The Vale Formation (Lower Permian) Its Vertebrates and Paleoecology Everett C. Olson James G. Mead Texas Memorial Museum Bulletin 29 ©1982 by Texas Memorial Museum, The University of Texas at Austin. All rights reserved Published March 24, 1982 Printed in the United States of America ISSN: 0082-3074-29 The Bulletin is an irregularly published series of technical monographs deriving primarily from research done on Texas Memorial Museum collections and projects. Texas Memorial Museum • The University of Texas at Austin • 2400 Trinity • Austin, TX 78705 Contents Abstract Acknowledgments THE VALE FORMATION The Arroyo-Vale Contact The Vale-Choza Boundary Vale equivalent in Oklahoma THE VALE FORMATION OF KNOX, BAYLOR AND FOARD COUNTIES, TEXAS Geology Fossil remains Vertebrates Plants THE SID McADAMS LOCALITY Geology Site 1 Site 2 Site 3 Site 4 Other Sites Deposilional environment Stream channel deposits Standing-water deposits Systematic paleontology Flora Fauna LOCALITY BLACKWOOD THE Geology Age of deposits Sediments Deposilional environment Taphonomy Systematic paleontology—Fauna THE STAMFORD LOCALITY Geology Occurrence of vertebrates Systematic paleontology—Fauna OTHER SITES GEOGRAPHICAL AND TEMPORAL MODIFICATIONS OF THE VALE FAUNA The Arroyo-Vale boundary and the lower part of the Vale Formation Lower part of the Vale Formation Southern outcrop Northern outcrop Middle and upper parts of the Vale Formation Northern outcrops of Vale Formation Central and southern outcrop regions SUMMARY AND DISCUSSION REFERENCES I 1 2 J 6 6 6 7 9 9 11 11 12 12 13 13 13 13 14 13 13 15 16 27 27 28 28 28 29 30 35 35 36 36 38 39 40 40 40 41 42 42 43 44 45 ILLUSTRATIONS Figure 1. Stratigraphic chart of Wolfcampian, Leonardian and Guadalupian 2. Map of Leonardian Series (Permian), Texas and Oklahoma 3. Detailed maps of areas covered by this study: A. northern area B. central area C. southern area 4. Cross section of Vale Formation in Texas 5. Distribution of Clear Fork Group in Texas 6. Stratigraphic distribution of vertebrate species in upper part of Arroyo and Vale Formations 7. Sketch map of Sid McAdams locality 8. Diplocaulus and dissorophid, Sid McAdams locality 9. Scatter diagrams, skull features of Trimerorhachis 10. Captorhinus aguti from Sid McAdams locality 11. Ophiacodon from Sid McAdams locality 12. Distributions of measurements of Dimetrodon 13. Femur and humerus of Dimetrodon giganhomogenes 14. Sketch map of Blackwood locality 15. “Horns” of specimens of Diplocaulus recurvatus from Blackwood locality 16. Vertebrae tentatively assigned to Lahidosaurikos, Blackwood locality 17. Maxillary teeth ofLabidosaurikos sp., Blackwood locality 18. “Horns” of specimens of Diplocaulus recurvatus from Stamford locality Page 3 4 4 5 5 7 8 10 13 18 20 21 22 23 24 27 31 34 35 36 TABLES Table 1. Faunal list, Vale Formation, Baylor, Knox and Foard Counties 2. Taxa from Sid McAdams locality, Craddock bone bed, northern outcrop of lower Vale Formation 3. Measured section, site 1, Sid McAdams locality 4. Measurements of Gnathorhiza samples 5. Specimens assigned to Trimerorhachis insignis from Sid McAdams locality 6. Measurements of large specimens of Trimerorhachis insignis 7. Dimetrodon giganhomogenes, TMM 30966-356, measurements of vertebrae 8. Dimetrodon giganhomogenes, measurements of femora and humeri 9. Measured section at site 2, Blackwood locality 10. Xenacanthus platypternus, Blackwood locality 11. Diplocaulus recurvatus, Blackwood locality 12. Eryops megacephalus, Blackwood locality 13. Seymouria grandis, Blackwood locality 14. Dimetrodon giganhomogenes, Blackwood locality 15. Diplocaulus recurvatus, measured specimens, Stamford locality 16. Measurements and ratios of skulls of Trimerorhachis Page 11 12 14 17 19 20 24 25 28 30 32 33 34 35 37 38 The Vale Formation (Lower Permian): Its Vertebrates and Paleoecology University Everett C. Olson of California at Los Angeles James G. Mead Smithsonian Institution National Museum of Natural History Washington, D.C. Abstract The Vale Formation is a wedge of predominantly terrestrial sediments of Leonardian age (Permian), overlying the Clear Fork Group and underlying the Choza Formation. Studies of the terrestrial vertebrates and stratigraphy of the Vale Formation, carried on since the late 19305, provide the basis for an initial synthesis of the faunas, stratigraphic distribution, paleoecology, taphonomy, and evolution of the vertebrates. Attention is focused on four principal outcrop areas: the northern Vale outcrop, with a complete section, largely in Knox County, Texas; the Sid McAdams locality, in the lower part of the Vale in southern Taylor County, Texas; the Blackwood locality, middle part of the Vale in central Taylor County, Texas; and the Stamford locality, middle part of the Vale in southern Haskell County, Texas. In addition, numerous other small sites in Texas and localities of equivalent age in Oklahoma were used in the analyses. Except for the northern Vale outcrops, the fossil vertebrate localities have been treated only briefly or not at all in previous publications. A study of fossil occurrences over a north-south distance of about 190 km in Texas, and of sections ranging up to 150 m thick, has provided a basis for tentative conclusions. The primary differences between the Vale and Arroyo faunas are the result of loss of genera and species in the upper part of the Arroyo Formation. New species were introduced early in the north and they, or their derivatives, appeared later to the south. Conditions of deposition pass from paralic in the south to strictly terrestrial with stream action predominant in the central and northern parts of the area. During deposition of the Vale Formation, especially in the north, climates were marked by increasing dryness and seasonality of rainfall. Most vertebrate remains were transported into the areas where they have been found. Only in ephemeral lakes and ponds were the organisms preserved where they lived. There is little evidence of pronounced evolutionary change during deposition of the Vale. Some speciation seems to have taken place, but faunal differences between areas were primarily related to climatological and physiographic differences. The fauna of the Vale does not differ significantly from that of the overlying Choza Formation. The report brings together all available information on the Vale faunas. Firmer interpretations of the evolution of the faunas will become possible when wider chronological and spatial distribution of vertebrates is known and after additional stratigraphic and paleoecological field studies have been completed. Acknowledgments Early collecting in the Vale Formation during the 1930s was carried out by Mr. Raymond Miller and Mr. Edgar Gardener at the Sid McAdams locality in Taylor County, Texas. This was directed by the Bureau of Economic Geology of The University of Texas at Austin and supported by the Works Progress Administration (WPA) of the United States Government. E.H. Sellards and Glen U. Evans directed the scientific and technical aspects of the WPA program. John A. Wilson and Wann Langston, Jr., made the collections from this locality available for study; and Langston supervised the study by James Mead. Mead’s results were presented in a Master of Science thesis, the Department of Geological Sciences (1971), The University of Texas at Austin. Mead’s work is incorporated into the present study. Robert Kier, formerly of the Texas Bureau of Economic Geology, provided consultation on the stratigraphy of the Clear Fork Group. Sergius H. Mamay, United States Geological Survey, provided information on the flora and stratigraphy of the Vale Formation, in particular on the Stamford locality. Studies of the northern Vale outcrops, mostly in Knox County, Texas, were aided by a number of 2 graduate students at The University of Chicago: among these were Nicholas Hotton 111, United States Museum of Natural History; Richard Seltin, Michigan State University; Richard Konizeski, University of Montana; Ernest Lundelius, The University of Texas at Austin; James R. Beerhower, New York State University, Binghamton; Robert Sloan, Llniversity of Minnesota; and the late Ralph G. Johnson and Robert L. Miller, formerly of The University of Chicago. Wann Langston, Jr., John A. Wilson, and Ernest Lundelius, The University of Texas at Austin, reviewed the manuscript and made many helpful suggestions. David Berman, Carnegie Museum, supplied field data and made his collections of fossils available for our studies. Later phases of the field work were supported by Mr. Richard Lassen and Mrs. Lila Olson. Richard Lassen also prepared much of the vertebrate material now housed at The University of California at Los Angeles and aided in the preparation of the illustrations. The figures were prepared by Kathryn Bolles, Museum Scientist, The University of California, Los Angeles. Many organizations and landowners helped witii the work in a variety of ways. In particular among them are the Waggoner Estate, Vernon, Texas; Mr. and Mrs. C.O. Patterson, Lawn, Texas; and Mr. L.N. Blackwood, landowner near Buffalo Gap. Texas. For many years Mr. and Mrs. Wade Barker, of Seymour, Texas, aided us by providing campsites, good water, and unerring advice and friendship. We express our appreciation for financial aid supplied by two research grants from The University of California at Los Angeles, and by several grants from the National Science Foundation. Of specific help in the late phases of the work were National Science Foundation grants DEB 71-01439 and DEB 78-08759. Support under NDEA Title IV made possible the study by James Mead. AMNH CM CNHM FMNH MCZ TMM No. 29 Texas Memorial Museum Bulletin Abbreviations American Museum of Natural History, New York Carnegie Museum of Natural History, Pittsburgh Chicago Natural History Museum. Now Field Museum of Natural History (FMNH). Specimens cited under CNHM are so specified in cited references. Field Museum of Natural History, Chicago Museum of Comparative Zoology, Harvard University, Cambridge T exas Memorial Museum, The University of Texas at Austin UCLA VP University of California, Los Angeles, vertebrate paleontology National Museum of Natural History USNM THE VALE FORMATION The Vale Formation is a wedge of primarily terrestrial strata lying between the subjacent Arroyo Formation and the overlying Choza Formation. These three formations compose the Clear Fork Group of the Lower Permian Leonardian Series (Fig. 1). Vale strata crop out more or less continuously in Texas from Knox, Baylor, and Foard Counties in the north to Tom Green County in the south (Fig. 2). Remains of terrestrial and freshwater vertebrates occur along most of the north-south outcrops, hut they have not been found in the southernmost exposures. Vertebrates were first reported from Taylor County, Texas, by Wilson (1948) and from Knox County, Texas, by Olson (1948). Wilson (1953) described a new paleoniscoid. Lawnia taylorensis, and presented a brief description of a rich vertebrate site known as the Sid McAdams locality. In an unpublished master’s thesis (UT—Austin, 1971) Mead analyzed the geology and fauna of the Sid McAdams locality. With some modifications, his results have been incorporated into this paper. Localities are shown in Figure 3 along with unfossiliferous sites that have been studied. Vertebrates from the Vale Formation in the northern outcrops were described and interpreted in a series of papers by Olson (1951.1952a, 1954a,h, 1955, 1956a,h and 1958). The 1958 paper contains a summary. The Vale fauna was also used to formulate the chronofaunal concept (Olson, 1952b). Recent explorations for vertebrates along the Vale outcrops indicate that these beds and their fossils provide the opportunity to study lateral faunal and ecologic changes for 300 km. Depositional environments range from near shore in the south to upland conditions in the north. Stratigraphic sections vary in thickness from about 15 m in the extreme south to about 150 m in the north (Fig. 4). Throughout the fossiliferous outcrops, strata representing more or less uniform deposition range from 100 rn to 150 m thick. In the northern part of the outcrop, earlier studies of faunal succession were concerned principally with the faunas and faunal changes between sections of limited extent along the depositional strike of the beds. Mead’s study was limited to vertebrates in a stratigraphic section about 15 m thick and over an area of about 1.5 km 2 . Integration of the information from Mead’s work with studies of the northern localities and those from intermediate sites makes it possible to initiate an analysis of contemporary faunal Figure 1.—A general stratigraphic chart of the Lower and lower Upper Permian of Texas and Oklahoma. The Oklahoma Geological Survey places the Permo-Carboniferous boundary at the top of the Oscar Formation. Texas Memorial Museum Bulletin 4 Figure 2.—General distribution of beds of Leonardian Age in Texas and Oklahoma. and ecological differences along the depositional strike of the Vale. It is also possible to determine modifications in time by studying sections at a number of localities along the strike of the beds. The total synthesis will require additional field work to acquire more faunal and floral materials and additional stratigraphic and depositional data. Description of the geology' and vertebrates of the principal localities is the major objective of this paper. Preliminary conclusions concerning the broad aspects of the faunas and ecologies are also presented. Geology The Vale Formation was named by Beede and Waite (1918) to replace an earlier, preoccupied name, Ty e, given by Wrather (1917). The type locality is the long-abandoned Vale Post Office near the BallingerMaverick Road (Texas Stale Highway 158) on Valley Figure 3.—Areas in Texas covered by this study: a. northern area; b. central area; c. southern area. Symbols: AE—exposures of Arroyo Formation visited south of Baylor County, including both those with and those without vertebrate fossils; PgW exposures of the Bullwagon Dolomite, the base of the Choza Formation; Pgp—exposures of the Standpipe Limestone, the top of the Arroyo Formation; VE—unfossiliferous - exposures of the Vale Formation; VF—fossiiiferous exposures of the Vale Formation; VFg—fossiiiferous exposures of the Vale Formation noted by David Berman in field notes which he supplied for this study. Localities indicated by diamond outlines: S—Sid McAdams (southeast Taylor County); B Blackwood (east-central Taylor County); ST—Stamford (south Haskell County). No. 29 Creek in Runnels County, Texas (Sellards, Adkins, Plummer, 1932). As originally defined, the Vale consists of shale, sandy shale and sandstone (Tye of Wrather, 1917). Sellards, Adkins and Plummer (1932) modified the definition to include the Bullwagon Dolomite. The differentiation of the Clear Fork Group into -Arroyo, Vale and Choza formations was not made on the map of the Geological Atlas, Abilene Sheet (Barnes, 1972), but the original definition of the Vale, which excludes the Bullwagon Dolomite, is implied in the legends and this interpretation is followed in the present paper. So defined, the Vale is 15 m thick at its southernmost outcrops where it passes beneath Cretaceous strata in Tom Green County (Fig 5). The Vale is 150 m thick in Knox County. It thickens rapidly north of the type locality to about 120 m in southern Taylor County; the Vale then thickens very gradually to its northern outcrop termination. The Vale contains clay, clay shale, siltstone, sandstone, mudstone, and conglomerate. These occur in a wide variety of patterns of association, but throughout the Vale, finer elastics make up the bulk of the beds. Sandstones, mudstones, and conglomerates occur both as sheet deposits and as linear deposits 1982 The Vale Formation: Vertebrates and Paleoecology with lenticular cross sections. Vertebrates are commonly concentrated where the coarser elastics are an important part of the formation. At these localities and through the entire Vale, the sedimentary composition is highly variable both laterally and vertically. Precise lithostratographic correlation between even closely spaced sections is commonly impossible. Lateral and vertical lithic variability are more highly developed in the Vale Formation than in the underlying Arroyo Formation and the overlying Choza Formation. Locally, however, both the Arroyo and Choza Formations may he lithologically so similar to the Vale Formation that differentiation of the three Clearfork formations on the basis of relatively limited exposures is generally impossible. This similarity has made recognition of the base and top of the Vale difficult in areas where carbonate beds are absent. The The Arroyo-Vale Arrayo-Vale Contact.— Contact.—The Standpipe Limestone member by definition is the uppermost bed of the Arroyo Formation and is overlain by fine clastic beds of the Vale. Where the limestone is present, separation of the Arroyo and Vale is obvious. This marker bed extends from the southernmost exposures of the Vale Formation to the northern part of Abilene, Texas (Figs. 2 and 3). For about 100 km northward from Abilene, the Arroyo and Vale 5 are not separated by the stratigraphic marker. In the northern outcrop area, as explained in some detail hy Olson (1958), the top of the Arroyo is placed at the base of a complex of coarse channel fills which eroded into a persistent red shale about 15 m thick. This channel unconformity can be traced with confidence south to the southern bank of the Brazos River north of Rhineland, Knox County, Texas. From Rhineland south to Stamford and disconlinuously to Abilene, the topographic relief is low and Permian exposures are generally covered by Pleistocene and Recent deposits. Tbe lack of adequate exposures makes it difficult to follow the contact between the red shale and the coarse channel-fill elastics. It is difficult to determine whether it persists south of the Brazos River. A uniform red shale underlying coarse elastics occurs at several sections in this area, but whether this shale represents a persistent bed or is merely local has not been determined. Deposition of the Arroyo and Vale Formations was by rivers, specifically in channels, on pointbars, in overbank and flood-plain environments, and in flood-basin lakes. Consequently, sediments and sediment distribution in the two formations are very similar. In well-exposed outcrops in the northern part of the area, the Arroyo conglomerates are composed predominantly of small, well-rounded hematite pellets in a grit-sized matrix. On the other hand, lower Vale 6 Texas Memorial Museum Bulletin conglomerates are larger, more angular, and lightercolored fragments in a sandy matrix. In upper Vale beds, clay-gall conglomerates predominate. It is not clear whether these differences occur to the south along the outcrop. In any case, it is not certain that the Arroyo-Vale contact in the north corresponds to the contact established to the south on the basis of the Standpipe Limestone. The Kirby Lake Limestone of the .Arroyo Formation, which lies beneath the Standpipe Limestone and is separated hy intervening red elastics, extends somewhat farther north; and the underlying Lytle Limestone is even more extensive. If it is assumed that the thickness between these limestones and the Standpipe Limestone is about the same northward, beyond the Standpipe pinchout, then some general estimate of the position of the Arroyo-Vale contact may be made. At this time, the Arroyo-Vale contact north of Abilene must be inferred (Fig. 2). Extrapolation of the contact from Abilene northward, on the basis of fairly uniform thickness of beds, agrees with the stratigraphic position of the Arroyo-Vale contact defined in Knox County. It is not possible to distinguish Arroyo from Vale along the 110-km outcrop from Abilene to Knox County. Additional study may improve correlation. The Vale-Choza Boundary.— Boundary.—Where present, the Bullwagon Dolomite marks the upper contact of the Vale Formation. The Bullwagon consists of one to several dolomite beds which thicken southward along the outcrop. Because of discontinuous exposures, the exact equivalency of the dolomite beds along the outcrops is difficult to determine. Occasionally, thick calcareous zones marked by intermittent dolomitic siltstones are found just below the dolomite. Whether these are part of the Vale or Choza is uncertain, but it poses no problems since we are concerned about the stratigraphic positions of the underlying vertebratebearing beds. Northernmost exposures of the Bullwagon Dolomite are near Dudley Creek, to the west of a line between the towns of Stamford and Haskell (Fig. 3), and about halfway between them. Northward to this latitude it is possible to separate the Vale and Choza Formations. North of Dudley Creek, exposures of Bullwagon Dolomite are poor and intermittent. Rather consistently, however, a band of evenly bedded, pastel sediments overlies a series of deeper red, more irregular shales. This zone appears to be a northward continuation of the dolomite-shale sequence observed to the south. Traced north, the zone appears to merge with evaporite deposits that have been used to mark the Vale-Choza contact in Knox County (Olson, 1958). This “contact” zone in Knox County is about 15 m thick and the evaporite content of the No. 29 sediment increases rapidly. Within this general interval, it appears that the Vale-Choza contact can be correlated with considerable confidence from the vicinity of Dudley Creek to the outcrop termination of the Clear Fork Group in Texas. Equivalent in Oklahoma.—The Hennessey Vale Vale Equivalent Oklahoma.— Group of Oklahoma consists of widespread red shales, sandstones, and some evaporites which crop out in a belt from near Red River on the south, northward around the Wichita Mountains, and eastward and northward in a broad arc that passes through Oklahoma City and then swings somewhat to the northwest (Fig. 2). Along the outcrop, the Hennessey Group includes beds inferred to he equivalent in age to those of the Clear Fork Group of Texas. Exposures in the southwestern Oklahoma area are approximate time equivalents of beds in the Arroyo Formation in Texas. Near Norman, Oklahoma City, and to the north, the Hennessey Group is underlain by the Garber Sandstone which, on the basis of its scant vertebrates, appears to be equivalent in age to the Hennessey along its southwestern outcrops. Vertebrates have been collected from the Fairmont Shale of the Hennessey Group in the Norman-Oklahoma City area and from a locality about 65 km to the northwest, near Navina. These Fairmont fossils occur within an interval of about 40 m, from about 20 m above the base of the Fairmont Shale. On the basis of the vertebrates, this section is inferred to be equivalent in age to the Choza Formation of Texas (Vaughn, 1958, Olson and Barghusen. 