Decie Ranch, SP = Sullivan Peak, both members of the Skinner Ranch Formation, L-ls. 3 & 4 = Leonard limestones 3 and 4 of P.B. King). Identification of the fauna of King's Second Limestone in the Lenox Hills with the Fifth Limestone of Dugout Mountain dispels this idea. Although we have assigned King's First Limestone Member to the Skinner Ranch Formation, the other limestones remain in the Cathedral Mountain Formation. For the sake of discussion, we do not alter the numbering of King. In the Lenox Hills the Second Limestone is well exposed at several places (USNM 714w, 727p), and it can be followed to hill 5300 where it pinches out. Its thickness is variable and amounts to 20 feet just below the west knob of hill 5300. This limestone, the Wedin Member of the Cathedral Mountain Formation, contains an unusual fauna and is therefore singled out for further discussion. Separated from the Second Limestone by about 80-90 feet of yellow siliceous shale is King's Third Limestone in hill 5300, consisting of 15 feet of small pebble conglomerate and 32 feet of gray limestone that contains numerous specimens of Perrinites. On the north dip slope of the west knob of hill 5300, the limestones contain numerous specimens of Institella in biohermal beds. The conglomerate is variable and thins to the east (Plate 7: figure 3). The Fourth Limestone Member (P. B. King, 1931:135, section 11) is recorded as 12 feet thick in section 11, but on the east side of the Lenox Hills it merges with the Third Limestone to form a combination of sandy limestones that become conglomeratic near the top. This combined limestone is characterized by abundant Perrinites and extends onto the north slope of hill 5021. P. B. King and R. E. King (in P. B. King, 1931) mapped the First Limestone of the Leonard Formation on the northwest slope of hill 5021 and on the next hill to the northeast. It forms a broad belt in the first hill but only a narrow band in the second one. Actually these bands belong to the combined Third and Fourth Limestones of the Leonard rather than to the First Limestone. Partly dolomitized, they abound in large ammonites of the genus Perrinites, a feature quite unlike that of the First Limestone ( = Sullivan Peak Member of the Skinner Ranch Formation). It is now known that the First Limestone unites with the Decie Ranch Member of the Skinner Ranch Formation. This combination appears in the Kings' map as the "Hess Limestone." The Leonard Limestone appearing in the hill between hills 5021 and 5280 is the last appearance of any of the limestones of the Cathedral Mountain. Summarizing the above, we see that the four limestones King individualized by number pinch out to the east by union with lower beds. The First Limestone is the Sullivan Peak Member of Cooper and Grant; the Second Limestone, the Wedin Member, pinches out in hill 5300 near the middle of the Lenox Hills; the Third Limestone unites with the Fourth at the east end of the Lenox Hills and pinches out east of hill 5021. This was misidentified as the First Limestone on the east side of the Sullivan (Yates) Ranch Road. WEDIN MEMBER.—The Second Limestone Member of the Leonard of P. B. King noted above is an important datum for the base of the Cathedral Mountain Formation. It was individualized, therefore, by Cooper and Grant (1966) as the "Wedin Member"—named after the Ava Scribner Wedin No. 1 Well on the Decie Ranch about 2.25 miles southeast of the type section. Since the Decie Ranch was at one time owned by the Wedin family, the name is doubly appropriate. The type section of the member is under the west knob of hill 5300, which is defined by the closed 5250-foot contour. 62 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Here the member is 20 feet of biohermal limestone abounding in Institella and Agelesia. The member pinches out about a mile to the east, but to the west it forms conspicuous ledges on the north slope of the west end of the Lenox Hills. It is also well exposed but somewhat thinner on the northwest side of Dugout Mountain, where it forms a belt almost a mile long. It is the highest rock mapped by P. B. King and R. E. King (in P. B. King, 1931) in the Dugout Mountain sequence, where it was called the Fifth Limestone Member of the Leonard. The member thus extends from the center of the Lenox Hills westward to the northwest side of Dugout Mountain, a distance of 6 miles. The Wedin Member contains the first appearance of the Institella and Agelesia assemblage, the first of the Cathedral Mountain faunas traceable across the entire mountain front except for the gap between hill 5300 and Leonard Mountain (Plate 7: figure 1). IN LEONARD MOUNTAIN.—The bulk of Leonard Mountain is composed of the Skinner Ranch Formation. This is overlain by yellow shaly beds of the Cathedral Mountain Formation, which appear on the mountain in the knob 0.25 mile northwest of bench mark 5860. Most of the north slope from this point is composed of the upper beds of the Skinner Ranch Formation, the type section of that formation appearing at the west end of the mountain, where it forms a conspicuous knob and extends into the valley floor. The Cathedral Mountain Formation is exposed at places in the valley, especially where small bioherms are resistant enough to stand above the floor. The upper part of the Cathedral Mountain Formation is exposed in the lower slopes of the hill north of Leonard Mountain, where they are overlain by the Road Canyon Formation. Excellent exposures with abundant fossils appear in the saddle between Leonard Mountain and the hill to the north. The ravines near the base of the mountain to the north (5674) contain good exposures of the yellow shale, and the thin limestones and bioherms contain numerous fossils. Good specimens of large Perrinites are fairly common. Perhaps the best collecting place is R. E. King's locality 123 (=USNM 71 Iq). This and the few exposures of the upper Cathedral Mountain Formation that appear in the ravines at the base of the slopes of hill 5801 and to the mouth of Road Canyon are the last exposures of the typical Cathedral Mountain Formation. The Cathedral Mountain has been cut out by the Hess Ranch Horst, but the interval appears again several miles to the east in a different facies. Institella is perhaps the most significant brachiopod of the Cathedral Mountain and seems to be confined to it. It occurs at several levels in the formation, but on Leonard Mountain it is at the very base. On top of Leonard Mountain the knob 0.25 mile northwest of bench mark 5860 is composed of about 80 feet of yellow, silty shale, followed by 5.5 feet of thick-bedded calcarenite containing Institella and other fossils. This is followed by 1-2 feet of chert with the same fossils, which in turn is succeeded by thin-bedded limestone that forms the top of the knob. The thin plates of limestone at the top are composed almost completely of Institella, with many of the valves crushed together and in great profusion. The calcarenite capping the west knob of Leonard Mountain contains Institella and is assigned to the Cathedral Mountain. This same bed appears midway between Leonard Mountain and the hill to the north (USNM 709), but it does not extend to the highest part of the mountain to the south and east. IN DUGOUT MOUNTAIN REGION.—The Cathedral Mountain Formation has a considerable development in the Dugout Mountain region, but much of it, which is soft shale, is buried under the plain on the north side of the mountain. In our discussion of the Skinner Ranch Formation, we showed that the lower part of the Leonard of P. B. King contained several limestones, numbered 2—4 by him, separated by beds of yellow, silty shale and chert. These limestones and shales, although lithically like the Cathedral Mountain Formation, contain fossils like those of the Skinner Ranch below. We, therefore, have named this sequence the "Dugout Mountain Member of the Skinner Ranch Formation" (Cooper and Grant, 1966). The Cathedral Mountain Formation thus begins with P. B. King's Fifth Limestone Member at its base. This limestone is here called the "Wedin Member of the Cathedral Mountain Formation" and introduces the Institella fauna (see above). The Wedin Member at the base of the section forms a narrow northeast trending belt at the NUMBER 14 63 GAP- CAPITAN CM- CATHEDRAL MOUNTAIN cp- conglomorote with PERRINITES DC-OECIE RANCH DM- DUGOUT MOUNTAIN &•GAPTANK LH-LENOX HILLS PT-POPLAR TANK RC- ROAD CANYON SP- SULLIVAN PEAK W-WORD 5-5th llmulone mlmblr of King with INSTITELLA-WEDIN MEMBER TC VERTICAL a HORIZON FIGURE 21.—Section through Dugout Mountain, past the site of Old Payne Ranch, through hill 4861, to hill capped by the 5600-foot closed contour in the northwest corner of the Monument Spring quadrangle. northwest base of the mountain. Here the member is only 12 feet thick (P. B. King, 1931:133), but it has numerous bioherms that contain abundant Institella and Agelesia exactly like those on the north slope of the Lenox Hills and near the old Word Ranch. On Dugout Mountain the silicification of fossils proved disappointing, but numerous specimens were obtained in the Wedin Member by conventional means. Above the Wedin Member the section is covered, concealing more than 500 feet of section (P.B. King, 1931:133). Above this are nearly 300 feet of shale and sandstone, the latter forming conspicuous brown masses in the valley. Perrinites is reported (P. B. King, 1931:133) in the upper beds, which are like those at the Clay Slide and Cathedral Mountain. The thick sandstone also resembles that occurring south of Sullivan Peak. P. B. King's (1931:133) section aggregates 987 feet from the top of the Dugout Mountain Member of the Skinner Ranch Formation to the base of the Road Canyon Member. The upper few hundred feet of the Cathedral Mountain Formation near the Old Payne Ranch site occur in situations suggestive of the Clay Slide. Indeed the conspicuous conical hill 0.25 mile northwest of the Old Payne Ranch site contains a long gash in the shales under the massive limestone of the Road Canyon Formation strongly reminiscent of the Clay Slide. Here fossils are fairly common and Perrinites occurs in a thin band of orangebrown limestone. Exactly the same sequence, but in a slightly different setting, can be seen in hill 4861 and in the one immediately to the northwest, which includes the 4750-foot contour. In hill 4861 a conspicuous round-pebble conglomerate occurs about one-third the height from the base. This conglomerate abounds in well preserved Perrinites and other Leonardian cephalopods. The hill and the other immediately to the northwest are capped by the Road Canyon Formation (Plate 13: figures 1,2). The presence of the Road Canyon Formation at several localities in the Sierra del Norte from a point about 2 miles west-northwest of the Old Payne Ranch site indicates that the lower slopes of this range west of Dugout Mountain belong to the Cathedral Mountain Formation. Due west of the new house along the road (USNM 737s) there occur dolomitic beds with large pebbles and an abundance of Perrinites. Farther south on the mountain front at USNM 741s, thin detrital beds contained Institella and Agelesia, indicating Cathedral Mountain, but we were unable to identify Road Canyon in this section. IN OLD WORD RANCH ( = SPLIT TANK AREA).—Cathedral Mountain Formation rocks are exposed in the canyon (Comanche Canyon? = e a s t branch of Hess Canyon of some inhabitants) from the igneous body on the east side of the Hess Ranch Horst, past Old Word Ranch, the Appel Ranch, Split Tank, and beyond. It is rather thin in this region, 301 feet thick near Split Tank and 237 feet thick in P. B. King's (1931:145) section 27. The base of the sequence is conglomerate with numerous small quartz pebbles (Plate 12: figure 4), followed by Hess-type limestones, then bioherms overlain by yellow siliceous shale. Above the shale there is a thick sequence of biohermal limestone, which is succeeded finally by a considerable thickness of yellow shale with abundant Rugatia and Peniculauris mckeei (see Figure 22 for details). Institella is common to rare in parts of the lower half, but it was not found in the upper shaly part. The Split Tank section has furnished many of the finest silicified fossils from these mountains, 64 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY SPLIT TANK _£5£]=_ 726y Feet r 6 0 - 30 ^ - ^ ^ ^ - 726w 702a' L 0 702a 'r^3 702 ent C\J O FICURE 22.—Columnar section of the Cathedral Mountain Formation 0.5 mile east of Split Tank showing stratigraphic level of locality numbers (RC = Road Canyon Formation). and the supply has not been depleted. The section is noteworthy for its bioherms, some of which are characterized by an abundance of particular brachiopods. Institella and Torynechus (formerly Uncinuloides) bioherms occur in the lower part of the section. Above them are bioherms replete with Enteletes and Hercosia. In the upper part, just below the thick shales, there occur remarkable bioherms composed almost entirely of Collemataria. Southwest of the old Word Ranch site the Cathedral Mountain Formation is not well preserved; just east of the Hess Ranch Horst it is greatly dolomitized, and many of the fossils have been destroyed. The map by P. B. and R. E. King (in P. B. King, 1931) indicates a belt of Wolfcamp limestone on the south side of the igneous body at the base of the Hess Ranch Horst. These limestones are indicated by Ross (1963a: map [Plate I]) as of Leonard age. We corroborated their post- Wolfcamp age by collecting Institella from their upper part at several places, thus clearly establishing their position at the base of the Cathedral Mountain Formation. The beds with Institella overlie heavy bedded limestone of the upper Skinner Ranch Limestone Formation. The Skinner Ranch part of this limestone lies against the igneous body at a high angle. The basal conglomerate of the Cathedral Mountain Formation occurs at the base of these Institella limestones and can be traced northeastward up the canyon for several miles. It is especially well developed near the Old Word Ranch on the steep dip slope forming the south side of the canyon. ROAD CANYON FORMATION This formation forms the top of the Leonard Series throughout the mountains from the Split Tank area to Dugout Mountain. It is an excellent datum in the stratigraphy of the Glass Mountains. Its name originally was proposed by Cooper and Grant (1964) as a member of the Word Formation. Previously it had been designated as the First Limestone Member of the Word Formation by P. B. King (1931:71) but, because of the predominantly Leonardian character (Miller, 1945a) of the fauna, we have transferred it to the Leonard Series. The changeable character of the formation from place to place makes its discussion on a geoNUMBER 14 65 graphic basis necessary. Five areas are considered: (1) the triangular hill north of Leonard Mountain, which contains the type section, (2) the Old Word Ranch area, (3) the Sullivan Peak region, (4) Gilliland Canyon, and (5) the area northwest of Old Payne Ranch, northwest of Dugout Mountain. IN HILLS NORTH OF LEONARD MOUNTAIN.—North of Leonard Mountain lies a triangular group of hills, which are bounded on the north by the eastwest Road Canyon and on the west and east, respectively, by the convergent canyons Gilliland and Hess. The type section of the formation is located at the south angle of the triangle, approximately on the 103' 15" parallel. The formation crops out on the south side of the mouth of Road Canyon and forms a conspicuous ledge on the hillsides facing Hess and Gilliland Canyons, but it passes under the Gilliland Canyon floor about 1.25 miles south of the western mouth of Road Canyon. The formation in these hills displays a great variety of lithologies; in fact, all types occur in the formation except the thin-bedded black shales, which appear to the east and west of the type area (Plate 16: figure 4; Plate 14: figure 2). At the type section, the formation is 228 feet thick and consists of bioherms with Coscinophoro at the base, followed by bioherms abounding in Hercosestria, more limestone with Coscinophora, and then a series of varied thin beds followed by a thick biohermal mass. No two sections of the formation in these hills are alike, a fact that is to be expected when dealing with bioherms. Coscinophora bioherms of considerable size can be seen on both sides of the triangular block of hills. Other fossils also are abundant; the area is one of the best in the mountains for collecting silicified fossils. The variety of animals is great. Not only are brachiopods abundant, but also a variety of gastropods and pelecypods may be retrieved in nearly perfect preservation. Perrinites, found in this area south of the eastern mouth Road Canyon (USNM 726c), attests to the Leonardian age of the formation. It is difficult to determine a faunal sequence in the formation, but a crude succession of forms seems to be manifest. The bioherms near the base on both sides of the hills are characterized by Hercosestria and a variety of other forms such as Edriostegcs, Texarina, and Cyclacantharia, which also are seen to the east in the conical knob (USNM 702c) 1.25 miles southwest of Old Word Ranch. One of the best of the bioherms (Plate 16: figure 4) occurs at the base of the type section on the east nose of hill 5779 about 2.5 miles due north of the Skinner Ranch house. Over 82 feet thick, it begins with a thick limestone conglomerate at the base, followed by fine-grained calcarenite, which is succeeded by the biohermal limestone. This large rounded mass, visible for a considerable distance, is one of the largest bioherms in the mountains. The beds above the bioherms contain one conspicuous bed of yellow shale 14 feet thick. The top boundary of the section is uncertain because of the appearance of yellow shale in the upper part of the section. The uppermost thick calcarenite and bioherms (bed 10) are followed by yellow shale with thin, inter-bedded limestone, but these contain fossils like those of the Road Canyon below. A similar situation exists at the top of the formation on both sides of the triangular hill. Lenses and thin layers of limestone with Road Canyon fossils appear just above Road Canyon on the west side of Hess Canyon and on the east side of Gilliland Canyon. On the former, the lenses and layers are usually only a few feet above the main mass of the Road Canyon Formation, but on the east side of Gilliland Canyon, lenses occur 25 feet above the main body of the formation (USNM 720d). Nevertheless, the fossils in them are Road Canyon types with affinities with the Road Canyon below. We have, therefore, included these beds in the Road Canyon Formation. IN OLD WORD RANCH AREA.—This area extends from the east end of the Hess Ranch Horst northeastward along the north side of the canyon that runs from the Hess gate past the Old Word Ranch to the Appel Ranch house. The vicinity of the Old Word Ranch is the type area for the Word Formation; the Road Canyon was originally the First Limestone of the Word of P. B. King. The best and most accessible sections are at the site of the Old Word Ranch, where P. B. King (1931:143, section 24) records 140 feet, including 60 feet of thin-bedded bituminous limestone weathering light gray and, above it, 80 feet of dark gray dolomite (Plate 13: figure 3). The section overlies siliceous shale of the Cathedral Mountain. P. B. King made no mention of the bed of bioherms at the base of 66 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY the thinly laminated beds (USNM 703a). These have the same fossils as the bioherms at the base of the Road Canyon Formation just southwest of the mouth of Road Canyon (USNM 719x). At the Old Word Ranch site the thin-bedded dark limestones contain several lenses teeming with fossils (USNM 703, 703c, 703d). The first of these contains a remarkable fauna of ammonites including Perrinites (Miller, 1945a); the second yielded a remarkable sponge and molluscan fauna, and the third produced some unusual spirifers and other fossils. Northeast of the Old Word Ranch, the Road Canyon Formation becomes dolomitic and merges with the higher limestones of the Word Formation by pinching out of the shale between them (Plate 13: figure 3). Southwest of the Old Word Ranch, the Road Canyon can be traced to the fault bounding the Hess Ranch Horst. Near the horst the rocks are greatly dolomitized and difficult to identify, but, near the crossing of the south branch of Hess Canyon with the road from the Old Word Ranch, two important localities have yielded fine collections. One of these, USNM 702c, forms the top of a small knob on the inside of a wide loop of the road 1.35 miles southwest of the Old Word Ranch. Here Hercosestria and many other species characteristic of the bioherms at the base of the Road Canyon Formation were taken. IN SULLIVAN PEAK AREA.—The thickest development of the Road Canyon appears in Cathedral Mountain and on the hills just east of it to Clay Slide. The best place to study the member is on the spur terminating slightly less than a mile south of Sullivan Peak (Plate 1: figure 5; Plate 5: figure 3). The section was measured to be 367 feet by P. B. King (1931:70, section 12, Word, beds 1-6) and is composed largely of thin- and thick-bedded bituminous limestone. The rocks are very fossiliferous, with characteristic bioherms at the base. The fauna from the dark limestone includes many species not seen elsewhere or extremely rare in other parts of the formation. Along with the rare specimens, characteristic Leonardian types have been collected, such as Perrinites, Peniculauris, and Rugatia. Clifton (1945:1770) records Waagen.oceras dieneri Bose "from strata of the Word First Limestone Member (— Road Canyon Formation) in outcrops near Sullivan Peak." Clifton's Waagenoceras is now called Stacheoceras normani Miller and Furnish (1957). On the east side of the Sullivan (Yates) Ranch road, two long hills extend to the east; the farthest one contains the Clay Slide. These two hills are capped by the Road Canyon Formation and are lithically like the section below Sullivan Peak. Perhaps the most interesting feature of these hills, especially the one nearest Sullivan Peak, is the occurrence in the bituminous limestones of large bioherms composed mostly of the peculiar lyttoniid Coscinophora. In places the rock is literally made up of the pedicle valves of this ostreiform brachiopod. Similar bioherms also were seen on the east side of Sullivan Peak toward the Yates place. Although Coscinophora occurs at other levels in the Skinner Ranch-Cathedral Mountain interval, it is most abundant in the Road Canyon Formation (Plate 17: figures 3 and 4). Traced westward along the lower slopes of Cathedral Mountain, the Road Canyon Formation seems to be split by insertion of shaly wedges (P. B. King, 1931:135, section 11). King reports 127 feet of bituminous limestone at the base, followed by 50 feet of shale and another 50 feet of limestone. Above this there are two thinner limestone beds (15 and 16 feet thick) that may belong to the Road Canyon Formation, the Word beginning at the base of King's bed 8, which is composed of 235 feet of siliceous shale and sandstone. IN GILLILAND CANYON.—About 1.5 miles north of Clay Slide, King's map (in P. B. King, 1931) indicates an isolated mass of Road Canyon Formation faulted down and appearing inconspicuously in the plain, where it has been exposed by an intermittent stream (USNM 724b). The map also shows a band of the First Limestone of the Word extending along the west side of Gilliland Canyon and forming a conspicuous bench along the hillside. The limestone is mapped as overlying Leonardian rocks. We sampled this limestone in three places (USNM 723t, 723w, 731m) and found it to belong to the Third Limestone of the Word ( = Willis Ranch Member) rather than to the First Limestone ( = Road Canyon Formation). It contains a great abundance of Waagenoceras dieneri, Echinosteges tuberculatus (R. E. King), and other characteristic Word fossils (USNM 723t). Furthermore, the limestone generally is light gray, is very NUMBER 14 67 sandy, and also contains much interbedded sand in the form of brown layers and lenses. The lithic character of the rock is exactly like that seen in the Willis Ranch Member at the junction of Road and Gilliland Canyons, a well-known location for the Third Limestone of the Word and its fossils. Examination of the map will thus indicate that the Third Limestone of the Word crosses the canyon near bench mark 4973 and rises along the west side of Gilliland Canyon. The Road Canyon Formation beneath it descends below the canyon floor 1.25 miles south of the mouth of Road Canyon, but it does not rise above the floor on the west side until the isolated mass northeast of Clay Slide is reached (USNM 724b). NORTHWEST OF DUGOUT MOUNTAIN.—In the low hills west of the Old Payne Ranch (no longer in existence) on the northwest side of Dugout Mountain, the Road Canyon Formation is reduced to two layers of limestone separated by siliceous shale. The lower limestone is recorded by P. B. King (1931:131, section 5) as 40 feet thick and the upper one as 20 feet thick; the two are separated by 60 feet of shale. Our measurements show the thicknesses to be variable. We measured 70 feet of the lower limestone of the Road Canyon Formation in the conical hill 0.25 mile northwest of the Old Payne Ranch. The lower part of the hill is composed of yellow-orange siliceous shale of the Cathedral Mountain. The cap of the hill is composed of some biohermal limestone in the lower part, but mostly of dark bituminous limestone abounding in fusulinids. In the low hills 0.75 mile southwest of the Old Payne Ranch, we found the lower limestone to be about 20 feet thick, but the upper limestone is only 8 feet thick and is underlain by 15 feet of sandstone. The upper limestone is crumbly and a veritable mass of fusulinids and other less abundant fossils. The brachiopods proved to be characteristic Road Canyon species. Above this thin limestone, there occurs a thick sequence of massive sandstone and siliceous shale of the Word Formation. The Road Canyon thus appears to thin significantly west of Sullivan Peak. Other and hitherto unidentified exposures of the Road Canyon Formation occur 1.5 miles northwest of Old Payne Ranch. In hill 4861, which is conical and with a section like that 0.25 mile northwest of Old Payne Ranch, the Road Canyon forms the crest of the hill. It is biohermal and has much detrital material in the form of fusulinids between the bioherms. Coscinophora was found here and in the next hill to the northwest. This is a low, flat hill, the lower part of which is marked by the 4750-foot contour. Here the Road Canyon formation consists of two limestones with intervening shale that contains some thin beds of limestone. The lowest bed, 20-30 feet thick, is variable but with bioherms at the base. The hill is capped by a 5-foot bed in two tiers, mostly of fusulinid limestone. The Road Canyon in this hill overlies Leonardian beds with Perrinites and is like the section in hill 4861, with the cephalopod-bearing conglomerate of the Cathedral Mountain Formation appearing at the base of the hill. The Leonardian sediments are in fault relation with the Word limestone south of it. After discovery of the Road Canyon Formation in the foothills of the Sierra del Norte, we were able to trace the formation southward for a few miles along the mountain front. Throughout the observed extent of the formation it is very variable, consisting mainly of thin limestones separated by yellowish shales. The tracing was facilitated greatly by the discovery of a cobbly, dark bluish limestone abounding in ammonites of the genera Paraceltites and Texoceras. This ammonite bed was found along the mountain front at USNM 737c, 737g, 737n, 737y, and 739d. Opposite the abandoned house (NW, NW 3; USNM 737g) the ammonite bed is well developed. The same bed was seen about 2 miles farther south, 3.5 miles due west of Dugout Mountain, where the following section was measured: feet K. Yellow shale at top of section J. Fine-grained limestone with fusulinids and small Enteletes like those of USNM 732j 7 I. Bluish-black limestone with ammonites 2 H. Fine-grained limestone 1 G. Blocky, bluish-black limestone breaking into lumps and with many ammonites (like bed at USNM 732z —737n) 20 F. Granular limestone with fusulinids 12 E. Siliceous, platy, buff-colored shale 20 D. Seven thick beds of dark limestone separated by yellow shale, fusulinids 75 C. Covered 50 B. Sandstone of the Cathedral Mountain Formation 20 A. Covered 68 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY A mile still farther south, just southeast of the "s" in "mountains" (SE, NW 3: USNM 737y), another section of the Road Canyon is revealed as follows: feet G. Fusulinid limestone 4 F. Limestone with the rare brachiopod Collumatus and showing relationship to USNM 732j 3 E. Dark, blue-black shale and limestone with ammonites similar to those at USNM 732z 20 D. Yellow siliceous shale 40 C. Yellow siliceous shale with scattered limestone layers 25 B. Biohermal limestone with Coscinophora 35 A. Covered No two sections on the mountain front are alike, but the presence of the ammonite zone, though variable in thickness, and the presence of the upper fusulinid bed with Collumatus relates these Sierra del Norte localities to those in the foothills, such as USNM 732j, where the Collumatus and ammonite beds also can be seen. It is believed that west of Cathedral Mountain the Road Canyon thickens by introduction of shale and concentration of detrital material in lenticular beds that tend to fray the formation in a western direction. The Road Canyon was not successfully identified south of USNM 737y. Guadalupe Series In the Glass Mountains the Guadalupe Series consists of the Word and Capitan formations. Only the Word Formation will be discussed at length here because most of the Capitan Formation and its members, as well as some members of the Word Formation, are dolomitic or otherwise poorly fossiliferous and, therefore, not collected or studied in detail. So far as known, none of the dolomites yield silicified fossils; however, three limestone members of the Word Formation yield fine fossils in abundance. WORD FORMATION The Word Formation was named by Udden, Baker, and Bose (1916:52). Udden (1917) further enlarged on the formation, indicating its type section on the Word Ranch, where its massive facies is exhibited. P. B. King (1931) redescribed the formation, individualized its limestone members by numbers, and explained the two facies in the western part of the mountains and the carbonate facies in the eastern part. The two facies begin in Gilliland Canyon north of Iron Mountain, approximately in the middle of the mountains. A western facies, predominantly of silty shale and sandstone, is recognized west of Gilliland Canyon, while the rocks to the east are characterized by carbonate (dolomite and limestone), which dominates the section in the area east of Split Tank. The western or shaly-sandy facies forms the thickest part of the formation, about 1470 feet, and thins to about 450 feet in the eastern part of the mountains. East of Gilliland Canyon, and especially on the Hess Ranch in the intermediate zone of the facies P. B. King (1931:71) described four limestones, which were given consecutive numbers. The first or lowest limestone that is best exposed near Old Word Ranch is lithically and faunally unlike the others, but it can be traced across the mountains. This limestone not only is unique lithically, but also it contains an unusual and characteristic fauna, with strongest ties to the Leonardian rocks below. Cooper and Grant (1964:1586), therefore, recognized it as a member of the Word Formation (the Road Canyon Member). Later, theyj (1966) raised it to formation rank and placed it in the Leonard Series, as explained above. Separation of this First Limestone Member from the Word Formation does not affect the conception of the formation as outlined by King, because his overall view of the facies remains unchanged. The removal of this lowest limestone, however, does affect the numbering of the other Word limestones. Therefore, in order to make our discussion clearer and the location of the collections unambiguous, Cooper and Grant (1966) named the other limestones as members of the Word Formation. The sections west of Gilliland Canyon above the Road Canyon Formation (=First Limestone of the Word) consist mainly of yellow platy shale, with some thin sandstone and thin limestone beds. Northwest of Dugout Mountain, thick sandstones appear in the sequence. Inasmuch as the shales are poorly fossiliferous and the thin limestones in this part of the area yield few good silicified fossils, little collecting was done in the Word west of Gilliland Canyon. NUMBER 14 69 The hills just north of Leonard Mountain aro bounded on the west by Gilliland Canyon, on the north by Road Canyon, and on the east by Hess Canyon. The hilltops are capped by a thick limestone, but between the Road Canyon Formation and the hillcrest the section is mainly yellow siliceous shale, like that of the western facies. Shale and limestone in this area do not yield abundant good silicified material. The fossils are not so strongly concentrated as they are in sections farther to the east. The area with the greatest abundance of fossils is on the east side of the Hess Ranch horst and in fault contact with it. Here four limestones of varying thickness are separated by siliceous shale of the western facies. These limestones, with the Road Canyon Formation, were designated by P. B. King (1931:71) as First to Fourth Limestones. The Second, Third, and Fourth Limestones are hereby designated by names but retained as members of the Word: China Tank, Willis Ranch, and Appel Ranch Members. CHINA TANK MEMBER (=SECOND LIMESTONE MEMBER OF P. B. KING).