Life Sciences Contributions 1 ) Royal Ontario Museum A Protorothyridid Captorhinomorph Reptile trom the Lower Permian of Oklahoma Robert R. Reisz ROIiM ROYAL ONTARIO MUSEUM LIFE SCIENCES PUBLICATIONS INSTRUCTIONS TO AUTHORS Authors are to prepare their manuscripts carefully according to the following instructions. Failure to do so will result in the manuscript’s being returned to the author for revision. All manuscripts are considered on the understanding that if accepted they will not be offered for publication elsewhere. jm GENERAL Papers for publication are accepted from ROM staff members, Research Associates, or from researchers reporting on work done with ROM collections. In exceptional cases,monographic works on the flora and/or fauna of Ontario will be considered for publication by authors not affiliated with the ROM. Authors are expected to write clearly and concisely, and to omit all material not essential for an understanding of the main theme of the paper. . FORMAT Manuscripts are to be typed double-spaced (including captions, synonomies, literature cited, and tables) on 11’’ X 8%”’ paper with a 1%’’ margin on all sides. Three xerox copies are to be submitted to the Chairman of the Editorial Board, and the original retained by the author(s). A separate sheet is to be submitted giving author(s) names, affiliation, title of publication, series in which it is to appear, number of typed pages, number of tables, and number of figures. Manuscripts should normally be organized in the following order: Table of Contents, Abstract, Introduction, Materials and Methods, Results, Discussion, Conclusions, Summary (if paper is long), Acknowledgements, Literature Cited, and Appendices. Authors are encouraged to include foreign language translations of the Summary where appropriate. Headings of sections are to be left-justified to the text margin. The first line of the first paragraph in each new section should not be indented. Text-figures are referred to as ‘‘Fig. 1”’. Literature cited in the text is in the form ‘*Jones (1972)’’ or ‘‘(Jones, 1972)’’ or ‘‘(Smith, 1960:71-79, fig. 17)’’. . STANDARD SOURCES The primary source for decisions on format and style is A Guide for Contributors and Editors of ROM Life Sciences Publications, available from the Chairman of the Editorial Board. Otherwise, consult CBE (AIBS) Style Manual (3rd Edition). Other standard sources are as follows: for English spelling (Concise Oxford Dictionary), for Canadian place names and coordinates (Gazetteer of Canada), and for spelling of geographic names (Times [London] Atlas). . ABSTRACT All papers are preceded by a short and factual abstract, about 3 per cent as long as the text, but not longer than 400 words. The abstract is to be followed by four to six keywords enclosed in brackets. . TAXONOMY The name of a taxon is given in full in headings, where it appears for the first time, or when the name begins a paragraph. Use authority and date if appropriate, with first mention of each taxon and not thereafter. Taxonomic papers follow the layout in Life Sciences Contribution 99, particularly the synonomies. . LITERATURE CITED References in the text cite author and date and are enclosed in parentheses (Smith, 1978). Complete references are listed in alphabetical order by author at the end of the paper. When there are two or more citations for an author, the works are listed chronologically. Names of journals are not abbreviated. Consult Life Sciences Contributions beginning with 117 for correct bibliographic form. . TABLES All tables are numbered consecutively in arabic numerals in numerical order of their first mention in the text. Mark the appropriate text location of each table with a marginal notation. Each table is typed on a separate sheet. Avoid footnotes etc., to tables by building them into the title. . FIGURES All figures are numbered consecutively in arabic numerals. Component photographs or drawings are labelled sequentially in upper case letters. Mark the appropriate text location of each figure with a marginal notation. The intended reduction for figures is ideally one and a half to two times. All labelling on figures is in blue pencil and not inked or letraset. Halftones must be photographic prints of high contrast on glossy paper. Authors are to submit 10°’ x 8’’ copies with the MS and retain originals until they are requested. Figure captions are to appear grouped together on a separate page at the end of the MS. LIFE SCIENCES CONTRIBUTIONS ROYAL ONTARIO MUSEUM NUMBER 121 ROBERT R. REISZ A Protorothyridid Captorhinomorph Reptile from the Lower Permian of Oklahoma ROM ROYAL ONTARIO MUSEUM PUBLICATIONS IN LIFE SCIENCES The Royal Ontario Museum published three series in the Life Sciences: LIFE SCIENCES CONTRIBUTIONS, a numbered series of original scientific publications including monographic works. LIFE SCIENCES OCCASIONAL PAPERS, a numbered series of original scientific publications, primarily short and usually of taxonomic significance. LIFE SCIENCES MISCELLANEOUS PUBLICATIONS, an