<<

A FROM THE OF KANSAS

ROBERT R. REISZ

Department of Biology University of Toronto Erindale Campus Mississauga, Ontario Canada L5L 1C6

SPECIAL PUBLICATION

OF THE

MUSEUM OF NATURAL HISTORY, UNIVERSITY OF KANSAS

NUMBER 7

1981 1^\ EmstWIayrLfcrary

. / UNIVERSIT^BI^'XANSAS PUBLICATIONS ^ <) I ' ( MUSEUM OF NATURAL HISTORY

Copies of publications may be obtained from the Publications Secretary, Museum of Natural History, University of Kansas, Lawrence, Kansas 66045. A list of available Special Publications is provided on the inside back cover. University of Kansas Museum of Natural History

Special Publication No. 7

February 18, 1981

A Diapsid Reptile From the Pennsylvanian of Kansas

BY

Robert R. Reisz

Department of Biology University of Toronto Erindale Campus Mississauga, Ontario Canada L5L 1C6

University ok Kansas Lawrence 1981 UNIVERSITY OF KANSAS PUBLICATIONS MUSEUM OF NATURAL HISTORY

Editor: Joseph T. Collins Editors for this issue: E. O. Wilev and Larry D. Martin

h/[r'7 Special Publication No. 7 LIBRARY pp. 1-74; 26 figures 1 table Published February 18, 1981 II II o Q ?nnR

- RD I U SiTY

Copyrighted 1981 BY Museum of Natural History University of Kansas Lawrence, Kansas 66045 U.S.A.

Printed by University of Kansas Printing Service Lawrence, Kansas ISBN: 0-S933S-011-3 This is the first of a series of Special Publications honoring the contributions of Dr. Theodore H. Eaton to the fields of paleontology and zoology. The series will center around the Pennsyl\anian fauna of eastern Kansas, and will include papers on fishes, and , as befits the wide-ranging interests Dr. Eaton has shown in his research. E. O. Wiley Labry D. Martin 15 June 1980 Lawrence, Kansas

CONTENTS

INTRODUCTION - 1

ACKNOWLEDGEMENTTS - 2 DESCRIPTn'E AXD SYSTEMATIC HiSTORY 2 SYSTEMATIC DESCRIPTION 4 Family Petrolacosauridae Peabody, 1952 4 Petrolacosauriis Lane, 1945 4 Pefrolacosmirus kansensis Lane, 1945 4

OSTEOLOGY - - 5 5

Dermal Bones of the 9

Dermal Bones of the Palate 21

Ossifications of the Palatoquadrate Cartilage 23

Ossifications of the Braincase 24 26 Dentitiox 27 Sclerotic Plates — 28 Vertebrate 28 29 Dorsal Vertebrae 33 Sacral Vertebrae 34 Caudal Vertebrae 35 Ribs 36 Pectoral Girdle 38 Pelvic Girdle 40

Limbs .-. 43 PHYLOGENETIC REL.4TIONSHIPS OF 54 Hypotheses of Relatio.vships 57 Basic Taxa 58 Shared Derived Characters Testing the Hypothesis of Relationship Betv^^een Petrolacosaurus and 61 Shared Derived Char.acters Testing the Hypothesis of Relationship Between Petrolacosaurus and 62 Petrolacosaurus Compared with Aracoscelis 64

HEARING IN PETROLACOSAURUS AND OTHER EARLY REPTILES .. 66 CONCLUSIONS 69 SUMMARY 71 LITERATURE CITED 72

)

INTRODUCTION

AH living reptiles can be grouped into , and therefore , could not four orders: Chelonia, , be derived from captorhinomorphs. and . The latter three Romer (1966) rejected Watson's theory of can be associated with the largest assemblage the of the ear and reiterated his of fossil reptiles, collectively called diapsid belief that all true reptiles can be derived reptiles. The diapsid condition refers to the from primitive captorhinomorphs. In a re- presence of two pairs of temporal openings view of early reptilian evolution, Romer on the skull roof behind the orbits. This con- (1967) suggested not only that the sauropsid- dition has been retained in a primitive form theropsid dichotomy, insofar as it exists, has in crocodilians and in the sole living rhyncho- little phylogenetic significance, but also that cephalian Sphenodon ptinctatum but has been the term Sauropsida has no systematic or modified in the squamates by the progressive phylogenetic meaning. Furthermore, he em- loss of dermal bones in the temporal region. phasized that the diapsids may not belong to Diapsid reptiles have an evolutionary his- a single phyletic unit. He argued that two tory long thought to extend only into the late discrete diapsid groups should be recognized, . The early evolution of this assem- the Archosauria and the . He blage is not well documented in the fossil proposed that the Lepidosauria (the squa- record. Consequently, the origins and phy- mates, rhynchocephalians and the ancestral letic relationships of member groups have eosuchians) on the one hand, and the Archo- been subject to dispute. (the two orders, ,

On several occasions, Watson ( 1951, 1954 crocodilians and the basic group, the reiterated his belief in Goodrich's concept of thecodonts) on the other, evolved separately a basic dichotomy among reptiles. According from the ancestral captorhinomoqjhs. It was to this theory, one group called the Therop- his suggestion that the miileretid skull pattern sida led to and included the mam- represents an intennediate stage between the mal-like reptile orders Pelycosauria and The- ancestral captorhinomorphs and the lepido- rapsida. The other group, the Sauropsida, saurs. ' also called "true" reptiles, included the diap- In 1967 Reig set forth, and later (1970) sid reptiles and the (Chelonia). Wat- elaborated, a radically different hypothesis for son argued for independent origins of the the origin of . He proposed that Theropsida and Sauropsida from anthraco- the ancestors of archosaurs should be sought saurian amphibians via some unknown inter- among varanopsid . Most of the mediate forms. He based this conclusion on comparable features cited by Reig as indica- his theory of the evolution of the ear (Wat- tive of phylogenetic relationships are, how- son, 1953). He discounted the possibility of ever, primitive characteristics found in the captorhinomorphs having given rise to the ancestral captorhinomorphs (Romer, 1971). Sauropsida and, hence, to the diapsid rep- Gow (1972), in his reconsideration of the tiles. He actually placed the Captorhino- milleretids, contended that this group gave morpha among the Theropsida. In 1957 Wat- rise directly to the Squamata. He felt on the son suggested that the diapsids evolved from other hand, that the Rhynchocephalia and the seymoriamorph anthracosaurs via the Millero- Archosauria had an entirely independent ori- sauria. gin.

Vaughn ( 1955 ) also accepted the saurop- The above workers have based their theo- sid-theropsid dichotomy and felt that the ries of the origins and early evolution of the SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7 diapsid reptiles mainly on the meager evi- primitive captorhinomorphs and to more ad- dence provided by the known Permian coty- vanced diapsid reptiles. The problem of hear- losaurs and on the Upper Permian and Trias- ing in Petrolacosaurus and other early reptiles sic diapsids. Thanks mainly to the labors of forms an integral part of this study. Robert L. Carroll and the late Frank E. Pea- body, the problem of the origins and early Acknowledgements evolution of diapsid reptiles can be reconsid- I am greatly indebted to T. H. Eaton of ered. Carroll, during a period of about ten the University of Kansas Natural History Mu- , described and reconsidered all of the seum, at whose suggestion this study was available Pennsylvanian and Lower Permian undertaken and who kindly allowed me to specimens of primitive captorhinomorphs. One borrow the Petrolacosaurus specimens at his significant conclusion became evident: the disposal. I am particularly grateful to Dr. and described members of this group conform to Mrs. Eaton for courtesies extended to mc a single conservative morphological pattern during the time spent studying the Kansas suflBciently to indi- and remained generalized collection. cate that i^rimitixe captorhinomoqohs could Thanks are due also to R. L. Carroll whose be ancestral to most subsequent reptilian line- wise guidance, friendliness, and unrelenting ages (Carroll and Baird, 1972; Clark and Car- support were invaluable. I am very grateful roll, 1973). to P. L. Robinson of University College for In 1952, Peabody published a partial de- her many valuable and instructive sugges- scription of Petrokicosaiinis kansensis and tions, for numerous stimulating conversations, suggested that this reptile from the LTpper and for making the time spent studying in Pennsylvanian was a member of the most London so pleasant. I ha\'e profited also from primitive diapsid order, the "." The several interesting discussions with Donald fragmentary nature of the specimens available Baird of Princeton University and Malcolm for this study led others, however, question to Heaton of University of Toronto. Peabody's contention. Subsequently Peabody Thanks are due to Bobb Shaeffer and E. S. spent two grueling summers 1953 and 1954 — Gaffney, both of the American Museum of collecting more specimens but died before — Natural History, New York, for extended could undertake reconsideration of the he a loans of specimens. anatomy of Petrolacosatirus. This collection fonus basis of present These the the study. DESCRIPTIVE AND SYSTEMATIC show that Petrolacosaiirus is a specimens HISTORY diapsid reptile with affinities to both primitive captorhinomorphs and the earliest "eosuch- Lane first described Petrolacosaurus kan-

iaiis." sensis in 1945 and referred the genus to the My interest in Petrolacosaiirus was aroused family Si)henacodontidae of the order Pely- during a visit to the Kansas University Mu- cosauria ( 1946) on the basis of a tarsus found seum of Natural History in Lawrence, Kansas, with an isolated hindlimb and peKis. An where T. H. Eaton invited me to examine this isolated iorelimb and some other skeletal re- interesting reptile. It became apparent that mains were- described as Podar'^osaurus hih- the available specimens of this genus, much hardi and referred to the family Araeoscelidae superior to those described in 1952, warranted of the order Protorosaiiria ( 1946) on the basis a thorough reeonsideration. The purpose of of the slender humerus, , , and the the present study, undertaken with Dr. Eaton's ribs. kind permission, is to describe the earliest In 1952 Peabod\ publisiicd a detailed de- known diapsid reptile, Pctrohicosaurus kansen- scription of (he CJarnett reptile and placed sis, and to consider its relationship to both Podargosaurus in synon)iny with Petrolaco- 1980 PENNSYLVANIAN DIAPSID REPTILE saurus. He concluded that "Petrolacosaurus and that the former genus may become the represents a new family of primitive reptiles type of a new family of . showing relationships which place it at the Tatarinov (1964) reassigned the Family base of the Eosuchia while also evidencing Petrolacosauridae to the Order Araeoscelidia; strong relationship with primitive cotylo- he based his conclusions on the similarities saurs." The fragmentary nature and imma- seen in the palate and postcranial skeleton of turity of the known specimens led others to Petrolacosaurus and . Tatarinov reject his conclusions. Peabody's association followed Romer (1956), however, in placing of an isolated edaphosaurian pelvis (KUVP the Araeoscelidia within the Subclass Eury- 1425) with the Petrolacosaurus material add- apsida. ed to the confusion. Romer (1956) tentatively placed Petrola-

Watson ( 1954 ) , after an examination of cosaurus within the Family the known specimens concluded that this (Pelycosauria). In 1966 in a joint paper with genus was a theropsid. He based his conclu- Stovall and Price, Romer maintained that sion on the nature of the quadrate in the "Apart from the possible but unproven diap- small, immature skull. He felt that this quad- sid nature of the temporal region, there is no rate was similar to those in pelycosaurs and reason to assign Petrolacosaurus to the Eo- captorhinomoiphs. ." He felt, however, that certain diag- Although he reserved final judgment until nostic features in the post-cranial skeleton

". a more complete description was undertaken, . . strongly indicate that this genus belongs Vaughn (1955) was inclined to agree with to a group of archaic edaphosaurians from Watson's diagnosis. He also suggested that which both Edaphosaiinis and, at a much Petrolacosaurus may be related to Araeoscelis later time, tlie caseids may have arisen." —

SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

SYSTEMATIC DESCRIPTION

Family PETROLACOSAURIDAE Peabody, processes on neural spines and excavation of 1952' neural arches of dorsal vertebrae, longer tail, and greater relative length and slenderness of Genus Petrolacosauriis Lane, 1945 limbs. Type .—Petrolacosaurus kansensis Separated from Araeoscelis by the inferior Lane, 1945. temporal fenestra, more lightly built skull and Si/noni/m. Podarg^osaurus Lane, 1945. dentition and straight ventral margin of Occurrence.—Upper Pennsylvanian of cheeks. Kansas. Differs from other diapsid reptiles by sim- Diagnosis.—Same as for species. ple lateral cmbayment of parietal rather than a ventrolateral flange in addition to the em- Petrolacosaurus kansensis Lane, 1945 bayment, larger lacrimal and squamosal, Figs. 1-25 larger quadratojugal, lack of retroartieular Synonym.—Podargosaurus hibbardi Lane, process, denticulate parasphenoid and mas- 1945. sive stapes.