—This member consists of 113 feet (P. B. King, 1931:72) of siliceous brownish- to yellowish-gray limestone. The type section is the west side of hill 5611 not far east of the China Tank of the Hess Ranch (Cooper and Grant, 1966: 7). The tank is in Hess Canyon, 2.2 miles north, 75° east, of the Old Willis Ranch on the P.B. and R. E. King map (1931). The tank takes its name from the luxuriant chinaberry trees that form a cattle shelter on its south side. The member at the type section is composed of highly fossiliferous limestone, nearly all of it containing silicified material (USNM 706c). It contains much echinoderm debris and some fine sand. The China Tank Member is well displayed in the bluffs on the east side of Hess Canyon about 1.5 miles south of the head of its south branch, due west of Old Word Ranch. Here the member is massive and in large part dolomitized, but fusulinids are abundant in limy parts. Patches of silicified brachiopods and other fossils are rare and difficult to find. East of Old Word Ranch and in the vicinity of the Appel Ranch, the member loses its individuality and becomes mostly massive dolomite. WILLIS RANCH MEMBER (^THIRD LIMESTONE MEMBER OF P. B. KING).—The type section selected for this member by Cooper and Grant (1966:7) is near the eastern mouth of Road Canyon, 1 mile southwest of the Willis Ranch site (R. E. King 243 = USNM 724u) (Plate 14: figure 1). Here P. B. King (1931:71) measured 308 feet of limestone, as follows: feet Third Limestone Member D. Oolitic gray limestone in thin ledges 69 C. Gray limestone, containing fossils, and several seams of small pebbles 11 B. Brown calcareous sandstone 10 A. Light gray oolitic limestone, containing some chert masses and fossils including Waagenoceras, Cyclacantharia 218 Good silicified fossils appear in the lower part of the member at the type section, but they are not as abundant as they are to the east. The Willis Ranch Member extends westward to the east side of Gilliland Canyon, where it is well exposed near the junction of this canyon with Road Canyon. This place is well known for its abundance of ammonites, especially Waagenoceras (Bose, 1917). It is also exposed in a broad patch opposite the mouth of Road Canyon on the west side of Gilliland Canyon, although it is mapped on the west side of this canyon as the Lower Member of the Word limestone, with Leonardian shale below it. From this place it can be seen as a conspicuous ledge on the western slope of Gilliland Canyon. It disappears underground 1.5 miles north of hill 4910 at the southwest end of Gilliland Canyon. It is not known west of here (Plate 14: figure 4). Everywhere that it has been examined, the Willis Ranch Member is very sandy. When dissolved in acid, the limestone leaves a residue of sugary quartz sand. In Gilliland Canyon the formation is very sandy and contains lenses and layers of fine brown sand (Plate 14: figure 3). Fossils are not common in Gilliland Canyon, but one locality (USNM 723t) yielded an abundance of typical specimens. On the west side of Gilliland Canyon thin lenses of richly fossiliferous dark limestone occur just beneath the main mass of the member. These are unusually fossiliferous and contain a variety of species not seen at any other level (USNM 723w). The same types of fossils were found 30 feet below the member at its southernmost occurrence (USNM 731m) 70 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY 1.5 miles north of hill 4910. Similar fossils were also taken under Sullivan Peak (USNM 73lu; Plate 15: figure 4) and northwest of Old Payne Ranch (USNM 732s). These isolated localities are thought to mark the approximate position of the Willis Ranch Member, which disappeared at the southwest end of Gilliland Canyon. The best place to collect the fossils of the Willis Ranch Member is on the east side of the Hess Ranch Horst, on the long slope marking the south side of Hess Canyon opposite the divide in the Canyon. There the member is 95 feet thick (P. B. King, 1931:72) and consists of yellowish-gray sandy limestone containing thick patches of fossils as death assemblages. The fossils are beautifully silicified, and USNM 706 and 706e have yielded enormous quantities of specimens. Waagenoceras is especially abundant near the top of the section. USNM 706 is in the lower part of the member while USNM 706e is at the top. East of the Old Word Ranch, on the Appel Ranch, the Willis Ranch Member merges with the other massive members of the Word and becomes strongly dolomitic and difficult to separate from the others. APPEL RANCH MEMBER (=FOURTH LIMESTONE MEMBER OF P. B. KING).—This member is characterized by the great amount of chert it contains. North of the Appel Ranch P. B. King (1931:143) recognized three beds as follows: feet Fourth Limestone Member C. Light gray crystalline limestone with abundant, small, brown chert nodules as well as an abundant fauna of brachiopods, fusulinids, and other fossils 260 B. Gray and brown limestone and dolomitic limestone, containing some chert nodules and many silicified fossils, interbedded with sandy brown limestone and some siliceous shale 255 A. Fossiliferous cherty dolomite, weathering to dirty gray, deeply pitted surfaces 40 We designated the area north of the Appel Ranch as the type section for the member (Cooper and Grant, 1966:8). Here it is well displayed and nearly every layer exposed. The member is also well displayed on the north side of Hess Canyon near the point where the canyon bends to the north, and the member can be followed westward to the site of the Willis Ranch, where it frays out into the shale (Plate 15: figure 1). On the east side of the bend in Hess Canyon, the Appel Ranch Member is exposed in a strongly dissected area that affords excellent collecting and opportunity to study the sequence. In places the rock is composed almost wholly of masses of large Parafusulina. In other places the brachiopods occur in death assemblages of shells matted together so closely that some of them are crushed and distorted. Nevertheless, the variety is great, and many species not seen in the lower members of the Word Formation appear in this member. Before closing the discussion of the Word Formation, we must note a lens that lies not far above the Willis Ranch Member in the shale separating that member from the Appel Ranch Member. This is USNM 706b, which consists of about 2 feet of sandy limestone with abundant fossils thickly matted together. The fauna has some elements of the Willis Ranch Member and some of the overlying Appel Ranch Member, yet it is sufficiently distinctive to be recognized as a separate level. Our operations unfortunately have removed most of this thin local layer. Northwest of Old Payne Ranch the section above the Road Canyon is poorly preserved, but occasional lenses and layers of limestone replete with fusulinids were seen. One of these (USNM 732s) contained Waagenoceras, Popanoceras, and brachiopods in addition to fusulinids. The faunal content suggested the fossiliferous zone just under the Willis Ranch Member on the west side of Gilliland Canyon (USNM 723w). Above this lens comes a conspicuous, thick sandstone. Above the sandstone there was found a fault block containing Cathedral Mountain Formation, with Perrinites in a thick conglomerate near the base exposed in a conical hill at an elevation of 4861 feet. This conglomerate contains rounded quartz pebbles ranging up to 3 inches in diameter. The hill is capped by Road Canyon Formation. The section also is well displayed in the next hills to the west and northwest. Above the Road Canyon in the section of the Word there are mainly sandy shale and sandstone with occasional thin limestone and blue shale. The section is similar to that seen under Sullivan Peak, but no characteristic Word fossils were found other than fusulinids (Plate 1: figure 4). South of hill 4861, about a mile, hill 4806 is a sandstone hogback (see Plate 15: figure 3) with NUMBER 14 71 Word fossils on the west slope. On the northwest side of this hill, Word ammonites (Waagenoceras) are common, but a low hill a little farther to the west is capped by Cretaceous limestone. It is the last limestone band on the King (1931) map due west of the road junction in NE, NW 3, Monument Spring quadrangle. This is the top of a downfaulted block (see Figure 21). CAPITAN FORMATION Although no detailed collecting was done in the Capitan Formation of the Glass Mountains, it is important to record one locality that was discovered by chance and that yielded some good brachiopods. This is a small downthrown block 1.1 miles north, 33° west, of Old Payne Ranch site, Altuda quadrangle (USNM 732q). This is in cream-colored dolomitic limestone containing an abundance of the fusulinid Polydiexodina and of brachiopods, indicating the level of the Hegler Member of the Bell Canyon Formation in the Guadalupe Mountains. This place offers possibilities for further collecting. The Capitan should be searched for undolomitized patches as possibilities for collecting and dating. Faunal Zones in the Glass Mountains A zone fossil, to be useful, must be distinctive and easy to recognize, fairly abundant, and widespread. Few genera in the Glass Mountains answer to these requirements. The method of collecting also has so increased the ranges of some hitherto "good" guide fossils that their value as zone indicators has become limited. Ombonia, for example, was regarded as restricted to the Capitan Formation and the Lamar Member, but now it is known from the Road Canyon and Cherry Canyon Formations. The very rare Word precursors in the Road Canyon, such as Echinosteges and Yakovlevia, limit the value of these genera, which otherwise are identified most with the Word Formation. Only a few genera in the Glass Mountains have the qualifications of good zone fossils, and these with their present ranges are listed below. Parenteletes: Although this is not so common as might be desired, it is identified readily by virtue of its strong plication and its sulcate anterior commissure. The genus has its roots in the Pennsylvanian, as it occurs in part of the Gaptank identified as of Canyon age. It extends into the Uddenites- bearing Shale and on through the Neal Ranch to the top of the Lenox Hills Formation and into the Poplar Tank Member. It is, thus, a guide in its Wolfcampian range to that part of the section that was described as the Wolfcamp Formation by P. B. King (1931) minus the Uddenites-bearing Shale Member. Spyridiophora-Glyptosteges: These two highly sculptured aulostegids have nearly the same range, and one often occurs where the other is absent. Spyridiophora has the longer range, as it occurs first in the Neal Ranch Formation and extends through the Lenox Hills Formation into the top of the Skinner Ranch. It is also one of the fossils that correlates the Taylor Ranch to the top of the Skinner Ranch. Glyptosteges is a new genus to be described in a subsequent volume. It does not occur in the Neal Ranch or Lenox Hills, but it appears first in the Decie Ranch Member and becomes moderately common in the remainder of the Skinner Ranch Formation. It is the best fossil with which to identify the Dugout Mountain Member of the Skinner Ranch Formation. The two used together thus cover the entire Wolfcamp as defined in this monograph (excepting the Uddenites-bearing Shale Member). Both of these genera also occur in the Skinner Ranch equivalent of the Bone Spring Formation, but neither one has been seen in the Hueco Group. Scacchinella: Some paleontologists object to the use of Scacchinella in stratigraphy because it is thought to be facies restricted. It is a facies fossil only in the sense that it really creates its own facies. As a cluster builder it helps to determine bioherms, but these may be in more than one type of environment. Furthermore, even facies fossils have value as horizon-markers in their own right. Scacchinella is such a fossil. Several of the levels in which this brachiopod appears are characterized by different species of the genus: S. primitiva, new species, occurs in the Pennsylvanian (Cisco); S. triangulata, new species, is found in the Uddenites-bearing Shale Member of the Gaptank Formation; S. exasperata, new species, appears in the Lenox Hills Formation; and S. titan, new species, characterizes the Skinner Ranch Formation. Each of these occurrences is in a biohermal environment, but in each 103° 30' 103°20 103 30 103°20 Volcanic Post-Word 8 Pre-Gaptank China Tank Word Shale Appel Ranch Word (WillisR.) Road Canyon ^ 3rd ft 4th Leonard limestones Dugout Member of Skinner Ranch Skinner Ranch Lenox Hills Neal Ranch Gaptank 103 20 NUMBER 14 103°20' £ v !l L ^o Poplar Tank _o u_i . 103°20' 103°15' FIGURES 23-25— Revised geological maps (after P.B. King): 23, Dugout Mountain area; 24, Lenox Hills; 25, Leonard Mountain. 103 10' 74 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY one of them Scacchinella is a conspicuous form, specifically diagnostic of that level. In the aggregate the genus thus becomes, in the Permian, an excellent guide to the Wolfcampian as defined in this monograph. Orthotichia: This is not a conspicuous fossil, but it is easily recognized by its two finely costellate valves and uniplicate anterior commissure. Its range is that of the Wolfcamp, as defined herein, from the Uddenites-bearing Shale Member of the Gaptank Formation through the Skinner Ranch Formation. Antronaria: A new genus to be described in a subsequent volume, these large rhynchonellids with their depressed costa or costae in the fold are recognized easily and are characteristic of the Skinner Ranch, beginning in the Decie Ranch Member and occurring through the Sullivan Peak Member and in the Taylor Ranch Member of the Hess Formation. A slight spillover into the Cathedral Mountain Formation occurs, but the genus is too rare there to lessen its value as a guide fossil to the Skinner Ranch Formation. Teguliferina: Although this is a very distinctive fossil, it does not have a clean-cut range. It appears in the Pennsylvanian, but it flourished in the Wolfcamp and extended into the lower part (Decie Ranch equivalent) and Poplar Tank Member of the Skinner Ranch Formation. It is rare in the upper part of the Skinner Jlanch Formation and the Taylor Ranch Member of the Hess Formation. The Wolfcampian is thus well provided with genera that are characteristic of all or parts of it. Institella: This well-sculptured aulostegid, with its leptaenoid form, is*one of the most easily recognized fossils in the Glass Mountains. It appears in abundance from the Dugout Mountain area to the Split Tank region. One interruption occurs in this wide lateral range in the vicinity of hill 5021 west to approximately hill 5300 in the Lenox Hills. It is the best indicator of the base of the Cathedral Mountain Formation, but these honors are shared almost equally by the little aulostegid Agelesia, whose size makes it a more difficult fossil to find. Hercosia: This distinctive richthofeniid with its knife-blade median septum is recognized easily and is about the best guide fossil to the Cathedral Mountain, as it appears first in the Wedin Member in the western part of the mountains and in the base of the Cathedral Mountain Formation at the east end. Combined with Hercosestria, it becomes a guide to the Leonardian. Hercosestria: This richthofeniid has the same interior as Hercosia, but the cone is covered by a network, the coscinidium. Hercosestria occurs in the lower part of the Cathedral Mountain Formation (USNM 72lu), but it is very rare there and elsewhere at this level. It is abundant in the Road Canyon Formation, where some bioherms are almost completely made up of the genus. Taken with Hercosia, which is common in the Cathedral Mountain Formation, the two genera combine to define the Leonardian. Rugatia: This is a very easily identified productid of fairly large size, but it is restricted to the Cathedral Mountain and Road Canyon Formations. It is uncommon in the lower part of the Cathedral Mountain, is fairly frequent in the upper part, but rather rare in the Road Canyon. It is, nevertheless, a good guide to the Leonard rocks. Peniculauris: The large size and general resemblance to Reticulatia or Dictyoclostus make Peniculauris an easy brachiopod to identify. It appears in the lower part of the Skinner Ranch (Decie Ranch Member), but it is rare throughout the Skinner Ranch. It is abundant in the Taylor Ranch Member and is rare to fairly common throughout the Cathedral Mountain. It is rare in the Road Canyon, where it reaches a very large size. The genus is thus a guide to the late Wolfcampian (as defined herein) and the Leonardian. Liosotella-Paucispinifera: These two, which are similar internally but differ chiefly in degree of ornamentation, are highly characteristic of the Word Formation from the China Tank through the Appel Ranch Members. Their value is somewhat marred by the fact that both of them appear first in the Road Canyon Formation. Yakovlevia: This genus has the same range as the previous two and is equally valuable. Its four major spines, two on the ears and two at the place of geniculation, are unmistakable. Spiriferella: This easy-to-recognize spiriferid is common in the Willis Ranch Member of the Word Formation, less so in the China Tank and Appel Ranch Members. The genus is not an ideal guide fossil because it appears first in the top of the Road Canyon Formation. It is, however, extremely NUMBER 14 75 rare there and is not likely to be found except by intensive collecting. Facies in the Glass Mountains The Glass Mountains section of the Permian has long been a classic one in which to study facies. Although the types of rocks in the section are many, the column is predominantly one of coarse elastics with the fine-grained rocks generally in the minority. Facies may be indicators of environment, especially when the lithic characters and fossil contents are taken together. The oscillating environments indicate movements of land or sea, or changes in climate that affected the kind and amount of sediment delivered to the site of deposition. The Glass Mountains section may be generalized as a largely clastic sequence, with coarse sediments (Wolfcamp and Leonard) in the lower two-thirds and with the upper one-third consisting of finer grained sediments. LIMITATIONS OF THE DISCUSSION.—This discussion of facies is not bolstered by detailed petrological analyses of the sediments, which are beyond the scope of this study and the capabilities of the writers. It is believed, however, that a general discussion of the facies, with an effort to record the fossil occurrences in relation to the sediments, will be of value to our consideration of the paleoecology of the region and will also be of value to sedimentationists who may try to read the story of the rocks without considering the fossils. CONGLOMERATES.—These are very conspicuous in the lower parts of the section, but they are rare or absent from the Word Formation. They are thick and common in the Wolfcampian but thinner in the Leonardian, although occurring in all parts of the section. The conglomerates may be divided crudely into those composed of transported material and those thought to be essentially residual. The most conspicuous conglomerate made up of transported cobbles is that of the Lenox Hills Formation from Dugout Mountain eastward to hill 5280, where it thins to disappearance on the west flank of the Hovey anticline. The same type of conglomerate appears on the east side of this anticline, but it is never so thick as on the west and is discontinuous in its occurrence. It attains a thickness of 300-400 feet in its western range, where it forms most of the Lenox Hills Formation. At its base it interfingers with limestone and bioherms containing the fauna of beds 12-14 (of P. B. King) of the Neal Ranch Formation. It thus represents an uplift that started late in Neal Ranch time and continued nearly to the end of Lenox Hills time. The rock is composed mostly of rounded limestone cobbles up to several inches in long diameter and of some angular cherty fragments believed to have been derived from the Marathon Basin. Fossils are rare or absent in this conglomerate except in the bioherms and thin clastic aprons at the base. A 40-foot thick conglomerate, without such large cobbles but with chert and quartz pebbles and some coarse sand, appears in the upper part of the Cathedral Mountain Formation. This occurs near the site of the Old Payne Ranch, where it appears at the base of the C Member of Ross. A thinner band of conglomerate, with cobbles up to 3 inches in long dimension, and considerable coarse sand appears near the top of the Cathedral Mountain Formation in hill 4861 on the Monument Spring quadrangle. It is especially mentioned for the occurrence in it of pockets made up almost wholly of large shells of the ammonite Perrinites. This conglomerate can be traced for several miles along the front of the Del Norte Mountains. It is noteworthy for the abundance of ammonites and other Leonardian fossils (Plate 13: figures 1, 2). Another type of conglomerate probably should be called a limestone conglomerate or conglomeratic limestone. This has a fine-grained limestone matrix enclosing small pebbles, ragged chert debris, and a variety of small fossils ranging from small foraminifers to sponges and ammonites. This type of conglomeratic limestone is often the most favorable lithology in which to find ammonites. Examples in the Wolfcampian may be seen at USNM 707j and 715, in the various limestones of the Dugout Mountain Member of the Skinner Ranch Formation, and in some limestones of the Third and Fourth Limestone Members of the former Leonard. Many beds of conglomeratic limestone occur in the Poplar Tank Member of the Skinner Ranch Formation. In some places the beds are fairly thick. A thick conglomerate occurs at the base of the Third Limestone of the Leonard of P. B. King in knob 5250, just west of hill 5300. The conglom76 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY FIGURE 26.—Diagram for the Permian of the Glass Mountains, showing shift from predominantly silicious facies in the west to carbonate facies in the east. Sections: (1) Dugout Mountain along the line of P. B. King's (1931:32, 132) section 7: (2) starting in the middle of the Lenox Hills and following the line of P. B. King's (1931:32, 135) section 11; (3) just west of Sullivan (Yates) Ranch road and essentially following P. B. King's (1931:32, 135, 136) section 12; (4) through hill 5021 (Decie Brothers Hill) on line of P. B. King's (1931:32,136) section 14; (5) through hill 5280 approximately along the line of P. B. King's (1931:32. 137) section 15; (6) through Leonard Mountain and hills to the north along the line of P. B. King's (1931: 32, 139) section 17; (7) in vicinity of Hess Ranch house and hills to northwest; (8) through Hess Ranch Horst (Word Formation projected from east); (9) through the Wolf Camp Hills essentially along the line of P. B. King's (1931:32, 143) section 24; (10) on Conoly Brooks Ranch along the line of P. B. King's (1931:32, 145) section 27; thicknesses are approximate, some from King's sections, some original (AR = Appel Ranch Member of Word Formation, C = covered, CM z= Cathedral Mountain Formation, CT = China Tank Member of Word Formation, DM = Dugout Mountain Member of Skinner Ranch Formation, DR = Decie Ranch Member of Skinner Ranch Formation, H^Hess Formation, L 3-4 = Third and Fourth Limestone Members of the Leonard of P. B. King, LHc = conglomerate of Lenox Hills Formation, LHls = limestone of Lenox Hills Formation, LHls —H — Hess Formation equivalent of Lenox Hills Formation, LHsh=shale of Lenox Hills Formation, NR — Neal Ranch Formation, PT = Poplar Tank Member of Skinner Ranch Formation, RC = Road Canyon Formation, sc=zScacchinella, SP = Sullivan Peak Member of Skinner Ranch Formation, SR = Skinncr Ranch Formation, TR = Taylor Ranch Member of Hess Formation, U=r Uddenites-bearing Shale Member of Gaptank Formation, W —Wedin Member of Cathedral Mountain Formation, Wd —Word Formation, WRzz Willis Ranch Member of Word Formation; in section 1 numbers l-5 = limestone members of Leonard Formation of P. B. King, in section 9 numbers 12-14 = beds 12-14 of P. B. King in Neal Ranch Formation [ = beds 9-12 of Cooper]; see text Figure 5). NUMBER 14 77 erate referred to above and most of those in the western part of the Glass Mountains often are composed of rounded limestone cobbles. An important conglomerate made up largely of small quartz pebbles, however, occurs at the base of the Cathedral Mountain Formation in the vicinity of Split Tank and extends as far west as the faulted blocks of the Skinner Ranch-Cathedral Mountain Formations on the south side of the Hess Ranch Horst. The ammonite-bearing conglomerates and conglomeratic limestones often contain numerous other fossils, but they are generally in a poor state of preservation. Those in the Dugout Mountain Member are usually broken and fragmentary, making efforts to obtain a fauna either by breaking rock or by dissolving it rather a frustrating experience. Some fragments such as those of Torynechus can be recognized, but most of the debris is a loss. In summary, it may be said that the fossils of the conglomerates are transported material, often fragmented, but, in the case of buoyant ammonites, frequently in very good preservation. All of the fossils probably represent transports from other environments except the examples that are found in bioherms in the conglomerates such as those at USNM 715b. By residual conglomerates we mean the beds of boulders with little transported material that surround some of the bioherms, especially those at the west end of the Lenox Hills. The boulders around or between the bioherms in that location often attain a length in one dimension of 4 or 5 feet. The boulders are ragged, helter skelter, but appear to be derived from the bioherms. All of the fossils seen in the boulders at the west end of the Lenox Hills contained Permian fossils. We did not see any boulders in which we recognized exotic species. Indeed, one boulder only a short distance from the side of a bioherm (hill 4801) abounded in Scacchinella of the same type common in the Sullivan Peak Member (Plate 11: figures 1, 4). It has been suggested that this material and the large bioherms represent reef slide. A similar occurrence in the Chinati Mountains was so interpreted by Rigby (1958:308) (Plate 23: figure 3). We believe that this is not reef slide in the Glass Mountains or in the Chinatis because the large biohermal masses are conglomerate based, and, in one instance, part of the bioherm has grown over the conglomerate (see discussion under "Bioherms"; Plate 18: figure 3). SANDSTONES.