unnumbered series of publications of varied subject matter and format. All manuscripts considered for publication are subject to the scrutiny and editorial policies of the Life Sciences Editorial Board, and to review by persons outside the Museum staff who are authorities in the particular field involved. LIFE SCIENCES EDITORIAL BOARD Senior Editor: J. H. McANprRews Editor: R. D. JAmMEs Editor: C. McGowan ROBERT R. REISZ is an Assistant Professor in the Department of Zoology, Erindale College, University of Toronto, and a Research Associate of the Department of Vertebrate Palaeontology, Royal Ontario Museum. Canadian Cataloguing in Publication Data Reisz, Robert R., 1947- A protorothyridid captorhinomorph reptile from the Lower Permian of Oklahoma (Life sciences contributions ; 121 ISSN 0384-8159) Bibliography: p. ISBN 0-88854-248-8 pa. 1. Reptiles, Fossil. 2. Paleontology — Oklahoma. 3. Paleontology — Permian. I. Royal Ontario Museum. II. Title. III. Series. QE862.C7R45 567.9°2 C79-094920-2 Publication date: 11 January 1980 ISBN 0-88854-248-8 ISBN 0384-8159 © The Royal Ontario Museum, 1980 100 Queen’s Park, Toronto, Canada M5S 2C6 PRINTED AND BOUND IN CANADA BY MacKINNON-MONCUR Contents Abstract 1 Introduction 1 Abbreviations Used in the Figures 3 Description and Comparison 3 Humerus 3 Radius 7 Ulna 8 Femur 10 Tibia 11 Fibula 12 Discussion 13 Acknowledgements 15 Literature Cited 15 Digitized by the Internet Archive in 2011 with funding from University of Toronto http://www.archive.org/details/protorothyrididcOOreis A Protorothyridid Captorhinomorph Reptile from the Lower Permian of Oklahoma Abstract A new primitive captorhinomorph reptile has been found near Fort Sill, Oklahoma, in fissure fill deposits believed to be contemporaneous with the lower part of the Arroyo Formation, Clear Fork Group (Leonardian) and possibly the upper part of the Lueders Formation, Wichita Group (Wolfcampian) of northcentral Texas. This find extends the fossil record of the oldest group of reptiles, the family Protorothyrididae, into the upper part of the Lower Permian. Although many superbly preserved individual limb elements have been recovered, the lack of any specimens in articulation and the absence of diagnostic features below the familial level precludes nam- ing the animal at this time. The proportions of the limb elements and the concentration of the origins and insertions of the muscles important in the power and recovery strokes of the walking motion near the ends of these bones indicate that this reptile was an agile, lightly built animal. Introduction The Suborder Captorhinomorpha occupies a unique position in the phylogeny of reptiles. The central captorhinomorph stock, represented by the family Protoro- thyrididae, includes the oldest known reptiles and the ancestors of three of the four orders of extant reptiles (Reisz, 1977). Recent studies of Pennsylvanian (Carroll, 1964, 1969; Carroll and Baird, 1972) and Lower Permian (Olson, 1962, 1967, 1970; Fox and Bowman, 1966; Holmes, 1977; Heaton, 1979) captorhinomorphs have shown that two distinct families (Pro- torothyrididae and Captorhinidae) can be recognized. The Protorothyrididae, are characterized by high narrow skulls; incompletely ossified opisthotics; presence of ectopterygoids and tabulars; unhooked premaxillae; lightly built skeletons; 29 to 32 presacral vertebrae; unswollen neural arches; moderately tall neural spines; presence of cleithra; slender limb elements including humeri with well-developed supinator proc- esses; narrow manus and pedes. They include six Pennsylvanian and two Lower Permian genera. In the Lower Permian only Protorothyris and an undescribed species, represented by MCZ 1474, are known to conform to the protorothyridid morphological pattern. The Captorhinidae, known only from the Lower and Upper Permian, are advanced over the Protorothyrididae in many significant features. The Captorhinidae, characterized by low, wide, massive skulls; hooked premaxillae; ectopterygoids and tabulars replaced by the jugals and postparietals respectively; fully ossified paroccipital / processes of the opisthotics; heavily built postcranial skeletons; 25 presacral vertebrae with swollen neural arches and low neural spines; absence of cleithra; short massively built limbs; no distinct supinator process on the distal head of the humeri; wide manus and pedes, include 14 Permian genera. The major osteological differences between the two families of captorhinomorphs recognized by Clark and Carroll (1973) require the placement of Romeria within the family Captorhinidae (Heaton, 1979), which unfortunately necessitates, in accordance with the International Code of Zoological Nomenclature, the abandonment of the family designation Romeriidae and its replacement by the familial designation Pro- torothyrididae (Price, 1937.) The fissure fill deposits exposed in the Dolese Brothers’ Quarry near Fort Sill, Oklahoma have produced thousands of fragmentary specimens of early Permian amphibians and reptiles. Most of the specimens