Revised diagnosis.—Medium sized early HoloUjpe.—KUVP ( Kansas University' Mu- diapsid reptile (Fig. 1) with well developed seum of Natural History) 1424, adult right superior and inferior temporal fenestrae and hindlimb consisting of most of the femur, the elongate, narrow suborbital fenestra. Parietal distal part of the tibia and in articula- without the ventrolateral flange that extends tion with a complete pes. into the superior temporal fenestra in younger Hypodigm.—KUVP 1423, complete right diapsids; posterior splenial bone present in forelimb, immature. KUVP 1426, right hind- mandible; marginal dentition unusually thin- limb lacking femur, immature. KUVP 1427, walled. Six elongate cervical vertebrae; twen- splenial, \ertebral column, scattered ribs, fore- ty six presacral vertebrae; a pair of mam- limbs, partial pectoral girdle, partial hind- millary processes on the neural spine of first limb, immature. KUVP 142S, questionable sacral ; large dorsal isciadic notch; skull fragment, partial , partial verte- slender forelimbs equal in length to more bral cohnnn, scattered ribs, incomplete fore- massive hindlimbs; propodials equal in length limbs and hindlimbs, partial pelvic girdle, to epipodials. scattered gastralia, immature. KU\T 1429, Distinguished from captorhinomorph rep- left forelimb, immature. KUVP 8351, partial tiles by relatively smaller skull, superior, in- skull, five cervical vertebrae, immature. KUVP ferior temporal fenestrae and suborbital fe- 8,355, right carpus, immature. KU\"P 9950, nestra, greater length of neck, partial rather isolated right parietal, mature. KUVP 9951, than complete coossification of atlantal cen- partial skull, vert<>bral column, ribs; on sepa- trum and axial intercentrum, mammillary rate blocks femora and peKes, left lower hind- limb, right lower hindlimb, slightly immature. KU\'P 9952, partial .skull (mainly dermal ' The family Petrolacosauridae has been formerly skull roof), mature. KUVP 9956, a series of included in the order Eosuchia (Pcal)ody, 1952; Rcisz, 1977). Study of the ahove order indicates tliat it is articulated \ertebrae from C3 to S2, with of and Mcso- an artificial assemblage Late some closely associated ribs, a short series of zoic diapsid reptiles. The "Eosuchia" has no ta.xo- articulated caudal vertebrae, mature. KUVP nomic validity because there are no known diag- nostic features that can distinf^i'sh it from other diap- 9957, three separate blocks containing left sid reptiles. The diagnostic features of the mono- scapulacoracoid with hinnerns, radius and generic family Petrolacosauridae are the same as for ulna in articulation and partial carpus, slight- the genus. —

0.75. Composite. Fig. 1. Petwlacosaunis kansensis. Reconstruction of skeleton, X c

1980 PENNSYLVANIAN DIAPSID REPTILE ly immature. KUVP 995S, left , slightly but also shows a number of significant de- immature. KUVP 9959, seven separate blocks partures from this pattern. containing badly preserved subadult skull, In dorsal aspect (Fig. 3a) the skull is long interclavicle, immature, lower jaws, scat- and relatively narrow. The width at the quad- tered \ertebrae and ribs, forelimbs, hindlimbs, rate exceeds, only slightly, one-half the length both ilia, very immature. KUVP 9961, left of the skull. Maximum width is reached just pelvis, mature. KUVP 9962, ilium with first behind the orbits in the region of the tem- sacral rib attached, immature. KU\T 10320, poral openings. In the antorbital region the badly preser\ed specimen containing skull skull table extends far laterally, obscuring the fragments, vertebrae, hindlimbs, immature. posterior part of the narial opening and the KUVP 33602, isolated left maxilla, slightly maxilla when seen in dorsal view. In the immature. KUVP 33603, fragments of skull postorbital region both temporal fenestrae are material, distal end of humerus, immature. visible from above. The summit of the skull

KUVP 33604, scattered skull material, atlas- (in cross-sectional view) in this area is at the axis complex, cleithrum, distal end of hu- level of the medial suture of the postfrontal merus, radius, ulna, mature. KUVP 33605, wedge. Laterally the skull roof slopes gently seven posterior dorsal vertebrae and first sac- down to the level of the upper temporal ral vertebra, first sacral rib, all articulated, fenestra. From here the grade of the down- slightly immature. KUVP 33606, partial skull, ward slope increases greatly. The curvature posterior dorsal, sacral and anterior caudal between the skull table and the cheek in the vertebrae, some associated ribs, clavicle, par- temporal region is thus formed by the ventral tial scapulacoracoid, humerus, partial pelvis, part of the postfrontal that lies beneath the right hindlimb with scattered pes, mature. postorbital and the dorsal part of the post-

KUVP 33607, slightly dissociated skull, first orbital and squamosal. The posterior margin three cervical vertebrae, cleithrum, right of the skull table is deeply emarginated where scapulacoracoid, olecranon, partial carpus, the dorsal aspect of the parietal ends. mature. KUVP 33608, poorly preserved skull The antero-posterior slope of the skull and cervical vertebrae. KUVP 33609, right table is illustrated in Fig. 2a. The skull is scapulacoracoid, mature. long and shallow in lateral \iew. The maxi- mum height, reached in the orbital region OSTEOLOGY represents only about 35 per cent of the length of the skull. The gently convex pos- Skull terior margin of the skull is nearly perpen- dicular to the ventral margin. The jaw articu- By using the available material, it has lation lies only slightly below the level of the been possible to obtain a composite, fairly tooth row. The marginal tooth row occupies accurate reconstruction of the skull, in which 58 per cent of the skull length. most of the external details of the dermal The outline of the palate is defined by the roof, occiput, palate, the lateral and medial \entral edges of the cheeks and by the pos- aspects of the mandible can be discerned. The terior margin of the braincase (Fig. 3b). The quality of preservation prevents any recon- internal nares, bounded by the premaxilla, structions of the braincase beyond those seen maxilla, palatine and vomer, are long and in Figs. 3b and 4. narrow. Both the vomer and palatine extend The skull as reconstructed has the princi- ventrally at the margin of the opening form- pal dimensions listed in Table 1. These meas- ing ridges. The interpterygoid vacuities are urements indicate that Petrolacosaurus has of moderate size. The suborbital fenestrae retained many of the primitive cranial pro- are well developed, elongate openings, albeit portions seen in primitive captorhinomorphs smaller than in Youngina (Olson, 1936; Gow, —

SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

Fig. 2. Pctwlacosaurti.i kansensis Lane. Rpconstniction of slaiU and niandihle in lateral view (A), and of mandible in medial view (B) based mainly on specimen.s KUVP 9952 and KUVP 33607, X 2. Abbreviations u.sed in Figs. 2-13: a, angnlar; ae, atlantal centrum; an, atlantal neural arch; art, articular; a.\, ;i.\ial centrum;

a.\i, axial intercentrum; bo, basioccipital; bs, basisphenoid; bt, basipter\goid process; co, coronoid; ct, cultri-

form process; d, dentary; ec, ectopterygoid; ex, exoccipital; f, frontal; j, jugal; 1, lacrimal; m, maxilla; n, nasal; o, opisthotic; p, parietal; pf, postfrontal; pi, palatine; pm, preniaxilla; po, postorbital; pop, paroceipital process; pp, ; pra, prearlicniar; prf, prefrontal; i^ro, proolic; ps, parasphenoid; psp, postsplenial; pt, pterygoid; q, quadrate; qf, (juadrate foramen; qj, (juadratojugal; rl, atlantal riii; r2, axial rib; s, stapes; sa, surangular; sc,

sclerotic plates; sni, septomaxilla; so, supraoccipital; sq, scpiamosal; st, supratemporal; t, tabular; tr. fl. pt, transverse flange of pterygoid; v, vomer.

1975). With the exception of the ectoptery- margin of the jaw articulation. The subtem- goid, the paired palatal l)Oiies hear denticles; poral fossae are quite large, occupying about in addition, three major rows of large teeth 23 per cent of the palatal area. radiat(> laterally, diagonally, and anteriorly The overall impression from the occipital from the hasicranial artienlation. Small den- aspect (Fig. 4) is of a fairl\- low, wide skull. ticles arc also present on the parasphenoid. Part of the skull roof is \isible in this view.

The palate is anchored directly to the dermal \ well developed nuchal ridge extends from sknll roof at three points: anteriorly to the the dorsal occipital border to the foramen premaxilla and laterally to the maxilla and uKignum, which has the same height as width. to the jngal. Posteriorly, the palate is indi- The post-temporal fenestrae are large. The rectly anchored to the skidl roof by the (jnad- occipital cond\le is located low on the occipi- rate and the braincase. The oceijiital condyle tal face, only 4 nun Irom the line joining the is set only slightly anterior to the posterior \entral borders of the (juadratojugals and —

1980 PENNSYLVANIAN DIAPSID REPTILE

B

Fig. 3. Petrolacosaurus kansensis Lane. Reconstruction of skull in dorsal (A), and ventral (B) views, based mainly on specimens KUVP 9952, 33606 and 33607, X 2. See Fig. 2 for key to abbreviations.

about 13 mm below the posterior edge of the frontal, postorbital, and jugal is slender; the parietal. orbit and temporal fenestra are, therefore, not

The largest openings in tlie are the widely separated.

orbits. Each occupies 38 per cent of the total The external naris is relatively large in skull length. The orbits incise the skull table relation to the skull. This opening faces an- deeply. terolaterally and only slightly dorsally. The

The upper temporal fenestra faces mainly external naris is floored laterally by a shelf

dorsally, but is also visible in lateral view formed by the premaxilla and maxilla. because of the curvature of the circumtcm- A large pineal foramen opens dorsally on poral bones. The lower temporal fenestra, on the midline, approximately at the level of the

the other hand, faces mostly laterally, but is anterior margin of the upper temporal fe- also visible from the dorsal aspect. Its pos- nestra.

teroventral corner is not preserved. The post- The lower jaw, as reconstructed (Fig. 2),

orbital bar, composed of parts of the post- has the principal dimensions listed in Table 1. —

SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

Fig. 4. Petrolacosaurus kansensis Lane. Reconstruction of skull in occipital \ie\v, mainly based on speci- mens KUVP 9951 and 33607, x2. See Fig. 2 for key to abbreviations.

Table 1.—Measurements of the skull and lower jaw of Petrolacosaurus kansensis, made primarily on the basis of the restoration of KUVP 33607. Measure- ments of the captorhinomorph archeri are based mainly on the type specimen (Clark and Carroll, 1973). Measurements in millimeters. 1980 PENNSYLVANIAN DIAPSID REPTILE normal sculpturing was exaggerated by the the dorsal, lateral and palatal sides of the superposition of the internal bone structure skull to constitute the region of the snout. onto the surface. On the lower jaw, only the The unusually slender nasal ramus of each anterior one-third of the dentar\' is sculptured. premaxilla, exposed in both external and in- Near the tip the lateral surface of the dentary ternal views in KUVP 9952, join at the mid- has delicate rounded pitting. More poste- line to form the anterior border of the snout. riorly the pitting takes on the shape of elon- The dorsal tips appear as slight posterior gate grooves. projections on the dorsal surface of the skull

Exposure of the internal surface of the and fit into the bifurcated anterior ends of skull roof of Petrolacosaurus revealed the the nasals. The dorsal process reaches its presence of a series of ridges or thickened greatest width at the point where it becomes areas along the major lines of cranial stresses. visible on the superior surface of the skull. These ridges (Fig. 25) form a framework of The maxillary ramus contributes to the strengthened areas that connect the orbital ventral border of the external naris and meets margin with the snout, maxilla and suspen- the maxilla in a long diagonal suture. In sorium. What makes the ridges seen in dorsal view the maxillary ramus gives a

Petrolacosaurus significant is their relative rounded outline to the snout. thickness when compared to the much thinner The palatal ramus forms a continuous areas of the dermal skull roof and palate. The shelf anteriorly that extends between the ex- use of the resin imbedding compound during ternal nares; it also forms the rounded an- preparation of the skulls made complete re- terior border of the internal naris. At the moval of the matrix possible. This revealed midline the premaxilla forms a wedge-like that in the non-ridged areas the skull is made posterior vomerine process best seen in KUVP of very thin bone. The anteromedial part of 8351 (Fig. 5), that fits into a corresponding the prefrontal, the medial half of the nasal, slot in the vomer. the anteroventral portion of the squamosal Anteriorly and laterally, the premaxilla is and most of the quadratojugal, are less than sculptured and is pierced by small foramina. one-tenth the thickness of the ridged areas In lateral view, the generally slender appear- of the same skull. Such differences in the ance of the premaxilla is accentuated by the thickness of the skull elements cannot be height and extreme slenderness of the dorsal attributed to some unusual type of preserva- process and by the emargination of the maxil- tion because other genera, found in the same lary process, where a medially directed flange deposits do not show such variation in bone provides support for the nasal capsule. There thickness. In and are from three (KUVP 9952) to five (KUVP (Heaton, 1979; Romer and Price, 1940), 33607) slender, conical premaxillary teeth on where the external and internal surfaces of either side of the midline. the skull are known in great detail, the dif- The premaxillae of the protorothyridids ferences in bone thickness do not approach Hijlonomus, Paleothijris, and Protorothyris are the range seen in Petrolacosaurus. The in- similar to but slightly more massive than ternal surfaces of the skull roof and palate in those of Petrolacosaurus. Reconstructions of Permian and Lower Triassic "eosuchians" are these protorothyridids (Carroll and Baird, inadequately known for comparative pur- 1972; Clark and Carroll, 1973) show broad poses. butt joints between the premaxillae and vo-

mers. This is unlikely, especially since the Dermal Bones of the Skull Roof anterior end of the vomer in Paleothijris is Premaxilla.—The paired premaxillae form narrow and bifurcate. The vomerine process the anterior margin of the skull. Each ele- of the premaxilla is poorly known in protoro- ment has three slender processes that connect thyridids. —

10 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

Fig. 5. Pctrolacosaurus kanscnsis Lane. Immature skull and lower jaws, KUVP 8351, X 4. (A) Dor- sal view of palate exposed by loss of skull roof, oblique view of lower jaws. (B) Ventral view of palate and oblique view of lower jaws. For key to abbreviations, see Fig. 2.