—Lenses of quartz sand appear in places in the Skinner Ranch Formation. These form part of Member C of the Leonard of Ross. In the upper part of the Word Formation, below Sullivan Peak and near the site of the Old Payne Ranch house, thick layers of sandstone occur. These sands, like the conglomerates, represent near-shore conditions in shallow, probably strongly moving, waters. Fossils are almost unknown in these sandstones, although a few Word species were taken from the long dip slope of the sandstone hogback west of the old Payne Ranch site (hill 4806). As explained under the heading of "Bioherms," a single anomalous example was found in a sandstone mass south of Sullivan Peak (USNM 727x). Sole-marked sandstones occur in the Lower Cathedral Mountain Formation at USNM 723y. The sands are not an important factor in the preservation of fossils in the Glass Mountains. Sands are commonest in the western part of the Glass Mountains and indicate an approach to the source of the sediments. In the limestone members of the Word Formation, sand forms an important part of the rock. This is exhibited especially by the Willis Ranch Member. This member contains an increasing amount of sand to the west, as shown in Gilliland Canyon, where the limestones contain stringers of sand (Plate 14: figure 3), and, with the increase in sand, there is a diminution of layers of fossils. Farther west of Gilliland Canyon, the sand becomes more and more abundant. Although quartz sands are the most familiar type of sandstone, they are much less developed in the Glass Mountains, than are the lime sands, calcarenites, or calcirudites. The former range from finegrained to coarse-grained and are common in all parts of the section, but they are particularly conspicuous in the Skinner Ranch Formation, where they probably represent a former sand bar. As noted in the stratigraphic section, the Decie Ranch and Sullivan Peak Members pinch out the predominantly shaly Poplar Tank Member to form the Skinner Ranch Formation undivided. This consists largely of calcarenite that forms a barrier interfingering on the east with the fine calcarenites, oolites, and other lagoonal types of limestone of 78 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY the Hess Formation but sending its Decie Ranch and Sullivan Peak tongues to the west. The aprons of clastic material, mostly broken shells and crinoid debris surrounding the bioherms, are other examples of lime sand. These occur in the lower part of the Skinner Ranch Formation at USNM 705a and 720e. Coarse lime sands or calcirudites are common in the Road Canyon Formation but less so in the Word Formation. A bioclastic sand facies of considerable interest is that of the fusulinid sands or fine gravels that are common in the Road Canyon in the Monument Spring quadrangle. Here, as at USNM 732j and 736x, beds of limestone 1—3 feet thick are composed almost wholly of fusulinid shells. These represent lag deposits in which the finer material has been winnowed out and the heavier and coarser debris left behind. These beds are common in the Road Canyon Formation from Sullivan Peak westward, the part of the region where the Road Canyon begins to fray out in the thickening Leonard yellow shales. How many living brachiopods existed in these bioclastics is a question. Most of the specimens dissolved from them are worn, broken, or single valves. Good specimens are uncommon, but some of them survived the obviously rough conditions that existed at the time. SHALE.—For convenience of discussion, the shales are divided into the blue to blue-black shales and the yellow shales, trie latter characterizing parts of the Leonard and Word columns. The Neal Ranch Formation and the upper part of the Lenox Hills Formation contain thick sequences of bluish shale. In the Neal Ranch Formation shale is predominant and separates thin layers of limestone and some limestones near the middle of the section that swell into bioherms. The lower shale of the Neal Ranch below bed 12 of King (= bed 9 of Cooper) is blue-black and contains few fossils; those that it does contain are predominantly Pennsylvanian types. The same may be said of the shale of the Uddenites-bearing Shale Member. It is the biohermal beds that contain the Permian types in both sequences. The upper shale of the Lenox Hills Formation in the Lenox Hills, which disappears eastward, is bluish-gray in color, and the megafossils it contains are generally Pennsylvanian types or like those of the Uddenites-bearing Shale Member. The fusulinids, however, betray their Permian affinities. This shale also appears on Dugout Mountain. Blue shale as exposed at Clay Slide forms a substantial part of the upper Cathedral Mountain Formation. The shale is rather sparsely fossiliferous, but fine specimens of Perrinites and Peniculauris are washed out of it (Plate 7: figure 2). A shale unique in the area is that which comprises the bulk of the Poplar Tank Member of the Skinner Ranch Formation. It is dull brown, blocky, and breaks into angular pieces. It contains few fossils, but the thin conglomeratic limestones enclosed by it are usually fossiliferous, although the preservation is not good. The somewhat chocolate color of the Poplar Tank shale is in marked contrast to the yellow- to orange-weathering shale of the Cathedral Mountain Formation. The Cathedral Mountain Formation is essentially a wedge of yellow shale interfingering to the east with carbonate biohermal tongues at the thinning end of the wedge in the eastern part of the mountains. The shale is largely siliceous (radiolarian and spicule-bearing), breaking into ragged plates, some of which may be discolored pink or red in rings, the weathering following the outline of the plate. Blocky chert that contains radiolaria and sponge spicules is often present. Higher in the section, above the Wedin Member, the shale is less strongly colored and is mainly dull yellowish. These shales thicken westward and also may be seen as tongues in the Road Canyon Formation and as separators of the bioclastic beds in the Del Norte Mountains. Similar shale also separates the limestone members of the Word Formation. We have not seen macrofossils in the Cathedral Mountain and Road Canyon shales but the Word yellow shales contain rare specimens of Crurithyris and Leiorhynchoidea. These shales probably indicate quieter water deposition away from the shore, in bays, or the result of moderate depression of the shelf. CARBONATE FACIES.—The carbonates consist of dolomites and limestones, the former eractic in occurrence and of two kinds, the latter more concentrated and with a variety of types Dolomite: The dolomites may be divided into two kinds: bedded dolomites and massive, unbedded dolomites. The former type appears in the Hess Formation, where thin layers are often bedded NUMBER 14 79 with thin layers of limestone. The Hess also contains thick layers of dolomite that often have a considerable lateral extent. Alteration of the original limestone in some of the beds has left a porous rock, the holes of which represent fusulinids that must have occurred in countless numbers. The massive type of dolomite appears to have been formed subsequent to lithification and may have been induced by the intrusives that appear in places in the mountains. Leonard Mountain and the hill north of the Hess Ranch house are areas of considerable dolomitization, although the dolomite is patchy (Plate 8: figure 2). One can walk from dolomite into limestone, straddle the boundary, and see the alteration of the fossils from one lithology to the other. Perhaps the best example of this type of dolomite is the large mass on the southeast side of Leonard Mountain, where the thick dolomite has been mistaken for a reef. We have walked two levels, one high in the Lenox Hills, the other low in the Skinner Ranch, into this large mass. The presence in it of a large Omphalotrochus is evidence of Skinner Ranch affinities. Most of the hill on the north side of the Hess Ranch house is composed of dolomite but upper Lenox Hills Formation and lower Skinner Ranch Formation can be identified in undolomitized patches. The incompleteness of the dolomitization makes possible identification of the various levels, because good fossils may be obtained in them. Fossils can often be seen in the dolomite but, as a rule, they are unidentifiable ghosts. Incomplete dolomitization can be encountered in parts of the Skinner Ranch Formation as on the knob 0.8 mile northwest of the Hess Ranch house. The limestone of the Dugout Mountain Member of the Skinner Ranch Formation is dolomitized incompletely on the prominent knob (USNM 700r) about a half mile northwest of Dugout Mountain. Limestone: Perhaps the commonest type of limestone facies in the Glass Mountains was referred to under the heading of lime sands. These coarsely clastic limestones are most common in the lower two-thirds of the section, but two conspicuous areas of calcilutites occur in the Road Canyon Formation. These are the dark, nearly black, limestones that commonly are regarded as basinal in deposition and probably indicate fairly deep water. One area occurs on the north side of the Old Word Ranch site, and the other forms the long spur extending southward from Sullivan Peak. The locality near the Old Word Ranch is about 60 feet of thin-bedded, platy, bituminous limestone, black when fractured, but weathering to light gray. Fossils are fairly common in patches and in thin beds, but some of the layers are devoid of life remains. These thin beds rest on biohermal limestones at the base of the Road Canyon Formation. The occurrence at Sullivan Peak (USNM 707e) rests on biohermal lower Road Canyon, and the bituminous platy limestone fills depressions in the biohermal beds. The black limestone is more than 300 feet thick. In parts of the section it is without fossils, but in the upper part fossils are very common and varied. The presence of abundant fossils suggests that this is not a deep water deposit, but probably it is a bay of quiet water on the old shelf. The thin-bedded limestone of the Hess Formation, often interbedded with thin layers of dolomite, has been interpreted as a lagoonal facies. That it is unlike any other limestone in the Glass Mountains is clear. The rock is often a fine-grained calcarenite, oolite, or pellet limestone. It is light colored, and some layers, especially near the top, contain fusulinids coated with layered limestone, suggesting algal deposition. Two other facies of the limestones need mention: the bioherms and the shell heaps. Both require more extended discussion than the preceding facies because a large share of the collections were taken from them. Organic Accumulations in the Glass Mountains BIOHERMS Mound-shaped structures containing numerous fossils are abundant in many of the stratigraphic units in the Glass Mountains below the level of the Word Formation. They range in size from a few inches high and a foot or two across to more than 80 feet high and 100—200 feet across. Some of these in the small and middle size ranges are "currentdeposited" heaps of shells, which are characterized by stratification and well-packed shells and other organic debris that show some signs of abrasion. Others of all sizes have some parts bound together in interlocking frameworks of skeletal organisms 80 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY that almost certainly were self-supporting during their lifetimes. We use the term "bioherm" for the mound-shaped structures with coherent organic framework, rather than employ the ambiguous and perhaps inappropriate term "reef." This distinction follows the recommendation of Cloud (1952) that the term "reef" be used for relatively large organic buildups that headed near the surf zone and were wave resistant and the term "bioherm" for "reef-like organic masses of uncertain, potential, or doubtfully wave-resistant nature." Cumings (1932) proposed the term "biostrome" for bedded shell heaps, crinoid accumulations, or coral growths that did not produce mounds; this term is not properly applicable to certain bedded structures common in the Glass Mountains. These appear to be ordinary stratified beds, bounded top and bottom by bedding surfaces, but close inspection or leaching in acid shows them to be organically interlocking frameworks with interstitial micritic calcite, directly analogous to bioherms except that they failed to produce the mound shape. These are perhaps incipient bioherms that were inhibited by early burial or an unfavorable situation. They may coincide with the reefoid facies of Bain (1967), but, because they are prevalent in the Glass Mountains, we are using the name "zotikepium" from the Greek "zotikos" ( = lively or vigorous) and "kepion" ( = garden), in allusion to the lush garden-like nature of the bryozoan and sponge growths that form large parts of the structures. Both biohermal and zotikepial structures are common in the Glass Mountains in all of the formations and members, from the Uddenites- bearing Shale Member of the Gaptank Formation through the Road Canyon Formation. The Word Formation has none of these structures. This discussion does not take into account the Guadalupian formations above the Word, which are mostly dolomitized and not sufficiently studied to individualize important organic structures, except for the gross form of the Capitan "reef." DIFFICULTIES IN THE STUDY.—The chief problem in the study of these complex but interesting structures is the fact that they are seldom seen completely or in three dimensions. Cross-sections are common; partly exhumed bioherms revealing their tops also occur frequently, but the bases seldom show. Many bioherms occur as bevelled mounds with some part cut away, either horizontally, vertically, or obliquely, to show the interior. Fortunately, many of the bioherms of the Glass Mountains contain fine silicified fossils that reveal the framework of the structure in detail. Localities such as USNM 702un, 714w, 719x, 726o, and others yield large blocks, which, when decalcified, show their structure to be mainly a mass of bryozoans, sponges, or, more rarely, brachiopods. Bioherms range in size from a few feet in height to about 80 feet, which is the thickest we have seen, with no obvious pattern. Generally, however, the largest ones are in the Sullivan Peak Member of the Skinner Ranch Formation and in the Road Canyon Formation. Little has been published about the bioherms of the Glass Mountains. The King brothers did not discuss them, although P. B. King (1932) interpreted the dolomite mass on Leonard Mountain as a reef. The only serious studies, so far, of the bioherms are by Grant (1971), who analyzed their brachiopod faunas, and Bain (1967), who made a detailed investigation of the Cathedral Mountain bioherms in the vicinity of Split Tank (Hess Canyon quadrangle). Bain recognized a reef core with flanking beds of debris, the classic picture of the bioherm of the Paleozoic. He also recognized four other facies connected with the bioherm not hitherto described. In discussing bioherms here, we will consider them according to their stratigraphic sequence. PENNSYLVANIAN.—One area of bioherms appears in the Pennsylvanian of the Marathon Basin south of the Arnold Ranch. This bioherm is located just 1.25 miles south of the Arnold Ranch, Monument Spring quadrangle (Plate 16: figure 1). It is a low, rounded mass of hard, dark gray limestone about 15 feet long and 2—3 feet high, although its base is not exposed. The limestone contains an abundance of the fasciculate coral Amplexocarinio delicata Ross and Ross (1963b) and considerable laminated limestone of possible algal origin. The freshness of the rock prevents collecting good fossils; nevertheless, among others, a primitive species of Scacchinella and a Limbella were obtained. The bioherm is lodged in Gaptank Shale, dated by P. B. King (1938) as equivalent to his bed 10 of the type Gaptank sequence. One fusulinid, found after NUMBER 14 81 exhaustive search in the bioherm, was declared by Garner Wilde (letter of 20 March 1962) to be Virgilian in age, which disagrees with King's assignment of bed 10 of the Gaptank to the Canyon in the Texan nomenclature (i.e., Upper Missourian). This bioherm is anomalous because of the Permian-type fossils it contains. It occurs in a complexly folded and faulted orogenic belt and might be an exotic block. If so, its place of origin is unknown, because nowhere in the Gaptank Formation or the Lower Permian is a fauna like it known. In spite of these difficulties, it is composed of reef-core rock and has the characteristic form of a bioherm. WOLFCAMP SERIES.—Bioherms are abundant in the lower Wolfcamp formations, but infrequent higher up. They are a problem in stratigraphy, because they abruptly swell the thin limestone beds and make measurements difficult on poorly exposed slopes or in ravines. One of the astonishing characteristics of these bioherms is their variety of faunal composition; no two have the same fauna even though they may be in fairly close proximity. It is not possible to discuss all of the bioherms studied and collected in the Uddenites-bearing shale member and the Neal Ranch Formation; a few in different parts of the formation will serve as examples. Neal Ranch Formation: This formation is about 500 feet thick and consists of a matrix of dark shale enclosing many lenses and layers of limestone. The thickest layer of limestone is bed 2 of P. B. King, otherwise known as the Gray Limestone, occurring at the base of the section. This is a great lenticular mass of calcarenite having a maximum thickness of nearly a hundred feet and a lateral extent of slightly more than a mile. It is thickest near its middle, capping the cuesta called hill 5060, and is the most prominent ledge of the Wolf Camp Hills. It tapers laterally to a few feet at the west end, but it is fairly thick (135 feet) where last seen on the east. This large lens might be interpreted as a bioherm or as a lime sand bank, but it has never been studied in detail; consequently, its composition is not known. Some brecciation occurs on the west side of the main mass. We found no silicification in the main mass of this lens, where it is thickest, but we found bioherms in its north flank. Two bioherms of local importance occur on the long northward dipping slope, just before the lens plunges below the surface. The first is small (USNM 722x), only a few feet in lateral extent, and now mainly a remnant, weathered and rotted. It is composed mainly of a new species of Eolyttonia. Most of the brachiopods are surrounded by laminated limestone of undoubted algal origin, which welds the mass together. This bioherm certainly was larger originally. It is most important because it establishes the presence of Eolyttonia, a Permian genus, in the Gray Limestone of P. B. King, which Ross (1963a: 45) places in the Pennsylvanian (Virgilian). The second bioherm in the Gray Limestone of P. B. King occurs in the bed of Geologists Canyon at its junction with a gully from the north, at approximately the point where the 4650-foot contour (Hess Canyon quadrangle) crosses the main canyon floor (USNM 701). This structure can be seen only in plan, but it consists of a matted mass of bryozoans, sponges, and algal material containing an unusual assemblage of brachiopods (Hypopsia, Schuchertella, Eolyttonia, Tropidelasma). The brachiopods do not make up much of the mass of the bioherm, which is largely constructed of bryozoans. Perhaps the most nearly ideal bioherm in the Neal Ranch Formation is that at USNM 701c (Plate 16: figure 3), which forms the top of the knob represented by the easternmost ring that is formed by the 4900-foot contour in the hill just northwest of the "W" in "Wolf Camp Hills" (Hess Canyon quadrangle, edition of 1923, reprinted in 1946; prior editions of this quadrangle do not bear the name "Wolf Camp Hills"). This bioherm is a rounded mass about 20 feet thick which consists of amorphous unbedded micritic limestone strongly contrasting with the relatively thin-bedded calcarenites in which it is embedded. It rests on a 2—3 foot thick bed of limestone cobbles welded together by shale. The rock appears to be finegrained, but, when leached by acid, it shows a framework of sponges, bryozoans, and fine siliceous granular material and bioclastic debris. The remarkable feature of this bioherm is the amazing large specimens of Eolyttonia, nearly as large as a tea-cup saucer (Plate 157: figures 1, 2). Numerous other types of brachiopods are fairly common also. 82 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY" The fauna of this bioherm is in marked contrast to that forming the twin knob of this hill about 0.25 mile southwest of USNM 701c. The bioherm on the second knob (USNM 701h) is small and seen only in horizontal section. It was an irregular patch with height of two or three feet above the enclosing rock. Decalcified blocks of this limestone proved to be a welded mass of brachiopods, chiefly the genera Geyerella, Meekella, and Limbella, with abundant sponges and bryozoans. The rock appears to have no grain and, therefore, stands in strong contrast to the surrounding bioclastic calcarenite. In spite of the grainless appearance, it has fragments of echinoderms, small foraminifera, and some lumps of bedded limestone of possible algal origin. Decalcification revealed a variety of brachiopods other than the frame builders, especially a huge new species of Parenteletes with a full range of growth stages, from the earliest shell to large adults three inches in width. Although this bioherm is only 0.25 mile away from the one at USNM 701c, the two faunas are utterly unlike (Plate 11: figure 3). At the same stratigraphic level, another bioherm about 0.25 mile to the northwest (USNM 701k) produced an entirely different fauna in which Enteletes is abundant. This genus showed preference for bryozoan gardens. Some of the Wedin Member structures contained an abundance of this genus. Although the thin limestones above bed 14 of P. B. King swell from biohermal development, we found none that contained silicified fossils. Bioherms are not common in the overlying Lenox Hills Formation, but a fair development was discovered at the west end of the Lenox Hills. Lenox Hills Formation: No bioherms were seen in the Lenox Hills Formation between the Wolf Camp Hills and the Lenox Hills. These structures appear at the west end of the Lenox Hills westward from about hill 4902. At USNM 715b (Plate 12: figure 2) numerous bioherms occur at the base of the section, enclosed in coarse calcarenite and conglomerate. These are the beds that Ross (1963a: 24) identified as Neal Ranch Formation and claimed were in unconformable contact with the Lenox Hills conglomerate. The angularity is undoubtedly due to the irregular bedding usually associated with bioherms. But we found, near the base of the biohermal beds, interfingering layers of conglomerate that suggest the bioherms belong to the Lenox Hills Formation. At USNM 707j a few small bioherms occur 25 feet up in the conglomerate, and there is no question about their belonging to the Lenox Hills Formation. The fossils of the bioherms at USNM 707j and 715b are clearly of Neal Ranch (beds 12-14) affinity, but the stratigraphic setting indicates that they were contemporary with the initial phases of Lenox Hills deposition. They have a rich brachiopod fauna with some Bivalvia and ammonites. No bioherms were seen in the Lenox Hills Formation of the Dugout Mountain area. A unique bioherm in the Lenox Hills Formation is that on the southeastern nose of Leonard Mountain, just above the Lenox Hills conglomerate (USNM 705k; Plate 20: figure 2). This is a small mass, short in lateral extent but about 20 feet thick, containing Tropidelasma and a small Scacchinella as its most abundant constituents. Heliospongia and Parenteletes are also present. The rock is brownish gray with orange-brown, earthy patches. The bioherm is surrounded by fusulinid-bearing limestone conglomerate with shale matrix, the common base on which Wolfcampian bioherms rest. Around the base of Leonard Mountain are large slipped (toreva) blocks of Lenox Hills limestone, often containing bioherms. One of these, USNM 705m, contains Scacchinella and Parenteletes. Although only one Scacchinella bioherm was seen in place, that in the detached block indicates that they were scattered about in the Lenox Hills limestone above the conglomerate in Leonard Mountain. To summarize, two kinds of bioherms characterize the Wolfcampian: one is the classic, moundlike patch reef originating on a limestone gravel but building a solid mass of cementing organisms surrounded by debris from the bioherm; the second is rather formless, often in relatively thin beds representing patches or gardens of bryozoans, sponges, algae, and combinations that include accessory niche dwellers. The apparently amorphous, smooth-appearing limestone is fine calcareous debris that is sifted by precipitation in the quiet waters which occupy the spaces between the shells and the animal colonies that are formed by the life NUMBER 14 83 process of the animals. Lithically, this forms a micritic matrix for the numerous skeletons. These are the zotikepia defined earlier; USNM 72lg is an excellent example. These bioherms and zotikepia evidently existed in well-aeriated and agitated waters, but the bulk of the reefy bodies and the niches within them furnished quiet places where brachiopods could flourish. Skinner Ranch Formation: Some of the largest and most interesting of the bioherms are those of the Skinner Ranch Formation. More of the classic, mound-like masses are present, but so is the horizontal type. 1. Decie Ranch Member. This member in its extent from hill 5021 to its last appearance on Dugout Mountain has no well-formed mound-like masses. The member is composed mainly of conglomerates, often with large boulders, but within these conglomerates are patches of Scacchinella with other rugged,' cemented-reef types of brachiopods. Scacchinella clusters often occur with accumulations of huge crinoid stems, the latter often more than two inches in diameter, and so large and numerous that they make up local gravel patches. The Scacchinella community flourished on this gravel and developed into clusters or patch reefs. The rock in places is a mass of Scacchinella and, when rotted, is spongy and crumbly from the vesicular character of the large conical pedicle valves. Collecting from weathered masses seldom yields a good specimen, but, contrariwise, it is almost impossible to get good specimens from fresh rock. Scacchinella bioherms are difficult to locate, but the basal crinoid accumulations are often a clue to them. Because very little silicification has taken place in the Decie Ranch Member, the individual Scacchinella does not stand out conspicuously in these bioherms as it does in those of the basal Skinner Ranch Formation east of Leonard Mountain, north of the Hess Ranch, and on the north slope of the Hess Ranch Horst. The Decie Ranch Member occurrences suggest a rocky near-shore area, with communities of robust Scacchinella that are intimately grown with bryozoans (especially Meekoporella), which are scattered among boulders in shallow, perhaps turbulent water. 2. Poplar Tank Member. One bioherm that was studied and collected occurs in this member (USNM 708e; Plate 21: figure 2). This bioherm is 15-20 feet thick and about 60 feet laterally, underlain by and surrounded by limestone conglomerate. The rock is dark bluish gray, but it weathers to a light gray. Of fine-grained appearance, it contains considerable coarse debris. This bioherm is not silicified. It is characterized by abundant small productids (Oncosarina) and large Antronaria (new genus), along with rare Scacchinella. This seems to be more of an accumulation of biological debris in a conglomerate than a reefy mass. No reef core was seen, but it may not have been exposed (Plate 21: figure 2). 3. Sullivan Peak Member. Bioherms varying in size from small patches to large reef-like structures are fairly common in this member. These can be seen as light-colored, rounded masses weathered into relief on the hillsides. Several appear on the southeast slope of hill 5300, but the best examples are located in hill 4801 facing U. S. Highway 90 at the very end of the Lenox Hills (Plate 18: figures 1-4). In lateral extent, one of these is among the largest bioherms in the mountains: 110 feet horizontally and 38 feet thick. The wellrounded top has a thin silicious skin. The base is conglomerate to a height of 5 feet, with boulders up to a foot in one dimension, and contains huge crinoid stems (Plate 18: figure 4). Above the crinoid stems the rock becomes fine grained, with numerous silicified bryozoans protruding from the surface. Scacchinella was not seen, but it probably existed in the mass, as it was seen in boulders flanking the bioherm. Another, smaller bioherm occurs west of the large one. The two are separated by a conglomerate 30 feet thick, composed of large boulders, occasional ones having a long diameter of 5 feet. This mass contains Geyerella, a common reef form. Several specimens are visible on the upper surface of this bioherm. On the west side of the bioherm an extension of its fine-grained limestone overlies the conglomerate (Plate 18: figure 3). This lateral extension of the biohermal rock contains Acritosia and numerous sponges. Fossils contained in the boulders flanking or underlying the bioherm tongue proved to be types common in the bioherms. One block near the margin of the bioherm contains numerous Scacchinella. These fossiliferous pieces indicate that the boulders in the conglomerate were derived from bioherms. 84 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY A short distance west of the large bioherms there occurs a small one, which is located on the bench around hill 4801 formed by the Sullivan Peak Member (Plate 16: figure 2). This small mound, 10 or 12 feet on one diameter and 2 or 3 feet high, protrudes from the conglomerate. 