are the remains of the small cap- torhinids Captorhinus and Eocaptorhinus (Fox and Bowman, 1966; Heaton, 1979). In his study of the Middle Pennsylvanian captorhinomorph reptile Paleothyris, Carroll (1969) noted the similarity between the humerus of this protorothyridid and the distal fragment of an isolated humerus from the much younger fissure fill deposits of the Dolese Brothers’ Quarry. Many complete limb elements have since been recovered, including stylopodia (humeri, femora) and zeugopodia (radii, ulnae, tibiae, fibulae) of the fore and hind limb. Although these elements are dissociated, they clearly belong to a small, slenderly built protorothyridid captorhinomorph. The specimens in the figures in this paper represent the largest, most completely preserved limb elements available to the author, in addition to the hundreds of uncatalogued specimens found in the collections of the American Museum of Natural History, New York, the Field Museum of Natural History, Chicago, and the Museum of Comparative Zoology at Harvard University. The Fort Sill deposits appear, on the basis of their vertebrate fauna, to be of the same age as the lower part of the Arroyo Formation of the Clear Fork Group and possibly the upper part of the Lueders Formation of the Wichita Group of the Lower Permian of northcentral Texas (Heaton, 1979). This small captorhinomorph is the latest known survivor of the family Protorothyrididae and is a contemporary of the captorhinids Labidosaurus and Captorhinus. The scarcity of protorothyridid captorhinomorph remains makes the description of these fossils important, especially in view of the superb quality of the known specimens. Abbreviations Used in the Figures add cr adductor crest anc quart anconaeus quartus cap capitellum delt deltoideus dist art distal articular surface ect ectepicondyle ect gr ectepicondylar groove ent entepicondyle ent f entepicondylar foramen fib fibular surface of articulation int tr internal trochanter interc intercondylar fossa is tf ischiotrochantericus lat d latissimus dorsi Pp pectoralis pop popliteal area post r posterior ridge prox art proximal articular surface scor supracoracoideus sup supinator process t ‘“trochlea’’ tib tibial surface of articulation tr 4 fourth trochanter tric triceps Description and Comparison Humerus The humeri (ROM 21732 and 21739, Figs. 1 and 2) are remarkably similar to those in the Middle Pennsylvanian protorothyridids Hylonomus (Carroll, 1964) and Paleothyris (Carroll, 1969), retaining the tetrahedral configuration common to primitive reptiles. The shaft, however, is better developed than in any other known captorhinomorph, with the possible exception of Anthracodromeus. The poor preservation and ossifica- tion of the latter, however, make direct comparisons difficult. The width of the proximal end of the humerus is about 25 per cent of the bone’s length and the width of the distal end is about 28.5 per cent of the length. The shaft is exceedingly slender, only about 6.5 per cent of the length. These proportions indicate that this humerus is more slenderly built than that of any other captorhinomorph. In Captorhinus, for example, the shaft is about 11 per cent of the bone’s length, the width of the proximal and distal heads are 40 and 50 per cent of the length respectively (Holmes, 1977). The twist of the distal upon the proximal plane is 85 degrees. In all other pro- torothyridids and in most other primitive reptiles this angle can only be estimated 3 prox art, 10 mm Fig. | Humerus (ROM 21732) in (A) proximal dorsal, (B) distal dorsal, (C) proximal ventral, and (D) distal ventral views. because of crushing. The proximal articulation of the humerus is a long spirally twisting surface composed of an anterior concavity and a large posterior convexity separated by a slight transverse groove. The humeral surfaces are divided into proximal dorsal, proximal ventral, distal dorsal, and distal ventral surfaces (Romer and Price, 1940). Anteriorly, the proximal dorsal surface is separated from the anteroventral deltopectoral crest by a rugose edge. A proximal tubercle on the anterior rugose edge was the region of insertion of the M. deltoideus (Fig. 1). The posterior edge that separates the dorsal surface from the posterodorsal surface also bears a conspicuous ridge and tubercle for the insertion of the M. latissimus dorsi. In ROM 21732 the proximal articulation appears, because of the immaturity of the specimen, to extend onto the dorsal surface beyond its normal confines. As in most primitive reptiles the deeply concave ventral proximal surface of the humerus was apparently occupied by the insertion of M. coracobrachialis brevis. The anterior proximal area above the deltopectoral crest is rounded and lacks any rugosity (Fig. 1c). In modern reptiles, the M. supracoracoideus inserts in this area. Distally a prominent deltopectoral crest protrudes anteroventrally from the proximal head. At the summit of this crest is a relatively small rugose tubercle for the insertion of the M. pectoralis (Romer, 1922) (Fig. 2B). Both the deltopectoral crest and the tubercle for the insertion of the M. latissimus dorsi are also visible in Hylonomus (Carroll, 1964) and in Protorothyris (Clark and Carroll, 4 prox_art, U 10 mm ra SR Fig. 2. Proximal ends of mature limb elements. (A) Femur (ROM 21740) in proximal and ventral views; (B) Humerus (ROM 21739) in proximal and proximai ventral views. 