Maxilla.—As in primitive captorhino- illa is .suturally attached to the premaxilla

morphs and "(-o.suchian.s," the maxilla is the o\er a large surface area, as indicated by the

longest hone of the skull roof. It extt-nds from rugose area on the anterior medial surface the midpoint of the external naris past the of the maxilla in KU\P 33602 (Fig. 9). To-

midpoint of the orbit, over .53 per ecnt of the gether with the premaxilla, the maxilla forms

skull length. It forms the posteroxentral bor- not only the ventral border of the external

der of the external naris but is exeluded from nares but also a medially directed shelf for the orbit by the lacrimal and the jngal the nasal capsule. (KUVP 9951, Fig. 6). Anteriorly, the max- Posteriorly, the bone expands dorsally to —

1980 PENNSYLVANIAN DIAPSID REPTILE 11

Fig. 6. Pctrolacosaurus kansensis Lane. Mature skull and caudal vertebrae, partially scattered, KUVP 9951, X 2. Braincase of same specimen shown in Fig. 7. See Fig. 2 for key to abbreviations. reach its greatest vertical expansion immedi- branches of the superior alveolar nerve and ately behind the level of the caniniform teeth, the maxillary artery. The lower margin is one-third of the distance from the anterior nearly straight, as in primitive captorhino- end of the maxilla. From this summit, the morphs and younginid "eosuchians." dorsal border of the maxilla slopes gently The internal surface of the maxilla is more posteroventrally in an almost straight line. important taxonomically than the lateral. As

The lateral surface of the bone is pierced by in all primitive reptiles, the simple conical small labial foramina, probably for cutaneous subthecodont maxillary teeth are attached to —

12 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

Fic. 7. PetTolacosaums kansensis Lane. Braincase of adult individual, KUVP 9951, X 2. Braincase ex- posed in ventral (A) and dorsal views (B), with the lateral, occipital and ventral elements compressed into a single plane. See Fig. 2 for key to abbreviations.

the alveolar shelf. In KUVP 9952, the entire bone) to both the palatine and the ectoptcry- left maxilla is exposed. Thirty teeth are in goid.

place, with room for five others. Laterally There is no buttressing on the medial sur- the teeth arc covered by a thin sheet of bone face of the maxilla, above the caniniform

that extends up to one-third of the length of teeth. This area is somewhat strengthened, the teeth below the alveolar shelf. The me- however, by the ventral margin of the lacri-

dial surface of the alveolar shelf is smooth in mal which fits into a long longitudinal groove the region where it forms the lateral border along the dorsal margin of the maxilla. The of the internal naris and the lateral border of cross-section of the maxilla (Fig. 9b) in the the suborbital fenestra. Between these two region of the canines shows that the alveolar openings, directly beneath the anterior mar- shelf is hollowed out from above, giving the gin of the orbit, the alveolar shelf bears maxilla the outline of an inverted letter "h." a small striated process, that is exposed in Septomaxilla. — The septomaxilla, pre- both KUVP 9952 (Fig. 8b) and KUVP served only in KUVP 9952 (Fig. 8), is a thin 33602 (Fig. 9a). This process which occupies sheet of bone whose shape and relationships only 17 per cent of the length of this bone, cannot be determined because of crushing. is sutured to the palatine. It is significant It was probably attached to the anterior edge that, as in all "eosuchians," the maxilla in of the maxilla. In most primiti\'e reptiles the Petrolacosaiirus is attached directly to the septomaxilla, when preserved, appears as a palate only through this jirocess. Posterior to thin rectangular sheet of bone that has been this contact with the palatine, the maxilla curved around a simple cone (Heaton 1979). forms the lateral border of the suborbital fe- Only in pelycosaurs is the septomaxilla robust nestra. The maxilla is not attached to the with a thickened base and a large vertical ectoptcrygoid. In primitive captorhinomorphs sheet placed transversely in the posterior por- and pelycosaurs, on the other hand, the max- tion of the external naris (Romer and Price illa is sutured behind the internal naris (more 1940). than 50 per cent of the total length of the LacrinuiL—The lacrimal in KUVP 33607 —

1980 PENNSYLVANIAN DIAPSID REPTILE 13

Fig. 8. Petrolacosaiinis kanxctuis Lane. Mature skull, KUVP 9952. The e.\ternal (A) and internal (B) surfaces of the skull talilc and left cheek are shown, but the preserved part of the right cheek is exposed only in lateral view. See Fig. 2 for key to abbre\iations.

(Fig. 12 and 13) i.s a large shoet of bone tliat side of the skull in KUVP 33607. Conse- extends from the naris to the orbit. From its quently, less than half of the height of the broad contribution to the narial border, the suborbital ramus is seen in lateral view. laterally exposed surface of the lacrimal ex- The right lacrimal, in KUVP 33607, shows tends slightly upward, increasing its width three foramina that open into the orbit. Two posteriorly, and forming the anteroventral large lacrimal puncti are situated in a deep, border of the orbit. Its suborbital ramus ex- conjunctival groove near the lateral border tends far posteriorly and makes contact with of the lacrimal. The groove is partially ex- the anterior end of the jugal. The ventral posed in lateral view. A much smaller fora- border of the lacrimal is overlapped by the men is visible on the medial surface of the maxilla from the region of the canines to the suborbital ramus that may have carried the posterior tip of the bone, as seen in the right anterior orbital artery, much as in the primi- —

14 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

lb

C

Fig. 9. Petrolacosaums kanscnsis Lane. (A) Isolated left maxilla, KUVP 33602, exposed in medial view, X 2. (B) Section through maxilla at b-b, X 5. (C) Left parietal from a poorly preserved skull, KUVP 9959a, exposed in ventral view, X 2. (D) Quadrate and quadratojugal from poorly preserved, immature skull, KUVP 33603, exposed in partial ventral view, X 2. See Fig. 2 for key to abbreviations.

tive captorhinid Eocaptorhinus laticeps (Hea- process of the premaxilla. The remaining part

ton, 1979). The lacrimal canal is completely of the anterior border is free and forms the enclosed within the lacrimal bone. dorsal border of the external naris. Laterally,

The orbital margin of the lacrimal is the nasal is sutured to the lacrimal and the

strongly buttressed. At the level of the an- prefrontal. The lacrimal fits over a depression terior border of the orbit, the lacrimal reaches on the nasal, seen in KUVP 9952 (Fig. 8),

the alveolar shelf of the maxilla and reinforces whereas the prefrontal is attached to the ven-

the maxillary process where it is sutured with tral surface of the nasal.

the palatine. This is best seen in Fig. 8, The nasal is transversely convex (in cross-

where the suborbital ramus of the right lacri- section) in the region where it is attached mal is exposed in medial view. to the lacrimal, forming a dome over the nasal Nasah.—The nasals, exposed both ven- capsule. On the ventral surface, a prominent trally and dorsally in KUVP 9952 and 33607, ridge runs anterolaterally from the posterior are paired bones lying along either side of border. the midline of the skull, between the pre- Prefrontal.—The prefrontal, seen in lateral maxillae and the frontals; they form part of view in KUVP 9952 and 33607 and in dorsal

the skull roof in the antorbital area. The pair view in KUVP 33606, is a triangular bone are sutured to each other along most of their with an anterior plate that forms a wedge length. Each nasal has a posterior process between the nasal and the lacrimal, and a that is well preserved in KUVP 9959b, which thickened posterior ridge that forms the an- diverges from the midline and covers part of terodorsal border of the orbit.

the frontals dorsally. Anteriorly, the bone The anterior plate is a thin dorsoventrally curves strongly down to receive the dorsal convex sheet of bone, partially overlapped by —

1980 PENNSYLVANIAN DIAPSID REPTILE 15

Fig. 10. Petrolacosaurus kansensis Lane. Skull and lower jaws of young adult, KUVP 33608, X 2. The slightly disarticulated palate and braincase are e.xposed in dorsal view by loss of skull table. Part of the cheeks and the left parietal are exposed in internal view. See Fig. 2 for key to abbreviations. the lacrimal and nasal. A heavily .sculptured was probably attached. A well developed ridge that runs from the anterior plate to the ventral process of the prefrontal orbital ridge orbital margin separates the lateral and dorsal is strongly sutured to the medial surface of aspects of the skull in the antorbital region. the thickened orbital rim portion of the lacri-

A heavy posterior thickening of the prefrontal mal. The prefrontal is heavily sculptured near forms the anterodorsal angle of the orbit. the orbital rim. The thickened dorsal orbital rim portion Frontal.—The frontals are the third in the of the prefrontal extends far posteriorly and series of paired bones lying along the midline is sutured medially to the frontal in a well of the skull. Each is a long prominent bone developed tongue and groove joint. The dor- that lies on the dorsal surface of the skull. sal orbital margin is rugose and probably In KUVP 33606 (Fig. 11) the two frontals are ser\'ed as an attachment for part of the or- joined to each other along most of their bital fascia. length in a slightly wavy tongue and groove

The prefrontal extends medially along its suture. Each is bounded anteriorly by the anterior orbital rim portion to form the verti- nasal and laterally by the prefrontal and post- cal edge to which the orbitonasal membrane frontal. Between the latter two elements the —

16 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

Fic. 11. Pctrolacosaurus kanscnsis Lane. Mature .';!ai!l and left lower jaw, KU\'P 33606, X 2. Part of the dermal skull roof is exposed in donsal view, in almost perfect association; partial left lower jaw is seen in lateral view. See Fig. 2 for key to abbreviations.

frontal forms the orbital margin, and is slight- frontal and the postfrontal are attached in ly rugose to provide better attachment for the order to brace the anterodorsal and postero- orbital fascia. Posteriorly, the frontal inter- dorsal corners of the orbit. This ridge is most digitatcs with the anterior margin of the pronounced over the orbit. parietal. The frontal forms a wedge, postero- Parietal.—The parietal in Pctrolacosaurus laterally, which partially separates the post- is readily distinguishable from that of capto- frontal and jiarietal. The suture between the rhiuomorphs and pelycosaurs because it forms postfrontal and the frontal is also of the part of the border of a well developed su- tongue and groove type, similar to the suture perior temporal fenestra. It resembles the between the prefrontal and frontal. In lateral parietal of other "eosuchians" and of arae- view, the frontal is convex in outline, follow- oscelids. The description given here is based ing the curve of the dorsal margin of the on well preserved parietals in KUVP 9951, orbit. In cross-section, the frontal is slightly 9952, 9959a, 33606, and 33607, and on isolated convex over the orbit. As seen in KUVP 9952 single elements in KUVP 9950, 33603, and

(Fig. 8b), the frontal has a strongly devel- 33604. Along the anterior part of its lateral oped ridge on its ventral surface that runs the border the parietal fonns an "L" shaped su- length of the bone, roughly parallel to the ture with the postfrontal and a well developed

midline. It is to this ridge that both the pre- lateral process that joins the postorbital and a

1980 PENNSYLVANIAN DIAPSID REPTILE 17 forms the antcnidorsal rim of the upper tem- only partially preserved in KUVP 9952 (Fig. poral fenestra. The central portion of the 8). Its configuration can, however, be estab- lateral border of the parietal is deeply con- lished on the basis of the deep, narrow groove cave and forms the dorsal rim of the superior on the posterolateral process of the parietal temporal fenestra. The remaining portion of and on the posterodorsal aspect of the squa- the lateral border is formed by a narrow wing mosal, in which this element lay. As in all that is directed posterolaterally and gently captorhinomorphs were this element is pre- ventrally beneath the supratemporal to reach served, the supratemporal in Petrolacosaurus the squamosal. This wing is deeply excavated is a long, narrow, slightly curved bone that on the dorsal surface in order to receive the lies on a process separating the superior tem- elongate supratemporal. poral fenestra from the occipital surface of

The posterior margin of each parietal is the skull. Posteriorly, it is anchored by a deeply embayed between the posterolateral depressed facet on the posterodorsal corner of wing and the midline. At the midline the two the squamosal. This indicates that the supra- parietals form a posteriorly directed nuchal temporal probably serves to strengthen the crest. At the occipital border the parietal ap- parietal-squamosal contact, bracing the pos- pears to have a narrow sheet of bone that terodorsal corner of the skull roof against the extends ventrally onto the occipital surface pull of adductor muscles. to support the and tabulars Tabular.—The tabulars, as seen in KUVP (Fig. 8a). The nearly straight inteiparietal 9952, are relatively small, elongate bones, suture is interrupted by the large, round completely restricted to the occipital sm-face. pineal foramen. This element, although much reduced, re-

The parietal is gently convex anteropos- tains its primitive relationships through con- teriorly; the downward slope is evident in tacts with the postparietal, parietal, supratem- lateral view (Fig. 2). poral, and probably with the supraoccipital In one specimen KUVP 9959a (Fig. 9c), a and squamosal. As reconstructed, the tabular parietal, badly worn at the edges, has been forms a small part of the border of the post- preserved in ventral view. It reveals a slight temporal fenestra. longitudinal crest about halfway from the Postfrontal.—The postfrontal, seen in ex- midline to the upper temporal fenestra. This ternal and internal views in KUVP 9952 and crest presumably marked the position of the 33607, is a relatively large, "L"-shaped bone cartilaginous taenia marginalis of the chon- that foiTns the posterodorsal margin of the drocranium and the point of attachment of orbit and has a long ramus lateroventrally the supraoccipital. Lateral to this crest the beneath the postorbital. The postfrontal, ventral surface of the parietal has a gently as reconstructed, is a shaiply convex bone, convex ridge that follows the medial margin with most of the medial portion occupy- of the upper temporal fenestra and extends ing a dorsal position while most of the onto the posterolateral wing of the bone. lateral ramus extends down on the cheek. Its From this ridge to the temporal opening the anterior apex fits into a small slot on the parietal becomes progressively thinner. frontal. The medial border of the postfrontal Postparietal. —The postparietal, preserved extends far posteriorly, bound by the frontal only in KUVP 33606 (Fig. 11), is completely and the parietal. The shape of the area of restricted to the occipital surface. It is only attachment of these two bones can be seen slightly visible in dorsal view. The element best in KUVP 33606 (Fig. 11). The angle retains its primitive reptilian relationships— between the medial and posterior edges of paired bone in contact with the parietal, tabu- the postfrontal is close to 90 degrees. This

lar, and supraoccipital. corner of the bone is wedged deeply within

Supratemporal.—A right supratemporal is the parietal, and the articulation is reinforced —

18 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

Fig. 12. Petrolacosaurus kansenxis Lane. Mature skull and lower jaws with articulated cervical vertebrae, and an isolated right clethnim, KUVP 33607, X 2. Tlie component .skull elements are disarticulated, and com- pressed into a single plane. The riglit mandible and the \ertel)rac are exposed in lateral view. See Fig. 13 for opposite view of same specimen. See Fig. 2 for key to abbreviations. —