4. Skinner Ranch Formation Undivided. The Decie Ranch and Sullivan Peak Members pinch out the Poplar Tank Member and become a single unit in hill 5021, where the Skinner Ranch Formation becomes undivided. Nevertheless, the lower part of the formation retains many elements that identify it with the Decie Ranch Member. One of these is Scacchinella, which is very abundant and makes small reefs or bioherms almost completely composed of this genus. Three areas for the examination of these bioherms are: (1) the southwest side of Leonard Mountain about one mile north of the Iron Mountain (Skinner) Ranch; (2) one-half mile north of the Hess Ranch house; and (3) the base of the north slope of the Hess Ranch Horst. (1) Southwest Side of Leonard Mountain. Scacchinella occurs in the lowest beds, where small clusters may be found. On the east side of the ravine near the center of the mountain (Plate 19: figure 1) four characteristic biohermal mounds may be seen. Originally they may have occupied the same level and then been offset by slumping (Plate 19: figure 1). (2) North of Hess Ranch. Bioherms exposed here (USNM 705a) are all low, not more than two or three feet high. Scacchinella is very abundant, almost to the exclusion of all other fossils. These mounds are surrounded by layers of interbiohermal, bioclastic debris of worn and broken shells and crinoid stems. The Scacchinella are silica-filled, but, nevertheless, they cannot be freed from the limestone in good condition; usually they are recovered as inner fillings. Good free brachial valves are obtained from the interbiohermal beds, where they were washed after being freed from the pedicle valves that were anchored to the reefy mass. This locality is the site of numerous sponges, chiefly Defordia and Heliospongia, that grew about the bioherms. They are found usually in the interbiohermal beds along with a very large species of Omphalotrochus. At this locality the bioherms at the base of the hill are well preserved, but the limestone extending up the hill is usually dolomitized and the fossils destroyed. On the top of the hill overlooking the Hess Ranch house, limey patches in the dolomite contain good fossils and evidence of biohermal structures. The same types of bioherms with Scacchinella and Geyerella may be seen in the "amphitheater" in the north side of Leonard Mountain (Plate 11: figure 2). (3) North Side of Hess Ranch Horst. This setting is exactly like that north of the Hess Ranch House, but the area of exposure is much larger and the beds are not dolomitized. Laterally the area of bioherms extends from the fault on the east side of the horst to the west end of hill 5305, a distance of about 2 miles and involving about 200 feet of strata. Bioherms occur at several levels and contain different faunas. Some are masses composed almost wholly of Scacchinella in the lower part of the section and may represent reef-cores USNM 720e; Plate 19: figure 4). The silicification like that north of the Hess Ranch is not good, but some silicafilled young individuals are excellent for studies of variation. Above the Scacchinella bioherms there are others with the mound-like form. Although most are small and low, they have a rich fossil content. The compartmented bryozoan Meekoporella is common and offers niches wherein many of the smaller brachiopods dwelt. In these, Scacchinella is rare or lacking, but enormous specimens of Meekella and Derbyia are attached among the bryozoans; productids are uncommon. Many of these bioherms are surrounded by bioclastic debris of broken and disarticulated shells and layers of fusulinids (including Schwagerina crassitectoria Dunbar and Skinner). The upper part of the Skinner Ranch Formation is exposed poorly from east of Leonard Mountain to the east side of the Hess Ranch Horst. Furthermore, east of Hess Ranch the formation changes facies into that of the Hess Formation. Hess Formation (Taylor Ranch Member): One large bioherm occurs in the Taylor Ranch Member of the Hess Formation. At USNM 702d several masses of light, yellowish-gray limestone abound in the sponges Heliospongia and Girtyocoelia, together with a variety of other fossils, especially gastropods (Plate 19: figure 1). NUMBER 14 85 LEONARD SERIES.-A1 though Bain (1967:212) reports that "Reef masses range from 20 feet to 125 feet long and up to 100 feet high," most of these structures are much smaller. Many of them do not have the characteristic mound-like form, but they are micritic masses in relatively thin-bedded strata, probably representing local concentrations of organisms, usually bryozoans, alive for a sufficient time to cement the whole with finely sifted calcium carbonate, but with little binding algae. Bioherms occur throughout the Leonardian. They are common, especially the reefoid type of Bain, in the Wedin Member on the north slope of Dugout Mountain and in the west end of the Lenox Hills. They are scattered between Leonard Mountain and hill 5674 to the north. They are also concentrated in the area east and west of Split Tank on the Appel Ranch. Here they have been minutely studied by Bain (1967). Cathedral Mountain Formation (Wedin Member): These are of the reefoid type described by Bain and are usually composed of a framework of ribbon-like and fenestellid bryozoans. Every nook and cranny in these masses is occupied by cementing productids or other brachiopods: Hercosia, Chonosteges, Institella, and Agelesia. Plump Enteletes and an ocasional lyttoniid are present. Institella is uncommon in the bioherms at USNM 700x on the north side of Dugout Mountain, but it is abundant at USNM 714w at the west end of the Lenox Hills, where it and Agelesia make up a substantial amount of the organic mass. We look on these bioherms, which are not rounded mounds, as bryozoan patches resident for sufficient time to build matted masses of dead forms to which surviving generations attached. Finely precipitated calcite (forming micrite), formed by the life process of the animals from lime-saturated waters, sifted into the spaces between the bryozoans, dead and alive, cementing the whole into a solid mass unlike the surrounding sediments. These are the structures to which we apply the name "zotikepium." Bain (1967) examined the bioherms from Split Tank for about 2000 feet to the east (USNM 702, 702a, 702ent, 702inst, and 702un). He identified six facies types associated with, or comprising, the bioherms. His map and text do not show any definite pattern to the facies. Reef cores usually are surrounded by the reefoid facies, but they may be surrounded by breccia, flank-bank facies (rare), or inter-reef facies. The latter facies is characterized mainly by normal deposition of biomicrite and micrite containing fossils wafted by currents from their moorings. The strata range from a few inches to slightly more than a foot thick. This facies is the widest spread of all. The bioclastic facies comprises biomicrudite and biomicrite surrounding the bioherms. The map does not show clearly the extent of this facies. It contains many fragmentary fossils such as pieces of fenestellid bryozoans, crinoidal debris, echinoid spines and plates, as well as brachiopods and some molluscs. This facies was seen about the bioherms in the Skinner Ranch Formation north of the Hess Ranch and on the north side of the Hess Ranch Horst. The molluscan facies is dark gray micrite occurring as lenses. It contains numerous gastropods, pelecypods, and cephalopods, representing a local, quiet water environment in which fine mud collected. The flank-bank facies refers to concentrations of unbroken fossil material occurring on the margins of the reef cores or as isolated accumulations on bioclastic beds unrelated to reef growth. They may be zotikepia as described above. Concentrations adjacent to reefs are referred to by Bain as "flank deposits," and those unrelated to bioherms are termed "banks." These consist of dark brown to dark gray clastic limestone containing numerous brachiopods, bryozoans, corals, and echinoderms. The interstices between the fossils are filled with micrite. The banks are usually small, but the flank deposits may be large, up to 100 by 50 feet. The reefoid facies refers to "massive accumulations of limestone which circumscribe reef cores" (Bain, 1967:224). It usually is limited in size, a few feet in diameter, and not well bound. The rock is gray biomicrite and biomicrudite, grading into the bioclastic facies. Skeletal material of brachiopods, bryozoans, sponges, and algae are cemented by fine micrite of physicochemical or biochemical origin, resulting in a fine-grained rock. According to Bain, Prorichthofenia (•=. Hercosia) is one of the major frame-builders of the reefoid structures and is common in the Split Tank area. He postulates a depth of water not exceeding 60 feet for this facies. 86 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY The reef core facies is composed of light gray, dense, micritic masses ranging up to 125 feet long and 75 feet high. The framebuilding organisms are mainly fenestrate bryozoans and the bryozoanalgal consortium Acanthocladia. Of lesser importance are sponges and fistuliporoid bryozoans. Stromatolitic algae are the principal binding organisms and comprise the dominant portion of reef cores (Bain, 1967:226). Accessory organisms such as brachiopods, rugose corals, and gastropods often are found in the reef core. We dissolved blocks from the east of Split Tank about 0.25 to 0.5 mile from the bioherms at USNM 702un and 702inst. These suggest Bain's reefoid or flank-bank facies and are like those of the Wedin Member in their abundance of Institella. These are located between the fine pebble conglomerate and the first silicious shale stratum. They appear to be assigned to the Hess by Bain, but, occurring above the conglomerate and containing Institella, they clearly belong to the Cathedral Mountain, as defined by Cooper and Grant (1966), and above the Hess, by P. B. King's (1931) definition of the Leonard Formation. Probable reefoid facies with abundant Hercosia occurs above the shale at USNM 702ent and 702. Flank-bank or reefoid deposits abounding in the new lyttoniid genus Collemataria (USNM 702al) occur still higher in the section (Plate 20: figure 3). Above this level bioherms do not occur in the Cathedral Mountain Formation in the Split Tank area. Bioherms are common in the Old Word Ranch area southwest of Appel Ranch house. One of these, close to the site of the Old Word Ranch (USNM 703bs), composed of light gray micrite, abounded in Agelesia, thus relating the lower Cathedral Mountain of this area to the Wedin Member of the extreme western end of the mountains (Plate 20: figure 4). Still farther southwest, 1 mile below the Old Word Ranch site is USNM 726o (Plate 17: figure 1), one of the finest and most fossiliferous of the bioherms. It is a mass of an undescribed species of the brachiopod Hercosia, with abundant bryozoans, sponges, and some large pelecypods. The matrix is a smooth, medium gray micrite, undoubtedly of biochemical origin. The pelecypods are mostly Pectens, long scabbard-shaped Myalinas, and large spiny Pseudomonotis. One of the most interesting of the Cathedral Mountain bioherms occurs in the isolated hill about 1.5 miles south of Sullivan Peak (USNM 727x). This hill is slightly more than 100 feet high and is capped by 65 feet of fine-grained, brownish, sugary sandstone. In the midst of the sandstone there is a typical small bioherm of greenish- to bluish-gray limestone, about 15 feet wide and 6 feet high, containing Institella. Although we did not see evidence of displacement, this occurrence might be an exotic block, as it is surrounded by sandstone. Nevertheless, its internal structure is clearly that of a bioherm. Road Canyon Formation: These bioherms mainly comprise two kinds: one type abounds in a variety of fossils but has a framework composed mostly of bryozoans and sponges; the other type contains great masses of the bizarre oldhaminid Coscinophora. The first type generally is located at the base of the formation, while the Coscinophora patches are scattered at several levels. The first type, with a variety of fossils, is widespread. The bioherms are variable in dimension, but they attain heights of 15-20 feet. They occur from the Old Word Ranch site to the Sullivan Peak area. Our USNM 703a, 702c, and 719x are examples. Other bioherms can be studied on the east side of the low spur extending south of Sullivan Peak (Plate 17: figure 2), where they are overlain by the thin-bedded bituminous limestone of the main mass of the Road Canyon Formation. These, like most other bioherms, have the appearance of unbedded, smooth, sublithographic limestone (micrite) having a greenish- to bluish-gray color when fresh but weathering light gray and standing out in conspicuous contrast to the surrounding rocks. The upper surfaces, when exposed, are usually smooth and rounded. One of the commonest fossils in these bioherms is Hercosestria, a brachiopod that often nearly completely comprises small parts of the bioherm, but with Meekella and Derbyia attached within the mass. Edriosteges multispinosus Muir-Wood and Cooper is another common inhabitant of these bioherms. Most of the framework, however, is composed of bryozoans and sponges, which occur in countless numbers. At the Old Word Ranch site, we observed geopetal structures in a large Edriosteges (USNM 703a; Plate 21: figure 1), which were inclined parallel to NUMBER 14 87 the regional dip, suggesting the original horizontality and the normal growth position of the animal, good evidence that this bioherm is in place and undisturbed. At USNM 702c, in addition to the usual reefy structures of Hercosia and Edriosteges, fine-grained limestone, abounding in brachiopods with both valves intact almost to the exclusion of other fossils, occurs associated with the bioherms. These seem to be accumulations of shells piled up against the bioherms in very fine micritic matrix (Plate 20: figure 1). Another place where bioherms occur (Plate 14: figure 2; Plate 15: figure 2) in abundance is in the triangular hill north of Leonard Mountain. Bioherms can be found at the base of the Road Canyon Formation from the junction of Road and Hess Canyons, southward around the hill to Gilliland Canyon, and thence north to about bench mark 4869. An isolated area of bioherms occurs at the southwest end of Gilliland Canyon (USNM 724d). USNM 719x, 721x, and 721y contain fine bioherms, yielding Hercosia, Edriosteges, and a variety of other brachiopods. In the Dugout Mountain area the Road Canyon is frayed out into the shaly facies, but small bioherms still can be found. The largest of the Road Canyon bioherms is about 2 miles north of the Iron Mountain - Skinner Ranch house near the crest of the hill (USNM 724j; Plate 16: figure 4). It is about 80 feet thick, consisting of massive granular limestone of 44 feet, with a conglomeratic base, fine-grained calcarenite for 18 feet, and the upper 20 feet formed by massive, fine-grained limestone with bryozoans. It is overlain by yellow platy shale with thin-bedded limestone above that. The bioherm contains a great variety of fossils. Coscinophora bioherms are scattered about in the Road Canyon Formation at almost any level above the basal bioherms. One of the finest of these is at USNM 72lq (Plate 17: figures 2-4), with 35 feet of massive limestone rounded on top, forming a reefy mass. The lower 10 feet contains long slender corals and masses of Coscinophora. The middle 15 feet are mostly calcarenite, but the upper 10 feet again contain Coscinophora. Similar masses packed with Coscinophora almost to the exclusion of all else occur in this hill on the east side of Gilliland Canyon and on the hill just east of Sullivan Peak (USNM 709c and 710u). A Coscinophora bioherm was seen on the side of Sullivan Peak not far south of the Sullivan (Yates) Ranch. An occasional bioherm with Coscinophora occurs in the Del Norte Mountains, and loose specimens suggest the presence of others. A bioherm unlike all others in the Road Canyon Formation was found at the Old Word Ranch site (USNM 703c). This was a small patch composed almost completely of various types of sponges. Since only die top of this bioherm could be sampled, its inner form is not known. The sponges occurred in untold numbers and in considerable variety; Finks (1960:33) records 18 genera of siliceous sponges. In addition to the sponges, many and various molluscs were taken; Batten (1958) and Yochelson (1956 and 1960) described 25 species of gastropods from this bioherm. A large number of undescribed bivalves also were obtained. The occurrence strongly suggests the Cherry Canyon (Getaway Member) sponge bioherm discovered by N. D. Newell and party in the Delaware Basin on the south slope of the Guadalupe Mountains (AMNH 512 = USNM 728). A great variety of sponges occurs at this place, accompanied by a varied assemblage of gastropods and bivalves. We found no bioherms in the Word Formation of the Glass Mountains; most of the fossils occur in shell heaps or scattered in the sediment. SUMMARY.—We recognize two types of organic accumulations: (1) mounds or bioherms; and (2) the inconspicuous, unswollen beds of fossils that contain much material in positions of growth, termed here a "zotikepium" or lush garden. The bioherms generally consist of a core of hard, dense, micritic limestone (reef-core), surrounded by bioclastic material that is derived from the bioherm (bioclastic facies). The bioherm generally is based on conglomerate (often composed of huge crinoid stems in the Wolfcampian) onto which cementing forms attached themselves to begin building the framework, which usually consists of bryozoans, sponges, and algae (Plate 12: figure 1). Continued growth occurred in the same place, one generation living on the shells and skeletons of the preceding one. Into the interstices between the living and dead skeletons, coarse and fine animal debris sifted, along with the fine precipitated calcium carbonate from biochemical reactions, pro88 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY ducing the micritic matrix that binds the mass into a mound. In the zotikepium on the other hand, a mound was not produced, although the same processes went on over a broad area to form a laterally extended accumulation. Growth, fairly uniform throughout, did not concentrate to produce a mound. This type is essentially a biostrome, but the definition of that term covers any bedded accumulation of biological origin, with or without a reefy frame. SHELL HEAPS Under this terminology are included masses of shells or other organic debris not bound by micritic matrix and not giving evidence of having been an adjusted biologic community. These heaps are very abundant throughout the section. Many are undoubtedly bottom sweepings of currents piled into quiet areas and shallow depressions or heaps resulting from storms. Many of the fossils in the heaps were dead shells before their burial, as they show signs of wear and some are riddled by borings. On many specimens, however, the epifauna are well preserved with both valves intact. These undoubtedly were living when buried. In most of the shell heaps the brachiopod valves are disassociated, especially the loosely hinged enteletaceans, productids, orthotetaceans, and related forms. Spiriferids usually have the valves separated, but many of the rhynchonellaceans and terebratulaceans, which are tightly articulated in life, occur with their valves in contact. We found few conspicuous examples of this type of occurrence in the lower part of the Wolfcampian sequence. Here, fossiliferous beds are generally bioherms or zotikepia or debris derived from the bioherms (bioclastic facies). Beds of transported and dead shells are especially common in the Dugout Mountain Member of the Skinner Ranch Formation. Thin beds made up of small pebbles and shell debris are common in this member, but the fossils dissolved from them are mostly worthless because they are so fragmentary and worn. Exceptions do occur, as at USNM 722—1, where sparse shell heaps yield abundant, fine fossils. Dead shell assemblages (thanatocoenoses) appear higher in the Skinner Ranch Formation in the Poplar Tank and Sullivan Peak Members. These shells may cover bedding surfaces or be abundant in the siliceous cherty skins that cover some of the limestone beds. Some of the thicker cherts also contain fossils on the surface. Examples of shell heaps occur in the lower part of the Cathedral Mountain Formation at USNM 702b and 721u. Here, countless shells—mostly single valves of brachiopods, pelecypods, and gastropods, broken sponges, small cup corals, and battered bryozoans—are heaped together and cemented in a sandy lime matrix. Sponge spicules are abundant in the sediment, but fusulinids are rare. The faunas represent the Institella-Agelesia assemblage that characterizes most of the bioherms of the Wedin Member and the lower part of the Cathedral Mountain Formation. The shell heap at USNM 72 lu suggests current and storm sweepings from nearby or moderately distant bioherms, because "reef-dwelling" types appear in the debris. Another shell heap occurrence entirely unlike that just mentioned occurs in the upper part of the Cathedral Mountain Formation in hill 4861 about 1 mile northwest of the Old Payne Ranch site (USNM 732u; Plate 13: figures 1 and 2). The rock is a coarse conglomerate with pebbles attaining a long diameter of 2—3 inches. In the conglomerate there are patches almost completely made up of shells of the large ammonite Perrinites. Scattered through the conglomerate, and with the ammonite masses, are typical Cathedral Mountain brachiopods such as Peniculauris and Rugatia. Besides Perrinites, large nautiloids, Medlicottia, and a few other ammonites have been found. We suggest that these banks of ammonite shells represent dead conchs floated onto shoals or beaches as described by Hamada (1964) for Nautilus pompilius in the Pacific today. The Road Canyon Formation forming the spur on the south side of Sullivan Peak contains shell heaps and layers, mostly dead shells, in the thinbedded bituminous limestone making up most of the formation at this place. At the top of the Road Canyon south of Sullivan Peak there are thin beds that constitute a fusulinid sand or gravel. Similarly constituted beds attaining a thickness of 1—3 feet are common in the Road Canyon Formation in the Old Payne Ranch area and on the lower slopes of the Del Norte Mountains. A rich locality NUMBER 14 89 for brachiopods as well as fusulinids is at USNM 732j, where Collumatus occurs. The topmost bed of the Road Canyon, it contains a variety of brachiopods and other fossils, mostly badly battered but containing sufficient good material to make profitable collecting. The limestones of the Word Formation contain many layers that represent accumulations by currents. The China Tank Member has a number of levels containing such beds. The best and most typical locality is in the Willis Ranch Member at USNM 706e, where a foot-thick layer abounds in fossils. One block of 186 pounds (when decalcified) was estimated to have yielded more than ten thousand specimens. This did not include fragments and bits of bryozoans. The dimensions of the block were approximately 16 x 12 X 12 inches. The biological shell material constituted about 13 percent of the mass, and about 3 percent was sugary quartz sand. The variety of brachiopods in the layer totaled 84 species. Most of the specimens consisted of disarticulated valves, but only a small percentage, of complete shells. Abrasion had not affected most of them seriously. Most of the productids still retain all or part of their delicate spines, and some immature productids preserved the initial attachment rings and other early, very delicate spines. Although most of the larger specimens were dead when accumulated, many of the epifauna such as Heteralosia, immature Meekella, and Cooperina may have been living. Many of the concave and cavernous shells are occupied by spat or other small encrusting forms. Deposits similar to the above occur in the Appel Ranch Member at USNM 715i and 719x. Great numbers of fossils are packed into an incredibly small volume of rock. This was probably accomplished by constant movement of the shells by currents until each had accommodated itself to the least space it could occupy. Two other localities (USNM 706b and 737w) are lenses between the Willis Ranch and Appel Ranch Members, the former locality lying nearer the Willis Ranch and the latter, nearer the Appel Ranch Member. The lenses are about 1-1.5 feet thick and are closely packed with fossils. The rock also contains some sand and brownish silicious material. The faunas of these lenses are like that of USNM 706e. Organic Accumulations Outside the Glass Mountains SIERRA DIABLO.—Our studies in the Sierra Diablo were directed primarily to the collection of fossils for comparative purposes and did not include a study of the reef-like masses of the Victorio Peak Member of the Bone Spring Formation. The lower part of the Bone Spring, from which our collections are derived, are in part lag deposits, thanatocoenoses, and lime sand banks—all bioclastic. We did not recognize any bioherms in this part of the section. Stehli (1954) postulated a shallow water environment for his localities in the Sierra Diablo, with depths ranging from uncertain to 25 feet at one locality and to 100 feet at another. He comments on the enormous number of broken and separated brachiopod shells in proportion to other animal debris, which consists of broken coral clumps of Heritschia, worn fusulinids, occasional mollusc valves, and the usual abundance of crinoid debris. That the deposits are near-shore seems likely; if reef-marginal, as also suggested, the reef source is conjectural. DELAWARE BASIN.—Our collecting here was limited to a few localities of the Cherry Canyon and Bell Canyon Formations. The localities in the lowest part of the Getaway Member are in lenses of shell debris very much like those of the Word Formation in the Glass Mountains. These are clearly thanatocoenoses, in which most of the specimens consist of separated valves and the fauna is predominantly brachiopod. The most interesting and best known of the Getaway fossil accumulations is that at AMNH 512 (=USNM 728). This is a sponge-mollusc community like that of USNM 703c in the Glass Mountains, but it is richer in brachiopods than the latter occurrence. This appears not to be a bioherm but a disrupted community, which suggests the zotikepia described above. The Bell Canyon members collected in the Delaware Basin mostly contain brachiopod assemblages. Those of the Hegler Member contain numerous brachiopods, and in one locality (USNM 731), a variety of sponges. These localities contain numerous brachiopods that must have been living at, or near, where they were found. Many of the rhynchonellids, spiriferinids, and terebratulids have 90 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY both valves joined and show no signs of wear. Some of the less tightly hinged brachiopods, such as Spiriferella and the small productids Scapharina, Xenosteges, and Heteralosia, are disjointed. The Lamar Member in the Basin contains one bed (USNM 728p) that is very rich in fossil material (Plate 22: figure 3). Most abundant are the genera Astegosia, Anomaloria, and Martinia, the valves of which commonly are disassociated and often broken. Since these brachiopods were articulated only loosely in life, they attest to some agitation of the water or to transport of the shells, but it would not have required great abuse to have created the condition in which they are found. They are in contrast to the same genera on the "reef" front, where the majority or a large number of specimens have both valves joined. In the Lamar at USNM 728p, the terebratulids and rhynchonellids generally are conjunct, whereas the larger spiriferinids are more commonly disjoined. The evidence suggests moderately shallow water with some current or possibly wave action during storms. We can see no evidence that these shells were transported by gravity slides from the reef, because there is no lime sand or reef debris in the form of broken sponges or fragments of algae associated with them. GUADALUPE MOUNTAINS.—In this region as in the others of West Texas, our main objective was the collection of comparative material rather than studies in stratigraphy or ecology. Nevertheless, we did observe these conditions where we collected, and we have views based on the faunas derived from the acidizing program. We made no collections from the Bone Spring of the Guadalupe Mountains west front, but we have specimens from AMNH 658, which came from a "biohermal-like lens on the north wall" of the seventh canyon north of Indian Cave Canyon near Williams Ranch. This contained Institella, which usually preferred a "reefy" habitat. Most of the Bell Canyon collections were taken from Rader Ridge (Plate 23: figure 4), where some effect of reef slide is evident, but specimens taken from the dark limestones generally can be described in the same terms as those farther out in the basin. The rhynchonellids in the Pinery Member, for example, are generally complete specimens, with only a few disassociated valves. This is true especially of Bryorhynchus, which is a very delicate, thin shell and might be broken easily. Some of these are distorted but not broken, probably the result of compaction of the sediments. Slide from the reef into the Rader dark sediments consisted of lime sand with typical Rader fossils, but these were not silicified and were recovered by breaking the rock. The silicified Rader fossils presented the same features as the other members of the Bell Canyon Formation. They all suggest specimens that lived in moderately quiet water and later were entombed where they lived or near their original habitats. Nearly all of our fossils from the Capitan Formation were taken from the inclined beds along the mountain front. In these beds, brachiopods are concentrated in patches that may have been assemblages in situ or may have been transported a short distance. Many of the specimens have both valves intact, and this includes such loosely articulated forms as Martinia, Anomaloria, and Astegosia. Stenoscisma usually can be found with both valves together, as can many of the spirifers, spiriferinids such as Paraspiriferina, and the terebratulids. The small rhynchonellids such as Anteridocus sxuallovian.us (Shumard) usually are complete. The large productid Thamnosia capitanensis (Girty) usually is complete and often retains its tubular anterior. Where slightly rotted, the Capitan calcarenite of this portion of the "reef" offers some of the best collecting in West Texas. The occurrences suggest specimens essentially in place or with only a slight amount of transportation (Plate 22: figure 1). CHINATI MOUNTAINS.—Important bioherms occur in the Chinati Mountains especially in the Breccia Zone of Udden's Cibolo Formation. Although these occurrences have been interpreted as reef slide by Rigby (1958), we believe these structures to be true bioherms. The biohermal beds abound in Scacchinella and prove to be an exact duplicate of the bioherms of the Decie Ranch and Sullivan Peak members in structure. Furthermore, the faunal content is so similar that correlation with the Skinner Ranch Formation is obvious (Plate 22: figure 4; Plate 23: figures 1, 2). NUMBER 14 91 Fossils in the Various Facies It is generally believed that some fossils are closely restricted to certain types of sediment. This is partly true, but most fossils are distributed widely and usually occur in several kinds of sediment, e.g., Leiorhynchoidea, which is commonest in black or dark muds. In spite of such common occurrence, we have species from calcarenites, shell breccias, and yellow shales. The fossils mentioned below as the common inhabitants of certain kinds of sediment may be characteristic of that time and place, but they are not always restricted to that type of rock elsewhere. IN THE CONGLOMERATES.—In the Glass Mountains there does not seem to be any clear evidence of residual faunas in the conglomerates that are formed by transported materials. In the first place, the environment is not good to preserve fragile specimens, and the ones that are found in the conglomerates show evidence of transport by their fragmentary condition. The only fossils that were probably resident in a gravelly or bouldery sediment are those that occupied bioherms, and these are considered under that facies. The most significant transported fossils in the conglomerates are the ammonites. Mentioned previonsly, they are almost certainly specimens that floated onto sand and gravel banks. In addition to the ammonites and often associated with them, there are fragments of wood and fairly well preserved seeds. These are further testimony to the near-shore and shallow-water environment of the transported conglomerates (Plate 13: figure 2). We have not seen any indisputable specimens in the matrix of the boulder beds that surround the bioherms of the Sullivan Peak Member at the southwest end of the Lenox Hills. IN THE QUARTZ SANDS.