1962). The only known vertebrates from the basal part of the Fairmont shale occur in zones where the Garber sandstones and the Hennessey shales interfinger, in the “transition beds" (Olson, 1967). From north of Oklahoma City and in the vicinity of Crescent, Oklahoma, fossils in this zone are distinct from any that occur higher in the Hennessey Group and have affinities with species from the Vale Formation in Texas. Dimetrodon giganhomogenes, Diplocaulus sp., Seymouria grandis, and Labidosaurikos meachami are present from the basal Fairmont in descending order of abundance. The last two species in particular resemble the Vale fauna. Their presence indicates that the fauna found in the Texas Vale Formation existed northward about 250 km into Oklahoma. Thus, the Vale fauna can he identified over a distance of about 550 km. THE VALE FORMATION OF KNOX, BAYLOR, AND FOARD COUNTIES, TEXAS The geology and faunas of the Vale Formation of Knox, Baylor, and Foard Counties, Texas, were considered in a series of papers (Olson, 1951a, 1952a, 1982 The Vale Formation: Vertebrates and Paleoecology 1954a,b,c, 1955, 1956a,b) and were summarized in 1958 (Olson, 1958). Seltin (1959 and 1972 ms) has reported on his field studies of the northern Vale outcrops conducted subsequent to the 1958 report. He discovered several new sites and collected large suites of vertebrate fossils. These have not been thoroughly studied but preliminary analyses do not appear to alter significantly the general conclusions reached earlier. Geology The base of the Vale Formation in these northas previously noted, was placed at the level of contact of conglomerate channel-fill and sheet-wash deposits above a persistent Arroyo red shale bed about 15 m thick. The upper limit of the Vale was placed in a transition zone of about 15 m between typical Vale clay-gall conglomerates and overlying evaporite deposits characteristic of the Choza Formation. The lithotypes within the Vale Formation so defined in this northern area clearly set it apart from the Arroyo and Choza Formations. The problem of how these boundaries correlate with those in Taylor, Runnels, and Tom Green Counties in the south, where contacts are based on marine carbonate beds, is discussed above. ern counties, 7 In the northern outcrop area, Vale strata are structurally simple with approximate north-south strike and a westerly dip averaging about 9 m per km. Thickness in the area is estimated at about 150 m and outcrops are 16 to 21 km in width. Outcrops studied in the 3 counties extend about 18 km along strike. Similar deposits are exposed on the south side of the Salt Fork of the Brazos River south of the area studied, hut these outcrops have not been studied in detail. Vale deposits in the northern outcrop area consist entirely of red nonmarine clastic deposits. Most of these deposits, about 70 to 80 percent by volume, consist of red clays, shales, and siltstones, which are locally dolomitic. Sandstones, mudstones, and conglomerates make up the remaining 20 to 30 percent of the Formation. The lower third of the Vale Formation is characterized by lenses and sheets ofmoderateto-coarse conglomerate deposited in stream channels and bars and formed on levees and overbank areas adjacent to the channels. Coarsest of channel-fill conglomerates range up to 3 m in thickness and carry fragments of sandstone and siltstone up to 10 cm in diameter; these contrast sharply with the conglomerates encountered in the underlying Arroyo Formation. The finer sediments include clays, shales, siltstones, and fine conglomerates with fragments up to about Figure 4.—Cross section of the Vale Formation from its appearance in Runnels County northward to the Red River (Knox Co.). Sites and stratigraphic extent of southern and central fossil localities indicated by X. Important collecting localities; S—Sid McAdams; B—Blackwood; ST—Stamford. Figure 5.—Distribution of the Clear Fork Formations in Texas. 1982 The Vale Formation: Vertebrates and Paleoecology I cm in diameter. These are similar to typical deposits of the Arroyo Formation. In the middle portion of the Vale, the coarse, light-colored conglomerates are replaced gradually by red brown clay-gall deposits with pebbles derived directly from the local deposits into which the conglomerates are incised. These coarse sediments acted as homogeneous bodies, and mud cracks formed iu many of them during desiccation. This type of claygall conglomerate persists to the top of the Vale hut is rarely found in Choza deposits. Irregular, small lake and pond deposits occur throughout the Vale. Their sediments are uniformly fine-grained and predominantly clay shales. As a rule, hut not invariably, the lake and pond deposits are lighter in color than are the surrounding beds, as a result of reduction of the pervasive iron oxides in the presence of high concentrations of plant remains. The marginal beds of standing water deposits tend to dip toward the centers of the basins. Flood-plain deposits constitute the hulk of those formed in this area. They consist largely of red clays and siltstones, interrupted in places by thin hands of light-colored, usually slightly coarser materials marking periods of greater water velocity during flood splays of the rivers. In places, the red flood-plain deposits show remnants of the original bedding but, by and large, the layering has been destroyed by compaction. Fossil Remains Vertebrates, plants, and some invertebrates occur throughout the Vale sections in the northern counties. Fossils have come from all types of deposits hut are most frequent in the channel conglomerates and least frequent in the fine-grained flood-plain deposits. In the upper part of the section, vertebrates are confined largely to clay-gall conglomerates. Plant remains, usually poorly preserved, occur in the sandstones and green to gray clay shales. Carbonaceous detritus occurs in all types of deposits. Invertebrates are not well preserved and consist mostly of casts of small pelecypods, occasional concentrations of tests of chonchostracans and indistinct tracks and trails. Vertebrates.—A list of genera and species from Vertebrates.— the Vale Formation of Knox, Baylor, and Foard Counties is given in Table 1 and the distribution of the genera and species important in this study is shown in Figure 6. The following points summarize and augment the table and figure. 1. Nine genera and species that are persistent throughout the Arroyo continue into the Vale. Of th ese, Gnathorhiza serrata has not been found in the lower Vale of this area, hut its presence thereafter 9 suggests that this may he due to sampling error. Seven of the species persist through the Vale. One of these, Captorhinus aguti, has not been found in the middle Vale. It is not abundant at any level but this absence is probably due also to sampling error. As in the Arroyo Formation Dimetrodon giganhomogenes (D. gigashomogenes, Romer and Price, 1940) is the most abundant terrestrial animal and Xenacanthus, Lysorophus, and Diplocaulus are abundant in lake and pond deposits. 2. Two of the nine species that persist from the Arroyo Formation into the Vale have not been found above the lowest beds of the latter: Diplocaulus magnicornis and Diadectes tenuitectes. D. tenuitectes is a form species of the Arroyo. Perhaps the specimen from the lowest Vale—a single vertebra—pertains to this species, hut it is listed merely as Diadectes sp. in the table. The absence of Diadectes from all higher beds in this region probably has resulted from its failure to survive iu the area, for Rs remains are large and easily recognized from fragments. 3. Two species, Gnathorhiza dikeloda andLabidosaurus barkeri, appear for the first time in basal Vale beds. L. barkeri was synonymized with L. meachami from the Fairmont shale of Oklahoma by Seltin (1959). L. barkeri is retained here on the basis of its fewer rows of teeth and smaller size, as compared to L. meachami, but with the recognition that these may be ontogenetic differences and that the two may in fact be synonymous. Slightly higher in the Vale, Diplocaulus recurvatus appears in stream deposits. Its predecessor and probable ancestor/), magnicornis lived primarily in standing water and is apparently absent from all beds deposited soon after the appearance oiD. recurvatus. 4. Edaphosauruspogonia and Labidosaurus harnatus, common in Arroyo beds, have not been found in the northern Vale area and presumably did not exist after deposition of the Arroyo in this region. 5. Cacops aspidophorus, Casea broilii, and Varanops brevirostris are entered in the faunal list. They all are from a single locality, the Cacops bonebed, which is either uppermost Arroyo or lowermost Vale, depending upon where the boundary is placed. These species will not enter into the discussion because, as indicated elsewhere (Olson, 1971) and discussed on p. 43, they pertain to a chronofauna not commonly sampled in the areas under consideration. Seymouria baylorensis was associated with these species hut also is present at many Arroyo sites. 6. Captorhinikos valensis appears in the middle Vale and thereafter is fairly common. It was probably derived from Captorhinus aguti, but transitional forms are not known. This genus is prominent in the Figure 6.— Stratigraphic distribution of Vale species in the upper Arroyo and Vale Formations. Species from Cacops bone bed (located north of the western end of Lake Kemp) are not entered. These include Cacops aspidophorus, Varanops brevirostris, and Casea broilii. The age of the bone bed is either late Arroyo or early Vale, The other genera are part of a separate chronofauna and erratics in the usual Clear Fork vertebrate assemblages. Also it should be noted that Peronedon occurs in beds equivalent to the Arroyo in age in Oklahoma. 1982 The Vale Formation: Vertebrates and Paleoecology Choza Formation and occurs in the Fairmont Shale of the Hennessey Group of Oklahoma (Vaughn, 1958; Olson and Barghusen, 1962, pt. 11; Olson 1970). 7. A number of species in the faunal list (Table 1) and in Figure 6 are known only from one or a few specimens. Among these are Trematopsis seltini, Waggoneria texensis, Captorhinoides valensis, Peronedon primus, and Casea nicholsi. The first three give little information about the Yale fauna as a whole but could, as collections from other areas are made, become important in comparisons of the northern and southern Vale deposits. Two specimens of Casea nicholsi, known from a pair of associated skeletons, appear to represent “erratics” from the second chronofauna noted in item 5 above. Indeterminate materials include a toothed palate resembling Rothianiscus from tbe lower part of the Pease River-El Reno beds of the Guadalupian Series. Plants. — Plants.—Read and Mamay (1964) place a small llora from the upper Vale of Knox County, Texas, in their Zone 15, the zone of the younger Gigantopteris llora. The sparse llora includes Gigantopteris (species B of Read and Mamay) and is dominated numerically by walchian conifers. Read and Mamay also identified fronds that are very close to Pterophyllum Brogniarl. This genus, they note, gives a somewhat Mesozoic cast to the flora. Throughout the Vale of the northern counties Walchia and Gigantopteris are predominant, and remains of other plants are uncommon. Preservation is generally poor except at a few sites, and comparisons with floras from elsewhere are difficult. THE SID MCADAMS LOCALITY The Sid McAdams locality lies on the Patterson Ranch in southern Taylor County, Texas (Fig. 3c). It was first worked in August 1939 by a crew from the Statewide Paleontological and Mineralogical Survey of the Bureau of Economic Geology, The University of Texas at Austin (supported by the Works Progress Administration). Raymond Miller and Edgar Gardener directed field activities until July 1940. During this period collecting was supervised from Austin, Texas, by Dr. E.H. Sellardsand Glen L. Evans. The materials are housed at the Vertebrate Paleontology Laboratory of the Texas Memorial Museum, The University of Texas at Austin. All specimens are designated by the locality number TMM 30966 followed by an identifying number. Since the closure of the original quarry, fossils have been collected from the area by Dr. John A. Wilson, The University of Texas at Austin, Dr. Sergius H. Mamay, the United States Geological Survey, and 11 Table 1. Vertebrate faunal list—Yale Formation in Baylor, Knox, and Foard Counties, Texas. Chondrichthyes Xenacanthus cf. platypternus (Cope) Dipnoi Gnathorhiza serrata Cope Gnathorhiza dikeloda Olson Amphibia Lysorophus tricarinatus Cope Diplocaulus magnicornis Cope Diplocaulus recurvatus Olson Peronedon primus Olson Eryops megacephalus Cope Primerorhachis insignis Cope *Cacops asipodophorus Williston Trematopsis seltini Olson Waggoneria texensis Olson Diadectes sp. Cope Seymouria baylorensis Broili Keptilia Captorhinus aguti Cope Captorhinoides valensis Olson Captorhinikos valensis Olson Labidosaurikos meachami Stovall Dimetrodon giganhomogenes Case *Casea broilii Williston Casea nicholsi Olson *Varanops brevirostris (Williston) cf. Rothianiscus sp. *These species occur in the Cacops bone bed, which is of uncertain stratigraphic position, either uppermost Arroyo or lowermost Vale Formation. Seymouria baylorensis has been recorded from the Cacops bone bed (Williston, 1911), but the identification is questionable. by the writers of this paper. Previous publications on the geology, faunas, and flora are Wilson (1953) devoted primarily to the paleoniscoid fishes; Brooks (1962) on a new crustacean; Read and Mamay (1964), Mamay (1976) on the flora, and Olson (1979a) on a specimen of Trimerorhachis. The flora and fauna were collected throughout a stratigraphic interval of about 5 m over an area that measures about 500 m east-west and 400 m northsouth. Four sites, numbered 1 to 4 in Figure 7, have produced the bulk of the fossil material. The greatest number of specimens has come from sites 1 and 2. Other specimens have been found widely scattered over the rest of the area, on the flanks of the hill including sites 1 and 2, on the valley floor, and along the north wall of the valley. Vertebrates from these sites are listed in Table 2. Geology The vertebrate-bearing deposits are near the base of the Vale Formation (Wilson, 1953). The quarry at site 1 lies about 7 m above the Standpipe Limestone, Texas Memorial Museum Bulletin 12 No. 29 Table 2. Frequency of taxa from the SidMcAdams locality compared with frequencies atthe lowerVale localities in the northern area (Olson, 1958) and at the Craddock bone bed (Arroyo Formation) (Romer, 1928;Williston, 1911). Taxon Sid McAdams Northern lower Vale localities locality Xenacanthus platypternus Lawnia taylorensis Gnathorhiza serrata Gnathorhiza dikeloda Diplocaulus magnicornis Diplocaulus recurvatus Lysorophus tricarinatus Trimerorhachis insigriis Eryops megacephalus tTersomius sp. Trematops milleri Trematopsis seltini Seymouria baylorensis Diadectes sp. Captorhinus aguti Labidosaurikos barkeri Dimetrodon giganhomogenes fOphiacodon sp. Varanosaurus acutirostris Araeoscelis gracilis rare common common common Craddock bone bed common present ?common mod. common not recorded not recorded not recorded common rare not recorded not recorded common nearby mod. common not recorded not recorded mod common rare common common common common common not recorded not recorded rare not recorded* rare rare absent present not recorded common rare not recorded rare common common not recorded not recorded not recorded not recorded present absent not recorded present present ?present present not recorded present present common** not recorded present present *This species has been recorded from the Cacops bone bed but its identification is questionable (see text p. 40 and Figure 6). It is not found elsewhere in the lower Yale of the northern region, Other species from the Craddock bone bed are not entered in the table. **This is based on the discussion of the species of Dimetrodon on p. 26, where it is concluded that the Craddock species is D. gigashomogenes. the top bed of the subjacent Arroyo Formation (Table 3). Although lithosomes usually cannot be traced even into adjacent gullies, the sequence of beds at site I gives a fairly good representation of the stratigraphy throughout the area. The deposits consist of a complex of predominantly red brown clays, siltstones, sandstones, and conglomerates cemented by granular hematite. Bed number 2 in the measured section is light gray and the conglomerates range through brown and yellow to gray. Local reduction resulting from entrapped organic matter occurs at various places in the clays, siltstones, and sandstones. The siltstones and sandstones of beds 5 to 10 vary laterally. Conglomerates occur at several levels, mostly below bed 5. They consist of 20 to 40 percent poikiloblastic calcite cement surrounding well-rounded to angular fragments of siltstone, claystone, and carbonate clasts. Some of the claystone fragments appear to have suffered postdepositional deformation. The conglomerate deposits are lenticular in cross section and range from 3 or 4 m to as much as 100 m in width. Pebbles range from about 2 to 10 mm in maximum dimensions. The thickness of the lenses varies from about 10 to 50 cm. Site 1.— 1.—This site has yielded most of the fossils the Sid McAdams locality. Both plants and vertebrates were found in the gray, silty clay of bed 2 in the measured section. According to Wilson (1953) this layer contains about 60 to 70 percent detrital quartz with a mixture of halloysite and glauconite. Some vertebrates have come from bed 3; and above is a poorly indurated conglomerate carrying occasional and fragmentary fossils. Fragmentary remains of Trimerorhachis and a partial specimen of the paleoniscoid fish Lawnia taylorensis are from bed 8. Site 2.— 2.—This site lies about 100 m west of site 1. At approximately the level of bed 4, site 1, is an extensive complex of red brown and yellow brown conglomerate. Overlying this complex is a sequence of siltstones that together are approximately equivalent in position to beds 5 through 8 of the measured section at sitel. The siltstone sequence is about 4 m in thickness and is overlain by fine sandstones that continue to the top of the hill. The conglomerates contain a wide variety of vertebrate fragments and, occasionally, fragments of highly carbonized plants. The detrital clasts of the conglomerate are for the most part randomly distributed through the lenses with respect to size. In places, at 1982 13 The Vale Formation: Vertebrates and Paleoecology Figure 7.—Sketch of the ground plan of the Sid McAdams locality showing the principal collecting sites, 1-4. C—cattle tank; T—site for partial skull of Trimerorhachis. however, a coarse-to-fine gradation from the base to the top of individual layers is well developed. The overlying 4 m of siltstones and silty, finegrained sandstones is generally homogeneous hut in places carries small lenses of fine-to-medium-grained sandstone. The sediments are marked by irregular gray green reduction patches which appear to be the result of reduction of the hematitic content of the sediments by products of burrowing invertebrates (Olson, 1979a). Fossils are rare in these beds but several specimens of Trimerorhachis have been found both at site 2 and in the area of site 1. The overlying fine sandstones have not produced any fossils. Site 3.— 3.—This is a small lens of gray, clay-rich conglomerate about 200 m north of site 2 which has yielded a partial skull of Diplocaulus and a few large coprolites. Site 4.— 4.—This site lies at the west end of the locality and includes a red brown sandstone which is somewhat coarser than the other sandstones in the area. Mud cracks and ripple marks are distinguishing features, and fragmentary vertebrate remains occur. Burrows, probably those of lungfish, occur in this sandstone. Other sites.— No fossil concentrations have been found elsewhere on the Patterson Ranch. Coprolites the valley floor, which has been much disturbed by deep plowing. Fragments of bone occur on the hillsides but they are scarce and scattered. A fragment of skull and jaws of Trimerorhachis was found southeast of the cattle tank (C) at the point marked “T” in Figure 7. occur on Depositional Environment The marine Standpipe Limestone is exposed about 2.5 km to the northeast of the quarry at site 1. As noted, the lowest vertebrate-producing beds lie about 7 m above the limestone (Wilson, 1953). Bedrock between the limestone and the quarry is not well exposed but is predominantly red clay and siltstone. The rock throughout the area and especially at site 1 may represent both marine and nonmarine deposition hut, except where fossils are present or the sedimentary structures provide clues, differentiation is difficult. The presence of Mamayocaris (Brook, 1962), a crustacean similar to those found elsewhere in marine facies, may indicate the periodic incursions ofmarine or brackish-water environments. The proximity of the vertebrate-bearing beds to marine deposits at the Sid McAdams locality contrasts sharply with the conditions that existed during the deposition of the lower Vale strata to the north No. 29 Texas Memorial Museum Bulletin 14 Table 3. Measured section site 1, Sid McAdams locality. Bed 10 9 8 7 6 (meters) (meters) 0.16 m capping hill, contact at base covered red-brown fine sandstone, cross-bedded, ripple-marked. possible mud cracks, gradational at base purple-brown clay and silt, many small blue-gray areas (local reduction), gradational at base 0.16 5.30 0.24 5.14 3.05 4.90 blue-gray siltstone with small clay lenses, gradational at base mottled red-brown-yellow clay and siltstone, 0.22 1.85 0.46 0.46 1.63 1.17 0.16 0.71 0.24 0.55 0.31 0.31 0.00 0.00 5 4 blue-gray siltstone, gradational at base 3 mottled brown-yellow clay, becoming more silty near top, few vertebrate remains, gradational at base. . . . blue-gray silty clay with abundant plant and vertebrate remains, sharp contact at base 1 Cumulative Thickness blue-gray fine sandstone, cross-bedded, top-eroded, gradational at base 2 Bed Thickness poorly indurated limestone-clay pebble conglomerate with clay matrix, gradational at base. . mottled gray and yellow lime stone-clay pebble conglomerate forming quarry floor, few vertebrate remains, base not seen of Taylor County and conditions that prevailed during deposition of the middle and upper Vale throughout most of their outcrop areas. At the very top of the Vale, in the southern area, thin calcareous beds just below the Bullwagon Dolomite indicate a return to marine or at least brackishwater conditions. The nonmarine deposits of the Lower Permian Wichita and Clear Fork Groups are usually referred to as deltaic (see for example Homer, 1958; Olson, 1958). Rayner (1971) and Parrish (1977) among others have questioned this designation. The Wichita and southern part of the lower Clear Fork deposits characteristically consist of sequences of terrestrial clays, sillslones and shales that alternate with marine limestone or dolomites. No major drainage systems have been identified and, although portions of the deposits are somewhat similar to those of modern deltas (see Lewis and Vaughn, 1965), the use of the term with reference to the whole complex is questionable. Most of the terrestrial deposits were formed on Hood plains, often cut by small streams, and in places included bodies of standing water. For those parts of the system in which an alternation of marine and nonmarine conditions can be identified the term paralic is more appropriate than deltaic. In their northern exposures the Vale deposits, below the transitional beds leading to the Choza, were formed farther from the sea margin and under moderately upland conditions not far to the south of the Wichita prominence. The term deltaic is reasonably applicable to ibis portion, in contrast to the southern exposures, such as those at the Sid McAdams locality, which conform to the general concept of paralic. The vertebrate-bearing beds at the Sid McAdams locality appear to have been formed largely or wholly under nonmarine conditions. Stream-channel and standing-water deposits are readily identifiable. In addition, some of the sandstones appear to represent overbank deposits and, along with much of the nonfossiliferous sediments of the area, fall into the general category of flood-plain deposits. As noted, there may be some brackish or marine interfingering with the last type, but clearly identifiable tidal-flat accumulations sucb as those described by Parrish (1977) have not been confirmed. Stream-channel deposits.