1973) but they are located farther distally from the articulating surface than in the Fort Sill protorothyridid. The proximity of these processes to the articulating surface in the new specimens greatly restricts the area available for the insertion of the M. scapulohumeralis and M. subcoracoscapularis on the dorsal and posterior surfaces and the M. supracoracoideus on the anteroventral surface. A large entepicondylar foramen and a distally oriented supinator process are distin- guishing features of the distal head of the humerus. The humeralis inferior nerve probably ran along the deep groove on the dorsal surface of the entepicondyle (Fig. 1B) and passed through the elongate entepicondylar foramen, as in Sphenodon. The entepicondyle does not extend far laterally. Its rugose lateral and distal margin, unfinished in ROM 21732, furnished the areas of origin of the flexor musculature of the lower arm and foot. The lateral edge of the entepicondyle at the level of the entepicon- dylar foramen is formed by a sharp ridge for the probable insertion of the M. coracobrachialis longus. On the distal dorsal surface, the entepicondyle is separated from the ectepicondyle by a shallow concavity that widens distally. The poorly developed entepicondyle, which is much smaller than in pelycosaurs, is a low ridge that turns posteroventrally at its distal end. Anteriorly the ectepicondyle is bounded by the long ectepicondylar groove (Fig. 1A). The ectepicondylar groove is shorter in pelycosaurs than in this protorothyridid and is absent in captorhinids. Anterior to the deep ectepicondylar groove, which carried the radial nerve, the supinator process lies ventral to the level of the ectepicondyle and the general dorsal surface. As in Paleothyris the supinator process in the humerus of the Fort Sill protorothyridid extends far distally. The radial nerve was not fully surrounded by bone as it traversed the humerus, but the slight gap between the distal end of the supinator process and the ectepicondyle was probably bridged by cartilage. The supinator and extensor muscula- ture of the lower arm and foot originated from the rugose distal heads of the supinator process and ectepicondyle. Although the humerus of the best-known protorothyridid Paleothyris has a well- developed supinator process, this condition has been considered unusual instead of characteristic of the family. The structure of the humerus of protorothyridids was considered to be similar to that of their captorhinid descendants. All known cap- torhinids have a single prominent ridge anteriorly on the distal expansion of the humerus, instead of a separated ectepicondyle and supinator process. Examination of a humerus of Hylonomus, BM (NH) R. 4168 (Carroll, 1964, fig. 1), reveals, however, the presence of a supinator process similar to that in Paleothyris and the Fort Sill pro- torothyridid. The badly worn humerus of Protorothyris, seen in MCZ 1532 (Clark and Carroll, 1973, fig. 7) has a deep groove that runs along the anterior edge of the distal head. This groove corresponds exactly to the ectepicondylar groove of Paleothyris and the protorothyridid from Fort Sill; it therefore provides strong evidence for the presence of a supinator process in this species. The only other protorothyridids that have preserved humeri are the immature speci- mens of Cephalerpeton and Anthracodromeus. The region of the supinator process in Cephalerpeton is not ossified. Immature specimens of the Fort Sill protorothyridid are also unossified in this region. The only known specimen of Anthracodromeus 1s not only too immature but also is too poorly preserved to show the presence of a supinator process. The development of a distinct supinator process is a common occurrence in early tetrapods, but its shape and position relative to the rest of the humerus distinguishes protorothyridids from most other tetrapods. In pelycosaurs, diadectids, and limnos- celids, forexample, the stout supinator process extends anteriorly, roughly perpendicu- lar to the long axis of the humerus, usually at the level of the entepicondylar foramen (Romer, 1956). The ectepicondylar groove is usually poorly developed. In pro- torothyridids, by contrast, the supinator process extends far distally, close to the level of the elbow joint and does not project far laterally. The ectepicondylar groove separating the weakly developed ectepicondyle from the supinator process is long. The humerus of the Carboniferous eosuchian Petrolacosaurus (Reisz, 1977) has a similar type of supinator process to that seen in protorothyridids, but this eosuchian can be distinguished readily by its much greater