—Quartz sands in the Glass Mountains are not an inviting source of fossils. Although all of the sands were examined for fossils, the search was not exhaustive and the few that were found undoubtedly were transports from other environments, e.g., the specimens seen on the upper surface of the thick sandstone that forms hill 4806. Inasmuch as the calcarenites or lime sands are composed largely of bioclastic material that includes crinoid debris and broken shells, these do not contain resident fossils, except for the bioherms that they may surround. We know of no fossils that are restricted to such an environment. The faunas are usually like those in neighboring bioherms. IN THE SHALES.—The bluish shales contain brachiopod types that are common in similar sediments in other parts of the column. The Uddenites- bearing Shale Member contains genera that occur in many of the Pennsylvanian shales, especially the Gaptank shale underlying the Wolfcamp: Rhipidomella, Parenteletes, Isogramma, Neochonetes, Chonetinella, Hystriculina, Kozlowskia, Kutorginella, Linoproductus, Reticulatia, Rhynchopora, Crurithyris, and Neospirifer. All of these except Parenteletes are common in the Gaptank shale and in the shales of the Pennsylvanian in the Midcontinent region. None of these is restricted to shale as its habitat, but they are common shale dwellers. The bioherms of the Uddenites- bearing Shale Member contain a different assemblage, mostly of Permian types such as Limbella and Scacchinella. We have seen very few megafossils in the radiolarian and spicule-bearing shales of the Cathedral Mountain and Word Formations. In the latter, Crurithyris and Leiorhynchoidea occur, as they do elsewhere, in comparable inhospitable environments. Paranorella, which occurs in black limestone and shale in Coahuila and the Delaware Basin, was taken in the blocky shale of the Poplar Tank Member. The occurrence is rare, however, and is the only specimen of this genus found in shale in the Glass Mountains. IN THE LIMESTONE.—The dark limestones are often replete with fossils, a fact that makes identification of them as basinal deposits impossible. As evidenced by the fauna from USNM 707e, 703c, 703d, a great variety of types occurs in the dark limestone. These faunas include a few of the characteristic forms usually associated with bioherms. They do not include the common black limestone types such as Bryorhynchus, Leiorhynchoidea, and Paranorella, although one species of Glossothyropsis is present but very rare. Perhaps Liosotella costata, new species, is the most characteristic species in the dark limestones at USNM 707e. Other forms are the very wide-hinged Reticulariina subu92 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY lata, new species, and an abundance of various chonetids. A similar association appears in the thin-bedded limestones of the upper part of the Cibolo Formation (thin-bedded zone of Udden), in which Liosotella costata, wide-hinged Reticulariina bufala, new species, and wide chonetids occur. This portion of the Cibolo Formation has been likened to the Bone Spring, but it should rather be compared to the Road Canyon Formation, of which it is probably an extension. The variety and abundance of fossils in these limestones is unlike the general sparsity of fossils characteristic of genuine basin deposits. We regard these as shelf deposits in shallow water, near a low, possibly swampy, land. The water was not necessarily quiet, because there is some heaping of shells and disjoining of the valves. As noted previously, the lagoonal limestones of the Hess Formation contain few fossils other than fusulinids. We found a few macrofossils, however: a large colony of Waagenophyllum, an occasional ammonite, an occasional large Omphalotrochus. A piece of float that could have been derived only from the Hess Formation contained a few brachiopod species, including a new Martinia. It was found loose, just under the Hess Conglomerate ( = Lenox Hills Formation of Ross), and could not have been derived from the Neal Ranch Formation. We are not able to say whether it came from the Wolfcampian portion of the Hess ( = Lenox Hills Formation of Ross) or from the Skinner Ranch part. The assemblage is more suggestive of the Skinner Ranch than of the Lenox Hills. IN THE BIOHERMS.—Most of the bioherms are a framework of bryozoans and bryozoan fragments, among which the brachiopods found asylum. Many of the Permian brachiopods are cementing forms that flourished naturally only where a firm bottom for attachment was available. They, like the brachiopods today, favor a fairly quiet water situation. The abundance of niches provides living for a variety of brachiopods and accounts for the great diversity in faunas of bioherms. The bryozoan bioherm, therefore, offers the utmost for their requirements. Only in a few examples and only with a few genera do the brachiopods contribute notably to the formation of a bioherm. In the Silurian and Devonian, brachiopods often make mounds of their own (Pentamerus and Stringocephalus), but they are not bioherms in the sense of being consolidated masses. They make mounds simply because they lived clustered together, the living shells attaching to the dead, until shifting seas obliterated them. A few brachiopod genera, often called reef types, are identified with bioherms, but it is an interesting fact that the greatest, supposedly Permian reef, the Capitan "reef," has very few of these reef-type brachiopods. The following genera may contribute notably to the formation of bioherms or may produce clusters of their own that possess biohermal significance: Scacchinella, Hercosia, Hercosestria, Teguliferina, Derbyia, and Coscinophora. To a lesser extent, the following are moderately important contributors to bioherms or zotikepia: Edriosteges, Geyerella, Tropidelasma, Institella, Collemataria, and Agelesia. Scacchinella: This and the next genus are the nearest to reef-type brachiopods in the Permian, as they are large and sturdy; because of these qualities, they probably were current resistant. They offered also a haven for other, smaller brachiopods. Scacchinella is found usually in biohermal assemblages, and it often makes up a large percentage of the bioherm. In spite of this, it may be found in small clusters away from the main biohermal masses. Although Scacchinella properly may be dubbed a "facies fossil," it shows a significant evolution in the development of the amount of vesicular tissue in the pedicle valve, which possesses great value as a time indicator. As described below, several levels of Scacchinella species have been detected. Hercosia: This genus consists of much smaller species, but one actual bioherm (USNM 726o) was found in which H. delicata, new species, makes up most of the framework of the bioherm. Hercosia uddeni at USNM 702 often makes small clusters, but generally the richthofeniids only are dwellers in the bioherms. Cyelacantharia and the other Glass Mountains richthofeniids are rarer in bioherms. The former appears in bioherms (USNM 702c, 703a) of the Road Canyon, but it is more commonly found in nonbiohermal assemblages in the Word Formation. Hercosestria: This genus seldom makes bioherms in which it is dominant; rather it is a participant in the making of the bioherm. NUMBER 14 93 Teguliferina: This is an important contributor to Wolfcamp bioherms (USNM 701d). Derbyia: Derbyia is usually a participant in the making of bioherms, but in some parts of the Neal Ranch Formation (USNM 701d, 742c) it is a dominant element, at least in parts of the bioherm. Coscinophora: This genus may be the dominant element in the formation of some bioherms. It is a more common biohermal form than Collemataria, which usually is only a contributor. The flattish surfaces of Coscinophora and the curved inner surfaces of the interior of dead shells are favorite sites for small cementing forms (see Plate 129 for a reefy cluster of Coscinophora). Bioherms at Different Levels In the Permian column of the Glass Mountains, bioherms are predominantly bryozoan, but each has brachiopod elements characteristic of its level. A brief discussion of the bioherms at different levels follows and includes some of the accessory genera commonly found in bioherms. Uddenites-BKAKING SHALE MEMBER.—Inasmuch as these bioherms have not been found in silicified condition, we do not know a great deal about them. They contain Scacchinella as a constituent, but this genus is not abundant and it was nowhere found to dominate a bioherm at this level. An accessory element in these bioherms is Parenteletes cooperi R. E. King, which often is fairly common. NEAL RANCH FORMATION.—Three levels of bioherms have been seen in this formation. The first is in the upper part of the Gray Limestone of P. B. King (USNM 701). This biohermal or possibly zotikepial structure was characterized especially by orthotetaceans of rare and unusual character, in addition to the usual bryozoa. Another small bioherm at USNM 722x is characterized by common Eolyttonia cemented by laminar calcite, which suggests an algal origin. The relationships of this bioherm to its surroundings suggest that it was at one time a much larger structure. Bed 4 of P. B. King produced bioherms and zotikepia that are interesting for the abundance of Derbyia in the bioherms and the presence of Striatifera in the zotikepia. USNM 70Id is characterized by great abundance of small derbyias cemented into clusters with masses of Teguliferina. These are the only bioherms seen in which Derbyia and Teguliferina formed a substantial part of the mass. Parenteletes formed a prominent accessory element reminiscent of the bioherms of the Uddenites-bearing Shale Member. The zotikepium at USNM 721g is rich in several types of bryozoans; it has numerous Striatifera attached to the colonies and a species of Pseudoleptodus attached to the Striatifera. A variety of bioherms was seen in beds 12—14 of P. B. King ( = beds 9-12 of Cooper) all faunally different. The bioherm at USNM 701c contains, as important elements, Eolyttonia, Limbella, Teguliferina, but none seem dominant. So abundant is Geyerella, the bioherm at USNM 70Ih may be characterized as of this genus, which occurs cemented together in clusters and makes up a large percentage of the mass. Other reefy elements are Tropidelasma, Meekella, and Limbella. Important accessories are Parenteletes, Composita, Diplanus, Spyridiophora, and a host of rarer forms. The bioherm at USNM 701k has the same elements as the preceding, but it is specially noteworthy for a flood of Enteletes as an accessory. This genus is a common accessory in most of the bioherms and zotikepia. LENOX HILLS FORMATION.—The most important and interesting bioherm in the Lenox Hills Formation is the one on Leonard Mountain just above the conglomerate (USNM 705k). Most of the frame of this bioherm is made of a small new species of Scacchinella. This is assisted in frame building by abundant specimens of Tropidelasma and the accessory element Parenteletes. Several Lenox Hills bioherms occur at USNM 715b in the southwest end of the Lenox Hills. These have no silicified material, and our collections are not extensive. No special element stands out to characterize them, but Spyridiophora usually can be obtained. DECIE RANCH MEMBER.—The bioherms in the western and eastern parts of the mountains at this level are different in composition, but they are believed, nevertheless, to be contemporaneous. In the Lenox Hills the Decie Ranch bioherms are definitely Scacchinella bioherms, that genus being the dominant element in size and numbers. With it, and intimately attached, there are a huge species of Geyerella, a stout Tropidelasma, a huge Der94 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY byia, and an enormous Eolyttonia. Accessories in this assemblage are Oncosarina, Limbella, Enteletes, Antronaria, and Stenoscisma. The bioherms at the same level in the more eastern part of the Glass Mountains on Leonard Mountain and the north side of the Hess Ranch Horst are dominated by Scacchinella, but the individuals, although large, seldom reach the enormous size of those seen in the Lenox Hills. Geyerella is missing or rare, and its place is taken by Meekella, which attains its largest size here (USNM 720e). An elongate Cyclacantharia occurs along with a large conical Eolyttonia. Accessory genera are Enteletes, large Stenoscisma, and Antronaria. Bioherms in the Cibolo Formation (Breccia bed of Udden) are like those of the Decie Ranch Member. SULLIVAN PEAK MEMBER.—Most of these bioherms are not, or only partially, silicified. Their composition is like that of the Decie Ranch Member bioherms, but Scacchinella is no longer the dominant element. The one whose fauna is best known, USNM 733j, is composed partly of Coscinophora (Plate 5: figure 1), with Scacchinella rare, but Limbella fairly common. Another "reef-type" is Tropidelasma, but it is not abundant. Spyridiophora is present. Scacchinella, although widespread in the Sullivan Peak Member, is never so abundant as in the Decie Ranch Member and in the lower part of the undivided Skinner Ranch Formation. TAYLOR RANCH MEMBER.—Only one small patch of bioherms is known in this member, and it is conspicuous for its sponges, Girtyocoelia and Heliospongia. Brachiopods constitute only a minor element in the assemblage. Common bioherm-loving forms are present: a new notothyridid called Chondroma, huge Enteletes, Spyridiophora, Antronaria (common), and large Stenoscisma (Plate 19: figure 2). CATHEDRAL MOUNTAIN FORMATION.—Bioherms and zotikepia, common in the Wedin Member of this formation, generally are dominated by Institella and Agelesia. These are the commonest brachiopods, but brachiopods are not important in making these bioherms and zotikepia. Here the honors fall to the bryozoans. Contributing brachiopods in the Wedin Member are: Hercosia, Chonosteges, Compressoproductus, and Enteletes, which is common. In the eastern part of the Glass Mountains in the vicinity of Split Tank, bioherms similar to those of the Wedin Member contain Institella, although the associated forms are different from those of the Lenox Hills and Dugout Mountain areas. Agelesia is somewhat rarer, but one bioherm (USNM 703bs) is very rich in this genus. Hercosia is abundant and forms clusters and small bioherms containing the accessory Enteletes. Another accessory is Niviconia globosa (R. E. King). Rugatia is found suspended by long spines among the branching bryozoa. Other accessories are: Edriosteges, Chonosteges, and Collemataria. If any genus is dominant, it is usually Hercosia. Bioherms are not common in the Upper Cathedral Mountain, and the best was mentioned under the note on Hercosia. The frame-builder of the bioherm at USNM 726o is Hercosia; Edriosteges, Chonosteges, and Echinauris are accessories. Common in this assemblage is the large bivalve Pseudomonotis, spiny and ostreiform. A unique bioherm largely composed of the lyttoniid Collemataria occurs at USNM 702al. ROAD CANYON FORMATION.—Bioherms of the lower part of the Road Canyon usually have a different assemblage from those of the upper part of the formation. The lowest part of the Road Canyon forms a biohermal zone that can be followed from a point near the Old Word Ranch site over to Sullivan Peak. Localities of these bioherms are USNM 703a, 702c, 719x. Hercosestria, a common member of these bioherms, does not make up the main mass, but it usually is abundant. Another abundant element is Edriosteges. With these are large Meekella, Coscinophora, Thamnosia, large Cyclacantharia, large Stenoscisma, and large Dielasma. The bioherms of the upper part of the Road Canyon Formation may be described as Coscinophora bioherms; this abundant genus is usually the dominant element. An interesting accessory is the tiny notothyridid Chondroma, which has been found only in bioherms or in association with them. In summarizing the fossils of the bioherms, it may be said that only a few types are mound builders, although many species occur as accessories. Some, such as Chondroma, appear always to be associated in or near bioherms. It is probably fair to say that most of the fossils in Wolfcampian and NUMBER 14 95 Leonardian rocks, except those of the shales, more than likely lived on or around bioherms. Some of these, such as the assemblage in USNM 72lu, probably were swept together from biohermal sites. In the shell heaps and bottom sweepings of the Word Formation, richthofeniids and aulostegaceans, which normally occurred in bioherms in the earlier sediments, probably had somewhat different habitats. There more than likely were not big clusters, to initiate bioherms, and there were no bryozoan zotikepia from which bioherms could be developed. The absence of bioherms along with the presence of genera that normally inhabited bioherms suggests that the Guadalupian conditions were different from those preceding. It is possible that the water, although shallow, was deeper than that of the biohermal locations and that the animals lived farther from shore in a less oxygenated and nutritious environment. The above discussion also is valid for the Sierra Diablo region and the Delaware Basin. In each of these regions, biohermal and accessory forms dwelt, but they did not form bioherms. SIERRA DIABLO AND GUADALUPE MOUNTAINS.— Our collections from this region come only from the lower part of the Bone Spring Formation, which faunally is related to the Skinner Ranch Formation of the Glass Mountains, that part of the section containing most of the bioherms. Although bioherms have been illustrated and studied in the lower Bone Spring (Stehli, 1954), we did not find any silicified ones comparable to those of the Glass Mountains; rather, our collections and those of Stehli are from the sweepings from the bioherms and possible zotikepia. The Sierra Diablo are clearly "shell heap" occurrences or thanatocoenoses. The fauna includes Scacchinella, Limbella, Acritosia, Teguliferina (Planispina of Stehli), and Meekella, all genera that normally are generous contributors to bioherms. Scacchinella americana Stehli, however, is very rare and small, and the species is unlikely to have been an important contributor to bioherms. Teguliferina and Acritosia are abundant and may have been framebuilders of small bioherms. They are the only likely members of this fauna to have been colonial. Meekella is abundant and may have contributed to the Teguliferina clusters in much the same fashion as Geyerella did in the Neal Ranch bioherms. Eolyttonia and Collemataria are too rare to have been important contributors. Enteletes is the commonest accessory genus, although Stenoscisma is present, but rare. Much of this fauna consists of chonetids, small and large productids, rhynchonellids and spiriferinids, all of them brachiopods that probably occupied niches either in bioherms or zotikepia or even in shell heaps. The evidence suggests to us that there may have been bioherms at this time but that they probably were small and easily broken up. We have not studied the Victorio Peak Member, but what we know of its faunal content indicates that it contains bioherms that suggest those of the Cathedral Mountain Formation (Plate 8: figure 3). The great "reef" surrounding the Delaware Basin has few of the "reefy" brachiopods. Cyclacantharia and Sestropoma are present, but they are extremely rare. As a matter of fact, the "reef" contains none of the possible frame-building brachiopods. All of these except Geyerella became extinct before Capitan time. Collemataria is also a very rare animal in the reef rock. That brachiopods grew on the reef, in its niches and on its slopes toward the basin, is shown by the patches in the steeply dipping beds on the front of the mass. The brachiopods consist partly of niche dwellers taking advantage of the crannies in sponge and algal beds. A large number of the Capitan brachiopods evidently formed their own patches in which dead members of the assemblage furnished the anchorage for the next generation. Many of the loosely hinged genera such as Astegosia, Anomaloria, Reticulariina, Arionthia, new genus, Martinia, and Composita are found commonly with both valves intact and only a small number of single valves. This indicates fairly quiet water conditions on the "reef" slope. One can only conclude that the brachiopods were an insignificant element in the formation of the Capitan "reef" but that they found its surface and niches a favorable dwelling place. Bell Canyon sediments marginal to the "reef" mass contain large numbers of brachiopods that occupy patches representing small heaps. Most of the specimens are small, and productids are extremely rare. The most abundant species are those of the Spiriferinacea, small Strophalosiacea, cho96 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY netids, small rhynchonellids, and small terebratulids. In addition, stout, sticklike bryozoans and cup corals are abundant. Many specimens are found with both valves intact. Breakage is minimal, and the assemblage gives the impression of having dwelt where it was found. Although some of the species also occur in the Capitan Formation, the sediment contains considerable quartz sand, but no debris demonstrably reef derived was seen. It seems that these deposits of the Hegler, Pinery, and Rader Members represent sediments in front of the reef with resident faunas. Demonstrable reef slide appears in the Rader, but this is calcarenitic rock containing unsilicified brachiopods like those of the dark Bell Canyon limestone (Newell et al., 1953; Rigby, 1958). The story of the fauna of the Lamar Member of the Bell Canyon Formation is much the same. The fauna has an entirely different composition from that of the lower members of the Bell Canyon, but essentially it is the same as that of the faunas mentioned as occurring on the slope of the "reef," abounding in Composita, Anomaloria, Astegosia, and Martinia. Rather than being derived from the reef slope and being transported to the base of the reef, as commonly believed, the Lamar fauna, our studies lead us to believe, lived at the base of the reef, but current agitation caused considerable disassociation of the valves. In the Lamar, complete specimens of any of the genera mentioned above are very rare. This is not so of the rhynchonellids, terebratulids, and spiriferinids, which often are found with valves intact. The fauna is composed almost exclusively of brachiopods, corals, and bryozoa, but molluscs are extremely rare. Grant (1971) traced the decline of reef-dwelling and reef-associated brachiopods through the West Texas Permian section, pointing out that the supposedly greatest of all Permian reefs, the Capitan, contains an overwhelming proportion of the kinds of brachiopods that lived attached by a pedicle. The spiny brachiopods that give Permian faunas their distinctive aspect are scarce in the Capitan Limestone and its equivalent members in the Bell Canyon Formation. The decline in spiny brachiopods and the proportionate increase in pediculate forms began as low as the Skinner Ranch Formation. By the time the Capitan was deposited, the fauna had what might be considered a "Mesozoic aspect," in the sense that most Mesozoic brachiopods are pediculate. Paleoecology To understand the conditions that existed in the Permian of the Glass Mountains, it is necessary to examine the paleogeographic setting of the area. In addition, the rocks that make up the Permian of the Glass Mountains have their own story to tell of the times. Added to these two versions is the story told by the fossils. Determination of past conditions deduced from rocks may lead to ideas contrary to those indicated by the fossils. Interpretation of lithic data will not coincide necessarily with that determined by fossils. Deductions as to depth of water may be at variance because of conflicting evidence from sediments and fossils. PALEOGEOGRAPHIC SETTING.—Philip B. King (1942) brought together basic information on the paleogeographic setting of the Glass Mountains, the Sierra Diablo, and the Guadalupe Mountains. He depicts, in his now well-known and much-figured map (King, 1942: figure 18), a Marathon Folded Belt lying south of a Southern Shelf area on which the sediments of the Glass Mountains were laid. This shelf faces the Marfa Basin area to the southwest and west and the Delaware Basin to the north. The Delaware Basin is bounded on the west by the Diablo Platform and on the east by the Central Basin Platform, essentially submerged peninsulas extending south from the Northwestern Shelf Area of southern New Mexico. The Central Basin Platform forms the western margin of the Midland Basin, which is bounded on the east by the Eastern Shelf Area. The Marfa Basin is connected to the Delaware Basin by the Hovey Channel and the latter basin connects with the Midland Basin by the Sheffield Channel. The Glass Mountains occupy a very small amount of this vast area, and its deposits mainly are shelf deposits, mostly clastic limestones, conglomerates, and sandstone, with some important layers of shale. The source of the sediments was partly from the Marathon Folded Belt and partly from the west, as most of the section thickens in that direction. It seems likely that the Southern Shelf extended farther southwest than indicated by King, because NUMBER 14 97 NO RTHWESTERN SHELF CHINATI MTS 25 50 75 100 i MILES FIGURE 27.—Paleogeographic setting of the Permian of West Texas. part, at least, of the Chinati Mountain (Cibolo Formation-Breccia beds of Udden) sequence are shelf deposits identical to parts of those of the contemporary Skinner Ranch Formation (Rigby, 1958: diagram, page 299). SEDIMENTS OF THE WOLFCAMP SERIES.—In the Wolfcampian we include the Uddenites-bearing Shale Member of the Gaptank Formation of P. B. King, which is identical to the Neal Ranch Formation in gross lithic aspect. We also include the Skinner Ranch Formation as explained elsewhere. Like the Neal Ranch, the Uddenites-bearing Shale Member contains bioherms based on limestone gravel and surrounded by detrital limestone that is enclosed in blackish to bluish shale containing scattered fossils. The bioherms of the Neal Ranch mainly are bryozoan-sponge assemblages, with algae suggesting existence in warm, sunlit, shallow water. Temporary submergence of the shelf, uplift of the Marathon Folded Belt, or a wet climatic cycle may account for the ultimate burial of the bioherms in the shale. Except for the local development of limestone cobbles associated with bioherms, conglomerates are rare in the Neal Ranch sedimentary suite. Quite the opposite is the condition in the Lenox Hills Formation. Lenox Hills: In the western part of the Glass Mountains this formation contains a thick conglomerate that thins eastward to disappear on the flank of the Hovey anticline just west of Leonard Mountain. East of this fold, Lenox Hills conglomerate appears in Leonard Mountain and along the base of the mountains except for one place east of the Wolf Camp Hills. In the western part of the Glass Mountains, a shale separates the conglomerate from the overlying Skinner Ranch Forma98 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY tion. In Leonard Mountain, bioherms appear in the limestone overlying the Lenox Hills conglomerate. Bioherms also occur above the conglomerate just north of the Hess Ranch. East of this point the conglomerate is overlain by a variety of rocks consisting of red beds and thin dolomitic limestone like that of the Hess Formation. The Lenox Hills conglomerates point to an uplift of the Marathon Folded Belt that blotted out the early development of bioherms at its base. The shale succeeding the conglomerate in the Lenox Hills is a facies like that of the Neal Ranch, with a comparable fauna consisting of ammonites, gastropods, and pelecypods similar to those of the Uddenites- bearing Shale Member and the Neal Ranch Formation. The limestone above the conglomerate in Leonard Mountain is a coarse lime sand (calcarenite and calcirudite) fingering into the Hess calcarenite to the east. The red beds represent a local uplift that washed lateritic muds into the Southern Shelf Area. It is interesting to note that thin beds of conglomerate in the Lenox Hills area contain numerous seeds and fragments of wood together with concentrations of ammonites. These two occurrences argue for shallow-water, near-shore deposition. The ammonites probably are dead shells that floated into shallow water and were deposited with piles of pebbles and fragments of waterlogged plant debris. On the southwest side of Leonard Mountain, a fairly large tree trunk was seen in the Lenox Hills shale. Skinner Ranch: These sediments are very diverse, consisting of limestone conglomerate, calcirudite, calcarenite, and some fairly thick, but local, quartz sandstone beds. The conglomerates of the Decie Ranch and Sullivan Peak Members are notable for their coarseness. Ragged boulders up to five feet in one dimension occur and many over a foot long are present (Plate 11: figures 1, 4). In the Decie Ranch Member, the bioherms are not banks of reefy limestone, but rather they are tangled masses of bryozoans and large brachiopods, formed on a gravel bed or comprising an accumulation of large crinoid stems. This association suggests a near-shore environment, with high energy conditions. The uplifted Marathon Folded Belt supplied some of the conglomerate, but fragmentation of the bioherms is also evident. The Poplar Tank Member is of varied lithic character, which consists mainly of brown crumbly shale enclosing thin beds of limestone and conglomeratic limestone. In places, the calcareous layers contain a thin skin of pinkish chert containing numerous fossils and sponge debris. Sponges probably furnished much of the silica. The cherts contain members of the Spyridiophora fauna, demonstrably a shallow-water assemblage. The cherts here are not evidence of deep water, but rather of a concentration of spicular material. Large moundlike bioherms occur in the Sullivan Peak Member. The most conspicuous example is in hill 4801 (Plate 18: figures 1, 3). The boulders occupying the space between the two adjacent bioherms are exceptionally large and almost entirely biohermal in their origin, as they contain the common Permian fossils of the bioherms. On the west side of the easternmost bioherm there is a boulder containing numerous specimens of Scacchinella. The setting is one of close shore, with violent water conditions that pounded the bioherms to fragment them and pile the debris at their bases. The setting on hill 4801 of the Lenox Hills is like that of Udden's Breccia Bed in the lower part of the Cibolo Formation, on the northwest slope of the Sierra Alta Creek in the Chinati Mountains. For about 3 miles along the creek, }arge Scacchinella bioherms are based on gravels containing reworked Wolfcampian fusulinids. The bioherms are surrounded by coarse breccia, like that of the Glass Mountains. If these large masses were reef slide, as claimed by Rigby (1958), one would expect them to be at different levels in the conglomerate, helter skelter. Such is not the case; they occupy a general level and are not covered by the boulder conglomerate. We believe that these bioherms occupied an extension of the Southern Shelf Area in Skinner Ranch time and that conditions were like those postulated for the Sullivan Peak Member in the Glass Mountains, i.e., shallow water, near shore, with strongly agitated water. The Skinner Ranch Formation east of Leonard Mountain is replete with small Scacchinella bioherms surrounded by bioclastic aprons, clearly a shallow-water environment. On the east side of the Hess Ranch and on the north slope of the Hess Ranch Horst, the coarse calcarenites and calcirudites of the Skinner Ranch NUMBER 14 99 interfinger with the fine lime sands of the Hess Formation. The latter often abounds in lenses with fusulinids, the most abundant fossils. Occasional large fossils such as coral colonies, Omphalotrochus, and ammonites occur. Generally, however, the rocks are fine calcarenite, pellet limestone, and oolite. In places fusulinids are encased in laminated lime (algae). The ensemble suggests deposition in a quiet bay behind the great limesand bank or bar of the undivided Skinner Ranch Formation. This great limesand bank, 400 feet thick, formed a barrier between the western coarse elastics and the finer limesands of the Hess Formation. The Taylor Ranch Member of the Hess Formation consists of a limestone conglomerate, followed by cobbly fossiliferous limestone that is capped by thick-bedded cherty limestone. Occasional bioherms have a high percentage of the sponges Heliospongia and Girtyocoelia (USNM 702d; Plate 19; figure 2). Large productids (Peniculauris) are abundant as well as large snails (Omphalotrochus). Some crinoidal debris and numerous small brachiopods also are present. The member extends for a few miles near the top of the scarp, as far as the Conoly Brooks Ranch. The presence of the bioherm, the great abundance of rolled shells, and crinoid debris suggest a shallow-water accumulation. The Taylor Ranch essentially is a tongue of the Sullivan Peak Member extending beyond the main barrier of limesand. SEDIMENTS OF THE LEONARD SERIES.