1980 PENNSYLVANIAN DIAPSID REPTILE 19

Fig. 13. Petrolacosmirus kausemis Lane. Mature skull and lower jaws with articulated cervical vertebrae, and an isolated right cleithrum, KU\T 33607, x 2. The component skull elements are disarticulated, and compressed into a single plane. The left mandible lies over the palate. See Fig. 12 for opposite view of same specimen. See Fig. 2 for key to abbreviations. 20 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7 by a facet in the parietal that supports the and left postorbitals of KUVP 9952. In KUVP posterior edge of the postfrontal from under- 33607 the right postorbital was also flattened, neath. The ventrolateral ramus of the post- but the left retains much of its original shape frontal fits imder and follows the curve of the (Fig. 13). The posterior process of the post- upper half of the postorbital. orbital is sutured to the squamosal and forms The ventrolateral process of the post- the anterior half of the intertemporal bar. frontal appears to be massive and triangular Jugal.—The jugal, exposed in lateral and in cross-section. The postfrontal in KUVP medial views in KUVP 9951 (Fig. 6), 9952

9952 ( Fig. 8 ) has a well exposed ventrolateral (Fig. 8), 33608 (Fig. 10), and in KUVP ramus. In lateral view, a slight ridge is pres- 33607 (Figs. 12 and 13), forms part of the ent close to the orbital margin. The medially cheek region below and behind the orbit. inflected anterior margin of the postorbital Its triradiate configuration can be seen clearly fits over this ridge and covers the ramus of in all specimens and is typical of forms with the postfrontal. In medial view, the post- lower temporal fenestrae. The two ventral frontal shows a well developed ridge that rami form not (jnly the lower margins of the extends to the ventral tip of the bone. This orbit and the lower temporal fenestra, but

ridge on the postfrontal is equivalent to the also the ventral border of the skull behind the medially extending plate of bone seen on the maxilla. The suborbital ramus extends far prefrontal. Both ridges reinforce the orbital forward to reach the posterior projection of margin. the lacrimal. In both KUVP 9952 (Fig. 8) Postorbital.—The postorbital, exposed in and KUVP 33607 (Fig. 12) the lateral surface lateral view in KUVP 33607 and in both lat- of the suborbital ramus is deeply recessed eral and medial views in KUVP 9952, has the close to the anteroventral border, indicating triradiate configuration typical of diapsid rep- that in this region the jugal was partially cov-

tiles. Although slightly disarticulated from ered laterally by the maxilla. The lower tem-

the rest of the skull elements, its relationship poral ramus of the jugal, completely pre- to the surrounding bones can be established served only in KUVP 9951 (Fig. 6), is much with confidence. The ventral process of the wider than the other rami. It extends below postorbital extends around the postfrontal the lower temporal fenestra, and probably

covering much of its surface and entering the attaches to the quadrato-jugal. The dorsal posterior orbital border. This process reaches ramus of the jugal forms the lower part of far ventrally to attach to the anterodorsal the postorbital bar. The anterior border of facet of the jugal. The length and surface this ramus bears a long groove that slants detail of the ventral process of the right post- diagonally upward and backward. This groove orbital in KUVP 9952 indicates that almost receives the ventral process of the postorbital. one-half of this process was applied to the The orbital margin of the jugal is thick- inside surface of the jugal, and must have ened on the medial surface. Immediately extended near to the jugal-ectopterygoid su- above the posterior end of the maxilla, this ture. The concave anterior rim of the post- thickened region of the jugal has a well devel-

orbital is thicker than the rest of the bone oped medially projecting tubercle that is su-

and is medially inflected. The posterior mar- tured to a stout lateral process on the ecto- gin of the ventral process of the postorbital pterygoid. This tubercle, partially exposed forms part of the anterior border of the lower in KUVP 33607, but also seen in KUVP 9951 temporal fenestra. The dorsal process forms and 9959a, provides the only contact between the anterior border of the upper temporal the palate and the jugal. The posterior ramus fenestra and extends as a narrow slip of bone of the jugal is smooth and ver>' thin. between the postfrontal and parietal. This Quadratojugal.—The quadratojugal, which dorsal process was flattened in both the right is only partially preserved in KUVP 33603, 1980 PENNSYLVANIAN DIAPSID REPTILE 21

33607, and 3360S is a thin, elongate sheet of bone, probably wrapped around the dorsal bone that forms the ventral border of the process of the quadrate to present in occipital skull behind the jugal, and meets the squa- view a narrow strip that forms the lateral mosal behind the lower temporal fenestra. border of the posterior temporal fenestra and

Although this bone is poorly preserved in all may have also provided support for the tabu- the above specimens, there is evidence to in- lars. The posterior region of the squamosal dicate that the jugal and quadratojugal attach is, therefore, similar in Petrolacosaurus and to one another below the temporal fenestra. captorhinomorphs and more primitive than in In KUVP 33608 (Fig. 10), the badly pre- all other diapsid reptiles, where the squa- served jugal and quadratojugal appear to be mosal does not cover the dorsal process of the in contact. In addition, the height and length quadrate completely. In all known Permian of the quadratojugal, relative to the quadrate and Triassic "eosuchians" it is the quadrate in the immature specimen KUVP 33603 (Fig. that forms the concave posterior margin of 9c), indicates that the quadratojugal extended the skull and eventually the otic notch. There

far anteriorly below the lower temporal open- is no otic notch in Petrolacosaurus, Arae-

ing as a tall, thin sheet. The jugal also ex- oscelis or primitive captorhinomorphs.

tends posteriorly as a relatively tall, thin The shape and configuration of the pre-

sheet; its posterior outline is indicative of an served squamosals also indicates that, dor- overlapping suture with the quadratojugal. sally, this bone enters the superior surface of The quadratojugal, therefore, not only forms the skull, contributes to the medial edge of the ventral edge of the lower temporal fenes- the upper temporal fenestra, supports the tra, but also contacts the jugal. The quadra- supratemporal, and comes in contact with the tojugal wraps posteriorly around the dorsal parietal. Ventrally the squamosal overlaps process of the quadrate as indicated in KUVP the quadratojugal.

33603. The squamosal in KUVP 33607 is thickest Squamosal.—Although seen in many speci- at its center of ossification, which is located mens, the squamosal is incompletely pre- at the level of the intertemporal bar. From

served in all, and is therefore, difficult to here the bone thins anteriorly and anteroven- interpret. The size and shape of the squa- trally, but remains thick posteroventrally. The mosals in KUVP 33607 and 33608 indicate posterior border of the squamosal in KUVP

that most of the cheek region behind the tem- 33604 and 33608 is also thick. At margins of

poral fenestrae is covered by this large plate- the temporal fenestrae, the squamosal is very like bone. The anterior portion of the right thin. The anteroventral corner of the squa-

squamosal in KUVP 33607 (Fig. 12) shows mosal is not preserved in any specimen. very clearly that, as in all primitive diapsids, Dermal Bones of the Palate this bone forms the rounded posterior borders of the upper and lower temporal fenestrae Vomer.—The vomers are best preserved and extends an anterior projection between in KUVP 33608 (Fig. 10) and KUVP 33607 the fenestrae to form the posterior portion of (Figs. 12 and 13) where they appear as the intertemporal bar. Although incomplete, paired sheets of bone lying anteriorly along the posterior border of the squamosals in the midline of the palate. In ventral aspect

KUVP 9952 (Fig. Sb), 9959a and 33607 the vomer is a long roughly triangular ele-

(Figs. 12 and 13) is sufficiently preserved to ment, occupying 28 per cent of the total indicate that this element forms the rounded length of the skull. The entire lateral border posterior border of the cheek region, as in of the vomer forms the medial rim of the primitive captorhinomorphs. In KUVP 33607 internal naris. Anteriorly and posteriorly the

the long posterior strip of the right squamosal vomer is sutured to the premaxilla and pala-

which is partly separated from the rest of the tine, respectively. The two vomers are su- 22 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7 tiired to each other at the midline along t\vo- reptiles, forming the anteromedial border of thirds of their length. Posteromedially the the suborbital fenestra. The palatine is not vomers are separated along the midline by a ridged along the rim of the fenestra. Small narrow wedge-shaped anterior extension of denticles, similar to those seen on the vomer, the pterygoids. cover the surface of the palatine ventrally. In

Medially the vomer is strengthened by a addition, a ridge with larger teeth runs di- longitndinal ridge on the dorsal and ventral agonally across the palatine from its postero- surfaces (Fig. 10). The ventral ridge carries medial corner toward the maxillary process. a row of large teeth that approaches but does In dorsal view, the palatine overlaps the not reach the premaxilla anteriorly. The lateral edge of the pterygoid and forms a posterior three-fourths of the ventral surface stout process anteriorly over the dorsal sur- of the vomer is also covered by small denti- face of the vomer. The ectopterygoid appar- cles. The ventral surface is concave in cross- ently overlaps the posterior part of the dorsal section. The lateral border of each vomer surface of the palatine. curves ventrally but is not ridged. Ectopterygoid.—The left ectopterygoid is The dorsal surface of the vomer is more preserved in KUVP 9952 (Fig. 8). It is complex than the ventral, with a tall median slightly out of position relative to the jugal ridge, a flat central surface and a longitudinal and is seen in partial lateral view. A well depression near the lateral rim ( KUVP 33607, developed process of the cctopterygoid pro- Fig. 13). The median ridge probably jects laterally to articulate with a correspond- strengthened the contact between the vomers ing tubercle in the jugal, as it does in Youn- and supported the nasal septum. The flat dor- gina and other primitive diapsids. Anterior sal surface of the vomer is narrow anteriorly to this the lateral border of the cctopterygoid but widens posteriorly, where it is partially is free of sutural contact and forms the pos- covered by the large anterior process of the teromedial and posterior boundary of the palatine. The longitudinal depression is a suborbital fenestra. Posteriorly, the ectoptery- deeply concave surface, bound medially by a goid forms the anterior boundary of the sub- ridge that probably represents the medial temporal fossa; medially it is suturally at- boundary of the internal naris. tached to the pterygoid. Posteromedially this Palatine.—The palatine in KUVP 3360S attachment is formed by a massive rounded

(Fig. 10) and KUVP 33607 (Fig. 13) is a abutting suture. Anteromedially, the ecto- rectangular sheet of bone pressed against the pterygoid fomis a small sheet of bone that lateral edge of the pterygoid. The shorter extends onto the dorsal surface of the ptery- sides of the rectangle form the sutures with goid. Anteriorly, the ectopterygoid overlaps the vomer anteriorly and the cctopterygoid the dorsal surface of the palatine. posteriorly. The long lateral side of the pala- Ptenjgoid.—The pterygoid, seen in both tine extends in a wide process to the maxilla dorsal and \cntral \iews in KUVP 8351, anteriorly as it does in Youngina and other 33604, 33607 and 3.360S, is the longest bone diapsids. This massive process occupies one- of the skull, similar in most features to that third of the total length of the bone, its su- in protorothyridids (Carroll and Baird, 1972).

tural surface is concave and scarred thus in- Although partially covered by other skull ele- dicating that the palatomaxillary suture was ments, its configuration and relationship to heavily stressed. Between the maxillary proc- the surrounding bones can be established with ess and the anterior suture with the vomer, confidence. The pterygoid has four processes the palatine forms the posterior margin of common to all primitive reptiles: the palatal the internal naris. ranuis, the transverse flange, the quadrate Behind the maxillary process, the lateral ramus, and the small process for the basi-

edge of the palatine is free as in all diapsid pterygoid articulation. 1980 PENNSYLVANIAN DIAPSID REPTILE 23

The palatine ramus extends anteriorly Fig. 5a shows an extensive area of attach- from the region of the basiptcrygoid articula- ment to the epipterygoid.

tion to its acuminate temiination between the The basicranial process of the pterygoid is vomers; it is widest where it articulates with not well preserved in most specimens. In

the ectopterygoid. Along the posterior part KUVP 33604 this process is exposed in dorsal of its medial edge the palatine ramus bounds view. It shows that the anteromedial part of the interpterygoid vacuity. A dorsally in- the process has a small dorsally oriented ar- flected ridge is present on the anterior one- ticulating surface for the attachment to the third of the rim of the interpyterygoid va- basisphenoid. Most of the articulation, how- cuity. This inflected ridge continues anterior- ever, must have been provided by the epi- ly and helps to form a greatly strengthened pterygoid. suture between the pterygoids along the mid- line. The palatine ramus carries two longi- Ossifications of the Palatoquadhate tudinal ridges on the ventral surface. These Cartilage bear irregularly arranged teeth, up to 0.5 mm in diameter and not more than 1.0 mm in Epipterygoid.—A small triangular piece of height. The medial ridge runs to the anterior bone lying on the crumpled left quadrate

tip of the pterygoid and is continuous with ramus of the pterygoid in KUVP 8351 has the tooth bearing ridge on the vomer. The been identified by Peabody (1952), with

shorter lateral ridge is continuous with a some misgivings, as the epipterygoid. This

similar ridge on the palate. The rest of the identification is of doubtful validity, how-

palatal ramus of the pterygoid is covered ever, and the element is not included in the with numerous small denticles, about O.I mm reconstruction. In the same specimen, the in diameter. dorsal expanse of the quadrate ramus of the The transverse flange of the pterygoid right pterygoid is well preserved. Dorsome- extends anterolaterally at about 60 degrees to dially. the quadrate ramus of the pterygoid is the sagittal plane. The flange slants ventro- strongly grooved at the level where the epi- laterally and extends below the lower margin pterygoid should be located. This attachment of the cheek. The ventral surface of the area is extensive, running posteriorly from the

flange is covered with a large number of ir- region of the basiptcrygoid process to the regularly arranged teeth of varying sizes. The dorsal process of the quadrate. The conclu- largest teeth, located on the posterior rim of sion that the epipterygoid extended so far the flange, are up to 0.8 mm in diameter and anteriorly is supported by the fact that the 1.2 mm in height. The anterior portion of the articulating surface of the basiphenoidal proc-