— deposits.—The lenticular beds of sandstones and conglomerates at the Sid McAdams locality were formed in stream channels and are restricted to the lower part of the section. The channels appear to have had a roughly east-west trend, hut this may have been oidy local in view of the limited outcrops. Some of the delrital materials consist of locally derived clay which was soft at the lime of deposition. Fragments probably were torn from the hanks of the stream and deposited rapidly during flood stages. At least some of the calcareous pebbles appear to have been derived from caliche, and some of the hematitic concretions suggest derivation from lateritic soils. Graded bedding in places suggests that 1982 15 The Vale Formation: Vertebrates and Paleoecology some of the conglomerates represent a single depositional event. A picture emerges of rapid erosion and deposition over short periods of time, indicating reasonably strong seasonality during the time of deposition of the channel deposits. Some of the sandstones were formed during waning stages of channel deposition. In places they also appear to represent overhank deposits. Mud cracks and current ripples occur both in the sandstones and siltstones. Burrows, probably those of lungfishes, have been found at one site (see p. 27). These various lines of evidence point to a moderate seasonality, at least as far as rainfall is concerned, throughout the deposition of the lower part of the Vale Formation in the general area of the Sid McAdams locality. Standing-water deposits.—Criteria are less precise for standing-water deposits than for those formed by stream action. Type of sediment, dipping marginal beds, and extensive reduction of iron by contained organic matter are important (Olson, 1962). Two types of standing-water deposits have been identified, represented by the blue gray silly clay at site l(bed 2), the red brown to purple brown clays and siltstones as in beds 7 and 8 at site 1, and the equivalents of beds 5 through 8 of this site in the area of site 2. Below beds 7 and 8 of site 1 are mixed red, yellow and brown clays with some silt and an intervening conglomerate series. These appear to be of mixed origin. Conglomerate also underlies the predominantly lacustrine beds at site 2. Overlying the lake beds in both areas are sandy facies, beds 9 and 10 at site 1, which appear to have been laid down by more actively flowing waters, probably on floodplains rather remote from the controlling stream channels. The blue gray silty clay, bed 2 at site 1, is a distinctive local feature. It contained a large and concentrated fauna and flora. Remains of Dimetrodon representing at least 22 individuals predominate; but these are accompanied by 16 specimens of the paleoniscoid Laivnia, a partial humerus of an ophiacodont, a femur and tibia of Seyrnouria, and a cranial spine of Xenacanthus. Fragments that may be from amphibians also are present. A matured aquatic fauna does not appear to have been present. Rather, this deposit appears to represent a temporary “carnivore trap somewhat similar to the Craddock hone bed in tbe Arroyo Formation of Baylor County, Texas (Williston, 1911). Comparisons of the faunas are given in Table 2. The remains of Dimetrodon are unabraded and seem to have accumulated rapidly in place. Most of the bones were disarticulated, although one complete skeleton was recovered. The lack of association may be owing partly to poor collecting methods. No structure or bedding occurs in the clay, indicating probable formation in an isolated body of water which was relatively short-lived, perhaps a lew years. Temporary ponds of this general nature also are found in the northern Vale deposits (Olson, 1958). The siltstones of beds 7 and 8 at site 1 and equivalent beds elsewhere in the area, in contrast to the short-lived pond, appear to represent standing-water deposits formed over a period of at least several years and probably several decades (Olson, 19 1 9 a). Sediments are relatively uniform throughout, consistently fine-grained, and show little or no evidence of current action. They are marked by light-colored reduced zones, very irregular in shape and no more than 2 or 3 cm in maximum dimension. These zones appear to have resulted from reduction of iron oxide by the decay products of invertebrates that burrowed in the muds of the lake bottom (Olson, 1979a). No bedding is evident in the siltstones, and it is probable that the original bedding was destroyed by bioturbation and by compaction. The lake (there may have been more than one) was obviously extensive, but its limits are not discernible from the outcrop; the definitive feature of dipping marginal beds has not been found. The other physical characteristics and the presence of vertebrates Trimerorhachis and Lawnia lead to the conclusion that the deposits were formed in the large and persistent lake. There is no evidence of desiccation in the lake deposits and whatever seasonality may have existed is not reflected in the sediments. Perhaps the seasonality evident in the lower part of the section had modified or perhaps the lakes were large enough that they did not become seriously restricted during the dry seasons. Systematic Paleontology Flora.— moderately large collection of plants was Flora.—A obtained from the Sid McAdams locality during excavation of the quarry at site 1 and from a gray shale lens about 900 m north of the quarry. The first identifications were made by C.A. Arnold, The University of Michigan (see Wilson, 1953). Materials from the lens north of the quarry were obtained in 1955 and 1956 by S. Mamay, E. Yochelson, and C. Read (Mamay, 1976). Soon thereafter, deep plowing of the valley floor essentially destroyed the site. The flora was described in a preliminary fashion Read and Mamay and in more detail by Mamay by (1976). In the second publication he gives the following list of taxa: Gigantopteris (undesc. species, sp. B of Read and Mamay, 1964), Cycadospadix yochelsoni, Annularia, Aphlebia, Taeniopteris. Sphenophyllum.Pecopteris, Gomphostrobus, Ualchia, Callipterus, ' Odontopteris. In addition some unidentified seeds, some odd. hooked appendages, possible liverworts, and a new genus of conifer were noted as present. The flora was assigned to Zone 15. the highest of the Upper Paleozoic Floral Zones of Read and Mamay (1964). This zone is characterized as the Younger Gigantopteris Zone. The flora differs only in a few insignificant details from that of Zone 14 but contains some distinctive elements, especially from some slightly later Vale sites noted in subsequent sections of this paper. Fauna.— Fauna.—ln addition to the vertebrates, the Sid McAdams locality has yielded a few invertebrate fossils. Mollusca Class Pelecypoda Order Prionodesmacea Family Anthracosiidae Walaeanodonta sp. Wilson (1953) recovered a number of pelecypod remains (TMM 30966-413) from a conglomerate near site 1. These were sent to Dr. N.D. Newell of the American Museum of Natural History who made the above identification and stated that they were freshwater forms. Arthropoda Class Crustacea Order Pygocephalomorpha Family Pygocephalidae Mamayocaris jepseni Brooks 1962 An excellently preserved individual of this species was collected from site 1 of the Sid McAdams locality by Mamay in 1956. ILK. Brooks described it in a general treatment of the North American Paleozoic Eumalacostraca (Brooks, 1962). The specimen (USNM 133393) is the holotype and the only example of the species from Texas. A large number of referred specimens was found in a lagoonal limestone in the Opeche Formation (Permian) near Rapid City, South Dakota. Brooks (1962) has suggested that this species was possibly No. 29 Texas Memorial Museum Bulletin 16 euryhaline. Vertebrata Class Chondrichthyes Order Pleurocanthodii Family Xenacanlhidae Xenacanthus platypternus (Cope) 1884 In contrast with many Vale localities the Sid McAdams locality has yielded few remains of Xenacanthus. Specimens include: TMM 30966-418, part of a cranial spine from site 1; and UCLA VP 475, 3 teethfrom siltstone at site 2. The spine is entirely similar to those of Xenacanthus (Pleurae an thus) figured by Hussakof (1911) and found throughout much of the Lower Permian. It is not specifically determinable. The teeth have thin, plate-like bases, longer than wide, and the basal tubercle is well defined but small. The central cuspule is single, long, and slender and the lateral cusps are compressed and denticulated or crenulated. All of these characters are the primary features of teeth of X. platypternus as specified by Hotton (1952). Class Osteichthyes Order Palaeonisciformes Family Amblypteridae Lawnia taylorensis Wilson 1953 This genus was described by Wilson (1953) on the basis of several excellently preserved specimens from site 1 of the Sid McAdams locality. Except for scales in the conglomerates, only one additional specimen (UCLA VP 465) has been found since that time. It consists of a skull, lower jaw, and about threefifths of the trunk. It adds nothing to the very full descriptions made by Wilson. The new specimen comes from the siltstone of site 2. Order Dipnoi Family Lepidosirenidae Gnathorhiza dikeloda Olson 1951 This species was defined (Olson, 1951a) on the basis of the proportions of the anterior and posterior blades of the teeth which differ from the proportions of these structures in G. serrata known both from the Arroyo and Vale Formations. In the lower teeth the ratio of the anterior segment of the inner blade to the posterior segment is less than 2 in G. .serrata and greater than 2 in G. dikeloda (Table 4). On the basis of this criterion,, two of the lower teeth from the conglomerate of site 2 at the Sid McAdams locality, TMM 30966-422, -423, are assigned to G. dikeloda. These teeth are much smaller and have fewer denticles on the blades than do the specimens upon which the species was erected. The differences probably are a matter of individual age, with the Sid McAdams specimens belonging to juveniles. Gnathorhiza serrata Cope 1883 Among the small teeth of Gnathorhiza from site 2 are two lowers, TMM 30966-425, -430, which pertain to G. serrata on the basis of the ratios of the anterior and posterior segments of the blade. One measurable upper tooth, TMM 30966-428, also appears to belong to this species. One other lower tooth, TMM 30966-426, which could not be measured accurately, may also belong to this species. Except for TMM 30966-425, all of the teeth are considerably smaller than those used for specific comparisons by Olson (1951a). 1982 17 The Vale Formation: Vertebrates and Paleoecology Table 4. Measurements of Gnathorhiza teeth. Measurements were made along the occlusal surface of the teeth, using the intersections of the enamel ridges as reference points. Abbreviations are as follows; CNHM—Chicago Natural History Museum (Field Museum); AMNll—American Museum of Natural History; TMM—Texas Memorial Museum. Data for AMNH and CNHM from Olson, 1951a. Measurements in mm. Length of Specimen Total number length TMM 30966-422 TMM 30966-423 CNHM-UF.90 CNHM-UF.94 CNHM-UF.93 CNHM-UF.92 CNHM-UF.91 TMM 30966-425 TMM 30966-420 TMM 30966-430 AMNH 7258 AMNH 8014 TMM 30966-427 TMM 30966-424 Length of posterior segment Length of Ratio radiating ridge ant/post Gnathorhiza dikeloda lower teeth 6.20 4.65 28.00 27.00 30.00 — 30.50 8.59 5.20 est 1.85 est 10.80 12.20 TMM 30966-429 TMM 30966-428 anterior segment 1.95 — 5.05 3.78 21.50 20.00 24.00 1.60 1.33 9.00 9.00 10.00 8.80 25.20 10.20 Gnathorhiza serrata lowe■r teeth 6.26 3.59 3.16 2.85 2.40 2.20 2.40 2.35 - 5.20 5.80 — — segments 5.20 6.00 2.40 3.34 1.74 — — — - — 1.19 6.20 8.20 0.75 est 4.70 5.30 Gnathorhiza sp. lower teeth 0.68 Gnathorhiza cf. G. serrata upper teeth 1.09 0.79 Gnathorhiza sp. upper teeth 2.88 2.29 _ Gnathorhiza sp. Several specimens in the Texas Memorial Museum collection, TMM 30966-424,-427,-429, and unnumbered fragmentary specimens in the UCLA VP collections are so broken that specific determination is impossible. All are from site 2. Some of these are entered into Table 4 for size comparison. Subclass Lepospondyli Order Nectridea Family Keraterpetontidae Diplocaulus cf. D. magnicornis Cope 1882 A partial left “horn,” TMM 30966-419 (fig. 8) collected from a gray limestone-clay/pebble conglomerate about 200 m northwest of the APA quarry, at site 3, is the only specimen for which assignment to Diplocaulus can definitely be made. Three small jaw fragments, TMM 30966-434, -435, -436, were found in the residue of acidized conglomerate from site 2. The cross sections of the teeth in these have a definite subquadrate shape, suggesting that they belong to Diplocaulus. Olson (1951b) subjected an Arroyo sample of Diplocaulus to statistical analysis and concluded that two species were present, D. magnicornis Cope 1.60 est 1.30 1.50 — — - 0.81 ant post 0.89 0.73 2.31 2.05 1.38 _ and D. hrevirostris Olson. These species can be separated by comparison of the midline length of the skull with the orbito-snout length, premaxillary length, and horn length, and by the relative convexity of the parietal bone. None of these features is preserved in TMM 30966-419. The shape of the horn, however, resembles that of D. magnicornis. Olson (1952b) compared a sample of Diplocaulus from the Yale with samples of the two Arroyo species. It was found that the members of the Vale sample resembled D. magnicornis in most characters, hut that there was a difference in the number of individuals in the Arroyo and Vale populations that had the tip of the “horn” recurved. The name D. recurvatus Olson was proposed for the Vale sample. Because D. magnicornis and D. recurvatus are distinguished by the relative frequency of this character in the populations, a single specimen cannot he referred definitely to one or the other. This applies, of course, to the specimen from the Sid McAdams locality. Because of the resemblance of the “horn” to that of the majority of skulls of D. magnicornis and the absence of this species from the lowest Vale elsewhere, the specimen is more likely to belong to 1). magnicornis than to D. recurvatus. Texas Memorial Museum Bulletin 18 No. 29 15 more. A second, TMM 30966-421, is 42.5 mm long and carries 38 teeth plus gaps for about 20 more. UCLA VP 476 is 56 mm long and carries 24 teeth plus spaces for about 30 more, including covered areas. Figure 8.—Above, Diplocaulus cf. D. magnicornis, TMM 30966-419. part of skull from the Sid McAdams locality; below, a dissorophid lower jaw, TMM 30966-6, probably representing Tersomius, from the Sid McAdams locality. Subclass Labyrinthodonlia Order Temnospondyli Family Dissorophidae ?Tersomius sp. The genus Tersomius (Case, 1910) has proven be fairly widespread in Lower Permian deposits. A number of partially to nearly complete mandibles and maxillae with dentitions have come from the conglomerates at site 2 (Fig. 8). The known specimens are as follows: TMM 30966-6, a nearly complete dentary; -179, a partial maxilla; -421, a nearly complete dentary; and UCLA VP 576, a nearly complete dentary; 578, a fragment of lower jaw; 581, a poorly preserved partial dentary; 592, a partial maxilla; 604, a partial maxilla, small individual. All specimens have slender, conical, acrodont, labyrinthine teeth. The crowns are from 2.0 to 2.5 mm long. TMM 30966-6, -421, and UCLA VP 576 indicate that about 60 teeth were present in complete series. One nearly complete dentary, TMM 30966-6, is 47.5 mm long and has 41 teeth and spaces for about to The numbers of teeth, their shape and insertions into the jaw all indicate dissorophid affinities. The dentary is extremely slender, producing a narrow, tapering anterior portion of the ramus of the jaw. In all of these features, plus the sculpture patterns of the external surfaces of the dentary and maxilla, these specimens closely resemble the dissorophid amphibian Tersomius. Affinities with another genus— Broiliellus— cannot, however, be completely ruled out. No armor has been found in the conglomerates at site 2 and, inasmuch as Broiliellus is a heavily armored genus, plates would be expected if the remains were from members of that genus. Tersomius has long been known from the Wichita beds of the Lower Permian. It has now been described from beds equivalent to the Arroyo Formation in stratigraphic position from near Grandfield, Oklahoma (Daly, 1973), and from the Fairmont Shale in the vicinity of Norman, Oklahoma. The Fairmont Shale in this area is roughly equivalent to the Choza Formation, so the presence of Tersomius in the Vale is not unexpected. Family Trimerorhachidae Trimerorhachis insignis Cope 1878 The specimens assigned to this species and the types of sediment in which they occur are listed in Table 5. 01 these, UCLA VP 463 has been described and figured (Olson, 1979a). It is a nearly complete individual with a well-preserved, armored integumentum. The other specimens are fragmentary, except for complete lower jaws. T. insignis ranges through much of the Lower Permian in south central United States, appearing first in the Admiral Formation of the Wichita Croup and persisting into the highest Vale (Olson, 1955). The genus and species were reviewed by Olson (1955) and at that time only the species T. mesops from the .Arroyo Formation was recognized as separable from T. insignis in the pre-Choza formations. As discussed on p. 38, T. mesops may also be present at a middle Vale site. A distinct species, T. rogersi Olson, characterized by a deep skull, has been recognized in the Choza. Specimens from the Sid McAdams locality might pertain to T. insignis, T. rogersi, or a hitherto unrecognized species. The position of the orbits precludes assignment to T. mesops. Diagnostic features distinguishing species of Trimerorhachis are found primarily in the skull and 1982 19 The Vale Formation: Vertebrates and Paleoecology Table 5. Specimens from the Sid McAdams locality assigned to Trimerorhachis insignis. Number Part of animal Sediment TMM 30966-5 TMM 30966-431 TMM 30966-449 UCLA VP 591 UCLA VP 595 UCLA VP 463 UCLA VP 464 UCLA VP 466 UCLA VP 468 UCLA VP 469 UCLA VP 474 right mandible conglomerate conglomerate conglomerate interclavicle jaw fragments right mandible conglomerate skull fragment, girdle plates skull and skeleton part skull part skull postcranial scrap part skull postcranial scrap skeletal fragments part skull, jaw jaws. The skull and jaws of the best-preserved specimen, UCLA VP 463, conform closely in the identifying characters to T. insignis. Appropriate measurements for comparisons are given in Table 6 with those of previously identified specimen of T. insignis taken from Olson (1953, 1955) and illustrated therein. Table 6 shows considerable variation in the several measurements relative to skull length, which is considered heuristically as the independent variable. Distortion of the skulls is likely to be the principal cause of much of the variation. Figure 9 shows bivariate scatter diagrams of measurements that have proven useful in species analyses of pre-Choza Trimerorhachis. The point for UCLA VP 463 in each of the scatters conforms to the general scatter in T. insignis for the pair of measurements involved. On the basis of the data in Table 6 and the scatter diagrams in Figure 9, assignment of UCLA VP 463 to T. insignis rather than to an undescribed species is indicated. The measurements do not, however, differentiate T. insignis and T. rogersi, the primary difference between the two being the relative depth of the skulls. Figure 9 shows the scatter diagram of skull depth on skull length for T. insignis and T. rogersi (from Olson, 1955) with the point for UCLA VP 463 added. UCLA VP 463 falls within the distribution of T. insignis and not that of T. rogersi. The only possible diagnostic feature of the lower that has emerged from many studies of Trimerojaw rhachis is the structure of the posterior tuberosity. Case (1911) used the presence of a double tuberosity to distinguish species he termed T. bilobatus and T. insignis. Larger samples of the lower jaws of Trimerorhachis have shown that this character is not distinctive, hut rather that the extremes noted by Case are merely ends of a continuous, essentially normal distribution of the form of the posterior tuberosity (Olson, 1955). The lower jaws thus do not appear to possess characters that are useful for specific determination. conglomerate siltstone siltstone siltstone siltstone siltstone siltstone All of these considerations point very strongly to assignment of UCLA VP 463 to T. insignis. Critical characters for species assignment are not preserved in the other specimens of Trimerorhachis from the Sid McAdams locality. It is assumed, however, because of their close association and because of the fact that almost all of the determinable specimens of the Arroyo and Vale Formations belong to T. insignis, that this is the case for all of the materials from the Sid McAdams locality. Order Batrachosauria Suborder Diadectomorpha Family Diadectidae Diadectes sp. Diadectes sideropelicus Cope 1878 was the firstnamed species of the genus. Many others have subsequently been named; but two, D. sideropelicus from the Wichita Group and D. tenuitectes from the Arroyo Formation, have become the recognized species from the Permian of Texas. Unfortunately these are “stratigraphic” species and are not separable on morphological bases. Diadectes tenuitectes must be considered a junior synonym of D. sideropelicus. A single incisiform tooth, TMM 30966-321, provides the only indication of the presence of this genus at the Sid McAdams locality. It was collected from the surface at site I by the WPA crew. The tooth is indistinguishable from the characteristic anterior teeth of Diadectes, hut is not, of course, specifically assignable. Suborder Seymouriamorpha Family Seymouriidae Seymouria baylorensis Broilii 1904 The following specimens from the Sid McAdams locality are referred to this species: TMM 30966-176, right femur; -205, right tibia; -420, neural arch fragment; -432, dorsal vertebra; and UCLA VP 465, neural arches. The limb elements are from site 1 and 20 No. 29 Texas Memorial Museum Bulletin Figure 9.