—The sedimentation in the Leonardian is different from that of the Skinner Ranch Formation. The Cathedral Mountain Formation is a great mass of shaly-silty rocks with interbedded limestone and occasional conglomerate. The mass thins eastward and interfingers with a thinned, predominately limestone sequence. The Cathedral Mountain is capped by the lithically complex Road Canyon Formation. Cathedral Mountain: In the western part of the Glass Mountains the Cathedral Mountain Formation is divisible into two parts representing different conditions of sedimentation. The lower portion consists of a basal limestone member, the Wedin Member, largely biohermal and zotikepial (see below) in its makeup, overlapped by yellowweathering silicious shale, spiculite, and radiolarite. This material suggests more basinal deposition, as it has few macrofossils. These quieter water conditions are interrupted by deposition of clastic, conglomeratic limestone beds, the Third and Fourth Limestones of the Leonard of P. B. King, containing many ammonites (Perrinites) and many fragmentary and rolled fossils. These were deposited under shoal water conditions, and an occasional bioherm in them helps to bear this out. These conglomeratic limestones were followed by deposition of bluish shale and sandstone. The former contains large Peniculauris and large Perrinites in abundance. The sandstones extend from a point near Sullivan Peak westward to the foothills of the Del Norte Mountains, near the site of the Old Payne Ranch. Late in Cathedral Mountain time, local uplift shed conglomerates in the Del Norte Mountains (hill 4816), which suggest a shore zone with moderately coarse gravels that are composed of well-rounded pebbles, abundant transported ammonites, fragmentary brachiopods, and numerous plant stem fragments, producing the same type of environment as that in the Lenox Hills Formation. These sands and conglomerates are the end of extensive conglomerate deposition in the Glass Mountains. The Road Canyon Formation has few conglomerates, and these generally are associated with bioherms. The overlying Word Formation, in our experience, has little or no conglomerate. Thus, in late Cathedral Mountain time the Marathon Folded Belt and other sediment sources were lowlying and furnished only fine grained material. Cathedral Mountain in the Split Tank Area: One-half mile east of Split Tank the Cathedral Mountain Formation is only 300 feet thick, its base is a conglomerate of small quartz pebbles, and its lowest beds are of Hess-type lithology, with scattered bioherms. The lower third at Split Tank contains numerous bioherms surrounded by bioclastic debris. Bain (1967:233), as a result of his study of this area, maintains: "Water was probably no deeper than 55 to 60 feet over reef crests (the reefs 'rose fifteen feet above the sea floor'). Salinity was normal and turbulence moderate." The higher Cathedral Mountain Formation in this area consists of thin limestone beds in irregularly bedded, limy, yellowish shale abounding in Peniculauris. The occurrence is reminiscent of the Taylor Ranch Member. 100 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Road Canyon: This formation presents considerable lateral variation in sediment type. From the site of the Old Word Ranch to Sullivan Peak, the lowest beds of the Road Canyon are composed of biohermal masses (Plate 17: figure 2) that have well-defined outlines, rounded tops, and a thickness of 20 feet in places. The bioherms chiefly are bryozoan- sponge assemblages, but with a great variety of brachiopods occupying niches in tangled masses. This is a shallow-water environment favorable to bioherm development. A depression of the shelf submerged these bioherms and brought a cover of black limy mud (Plate 13: figure 3) over them between Appel Ranch and Sullivan Peak, except for the hills just north of Leonard Mountain. The black muds are not true basinal deposits because they contain numerous fossils and are lacking in Leiorhynchoidea and ammonites, which are common basinal fossils (as in Bone Spring Formation and South Wells Member of the Cherry Canyon Formation). In the hills north of Leonard Mountain, the Road Canyon Formation above the basal bioherms consists largely of bioclastic limestone with very little shale or black limestone. In the bioclastic beds, scattered bioherms of Coscinophora are present (Plate 17: figures 3, 4). Some of the bioclastic beds have high percentages of fossils. In the western area of Road Canyon, fusulinids may be present in such abundance as to constitute a sand or fine gravel. This area suggests a shoal on the shelf where agitated waters winnowed out fine debris to leave fossils of all sorts as a sand or gravel (USNM 732j). Some of the fossils probably were washed into the shoal from black mud areas to the east and west, because many of the species are those common in the black muds. West of Leonard Mountain on the spur south of Sullivan Peak, the Road Canyon Formation is exceptionally thick and is composed largely of dark to black limestone that, at several levels, is replete with fossils. The dark limestones fill cavities between the bioherms at the base (Plate 17: figure 2). At the top of the Road Canyon at this place, beds of fusulinids form a limesand. Similar beds can be seen in the Del Norte Mountains. Beds of the Road Canyon Formation in the Del Norte Mountains are thinned greatly, compared to the section at Sullivan Peak, and they fray out into the yellow shale. The sediments are mainly bioclastic limestone abounding in fusulinids. The bioclastic beds are separated by thin layers of yellowish shale like that of the Cathedral Mountain Formation. Occasional bioherms occur in the area. USNM 700v and 732j represent the topmost bed of the Road Canyon. The bed is a foot or two thick and is composed largely of fusulinids, but it contains a variety of fossils, mostly brachiopods, showing evidence of transport. Chief among the fossils is Collumatus pxatus Cooper and Grant, a richthofeniid that lived tightly fixed to the substrate. All but two of the specimens of this genus obtained by us have been broken free of their moorings. Few of them retain the coscinidium or the brachial valve. Freeing such a tightly cemented form indicates an environment of turbulent water that removed the fine debris and left only the heavier fossils. Just beneath this terminal layer, there is a bed of fine-grained calcareous shale abounding in ammonites. This is an excellent stratigraphic marker in the Del Norte Mountains. Ammonites are so abundant as to suggest that they are a concentration of dead shells in a local, quiet shoal area. SEDIMENTS OF THE GUADALUPE SERIES.—Sedimentation in the Guadalupian of the Glass Mountains is much less complicated than that of the preceding stages. In the Word Formation, that part of the Guadalupian that concerns us, there is a thick mass of sandstone and yellowish sandy shale in the western part of the Glass Mountains and the Del Norte Mountains. These contain tongues of limestone protruding into the mass from an eastern limestone unit. The limestone tongues, or members, are in ascending order: China Tank, Willis Ranch, and Appel Ranch Members. Small but important lenses appear between the last two members. The Willis Ranch Member is the most extensively developed unit and extends from the Old Word Ranch site to the west side of Gilliland Canyon. West of Gilliland Canyon its level is represented by lenses. The Appel Ranch Member extends west to the Old Willis Ranch site, but the China Tank Member ends against the Hess Ranch Horst. Each of these members is composed of calcarenite, containing sugary quartz sand and abundant fossils. These occur concentrated in layers in which incredible numbers of specimens are packed into a small space. The fossils are not graded in size, NUMBER 14 101 and many show evidence of transportation. The smaller, tightly articulated rhynchonellids and terebratulids usually retain both valves together, but the productids and other less strongly articulated forms occur as separated valves. Some genera are represented by several species in a single bed. The epifauna usually is well preserved, and some may have been living when buried. In a few places ammonites are common, although the Word sediments are devoid of conglomerates. The one band of conglomerate attributed to the Word on the flank of the Del Norte Mountains proved to contain abundant Perrinites and other fossils of the upper Cathedral Mountain Formation. Fusulinids of large size are common and frequently are current oriented. The Willis Ranch Member may be distinguished from the others by the large amount of sand contained in it. In Gilliland Canyon (Plate 14: figure 3), stringers and layers of quartz sand are common. This is near the place where the member loses its continuity. At USNM 706e, 4.1 miles northeast of the Hess Ranch, the limestone also is very sandy. A block weighing 186 pounds yielded 5 pounds of sugary sand. In places fusulinids are abundant enough to constitute a sand. This is especially true at the level of the Willis Ranch Member at USNM 732s, at the base of the Del Norte Mountains. The depth of water and other conditions of these members are thought to be the same. After Road Canyon time, water over the shelf deepened and sandy limestone formed on a bottom abounding in shelled animals. Either the water depth was too great or the area was too far off shore to permit biohermal development. Agitation of the water by storms or currents was sufficient to pile the shells, living and dead, into heaps and lenses. Although reef types such as Collemataria and Cyclacantharia are present, these never formed bioherms but existed in small clusters easily knocked down, broken up, and transported by storm waves and currents. Briefly, the story of conditions in the Glass Mountains may be depicted as follows: A narrow shelf existed along the margin of the Marathon folded belt, which produced in Wolfcamp through Leonard times an epineritic environment (Krumbein, et al., 1949:1869-1872). Bioherms and zotikepia of colonial forms developed in clear, sunlit water at a modest depth from tide zone to about 120 feet. Oscillations of water level, which altered the environment to infraneritic, 120-600 feet, resulted in burial of the bioherms in limy muds or shale (Wolfcampian and Leonardian) and black muds (Road Canyon). After deposition of the Road Canyon, conditions changed: no conglomerates formed and bioherms could not develop, although the bottom had a prolific fauna, often heaped into piles or rows by waves and currents. Epifauna Epifauna, as used here and by others, refers to the creatures that attach themselves to brachiopods and other shells. The Strophomenida of the Glass Mountains Permian are among the most numerous animals to attach themselves to the exoskeletons of living or dead contemporaries. The young of the Orthotetacea and Derbyiacea commonly attached themselves until they were big enough to break away, then lived loose on the sea bottom. The strophalosiids, by cementing the beak and strengthening their hold with struts in the form of spines, may occupy a host for a short time or throughout life. The Productacea may or may not cement the beak, but they commonly attached to crinoid stems or other cylindrical object by means of a spiny ring (Muir-Wood and Cooper, 1960; Grant, 1963, 1968). The oldhaminids also cemented to a host, dead or alive, and may have engulfed the host completely. The richthofeniids attached to any hard object and then, in reproducing, often showered themselves with their own larvae to the extent that the older and larger shells soon were engulfed. The small aulostegacean Cooperina and the strophalosiaceans Heteralosia and Ctenalosia were attached throughout their lives. Some of these occupied living shells, but a favored location for Cooperina was inside the dead valves of Cyclacantharia and Echinosteges (Plates 210 and 212). Such locations assured quiet water and seclusion from larger enemies. Xenosteges, in the Cathedral Mountain Formation, also used the interior of the pedicle valve of Hercosia as a favorite dwelling spot. Small specimens of Chonosteges also occupied this niche, but often it did not have room to come to full adulthood. The shape of the host shell often determined the ultimate form of the squatter. The pedicle valve of the lyttoniid Collemataria usually has a posteriorly 102 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY directed callus or attachment, the "flap" of Williams (1965). But if the spat settled on the concave outer surface of a productid brachial valve, the flap would have been forced to grow anteriorly rather than posteriorly, with the result that the adult Collemataria would have the irregular conical form of Eolyttonia. Similar growth aberrations occur in Hercosestria. When the spat settles on a flat or concave shell without being confined, the individual will have a broadly conical form with a large apical angle. Although brachiopods constitute a large amount of the epifauna that attached to brachiopod shells —and by weight or rapid growth may be lethal to a live host—other animals found the abundant brachiopod shells of the Permian to be convenient habitations. Foraminifera forming long slender tubes occur frequently, especially on Composita shells. Sponges are common epifauna, especially in their early stages. Living and dead shells of Coscinophora, in the Road Canyon Formation, and of Collemataria, in the lower Cathedral Mountain Formation, contain the small ball-like beginnings of Girtyocoelia and small colonies of Fissispongia. The spiral tubes of the worm Spirorbis occur through the sequence, but they are not common. Cup corals are common epifauna, especially in the Coscinophora bioherms, such as that in the Skinner Ranch Formation at USNM 733j and in the Road Canyon Formation at USNM 709c and 710u. Cup corals attached to Bryorhynchus (Plate 547: figures 38—40) occur frequently in the Bell Canyon Formation (Rader Member) at USNM 725f. The corals generally attached to the commissural edge of either valve on the side or front, suggesting that the brachiopod had an oblique living position. In one instance, the coral is more than twice the size of the brachiopod and engulfs part of the commissure, permanently sealing it. A variety of bryozoans attach to brachiopod shells. These may be funnel-shaped fenestellids, branching forms, or more massive types. In a number of examples the bryozoans have grown over the margin in such a way as to seal it. These may have been dead shells to begin with, but it is also possible that the weight of the bryozoan colony may have suffocated the brachiopod or that an already attached bryozoan just continued to grow after the death of the brachiopod. In the Road Canyon Formation (USNM 702c) plates of a barnacle like Turrilepas are fairly common; this animal probably used some of the Brachiopoda as host. Accidents of Settling of Larvae Although it has been demonstrated that the larvae of some marine animals have the ability to select a settling spot favorable to them, evidence from the Glass Mountains suggests that some of the spots selected may have been lethal. The favored habitat of the brachiopod is in still water, in grottoes, or in niches that afford protection as well as quiet. In two examples cited below, the quest for a favorable spot led the brachiopod to its doom. Specimen 150820 (USNM) is a large lyttoniid, Collemataria, with the richthofeniid Cyclacantharia attached to its ventral surface. The Cyclacantharia is anchored by abundant rhizoid spines that make a veritable thicket around the cup. Nestled at the base of the cup and completely surrounded by spines on three sides, there is a small specimen of Meekella skenoides attached by its ventral surface (Plate 116: figure 8). The specimen grew large enough to butt against one of the lower spines. This and the proximity of the anterior margin of the dorsal valve to the base of the richthofeniid make it impossible for the brachial valve of the Meekella to have opened. Starvation or suffocation must have been the result. The other specimen concerns a trapped Composita. A larva of Composita went through the mesh of the coscinidium of Hercosestria and attached to the brachial valve. Living was probably high for awhile, because the Composita had the advantage of the feeding streams of Hercosestria and must have grown fairly rapidly. But, alas, it became too big for its confined quarters and must have prevented both brachiopods from feeding normally. The death of the host as well as that of the intruder resulted. It is doubtful that the Composita could have attached itself or lived in the situation where it was found if Hercosestria had the valveflapping feeding habits postulated by Rudwick (1961) (see Plate 292: figures 1, 2; see also Plate 229: figure 34 for a trapped Derbyia). Color Color markings are regarded by Richter (1919) as evidence of shallow water. Several remarkable NUMBER 14 103 specimens preserving their color patterns were taken from the Cathedral Mountain, Road Canyon, and Word formations. Color bands frequently have been seen on Dielasma or Beecheria in the Mississippian, but color-banded specimens have not been described from the Pennsylvanian. At USNM 702c and 707e in the Road Canyon Formation and at USNM 706e in the Word Formation, numerous specimens of Dielasma were found with outwardly curved radial bands of color. These originate near the beak and curve laterally to the margins. The bands are now dark brown, but probably they were red or crimson. The color pattern is reminiscent of that of modern Laqueus rubellus (Sowerby), which has curving red bands (see Plates 749, 750, 753). In addition to the specimens of Dielasma, a brachial valve of Composita imbricata from USNM 721u in the Cathedral Mountain Formation has straight dark radii that extend from the beak to the anterior margin. They are very narrow at the beak, but they widen to the margin, where some have a width of about 2 mm (Plate 658: figure 31). Another genus not hitherto known to have color bands is N eophricadothyris (USNM specimen 155121). These bands are like the preceding except that they are confined to the anterior half. They are direct, without any curvature (see Plate 636). Malformations Brachiopods are gregarious animals and frequently live clustered together. Some modern genera, such as Laqueus and Liothyrella, may suggest bunches of grapes when different generations festoon the adults of previous generations. Such gregariousness causes extreme crowding, and the crowding prevents many specimens from growing normally. Specimens malformed by crowding are common in our present seas. They were even more abundant in the Permian, when large numbers of the brachiopods cemented to the substrate or to their parents. Malformations especially are common in the bioherms formed by Hercosia, Hercosestria, and the clusters of Cyclacantharia. The apical angles of specimens of the same species of Hercosia often vary by as much as 70 degrees. Specimens that attempt to grow between robust adults commonly are elongated greatly and may be as narrow as a lead pencil. Scacchinella, which is the foremost brachiopod bioherm builder, forms dense clusters in which many specimens are deformed. These commonly are small and narrow, but they may be twisted in any direction, and one cup (USNM specimen 153661g) is bent at an angle of 45 degrees. This may have resulted from part of the biohermal mass toppling over and the brachiopod being forced to grow in another direction. Another specimen from the same bioherm is twisted at a right angle. If the lower part of the pedicle valve of this specimen is held in a horizontal position, the upper half is at right angles to the lower half. Many other specimens from this bioherm (USNM 720e) are also distorted, but in less extreme ways (see Plate 278). Impingement against an unyielding object, such as a branch of a bryozoan or a stout productid spine, can cause malformation. A specimen of Edriosteges multispinosus Muri-Wood and Cooper (USNM specimen 154183d) has a deep reentrant in the anterior margin that divides the anterior into two strong lobes. This may have resulted from impinging against a branch of a bryozoan colony (Plate 218: figures 12-20). The way such a deformation may be brought about is shown by a specimen of Rhamnaria tenuispinosa, new species (USNM specimen 154217c), which is attached to a branching bryozoan in such a way that one branch has become embedded in the shell, producing a deep groove and anterior bilobation (Plate 261: figures 36, 37). Another source of such deformation is shown by a specimen of Rhamnaria kingorum Muir-Wood and Cooper, which found refuge on the concave surface of the brachial valve of Echinauris lateralis Muir-Wood and Cooper. The Rhamnaria spat has grown against one of the strong brachial spines that grow across the concave valve, presumably as defense against settling of larvae. The anterior margin of the Rhamnaria already has become anteriorly bilobate (see Plate 264: figure 22). Pathology Occasional specimens indicate possible pathological conditions. A specimen of Yakovlevia (USNM specimen 153979h; Plate 473: figure 17) shows a malformation of the muscle scar on the left (observer's) side. Not only is the left side affected, but also the right side is abnormal. The 104 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY diductor scar on the right side occupies its normal position, but also it extends over the median ridge that normally is occupied by the adductor scars and usually separates the adductors. The adductor impressions lie in the left side of'the median ridge, and much of the left diductor scar is aborted and misshapen. Similar deformities have been seen in the Devonian in specimens of Stropheodonta (USNM specimen 140852a), from the Romney Formation of West Virginia, and of Cariniferella tioga (Hall), from the Chemung Group of New York (Kaiser, 1964). As with Yakovlevia, the exterior of the pathological specimens is without injury; the deformation is concerned entirely with the interior. Another pathological condition is illustrated by the cardinal process of a large Tropidelasma culmenatum Cooper and Grant, from the Neal Ranch Formation at USNM 701h. The prong on the observer's right is aborted, and an excessive amount of calcareous tissue, in the form of a shallow cup, is laid around the anterior side of the prong. There is no sign of breakage of the prong, and the stub shows part of the characteristic slit that normally bears the diductor muscles (Plate 52: figure 28). Borings Shells with borings of one kind or another are fairly frequent in the Glass Mountains Permian. The rarest type is the perfectly round hole attributed to boring gastropods. A specimen of Torynechus caelatus Cooper and Grant from USNM 702 (USNM specimen 152595a) has both valves bored. The hole is slightly more than a millimeter in diameter and seems not to be bevelled. The boring on each valve is near the middle and consequently through the thinnest part of the shell. Schlaudt and Young (1960) and Rodda and Fisher (1962) described and illustrated burrows from the Glass Mountains Permian that are attributed to acrothoracic barnacles. Schlaudt and Young stated that, if the burrows penetrated the shell, they were not produced by barnacles, but these authors speculated that the thin protective film of shell over the inside of the burrows was not preserved during silicification or that it was lost during the etching process. Numerous specimens from the Glass Mountains show a thin protective film, indicating that the brachiopod defended itself against the borer. Some valves are riddled almost completely by these borings, but the shells obviously were dead. Completely riddled shells are fairly common around bioherms such as that at USNM 733j in the Sullivan Peak Member (Plate 56: figure 21; see also Tomlinson, 1969). A specimen of Derbyia texta, new species (USNM specimen 151176; Plate 87: figures 1, 2) and one of Meekella skenoides Girty (USNM specimen 153538) illustrate the reaction of the brachiopod animal to the boring barnacles. The larger borings in the Derbyia are about 0.75 mm wide and about 2.0 mm long, elliptical in outline. On the interior the brachiopod has formed a series of moundlike blisters to protect itself from the borers. None of the blisters is broken or open on the inside. Several of the blisters inside the Meekella are open, but the openings are irregular, and they suggest later breakage rather than openings produced by the boring organism. Another type of boring is less prevalent than the preceding, and, in at least one instance, it produced a response from the host. These borings are minute, and they may be confused with the exopuncta of Rhipidomella, in the shells of which they occur. The borings are most prevalent near the margins of the valves, and they can be detected by their nearly vertical penetration, unlike that of the exopuncta, which conform with the radial ornament and which are strongly oblique and do not affect the inner layer. The borings may penetrate the inner layer and may even protrude into the shell as points where the animal has laid shell around them. Many of the borings are open on the inside of the shell, but the thin protective coverings were probably not preserved (Plate 666; figures 28, 29). A boring suggestive of the preceding occurs in Phrenophoria subcarinata Cooper and Grant (USNM specimen 148385g; Plate 553: figures 35- 39). This is a minute opening on the margin that leads into a slightly elevated blister and then a needlelike tube about 0.75 mm in length. The tube is not open at the end, but this may be due to silicification of the fine point. This kind suggests the endoparasitic phoronid described in Devonian brachiopods by Biernat (1961). A tubular boring quite unlike those described above occurs in a specimen of Enteletes wordensis R. E. King. The boring is in the ventral valve of NUMBER 14 105 USNM specimen 153809r (Plate 686: figures 11- 13). The first hole appears on the right side of the shell on the margin, not quite halfway to the anterior. The hole on the exterior is 2.5 mm in diameter and produces a tube 7 mm long on the inside. This is jaggedly broken, but, when projected obliquely across the shell, it coincides perfectly with another tube bored on the costa that bounds the left side of the fold. The opening here is 2 mm wide, but the tube is only 4 mm long into the shell. This tube is open, but the edges are rounded and smoothed, suggesting that healing over the whole may have started. The borings are perfectly smooth inside. Faunas and Correlations of Glass Mountains Formations Faunal lists given in this section include some nomina nuda. This unfortunate circumstance comes about because of the economics of publishing such an extensive monograph. Originally it was intended that the entire work would appear at one time, but slender budgets for publication require the appearance of the monograph in several parts at varying intervals. To delete the nomina nuda would destroy or make ineffective some of our arguments on age, correlation, and faunal relationships. The nomina nuda will be validated in subsequent volumes. GAPTANK FAUNAS Any discussion of the faunas of the Glass Mountains must begin with the Gaptank fauna from which the Permian species were descended. Some of the characteristic elements of the Permian fauna appear first in the Gaptank (exclusive of the Uddenites- bearing Shale Member). Of these, Parenteletes and Diplanus were taken from the Gaptank bed 10 of P. B. King, said to be Upper Canyon = late Missourian, about 2 miles east of Gap Tank. Neither of these brachiopods is abundant, but the Parenteletes, usually crushed, is fairly common. Cryptacanthia, more typically Pennsylvanian but certainly the progenitor of Glossothyropsis, is fairly common at this place (USNM 700a; USGS 6705, 7085, 7088; P. B. King, 1937:77). Another common form in this bed is Teguliferina, which is an abundant Wolfcampian fossil. Other precursors of the Permian are huge Derbyia and Reticulata. The associates of these Permian heralds are characteristic Pennsylvanian types. In the western part of the Marathon Basin near the Arnold Ranch, other Permian types have been taken from a "block," possibly a displaced piece, but more probably a bioherm. This appears on P. B. King's map as "loc. b," but in reality it is locality C of his text (1938:81). The famous Prouddenites ledge on the map, mistakenly labelled "loc. C," is 0.5 mile south of the brachiopod bioherm in question. The bioherm (USNM 700g) is about 1.25 miles due south of Arnold Ranch (Monument Spring quadrangle; Plate 16: figure 1). It consists of a large limestone block in Gaptank shale, composed of a mass of the coral Amplexicarinia delicata Ross and Ross and algae, with brachiopods scattered throughout. These include Scacchinella, Limbella, and a large Teguliferina, suggesting T. boesei (R. E. King). The Scacchinella is the most primitive species of that genus known, as it has no vesicular shell tissue in the apex of the pedicle valve. Its outer form is very similar to Derbyella Grabau. The Limbella species is small, but otherwise it is characteristic of the genus. The Teguliferina attains as large a size as any seen in definite Permian rocks. The occurrence of these fossils is anomalous. The limestone block in which they occur is the only one of its kind in that part of the section. If it is an exotic block, its advanced fauna makes it difficult to postulate its place of origin or to explain its present location. The composition of the rock with its numerous corals and algae suggests a bioherm, rather than an exotic piece. The block is similar to the one that produced the Prouddenites ammonite fauna, which lies 0.5 mile to the south and has been regarded as in place. Further evidence as to the age of the block containing Scacchinella is a single fusulinid specimen, which has been pronounced Virgilian in age by Garner Wilde (letter of 20 March 1962). P. B. King suggested that the fauna identified by Girty from the same locality as this block (locality b=C; King, 1938:81) is similar to that of bed 10 of the Gaptank type section. This bed at the east end of the mountains produced Parenteletes, not present south of the Arnold Ranch, but it is dated generally on its fusulinids and molluscs as Missourian 106 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY GU A 0 A L U PE M O U N T AINS O t L A W J B E B A S I N S I E R R A D I A B LO CHINATI MOUNTAIN S G L A S S MOUNTAINS W E S T CENTRAL EAST RANGES OF SOME IMPORTANT FOSSILS C A P I T A N ( R A D ER GOAT u 2 2 S SPONGE SPICULES W E D I N > HUECO MOUNTAINS ALACRAN MOUNTAIN L O W E R M A S S IVE B O N E S P R I NG CERRO ALTO HUECO CANYON POWWOW G R O U P POWWOW MAG OALENA OROOVICIAN BRECCIATED ZONE TRANSITION BEDS ALTA (PART) S U L L I V A N i U G 0 U T < ^ MTN. \ ^TAYLOR RANCH P O P L A R T A N K D E C I E / R A N C H SCHWAGERINA CRASSITECTORIA L E N O X G A P T A N K UDDENITES ng SHALE MEMBER ?. = o u r_ FIGURE 28.