size and relative slenderness. The captorhinids, as already noted, lack a distinct supinator process (Holmes, 1977). The lateral edges of the distal ventral surface are formed by the entepicondyle and the ectepicondyle. Between these margins the ventral surface is relatively flat, but is pierced by the large, oval entepicondylar foramen. This foramen and its related depressions on the dorsal and ventral surfaces are relatively larger than in captorhinids or pelycosaurs. Most of the distal end of the humerus is occupied by the elongate convex radial, and the slightly concave ulnar, surface of articulation (Fig. 1D). The ventrally facing capitellum is continuous with the ventrodistally oriented ulnar articulation. In contrast to the condition seen in captorhinids and pelycosaurs, the distal expansion of the humerus in the Fort Sill protorothyridid does not extend far beyond the confines of the elbow joint. The slight development of the entepicondyle and of the ectepicondyle, 6 typical of all protorothyridids, greatly restricts the areas of origin and reduces the mechanical advantage of the flexors and extensors, muscles important in the power and recovery strokes of the walking motion in primitive reptiles. Although the shoulder and elbow joints were probably as restricted in the protorothyridids as in all other primitive reptiles, and therefore none of the significant evolutionary changes that freed the movement of the lizard forearm are evident in this or any other protorothyridid, the humeri have become similar in proportions to those of agile extant lizards (Holmes, 1977). The slenderness of the humerus may be a reflection of the light build of the reptile. Radius The radius is a long, unusually slender, nearly cylindrical element, with convex dorsal, partially flattened ventral surfaces and slightly expanded ends. In ROM 21733 (Fig. 3) the shaft is about 6 per cent of the bone’s length, the breadth of the proximal end measures only about 16 per cent of the length and the distal width only about 13 per cent of the length of the bone. The radii of this captorhinomorph are, therefore, slightly more slender than those of other primitive captorhinomorphs but are much more lightly built than are those of captorhinids and pelycosaurs of small size. In Captorhinus, for example, the shaft is about 11 per cent of the bone’s length, the width of the proximal and distal heads are 26 per cent of the length (Holmes, 1977). The proximal head of the radius has a mediolaterally elongated concave articular surface that matches the rounded surface of the capitellum. As in Paleothyris (Carroll, 1969) and in Captorhinus (Holmes, 1977), the articular surface extends slightly onto the flattened ventral surface (Fig. 3c). 10 mm Fig. 3. Radius (ROM 21733) in (A) anterior, (B) medial, (C) posterior views, and outlines of the (D) proximal and (£) distal ends. A longitudinal ridge that extends on the medial surface (Fig. 3B) along two-thirds the length of the bone is prominent only distally. This radius lacks the prominent tuberosity for the biceps tendon that is commonly found on the medial surface, near the proximal head of the radius, in captorhinids and pelycosaurs. Another longitudinal ridge extends on the lateral surface, as in many pelycosaurs (Romer and Price, 1940: 228), from the distal head onto the posterior surface. This ridge may represent the site of attachment of the M. pronator quadratus, a muscle that originated from the ulna. In contrast to the radius in captorhinids and pelycosaurs, this radius in only slightly arched and the distal articular surface is perpendicular to the long axis of the bone. Amongst protorothyridids, radii are known only in Paleothyris (Carroll, 1969) and Protorothyris (Clark and Carroll, 1973). They resemble the radius of the Fort Sill protorothyridid in general proportions, but are too poorly preserved for detailed comparisons. Ulna The ulnae (ROM 21734 and 21735, Figs. 4 and 5) of the Fort Sill protorothyridid resemble those of Paleothyris and Protorothyris in general proportions, in the config- uration of the prominent proximal and distal expansions, in the proximal subterminal sigmoid notch for articulation with the humerus, and in the distal articular surface for the carpus. Detailed comparisons with these protorothyridids is not possible because of the poor preservation of this bone. The olecranon is well ossified in the Forst Sill specimens ROM 21734 and 21735, but less mature specimens have incompletely ossified olecranons. In ROM 21734 (Fig. 4) ate distant 10 mm Fig. 4 Ulna (ROM 21734) in (A) lateral, (B) posterior, (C) medial, and (D) anterior views. 