—Scatter diagram of the skull features of Trimerorhachis. .411 entered specimens are T. insignis except for 7’. rogersi and the USNM specimen, which is only tentatively assigned. Note lire different ordinate scales on the right and the left. Symbols: open triangles are pre-Vale skulls, black triangle is UCLA VP 463 (mid-Vale); open circles are pre-Vale skulls, black circle is UCLA VP 463 (mid-Vale). Open squares are pre-Vale skulls, black square is UCLA VP 463. Other symbols are explained in caption. the vertebrae from site 2. All elements are indistinguishable in form and size from the homologous hones of Seymouria baylorensis. This is the only species known from the underlying Arroyo Formation. A larger, morphologically distinct species, S. grandis, has been described from the middle Vale of Texas and from beds of comparable age in Oklahoma (Olson, 1979h) as noted later in this paper. On the basis of known material, it is unlikely that the two species may be confused. Class Reptilia Order Araeoscelida Family Araeoscelidae Araeoscelis gracilis Williston 1910 Two specimens are referred to this species, associated right and left femora, right and left ischia, and the first sacral vertebra of at least one individual contained in a block of conglomerate, TMM 30966415, and a second sacral vertebra, TMM 30966-417. The block of conglomerate is unweathered and was was apparently collected from bed I or bed 4 at site 1. The second sacral vertebra was picked up on the surface about 100 m southwest of site 1. The femora of Araeoscelis are long and slender with a sigmoid curvature and are easily distinguished from those of other Clear Fork reptiles (Vaughn, 1955, p. 376). The femora of TMM 30966-415 are about 50 mm long, less than the average adult length (64 mm) given by Vaughn. In view of the close morphological resemblances of the Sid McAdams femora to those described in detail by Vaughn, in all other respects the moderate size difference is no basis for species separation. The ischia consist of the posterior portions of right and left blades, with no trace of the acetabulum. Their general size and shape agree with the description of the ischium of A. gracilis given by Vaughn. The same similarities hold for the sacral vertebra. Vaughn(l9ss) states that he could not distinguish on morphological grounds A. gracilis of the Clear Fork from A. casei of the Wichita. The species difference is based on their stratigraphic separation. Table 6. Measurements in mm of moderate-to-large specimens of Trimerorhachis insignis, data from Olson (1953, 1956) except for UCLA VP 463. Abbreviations: Skj—skull length; Skw —skull width; 0-S]— orhitosnout length; Oj—orbital length. Number Skj Ol 0-Si Sk w UCLA VP 463 160 138 123 120 139 175 110 25 23 20 26 20 53 44 38 40 43 67 30 160 145 146 118 AMNH 4557 AMNH 4570 MCZ 1508 MCZ 1725 FMNH UR 668 FMNH UR 667 _ 22 _ 150 112 1982 The Vale Formation: Vertebrates and Paleoecology Accordingly, A. casei (Broom, 1913) should be considered a junior synonym of A. gracilis (Williston, 1910) until such time that the contrary can he demonstrated on morphological grounds. Our materials are thus referred to A. gracilis. Subclass Anapsida Order Captorhinomorpha Family Captorhinidae Captorhinus aguti (Cope) 1882 The left half of a skull and an articulated mandible of Captorhinus aguti, TMM 30966-433, from the conglomerate at site 2 is the only specimen assigned to this species (Fig. 10). A few fragments of vertebrae may also belong here but they may equally be fragments of A raeoscelis. Captorhinus aguti is the only species of Captorhinus from the Arroyo and Vale Formations (Seltin, 1959). It is readily identified by its multiple rows of somewhat laterally compressed teeth which are set in the jaws in a somewhat en echelon pattern, strong anterior maxillary “canine” teeth, sharply recurved upper incisors and somewhat procumbent anterior lower teeth. The specimen from site 2 was exposed in the sagittal plane in a block of conglomerate. The inner surfaces of the skull and jaw were well preserved. Acid and mechanical preparation revealed features of the outer surface after impressions of the exposed areas were taken. Much of the maxilla—lacrimal and Figure 10.—Captorhinus aguti, TMM 30966-433, incomplete skull from the Sid McAdams locality. Above, lateral surface; below, inner surface of side of skull. 21 jugal—and part of the postorbital and quadratojugal present, along with the dentary and part of the angular and surangular of the lower jaw. Sutures are somewhat obscure but the approximate limits of the bones are determinable (Fig. 10). I he shapes of the skull and jaw and the included bones conform closely to those of C. aguti as figured by Fox and are Bowman (1966). Assignment of TMM 30966-433 to C. aguti can be made with confidence. The specimen is small compared to the average for skulls of adults of the species. The mean overall skull length in specimens described by Seltin (1959) is about 55 mm, whereas the specimen from site 2 can be estimated to have been about 35 mm long. In view of the close conformity of the specimen to others of C. aguti both in skull and dental characters, it is assumed that the differences in size are owing to age and have no taxonomic significance. Subclass Synapsida Order Pelycosauria Suborder Ophiacodontia Family Ophiacodontidae ‘(Ophiacodon sp. The distal half of a small humerus, TMM 30966277, from the quarry at site 1 of the Sid McAdams locality, appears to belong to Ophiacodon (Fig. 11). It is extremely small, though adult. The genus was named by Marsh (1878) and its rather complex history has been reviewed by Homer and Price (1940). Unlike the small elements from the quarry referred to Dimetrodon giganhomogenes, the bone has fully developed terminal portions and articular facets. The radial condyle and ulnar facet are well formed and the entepicondyle is completely ossified. Tendon scars are sharply defined. The bone is about 3 cm wide distally and its length can he estimated to have been about 6.5 cm, on the assumption that the proportions of the missing parts fit the general pelycosaurian pattern. The specimen has distinctly ophiacodontid characters. The entepicondyle is very broad and flat, the ulnar facet does not extend appreciably onto the dorsal surface of the bone, and the supinator process, which is broken, appears to have been narrow with a sharp anterior projection. The entepicondylar foramen lies rather high up on the shaft and close to the posterior margin of the bone. The shaft itself is relatively longer and more slender than in various larger species of Ophiacodon. This specimen cannot be referred to Dimetrodon nor can it be assigned to any other genus of sphenacodontid or varanopsid. Being small and adult, it 22 Texas Memorial Museum Bulletin Figure 11.-I Ophiacodon, TMM 30966-277. The distal half of a right humerus from the Sid McAdams locality. Left, ventral aspect; right, dorsal aspect. certainly does not belong to any described species of Ophiacodon. In view of the restricted morphological information available, assignment to the genus, although reasonable, is viewed as tentative. The presence of a member of the family Ophiacodontidae or even the suborder Ophiacodontia in the Vale is somewhat unexpected. A lone survivor in the Arroyo Formation is Varanosaurus. This was an animal of intermediate size and one that is not well known. It is possible that the Vale specimen pertains to this genus, but close comparisons of available material suggest that this is not the case. Heretofore, the latest known occurrence of Ophiacodon is the very large species from the pre-Arroyo Clyde Formation, 0. major. Some small, unnamed elements from the .Arroyo Craddock bone bed may possibly be ophiacodontids (\\ illiston, 191 l)but they are generally considered to be immature specimens of Dimetrodon. The small size is of interest, although there are apparently mature individuals of about the same size from the Pennsylvanian (personal communication from Peter P. Vaughn). The single, partial humerus from the Sid McAdams locality extends considerably the range of the ophiacodontids. Suborder Sphenacodontia Family Sphenacodontidae Dimetrodon giganhomogenes Case 1907 This species was named by Case (1907) as D. giganhomogenes. Although this spelling has been used in some subsequent publications, it has been common practice to use D. gigashomogenes, for example Homer (1937) and Romer and Price (1940). No. 29 The remains of D. giganhomogenes constitute the hulk of the Sid McAdams collection. At least 22 individuals are included, based on a count of left femora. Several hundred isolated skeletal elements were recovered. These are not listed in toto, but those that are important in the following systematic descriptions are recorded in Tables 7, 8, and 9. Nearly all of the specimens have come from the light-gray clay of the quarry at site 1 (bed 2 of the measured section). Fragmentary, isolated specimens have also been found in the coarse conglomerates. Most of the bones from the quarry are in excellent condition and have preserved distinct surface detail. Initially, greater association of elements than is now recorded may have existed. Some elements, especially some of the immature long bones, have been crushed. A nearly complete skeleton, TMM 30966-356, was recovered. Measurements of the vertebrae of this specimen are included in Table 7. One of the peculiarities of the quarry collection is that remains of most parts of skulls, even the massive braincases, are uncommon and fragmentary. About a dozen partial maxillae were recovered. All of the specimens, coming as they do from a single accumulation and being similar in their major morphological features, are presumed to pertain to a single species of Dimetrodon. Histograms of sizefrequency distributions of the length of the femora and humeri, however, reveal a strongly bimodal distribution (Fig. 12-2). Measurements used are illustrated in Figure 13 and are given in Table 8. The two modes appear to represent age groupings, for all of the members of the cluster of smaller elements have incomplete ossification of the proximal and distal ends, indicating immaturity. Preservation over a short period of time could account for this distribution, as could seasonal reproduction, whereas continuous accumulation over a period of ten-or-so years could not (Olson, 1957). Annular laminae are present in the diaphyses of the long bones, and an attempt was made to age the sample by counting these. In view of the probability of climatic seasonality in the region, herein to be discussed, these annulae may represent seasonal variation in the rate of bone growth. The central area of cancellous bone is too extensive to permit an accurate estimate of the total number of laminae, but those preserved in the outer layers of the bone could be taken to indicate that the animals in the smaller cluster (F ig. 12) would be about 5 years old, assuming that two periods of growth took place during the year. This poses a problem of population structure. If Dimetrodon was a very slowly growing animal, reaching reproductive maturity in no less than 10 years, 1982 The Vale Formation: Vertebrates and Paleoecology then 5 year olds might be juveniles, as the poorly ossified bones of the smaller animals indicate. The size-frequency distribution is somewhat similar to that seen in populations of various amphibians and reptiles today, with the smaller mode repre- 23 senting young of the year and the larger mode representing the adults of earlier generations. But these distributions involve annual populations with reproduction and attainment of adulthood over a relatively short period of lime. Otherwise the distribution is Figure 12.—Distributions of measurements of Dimetrodon: 1. Regressions of humerus width on humerus length from specimens of several species of Dimetrodon as indicated by the symbols; 2 through 8, Dimetrodon giganhomogenes, based on specimens from Sid McAdams locality. 2. Regression of proximal femur width on femur length; 3 through 8, size frequency distributions of various measurements of limb elements. Texas Memorial Museum Bulletin 24 Table 7. Dimetrodon giganhomogenes, TMM 30966-356. Measurements of the basal length of centra in mm. Abbreviations: No.—vertebra number in column (axis=l); CL— centrum length. Second listing includes measurements for other cervical vertebrae from Sid McAdams locality. Series Cervical No. CL 1 2 3 4 5 28 29 28 26 6 Dorsolumbar 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Series Sacral 25 24 Caudal 24 27 31 30 30 31 33 No. CL 26 27 1 2 3 1 2 3 4 27 26 34 33 30 25 22 20 18 17 16 15 14 14 5 6 n ( 8 9 10 11 — 35 37 34 34 33 33 33 33 31 31 30 12 13 14 15 16 17 18 19 20 21 — 15 17 16 16 16 16 16 15 15 15 15 In dividual Cervical Vertebrae TMM Number CL 30966-59 30966-59 48 54 50 37 44 30966-165 30966-331 30966-341 No. 29 adults of either sex would have as low a level of ossification of the long bones as that found in the smaller individuals. Efforts to assign the Sid McAdams specimens of Dimetrodon to a species have raised some interesting questions. As far as the postcranium is concerned, size and proportions of the various elements have been used for species determination along with special features of the elongated neural spines. Problems of species assignment based on the proportions of limb hones are illustrated in Figure 12-1 in which a composite regression of the length or width of the humeri for the several species indicates no differences between the named species except in size. Most of the other characters used are even less suitable for quantitative statistical investigation. Gould (1967) has made the most definitive analysis of allometry in pelycosaurs, hut this was at a broader level and not pertinent to our aims in this paper. In their classic monograph Romer and Price (1940) list 4 species of Dimetrodon from the Arroyo Formation; D. kempae (Romer, 1937); D. loomisi (Romer, 1937); D. gigashomogenes (properly /). giganhomogenes [Case, 1907]); and D. grandis (Case, 1907). The holotype of D. kempae is a small, though apparently mature, abraded humerus (MCZ 1361). Assigned specimens include a partial skeleton, a humerus, and various isolated (unspecified) elements Position in column anterior axis ?posterior 4th or 5th middle much less distinctly bimodal. It is difficult to imagine a reproductive program that would produce a population structure including only a group of 5 year olds and a series of much older adults, which is what the aging by annulae would suggest. A model in which a population consists of one generation of young and of adults, buried over a short span of time, does fit the evidence; hut the aging by annulae must he held suspect. An alternative argument is that the two clusters indicate the presence of a strong sexual dimorphism. This possibility has been suggested for species of Dimetrodon by Romer and Price (1940) hut is contrary to the osteologjcal evidence that the smaller forms were juveniles. It seems improbable that the Figure 13.—Outline drawings of the femur (above) and humerus (below) of Dimetrodon giganhomogenesto show positions of measurements as used in tables, figures, and text. Abbreviations: DW—distal width; L—length: PW—proximal width. 25 The Vale Formation: Vertebrates and Paleoecology 1982 exception (TMM 30966-277, described earlier as an opbiacodont), they clearly belong to the growth series we have interpreted as representing a single population. Remains of D. loomisi have come largely Irom the Craddock bone bed, a “carnivore trap” somewhat similar to the Sid McAdams locality. One ol the bases tor separation of D. loomisi and D. giganhornogenes is the number of dentary teeth, 30 to 32 in D. loomisi and from the Craddock hone bed. Many of the specimens are immature, and Romer and Price (1940) felt that they were sufficiently small that, even at maturity, they would not have attained the size of adults of other species. Those authors, however, questioned whether most of the specimens were in fact Dimetrodon, rather than some other small sphenacodont. Similar small bones occur in the assemblage in the Sid McAdams collection from the quarry and, with one Table 8. Measurements in mm of the femora and humeri ofDimetrodon giganhomogenes from the Sid McAdams locality, TMM 30966. Measurements taken as illustrated in Fig. IS. Abbreviations; A—adult; crushed lat—laterally crushed; crushed v—vertically crushed; DW—distal width; e—estimate; J—juvenile; L—length; I—left side; PW—proximal width; r—right side. FEMORA TMM 30966 -49 -129 -132 -140 -186 -188 -189 -200 -201 -209 -266 -270 -27b -291 -329 -330 -367 -385 -399 L PW DW Age Condition 220 107 130 56 216 126 102 140 142 72 91 191 209 125 209 93 208 108 200 e 71 33 43 20 75 34 34 44 41 22 53 38 36 22 70 44 36 50 50 28 A good good Side r 1 1 1 r r 1 1 1 1 1 r r i i r,l 1 1 1 — - 61 63 65 38 66 28 71 33 — 35 54 23 60 30 68 J J J A J J J J J J A A J crushed v good good good crushed v good good good damaged good crushed lat J good slightly crushed lat good slightly crushed lat good A damaged A J A HUMERI -50 -61 -126 -136 -187 -208 -214 -222 -226 -227 -248 -285 -340 -361 -362 -369 -377 -377 -380 -389 -391 -398 438 r r,l 1 1 1 1 1 219 188 _ — Ill 66 197 r — r 117 12 e 99 171 98 119 180 1 1 192 i i i i r 1 ] 129 e 96 106 94+ 94 42 - 46 63 35 98 - 45 35 94 33 - — — 46 87 40 53 e 81 52 e 91 r 199 96 1 1 1 1 1 128 104 100 108 e 186 52 - 41 41 87 e 57 53 54 78+ - — - - 98 106 50 e 59 51 47 101 A A J J J J good crushed lat prox. end only ends poor poor break in shaft A good J prox. end only .1 crushed v J J prox. end absent A good damaged J J A crushed v broken damaged J incomplete A A A slightly crushed lat J J J A good good good fair left side +2 mm slightly crushed v 26 Texas Memorial Museum Bulletui or 23 in D. giganhomogenes. Estimates using teeth and alveoli have been attempted for the Sid McAdams specimens, but these have not been sufficiently reliable to allow any conclusions as to specific affinities. According to Romer and Price (1940) the vertebral spines of D. loomisi are straight, neither wavy nor recurved in the posterior presacral region; whereas those of D. giganhomogenes are somewhat sinuous and strongly recurved. This may be a real difference. Some questions arise, however. First, the spine curvature as defined appears to represent the extreme of a somewhat intergrading series. Spines in smaller animals seem to be fairly straight and those in large ones more recurved. In both D. giganhomogenes and D. grandis, recurvature is marked. In the somewhat smaller Wichita species D. limbatus there is less curvature. In the rather small skeleton TMM 30966-356 from the Sid McAdams locality, the posterior lumbar spines are somewhat recurved and twisted. Curvature may be largely a feature of growth; also, there is a problem of distortion during preservation and the usual need of preparators to make restorations while mounting specimens. How much of this has entered, into the determinations of spine irregularity is not evident, but it could play an important role. In our opinion it cannot be clearly demonstrated that D. loomisi and D. giganhomogenes are different species. This leaves no alternative on the basis of priority to the assignment of the Sid McAdams material to D. giganhomogenes if the choice is only between these two species. The situation with respect to I), grandis is somewhat different. Here skull differences enter in, as well as those of postcrania. D. grandis was associated by Romer and Price (1940) with a group of species in which the lengths of vertebral centra were short relative to the diameter of the centrum. The species is described as short-necked with the lower long bones of the fore and hind limbs short relative to other body dimensions. D. giganhomogenes was assigned to a second group characterized by elongated cervical vertebrae. Romer and Price (1940) developed and applied a unit system for expression of proportions designed to reduce the absolute size factor. This unit system has been reviewed by Currie (1978) and its strengths and weaknesses evaluated. The basis of the system was the diameter of the vertebral centrum. Determination of proportions within species depended upon associated elements in individuals. Because of the lack of associations in the Sid McAdams material it has not been possible to apply this method. In addition to characteristic proportions of limb and vertebral features. D. grandis has somewhat quadrilateral neural spines not marked by a strong figure-8 22 No. 29 cross section except in terminal portions. The short vertebrae, relatively short lower limb segments and shapes of vertebral spines may well be valid species characters. Too few specimens are available, however, to rule out the possible effects of allometric growth. Large specimens recognized as D. giganhomogenes do show a moderate increase in the quadrilateral cross sections of the spines relative to smaller individuals, but they do not show the extreme condition seen in specimens assigned to D. grandis. Again, not enough material is available for the establishment of clear statistical differences and it is possible that a graded size-series of a single species has been sampled. The skull characters of D. grandis as outlined by Romer and Price (1940) seem to separate it from other species, but this does not aid in the present situation because of the paucity of skull remains in the Sid McAdams locality. The probability that D. grandis and D. giganhomogenes are valid species seems moderately high and, as far as definitive or suggestive characters are present, the Sid McAdams materials resemble D. giganhomogenes rather than D. grandis. Thus it is concluded that D. grandis does not occur in the Sid McAdams collection. It should be noted that were D. grandis and D. giganhomogenes synonymized then D. grandis would apply to the species, having page priority in Case (1907). Lebenspuren.