flange is covered by smaller teeth. ess faces dorsolaterally and that the engaging process on the pterygoid is too thin in all the The quadrate ramus, as seem in KUVP specimens, where this region is preserved, to 8351 (Fig. 5a and b) is essentially a large provide support on its own. In addition, the vertical plate running posterolaterally from epipterygoid in pelycosaurs and captorhino- the basiptcrygoid region to the quadrate. The morphs fonued partly, or wholly, the articu- ventral edge of the quadrate ramus is thick- lation with the basisphenoid (Heaton, 1979). ened to form a ridge, above which the inner The groove on the pterygoid for the base side is slightly concave. Posteriorly, the ra- of the epipterygoid is not excessively long in mus is attached to the inner surface of the PetrolacosauTus when comparison is made dorsal process of the contact quadrate. The with other primitive reptiles. In Dimetrodon between the two bones is extensive both (Romer and Price, 1940, PI. 7), for example, anteroposteriorly and dorsoventrally. The the base of the epipterygoid extends from laterodorsal surface of the quadrate ramus in above the dorsal process of the guadrate to 24 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7 the interpterygoid vacuities, well past the noid is covered with denticles at the level of level of the basicranial articulation. the basisphenoidal tubera. The flanks of this Quadrate.—In Petrolacosaurus, the quad- rostrum are not well preserved. The location of the anterior internal carotid foramina can- rate is a tall, well developed element, similar not established. sides of the para- in configuration to that in captorhinomorphs be The sphenoid are formed by two rounded crests and pelycosaurs. It is composed of a rela- (Figs. 5b, 7a, and that converge an- tively flat sheet of bone that extends dorsally 13) and anteriorly from the articular condyle. The teriorly and widen posteriorly. These crests, articular area of the quadrate seen in ventral the cristae ventrolaterales, leave a wide de- pression on the ventral surface them. view in KUVP 33603, 33607 and 33608, con- between The posterior end of these crests extend lat- sists of two subhemispherical surfaces, a lat- the basioccipital. ventral eral and a medial, separated by a groove. The erally beyond The between two lateral surface faces ventrally. The medial surface of the parasphenoid the is a thin sheet of surface, on the other hand, faces ventrolater- lateral ridges formed by bone underlying the anterior end of the ally and is located farther anteriorly and ven- basioccipital. trally than the lateral surface. The tall dorsal parasphenoid process of the quadrate preserved only in Exposed in dorsal view, the in thin sheet of bone on KUVP 33608, joins the quadrate process of KUVP 33608 has a the pterygoid anteriorly. The posterior mar- each side that probably extended dorsally (Fig. to attach to the prootie, contrib- gin of the dorsal process of the quadrate is 10b) covered by the squamosal and quadratojugal uting to the lateral wall of the braincase. basi- except in the region where the palatoquadrate Basisphenoid.—In ventral view the foramen emerges, precluding the possibility sphenoidal tubera project anterolaterally be- of fonuing the margin of a squamatc type of yond the dermal parasphenoid. Each process in otic notch. The lateral surface of the quadrate has a slightly constricted waist and ends a forms part of the posteromedial wall of the gently convex articular head. The articular adductor chamber. area faces mainly antcrodorsally and is little exposed in ventral view. Preservation of all the known basisphenoid-parasphenoid com- Ossifications of the Braincase plexes either in dorsal or ventral views, makes Varasphenoid.—The parasphenoid is a me- identification of the suture between these ele- dian dermal element that provides a ventral ments impossible. cover for the braincase. In ventral view, the The basisphenoid in Petrolacosaurus can region of the parasphenoid ventral to the be seen best in dorsal view (Figs. 5a, 7b, and basisphenoid has the general outline of an 10). The anterior portion of the basisphenoid isosceles triangle, typical of all primitive cap- forms a short rostral process. Its anterior

torhinomorphs and early "eosuchians." Tlie termination is deeply incised between the cultriform process, seen in both dorsal and cristae trabeculares to carry a small branch of ventral views in KUVP 9951 (Fig. 7) extends the anterior internal carotid artery. The cris- forward between the interpterygoid vacuities. tae trabeculares form the point of attachment

It is "V"-shaped in section and bears a row of the trabecula communis to the basisphe-

of tiny denticles on its posterovcntral aspect, noid. as in primitive captorhinomorphs. The cultri- Along the midline, just above the basi- form process probably supported the sphe- sphenoidal tubera, a small pit in KUVP 9951 nethmoid. and 33608 (Figs. 7b and 10) probabb- marks The triangular portion of the parasphenoid the exits of the anterior internal carotid arte- hides most of the basisphenoid from ventral ries, and their subdivisions, the pituitary ar- view. The ventral surface of the parasphe- teries. A small trans\erse ridge separates this 1980 PENNSYLVANIAN DIAPSID REPTILE 25 region from the more posteriorly located sella exoccipital forms the medial border of the turcica. vagus foramen. Below this foramen, and The basisphenoid supports a well devel- slightly medially, are two small hypoglossal oped dorsum sella, comparable in height to foramina. At the base, the pair of exoccipitals that in modern iguanid and nearly as approach each other but do not join at the tall as in Captorhitms. The dorsum sella in midline (KUVP 33608, Fig. 10). Petrolacosaurus, exposed in KUVP 8351 and Prootic. —The prootic is the least well

33608 (Figs. 5a and 10) is a thin trans- known element of the braincase in Petrolaco- verse sheet of bone strengthened laterally saurus. This element, preserved only in KUVP by the sellar processes. In both specimens 9951, is similar in outline to that of Capto- the dorsum sella and the sellar processes have rhinus (Price, 1935). been crushed postmortally down onto the Supraoccipital.—The supraoccipital, also dorsal surface of the parasphenoid obscuring exposed in KUVP 9951 and 33607, forms the the cavum cranii, the largest concavity of the broad occipital plate. It not only forms the basisphenoid, where the anterior body of the arched roof of the foramen magnum but also medulla rested. extends down on both sides to the exoccipi- Basioccipital.—The short, stout basioccipi- tals. tal forms the ventral surface in the posterior Dorsomedially, the postparietals overlie portion of the braincase and the major por- the process of the supraoccipital that extends tion of the occipital condyle. The ventral sur- to the parietals. Dorsolaterally, the tabular face of the basioccipital is fully exposed pos- covers the slightly ridged process of the su- teriorly; its anterior portion, however, is cov- praoccipital that extends to the squamosal- ered by the posterior end of the plate parietal suture. of the parasphenoid. The well developed Posteriorly, the supraoccipital bears a pro- basioccipital tubercle, which forms most of nounced median ridge separating the prob- the exposed ventral surface of the basioccipi- able insertions of the M. recti capiti posterior. tal, has a transverse ridge anteriorly between This ridge runs from the postj^arietals to the the cristae ventrolaterales of the parasphenoid, foramen magnum, where it fans out laterally. and a longitudinal ridge that extends along In KUVP 9951, the supraoccipital has been the midline to the occipital condyle. The crushed flat, exaggerating the lateral extent of margins of this tubercle probably formed the this bone. Normally the two dorsolateral anterior and medial limits of the M. rectus "wings" of the supraoccipital, exposed in Fig. anterior insertion. 7, face mostly laterally and only slightly pos- Unlike those of pelycosaurs and capto- teriorly; they formed the lateral wall of the rhinomorphs, the basioccipital in Petrolaco- braincase behind the prootic. saurus is not completely covered dorsally by Opisthotic.—The opisthotic, also found the exoccipitals. The latter do not join at the only in KUVP 9951 and 33607, is partially pre- midline to exclude the basioccipital from the served. The available material indicates that, floor of the foramen magnum. as in pelycosaurs, Araeoscelis and Youngina, Exoccipital.—In posterior view the ex- the opisthotic in Petrolacosaurus forms an os- occipital has a wide base that forms the sified paroccipital process. Unlike Araeoscelis, dorsolateral part of the condyle. Above the however, this process did not quite reach the condyle, an ascending process runs up along quadrate. In addition, a ventral process of the side of the foramen magnum and is ap- the opisthotic extended down along the lat- plied dorsally to the ventral edge of the su- eral margin of the exoccipital-basioccipital praoccipital. The posterior surface of the complex. The level of its ventral termination ascending ramus bears a facet for the proat- is, however, indeterminate. las. Between this and the main body, the Stapes.—The stapes preserved in KUVP 26 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

8351 and 33607 retain the pattern found in all portion of the mandible posterior to the den- primitive Paleozoic reptiles. It is a relatively tary and splenial, and continues back to the massive bone consisting of a large proximal end of the jaw as a deep keel below the sur- head and a gently tapering columellar process angular and prearticular. This bone is over- pierced by the stapedial foramen. The proxi- lapped aiiterodorsally by the dentary, but it the mal head is poorly preserved, but it probably overlaps the surangular and articular on consisted of a footplate and a dorsal process. lateral surface and the prearticular on the in- The distal tip of the columella in KUVP 33607 ternal surface of the mandible. Below the forms the floor is slightly expanded and its surface is un- prearticular fossa the angular finished. of the Meckelian canal and part of the lateral wall. Mandible Surangular.—The surangular, also exposed Dentanj.—This long, unusually slender ele- in KUVP 33607, is a simple sheet of bone that ment, best preserved in KUVP 33606 and occupies the posterior portion of the mandi-

33607, carries the mandibular tooth row and ble. Its lateral exposure is relatively small;

occupies a considerable area of the outer it is overlapped by both the dentary and the

surface of the mandible. Anteriorly, it ex- angular. Internally, it is overlapped by the cludes the splenial from the lateral surface. posterodorsal process of the coronoid. Pos-

It is bounded posteroventrally by a narrow teriorly, the surangular has a large exposure slip of the splenial. Posteriorly, the dentary along the lateral wall of the adductor fossa. covers the angular and surangular and nar- The upper edge of the bone is somewhat

rows dorsally. Its posterior tip lies near the thickened; posteriorly, it swings inward to dorsal margin of the mandible, covering most, attach to the articular and form the posterior

but not all, of the coronoid. The single lat- wall of the fossa. eral tooth row is supported by a medially Splenial.—The splenial, as seen in KUVP directed shelf on the internal surface of the 33607 and 33608, forms most of the anterior dentary, close to the dorsal edge of the bone. portion of the medial mandibular surface. The

The lateral margin of this shelf is, as in the lateral exposure of this bone is limited to a case of the maxilla, much higher than the narrow strip below the posterior part of the medial one. Anteriorly, the shelf forms most dentary. Anteriorly, the splenial is in contact

of the symphysial surface. This area is ex- with the dentary above and below the anterior

posed only in KUVP 3.3607; it is quite small opening of the Meckelian canal. This contact

and bears no marked ridges. Behind the sym- is maintained for some distance posteriorly

physial area, the dentary is completely ex- with the splenial forming the medial wall and posed medially to the level of the eighth floor of the canal. Posterodorsally, the splenial tooth. Further posteriorly, the medially di- covers the anterior projection of the coronoid.

rected shelf is strongly ridged, indicating the A posteriorly directed thin sheet of the splen- area of attachment of the splenial and coro- ial covers part of the prearticular and pos-

noid. Below the shelf, the dentary is strongly terior splenial and comes in contact with the excavated for the channel of the Meckelian angular. canal; the dentary forms the roof and lateral Posterior splenial.—A small elongate sheet

wall of the canal in its anterior portion. In of bone, that tapers to a point both anteriorly KUVP 33607 there are 28 teeth attached to and posteriorly, lies on the medial surface of the dentary with shallow alveoli for five more. the mandible in KUVP 33607. Anteriorly, a Anp,ular.—The angular, also preserved in large portion of this bone was probably cov- KUVP 33607, is a large "V"-shapcd element, ered by the splenial. It is sutured to the pre- occupying part of the medial and lateral sur- articular dorsally and the angular ventrally. faces of the mandible. It forms the ventral This bone may be comparable to the posterior 1980 PENNSYLVANIAN DIAPSID REPTILE 27 splenial seen in some amphibians; it presuma- KUVP 33607 is actually part of the rounded bly formed part of the medial wall of the posterior end of the articular that was pushed Meckelian canal. A posterior splenial has not posteriorly when the lower jaw was crushed. been reported in other Paleozoic reptiles. A short angular process extends medially and Prearticuhr.—The prearticular, a major somewhat ventrally from the articular surface element on the inner surface of the mandible, of the articular, probably for the insertion of is well preserved only in KUVP 33607. Pos- the superficial pterygoideus musculature. teriorly it is applied to the inner surface of the articular. The suture between these two Dentition bones is not visible. The conjoined edges of The teeth in Petrolacosaurus are simple the prearticular and articular are drawn out conical structures. In both the upper and medially to form a small shelf at the postero- lower jaws there is a marginal scries of sub- ventral corner of the adductor fossa. From thecodont teeth. There are also many palatal this area, the prearticular extends anteriorly teeth. The marginal teeth, arranged in a as a sheet of bone along the inner surface of single row, are slightly compressed and sharp- the mandible, forming the medial margin of ly pointed. Towards the tips, the teeth bend the adductor fossa and most of its medial slightly posteriorly and are marked by longi- wall. The prearticular is overlapped below tudinal grooves. This longitudinal grooving, or by the angular and the postsplenial. Anterior- fluting, occurs only towards the tip and is ly, the prearticular is covered by the posterior often seen in well-preserved Paleozoic rep- end of the splenial. Anterior to the adductor tiles. fossa, the prearticular is again drawn out medially to form a shelf ventral to the coro- On the prcmaxilla, the anteromedial tooth noid. In this region, the prearticular forms is the largest. Posterolaterally, the premaxillary part of the medial wall of the Meckelian teeth show progressive diminution in size. canal. The first three teeth tend to be inclined slight- Coronoid.—Only a single edentulous coro- ly anteroventrally. Two teeth in the anterior portion of the maxilla are larger than the noid is present in Petrolacosaurus. In KUVP others as ca- 33607 the anterior half of this bone is a thin and hence may be designated sheet, applied to the medial surface of the niniform teeth. This is a primitive dentary below the marginal tooth row, and condition that persists in many Permian rep- in contact with the splenial anteriorly and the tiles. In Petrolacosaurus, the caniniform teeth prearticular ventrally. A thicker area forms are only slightly longer than the first pre- the anterior margin of the adductor fossa and maxillary teeth. These relatively large teeth, to premaxillary teeth, part of its dorsal rim. The posterodorsal end the two four smaller of the bone lies beneath the surangular, on and the four pre-caniniform teeth create a the outer part of the mandible. seemingly efiBcient food trap. The post-ca- Articular. —The articular joins the lateral niniform teeth, although definitely smaUer and medial walls of the adductor fossa and than the caniniform teeth, remain quite large forms the articulating surface with the quad- until well posteriorly in the series; only the rate. This surface consists of two parallel, final eight to ten teeth show successive dimi- elongate concavities, the inner groove lying at nution in size. The upper tooth row continues a lower level than the other. Externally, the back to the posterior extremity of the maxilla. articular bone is almost completely covered The mandibular marginal dentition is ba- by the surangular and angular, but forms the sically homodont with reduction in tooth convex posterior end of the mandible. There height only in the last five to seven teeth is no distinct retroarticular process. What (KUVP 3.3606, Fig. 11). The lower tooth row looks like a small retroarticular process in does not seem to extend back to the extreme 28 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