—Chart showing correlation of Permian formations in West Texas. (Upper Canyon), a far older date than indicated by the Virgilian fusulinid in the Gaptank block or by its brachiopod fauna. In the Pennsylvanian of north-central Texas, the early Cisco (Graham Formation) contains Permian elements such as Waagenoconcha and Martinia. Some workers may object to the latter genus being regarded as a Permian element, but, except for this occurrence, it is not known from American Pennsylvanian formations (although it is rather rare in the American Mississippian). Consequently, its appearance in the late Pennsylvanian heralds a Permian development. Poikilosakos from the Graham of north-central Texas also is a prophetic element foreshadowing the considerable development of Pseudoleptodus in the Neal Ranch Formation and younger parts of the Permian. Gypospirifer, new genus, like Spirifer condor (d'Orbigny), of the Copacabana Formation in the Permian of Bolivia, occurs in the Cisco (Wayland Shale, USNM 728x). A significant Permian element said to appear in the Gaptank Formation is Spyridiophora, reported as coming from bed 3 of the Gaptank (R. E. King, 1931:71), about 1 mile south of Gap Tank. The specimen was identified as Productus gratiosus occiden talis Schellwien by R. E. King and reported as occurring in a part of the section with Mesolobus of Missourian age. Inasmuch as Spyridiophora is not yet known anywhere else from the higher Pennsylvanian, it was suspected that the specimen NUMBER 14 107 might not have been taken at the place from which it was reported. The matrix of the specimen was searched for fusulinids, and those found were submitted to Mr. Garner Wilde, who reported them to be Wolfcampian species. This clearly indicates that the specimen is not from the Pennsylvanian and that Spyridiophora has its roots in the Permian. Fifty-six genera pass from the Gaptank Formation into the overlying Permian. Some of these have not yet been found in the Uddenites-bearing Shale Member, but they do appear higher. In the list below, starred genera (*) are those passing from the Pennsylvanian to become the hard core of the Permian (Wolfcampian) fauna; genera marked with a degree sign (°) do not survive beyond the end of the Wolfcampian as defined herein; unmarked genera extend into, or terminate by, the end of the Leonardian. Meekella, for example, generally is not common in the Pennsylvanian, but it becomes one of the most frequent Permian forms in many environments. These Pennsylvanian elements of the Permian are: Antiquatonia0, Beecheria0', Calliproton,ia°, Cancrinella*, Chonetinella, Cleiothyridina*, Composita*, Cooperina*, Crurithyris*, Cryptacanthia°, Derbyia*, Derbyoides °, Dielasma*, Diplanus, Echinaria0, Echinauris*, Enteletes*, Eridmatus0, Fimbrin,ia°, Goniarina, Heteralosia*, Hustedia*, Hystriculina0, Isogramma, Juresania°, Kozlowskia0, Kutorginella, Limbella0, Linoproductus*, Lissochonetes°, Martinia*, Neochonetes°, Phricodothyris* [Condrathyrisf], Neospirifer*, Nudauris0, Orthotichia0, Parenteletes0, Petrocrania*, Poikilosakos°, Pontisia*, Psilocamara°, Quadrochon,etes°, Reticulariina*, Reticulatia0, Rhipidomella, Rhynchopora*, Roemerella, Scacchinella*, Schuchertella0, Gypospiriferu, Stenoscisma*, Streptorhynchus [correct genus uncertain], Sulcataria0, Teguliferina*, Waagenoconcha*, and Wellerella. WOLFCAMP FAUNAS Uddenites-BEARiNG SHALE MEMBER.—The brachiopod fauna of the Uddenites-bearing Shale Member was listed by R. E. King (1931:147-150), who recognized the truly Permian elements in it as well as those with a Pennsylvanian aspect. Our collections from this zone unfortunately are not extensive, but we are able to offer some revision of this fauna and to emphasize its Permian flavor. Little attention has been paid to the typically Permian kinds of brachiopods (and some other phyla) in the Uddenites-bearing Shale Member in spite of the fact that R. E. King made an excellent discussion of them and that P. B. King later (1937:78) repeated his brother's analysis. Innovations in the Uddenites-bearing Shale Member fauna are few. Orthotetella appears for the first time, as does Tropidelasma. Martinia is reported by R. E. King, but we did not find it. Although Kochiproductus is reported in the Pennsylvanian of Peru (as Buxtonia) by Chronic (1953: 83, 84), it appears so far as known, for the first time, in the Glass Mountains in the Uddenitesbearing Shale Member. King lists 36 species, and we record 45 species, from the Uddenites-bearing Shale Member. Combining the two lists, there are 54 species in all, many of which are strictly Permian types. For some of the forms assigned by R. E. King to Pennsylvanian species, we have found it necessary to erect new names. The brachiopod fauna is distinctive, and we think its Permian affinities are sufficiently striking to use it as the base of the Wolfcampian in the Glass Mountains as did R. E. King (1931). The 54 species in the combined list are distributed among 41 genera, of which 16 are definite Permian types, that is, they have their acme in the Permian: Parenteletes, Cenorhynchia, Diplanus, Tropidelasma, Orthotetella, Teguliferina, Pontisia, Scacchinella, Echinauris, Limbella, Kochiproductus, Kutorginella, Waagenoconcha, Gypospirifer, Martina, and huge forms of Meekella. In addition to these, three others may be considered as Permian forms: Fimbrinia, Nudauris (Productus semistriatus King, not Meek, whose type-species is a Pennsylvanian form), and Chonetinella. These last three and Kutorginella occur in the late Pennsylvanian. The first is rare in either period, the second is rare in the Pennsylvanian so far as known, the third becomes most abundant in the Early Permian, while the last is commonest in the Leonardian. According to R. E. King, 31 of his 36 listed species pass into the overlying Neal Ranch Formation. Of the 33 listed by us, 14 also are identified in the Neal Ranch Formation: Chonetinella biplicata (R. E. King), C. spinolirata (R. E. King), Diplanus redactus, new species, Hystriculina ventroplana, new species, Echinauris subquadrata, new species, 108 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Kutorginella uddeni, new species, Limbella wolfcampensis (R. E. King), Gypospirifer anancites, new species, Meekella texana R. E. King, Pontisia kingif, new species, Rhipidomella miscella, new species, Rhynchopora molina, new species, Teguliferina boesei (R. E. King), and Waagenoconcha prophetica, new species. There is no faunal break of any consequence. Innovations and expansions of stocks passing from the Uddenites-bearing Shale Member occur in the overlying Neal Ranch Formation. Age: We take the presence in this shale member of Parenteletes, Limbella, Scacchinella, Waagenoconcha, Kochiproductus, and Gypospirifer as indicating a Permian age for these beds. The first three occur in the Trogkofel Limestone of the Carnic Alps; the last three occur in the Permian (Copacabana Formation) of Peru; outside the United States, Scacchinella is known only from lower Permian rocks. This is also true of Parenteletes, which occurs in Permian formations in Sicily, Japan, and China. An aulostegid strongly suggestive of Limbella occurs in the Trogkofel Limestone of the Carnic Alps (Productus cancriniformis of Schellwien, not Tschernyschew). Waagenoconcha is regarded widely as a Permian indicator in the United States (Knight, 1940:1129). Along with these brachiopods, two specimens of Omphalotrochus, also regarded as an important Permian guide, were taken from the Uddenites-bearing Shale Member (Yochelson, 1954). Correlation: The correlation of this shale member has been confused for years because proponents using data from different phyla are in disagreement. The ammonites, fusulinids, and brachiopods indicate different ages to their practitioners, although the differences gradually are coming to be semantic rather than substantive. The ammonites at first were regarded as Permian by Bose (1917) and Smith (1929) but were transferred to Late Pennsylvanian by Plummer and Scott (1937:388) because of identity with species in the Cisco Group of north-central Texas. The fusulinids generally have been taken to indicate a late Pennsylvanian age, but Bostwick (1962), claiming to have discovered Schwagerina in the Uddenites-bearing Shale Member, advocates a Permian age for it. Presence of Omphalotrochus in the member suggests a Permian age as argued by Knight (1940) and Yochelson (1954). On the basis of brachiopods, R. E. King (1931:6) correlated the Uddenites-bearing Shale Member with parts of the Permian, citing the faunal similarity to the Copacabana beds of Bolivia and the Trogkofelschichten of Yugoslavia. Our findings indicate a closer similarity to the Permian of Bolivia than to that of Yugoslavia, which in our opinion correlates with Neal Ranch and Lenox Hills Formations rather than with the Copacabana. Despite the absence of Scacchinella and other reefal types from the Copacabana of Bolivia, the correlation of the fauna of this formation with that of the Uddenites-bearing Shale Member is tenable. Actually, few typically Permian genera of brachiopods occur in the Copacabana fauna: Productus boliviensis, Kozlowskia capaci (d'Orbigny), Kochiproductus peruvianus (d'Orbigny), Waagenoconcha humboldti (d'Orbigny), and Gypospirifer condor (d'Orbigny). Forms similar to these occur in the Uddenites-bearing Shale Member. In addition, the following unusual Copacabana types also are found in the Uddenites-bearing Shale Member: Hystriculina, Reticulata, Cancrinella, Fimbrinia, Echinauris boulei-type, Orthotichia, and Rhynchopora. Some of these also occur in the Cisco Group of the Upper Pennsylvanian. Some of the species of these genera are very similar in the two levels under consideration. The majority of common genera usually found in the upper Paleozoic are also specifically close in the formations being considered. NEAL RANCH FORMATION.—This formation brings more Permian elements into Texas to mingle with those inherited from the Uddenites-bearing- Shale Member and with these persistent, transient forms extending from the Pennsylvanian. First appearances are: Altiplecus, Atelestegastus, Camarelasma, Hemileurus, Hypopsia, Lepidocrania, Nothopindax, Acosarina, Pseudoleptodus, Eolyttonia, Ctenalosia, Geyerella, Striatifera, Spyridiophora, and Spuriosa. The following genera, all of them rare, are confined to the Neal Ranch Formation according to our present knowledge: Atelestegastus, Camarelasma, Hemileurus, Hypopsia, and Nothopindax, The genus Spuriosa also occurs in the Hueco Canyon Formation. Its Neal Ranch occurrence probably is earlier. A few genera coming in from the underlying Udden /to-bearing Shale Member or the Pennsylvanian have their last appearance in the Neal Ranch Formation: Beecheria, Eridmatus, NUMBER 14 109 Juresania, Lissochonetes, Psilocamara, and Neochon. etes. The remainder of the fauna is composed of a variety of orthotetids, derbyids, productids, lyttoniids, spiriferids, and athyridids. The orthotetids and derbyids are exceptionally well developed in the Neal Ranch. Some bioherms or small patch reefs almost completely comprise one or more of their genera. The bioherm at USNM 70lh contained great quantities of Geyerella, while those at USNM 70Id were replete with Derbyia of several species. Great abundance of Meekella characterizes USNM 701g. Of the productids, special mention should be made of the abundance of Striatifera at USNM 721g, where specimens are scattered among bryozoan colonies that form a zotikepium. Of particular interest and importance in the Neal Ranch Formation is the budding of the Oldhaminidina, which flower later in the Leonardian. Pseudoleptodus appears in fair abundance at USNM 721g and 727e, where it occurs attached to bryozoans and Striatifera. Eolyttonia usually is rare, but some local patches rich in this genus were discovered. The earliest oldhaminid patch occurs at USNM 722x either on, or in, the flank of King's bed 2 or Gray Limestone. Here, the specimens often are engulfed by concentrically banded limestone of probable algal origin. The large bioherm at USNM 701c has Eolyttonia of exceptionally large size. Age: In correlating the Neal Ranch Formation is is necessary first to establish the age of the Gray Limestone Member, which P. B. King placed in the Wolfcamp Formation but which Ross (1963a: 13, 45) returned to the Gaptank. The Uddenitesbearing Shale Member immediately underlies the Gray Limestone Member, which forms a conspicuous lenticular mass in the center of the Wolf Camp Hills. Although this member traditionally was a part of the Wolfcamp Formation, Ross placed it at the top of the Gaptank Formation on the basis of certain fusulinids. We are unable to agree with this assignment because the brachiopods and ammonites we have collected from bed 2 indicate a Permian age. The collection from USNM 701, from the upper 15 feet of the member as exposed in Geologists Canyon opposite the first arroyo from the north in about the center of the hills, includes many Permian genera, such as: Diplanus, Geyerella, Hypopsia, Orthotetella, Eolyttonia, Cooperina, Limbella, Spyridiophora, Waagenoconcha, Parenteletes, Altiplecus, and Camarelasma. In addition, species of Enteletes, Derbyia, Meekella, Cancrinella, Neochonetes, Crurithyris, Orthotichia, and Tropidelasma occur in the Gray Limestone and also extend into the overlying beds of the Neal Ranch Formation. The brachiopods cited clearly indicate a Permian age, and the ammonites tell the same story. From the uppermost part of the member (USNM 701), Artinskia artiensis Griinewaldt was taken along with species of Eoasianites, Neopronorites, and Marathonites. The fusulinid evidence for the age of this member is equivocal. P. B. King (1942:647) states, in referring to the Uddenites-bearing Shale Member, that the "overlying gray limestone bed at the base of the Wolfcamp formation (beds 2 and 3) contains a few specimens of Schwagerina." In footnote 182 on the page following the above reference, King states: "The occurrence of Schwagerina in the gray limestone has not been mentioned in any publication, and was reported to the writer by J. W. Skinner, September, 1938." The evidence of the brachiopods and ammonoids indicates a Permian age for this member. This evidence of the fusulinids needs further testing. Most of the brachiopods from the Neal Ranch Formation come from the middle part, i.e., beds 12-14 of P. B. King ( = beds 9-12 of Cooper). Most of the fossils from these beds are from bioherms and thus are restricted in distribution. Except for the lower part of the Cibolo Formation in the Chinati Mountains, we know of no Wolfcampian like the Neal Ranch. Some Neal Ranch elements such as Orthotetella, Acosarina, and Eolyttonia occur in the lower Bone Spring Formation, but they are regarded, on the basis of their associates, to be younger than the Neal Ranch. The most significant genera of the Neal Ranch Formation for use in correlation are: Acritosia, Derbyoides, Diplanus, Eolyttonia, Geyerella, Limbella, Nudauris, Orthotichia, Parenteletes, Reticulata, Spyridiophora, and Teguliferina. It will be noted that some of this list (numbers 3, 6, 8—11) already have been cited as leading members of the fauna of the Uddenites-bearing Shale Member. Most genera of the above list occur in the bioherms. Orthotichia, Nudauris, and Reticulatia also occur in the shales; the last genus mentioned seldom is seen in bioherms. Derbyoides and Reticulatia are 110 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY more common in the midwest in the more shaly facies of the Early Permian. Correlation: In the Permian correlation chart of the National Research Council (Dunbar, et al., 1960), the Neal Ranch Formation is correlated either by position, ammonites, or fusulinids, with formations elsewhere in the United States. The Neal Ranch brachiopods predominantly consist of forms confined to bioherms and a few others that occur in shale. No bioherms comparable to those of the Neal Ranch are known in Wolfcampian sediments in the Midcontinent or elsewhere. Consequently, correlation must be based on those forms of a facies comparable to that containing the genera that occur in the shales or nonbiohermal limestone. Another difficulty in correlation between the Glass Mountains and other areas is the fact that the brachiopods of the Pennsylvanian and Permian are known so poorly. No recent monographic work has been done on the brachiopods of the Pennsylvanian and Permian of Oklahoma, Kansas, and Texas. The faunas of the Council Grove and Chase Groups scarcely are known and the same is true of those of the Pueblo, Moran, Putnam, and Admiral Formations of north-central Texas. A few species from the Permian of Nebraska are described by Dunbar and Condra (1932), but nothing comprehensive yet has been prepared. The few species listed here as common to the Midcontinent region and the Glass Mountains do not help with definitive correlation. Although species of Hystriculina, Echinaria, Antiquatonia, Kozlowskia, Kutorginella, Neochonetes, Reticulatia, Neospirifer, and Juresania occur in the shales of the Neal Ranch Formation, all but two of the species are different from those of the Admire and Council Grove Groups in Kansas, Oklahoma, and Nebraska, or from those of the Permian formations of northcentral Texas. The brachiopods of the latter are more like those of the Midwest than they are like Neal Ranch brachiopods. On the basis of their fusulinids and ammonites, these Midcontinent formations are placed here in the Wolfcampian, but they are not correlated except in a general way. LENOX HILLS FORMATION.—This formation has not yet yielded a good supply of brachiopods or other fossils. Most of the formation is composed of conglomerate in the western part of the mountains and of dolomites, red shales, and conglomerate in the eastern part. In the Lenox Hills and Dugout Mountain a conspicuous shale occurs high in the Lenox Hills Formation, but intensive search has yielded only a few ammonites, gastropods, brachiopods, and some fusulinids. A problem exists as to whether or not the beds identified as Neal Ranch by Ross (1963a:24, pi. 1), which are rich in bioherms just east of hill 5300, are in reality a part of the Lenox Hills Formation (our USNM 715b). As noted above, the conglomerate of the Lenox Hills interfingers with these supposed Neal Ranch limestones. The brachiopod fauna in them is that of the Neal Ranch Formation (beds 12-14 of P. B. King = beds 9-12 of Cooper), but we believe that the field relations prove these limestones to be part of the Lenox Hills Formation. The Lenox Hills Formation in Leonard Mountain and some beds of the Hess Ranch Horst yielded a few specimens. Thin shales at the top of the Lenox Hills Formation just east of the Hess Ranch Horst (R. E. King 196 = USNM 716r) yielded species not seen elsewhere. We did not find any silicified limestone in the Lenox Hills Formation. The only innovation in the Lenox Hills Formation is the genus Dasysaria. In the Glass Mountains this genus is confined to that formation. It is very rare in the Glass Mountains, but it is abundant in the Hueco Canyon Formation and in other parts of the Hueco Group in the Sierra Diablo, Hueco, and Franklin Mountains. It also is common in the Chase Group of Kansas and the Putnam and Admiral Formations of north-central Texas. Two genera terminate in the Lenox Hills Formation: Sulcataria and Reticulatia. The first is a poorly known and rare chonetid, first seen in the lower Cisco and in the Neal Ranch Formation. The second is uncommon in the Glass Mountains, but it is frequent in the Midcontinent region and north-central Texas. Parenteletes, another diagnostic genus of the Lenox Hills, is not common in the Glass Mountains, but it attains a large size in the Neal Ranch Formation. It is very rare in the Lenox Hills Formation, but it was discovered in the base of the Poplar Tank Member of the Skinner Ranch Formation. The large detached slide block of the Lenox Hills Formation, about one-third the height of Leonard Mountain on the southeast nose, is a source of Lenox Hills fossils (USNM 705m). These rocks are biohermal in part, but mainly a coarse calcarenite or rudite. From them, we have taken NUMBER 14 111 Geyerella equal in size to those of the Decie Ranch Member as well as large conical Eolyttonia. Other associated species are robust, and all strongly resemble those of the Decie Ranch Member of the Skinner Ranch Formation. Of special interest in the Lenox Hills Formation on Leonard Mountain, just above the conglomerate, is the bioherm containing Scacchinella, Parenteletes, and Tropidelasma (USNM 705k). This is on the southeast nose of the mountain, and the horizon represented was mistaken early for the base of the Hess Formation. This is the only occurrence of abundant Scacchinella yet found in the Lenox Hills Formation. A few valves have been taken elsewhere, but the genus is rare in this formation. Correlation of the Lenox Hills Formation with strata elsewhere in the United States is easier and more reliable than that of the Neal Ranch fauna. Fusulinids, ammonites, and brachiopods of the Lenox Hills seem to combine to tell essentially the same story. Dasysaria, restricted to the Lenox Hills in the Glass Mountains, is abundant in the Putnam and Admiral Formations, the Wichita Group, and the Hueco Canyon Formation. Reticulatia huecoensis (R. E. King) occurs in the uppermost Lenox Hills (King 196 = USNM 716r). A Stenoscisma related to S. hueconianum (Girty) is present in the Glass Mountains and relates part of the Lenox Hills to the Alacran Mountain Formation of the upper part of the Hueco Group. Species of Neospirifer, Echinauris, Nudauris, and Kutorginella are present in the Glass and Hueco Mountains. Kochiproductus is common in the Hueco Canyon Formation, but it has not yet been found in the Lenox Hills Formation, although it should be expected. Williams (1963:31), on the basis of fusulinids, correlates most of the Hueco Group with the Lenox Hills Formation, but he assigns the part of the Alacran Mountain Formation that contains Schwagerina crassitectoria Dunbar and Skinner to the Leonardian. The brachiopods of the Hueco Group are a fairly uniform lot, and probably they are correlated best with the entire Lenox Hills. SKINNER RANCH FORMATION.—This formation in the western part of the Glass Mountains has three members, but it is undivided eastward from hill 5021. Each of the members has its own fauna and history. This is true of the undivided formation as well; the important events in each are mentioned below. In the Skinner Ranch Formation, certain persistent, transient elements from the Pennsylvanian finally are eliminated, but strong Wolfcampian flavor is maintained throughout the formation. Decie Ranch Member: A few newcomers appear in this otherwise largely Wolfcampian assemblage: Acolosia, Chonosteges, Coscinophora, Cyclacantharia, Antronaria, Cartorhium, Oncosarina; Rhamnaria, Rugaria, Thamnosia. The remainder of the 29 genera of the Decie Ranch fauna were inherited from the Lenox Hills fauna. A remarkable feature of this assemblage is the large size attained by many of its members. Scacchinella attains a length of 7 inches; Derbyia is 3—4 inches in width; Eolyttonia reaches the proportions of a teacup; Geyerella, Kochiproductus, Coscinophora, and Tropidelasma are all large forms. Perhaps the most striking feature of the assemblage is the presence of bioherms of Scacchinella. Intergrown with the Scacchinella are Geyerella, Tropidelasma, Eolyttonia, and Derbyia, which make a fairly strong mass. These bioherms commonly are based on a gravel of huge, crinoid-stem debris and are surrounded by coarse limestone pieces in the form of thick conglomerates. The whole member thus suggests a near-shore gravel bed in a zone of strongly moving waters. The patch reefs are strong, and they provided, in crannies and niches, abundant quiet shelter for the numerous small brachiopods that make up a large part of the Decie Ranch fauna. Poplar Tank Member: The fauna of the Poplar Tank Member is not well known because much of the member is composed of brown, crumbly shale that yields few good fossils. The thin bands of limestone in the shale are conglomeratic, and their fossil content usually consists of fragmentary specimens often identifiable only to genus. The sandstones and conglomerates do not yield good fossils. An occasional bioherm, such as that at USNM 708e, produces numerous fine but unsilicified fossils. Silicified material was not discovered in useful quantity. Many of the thin limestone beds are capped by an inch or two of brown chert, which usually is very fossiliferous, but the preservation is poor. The Poplar Tank fauna contains most of the same species as that of the underlying Decie Ranch Member and the overlying Sullivan Peak Member. 