—Many coprolites have been collected from the Sid McAdams locality. One collection, apparently from site 1, was made by the WPA crew and the specimens are catalogued under TMM 3096626, -44, -278. -437. The preservation suggests that they came from bed 3 rather than bed 2 which has yielded most of the vertebrates. Other coprolites from the floor of the valley and slopes of the adjacent hills have been numbered UCLA VP 471, 472. Lengths range from 2 cm to 15 cm. Shapes vary from long, thin, cigar-shaped objects to thick elliptical cylinders, rounded at one end and tapered at the other. Some appear to be remains of actual faeces and others correspond closely to the intestinal casts as described by Zangerl and Richardson (1963) and Williams (1972). The majority of coprolites are composed of macerated bone disposed in layers that are either sequential or formed into crude spirals. A few contain partially preserved bone fragments and paleoniscoid scales. .All of the coprolites were produced by carnivores. The wide range of sizes and shapes indicates that several different kinds of animals produced them. The largest, those 10 cm or more in length, almost certainly pertain to Dimetrodon, the only known member of the fauna large enough to have 1982 27 The Vale Formation: Vertebrates and Paleoecology surrounded by a concretionary mass of line sand- stone. No lops nor bottoms are preserved. The general shape and dimensions are within the range of lungfish burrows described from other places in the Permian (Romer and Olson, 1954; Olson and Holies, 1975). On the other hand, the burrows are in sandstone rather than in shale, as is typical elsewhere, and the tilling is not dolomitic as is usually the case. They may he lungfish burrows, hut the nature of the sediments argues somewhat to the contrary. Various aquatic invertebrates make similar burrows, although rarely of the dimensions of those at site 4. Assignment, in the absence of contained lungfish remains, must he considered tentative. THE BLACKWOOD LOCALITY Figure 14.—Sketch plan view of Blackwood locality showing principal collecting localities 1 to 4. Plan is based on paced measurements and field sketches, so distance and proportions are only approximate. produced them. The externally and internally spiraled coprolites, characteristic of xenacanth sharks (Williams, 1972), are absent or rare as are their teeth and skeletal remains. Any of the other carnivores could have produced the smaller coprolites, but there is no reliable way of making assignments. Ichnites. —Tilton (1931) established the ichnospecies Dimetropus herea to include two footprints, those of a right manus and a right pes, from the Permian Waynesburg Sandstone of West Virginia. He ascribed them to Dimetrodon. Two footprints, TMM 30966-412 from the Sid McAdams locality, preserved as negatives in a fine sandstone, are referred to Dimetropus. These were found near site 1 and represent the impressions of left fore and hind feet. They are considerably smaller than the prints described by Tilton, but are otherwise similar. The length of the hind foot from the heel to the end of the 4th digit is 24.8 cm in Tilton’s specimen and 14.0 cm in the Sid McAdams specimen. Both the Sid McAdams ichnites and those described by Tilton show definite claw marks. The size of the ichnites from site 1 is consistent with that of podial elements of Dimetrodon from the quarry. Burrows.—A few cylindrical structures from site 4 may be burrows of lungfish. They consist of holes 2 cm to 5 cm in diameter filled with red clay and Extensive exposures on the property of Mr. L.A. Blackwood, about 3.5 km south of Buffalo Gap, Taylor County, Texas, have yielded a large collection of fossil vertebrates. The locality was discovered by Dr. David Berman of the Carnegie Museum, Pittsburgh, Pennsylvania. He collected from it in 1970 and brought it to our attention soon thereafter. The specimens he obtained are housed at the Carnegie Museum. During three subsequent field seasons other collections have been made, and these are housed at the University of California, Los Angeles. Over much of the area in the vicinity of Buffalo Gap to the east, west, and north of the Blackwood locality, the Vale deposits are but sparsely fossiliferous and the specimens that do occur are fragmentary. However, the beds on the Blackwood property and immediately to the north are rich in fossils. Possible reasons for this concentration are discussed later in this section of the paper. The Blackwood locality lies about IB km almost due north of the Sid McAdams locality and about the same distance south of the Stamford site. To date relatively few vertebrates have been found in the areas between these three productive places. Geology The Blackwood locality (Fig. 14) covers an area that measures about 1.5 km north-south and 1 km east-west. It is flanked on the west by a steep bluff and to the east by lower hills limited by a railroad cut. Additional open exposures outside the area are found on the Carey Ranch to the west and on the Grimes property to the north. Over much of this area and passing northward well past Buffalo Gap. the outcrops are made up predominantly of red and maroon siltstones with irregularly distributed beds and lenses of sandstone, mudstone, and conglomerate. Lateral variation is rapid as elsewhere in the Vale Formation. No. 29 Texas Memorial Museum Bulletin 28 Table 9. Measured section at Blackwood locality (middle part of the Vale Formation) taken at site 2. Bed 6 soil zone, Quaternary 5 red siltstone, bright, evenly bedded; scattered fine sandstone lenses; no fossils found 4 thin, brown, fine mudstone; source of fragments local; thickens and thins laterally . 3 maroon shale and siltstone;evenly bedded; weathers to form limonitic nodules; occasional layers of thin green clay, 2 to 4 cm thick; main fossil-bearing bed; most 2 maroon clay-shale interbedded with fine, conglomeratic mudstone; pebbles local; dips up to 5 degrees in places, but irregular; 50 yds to east passes into conglomerate up to 3 m thick 1 maroon clay-shale and siltstone to creek bed; evenly bedded, with uneven contact with bed 2 Age of the Deposits.— Deposits.—Both to the east and west of the Blackwood locality the Permian is covered by marine Cretaceous deposits. To the east of the locality, both to the north and south of a westward-extending nose of the Cretaceous cover, the Standpipe Limestone is exposed and a westward measurement from an extrapolated line between the outcrops shows that the Blackwood locality lies about 12 km west from the base of the Vale. To the west of the Blackwood locality.. the position of the covered Bullwagon Dolomite is less definable, but by extrapolation it appears to he about 9 km distant. The Vale Formation at the level of this locality is about 120 m to 130 m thick and the Blackwood exposures appear to lie about 70 m to 80 m above the base of the Vale as marked by the top of the Standpipe Limestone. This estimate is based on the assumption of a fairly uniform dip of the beds to the west, which is quite consistent throughout the region. Elevations at the outcrops of the Standpipe Limestone and the expossures at the Blackwood locality are approximately the same, and hence the exposures on the Blackwood locality are considered to lie in the middle part of Vale Formation. Sediments.—Maroon and red siltstone, mudstone, Sediments.— and conglomerate are the dominant sediments at this locality. Clays and fine sandstones occur as discrete bands and lenses in the siltstones; conglomerates and mudstones form lenticular, elongated structures which contain interbedded sandstone. The general relationships of the sediments as they exist over the area are shown in the measured section at site 2, Table 9 (see sketch map, Fig. 14). Bed 4, a thin mudstone, is sporadic in occurrence and is absent in many of the exposures. The siltstones are evenly bedded throughout. The break between the lower maroon siltstone and the overlying red siltstone is generally Bed thickness Cumulative thickness (meters) (meters) - . . . . . . . 3.00 0.15 14.65 11.65 6.00 11.50 2.50 5.50 3.00 3.00 sharp. Both the red and maroon siltstones are composed of very fine quartz grains in a clay matrix. In the central valley of the Blackwood locality, complex conglomerate lenses as thick as 2 m, but usually less, crop out sporadically along the northsouth axis over an east-west hand about 0.3 km in width. They consist of interbedded mudstone, gray carbonate pebbles and yellow-to-hrown calcareous and limonitic pebbles. Little or none of this material seems to have come from caliche. The pebbles are rarely over 0.5 cm in maximum diameter. The cement is primarily calcite. Small lenses of sandstone and fine conglomerate also occur in the red siltstone, hut the major conglomerates are confined to the lower half of the maroon siltstone. Beyond the limits of the Blackwood locality, both to the west and north, red siltstones predominate and in places they are interbedded with lenses and sheets of fine- to medium-grained sandstone. About 5 km west of the Blackwood Ranch and 5 km southwest of Buffalo Gap, along the road to Lake Abilene, thick lenses of the Buffalo Hill Sandstone crop out, indicating a depositional stream regimen quite different from that at the Blackwood locality. These adjacent areas, where sandstones form a more significant part of the sediments, have yielded very little vertebrate material. In the Blackwood area fossil vertebrates occur in the maroon siltstones and in the conglomerates and mudstone lenses (beds 2 and 3 of the measured section). No vertebrate remains have been found in the red siltstones and none has been recovered in the section below the conglomerates of bed 2. Depositional Depositional environment.— Of the principal sedimentary types exposed in the Blackwood area, only the conglomerates and small sandstone lenses offer definitive evidence of the nature of their origin. 1982 The Vale Formation: Vertebrates and Paleoecology The major conglomerates of the central valley were formed in stream channels, deposited by swiftly running water. No graded bedding has been found, hut within the major lenses smaller lenses of conglomerate and sandstone cut earlier deposits, indicating a complex history of erosion and deposition. The width of the channel zone is about 0.3 km and the system has been traced for about 1.5 km northsouth. For this full length the conglomerates have incised the maroon siltstone. Thickness varies somewhat, being greatest in the center of the lenses, where it ranges up to 2 m. Fragments of fossils occur irregularly in the conglomerates. Most are broken and abraded and there are no concentrations. The stream deposits and their contained fossils show none of the usual features, such as mudcracks and multiple channeling, associated with deposition in a climate with sharply defined seasonality. Some sandstone lenses are found within the conglomerate complex, hut most sandstones in the area form small, lenticular bodies within the red siltstone (bed 5 in the measured section, Table 9). These appear to have been formed in small channels rarely more than a meter wide and 0.3 m thick. Linear dimensions are not ascertainable. No fossils have been found in these lenses. The maroon siltstones (beds 1 and 3) are uniform throughout. Bedding is rarely well preserved, but where it can be made out it is very even. No evidence of microbedding or other small internal structures has been observed. There can be little doubt that these siltstones were water laid but that current action was slight and left few marks in the sediments. Siltstone deposits of this type are characteristic of both lake and flood-plain sedimentation in which fine sediments settled from still or nearly still water. Often in the Vale, lake beds can be identified by the presence of dipping marginal deposits. None have been found in the Blackwood area. .Also, lake beds tend to be lighter in color than the surrounding deposits, due to the reduction of the iron oxide in the presence of concentrations of plant remains. Similar reduction is also characteristic of the thin beds formed close to stream channels at flood stages. Again, in the Blackwood Ranch area, this type of reduction does not occur. The physical aspects of the sediments themselves do not provide a basis for precise interpretation of depositional conditions, but the general setting within the wider area and the vertebrate fossils contribute some information on this matter. Extensive, relatively thick sandstone lenses and wedges are exposed to the west and north, for example near Lake .Abilene (p. 28) and on the Carey Ranch to 29 the immediate west. These sandstones indicate an area of major stream activity approximately equivalent in lime, or a little later, to the formation of the Blackwood deposits only a few kilometers away. The red siltstone at the Blackwood locality is lithologically and structurally similar to the siltstones associated with the sandstones and formed upon adjacent flood plains. The maroon siltstones are more local and at the base are related to the conglomerate channel deposits. It may be that the higher parts of these maroon beds were also related to similar channels which have shifted sufficiently in position that they are not now exposed in the relatively limited areas of outcrop. In any event, the channel conglomerates and the maroon siltstones do represent a unique situation with respect to both the sediments and fossils, one not reflected in adjacent areas. The overlying red siltstones at the Blackwood locality indicate the transgression of more usual conditions over the area of the vertebratefossil-producing region, presumably accompanying the development of a stream system including a fairly large, slowly flowing stream whose channels were sites of deposition of sand. Under this interpretation the maroon siltstones and conglomerates appear to indicate fairly local and temporary conditions under which deposits were formed in the channels of rather rapidly flowing, small streams with restricted flood plains lateral to them. The occurrence of vertebrates supports this interpretation. Taphonomy.—Xenacanthus, the freshwater shark, Taphonomy.— has an interesting but not unique occurrence. It is represented almost exclusively by the calcified cartilage of chondrocranial and visceral elements. Although each individual carried about 400 teeth at any one time (and many hundreds during its lifetime), only a few have been found in the deposits. In more typical Permian sites where shark remains arc present, teeth tend to be abundant and cartilaginous elements absent or scarce. In the Blackwood locality, fairly large pieces of chondrocrania are preserved and a major portion of the chondrocranium of a small individual was found in a large coprolite, probably produced by Eryops. As the soft tissue disintegrated after death, it may he supposed that teeth were freed and fell to the bottom. The chondrocranial parts, perhaps with some remaining soft tissue, being more buoyant, were rafted by gentle currents to be deposited beyond the usual life habitats. Gentle currents during flood stages of a stream could account for this segregation of the teeth and chondrocranial parts. Diplocaulus occurs both in stream channel deposits and in the maroon siltstones. Partial to nearly 30 complete skeletons have been found in the latter. This could support either a lake or flood-plain hypothesis for deposition of the siltstone. Elsewhere in the Vale, however, A recurvatus, the species identified here, is known only from stream deposits. In the very lowest beds of the Vale, D. magnicornis occurs in pond deposits, continuing the pattern seen in the Arroyo. If the stream habitat for D. recurvatus holds at the Blackwood locality, then either the animals were carried from the streams during times of flood to the less active waters of the flood plains and then buried, or perhaps the animals swam into quieter waters and perished as the ponded waters receded and disappeared. Predator action could account for the fragmentary nature of some of the specimens. Eryops and Dimetrodon are the other abundant vertebrates at the locality. Eryops occurs as fragmentary remains in the conglomerates and as single bones to partially articulated skeletons in the maroon siltstones. No complete skeletons have been found, but associations of various parts of individuals in gully washes suggest that they may have come from fairly complete individuals. It is possible that future work will reveal these. Elsewhere in the Vale, Eryops, although not abundant, appears to have dwelt along stream and lake margins. It was not a common inhabitant in clearly identified lake deposits. The presence of abundant coprolites at site 3, in association with Eryops, suggests that the site of deposition is not far from the area of habitation. Here, as for Xenacanthus, the biological evidence seems to point to deposition on a flood plain—in the case of Eryops, near the site of the life zone of the amphibian. Dimetrodon is present primarily as skeletal and skull fragments throughout much of the area, but fairly complete skeletons have been found at the northern extremity of the locality (site 4). Remains are common in the maroon siltstones and conglomerates. When found in other localities along lake margins, which appear to provide a common habitat for Dimetrodon, remains tend to be well preserved and partial skeletons are frequently encountered. OccaTable 10. Xenacanthus No. 29 Texas Memorial Museum Bulletin sional intact specimens have come from flood-plain deposits, but these are rare, particularly in the Vale. Much of the Dimetrodon material at the Blackwood locality appears to have been transported and many of the hones are strongly abraded. Even the more-orless complete skeletons may have been carried some distance by rapidly flowing water. Under these circumstances Dimetrodon gives no particularly definitive evidence on the conditions of sedimentation under which it was deposited. In summary, the evidence, although somewhat equivocal, suggests that the maroon siltstones were deposited in relatively quiet receding waters of flood plains. No evidence of rapid water flow beyond the limits of the channels is found, so flooding was probably the result of a gentle overflow of the streams and not of abrupt increases in velocities such as often accompany torrential rains. Vertebrates living in and about the streams were carried or swam into the flood waters and were either buried when transported, as currents subsided, or died as waters receded to form small temporary ponds. Following upon this type of deposition the maroon siltstones were covered by red siltstones which predominated through much of the broader area. As noted earlier, these siltstones appear to have been flood-plain deposits associated with much larger streams in which sands form the principal deposits in and near the channels. Such conditions, which were unfavorable for preservation of fossils, ended the special circumstances that produced the rich fossil locality. Systematic Paleontology-Fauna Class Chondrichthyes Order Pleurocanthodii Family Xenacanthidae Xenacanthus platypternus (Cope) 1884 Specimens of 5 individuals of Xenacanthus and few isolated teeth have been recovered (Table 10). The 5 all consist of chondrocranial remains and these a platypternus from the Blackwood locality in the collections of the University of California, Los Angeles, UCLA VP. Number Part of animal 555 556 560 566 587 chondrocranial fragments chondrocranial fragments partial chondrocranium partial chondrocranium partial chondrocranium in coprolite of Eryops (?); two coprolites of Xenacanthus several teeth — Site Sediment 2 2 2 2 siltstone siltstone siltstone siltstone 3 2 siltstone siltstone 1982 The Vale Formation: Vertebrates and Paleoecology 31 specifically determinable. Four of the specimens (UCLA VP 555, 556, 560, 566) came from the maroon siltstones at site 2 (Table 10). A few teeth came from this site. None of the specimens was in place in the sediment, hut each grouping was sufficiently isolated that it is safe to assume that a single individual is represented. A fifth (UCLA VP 587) consists of a partial chondrocranium preserved in a coprolite. At this site there was an abundance of moderately large coprolites, up to about 4 cm in length with rounded, oval cross sections. The site has yielded remains of Eryops megacephalus and little else. The coprolites probably pertain to this species, and presumably Eryops fed on Xenacanthus. Of about 40 coprolites collected, 2 have the characteristic shapes and marking of coprolites of Xenacanthus. are not The few isolated teeth of Xenacanthus have come from site 2. A thin flat base, a single central tuherculum and long, divergent lateral cusps show these to belong to X. platypternus. It is on the basis of these teeth that the specific assignment of all of the materials has been made. Class Amphibia Subclass Lepospondyli Order Nectridea Family Keraterpetontidae Diplocaulus recurvatus Olson 1952b Remains of Diplocaulus are the most abundant fossils at the Blackwood locality. Almost all have come from the maroon siltstones at site 2, hut a few are from the conglomerates at site 1 and the shales at site 4. Some specimens were collected by Berman from site 2, and these are housed at the Carnegie Museum, Pittsburgh, Pennsylvania. They include some good skull material and incomplete specimens. Thirty-eight numbered specimens are in the UCLA VP collection (Table 11). These include two nearly complete skulls, some jaws, and skeletons in various degrees of completeness. The articulated specimens UCLA VP 439 and 440 were found partly weathered out ol the sediments and partly in place. All others had been completely weathered out. Individuals range from suhadults to adults, and no juveniles have been found. Seven of the skulls in which the tabular area is well preserved have the “horns” recurved (Fig. 15). As in the case of the specimens from Stamford (p. 36), the presence of this character called “crook is the basis for assignment of the specimens to D. recurvatus. Lacking the critical region of the skull, the other materials are not morphologically distinguishable from D. magnicornis. They are, however. Figure 15.