end of the dentary. The teeth

teeth sit on a medially directed shelf and are between five larger, nearly rectangular plates. supported laterally by a thin sheet of bone. In In both specimens the sclerotic plates were medial view, the teeth show slight vertical flat, but this may be due to crushing. ridging close to the base. Cross-sectional views of the maxillary teeth (Fig. 9b) show Vertebrae that they are essentially thin walled eones with an open base. Large pulp cavities extend Notwithstanding certain specializations, the vertebral of Petrolacosaurus close to the tip of the teeth. The teeth sit column re- in shallow pits on the jaw. Implantation has tains most of the general characteristics seen been strengthened by a thin sheet of bone in the captorhinomorphs. Each segment of that extends from the circular margin of the the column includes a centrum, neural arch pit onto the lateral surface of the tooth base and intercentrum. Each centrum has the basic hourglass configuration, with expanded and is solidly fused to it. This fusion gives the appearance of ridging both on the outside hollow ends facing adjacent centra and a waist. and the inside of the teeth close to the base. narrow A ventral longitudinal ridge Palatal teeth occur on the pterygoids, pala- extends between the anterior and posterior tines, vomers and the parasphenoid. Rela- ends of the centrum to strengthen the hour- glass structure. Anterodorsally, tively large teeth (up to 1.2 mm in length), thickened irregularly distributed along ridges on the areas serve as points of attachment for the neural all palate, extend in three directions from the arches. Thus, the centra except the first are amphicoelous all basipterygoid area. A variable number of and are notochordal. large teeth extend laterally onto the trans- The notochordal canal is strongly constricted verse flange of the pterygoid. Another group at the middle of each centrum. The neural of slightly smaller teeth extend diagonally arches are fused solidly to the centra and su- tures across the pterygoid and palatine. A third have been obliterated. The articular group extends anteriorly along the medial areas of the zygapophyses are simple plane edge of the pterygoid and the vomer. surfaces, somewhat tilted from the horizontal. The intercentra are well developed. Numerous small teeth (about 0.3 mm in length) are rather evenly distributed over A presacral vertebral count of 26 is defi- most of the palatal plate between the rows nitely established on the basis of several ar- of larger teeth. The teeth on the parasphe- ticulated specimens. There are six cervical noid are also cjuite small. A niunber of broken \ertebrae that form a distinct neck region. teeth indicate that both large and small pala- The calculated length of the cervical series in tal teeth have well developed pulp cavities. KUVP 9951 is about 72 mm; the calculated There are no teeth on the inner surface of the length of tlie dorsal series in the same indi- . vidual is about 192 mm. There are two sacral vertebrae as is typical in most primitive rep- Sclerotic Plates tiles. An approximate count of 60 to 65 for

Series of thin translucent sclerotic plates the number of caudal vertebrae is based on lie in the orbit of the skulls in KUVP 9952 two partial specimens, KUVP 1427 and KUVP —

1980 PENNSYLVANIAN DIAPSID REPTILE 29

1428 (Peabody, 1952, Figs. 2 and 4). The The articulating surface of the posterior 2^'ga- distal part of the tail is composed of at least pophysis faces medially so as to be aligned twelve spool-shaped centra without neural almost with the saggital plane. An anterior arches. The calculated length of the caudal zygapophysis is also present but faces for- region based on these two immature speci- ward to articulate with the corresponding mens is about 315 mm. facet on the exoccipital. The proatlas in

Petrolacosaurus is similar in outline to the Cervical Vertebrae proatlas in the primitive captorhinomorphs Atlas-axis complex.—The proatlas (Fig. Hijlonomus and Paleothtjris; it is much short- 14) is a relatively small but robust element. er than the same element in the The two elements of the pair do not appear Ophiacodon. to be in contact. There is a rudimentary neu- The atlas in Petrolacosaurus includes four ral spine on each element and a small posterior elements: the intercentrum, two neural zygapophysis that articulates with the atlas. arches and the centrum. The intercentrum

A ii 1)

c:::^

Fic. 14. Petrolacosaurus Jcansensis Lane. Cervical vertebrae, all x 1.5. (A) Lateral view of left proatlas, KUVP 33607. (B) Ventral and anterior views of atlantal intercentrum, KUVP 33604. (C) Lateral and me- dial views of left atlantal neural arch, KUVP 33604. (D) Anterior and lateral views of atlantal centrum and fused axial intercentrum, KUVP 33604. (E) Outline of the atlas-axis complex, based on KUVP 33604 and 33607. (F) Lateral view of axis, KUVP 33604. (G) Lateral and dorsal view of third cervical vertebra, based on KUVP 9951 and 9956. (H-I) Lateral view of fourth and fiftli cervical vertebrae, KUVP 9951. (J) Lateral and dorsal views of the si.xth cervical vertebra, based on KUVP 9951 and 9956. 30 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7 has the usual crescentic shape, but is slightly notochordal canal is completely closed by a larger than its countcqoarts in the remainder more fully ossified centrum. In these mature of the column with the exception of the axial individuals the upper part of the anterior sur- intercentrum. A strongly concave anterior face surrounds the relatively small noto- surface receives the ventral portion of the chordal pit and articulates with the central condyle. The posterior surface, which is less portion of the condyle. The posterior surface deeply cupped, is apposed to the axial inter- of the centrum articulates with the upper half centrum. Laterally there is a small articular of the axial centrum and is of normal con- surface on each side for the capitulum of the struction. first cervical rib. This is a rather primitive Ventrally, the atlantal centrum sits on the feature seen in captorhinomorphs and pely- axial intercentnmi and does not reach the cosaurs. Ventrally the intercentrum has the ventral surface of the column. Even in ma- longitudinal ridge seen in the other cervical ture specimens, although the above two ele- intercentra. ments are fused, the areas of fusion are The atlantal neural arch (Figs. 12, 13 and marked by a wide band of unfinished bone.

14 ) is a paired structure. There is a thin shelf A somewhat similar pattern is seen in Heleo- of bone above the neural canal, but the two satirus (Carroll, 1976). sides apparently are not in contact. The post- The axial intercentrum is relatively large, zygapophysis extends far posteriorly to artic- exceeding in length all the other intercentra. ulate, by a medially directed face, with the This intercentrum is more plate-like than cres- prezygapophysis of the axis. A posteriorly centic and has the well developed longitudi- directed spine, external to this facet, extends nal ridge seen in the atlantal intercentrum. for a short distance beyond the point of artic- In lateral view, it has the shape of a right ulation with the axis. At the anterodorsal an- angle triangle, the sides of the 90 degree gle of the atlantal neural arch a roughly circu- angle facing posteriorly and ventrally. There lar surface faces laterally and slightly dorsally are no obvious facets for rib articulation at to connect with the proatlas. Anteriorly, the the lateral margins of the element as seen on neural arch articulates with the dorsolateral the atlantal intercentrum. portion of the condyle. A wide ventral proc- The axial centrum-neural arch complex, ess descends over the lateral surface of the well preserved in KUVP 33607 and 9951, is a atlantal centrum and articulates with it via a robustly built stnicture in Petrolacosaurtis, medially directed pedicel. The base of the lying at the functional end of the vertebral neural arch has a small ventrally directed column and linking it to the skull. Its dual process for articulation with the tuberculum function is clearly indicated by the form of of the atlantal rib. the axial neural arch (KUVP 33607 and

The atlantal centrum is well developed 9951), which posteriorly is not greatly differ- although considerably smaller than other pre- ent from that of the succeeding cervical verte- sacral centra. Dorsally, it is gently grooved brae. Anteriorly, it is highly modified to pro- to form the floor of the neural canal. Antcro- vide origins for muscles inserting on the occip- dorsally, there is on both sides a well devel- ital region of the skull and to support the oped articulating surface for the pedicel of skull through the remainder of the atlas-axis each neural arch. Behind this area the upper complex. The neural spine is strongly built, and lateral surfaces of the centrum are free. unusually thickened and elongated antero-

In immature specimens ( Peabody, 1952, Fig. posteriorly to a length about equal to that of 3), the centrum has, in anterior view, the the centnnn below. The posterior border is shape of an inverted "U" that sits loosely on heavy, and in one specimen small processes the axial intercentrum. In more mature speci- near the top project posteriorly, possibly mens (KUVP 33607, Fig. 12), however, the marking the origin of the M. spinalis capitis. 1980 PENNSYLVANIAN DIAPSID REPTILE 31

The massiveness and the anteroposterior elon- than the average mid-dorsal centrum, and in gation of the spine unquestionably provides vertebra 6, it is 1.35 times longer; the centra added support to such muscles as the M. in vertebrae 2 to 5 are of equal length, being rectus capitis and the M. obliquus capitis. 1.55 longer than the average mid-dorsal cen- The anterior portion of the neural spine proj- trum. All these elongate centra have a slightly ects anteriorly dorsal to the posterior part of smaller vertical diameter than the dorsal the atlantal centrum and neural arches, as in vertebrae. In all cervical vertebrae the an- Araeoscelis, and other primitive reptiles. terior centrosphenes are inclined anteriorly. The postzygapophyses of the axial neural This inclination is most marked on the an- arch are normally developed; the prezyga- terior cervical vertebrae, where the posterior pophyses are weak, project little from the centrosphenes are also tilted forward. This surface of the arch, and face laterally as well specialization in Petrolacosaurus would have as slightly upward. A ridge runs posteriorly permitted raising, and possibly maintaining, from the prezygapophyses. The transverse the head high above the ground. process or diapophysis is stoutly built, al- There is no bevelling on the centra for though not greatly elongated, and originates accommodating the large intercentra; the lat- far ventrally on the neural arch. It projects ter were probably continued dorsally in carti- ventrolaterally at about 45 degrees to the lage. In all cervical vertebrae, the longitudi- sagittal plane. The diapophysis is buttressed nal ridge at the bottom of the centrum ex- by converging ridges from the neural arch tends far ventrally, fomiing a strong keel, and centrum. The articular surface is small, with its lateral surfaces concave in section. however, and keyhole-shaped. The keel of the anterior cervical vertebrae, The neural arch contributes significantly forms a nearly straight line between the ends to the anterior centrosphene, which has a of the centrum, and is "V"-shaped in section. much greater vertical extent than the posterior On the posterior cervical vertebrae this keel articulation. This greater anterior articulating does not extend as far ventrally and tends surface provides suflBcient contact for the towards a more rounded ventral margin. combined height of the atlantal centrum and The cervical neural arches are attached by

axial intercentrum. This basic pattern is seen a pair of pedicels to the anterior half of the in such diverse forms as Ophiacodon. Ileleo- centra. Anteriorly, the margin of the pedicels saurus, Araeoscelis, Palcothyris. Platecarpus, form the dorsal half of the centrosphene. and . The construction of the atlas- Above the centrosphene, the anterior edge of axis complex indicates that movement of the the arch forms a slight notch before curving head was concentrated at the articulation be- up and anteriorly beneath the prezygapophy-

tween the condyle and the atlantal elements. sis; this notch forms the posterior boundary of

There appears to have been little movement the intervertebral opening. The dorsal margin

between the atlantal centrum-axial intercen- of the posterior end of the pedicel is far for- trum and the axial centrum. ward on the upper surface of the centrum. The cervical vertebrae in Petrolacosaurtis From this point the posterior margin of the (KUVP 9951, 9956) are conspicuous in their arch curves strongly to form the anterior degree of elongation. A number of later but boundary of the intervertebral opening. The still relatively primitive reptiles, such as Arae- elongate shape and anterior extent of this

oscelis, , Macrocnemits, and Tanij- opening is in contrast with the situation in stropheus are also long-necked. the dorsal region of the column where this

In the anterior part of the vertebral col- opening is rounder and considerably smaller. umn, an abrupt elongation of centra starts The transverse jjrocesses on vertebrae 2 with vertebra 7 and continues to the axis. The to 7 show the following gradual changes from

centrum in vertebra 7 is 1.25 times longer the pattern in vertebrae 2: (1) shift to a 32 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7 higher position on the lateral surface of the are large and are curved transversely. The neural arch; (2) increased lateral elongation; plane of the articular surfaces is close to the (3) change of angle of projection of the horizontal, but the exact angle cannot be process from 45 degrees to the sagittal plane established. Because of the large articular in vertebra 2 to almost 90 degrees in verte- surfaces, the members of each pair of zyga- bra 7; (4) increase in the size of the articular pophyses are not widely separated. surface with the rib tuberculum and accom- The neural spines are situated well toward panied change in shape to an oval head plus the posterior end of the vertebrae, centered an attached thinner ridge than runs antero- between the postzygapophyses. Each spine is ventrally to the anterior centrosphene (there bordered both anteriorly and posteriorly by is apparently no gap on the process to sepa- paired ridges curving dorsally from each zy- rate the oval head from the thinner ridge for gapophysis. These ridges continue dorsally

) the passage of the segmental artery ) ; ( 5 de- for some distance. Between these ridges, the velopment of a ridge connecting the prezyga- base of the spine shows signs of lateral exca- pophysis with the dorsal margin of the dia- vation, but the fossa thus fonncd does not pophysis. Similar changes are seen in Araeo- liecome really conspicuous until the seventh sceJis, but the shift is not as gradual as in vertebra. The neural spine of the third cervi-

PetrolacosauTus and occurs between vertebrae cal vertebra is low in lateral aspect. In dorsal