8 the width of the olecranon, measured from the medial end of the sigmoid notch, forms 20 per cent of the length of the ulna. Similar proportions are found in some of the sphenacodontine pelycosaurs (Romer and Price, 1940 : 147) and in protorothyridids (Carroll and Baird, 1972), but it is the small diameter of the shaft, 6 per cent of the length of the bone, which gives the ulna of the Fort Sill protorothyridid its unusually slender appearance. The proximal end of the ulna is capped by a rugose ridge that curves over the apex of the bone. Distally from this ridge, to which the tendon of the M. triceps attached, the rugose lateral surface of the olecranon forms a triangular area for the insertion of the M. anconaeus quartus (Fig. 4A). The sigmoid notch retains the general pattern seen in most primitive reptiles, a strap-shaped surface composed of a small anteroventral region articulating with the medial surface of the capitellar protuberance of the humerus and a larger posterodorsal surface separated by a curved ridge (Fig. 4c). This ridge fits into a deep groove located medial to the capitellum of the humerus. The anterior and posterior surfaces of the ulna are separated medially by a gently rounded ridge that carries a rugose protuberance, in Captorhinus (Holmes, 1977), for insertion of the M. biceps tendon. The ulna of the Fort Sill protorothyridid lacks this process (Fig. 4C). Distally the expansion of the ulna ends in an elongate convex surface of articulation with the carpus. The broad, slightly convex surfaces of the distal head face posterolat- erally and anteromedially, in accordance with the torsion of the shaft of the bone. A pathological specimen (Fig. 5) has been found among the dozens of ulnae exam- ined. The highly distorted head of this slightly immature ulna appears to be the result of inadequate repair following a fracture. 10 mm Fig. 5 Pathological ulna (ROM 21735) in anterior view. Femur Except for its lighter build and larger size this femur (ROM 21740 and 21736, Figs. 2A and 6) resembles that of Paleothyris and Protorothyris. The width of the proximal end of the femur forms 17 per cent of the length of the bone. The width of the distal end is about 21 per cent of the length. The shaft is exceedingly slender, only about 6.5 percent of the length. In all these proportions this femur is much more slenderly built than those of any captorhinid. In Captorhinus, for example, the shaft is about 13 per cent of the bone’s length, the width of the proximal and distal heads are 30 and 37 per cent of the length respectively (Fox and Bowman, 1966). There is a definite curvature to the bone (Fig. 6B), the proximal head turned dorsally and the distal head turned slightly ventrally. This curvature, also seen in Paleothyris and most extant reptiles, is not found in either captorhinids or pelycosaurs (Romer and Price, 1940). On the ventral surface the deep intertrochanteric fossa is bound post- erodistally by a slightly rugose limiting ridge that extends to the prominent internal trochanter. The proximal end of the internal trochanter, set off from the head of the femur by a slight notch as in Paleothyris, has a rugose process for the tendinous a toby Sealy oe ay SF LD Secu EPR aan ae a LER ay ENR ai on Ped S rey we -Age ‘3 oe = aS ay as ae ria (S 1 ‘ t Fig. 6 Femur (ROM 21736) in (A) dorsal, (B) posterior, (C) ventral, and (D) anterior views. 10 insertion of the M. puboischiofemoralis (Fig. 2B). Distally from this process, the ridge that extends diagonally along the ventral surface of the shaft to the popliteal area has a short slightly rugose surface that probably represents the fourth trochanter for insertion of the M. caudofemoralis, and a long adductor crest that extends nearly to the middle of the bone (Fig. 6C). On the dorsal surface the areas of attachment of the M. ischiotrochantericus and the M. puboischiofemoralis internus form a conspicuous swelling near the proximal surface of articulation (Fig. 6A, B). On the distal head of the femur the intercondylar fossa is long and quite deep. Tibia Except for the smaller size of the cnemial process for the attachment of the M. triceps, the tibia closely resembles that of Paleothyris. The bone is strongly arched, much as in other captorhinomorphs and in pelycosaurs; it is deeply concave posterolaterally. The width of the proximal end of the tibia forms 28 per cent of the length of the bone. The width of the distal end is about 16 per cent of the length. The shaft is not unusually slender, about 9.5 per cent of the length. The much expanded proximal end is occupied by the two femoral articular surfaces, separated by a groove which extends posteromedially from the anterior surface and ends in a deep pocket between the articular surfaces. In ROM 21737 (Fig. 7) a narrow strip of bone extends from the socket to the posterior margin, separating the proximal surface of the bone into distinct medial and Jateral articular areas, as in the case of the distal head of the femur. Such a high degree of ossification indicates that this tibia belonged to a mature, adult individual. The anterior and posterior (extensor and flexor) surfaces are partially separated medially (Fig. 7D) and laterally (Fig. 7B) by a pair of well-developed ridges. The medial ridge commences near the proximal head and extends diagonally across the medial surface onto the posterior surface of the distal head. A well-developed ridge