—“Horns” of skulls of Diplocaulus recurvatus from the Blackwood locality, middle part of Vale Formation. 1. UCLA VP 497; 2. UCLA VP 439; 3. CM 24962; 4. CM uncat.; 5. CM 24964; 6. UCLA 440. assigned to D. recurvatus on the basis of association with those specimens showing the distinctive “crook.” Subclass Labyrinthodontia Order Temnospondyli Family Eryopidae Eryops megacephalus Cope 1877 Twenty-two specimens are present in the UCLA collections VP (Table 12); all are from the 4 sites shown on Figure 14. The specimens are fragments of skulls, jaws, and postcrania, hut Berman (personal communication) reports substantial parts of skulls and articulated vertebral columns. Most of the materials have come from site 2, hut some associated parts of individuals are from site 3. The fragmentary condition of the materials would normally make species assignment difficult or impossible. Eryops megacephalus, however, is a form species, and the name is 32 No. 29 Texas Memorial Museu m Bulletin applied to most specimens of the genus from the Lower Permian. All the specimens from the Blackwood locality conform in morphology to various individuals included in this form species. One extremely large individual from site 4 is known only from partial limh bones; but at present no basis for discrimination exists except the questionable criterion of size. Order Batrachosauria Suborder Diadectomorpha Family Diadeclidae Diadectes sp. Only 2 fragments of Diadectes have been found in the Blackwood deposits. Part of a quadrate, UCLA VP 558, is from the conglomerate at site 1 and an incomplete vertebra has come from a siltstone at an unnumbered site at the southern end of the exposure. Diadectes has always proven laxonomically difficult at the species level (Olson, 1947). The two commonly recognized species of the Texas Lower Permian are D. sideropelicus and D. tenuitectes. As discussed on page 19, these are “stratigraphic” rather than morphological “species.” Efforts at interpretation are hampered by the apparent great variability of the skulls and the relative constancy of poslcranial structures. Diadectes is a common constituent of vertebrate collections from beds of the Wichita and lower Clear Fork Permian. It is fairly abundant through the lower part of the Arroyo Formation hut diminishes rapidly in frequency to become rare in the upper beds. Except for the Blackwood site, only 2 specimens, a vertebra from the lowest Vale of Knox County (Olson, 1956b) and the tooth noted from the Sid McAdams locality (p. 19), are known. The specimens from the Blackwood sediments are thus the highest so far known from the Lower Permian. Table 11. Diplocaulus recurvatus from Blackwood locality in collections of the University of California, Los Angeles, UCLA VP. Number 439 440 441 443 444 445 446 459 460 461 492 493 494 495 496 497 498 529 530 538 539 540 541 542 543 544 545 546 547 548 549 550 574 607 609 611 612 Part of animal skull, jaws, postcranium skull, jaws, postcranium part skull part skull, vertebrae part skull, vertebrae part skull ( 'Wiplocaulus) part skull, vertebrae part skull part skull part postcranium fragment of skull, lower jaw part skull, vertebrae part large skull, lower jaw, vertebrae part small skull, lower jaw, vertebrae part skull, part jaw, vertebrae fragment of skull, vertebrae fragment of skull, vertebrae fragment large skull part small skull, fragment of large skull part skull, vertebrae 2 lower jaws part skull, vertebrae 2 vertebrae fragments of skull, vertebrae several articulate vertebrae fragment of skull part large skull part skull part small skull Vi skull, jaw, vertebrae part skull, ?vertebrae fragments 2 large skulls part skull, caudal vertebrae part skull part skull vertebrae, part skull vertebrae Site Sediment 2 siltstone siltstone siltstone 2 2 1 2 2 2 2 2 2 2 2 2 2 2 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 4 4 4 4 conglomerate siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone conglomerate conglomerate siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone siltstone 1982 33 The Vale Formation: Vertebrates and Paleoecology Suborder Seymouriamorpha Family Seymouridae Seymouria grandis Olson 1979b Several large vertebrae, a skull fragment, part of femur and an interclavicle show the presence of a large seymouriamorph at the Blackwood locality. The specimens and measurements are given in Table 13. The specimens have been considered elsewhere (Olson, 1979b) along with specimens of the same species from the level of interfingering of the Garber Sandstone and Fairmont Shale of the Hennessey Group near Crescent, Oklahoma. The dorso-lumhar vertebrae of S. grandis resemble th ose of S. baylorensis, hut are larger by about onefifth. Correlated with the larger size, the neural arches and spines are relatively more robust in S. grandis. The cervical vertebrae, however, known oidy from Oklahoma, have proportionately broad arches and the transverse processes are thin anteroposteriorly. The partial skull from Oklahoma shows various distinctive features of the brain case and skull table. A maxilla is the only skull part found at the Blackwood locality. A femur (UCLA VP 572) from the Blackwood locality, tentatively assigned to Seymouria, is very different from that of S. baylorensis but resembles the femora of some Old V'orld genera such as Kotlassia. Although it was found associated with the maxilla, its taxonomic position is uncertain. Assignment of the femur to any other genus known from a the locality is out of the (juestion, hut the possibility that it belongs to a dissorophid rather than a seymouriid remains open. The Blackwood specimens have come from the conglomerates at site 1 and from the maroon siltslone at site 2. All have been found as single hones; although the maxilla and femur were found in proximity. This association, however, may not he significant because the specimens are from a conglomerate. Class Reptilia Subclass Anapsida Order Captorhinomorpha Family Captorhinidae Labidosaurikos sp. Remains of captorhinomorphs have been encountered only rarely in outcrops of Vale Formation south of Knox and Foard Counties, Texas. One specimen from the Sid McAdams locality is noted on page 19, and two have come from the Blackwood locality: UCLA VP 577, four and one-half presacral vertebrae; and UCLA VP 601, a badly crushed and damaged skull with maxillary dentition. These two specimens were not associated and they represent individuals of different sizes. For reasons pointed out below they are assigned to Labidosaurikos and probably belong to the same species. Labidosaurikos meachami (Stovall, 1950), the first named species, came from the basal part of the Table 12. Eryops megacephalus from the Blackwood locality in the collections of the University of California, Los Angeles, UCLA VP. Site Sediment several vertebrae partial lower jaw partial lower jaw limb bones, etc. skull parts, vertebrae, ribs girdle, limb parts vertebral fragments 2 2 2 2 3 2 siltstone siltstone siltstone siltstone siltstone siltstone 1 conglomerate vertebral spines, coprolites skull fragment 3 1 2 1 2 2 1 2 1 2 1 3 4 4 4 Number Part of animal 422 448 449 452 457 532 533 534 535 536 552 553 554 562 563 564 565 567 578 608 614 617 various fragments rib, intercentra, etc. vertebrae skull parts spine, part vertebral column vertebrae skull fragment scrap part pectoral girdle approx. 40 coprolites skull parts part jaw limb bones, vertebral spines siltstone conglomerate siltstone conglomerate siltstone siltstone conglomerate siltstone siltstone siltstone conglomerate siltstone siltstone siltstone siltstone No. 29 Texas Memorial Museum Bulletin 34 from anterior to posterior in the series are as follows: 14.2, 16.2, 16.3, 16.3, respectively, with a mean of 15.8. The means for comparable measurements of series of vertebrae in specimens of other Clear Fork captorhinids are: x Species 8.8 17.0 18.0 22.4 Captorhinus aguti Captorhinikos chozaensis Labidosaurikos barkeri Labidosaurus hamatus Figure 16.—Vertebrae tentatively assigned to Labidosaurikos, from the Blackwood locality, middle part of Vale Formation. Anterior to the left. Above, lateral view of 6 vertebrae; below, dorsal view of 2 vertebrae. UCLA VP 577. Fairmont Shale where it interfingers with the Garber Sandstone near Hennessey, Oklahoma. Seymouria grandis has come from the same site. Labidosaurikos barkeri (Olson, 1954c) was based on specimens from the Vale Formation in Knox County, Texas. It was synonymized with L. meachami by Seltin (1959), but L. barkeri is being retained in this paper because of differences indicated in the original description. The four and one-half vertebrae (Fig. 16) are fairly well preserved hut the centra have been damaged. A measurement of the posterior zygapophysial width along the series gives an idea of the general size of the individual. Measurements in mm of UCLA VP 577 mm The vertebrae of Captorhinus aguti are much smaller than those of the Blackwood specimen, and even the largest leave a marked gap in dimensions between the two. Labidosaurus hamatus, based on normal-sized adults, was a much larger animal. The vertebrae cannot, of course, be excluded from L. hamatus on this basis, but the size difference suggests a taxonomic distinction. The vertebral measurements of Labidosaurikos barkeri and Captorhinikos chozaensis are similar to those of UCLA VP 577 and provide no basis for separation. The Vale Captorhinikos valensis, for which vertebrae are not known with certainty, appears to have been much smaller on the basis of skulls and jaws, more in the range of Captorhinus aguti. The morphology of the vertebrae does not differentiate the captorhinid genera or species, but the size of the Blackwood vertebrae suggests probable affinites with either Captorhinikos or Labidosaurikos. The skull UCLA VP 601 is badly crushed so that the existence, not to mention shape, of most elements, cannot be determined. The length of the skull is approximately 60 mm and well within the range of Captorhinus agati. A fairly well-preserved maxilla with teeth, however, precludes assignment to this genus (Fig. 17). The teeth are arrayed in 5 regular rows posteriorly and a single row of slightly larger teeth anteriorly. The regularity of the rows of teeth, their lack of curvature, and their number are characteristic of Labidosaurikos and not Captorhinikos or Captorhinus. Table 13. Seymouria grandis from the Blackwood locality in the collections of the University of California, Los Angeles, UCLA VP. Measurements in mm. Abbreviations; Pzw-postzygophysial width; e-estimated; a-approximate. Number 458 559 570 571 572 595 613 Part of animal 2 dorsal vertebrae 2 presacral vertebrae 1 dorsal vertebra 1 dorsal vertebra premaxilla, part femur interclavicle presacral vertebra Site Sediment 1 2 2 2 1 1 2 conglomerate siltstone siltstone siltstone conglomerate conglomerate siltstone Pzw 4.8 4.3 a 4.7 3.6 e — - 4.4 a 1982 35 The Vale Formation: Vertebrates and Paleoecology The skull, however, is much smaller than any known for Labidosaurikos barkeri or L. meachami. The one known skull of the latter is 259 mm long (Stovall, 1950) or about 5 times the length of the Blackwood specimen. The latter, of course, might be a young individual of either species, hut moderately full development of the maxillary tooth row argues to the contrary. On the basis of what is now available, specific assignment cannot be made. It is assumed that UCLA \ P 577 and 601 belong to the same genus and species on the bases of morphology and their occurrence in the same beds. Subclass Synapsida Order Pelycosauria Family Spbenacodonlidae Dimetrodon giganhomogenes Case 1907 Remains of Dimetrodon, as in most vertebrateproducing Lower Permian terrestrial deposits, are common in the exposures at the Blackwood locality. Specimens in the UCLA collections are listed in Table 14 and Berman (personal communication) has obtained other specimens from this locality. Most specimens are fragmentary or consist of a few associated bones. UCLA VP 605 from site 4, however, was deposited as a nearly complete individual but had largely weathered out from the rock when it was found. All materials are assigned to the common Clear Fork species D. giganhomogenes. This is on the basis of size, the strong “figure-8” cross section of tbe long Figure 17.—Maxillary teeth of Labidosaurikos sp. From the Blackwood locality, middle part of Vale Formation. UCLA VP 601. vertebral spines, elongated vertebral centra, and general overall lightness of the structure of all elements. The Clear Fork Dimetrodon species has been considered in detail pages 22-26. The only other species that could be present in the Clear Fork is D. grandis, and none of the characteristics of the Blackwood materials are those attributed to it. THE STAMFORD LOCALITY Geology This site was collected and reported upon by Dal(1963). The geology was discussed as fully as possible in theirreport and a detailed analysis ol the probable conditions of deposition was included. Visits to the area since 1969 and additional (juest and Mamay Table 14. Dimetrodon giganhomogenes from the Blackwood locality in the collections of the University of California, Los Angeles, UCLA VP. Number Part of animal Site Sediment 450 458 480 481 482 483 484 485 486 487 488 489 490 491 583 584 585 586 605 606 610 615 bl6 jaw fragments 1 2 2 2 siltstone siltstone siltstone siltstone siltstone siltstone limb elements, centrum 2 vertebrae, spine fragments tibia and scrap large sacral vertebra 3 centra, scrap 5 centra, skull, girdle fragments 3 centra, spine fragments 2 centra, scraps centrum and scrap parts of vertebral arches ribs, fragments part scapula centra, girdle parts, limb elements part femur, sacral vertebrae part jaw, skull fragments limb and vertebral parts axis, centrum, ribs, carpals part skeleton and jaws vertebrae, scrap vertebrae and scraps vertebrae, humerus vertebrae, limb bones — i i i 2 1 2 1 2 2 2 2 1 2 4 4 4 4 4 conglomerate conglomerate siltstone conglomerate siltstone conglomerate siltstone siltstone siltstone siltstone conglomerate siltstone siltstone siltstone siltstone siltstone siltstone 36 Texas Memorial Museum Bulletin No. 29 shaped lumps of clay. The siltstones form a complex of thin, semilenticular bodies nested in a reddish brown sandy clay. Dalquest and Mamay (1963) interpreted the deposit as having been formed in shallow channels of a drying stream into which the siltstone was carried by renewed rainfall after the associated organisms had perished due to drought. This concept was advanced to explain the high percentage of aquatic animals and the absence of most types of semiaquatic amphibians, fish, and terrestrial animals that presumably existed in this general area at the time. Occurrence of the vertebrates Figure 18.—“Horns” of specimens of Diplocaulus recurvatus from the Stamford locality, middle pari of Vale Formation. 1. UCLA VP 588; 2. UCLA VP 589; 3 through 7, USNM uncat.; 8. USNM 22872. collecting have substantiated the descriptions only a short summary need be included here. so that The locality lies just north of the Jones-Haskell county line, Texas, about 3 km northwest of the center of Stamford, Texas (Fig. 3b). The approximate boundary between the Arroyo and Vale Formations lies about 4 km to the east, and the Bullwagon Dolomite crops out about B km to the west. The locality thus lies about one-third of the distance between the limiting boundaries of the Vale. Inasmuch as the topography is moderately flat and the strata dip fairly uniformly from east to west, it is assumed that this locality lies near the top of the lower third of the Vale Formation. Being about 45m to 50m above the somewhat indefinite Arroyo-Vale contact, it is approximately equivalent to the transition from the lower to middle Vale Formation in Knox County, Texas. The sediments that carry most of the vertebrates this locality form a roughly heart-shaped area about 15 m to 20 m in length and width. The matrix is a hard, greenish gray siltstone containing irregularly at Vertebrates occur throughout the siltstones but concentration varies widely over the area, tending to be low near the margins of the siltstone bodies. The bones on the whole are poorly preserved, having suffered solution and recrystallization as a result ol percolating waters and, in many cases, having been deformed during compaction of the siltstone. Remains of Diplocaulus are by far the most abundant, and skulls and jaws make up the preponderance of specimens of this genus. Some postcranial elements are present and in some instances they are associated with the skulls. Intact skulls occur in various orientations but most are preserved with their leading edge deflected strongly downward toward the base of the siltstone. Trimerorhachis is present but widely scattered. Skulls, jaws, and some postcranial material have been recovered. No fish have been found except for a single tooth of Xenacanthus in the sandy clay. At some distance from the siltstones, in the red brown sandy clay, burrows—presumably those of Gnathorhiza—have been found by Berman (personal communication). Otherwise the sediments adjacent to the channels are barren of fossils. Systematic Paleontology-Fauna Class Amphibia Subclass Lepospondyli Order Nectridea Family Keraterpetontidae Diplocaulus recurvatus Olson 1952b Literally hundreds of specimens of Diplocaulus the Stamford locality. The collection at the National Museum of Natural History (USNM) includes more than 50 and there are several in the collections of the University of California, Los Angeles (UCLA VP). Five partial skulls in the collections of the USNM preserve gross features, and critical dimensions can be measured (Table 15; Fig. 18). The skulls range from 65 mm to 155 mm in length (measurements as in Olson, 1953). The ratios of skull occur at 1982 37 The Vale Formation: Vertebrates and Paleoecology width to skull length and of orhito-snout length to skull length (Table 15) fall within the range of those for D. magnicornis. The ratio oforhito-snout length to skull length ranges from 1:4 to 1:5, which is also within the range of D. magnicornis and rules out assignment to the Arroyo “river” species D. brevirostris in which this ratio is not less than 1:7 (Olson, 1951b). The distinction between D. magnicornis and D. recurvatus, as also noted on page 17, lies in the configuration of the tabular projection of the skull. In the majority of individuals of D. magnicornis the tabular extends laterally, whereas in D. recurvatus it is reflected rather sharply posteriorly, producing the character called “crook” (Olson, 1952b). In a sample of 25 specimens of D. magnicornis from the Arroyo Formation, crook is well developed in 16 percent of the sample, partially developed in 28 percent and absent in 56 percent. In a sample of 7 specimens of D. recurvatus from the Vale Formation, crook is fully expressed in 86 percent and partially expressed in 14 percent. Thus, a single skull of Diplocaulus cannot be used for specific assignment. In 6 skulls of Diplocaulus in the USNM collection and 2 partial skulls in the UCLA VP collection, the only ones in which the tabular “horn” is well preserved, the character of crook is strongly developed . On the basis of this, even though the percentage of crook in the sample cannot be ascertained, assignment of the specimens to D. recurvatus has been made. It is assumed that all specimens from this site pertain to a single species. The fact thatDiplocaulus from the Stamford site appears to have been a stream dweller—not a pond dweller as D. magnicornis has been found to be—is consistent with the assignment to D. recurvatus. Subclass Labyrinthodontia Order Temnospondyli Trimerorhachis 9 .insignis Cope 1878 Collections from the siltstones at the Stamford locality include several incomplete skulls and jaws that are referable to Trimerorhachis. The problem of specific assignment arises because the necessary details of skull proportions, dentitions, and sculpture patterns are not shown in most of the specimens. Numerous sharp teeth form a long row along the jaws,changing only in a decrease in size at the posterior end. The teeth plus the surface ornamentation, where visible, are typical of Trimerorhachis, but neither feature is sufficient for specific determination. Skull measurements can be obtained for two trimerorhachid skulls in the United States National Museum collections. One of these is clearly Trimerorhachis and the other, of less certain affinities, is considered in the following section. Measurements of the Table 15. Diplocaulus recurvatus from the Stamford locality, based on skulls in the collections of the National Museum of Natural History. The specimens are entered by sequential numbers because they have not been numbered in the collection. Abbreviations: Ski—skull length; Skw —skull width; 0-Si—orbitosnout length. Measurements Skulls 1 2 3 4 5 Ski Sk w O-Si 98 65 95 85 115 165 88 20 18 — 144 240 — — 28 former are given in Table 16a along with those of the skulls of large individuals of Trimerorhachis insignis for comparison. The Stamford skull is about onefifth larger than any other known complete skull of the species. Large jaws, suggesting the existence of equally large individuals, are known from both the Wichita and lower Clear Fork beds, but their specific relationships are uncertain. As Table 16c shows, the ratios of various measurements to skull lengths of the Stamford specimen fall within the range of the other Lower Permian specimens. The range, however, is great and data are few. The position of the orbits, as indicated by the ratio of orhito-snout length to skull length, expresses an important aspect of T. insignis, the forward position of the orbits. The ratio in the National Museum skull falls within the distribution for the species. From the available data, the large National Museum specimen appears to be referable to T. insignis. The only other known species from the Clear Fork to which it might belong is T. rogersi from the Choza Formation. This species was based primarily upon the ratio of skull depth to skull length, which is consistently greater than that in T. insignis. A measurement of skull depth near the posterior end is possible in the National Museum specimen but distortion is such that it must be considered somewhat unreliable. Figure 9 shows the regression of skull width on skull length for specimens of T. insignis and T. rogersi, from Olson (1955), with one specimen of T. rogersi and one of T. insignis added. The point for the National Museum specimen falls between extrapolated lines for the two species, somewhat closer to that of T. insignis. Again, assignment to T. insignis is suggested, but the basis is tenuous. If the National Museum specimen is, as appears probable, T. insignis, then most of the other specimens from the siltstone probably also pertain to it, but they could pertain to T. rogersi or to a hitherto No. 29 Texas Memorial Museum Bulletin 38 Table 16. Measurements in mm and ratios of dimensions of skulls of Trimerorhachis. Abbreviations; FMNH—Field Museum of Natural History; UC-University of Chicago; Fq—frontal length; Ina w -internarial width; loi-interorbital length; loyy-interorbital width; Ipj—interparietal length; MCZ—Museum of Comparative Zoology; oW—orbital0 W —orbital width; O-Sl—orbitosnout length; depth (posterior); Ski-skull length; Sk w —skidl Paj—parietal length; Pa w -parietal widtli; Pi-Oj-pineal orbital length; width (posterior); USNM—United States National Museum. Table 16a. Measurements of T. Specimen FMNH UC668 FMNH L1C667 MCZ 1105 USNM Trirnerorachis insignis as in Olson (1955) with USNM specimen added. Ski Skw Sk d 175 110 105 220 150 112 118 240 25? 