112 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Large Scacchinella is present, and Spyridiophora is fairly common in contrast to its extreme rarity in the Decie Ranch Member. Parenteletes, at the bases of the member, is a Wolfcampian hangover like Spyridiophora. Nudauris, a Gaptank and Neal Ranch genus, was found in the Poplar Tank Member. The only novelty appearing in the Poplar Tank fauna is Paranorella, very rare and poorly preserved. This probably is predated by occurrence of the same genus in the lower Bone Spring Formation in the Sierra Diablo. Dugout Mountain Member: The fauna of this member, which consists of the Second, Third, and Fourth Limestones of the Leonard of P. B. King (only in the vicinity of Dugout Mountain) and their intervening shales, is known very poorly. The limestones mainly are detrital, consisting of small pebbles, chert fragments, broken shells, and other debris. The result of collecting or dissolving blocks from these beds usually is a large quantity of fragmentary specimens difficult to identify. The known fauna is Wolfcampian in aspect, but new elements appear in it: Sceletonia, Torynechus, Glyptosteges, and Lepidospirifer. Another element of the Dugout Mountain fauna is that of the ammonites, which are abundant in the limestone beds. These include a new genus (formerly Propinacoceras) and small species of Perrinites. Other ammonites are: Eothinites hessensis Miller and Furnish, Metalegoceras, Peritrochia dunbari Miller and Furnish, and a species of Neocrimites. The Dugout Mountain fauna is closest to that of the Sullivan Peak Member of the Skinner Ranch Formation. The two members share the rare genera Sceletonia, Torynechus, and Glyptosteges. We have not found Institella and its associates in this member. The fauna seems to have close ties below, and we have found its characteristic ammonites in the upper part of the Sullivan Peak Member at hill 4801 at the south end of the Lenox Hills. Outside of the Glass Mountains, Sceletonia was found 130 feet above the base of the Bone Spring Formation at the north end of the Baylor Mountains (USNM 725c). Sullivan Peak Member: This is the uppermost member of the Skinner Ranch Formation, and some major faunal changes take place in it. Most of the Wolfcampian elements and the last of the persistent holdovers from the Pennsylvanian disappear. Above the Skinner Ranch a number of new genera arrive to inaugurate the Leonardian. Genera appearing for the first time in the Sullivan Peak are Phrenophoria and Plectelasma. Genera seen for the last time in the Sullivan Peak are discussed under "Skinner Ranch Formation Undivided." The most striking feature of the Sullivan Peak fauna is its similarity to the underlying Decie Ranch, Dugout Mountain, and Poplar Tank Members. Bioherms, often of large size, are fairly frequent in the Sullivan Peak Member, especially at the south end of the Lenox Hills in hill 4801. These contain large Scacchinella and Geyerella that almost attain the size of the large specimens from the Decie Ranch Member. Tropidelasma and other reef-dwelling types occur in the formation. Coscinophora of large size occurs in several bioherms, especially on the northeast side of Dugout Mountain (USNM 733j). Perhaps the most characteristic elements of the fauna are Spyridiophora and Glyptosteges. The former is abundant in places, but the latter, although distinctive, is much less common. Torynechus also is a marker of the upper Sullivan Peak, but it is transient into the overlying Cathedral Mountain Formation. This is true also of Spyridiophora and Glyptosteges, but neither they nor Torynechus established themselves in the Cathedral Mountain fauna; they are extremely rare in the lower part of that formation. Spyridiophora reticulata (R. E. King), the abundant species of the Sullivan Peak Member, occurs in the Sierra Diablo a short distance above the base of the Bone Spring Formation on the south side of the mouth of Victorio Canyon (AMNH 625 = USNM 725e). It also occurs with Glyptosteges in Apache Canyon at the north end of the range. The rocks in Apache Canyon containing these species, although near the base of the formation, represent a higher horizon than that at Victorio Canyon. Spyridiophora reticulata also helps to tie the Sullivan Peak fauna to that of the Taylor Ranch Member of the Hess Formation. Presence of Parafusulina spissiseptata Ross in both of these members is another help in establishing this link across the mountains between two very unlike facies (see "Taylor Ranch Member" under "Hess Formation"). NUMBER 14 113 Skinner Ranch Formation Undivided: From hill 5021 (west of Iron Mountain) eastward to its merger with the Hess Formation, the Skinner Ranch Formation cannot be separated satisfactorily into its component members. It is possible, however, to recognize faunas of the Sullivan Peak in the upper part and faunas of the Decie Ranch at the base. Some genera appear for the first time in the lower 200 feet of the Skinner Ranch Formation, which is the part believed to be correlative to the Decie Ranch Member. These newcomers are: Anomalesia, Crenispirifer, Elliottella, Metriolepis, Peniculauris, Spinifrons, and Tricoria. All of these except Peniculauris and Tricoria appear also in the lowest part of the Bone Spring Formation in the Sierra Diablo on the south side of Victorio Canyon (USNM 728e). Several genera, mostly Pennsylvanian-Wolfcampian types, make their last appearance in the Skinner Ranch Formation: Anomalesia, Antiquatonia, Chonetinella, Elliottella, Fimbrinia, Glyptosteges, Hystriculina, Kozlowskia, Limbella, Oncosarina, Orthotichia, Parenteletes, Rugaria, Scacchinella'*, Spyridiophora*, Teguliferina, and Tricoria. The disappearance of these genera, in our opinion, marks the end of the Wolfcamp Stage. Occasional specimens of the starred genera (*) have been found low in the Cathedral Mountain Formation. Exceedingly rare, they have never established themselves as members of the Cathedral Mountain fauna. A single fragment of Scacchinella probably is a reworked fragment. Interesting developments involving the upper and lower parts of the Skinner Ranch fauna have been detected. It is our belief that the lower 100 to 200 feet of the Skinner Ranch near the Hess Ranch house and the Hess Ranch Horst represent the equivalent of the Decie Ranch Member. It is in this area that the Decie Ranch fauna is enriched by new elements. Variation in the Decie Ranch fauna may be detected along the Glass Mountains front from Dugout Mountain to the Hess Ranch and Hess Ranch Horst. In the western part of the mountains, in the Lenox Hills, the Scacchinella and associated Derbyia, Geyerella, and Eolyttonia are exceptionally large. Eastward these appear to become smaller and more slender, while Geyerella and Tropidelasma nearly disappear from the numerous bioherms on the Hess Ranch Horst, although they are present just north of the Hess Ranch house. Beginning on the west side of Leonard Mountain, a number of small productids, such as Elliottella, Hystriculina, and Kozlowskia, reappear in the Decie Ranch Member. In the bioherms that dot the base of the nordi slope of the Hess Ranch Horst, Orthotichia is common and Cyclacantharia appears in some abundance. Although present, neither of these is conspicuous in the western part of the mountains. North of the Hess Ranch house, Tricoria is fairly common, but it is rare on the horst. Tropidelasma is rare or absent from Leonard Mountain, Hess Ranch, and Hess Ranch Horst exposures. Spyridiophora is rare in the western part of the mountains, but it was not seen in the equivalent of the Decie Ranch Member in the eastern part. Another faunal aberration in the base of the Skinner Ranch Formation is the distribution of the fusulinid Schwagerina crassitectoria Dunbar and Skinner. This is common in the basal part of the Skinner Ranch Formation, referable to the Decie Ranch Member, but it has not been taken from the Decie Ranch Member proper in the Lenox Hills. It is reported by Ross (1962b:3, 9), however, at his locality 6A in the center of the Lenox Hills, 245 feet above the base of the Leonard ( = the base of the Decie Ranch Member). This level is near the top of the Poplar Tank Member of the present classification. It has not been seen by Ross or anyone else in the Decie Ranch Member. This species has a vertical range of more than 200 feet in the Hess Formation, and the occurrence recorded by Ross may indicate only the upper part of its range in the western part of the mountains. The above data might be taken to indicate a difference in age of the Decie Ranch Member and the basal part of the Skinner Ranch Formation (undivided) between the western and eastern parts of the mountains. We have not been able to satisfy ourselves that this is the case. Too many other species and genera are common to the members in the two areas, and their Scacchinella bioherms always can be found just above the Lenox Hills Formation. It seems to be a fact, as abundantly seen in the Neal Ranch and higher in the Cathedral Mountain Formations, that the faunas of individual bioherms, even of ones in close proximity, may be startlingly different. The bioherm is an area of microenvironments that enhance the faunal diversity of any member or formation. 114 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY HESS FORMATION.—This formation is notorious for the scarcity of megafossils and the poor preservation of those found in it. This is true of the whole formation except for two levels, one near the top and the other at the very top. The Hess Formation spans a considerable interval of the geological column in the Glass Mountains. The lower 400—500 feet of the formation contains fusulinids that date this part as Wolfcampian. Ross has applied the name "Lenox Hills Formation" to this part of the Hess section, in spite of the fact that the conglomerate at its base is thin or absent and the remainder of the formation consists of red shales, dolomites, and thin-bedded limestone totally unlike Lenox Hills lithology elsewhere. The Wolfcampian or "Lenox Hills" part of the Hess Formation continues up to the Schwagerina crassitectoria zone, which marks the level of the lower Skinner Ranch (Decie Ranch Member). This zone spans at least 200 feet of rock that we place in the Hess Formation above the Lenox Hills Formation (Cooper and Grant, 1966:3). The zone is the only marker for the lower part of the Skinner Ranch equivalent in the Hess Formation. The fauna of the Taylor Ranch Member marks the upper part of the Skinner Ranch level (Sullivan Peak Member) in the Hess Formation. Taylor Ranch Member: The Taylor Ranch Member is a thin band of shaly limestone underlain by a limestone conglomerate (Cooper and Grant, 1966:3). The member can be traced near the crest of the mountains for several miles, from a mile or two east of the Hess Ranch nearly to the Conoly Brooks Ranch house. Although the member is barely 40 feet thick, its upper half is very fossiliferous and contains a variety of brachiopods. It also contains interesting bioherms (USNM 702e above the Bill Neal Ranch house), which contain large numbers of Heliospongia and Girtycoelia with Spyridiophora and other brachiopods. Although these sponges make up a large part of the main bioherm, brachiopods are abundant and some are known only from this reefy mass. Two genera appear for the first time in the Taylor Ranch Member: Elassonia and Tschernyschewia. The former is a small rhynchonellid that appears in some abundance in higher formations. The latter is an exceptional brachiopod, and its discovery was completely unexpected because the genus is known elsewhere only in rocks that are dated as uppermost Permian. It is reported from the Dzhulfian of Armenia, from the Upper Productus Limestone of Pakistan, and from the Upper Permian of Yugloslavia. The genus first appears in the Taylor Ranch Member, but it also occurs in the Road Canyon Formation and in the lower Bone Spring Formation of the Sierra Diablo, each occurrence being in the Lower Permian as that period is now classified. The Taylor Ranch Member contains (besides the newcomers mentioned above) Chonosteges, Diplanus, huge Enteletes, Thamnosia, Linoproductus, Oncosarina, abundant Peniculauris, Pontisia, abundant large Rhipidomella, Rugaria, Stenoscisma, and Spyridiophora. Elements of the Taylor Ranch fauna occur on Leonard Mountain in the notch near the top of the southeast nose, where large Rhipidomella hessensis and Spyridiophora were found. Several specimens of Scacchinella were found loose on the slopes of the Taylor Ranch Member. These were traced by their lithology to a dolomite bed about 25 feet above the Taylor Ranch Member, but none was found in place. The Taylor Ranch fauna is difficult to correlate in the Glass Mountains because of its isolation and because so many of its members are unknown elsewhere. The presence of Spyridiophora reticulata (R. E. King), Scacchinella, and Rhipidomella hessensis R. E. King indicate relationship to the Sullivan Peak Member of the Skinner Ranch Formation. Other ties to the Sullivan Peak are Oncosarina, Peniculauris, and Meekella hessensis R. E. King. Antronaria speciosa, new species, strongly suggests A. mesicostalis (Girty) of the Bone Spring Formation, which correlates with the Skinner Ranch. A further tie to the Sullivan Peak is the presence of Parafusulina spissiseptata in the two members. The uppermost fauna of the Hess Formation was found near Old Word Ranch (USNM 726n), where a few species occur in excellently silicified condition. This locality is noteworthy for Plectelasma kingi, new species, which also was found in the upper part of the Sullivan Peak Member in Dugout Mountain (USNM 727a). In summarizing, we believe that the Taylor Ranch Member and upper Hess Formations are correlated most satisfactorily with the Sullivan Peak Member of the Skinner Ranch Formation. The Schwagerina crassitectoria zone establishes the NUMBER 14 115 position of the lower Skinner Ranch or Decie Ranch Member equivalent in the Hess Formation. Thus, the Hess above the Lenox Hills of Ross is the equivalent of the Skinner Ranch Formation. LEONARD FAUNAS As now conceived, the Leonard Series in the Glass Mountains consists of two formations: the Cathedral Mountain and the Road Canyon. The latter is the First Limestone Member of the Word Formation of P. B. King (1931), raised to formation rank by Cooper and Grant (1964) and placed in the Leonardian by them (1966). The Cathedral Mountain Formation, lower of the two, contains one member, the Wedin Member. The most diagnostic fossil to appear in the Cathedral Mountain Formation is Institella, which can be found at, or near, the base of the formation in all parts of the Glass Mountains except from the east end of Lenox Hills to hill 5021 (west of Iron Mountain). There are 47 new genera that appear in the Leonardian and successfully blot out all of the Wolfcampian elements; 35 genera terminate during, or at the end of, the Leonardian (Road Canyon), to be replaced by the Word fauna. CATHEDRAL MOUNTAIN FORMATION.—This formation in the Lenox Hills and Dugout Mountain areas makes a striking color contrast to the underlying Skinner Ranch Formation. It usually begins with orange-colored, somewhat fissile, siliceous beds. The color contrast is striking, but in the eastern part of the mountains, from Hess Ranch east, the siliceous sediments are reduced to minor tongues and are no longer basal. The formation is introduced, however, at most places in the eastern part of the mountains by a small-pebble conglomerate. Where this is lacking, Institella usually can be found to establish the identity of the formation. The Cathedral Mountain contains several limestone bands that originally were numbered from 1 to 5 (P. B. King, 1931). The First Limestone of the Leonard was separated by Cooper and Grant (1964) and made the Sullivan Peak Member of the Skinner Ranch Formation. The Second Limestone later was named "Wedin Member" by Cooper and Grant (1966), but the higher limestones were not named. These limestone beds contain few brachiopods, but, in places, they are fairly rich in ammonites, especially Perrinites. The Wedin Member proved to be the key to the stratigraphy of the Cathedral Mountain Formation in the west end of the mountains. The Second Limestone of the Leonard Formation in the Lenox Hills proved to be faunally identical to the Fifth Limestone of the Dugout Mountain region. These two limestones are richly biohermal and contain Institella and other characteristic fossils in abundance. Wedin Member: The correlation of this member is based on the tracing of a distinctive assemblage of fossils that occurs near the base of the Cathedral Mountain Formation in its western part but that occurs at the base in the eastern part. This fauna contains two brachipods in particular, Institella and Agelesia, that are most distinctive, but, along with these, there are many other genera that make up an easily recognized assemblage. The Institella beds contain numerous small bioherms, each of which usually has its own assemblage of the characteristic fossils. The Wedin Member is known only on the north slope of Dugout Mountain and the west half of the Lenox Hills to about hill 5300. The same fauna, however, occurs from the top of Leonard Mountain to east of Split Tank. New genera that make their appearance in the Wedin Member and its correlate in the eastern part of the mountains are: Agelesia, Amphipella, Anemonaria, Chaeniorhynchus, Choanodus, Dyoros (Lissosia), Dyoros (Tetragonetes), Edriosteges, Grandaurispina, Echinauris, Hercosestria, Hercosia, Heteraria, Holotricharina, Institella, Loxophragmus, Nucleospira, Petasmaia, Ptygmactrum, Rallacosta, Rugatia, Scenesia, Siphonosia, Texarina, Thedusia, Trophisina, Xenosteges, and Xestosia. Some transient elements from the underlying Skinner Ranch Formation, such as Lepidospirifer, and Torynechus, appear in the lowest part of the Cathedral Mountain and Wedin Member. Spyridiophora, Glyptosteges, and Scacchinella have been found as great rarities in the lower Cathedral Mountain Formation (USNM 721u), the latter possibly a pebble reworked from below. No trace of the former two was seen anywhere else in the Cathedral Mountain. Peniculauris is inherited from below, but it is not a conspicuous part of the lower Cathedral Mountain fauna in the western part of the mountains, although it is common in the eastern part. The Echinauris that appears in the Cathedral Mountain Formation differs from that in the Wolfcamp in 116 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY having long halteroid spines on the lateral slopes over the ears. Most of the Cathedral Mountain Formation consists of shale or fissile, siliceous rocks or cherts in the western part of the mountains, where it attains a thickness estimated at 1500 feet. In the eastern part of the mountains, this great thickness of shale is fingered into a section about 300 feet thick, mostly limestone. Naturally the fossils in the western part of the formation differ, by virtue of facies conditions, from those in the eastern part. Nevertheless, the generic assemblage is similar in the two regions. Peniculauris is common in the upper part of the Cathedral Mountain Formation all across the mountains. Rugatia also is frequent in the upper beds, but Institella is extremely rare, usually appearing only in occasional bioherms in the western part of the mountains. The following genera are confined to the Cathedral Mountain Formation and therefore are diagnostic of it: Agelesia, Anemonia, Choanodus, Hercosia, Heteraria, Institella, Loxophragmus, Scenesia, Siphonosia, Trophisina, and Xestosia. Genera terminating in, or at the top of, the Cathedral Mountain are: Agelesia, Anemonaria, Choanodus, Glyptosteges, Hercosia, Heteraria, Institella, Lepidospirifer, Loxophragmus, Nucleospira, Nudauris, Rhipidomella, Scenesia, Siphonosia, Spyridiophora, Torynechus, Trophisina, and Xestosia. Correlation: Outside of the Glass Mountains, the Bone Spring Formation in the Sierra Diablo— exclusive of the thick beds equivalent to the Skinner Ranch Formation at the base but including the Victorio Peak Member at the top—can be correlated with the Cathedral Mountain Formation. Institella is present in collections from the Victorio Peak Member. Also included in this correlation is the Bone Spring along the west side of the Guadalupe Mountains, where Institella also occurs (USGS 7677, 7700, 7722). The Kaibab fauna, as published by McKee (1938), correlates with the upper part of the Cathedral Mountain Formation. Abundant large Peniculauris, Rugatia paraindica, and R. occidentalis are important elements in indicating this correlation. Because of their poor preservation, we have been unable to identify other species illustrated by McKee. Some of these bear names of species that are more characteristic of the Capitan Formation and that need revision when better specimens are found. Identification of Cathedral Mountain correlatives in the shaly formations east of the Glass Mountains in the midcontinent is fraught with the same difficulties as those connected with identification and correlation of Wolfcampian rocks. The Leonardian of the midcontinent and of north-central Texas is based on fusulinids or ammonites, but it contains numerous brachipods of a few kinds: dictyoclostids, linoproductids, chonetids, Derbyia, Meekella, Composita, and a few others, all of which have a Pennsylvanian rather than a Permian aspect. ROAD CANYON FORMATION.—From northeast to southwest, parallel to the Glass Mountains front, this formation exhibits a variety of facies that indicate many special environments. The fauna is highly varied as a result of this great lithic variation. The brachiopod assemblage also is rich and varied, including many unusual species and occurrences of generic types quite unexpected at this level. The fauna is derived clearly from the Cathedral Mountain fauna, but it contains many innovations: Allorhynchus, Ametoria, Bothrostegium, Chonetinetes, Cactosteges, Collumatus, Costispinifera, Echinosteges, Horridonia, Liosotella, Mesolobus, Notothyris, Ombonia, Paucispinifera, Petasmatherus, Holosia, Rhytisia, Simplicarina, Spinarella, Spiriferinaella, Taphrose stria, Undulella, and Yakovlevia. Genera confined to the Road Canyon Formation are: Ametoria, Bothrostegium, Collumatus, Horridonia, Mesolobus, Rhytisia, Simplicarina, Spinarella, and Taphrosestria. Several of the newcomers in the fauna are transient into the overlying Word Formation, and they become important members of that fauna: Allorhynchus, Cactosteges, Costispinifera, Echinosteges, Leurosina, Liosotella, Notothyris, Ombonia, Paucispinifera, Spiriferinaella, Undulella, and Yakovlevia. Genera terminating at the end of the Road Canyon Formation are: Acosarina, Ametoria, Amphipella, Anteridocus, Bothrostegium, Chaeniorhynchus, Chondroma, Chonosteges, Collumatus, Coscinophora, Edriosteges, Elassonia, Goniarina, Hercosestria, Kochiproductus, Kutorginella, Mesolobus, Peniculauris, Petasmaia, Rhytisia, Rugatia, Simplicarina, Spinarella, Taphrosestria, and Tschernyschewia. NUMBER 14 117 One of the astonishing appearances in the Road Canyon Formation is Mesolobus. The specimen has all the features of this unusual chonetid. It is likely, however, that the specimen represents a parallel chonetid stock rather than a stock in the same line as the well-known Pennsylvanian genus. Other unusual occurrences are Geyerella and Ombonia, which is remarkable because they were taken from the same piece of rock. These two genera are similar but not too difficult to separate, because their spondylia are different and their cardinalia quite unlike. The Ombonia is prophetic of the occurrence of this genus higher in the column. It is very rare in the Cherry Canyon Formation, but common in the Capitan and the higher Bell Canyon Formation (Lamar Member). Geyerella is a holdover from the Wolfcampian, but it also is known as a great rarity in the Capitan and Bell Canyon Formations (Lamar). Although nine Word genera appear in the Road Canyon Formation, the fauna is predominantly Leonardian. This is seen not only in the large number of Leonardian species that it contains, but also in the lingering of Perrinites in the formation. Leonard elements of importance are: Peniculauris, Rugatia, Chonosteges, Acosarina, Hercosestria, and Goniarina. Another feature of considerable interest in the Road Canyon is the last appearance of the small bioherms or patch reefs, largely made up of bryozoans, sponges, and brachiopods. We have not found any bioherms in the Word succession. The Word fossil assemblages seem to be death assemblages, whereas, from the Uddenites-bearing Shale Member through the Road Canyon Formation, bioherms and conglomerates are common. The Word can be distinguished from the Leonard not only by widespread conglomerates in the latter, but also by their almost complete absence from the Word Formation. Correlation: The intermediate character of the Road Canyon Formation makes it difficult to correlate. As mentioned above, it is predominantly a Leonard fauna and includes some of the best guides to the Leonard, such as Peniculauris and Rugatia, among the brachiopods, and Perrinites, the ammonite said to be most characteristic of the Leonard. All of the Word or Guadalupe elements in the Road Canyon are rare fossils, some having been found as single specimens after much collecting. Strong elements of the Road Canyon fauna occur high in the Cibolo Formation in the Chinati Mountains in the thin-bedded zone of Udden. The coarse-ribbed Liosotella is very common in the two formations. Wilde (1968), in his discussion of the Cutoff (Shale) Formation in the Apache Mountains and Guadalupe region, concludes that this formation is to be correlated with the Road Canyon Formation. The Cutoff in the Sierra Diablo has been difficult to place, but the occurrence of Perrinites in it has led to is placement in the Leonard Series (P. B. King, 1965:78). This has been contradicted by the presence of fusulinids usually assigned to the Guadalupe Series. The same situation also exists in the Glass Mountains region, in which the fusulinids of the Road Canyon usually are dated as early Word (Wilde, 1968:12). Correlation of the Road Canyon Formation with the Cutoff indicates other correlations as outlined by Wilde (1968). The Cutoff has been correlated and traced by Boyd (1958) into the lower part of the San Andres Formation. The lower part of the San Andres was correlated by R. E. King (1931:26) with the Leonard because of the presence of "Productus ivesi" (= Peniculauris) and "Productus occidentalis" (= Rugatia). Both of these are common in the Cathedral Mountain and occur still higher, in the Road Canyon Formation. They are good indicators of the Leonard Series. So little is known about the faunas of the San Andres Formation, other than its fusulinids, that only the most tentative conclusions can be drawn. Some elements of the Permian fauna of the Coyote Butte Formation in central Oregon suggest correlation with the Road Canyon Formation. In spite of the European and Asiatic types in the fauna, the small "Muirwoodia" (= Yakovlevia) are very similar to those of the Road Canyon. Antiquatonia, Kochiproductus, Martinia, small Waagenoconcha, small "Leptodusf," costellate chonetids, and small alate Stenoscisma all suggest the Leonard. Small Spiriferella is rather suggestive of later affinities, but it also occurs in the Road Canyon. Cooper (1957a: 18) correlated the Oregon (Coyotte Butte) fauna with that of the lower Word of the Glass Mountains, now the Road Canyon; relationship to the Cache Creek fauna of British Columbia also was suggested by Cooper. 118 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY GUADALUPE FAUNAS WORD FORMATION.—The Word Formation in the western part of the mountains consists of a thick silicious shale sequence with several limestone members. The shale contains some fossils, usually Leiorhynchoidea and Crurithyris, but they are infrequent. The bulk of the fauna is confined to the limestones. Cooper and Grant (1964) named the First Limestone of the Word of P. B. King the "Road Canyon Formation" and thereby upset the number system established by King. Later, Cooper and Grant (1966) substituted names for King's numbered limestones. The Second Limestone became the "China Tank Member," the Third Limestone became the "Willis Ranch Member" and the Fourth Limestone was named "Appel Ranch Member." In each of these members the fossils occur as death assemblages, heaps of shells, concentrated in small or large patches, but not in bioherms. China Tank Member: This is the least extensive of the three members, but it is very rich in fossils that have fine preservation. Only two genera appear in this member for the first time: Ectoposia and Pseudodielasma. The member contains a total of 65 genera, most of which also occur in the higher Word members. No genera terminate in the member. Many species, as well as genera, are also transient into the higher Word. The fauna of the lowest part of the Getaway Member (AMNH 600, USNM 732) of the Cherry Canyon Formation strongly resembles it. Willis Ranch Member: The Third Limestone Member of the Word Formation (P. B. King, 1931) has a much wider distribution than the preceding member. It can be traced from the Appel Ranch westward to the west side of Gilliland Canyon, where it disappears except for lenses that can be found high on the slope of Sullivan Peak and on the northeast side of Dugout Mountain. This formation, especially at USNM 706e, contains some of the finest fossils found in the Glass Mountains. Despite the great abundance of fossils, only three newcomers are recorded: Bothronia, Leiorhynchoidea, and Polymorpharia. The first two continue higher; the last occurs also in the Getaway Member of the Cherry Canyon Formation. Four genera terminate in the Willis Ranch Member: Acolosia, Cactosteges, Enteletes, and Undulella. Outside the Glass Mountains the fauna most like that of the Willis Ranch appears in the lower part of the lower Getaway Member of the Cherry Canyon Formation (USNM 732, AMNH 600). Diagnostic genera are: Echinosteges, Liosotella, Grandaurispina, Cyclacantharia, Leurosina, Paucispinifera, Pseudo dielasma, Rhamnaria, Undulella, and Xenosteges. In regard to the termination of Enteletes in the Willis Ranch Member it should be emphasized that this concerns West Texas only. Enteletes has not been identified in the Capitan or its equivalents, but it is reported from highest Permian in Dzhulfa, Armenia, and in the Salt Range of Pakistan. This variation in range and that reported below for Tschernyschewia make intercontinental correlation difficult indeed. Lenses Between Willis Ranch and Appel Ranch Members: USNM 706b is a lens about 200 feet above the Willis Ranch Member that produced an enormous supply of fine fossils of genera also occurring in those two members. The specific composition is quite different, however, and the fauna very distinctive. It is noteworthy for the last appearance, in the Glass Mountains of Cenorhynchia, Costispin.ifera, Paranorella, and Petasmatherus. Although the generic composition of this lens and the superjacent and subjacent members are almost identical, the abundance of the genera is very different. The fauna of the lens, for example, abounds in Stenoscisma, which is very rare in the members above and below. We do not know any fauna outside of the Glass Mountains like that of the lens. Another lens or series of lenses (USNM 737w, 742b) occur about 325 feet above the Willis Ranch Member. These contain species similar to those of the Willis Ranch and prophetic of some in the Appel Ranch Member, but notable for their robust character. The most noteworthy species is a large Pseudoleptodus. Appel Ranch Member: This is the uppermost member of the Word Formation to produce good silicified fossils. Consequently, it is the place of disappearance of a number of genera, but with only two innovations, namely, Bryorhynchus and Divaricosta. The following last appearances are recorded: Bothronia, Cancrinella, Cooperina, Cyclacantharia, Dyoros (Tetragonetes), Eolyttonia, Grandaurispina, NUMBER 14 119 Leurosina, Linoproductus, Neophricadothyris, Rhynchopora, Spiriferinaella, Texarina, Waagenoconcha, and Yakovlevia. It is possible that a species of Cyclacantharia occurs in the Carlsbad Formation in the Guadalupe Mountains and that the genus is not terminal in the Appel Ranch Member. The fauna outside of the Glass Mountains most like that of the Appel Ranch Member is that of the upper part of the lower Getaway Member of the Cherry Canyon Formation. These two members share the following genera: Bothronia, Ctenalosia, Cooperina, Divaricosta, Dyoros (Tetragonetes), Grandaurispina, Leurosina, Reticulariina, Spiriferella, Spiriferinaella, Texarina, Xenosteges, and Yakovlevia. CAPITAN LIMESTONE FORMATION.—The beds above the Appel Ranch Member (the Vidrio Member and the Capitan Formation) generally are dolomitized and do not yield good fossils. Nevertheless, on the northwest side of Dugout Mountain a small fault block near the foot of the Sierra del Norte (USNM 732q) furnished good exposures of fossiliferous Capitan. The rock is not dolomitized at this place. The fossils are not well preserved, but the fusulinid Polydiexodina is fairly common, with a few brachiopods such as Stenoscisma, Pseudoleptodus, Liosotella, large Collemataria, Cartorhium, Elivina, and Echinauris. The assemblage suggests correlation with the Hegler Member of the Bell Canyon Formation. Rocks lying above the Appel Ranch Member cover a great area in the northern part of the Glass Mountains. These have never been searched for fossils except for a locality northeast of Altuda (USNM 718a), where King reports some Capitan species. Search of this vast region for undolomitized parts of the sequence might yield interesting collections. Stratigraphy and Fossils of Other West Texas Areas In order to understand the correlation and fossils of the Glass Mountains with other areas in West Texas, a brief review of the stratigraphy of the Hueco Mountains, the Sierra Diablo, the Guadalupe Mountains, and the Chinati Mountains is given. We have made collections in all of these