extends along the lateral surface from near the posterior surface of the proximal head diagonally across to the anterior surface of the distal head of the tibia. In the middle of the shaft a pronounced tuberosity is associated with the lateral ridge. The distal end of the tibia has an oval outline and a small concavity in the center of the articular surface. The long axis of the distal articular surface extends anteroposteriorly, whereas the long axis of the proximal double articular surface is directed mediolater- ally. This is in accordance with the general torsion of the tibia, much as in the case of the ulna. yy ehh 10 mm Fig. 7 Tibia (ROM 21737) in (A) anterior, (B) lateral, (C) posterior, (D) medial, (E) proximal, and (F) distal VIEWS. Fibula The only known fibula (ROM 21738, Fig. 8) is incompletely ossified, without any rugosities or ridges on the finished surfaces of the bone, and with poorly differentiated articular surfaces. The shaft is very narrow, only 6 per cent of the length, but expands proximally and distally to 18.5 per cent of the length. The distal head of the fibula is much expanded mediolaterally to form an elongate surface of articulation with the astragalus and calcaneum. The fibula shows little twisting or arching, in strong contrast to the condition seen in both captorhinids (Holmes, 1977), and pelycosaurs (Romer and Price, 1940), where the medial margin of the bone is strongly concave, and the twist of the distal plane upon the proximal plane is at least 45 degrees. Amongst other protorothyridids, the fibula is well preserved only inPaleothyris. With the exception of a somewhat larger proximal head there is little to differentiate the fibula of the Fort Sill protorothyridid from that of Paleothyris. 10 mm Fig. 8 Fibula (Rom 21738) in (A) posterior, (B) medial, (C) anterior, and (D) lateral views. Discussion The limbs of protorothyridids are poorly known, with complete well-ossified stylopodia and zeugopodia preserved only in Paleothyris. All other protorothyridids (Carroll and Baird, 1972; Clark and Carroll, 1973) have either incompletely preserved or ossified limb elements. The excellent quality of preservation of the protorothyridid limb elements from Fort Sill, in which all surfaces were exposed when the matrix was completely removed by washing with water, allows a more complete description than in any other protorothyridid. The specimens show that, as in other protorothyridids, the areas of origin and insertion of the muscles important in the power stroke and recovery are concentrated nearer to the ends of the bones than in either captorhinids or pelycosaurs. These muscles acted in protorothyridids as Class III levers with short moment arms and produced rapid movement of small force at the distal end of the humerus and femur. On the humerus, for example, the areas of insertion of the M. pectoralis, M. deltoideus and M. supracoracoideus were concentrated on the proximal 17 per cent of the bone. In both Captorhinus (Holmes, 1977) and Archeothyris (Reisz, 1972) power stroke and recovery muscles inserted on the proximal 29 per cent of the humerus. On the femur, the areas of insertion of the M. caudofemoralis, M. puboischiofemoralis externus and internus, and M. ischiotrochantericus were all concentrated on the proximal 24 per cent of the bone. In both pelycosaurs and captorhinids these muscles inserted on the proximal 35 to 50 per cent of the femur (Fox and Bowman, 1966; Romer and Price, 1940). The concentration of the above muscle origins and insertions close to the ends of the bones and the proportions of the stylopodia and zeugopodia indicate that this small reptile from Fort Sill, like the earlier protorothyridids, was a slender, agile, lightly built animal, in strong contrast to the massive, heavily constructed relatives, the cap- torhinids. {3 The limb elements described here can be readily assigned to the Protorothyrididae, and can be distinguished from all captorhinids and pelycosaurs on morphological grounds. The humerus and femur of the Upper Pennsylvanian eosuchian Pet- rolacosaurus and its Lower Permian relative Araeoscelis are surprisingly similar to those of the Fort Sill protorothyridid (Vaughn, 1955; Reisz, 1977). In only two significant characteristics are the stylopodia and zeugopodia of these two reptiles advanced over the pattern seen in earlier protorothyridids: the first character is limb proportions, the second is zeugopodial to stylopodial ratios. 1) The limb elements of mature individuals of both Petrolacosaurus and Araeoscelis are at least two and a half times larger than those of protorothyridids. Despite these size differences, the humeri and femora of Petrolacosaurus and Araeoscelis retain similar proportions (shaft to length ratio, proximal width to length ratio, distal width to length ratio) to those of protorothyridids. Since the shaft diameter of these limb elements is directly proportional to the volume of the animal, the two genera can be considered to be of relatively lighter build, or to have relatively longer limbs than protorothyridids. 