0-Si Ow 67 30 39 75 _ 22 21 14 40 — 59 Table 16b. Measurements in mm of the skull of USNM Trimerorhachis cf. T. mesops, as in Olson (1955). Ski Sk w Fri Pai Iow O-Sj Ina w Pa w Pi-Oi IPl 163 156 46 50 27 68 22 37 38 16 Table 16c. Ratios of various measurements and Skj as above. Specimen FMNH UC668 FMNH UC667 MCZ 1105 USNM Trimerorhachis USNM Trimerorhachis cf. T. Mesops Skw /Skj Sk d/Ski 0.86 1.02 0.97 0.26 0.20 1.09 0.27 0.97 unknown species. The skull depth to skull length ratio (Fig. 9) is closer to that of T. rogersi than of T. insignis. In addition, the specimen discussed in the following section reduces the degree of confidence with which the fragmentary remains can be referred to T. insignis. Trimerorhachis cf. T. mesops Cope 1896 A single skull with a well-preserved left side and partially preserved right side has come from the siltstones at the Stamford locality. It is part of the United States National Museum collections and is being studied by Nicholas Hotton 111. Measurements for this skull and ratios of important dimensions are entered in Table 16b and c for comparison with those of T. insignis. Although in general size and shape this specimen falls well within the range of T. insignis, two notable differences exist. First, it appears to lack an intertemporal hone. This element is well developed and readily identifiable in most skulls of Trimerorhachis. More importantly, the ratio of orbito-snout length in the National Museum specimen is 0.48, whereas in T. insignis it ranges from 0.27 to 0,34. The orbits in the former are situated well posterior to the position that they occupy in T. insignis. The orbital position in T. mesops from the Arroyo Formation, known from only a few specimens, is comparable to that in the National Museum specimen. a — - Ow/Skj O-Si/Ski 0.20 0.14 0.18 0.33 0.27 0.37 0.34 0.14 0.48 _ Except for lack of an intertemporal bone, the specimen has a dermal skull pattern very similar to that of Trimerorhach is. One other very large skull from the Vale Formation of Knox County (Olson, 1955) also lacks this bone whose presence is questionable in a few other specimens. The meaning of this feature and the significance of the position of the orbits as far as taxonomy is concerned are uncertain. The specimen might be a large individual of T. mesops, but data are insufficient for making such an assignment. OTHER SITES To date, approximately 60 sets of exposures have been examined in addition to those discussed in the preceding pages. They lie to the south of the Clear Fork of the Brazos River which flows north of Abilene and east of Stamford, Texas. Many of these are indicated on the maps in Figure 3. Except for some of the most easterly and southerly sites, the sections all include rocks of the Vale Formation, South of the town of Winters in Runnels County, the sediments are predominantly marine, consisting of limestones and red and gray shales. To the west of Winters a thin wedge of red Vale siltstone and clay crops out and can be traced in isolated exposures to the south. fifteen of the Vale localities have yielded fragments of vertebrates, as indicated on the map. Dimetrodon, Diplocaulus, Gnathorhiza, Eryops, and 1982 The Vale Formation: Vertebrates and Paleoecology Ctenacanthus have been identified. Lysorophus, an aestivating amphibian abundant in outcrops to the north, has not been found. Of the five genera, Dirnetrodon and Diplocaulus are the most common and Ctenacanthus, Gnathorhiza, and Eryops are rare. The remains have come from mudstones, siltstones, and small lenses of conglomerate. No complete skeletons or skulls have been found. Several of the finds, as keyed on the detailed map of Figure 3, were made by David Berman who generously supplied a copy of his field notes to aid in the preparation of this paper. It will be noted on the map that three of the fossiliferous sites lie close to the Sid McAdams locality, and five cluster around the area of the Blackwood Ranch near Buffalo Gap, Texas. Exposures in all instances are small and scattered. Further exploration is planned, and of course it is hoped that additional, more productive localities will be found. At present the small fossil sites noted by Berman do not add to an understanding of the evolution of the vertebrates during the time of deposition of the Vale Formation. The exposed sections are mostly restricted both laterally and vertically. The absence of extensive clay-gall conglomerates derived from local sediments, common in tire middle and upper Vale Formation in Knox, Foard, and Baylor Counties, Texas, indicates that to the south there was less seasonality attended by intense periods of rainfall and intervening periods of drying. Conditions appear to have been somewhat similar to those of the lower Vale deposition to the north, with major difference being less energetic stream action to the south. The uppermost Vale beds over this area are composed primarily of red clays, siltstones, and calcareous layers. They presage the coming of the marine conditions during which the Bullwagon Dolomite of the Choza Formation was laid down. This is in contrast with conditions of deposition of the uppermost Vale and lower Choza deposits to the north, where the transition is marked by increased increments of evaporites, especially well shown in Knox County, Texas. The genera present in these “other sites” are übiquitous throughout the Arroyo and Vale Formations and continue into the Choza. Their highly fragmentary condition and the absence of concentrations suggest that they were washed into the sites of deposition and that the animals may have lived at some distance from where their remains have been found. The highly fossiliferous localities of the Vale Formation are all associated with well-defined stream systems and lake beds. These appear to represent the primary habitat areas during Vale deposition. 39 Presumably they provided the sources of the scattered remains which were carried into deposits away from the places where the faunas were concentrated. GEOGRAPHIC AND TEMPORAL MODIFICATIONS OF THE VALE FAUNA Environmental changes over the north-south of the Vale Formation and in vertical sections along the strike offer an unusual opportunity to examine the coordinate effects of time, space, and environment on the faunas preserved in the rocks over a relatively short time span. This is the ultimate aim of the study of this formation. The work reported in the earlier part of this paper and interpreted in this section is part of a continuing program. Although collecting sites, stratigraphic analyses, and the fossils known to date are insufficient for a full analysis, some tentative conclusions can now be made. extent The general chart showing temporal distributions of genera and species of vertebrates in the Arroyo and Vale Formations, omitting the Arroyo taxa not present in the Vale, was presented earlier (Fig. 6). Although based on the northern Vale assemblages, the chart will serve as a general summary and point up the marked similarity of the dominant species of the two formations. What is not shown are the changes that took place during the deposition of the Arroyo Formation, primarily the reduction in frequency and loss of a number of species from the base to the top. This is most important in interpretations of the Vale faunas. Most striking is the loss of Edaphosaurus, a genus with a heritage extending well back into the Upper Carboniferous. Diadectes, another representative of an ancient lineage, was severely reduced in frequency. A substantial number of other genera typical of the lower Clear Fork, among them Euryodus, Cardiocephalus, Broiliellus and other dissorophids Trematops and Labidosaurus, have not been found in the uppermost Arroyo beds. These trends and changes during the Arroyo are seen mainly in collections made in Baylor and Wilbarger Counties, Texas, where collecting has been carried out for over 100 years. To the south, extending as far as Runnels County, the sporadic finds in the .Arroyo beds are too incomplete for a similar analysis. In these beds, as throughout the Arroyo, Dirnetrodon and Diplocaulus make up most of the small samples, along with occasional remains of Xenacanthus, Gnathorhiza, Eryops, and Diadectes. More intensive collecting of the rather limited and scattered sites should produce more definitive samples. 40 Texas Memorial Museum Bulletin THE ARROYO-VALE BOUNDARY and the Lower Part of THE VALE FORMATION Because the Arroyo-Vale boundary, as explained in the introductory section and summarized below, was fixed on the basis of a convenient marker bed—the Standpipe Limestone—which separates a section of red siltstone from underlying predominantly marine beds, there is no a priori reason to expect a faunal change from thelower to the higher formation. The assignment was made on the basis of a section in Runnels County, Texas, where the Standpipe Limestone was designated as the uppermost bed of the Arroyo Formation. There the Vale Formation is a thin wedge of red clastic sediment. In the northern area, however, the base of the Vale Formation was designated at the level where extensive sandstones and conglomerates replaced a series of even-bedded clays and siltstone. The change indicated that after a return to a high flow regime following a period of deposition in quiet or slowrunning water, a faunal change might be expected. This level was taken to be the approximate time equivalent of the deposition of the Standpipe Limestone (Olson, 1958). Stratigraphic studies to date have not physically bridged the geographic gap between the northern and southern outcrop areas.lt has not been clearly demonstrated that the base of the Vale Formation as designated in the north is the precise time equivalent of that marked by the top of the Standpipe Limestone, but it is assumed that this correlation is approximately correct. Lower Part of the vale Formation Southern outcrop.— outcrop.—The area from which vertebrates have been derived in the southern part of the Vale Formation appears to have been heavily vegetated and to have lain close to the margin of the Leonard Seaway. The climate was subtropical to warm-temperate and the general conditions may be considered paralic with a fluctuating shoreline. Brief marine incursions into the vertebrate-producing area appear to have taken place. The only area rich in vertebrate fossils is in southern Taylor County, Texas, at the Sid McAdams locality. The deposilional environments in which vertebrate remains were preserved include ponds, streams, and lakes, some of which were ephemeral. Vertebrates deposited in the standing water sediments appear to have been buried in situ. Other sites in the southern Vale area have yielded only fragmentary remains from local carbonate conglomerates and flood-plain siltstones. No. 29 The greatest number of individuals was obtained by quarrying operations from gray silty sediments formed in an ephemeral lake at the Sid McAdams locality. Three genera, and perhaps a fourth, were obtained: Dimetrodon, Seymouria, tOphiacodon, and Diadectes, the last either from the quarry or close to it. The remaining 8 genera from this locality came from stream deposits and from the siltstones formed in a permanent lake. Although this assemblage suggests considerable ecological diversity, its closest physical analogue in the Arroyo Formation appears to be the restricted fauna of the Craddock hone bed north of Seymour, Baylor County, Texas. Table 2 gives a faunal comparison of the Sid McAdams locality with the Craddock bone bed, the fauna of the latter based on Williston (1911), Romer (1928), and Langston (personal communication). Also shown for comparison are the vertebrates from the lower Vale of the northern area from Olson (1958). Dimetrodon is numerically dominant and Seymouria is represented by several specimens in the Sid McAdams collection. Both occur in somewhat similar proportions in the Craddock bone bed. The curious, single specimen of represents a genus not known at the Craddock locality. A small ophiacodont, Varanosaurus, occurs there, hut the affinities of the Sid McAdams humerus are definitely with Ophiacodon and not Varanosaurus. Diadectes, known from a single tooth at the Sid McAdams locality, is not recorded from the Craddock hone bed. Some of the specimens in the Craddock collections are eroded, suggesting that they were transported (Williston, 1911). They probably came from marginal environments similar to those tapped by the various stream deposits at the Sid McAdams locality. If the additional laxa from this locality are added to the pondquarry taxa, resemblance to the Craddock assemblage is quite strong. Only Lawnia taylorensis and Gnathorhiza are absent from the Craddock deposits and neither can he considered of much significance, inasmuch as Arroyo paleoniscoids have been insufficiently studied for comparisons, and the introduction of Gnathorhiza, an aestivatinglungfish, intolake deposits such as those at Craddock would clearly he accidental. The presence of ? Ophiacodon in the Vale lake may be significant as representing an ancient lineage otherwise not known since the end of deposition of the Clyde Formation which underlies the Arroyo formation in Baylor County, Texas. ? Ophiacodon, however, would appear to be an ecological equivalent of Varanosaurus, which is rare in the Arroyo. The presence of Gnathorhiza dikeloda at the Sid McAdams locality marks a positive difference from 1982 The Vale Formation: Vertebrates and Paleoecology the Arroyo fauna, from which only G. serrata is known. Gnathorhiza dikeloda has been considered an “index fossil” of the post-Arroyo Clear Fork in the northern area. It is important that a substantial number of letrapods that are moderately abundant in the .Arroyo Formation are absent in the lower Vale, especially th ose that might be expected to have lived in the environments in which the lower Vale deposits were formed. Foremost among these is Edaphosaurus. This genus occurs in deposits from streams, lake margins, and flood plains. Occasionally in the Wichita beds it occurs in concentrations in swamp deposits. Oddly, it also is unrecorded at the Craddock bone bed where it might be expected on the basis of physical environment. Edaphosaurus is fairly abundant in the Wichita Group of Texas and is known also from the Dunkard, Lower Permian, of eastern United States and from equivalent and older beds in both Europe and the United States. It is present to some degree in the early Arroyo sediments but becomes increasingly less frequent in the later deposits. This was a time of slowly increasing aridity and seasonality, and the reduction in numbers is likely to be associated with this change. The absence of Edaphosaurus in lower Vale and all later deposits suggests that the trends noted in the Arroyo led to an actual extinction over the areas sampled. Diadectes, also with an ancient heritage, shows a similar pattern in the Arroyo. A specimen in the southern lower Vale and one in the north show that it did not go extinct at this time hut, as noted earlier, persisted at least into the time of the middle Vale deposition. Armored dissorophids, such as Broiliellus, trematopsids and gymnarthrids are absent as far as is known from the Vale fauna. These were terrestrial animals, except possibly the gymnarthrids, and the reasons for their waning are not at all clear. An unarmored dissorophid, Tersomius, extends into the Vale Formation in moderate numbers. Among the reptiles, Labidosaurus, common in lower and middle Arroyo lake deposits, is not known from the Vale Formation. Casea, Varanops, and the dissorophid Cacops, which occur at a single site near the Arroyo-Vale boundary in the north, are not known in the southern area. However, they have been interpreted as pertaining to another chronofauna (Olson, 1971) and their absence is to be expected. From these considerations, we can find no critical positive differences between the vertebrate assemblages of the Arroyo Formation and the southern Vale Formation, except possibly the presence of 41 Gnathorhiza dikeloda in the latter. The primary difference between the faunas as a whole is the absence of a number of genera, a result of changes that went on during the deposition of the Arroyo. These changes appear to be concomitant with slowly increasing aridity marked by some increase in the seasonality of the rainfall. The conditions of deposition of the southern lower Vale were very similar to those during deposition of the upper part of the Arroyo, except that deposition of the latter look place considerably farther from the sea margin. The low, coastal environment of the southern Vale is best matched by the pond and pond-margin environment exemplified by the Craddock bone bed in the Arroyo Formation. It is, of course, important that in the work to date the sampled area of the Arroyo Formation is much greater than that of the southern lower Vale. Similarly, the sample from the lower Vale is from a rather restricted environment. The apparent depauperate nature of the Vale taxonomic list may be partly a result of these factors. On the other hand, the trends seen during Arroyo deposition suggest that what the samples show may be a fairly accurate representation of what in fact was there. outcrop.— Occurrences of fossils in the Northern outcrop.— of the lower Vale between the Sid McAdams beds and the localities locality in Baylor, Knox, and Foard Counties are few, and the specimens are fragmentary. It is hoped that future work may remedy this lack, hut chances do not look promising because of the flat topography and limited exposures. The principal genera found are Dimetrodon and Diplocaulus, with rare remains of Eryops, Gnathorhiza, and Ctenacanthus. The known fossils provide no basis whatsoever for separating the lower Vale from the upper Arroyo faunas. Thus comparisons must be made between the two well-separated fossiliferous areas, one near the shoreline and the other 180 to 200 km remote from it. Within determinable limits the climates were only moderately different, with rainfall somewhat more intense and periodic in the north. The topography in the north probably was somewhat rougher and elevation above sea level somewhat greater. These conclusions are based upon the evidence of high-energy stream action and short periods of intense flooding that has been found in the thick conglomerate channels and coarse overhank deposits. The principal features of the lower Vale deposits in the north have been described earlier, pages 6-11, and the genera and species present are noted in Figure 6. In general the fauna represents a continuation of that of the .Arroyo, with loss of several important species, much as that to the south. The most Texas Memorial Museum Bulletin significant positive faunal difference between the north and the south is the presence of Labidosaurikos in the former. No obvious source of this fairly common reptile is known among the Arroyo species. Presumably it evolved elsewhere from a Captorhinuslike ancestor, hut not directly from Captorhinus aguti. Its presumed invasion at the onset of the Vale deposition did not, as far as is known, extend to the more southerly areas at this time. Diplocaulus magnicornis is an abundant fossil in the northern lowermost Vale deposits. Somewhat later it is replaced hy D. recurvatus. Single specimens and poorly preserved specimens of Diplocaulus, as discussed on page 37, cannot he assigned to one or the other of these species, rendering useless for this purpose distributional data based on small samples. In the very lowest northern Vale deposits, essentially at the transition level between Arroyo and Vale, Trematops seltini is known from a single specimen. No evidence of Seymouria baylorensis, Araeoscelis sp., or Tersomius sp. has been found in the north, although all three are known from the Sid McAdams locality. Captorhinus aguti likewise has not been found, but it occurs in the middle Vale of the northern area. Overall, the southern lower Vale fauna has a more consistent Arroyo cast than does the fauna from the northern beds. A sharp change in sedimentation, marked hy the increasing importance of coarse conglomerates in the north, contrasts with the south, where essentially no change from the earlier terrestrial upper Arroyo beds can be seen. Climatic change, which intensified during the deposition of the later Vale and Choza beds, appears to have begun earlier in the north than in the south. Middle and upper parts of the Vale Formation Northern outcrops Formation.— the outcrops of Vale Formation.