7 and 9. view, the spine is unusually wide and bears In response to the greatly elongated cen- two deep longitudinal grooves along the dor- tra, the zygapophyses in the cervical region sal surface. The neural spines change in are also unusually elongate and well-devel- shape and gradually increase in height pos- oped. Both pre- and postzygapophyses are teriorly along the column. The spine on strengthened by longitudinal ridges. In addi- vertebra 5 is roughly rectangular in lateral tion, the prezygapophyses are supported by aspect and bears a distinct mammillary proc- buttresses directed dorsally and anteriorly. ess on each side. Posterior to the fifth verte- The postzygapophyses are braced above by bra the neural spines have a pair of mammil- paired supports descending and diverging lary processes; these persist until vertebra 11. posteriorly. These supports, when viewed The shape of these processes and their posi- from the rear, have a convex margin laterally tion on the neural spine changes along the giving the neural arch its slightly swollen colunm: in vertebra 5, the pair of mammil- appearance. Posteriorly from the third cervi- lary processes occurs anteriorly, in vertebra cal vertebra, the neural arches show a gradual 6 it lies about midway along the spine and increase in this convexity, which reaches its by vertebra S it is near the posterior end. maximum on the posterior dorsal region. This The mammillary processes, which are also convexity represents additional bone above seen in Araeoscelis and Prohicerta probably the articular surface to withstand the pressure provided attachment for strong intervertebral acting on it. muscles or ligaments (Vaughn, 1955; Cow, The prezygapophyses of the cervical ver- 1975). tebrae extend far beyond the anterior end Intercentra in the cervical region are rela- of the centra. The postzygapophyses extend tively large, well-developed elements; they only as far as the posterior end of the centra. are up to 1.5 mm long in KUVP 9951 and This is in contrast to the situation in the 33607. The exposed ventral surface bears a dorsal region where both zygapophyses ex- well-developed longitudinal ridge; there are tend slightly beyond the ends of the centra. Both zygapopliyses extend far laterally, no capitular facets at the lateral extremities well beyond the lateral limit of the centrum. of the intercentra. Each intercentrum extends The articular surfaces of the zygapophyses far dorsally between the centra. —

1980 PENNSYLVANIAN DIAPSID REPTILE 33

Fig. 15. Petrolacosaunis kansensis Lane. Dorsal, sacral and caudal vertebrae with some associated ribs, all X 1.5. (A-C) Lateral view of vertebrae 7, 8, 13 and 14, KUVP 9951. (D) Lateral view of vertebrae 17 and 18, KUVP 9951. (E) Lateral view of vertebrae 25 and 26, KUVP 33606b. (F) Dorsal and lateral views of first sacral neural arch and ribs, KUVP 33605. (G) Partially preser\'ed second sacral vertebra and rib in anterior view, KUVP 33606b. ( H-I ) Lateral and anterior views of anterior caudal vertebrae and associated ribs, KUVP 33606c.

Dorsal Vertebrae ratio is 2.5:1 (vertebra 3). The absolute iieight of the centra increases slightly in the Dorsal vertebrae (Fig. 15) are consider- dorsal region, and there is bevelling of the ably shorter than the cervical vertebrae. The ventral surface anteriorly and especially pos- centra in KUVP 9951 decrease gradually in teriorly to accommodate the intercentra. In- length from 9 mm in anterior dorsal verte- terccntral gaps are, therefore, small. Between brae, to 8 mm in typical mid-dorsals and the anterior and posterior ends of the centrum down to 7 mm in posterior dorsals. The the lateral surface is strongly concave centra in the dorsal vertebrae are stoutly when built, with a length to posterior height ratio viewed in cross-section. The keel, seen in of 1.3:1 in vertebra 14, in contrast to the cervical vertebrae, is retained in dorsal verte-

condition in the cervical region where the brae, but the ventral border of the keel is 34 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7 broadly rounded as the longitudinal ridge are tilted medially at about 10 degrees to the does not descend far vcntrally. horizontal. The zygapophyseal surfaces are The neural arches of the dorsal vertebrae smaller than in the cervical region. contribute much less to the anterior centro- The neural arches of all the dorsal verte- sphcne than in cervical vertebrae. The trans- brae are excavated by deep lateral fossae. verse processes in the dorsal region are not The fossae are found by the prezygapophyscal only (juite different from those in the neck buttress that forms the anterior margin of the but also undergo considerable change. The neural spine and by the ridge that extends transverse processes in the anterior dorsal anteroventrally from the postzygapophysis. region have undergone a number of changes The fossae are most pronounced on mid-dor- from the pattern seen in the posterior cervical sal and posterior dorsal vertebrae. The neural vertebrae: (1) the transverse processes have arches arc slightly swollen above the post- migrated anteriorly onto the prezygapoph- zygapophyses, but only in the dorsal verte- yses; (2) the lateral elongation of the trans- brae. although verse process has decreased; (3) The gradual increase in neural spine configuration the articular surface changes in height, noted in the cervical region, continues of to correspond to the cross-sectional outline throughout the anterior dorsal region until there is no gap the head on the dorsal ribs, vertebra 18. The spines on vertebra 7 to 11 as between two oval areas of articulation bear short dorsolaterally directed mammillary there is in Araeoscelis, but a single elongate processes; these are located on the posterior articulation; the transverse surface of (4) half of the spine, close to the tip. The mam- process, because of its position on the pre- millary processes disappear completely by is the same ridges zygapophysis, braced by vertebra 12. In a number of articulated pos- however, a as the latter; (5) in addition, terior dorsal vertebrae (KUVP 9951, 33605 well-developed ridge runs posteromedially and 33606 ) , the spines vary in size; every sec- the top of the transverse process to the from ond spine is either somewhat wider antero- body of the arch. posteriorly (KUVP 9951 and 33606) or taller Caudad from vertebra 8 the transverse (KUVP 33605) than the intervening one. The their processes gradually decrease both in function of such alternations in the size of projection the surface of the lateral from the neural spines is not understood. Similar articular sur- arch and in the length of the ]>atterns of neural .spine size alternation have articular surface is face. By vertebra 20, the been observed by the author in Captorhinus, quite short but still retains both "heads" on Eocaptorhimts, Araeoscelis and Lahidosaurus. the neural arch. Posterior to this vertebra, the Well-de\elopcd intercentra are present articular surface is located further antcro- throughout the dorsal region of the column. ventrally and continues to decrease in size They are about 1.5 mm long in KUVP 9951. so that by vertebra 23 only the diapophyseal

area of attachment is present; it lies close to Sacral Vertebrae the centrum near the anterior centrosphene. Presumably the rest of the articular surface Sacral \ertebrae are preserved in KUVT

for the rib is now part of the centrosphene. 9951, 33605 and 3.3606. The first sacral verte-

In vertebra 26, the tiny articular surface sits bra is distinguishable from the posterior dor- on a small knoblike projection at the antero- sal vertebrae by the following features: the vcntral corner of the neural arch as in Araeo- two prezygapophyses extend as far laterally scelis. as on the dorsal vertebrae, but the postzyga- The pre- and postzygapophy.ses of all pophyscs extend only about half as far from dorsal vertebrae extend laterally slightly be- the midline. The extremely massive dia- yond the centra. The zygapophyseal planes pojihyses extend anterolaterally from high on 1980 PENNSYLVANIAN DIAPSID REPTILE 35

the neural arch. The neural spine is inclined ventral edge is not as broadly rounded. The posteriorly and has a maniniillary-like process lateral excavations above the ridge are not as (KUVP 33605) on each side at its postero- deep as in the dorsal vertebrae. The zyga-

dorsal tip. Anterodorsally, the spine bears two pophyses are lightly built with considerably longitudinal grooves along the crest, marking smaller zygapophyseal surfaces than in the the probable site of the strong interspinous dorsals. The articular surfaces are tilted at ligament that runs to the last dorsal vertebra. about 25 degrees to the horizontal plane. lateral of The centrum is slightly smaller than those The surface the neural arch of of the posterior dorsal vertebrae. Ventral to the anterior caudal vertebrae is less deeply the diapophysis, but separated from it by a excavated than in the dorsal vertebrae. No swelling is arch, deep groove, is the small longitudinal para- apparent on the above the pophysis. It lies on the lateral surface of the postzygapophyses. The spines are also much centrum at the anterior end. The intercen- smaller and alternation of width is not evi- trum does not fuse with the centrum as re- dent. The transverse process, on the other ported in Araeoscelis (Vaughn, 1955) and a hand, is more massive than on the dorsal number of other Paleozoic reptiles (Romer vertebrae and only slightly less well-devel- and Price, 1940). oped than in the second sacral vertebra. It is difficult to establish whether the diapophysis Less information is available on the sec- is distinct from the parapophysis. The in- ond sacral vertebra. The pre- and postzyga- completely exposed articular surfaces suggest pophyscs conform to the width set by the that only one, large, roughly oval surface pro- postzygapophyses of the first sacral vertebra. vides the area of attachment for the rib. The anterodorsal edge of the neural spine is The most anterior caudal intercentra are also grooved on the second sacral vertebra. similar to those of the sacral and presacral This suggests that the first and second sacral vertebrae anterior to them. Posterior to the vertebrae were also connected by well-devel- second caudal vertebra the centra gradually oped interspinous ligaments. The neural spine decrease both in width and height, but the on the second sacral vertebra is considerably length remains relatively constant well into smaller than the one on the first. The dia- the mid-caudal region. Centra in this region pophysis is also smaller, although it is still of the appear relatively long considerably larger than the processes on the and laterally compressed. At the fourteenth mid-dorsal vertebrae. The distinct diapoph- caudal vertebra, the length to height ratio is ysis and parapophysis form the articulation 1.8 to 1; by the 39th caudal vertebra, the with the rib. No infonnation is available on ratio is 2.2 to 1. Both the ventral longitudinal the second sacral intercentrum or centrum. ridge and the lateral excavation of the centra diminish gradually; by the sixteenth caudal Caudal Vertebrae the ventral border of the centrum is gently The anterior caudal vertebrae in Petrola- rounded. The neural arches also diminish cosaurus are somewhat similar to the second gradually in size. The diapophyses disappear sacral vertebra and are considerably altered by the thirteenth caudal vertebra. The neural from the pattern seen in the dorsal series. spines diminish both in height and lateral They are roughly 25 per cent smaller in all extent, disappearing at about the twenty- dimensions than the mid-dorsals and conse- fourth caudal vertebra. The spines of the quently are less stoutly built. The width of anterior and mid-caudal vertebrae follow the

the centmm is slightly greater than the height general tendency of other primitive reptiles

and is equal to the length. The longitudinal to a slight backward inclination. This postero- ventral ridge on the centrum does not extend dorsal inclination increases as the spines de-

as far ventrally as in the mid-dorsals and its crease both in height and lateral extent pos- —

36 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

teriorly along the series. Zygapophyses persist tally to enclose a vertically elongate foramen.

beyond the fortieth caudal as simple anterior The ventral projection is "Y"-shaped in pos- and posterior projections above the spool-like terior view. The first chevron appears to be

centra. They are, however, definitely lost by small and delicately built. The second is much the fiftieth caudal. larger, with a long ventral projection beyond Associated with the five anteriormost cau- the foramen. As expected, the chevrons de- dal vertebrae but not preserved in position, crease in length posteriorly, albeit gradually. are two chevron bones. The first chevron, in KUVP 336()6c, probably fits between the third Ribs and fourth caudals. As seen in several speci- .\s in all known primitive reptiles, ribs are mens, they continue posteriorly at least as far present on every vertebra from the atlas to, as the fortieth caudal vertebra. They are apparently outgrowths of the intercentra and and including, the proximal caudal vertebrae. consist of a crescentic intercentral portion, The head of most ribs includes capitular and and two thin ventral arms that converge dis- tubercular portions, but these vary consider-

FiG. 16. Pctrolacosaunts kanxctisis Lane. Ribs, all X 1-5. (A-B) Atlanta! and axial ribs in lateral view, KUVP 9951 and 33604. (C) Ventrolateral view of fifth rib, KUVP 9956. (D) Medial view of sixUi (cerv- ical) rib, KUVP 9956. (E) Medial view of .seventh (first dorsal) rib, KUVP 9956. (F) Anterodorsal view of eighth rib, slightly enished, KUVP 9956. (G) Anterior view of rib 15 slightly crushed, KUVP 9951. (H) Proximal ends of anterior dorsal ribs 9-11, KUVP 9956. (I) Posterior dorsal ribs 23 and 25, KUVP 33606b. first first sacral ( J ) Ventral view of sacral rib and lateral view of vertebra, KUVP 33606b. 1980 PENNSYLVANIAN DIAPSID REPTILE 37 ably along the column both in the degree of In a typical dorsal rib (Fig. 16), the shaft their development and in their relation to one extends dorsolaterally from the capitulum, another. In the anterior cervical, posterior whose articular surface is oval in outline and dorsal and sacral regions, the ribs have dis- occupies the head of the rib. The tuberculum tinct gaps between the tubercular and capitu- does not project markedly from the shaft. Its lar surfaces of articulation for the passage of articular surface is also oval in outline and the segmental artery. Along the rest of the broader distally. The two heads are con- presacral column the rib heads have a single nected by a narrow web of bone whose edge continuous hourglass-shaped area of articula- apposes the corresponding articular surface tion, with rounded dorsal and ventral margins on the neural arch. Distally from the tuber- and a thinner spanning portion. The dorsal culum, the rib curves strongly laterally and and ventral ends of the area of ai+iculation ventrally. This curvature indicates that the probably correspond to the capitulum and trunk was relatively high and narrow. tuberculum of dichocephalous ribs. None of Except for the rather flattened promixal end, the ribs is immovably fused to the body of the typical dorsal ribs are oval in section. A small vertebrae. ridge is usually present along the postero- In the cervical region (Fig. 16) the shafts dorsal margin. Proceeding posteriorly along of the ribs typically run without cur\'ing dis- the dorsal series, the posterior dorsal ribs as tally from the expanded heads. The atlantal far as rib twenty-one appear to be essentially rib (KUVP 3.3607) is the smallest of the cervi- similar to more anterior dorsal ribs but show cal ribs. The tuberculum and capitulum are a gradual decrease in length and in the size joined by a thin connecting web. The tu- of the head. From the twenty-second rib berculum articulated with the atlantal neural back, this decrease becomes very pronounced. arch, while the capitulum reached vcntrally Rib twenty-three is about half as long as a to the atlantal intercentrum. Distally from typical dorsal rib. The shaft is slender and the expanded head, the rib narrows rapidly almost straight. The tubercular and capitular to a point. The second (axial) and third heads are separated by a small depression for cervical ribs are each more than twice the the passage of the segmental artery. The tu- size of the first. Although they retain the berculum of this rib still attaches to the neural same structural pattern as the first rib, they arch. The capitulum, however, attaches to a do not taper to a point distally. Unlike that small facet on the centrum, which is confluent of the first three ccrvicals, the distal end of with the anterior centrosphene. Rib twenty- the fourth, fifth and sixth ribs is flattened and five is still shorter and more slender. expanded to a paddle-like blade, probably to The first sacral rib preserved in KUVP provide additional support for the levator and 9951, 99.56, 33605 and 33606b consists of a serratus musculature that runs to the scapula. short, massive, cylindrical proximal portion