2) In all protorothyridids, including that from Fort Sill, the zeugopodia are consid- erably shorter than the stylopodia. This pattern represents the primitive reptilian condition, where the zeugopodial length is in general equal to about two-thirds of the stylopodial length. In Petrolacosaurus the zeugopodia are equal in length to the stylopodia, whereas in Araeoscelis the zeugopodia are even slightly longer than the stylopodia, a very specialized condition. In both of these genera the zeugopodia are exceedingly slender, with shaft diameters of 3.5 to 5 per cent of the length. The limb elements of the Fort Sill protorothyridid are not sufficiently diagnostic below the familial level to warrant the naming of this reptile. Several cranial fragments from Fort Sill have already been named. Of these Delorhynchus, based on fragmentary maxillae (Fox, 1962) and Colobomycter, based on a partial right cheek and skull table (Vaughn, 1958a) have been placed among the Pelycosauria. Their similarity to primi- tive captorhinomorphs that were subsequently described (Carroll, 1964; Carroll and Baird, 1972) suggests that these two genera may be protorothyridids, but articulated specimens are needed before their identity can be established. Relatively small sphenacodont pelycosaurs have also been found at Fort Sill. Both Thraumosaurus (Fox, 1962) and Basicranodon (Vaughn, 1958b) are based on cranial fragments, but their postcranial skeletons, when found, would be readily distinguish- able from the protorothyridid remains, described here, on morphological grounds. In addition, the skull fragments indicate that these sphenacodonts would be considerably larger than the protorothyridid captorhinomorph from these deposits. 14 Acknowledgements I wish to express my thanks to Dr. P. P. Vaughn who donated most of the specimens illustrated in this paper. I am also much indebted to Dr. John Bolt of the Field Museum of Natural History in Chicago and Dr. Farish Jenkins, Jr., at the Museum of Compara- tive Zoology at Harvard University for allowing me to borrow many specimens from their collections. I also extend my gratitude to my colleague, Dr. Malcolm Heaton, for his frequent advice and for his comments on the manuscript. This work was supported by a grant from the National Research Council of Canada. Literature Cited CARROLL, R. L. 1964 The earliest reptiles. Journal of the Linnean Society of London (Zoology) 45 : 61-83. 1969 A middle Pennsylvanian captorhinomorph, and the interrelationships of primitive reptiles. Journal of Paleontology 43 : 151-170. CARROLL, R. L. and D. BAIRD 1972 Carboniferous stem-reptiles of the Family Romeriidae. Bulletin of the Museum of Comparat- ive Zoology 142 : 321-364. CLARK, J. and R. L. CARROLL 1973 Romeriid reptiles from the Lower Permian. Bulletin of the Museum of Comparative Zoology 144 : 353-407. FOX, R. C. 1962 Two new pelycosaurs from the Lower Permian of Oklahoma. University of Kansas Publica- tions, Museum of Natural History 12 : 297-307. FOX, R. C. and M. C. BOWMAN 1966 Osteology and relationships of Captorhinus aguti (Cope) (Reptilia: Captorhinomorpha). Uni- versity of Kansas Paleontological Contributions: Vertebrata il : 1-79. HEATON, M. J. 1979 The cranial anatomy of primitive captorhinid reptiles from the Late Pennsylvanian and Early Permian, Oklahoma and Texas. Bulletin of the Oklahoma Geological Survey 127 : 1-84. HOLMES, R. B. 1977 The osteology and musculature of the pectoral limb of small captorhinids. Journal of Mor- phology 152 : 101-140. OLSON, E. C. 1962 The osteology of Captorhinikos chozaensis Olson. Oklahoma Geological Survey, Circular 59 : 49-68. 1967 —_ Early Permian vertebrates of Oklahoma. Oklahoma Geological Survey, Circular 74 : 1-111. 1970 New and little known genera and species of vertebrates from the Lower Permian of Okla- homa. Fieldiana: Geology 18 : 359-434. PRICE, L. I. 1937 Two new cotylosaurs from the Permian of Texas. New England Zoological Club Proceedings 16: 67-102. [5 REISZ, R. R. 1972 Pelycosaurian reptiles from the Middle Pennsylvanian of North America. Bulletin of the Museum of Comparative Zoology 144 : 26-62. 1977 — Petrolacosaurus, the oldest known diapsid reptile. Science 196 : 1091-1093. ROMER, A. S. 1922 The locomotor apparatus of certain primitive and mammal-like reptiles. Bulletin of the American Museum of Natural History 46 : 517-606. 1956 The osteology of the reptiles. Chicago, University of Chicago Press. 777 pp. ROMER, A. S. and L. I. PRICE 1940 Review of the Pelycosauria. Geological Society of America, Special Paper 28 : 1-538. VAUGHN, P. P. 1955. The Permian reptile Araeoscelis restudied. Bulletin of the Museum of Comparative Zoology 113 : 305-467. 1958a Ona new pelycosaur from the Lower Permian of Oklahoma, and the origin of the family Caseidae. Journal of Paleontology 32 : 981-991. 1958b A pelycosaur with subsphenoidal teeth from the Lower Permian of Oklahoma. Journal of the Washington Academy of Sciences 48 : 44-47. 16 ISBN 0-88854-248-8 ISSN 0384-8159