—ln northern areas, where the middle and upper Vale occur in continuous sequence, no notable changes in the faunas occur throughout the sequence. Since 1958 (Olson, 1958) Richard Seltin of Michigan State University (Seltin, 1959, 1972, unpublished) has discovered and reported on a number of additional sites in the northern outcrops of the Vale Formation. Preliminary analyses suggest that the species composition of the fauna as outlined in 1958 are essentially correct, but relative abundances, which have not been considered, may differ. As in the lower Vale, Xenacanthus platypternus and Dirnetrodon giganhomogenes are the most abundant species. Diplocaulus recurvatus has replaced D. magnicornis and is abundant. All of the specifically assignable material belongs to the former species. No. 29 Trimerorhachis insignis continues to he relatively abundant as well. Captorhinus aguti is present although not common. Labidosaurikos barkeri is somewhat increased in relative abundance. A new genus and species, Captorhinikos valensis, is added to the fauna near the base of the middle Vale and becomes somewhat more abundant in later deposits. The batrachosaur Waggoneria texensis, an odd seymouriamorph, is known from a few specimens in the middle Vale. Species associated with temporary lakes and ponds, Lysorophus tricarinatus, Gnathorhiza serrata, and G. dikeloda are present. The first and last are abundant hut G. serrata is uncommon. A small keraterpetonlid, Peronedon primus, occurs in abundance in one locality, associated with burrows of Gnathorhiza dikeloda and Lysorophus tricarinatus. This species is also known from the lower part of the Hennessey Group in Oklahoma, in beds equivalent in age to those of the Arroyo, and also in the Fairmont Shale of the Hennessey Group in a site near Norman, Oklahoma (Olson, 1967; Haglund, 1977). The absence of Diplocaulus magnicornis, Diadectes sp., and Seymouria baylorensis is probably an expression of a faunal change and not merely sampling insufficiency, to judge from the faunal trends noted for earlier beds and the continued absence of those species in the Choza. Eryops sp. is known from a skull in the middle Vale and it probably pertains to E. megacephalus. This it was not a major constituent of the fauna is shown by the fact that its readily recognized vertebrae are rarely found in assemblages from large concentrations of vertebrates. It was abundant in the lowest Vale beds and thereafter much reduced in relative numbers through the middle Vale, after which it has not been definitely identified. The segment of the Clear Fork chronofauna (Olson, 1971) found in the middle and upper Vale beds in the northern area reflects climatic changes toward increasing dryness and seasonal distribution of the rainfall. The strong increase in relative numbers of aestivators and the reduction or absence of primarily lake-dwelling species such as Diplocaulus magnicornis and Eryops megacephalus testify to this. There was, however, sufficient permanent water for such aquatic vertebrates as Xenacanthus platypternus and Trimerorhachis insignis to thrive. Of the strictly terrestrial vertebrates, Captorhinus aguti, Captorhinikos valensis, and Labidosaurikos barkeri formed a land-living suite of animals well suited to moderately xeric conditions. Only the last was abundant. The taxonomy and relationships of these forms along with Captorhinoides valensis (Bolt and DeMar, 1978) are being restudied by John 801 l 42 1982 The I ale Formation: Vertebrates and Paleoecology and Robert DeMar with additional materials supplied by Seltin, and these studies may change the current concepts of the genera and their species, which were based on limited samples and rather fragmentary materials. Two specimens of Casea nicholsi and two unassigned specimens from the upper Vale require note. The well-preserved specimens of C. nicholsi (Olson, 1954a, 1958) were recovered from a single nodule. They were not associated with any other vertebrates, and not a trace of this species has been found elsewhere. A large-toothed plate, FMNII UR 29, and a very large, badly preserved scapulocoracoid, FMNH UR 268. are also unique (Olson, 1965). Casea, as noted on page 11, has come from the Cacops bone bed near the Arroyo boundary with the \ ale, and it is known from the single specimen from the middle Choza. All of the specimens appear to be erratics carried in from another chronofauna. The toothed plate, when first described, was tentatively referred to the Edaphosauria. With information on other faunas, particularly those of the Guadalupian San Angelo Formation, now available, it seems clear that this specimen is much closer to Rothianiscus from the San Angelo and temporally equivalent parts of the Chickasha of Oklahoma. It has eight welldeveloped rows of teeth, more than any known specimen of Rothianiscus, although R. robustus has six rows and an incipient seventh. The toothed plate, although considerably smaller, is most similar in shape and dentition, to that of Moradisaurus, a giant captorbinomorph from the Republic of Niger (Taquet, 1969). It appears that there existed during the time of deposition of the Vale Formation a chronofauna which included Rothianiscusdike animals and presumably this was the same fauna that included Casea. The large scapulocoracoid is not well enough preserved to allow any statement on its affinities. It is not a normal member of the Vale portion of the Clear Fork chronofauna. Central and southern outcrop regions.—Two areas, as described on pages 27-38. have produced significant collections of middle Vale vertebrates from south of the Clear Fork of the Brazos River. One is near Buffalo Gap in Taylor County, Texas, and the other is near Stamford in southern Haskell County, Texas. The Stamford site has yielded Diplocaulus recurvatus, Trimerorhachis ? msignis, and Trimerorhachis cf. T. mesops from siltstone channel deposits. Adjacent soft, silty clays have produced a single tooth of Xenacanthus and fragments of Lysorophus. The significance of the site for tbe present study is the presence of D. recurvatus, which is characteristic of the middle and upper Vale Formation to the north. As in the north, all specimens occur in river-channel deposits. 43 The absence of Dimetrodon illustrates tbe highly selective nature of the deposition. Dimetrodon and Diplocaulus, not determinable to species, occur in hard, gray green sandstone exposures on the property of Mr. Wedeking, about 3 km north of Abilene, and similar scrap hascome from Dudley’s Creek in Haskell County (Fig. 3a). The beds are approximate time equivalents of the Stamford deposits. The area southeast of Buffalo Gap, especially the exposures at tbe Blackwood locality, has yielded excellent fossils as detailed on pages 30-35. Several points of ecological importance emerge from the study of these fossils. First, a great majority, and all of the well-preserved specimens, have been found along or near the courses of small rivers. Those few found away from such sites are fragmentary and predominantly Dimetrodon and Diplocaulus. Lysorophus, Eryops, and Gnathorhiza are encountered occasionally. The badly preserved, very small specimen of Labidosaurikos from the Blackwood Ranch indicates that by the time the middle Vale beds were being formed, conditions had developed in which this genus could live. Its earlier and continuing existence to the north suggests that these conditions involve dry, seasonal climates. Such conditions did not hold, it would seem, during formation of the beds of the lower Vale in the area of the Sid McAdams locality, some 22 km to the south. The species of the BlackwoodLabidosaurikos. based on size audits adulthood, is probably different form the northern L. barkeri. It was perhaps adapted to somewhat less advantageous climatic and trophic conditions, but of course such speculation based on a single individual is at best hazardous. Seymouria grandis is not known elsewhere in the Vale. It is a large species probably derived from S. baylorensis. It also occurs in beds of approximately the same age as those at the Blackwood site in the Fairmont Shale of the Hennessey Group near Crescent, Oklahoma (Olson, 1979b). There, it is associated with Diplocaulus sp. indet., Labidosaurikos meachami, and Dimetrodon giganhomogenes. To the west of Crescent are marginal marine beds of equivalent age, and in the Bison area to the northwest are evaporites, suggesting at least incipient aridity in the region. Between Crescent and Buffalo Gap lay the Wichita positive area, but in spile of the seeming discontinuity, similar faunas appear to have existed in the two areas. The Fairmont Shale, however, also includes specimens of the caseid Cotylorhynchus romeri, apparently floated in as carcasses (Stovall, Price, Romer, 1966). This genus has not been found in the Taylor County, Texas, sites. 44 Texas Memorial Museum Bulletin The specifically assignable specimens of Diplocaulus in the Blackwood locality have all been placed in D. recurvatus, but these occur in still-water deposits rather than the usual stream habitat, as discussed on page 29. Two specimens of Diadectes at this locality show that the genus persisted to the south, although extensive collecting has not revealed it above the very lowest Vale sites to the north. Its habitat is not determinable. Xenacanthus platypternus andDimetrodon giganhomogenes are not good climatic indicators. Both taxa require persistent water; in lakes or streams, as a source of food, and—for the shark—as a habitat. Their wide range and abundance, however, indicate a broad tolerance for differing ecological conditions. The relative abundance of Eryops megacephalus at the Blackwood locality shows a persistence of more humid and equable conditions than existed to the north where the species was rare and diminished rapidly in relative numbers after deposition of the initial Vale beds. It is also of interest that this genus does not occur in the Sid McAdams locality where sedimentary conditions suggest an appropriate environment. This may, of course, result from the restricted suite of environments sampled, as seems to apply in the case of the Craddock bone bed. From the available information, it appears that the climate during the time of deposition of the middle Vale beds in the Buffalo Gap area was less seasonal than that to the north and that there was sufficient moisture to support such large, aquatic amphibians as Eryops. On the other hand, the presence of such aestivators as Lysorophus and Gnathorhiza precludes the interpretation that this was a truly humid region with ample rainfall throughout the year. As yet, nothing can be said about the faunas of the highest Vale south of the Brazos River. Whether the xeric conditions of the north continued the gradual shift to the south, as might be anticipated, is not determinable from the sediments or fossils now known. SUMMARY AND DISCUSSION Sedimentological and faunal analyses have led to the establishment of a complex of hypotheses concerning the changes of the environment and ecology in space and time during the deposition of the Vale Formation. The hypothesis to date is based on a less-than-desirable amount of data. In cross terms the hypothesis is as follows. The climate during the first half of the deposition of the Arroyo Formation was warm, moderately humid, and only mildly seasonal over the areas where terrestrial deposits have been studied. It undoubtedly varied from the north to the south, away from the No. 29 shifting seaway margin, but as yet no faunal effects have been observed. This may be the result of poor knowledge of the faunas to the south. Gradually, the amount of rainfall was reduced and seasonality increased. Concomitantly, a wide range of species, abundant in the lower Arroyo, decreased in numbers, and some disappeared. A few, aestivatorsin particular, increased in relative numbers. The oldest beds of the Vale Formation contain evidence of climatic changes trending to greater seasonality of rainfall. These conditions persisted and intensified throughout the lime of the Vale deposition in the northern area. The more southerly outcrops of the Vale Formation initially showed less change and remained similar to the upper portions of the Arroyo, both faunally and climatologically. During the deposition of the Vale, as known in central Taylor and southern Haskell Counties, Texas, a shift to greater seasonality set in and this was mildly expressed by the lime of the formation of the middle Vale. The sediments of the upper Vale, which have not yet yielded vertebrate fossils, suggest that in the southern and central areas conditions similar to those in Taylor and Haskell Counties persisted nearly to the time of deposition of the Bullwagon Dolomite. This, however, remains to be determined by detailed study. Faunal changes between the lower and middle Vale in the northern area correlate well with the sedimentological changes and the climatic conditions that they imply. Lesser changes occur to the south, but some genera and species first noted in early beds to the north appear in the middle Vale, suggesting a gradual penetration as climate permitted. We assume that in the case of Labidosaurikos this was by speciation, whereas in the case of Diplocaulus recurvatus it was probably by expansion of the range of the species. Eryops megacephalus, which was relatively rare after the onset of Vale deposition in the north, persisted under more equable conditions to the south with abundance comparable to that recorded earlier during formation of the beds of the Arroyo. Similarly, the aestivators Lysorophus and Gnathorhiza, both in their scattered occurrences and only moderate abundance, reflect persistence of conditions differing little from those of the uppermost Arroyo. Some species that developed under more xeric conditions in the northern area were able to penetrate the more southerly areas as suitable environments began to develop. This sketchy history covers a short span of geological time and a limited geographic area. It shows the potential of the region studied for a fairly detailed knowledge of short-term evolutionary changes and their relationships to modifying environments. 1982 The Vale Formation: Vertebrates and Paleoecology Pub. 146:153-175. Lewis, G.E., and P.P. Vaughn, 1965. Early Permian vertebrates from the Cutler formation of the Placerville area, Colorado. U.S. Geol. Surv. Prof. Pap. 503-C. Mamay, S.H., 1976. Paleozoic origin of the cycads. U.S. Geol. Surv. Prof. Pap. 934. Marsh, 0.C., 1878. Notice of 195 , new fossil reptiles. Amer. J. Sci. 15:409-411. Mead, J.G., 1971. Paleontology of the Sid McAdams locality (Permian) southern Taylor County, Texas. Unpublished master’s thesis, Univ. Texas—Austin. Olson, E.C., 1947. The family Diadectidae and its bearing on the classification of reptiles. Fieldiana: Geol. 11:2-53. , 1948. A preliminary report of vertebrates from the Permian Vale formation of Texas. J. Geol. 56:186-198. a. Fauna of the Vale and Choza, 1-5: 1. A new family of Parareptilia; 2. A new captorhinomorph reptile; 3. Lungfish in the Vale; 4. The skull of Gnathorhiza dikeloda Olson; 5. An eryopid amphibian. Fieldiana; Geol. 10:89-218. , 1951b. Diplocaulus, a study in growth and variation. Fieldiana: Geol. 11:57-154. The evolution of a Permian vertebrate chronofauna. Evolution 6:181-196. , 1952b. Fauna of the Vale and Choza, 6. Diplocaulus. Fieldiana; Geol. 10:147-166. 1953. Integrating factors in amphibian skulls. J. Geol. 1952 , a. ——, 61:557-568. a. 1954 , 1954 , Fauna of the Vale and Choza, 7. Pelycosauria, family Caseidae. Fieldiana: Geol. 10:193-204. , 1954b. Fauna of the Vale and Choza, 8. Pelycosauria, Dimetrodon. Fieldiana; Geol. 10:205-210. c. Fauna of the Vale and Choza, 9. Captorhinomorpha. Fieldiana: Geol. 10:211-218. , 1955. Fauna of the ValKand Choza, 10. Trimerorhachis, including a revision of pre-Vale species. Fieldiana: Geol. 10:225-312. a. Fauna of the Vale and Choza, 11. Lysorophus: Vale and Choza; Diplocaulus and Eryopidae: Choza. Fieldiana; Geol. 10:313-322. , 1956b. Fauna of the Vale and Choza, 13. Diadectes, Xenacanthus and specimens of uncertain affinities. Fieldiana; Geol. 10:329-334. , 1957. Size-frequency distributions in samples of extinct organisms. J. Geol. 65:309-333. , 1958. Fauna of the Vale and Choza, 14. Summary, review and integration of the geology and faunas. Fieldiana: Geol. 10:379-448. , 1962. Late Permian terrestrial vertebrates, U.S.A. and U.S.S.R. Amer. Phil. Soc. Trans. 52, pt. 2. , 1967. Early Permian vertebrates from Oklahoma. Okla. Geol. Surv. Circ. 74. , 1970. New and littleknown genera and species from the Lower Permian of Oklahoma. Fieldiana: Geol. 18:359-434. , 1971. Vertebrate Paleozoology. Wiley Intersci., N.Y. a. Aspects of the biology of Trimerorhachis (Amphibia: Temnospondyli). J. Paleontol. 53:1-17. 1956 , Barnes, V.E., 1972. Geological Atlas of Texas, Abilene Sheet. Bur. Econ. Geol., UT—Austin. Beede, J.W., and V.V. Waite, 1918. The geology of Runnels County. Univ. Texas Pub. 1810. Bolt, R., and R. DeMar, 1978. The taxonomic position of Captorhinoides valensis Olson (Reptilia; Captorhinomorpha). J. Paleontol. 52:934-937. Broili, F., 1904. Permische Stegocephalen und Reptilien aus Texas. Paleontolographie 51:1-49,51-120. Brooks, H.K., 1962. The Paleozoic Eumalacostraca of North America. Bull. Amer. Paleontol. 44:163-357. Broom, R., 1913. On the structure and affinities of Bolosaurus. Amer. Mus. Natur. Hist. Bull. 32:506-516. Case, E.C., 1907. Revision of the Pelycosauria of North America. Carnegie Inst. Wash. Pub. 55. , 1910. New or little known reptiles and amphibians from the Permian (?) of Texas. Amer. Mus. Natur. Hist. Bull. 28:163-181. , 1911. Revision of the Amphibia and Pisces of the Permian of North America. Carnegie Inst. Wash. Pub. 146. Cope, E.D., 1877. Descriptions of extinct Vertebrata from the Permian and Triassic formations of the United States. Amer. Phil. Soc. Proc. 17:182-193. , 1878. Description of extinct Batrachia and Reptilia from the Permian formation of Texas. Amer. Phil. Soc. Proc. 17:505-530. , 1882. Third contribution to the history of the Vertebrata of the Permian formation of Texas. Amer. Phil. Soc. Proc. 20:447-461. , 1883. Fourth contribution to the history of Vertebrata from the Permian formation of Texas. Amer. Phil. Soc. Proc. 20:627-633. , 1884. On the structure of the skull in the elasmobranch genus Didymodus. Amer. Phil. Soc. Proc. 21:572-590. , 1896. Second contribution to the history of the Cotylosauria. Amer. Phil. Soc. Proc. 35:137. Currie. P.J., 1978. The orthometric linear unit. J. Paleontol. 52:964-971. Dalquest, W.W., and S.H. Mamay, 1963. A remarkable concentration of Permian amphibian remains in Haskell County, Texas. J. Geol. 17:641-644. Daly, E., 1973. A Lower Permian vertebrate fauna from southern Oklahoma. J. Paleontol. 47:562-589. Fox, R.C., and M.C. Bowman, 1966. Osteology and relationships of Captorhinus aguti (Cope) (Reptilia: Captorhinomorpha). Univ. Kansas Paleontol. Contrib.: Vertebrata 11. Gould, S.J., 1967. Evolutionary patterns in Pelycosaurian reptiles. A factor-analytical study. Evolution 21:385-401. Haglund, T., 1977. New occurrences and paleoecology of Peronedon Olson (Nectridea). J. Paleontol. 51:982-985. Henderson, G.G., 1928. The geology of Tom Green County. Univ. Texas Bull. 2807. Hotton, N. HI, 1952. Jaws and teeth of American xenacanth sharks. J. Paleontol. 26:489-500. Hussakof, L., 1911. The Permian fishes of North America. In Case, E.C., Revision of the Amphibia and Pisces of the Permian of North America. Carnegie Inst. Wash. 197 , REFERENCES 45 , 1979b. Seymouria grandis n. sp. (Batrachosauria: Amphibia) from the middle Clear Fork (Permian) of Oklahoma and Texas. J. Paleontol. 53(3):720-728. and H. Barghusen, 1962. Vertebrates from the Flowerpot formation, Permian of Oklahoma, Pt. II: The osteology of Captorhinikos chozaensis Olson. Okla. Geol. Surv. Circ. 59:5-48. , and K. Bolles, 1975. Permo-Carboniferous fresh water burrows. Fieldiana: Geol. 33:271-290. Parrish, W.C., 1978. Paleoenvironmental analysis of a Lower Permian bonebed and adjacent sediments, Wichita County, Texas. Palaeogeogr., Palaeoclimatol., Palaeoecol. 24: —, 209-237. 46 Texas Memorial Museum Bulletin Rayner, D., 1971. Data on the environment and preservation of the Late Paleozoic tetrapods. Yorkshire Geol. Soc. Proc. 38:437-495. Read, C.8., and S.H. Mamay, 1964. Upper Paleozoic floral zones and floral provinces of the United States. U.S. Geol. Surv. Prof. Pap. 454K. Romer, A.S., 1928. Vertebrate faunal horizons in the Texas Permo-Carboniferous red beds. Univ. Texas Bull. 2801: 67-107. Romer, A.S., 1937. New genera and species of pelycosaurian reptiles. New Engl. Zool. Club Proc. 16:89-96. Romer, A.S., 1958. The Texas Pemian redbeds, and their vertebrate fauna. In Studies on fossil vertebrates, T.S. Westoll, ed., p. 157-179. London, 1958. Romer, A.S., and E.C. Olson, 1954. Aestivation in a Permian lungfish. Breviora 30. Romer, A.S., and L.W. Price, 1940. Review of the Pelycosauria. Geol. Soc. Amer. Spec. Pap. 28. Sellards, E.H., W.S. Adkins, and F.B. Plummer, 1932. The Geology of Texas, I. Stratigraphy. Univ. Texas Bull. 3232. Seltin, R.J., 1959. A review of the family Captorhinidae. Fieldiana: Geol. 10:461-509. Seltin, R.J., 1972. Report of new fossil localities in the Vale formation (Early Permian) of Texas. Unpublished. Stovall, J.W., 1950. A new cotylosaur from north central Oklahoma. Amer. J. Sci. 248:46-54. Stovall, J.W., L.l. Price, and A.S. Romer, 1966. Cotylorhynchus romeri, a giant pelycosaur from the Permian of Oklahoma. Bull. Mus. Comp. Zool, Harvard. 135(l):l-30. No. 29 Taquet, M.P., 1969. Premiere decouverte en Afrique d’un reptile captorhinomorphe (Cotylosauria). Acad. Sci. Comptes Rendus Hebdom. des Seances, Paris. 268(D); 777-781. Tilton, J.L., 1931. Permian vertebrate tracks inWest Virginia. Geol. Soc. Amer. Bull. 42:547-556. Vaughn, P.P., 1955. The Permian reptile Araeoscelis restudied. Bull. Mus. Comp. Zool. Harvard. 113:305-467. Vaughn, P.P., 1958. A specimen of the captorhinid reptile Captorhinikos chozaensis Olson 1954, from the Hennessey formation. Lower Permian of Oklahoma. J. Geol. 66; 321-332. Williams, M.E., 1972. The origin of “spiral” coprolites. Univ. Kansas Paleontol. Contrib. 59. Williston, S.W., 1910. New Permian reptiles: rhachitomous vertebrae. J. Geol. 18:586-600. Williston, S.W., 1911. American Permian vertebrates. Univ. Chicago Press, Chicago. Wilson, J.A., 1948. Vertebrate fauna of a Permian locality in Taylor County, Texas. Bull. Geol. Soc. Amer. (Abstr.) 29:1363. Wilson, J. A., 1953. Permian vertebrates from Taylor County, Texas.]. Paleontol. 27:456-470. Wrather, W., 1917. Notes on Texas Permian. Southwest. Assoc. Petrol. Geol. Bull. 1:93-106. Zangerl, R., and E.S. Richardson, 1963. The paleoecological history of two Pennsylvanian black shales. Fieldiana; Geol. Mem. 4.