The seventh cervical rib is almost twice as that expands laterally to reach out towards long as the sixth, but unlike that of the dorsal the inner surface of the ilium. Its distal mar- ribs, its shaft is nearly straight. gin is applied to the inner surface of that

The fourth and fifth cervical ribs have a bone but not fused to it. The first sacral rib small anterolaterally directed process close to articulates with its vertebra, in a large round- the anteroventral end of the head. The sixth ed tubercular articulation that begins on the and seventh ribs have a thin dorsally directed neural arch and passes ventrally onto the flange, running along the external edge of the centrum. The capitulum is separated from shaft. On the sixth this flange runs along the tuberculum by a small groove and, as in most of the length of the shaft, whereas on Araeoscelis (Vaughn, 1955) extends antero- the seventh it is restricted to the proximal ventrally to attach to the intercentrum. Be- half. yond the articular area, the neck of the rib 38 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

narrows slightly, then aljriiptly flares latoralTy .specimen, KUVP 1427 (Peabody, 1952, Fig. as a cupped expansion that widens below, and 2). The study of much better material, be-

is boundt^d anteriorly and posteriorly by two longing to mature individuals in KUVP 9957,

heavy ridges. Between these ridges the rib is 33606 and 33607 indicates that Peabody's left

strongly concave with its lateral surface al- scapulocoracoid is in fact an incompletely most vertical. The rib terminates laterally in preserved scapula with no trace of either a broad, roughly "U"-shapcd strongly curved anterior or posterior coracoids. In addition, border, which articulates with a correspond- the elements tentatively identified as the ing groove on the ilium. clcithrum and clavicle are quite different from

The second sacral rib is much more slen- the well preser\'ed, complete cleithrum in

der than the first. The smaller tuberculum KUVP 33607 and in KUVP 9958 and and capitulum articulate with the neural arch 33606. They may, in fact, be interpreted as and centrum respectively. The shaft of this is(}lated lower jaw elements since a splenial

rib extends as far laterally as the first, but its lies nearby.

distal expansion against the ilium is less pro- Cleithrum.—The cleithrum in KUVP 33607

nounced. Its distal end, as in Araeoscelis (Figs. 12 and 13) is a long, narrow bone (Vaughn, 1955), has a dual function: bracing with a slightly expanded head, quite similar

the posterior end of the first rib from below in general proportions to that found in Penn- and articulating with the ilium. sylvanian captorhinomorphs (Carroll and The caudal ribs are markedly different in Baird, 1972). The head consists of an oval character from those preceding them. They plate that was probably applied to the an- extend laterally, curve slightly forward, then terior end of the cartilaginous extrascapula.

turn sharply posteriorly in a horizontal plane. Its outer surface is only slightly sculptured.

The proximal part of each rib in KUVP The slender shaft is flattened. The medial

33606c has a longitudinal groove on the ven- surface of tlie bone is strongly grooved to re- tral surface, reflecting the probable dicho- ceive the anterior margin of the scapula. The cephalous character of the head, yet examina- lower end appears to ha\e o\erlapped the tion of the proximal surface of the rib reveals clavicle. The M. levator scapulae inferioris no distinct tubercular and capitular surfaces presumably attached to the well developed for attachment to the vertebra. A single, oval ridge that runs along the anterior margin of surface serves this function. The shaft of the the cleithrum (Holmes, 1977).

rib is circular in section. Posteriorly the rib Clavicle.—The clavicle in KUVP 33606, tapers to a pointed tip. The first and second found in close association with the .skull and

caudal ribs are unusual in having a laterally other postcranial elements, is exposed in me- directed portion of the rib. This flange bears dial view. A second, isolated clavicle belong-

longitudinal ridges, probably for muscle at- ing to a juvenile indi\'idual is exposed in tachment. lateral ^icw (KUVP 9958, Fig. 17c). The The anterior caudal ribs are elongate. The clavicle consists of a narrow dorsal process first of the series reaches far beyond the pos- and an expanded plate. The outer .surface of terior end of the corresponding vertebra. Suc- the dorsal process shows two groo\'ed areas. ceeding ribs run posteriorly parallel to, and Near the upper end, the anterior margin of

medial to the first. They diminish rapidly in the clavicle is longitudinally striated to re- length and disappear by the eleventh or ceive the overlapping lower end of the cleith- twelfth caudal vertebra. rum. The whole of the posterior margin of

the shaft is strongly grooved to receive the Pectoral Giixdle anterior edge of the scapula. At the end of

The first description of the shoulder girdle the shaft the external lip of the groove broad- of Petrolacosaurus was based on an immature ens posteriorly and conceals the groove from —

1980 PENNSYLVANIAN DIAPSID REPTILE 39

Fig. 17. Petrolacosatirus kansensis Lane. Pectoral girdle. (A) Lateral view of left scapulacoracoid in ar- ticulation with limb, KUVP 9957a, X 0.75. (B) Ventral view of .slightly immature interclavicle, KUVP 9959b, X 1.5. (C) Lateral view of left clavicle, immature, KUVP 9958, X 1.5- (D) Lateral and medial views of right scapulacoracoid, KUVP 33609, X 2. See Figs. 12 and 13 for cleithrum. lateral view. The narrow dorsal process of medial margin. The ventromedial margin is the clavicle probably provided excellent at- gently convex. It appears that the clavicles tachment for the clavicular portion of the covered the anterior part of the head of the deltoid and trapezius muscles. The plate-like interclavicle head but did not meet at the portion of the clavicle is relatively massive in midline. comparison to that in primitive captorhino- Interclavicle.—The isolated, immature, in- morphs, but is comparable to the clavicle of terclavicle found in KUVP 9959b has a dia-

Araeoscelis (Vaughn, 1955). This plate is in- mond-shaped head and a long stem (Fig. flected medially on the shaft at an angle of 17b). The exposed ventral surface of the about 45 degrees. The expansion of this por- head is strongly convex. The greater part of tion of the clavicle is formed largely by a the head is depressed, bears no sculpturing, strong anterior curvature of the thickened and lies internal to the clavicles. This iso- anterior margin. The lower portion of the lated element has been associated with Petro- clavicle has striations running to the ventro- lacosaurtis partly because the depressed areas 40 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7 on the head correspond closely in outline to As in Paleothyris and Protorothtjris the supra- the medial end of the clavicular blade. In glcnoid foramen opens externally on this addition, the rai.sed portion of the head, which second ridge, just above the tip of the supra- forms the posterior quarter of the diamond glcnoid buttress. and extends anteriorly between the clavicles, The glenoid cavity has the usual screw- bears the distinctive type of sculpturing seen shape seen in most early tctrapods (Romer, on the clavicle and the dermal skull roof 1956; Holmes, 1977). The anterior part of elements. The stem of the interclavicle is a the glenoid projects far laterally, supported flat straight blade that runs along the ventral by the massive anterior glenoid buttress. The midline of the pectoral girdle. It lies on the external opening of the coracoid foramen lies external surface of the bony scapulacoracoid within a deep pocket, beneath this buttress. and the unossificd sternum. The ventral sur- Behind the glenoid there is a prominent proc- face of the blade is missing in KUVP 9959b, ess for the origin of the coracoid head of the except for the posterior third, where a series triceps muscle. This structure, common in of bilaterally arranged longitudinal grooves is pelycosaurs, is also found in Protorathyris and seen. The anterior and posterior tips of the Araeoscelis. The large external surface of the bone bifurcate. anterior coracoid area is smoothly confluent Scaptdocoracoid. — The scapulocoracoid with that of the scapular blade. Close to the

( Fig. 17d ) is present in four adult specimens, ventral margin, the external surface of the KUVP 9957a, 33606c, 33607 and 33609, and coracoids is unfinished. appears to be ossified as a unit with no su- The medial surface of the right scapulo- tures separating the anterior and posterior coracoid in KUVP 33609 closely resembles coracoids and scapula. The only indication that of Protorathyris (Clark and Carroll, of a division between the anterior and pos- 1973). A system of two ridges has the outline terior coracoids is a delicate crack that runs of an inverted "T." One of these ridges dorsoventrally through the glenoid cavity in extends vertically onto the scapular area me- the right scapulacoracoid of KUVP 9957a. In dial to the supraglcnoid buttress. The second immature specimens, however, the component diverges anteriorly and posteriorly from the elements show postmortem separation. base of the first and supports the anteroven- In general proportions the scapulocoracoid tral part of the anterior coracoid and the resembles those of the primitive captorhino- glenoid respectively. In contrast to the con- morph Protorathyris archeri (Clark and Car- dition in Protorathyris, where there are two roll, 1973, Fig. 6) and Araeoscelis (Vaughn, internal openings of the supraglcnoid fora-

1955) except that the- scapular blade in Petro- men, there is only a single internal opening in lacosaurus is much wider. The straight dorsal Petrolacosaurus, in the upper part of the deep margin of the scapula is unfinished, indicating subcoracoscapular fossa. The coracoid fora- that the endochondral girdle continued dor- men probably opened into the lower part of sally in cartilage. The anterior portion of the the subcoracoscapular fossa, as in other early blade is very thin transversely. The posterior reptiles, but this area of the fossa is slightly margin of the; scapula curves posterodorsally damaged in KU\'P 33609. as in Araeoscelis. Ventrally this margin is Pelvic Girdle continued as two ridges. One forms the pos- terior boundary of the supraglcnoid buttress Both Lane (1945) and Peabody (1952) and extends posteriorly onto the posterior based their description of the pelvic girdle on coracoid. A second, more prominent ridge the isolated left half of a pelvis (KUM' 1425)

extends straight down to the massive anterior tliat is definitely tliat of an edaphosaurian

glenoid buttress and forms the anterior boun- pelycosaur. There is now conclusive evidence dary of the triangular supraglcnoid buttress. that this element does not belong to Petrola- —

1980 PENNSYLVANIAN DIAPSID REPTILE 41

Fig. 18. Pctrolacosatirus kansensis Lane. Pelvic girdle, X 1.5. (A) Lateral \iew of slightly crushed left pelvis, KUVP 9961. (B) Medial view of right iliac blade, uncrushed, KUVP 33606c. (C) Medial view of shghtly crushed left pelvis, KUVP 9951. cosaurus. In KUVP 33606 part of the left which has also been found in the Gamett side of a mature pelvis, exposed in medial quarry. view, is closely associated with a series of In its general proportions and in the struc- posterior dorsal, sacral and anterior caudal ture of the component elements, the pelvic vertebrae, on a small block which also con- girdle (Fig. 18) resembles that of Araeoscelis more closely than the girdle of any other rep- tains other remains of Petrolacosaurus. In tile, but also retains a number of primitive KUVP 9951c the complete pelvis, exposed in features seen in primitive captorhinomoriDhs. lateral and medial views, is closely associated In contrast to the fectoral girdle, which in with both hindlimbs and lies near a series of Petrolacosaurus is lightly built and is ossified articulated presacral vertebrae and a scat- as a unit in mature individuals, the pelvic tered skull of the same genus. These pelves girdle is massive and the sutures between the are unique in several features and can be component elements are always easily dis- readily distinguished from those of other cernible in medial view. The acetabular cav-

Paleozoic reptiles. The pelvis mistakenly as- ity is primitive in its configuration, with the cribed by Lane and Peabody to Petrolaco- three raised comers of the cavity supported saurus probably pertains to Edaphosaurus sp. by prominent buttresses on each of the three 42 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7 pelvic elements. In both KUVP 9951 and probably served as the origin for the ambiens 9961 the unfinished articular surface of the and pubotibialis muscles. The true pubic tu- acetaijulum extends to the posterior margin of bercle, located at the distal end of the ridge is the pelvis, as in Araeoscelis (Vaughn, 1955). also massive. As in other early reptiles the Ilium.—The ilium has a long blade which pubic part of the puboischiadic plate faces extends far posterodorsally and only slightly more or less vcntrally and is not fully seen anteriorly. This condition, also seen in all in lateral view. The external opening of the captorhinomorphs, Araeoscelis and ophiaco- obturator foramen, which is much larger than

dont pelycosaurs, is considered primitive. The in primitive captorhinomorphs, lies directly posterior end of the blade in Petrolacosaurus beneath the anterior end of the acetabulum.

is thin in cross-section, and its lower corner On the internal surface of the pubis, the un- tends to have been lost during preservation. finished symphysial surface is relatively nar- The internal surface of the ilium, superbly row, running along the ventromedial margin. preserved in KUVP 33606c, is more significant The internal opening of the obturator foramen than the lateral. Near the dorsal edge a well is obscured in all the preser\ed specimens, developed shelf of bone extends medially but probably lies on the upper part of the along much of the length of the iliac blade. large pubic surface which faces not only This process, not found in captorhinomorphs, dorsomedially but also anteriorly. is well developed in Araeoscelis (Vaughn, Ischium.—The ischium in Petrolacosaurus 1955), and in ophiacodont pelycosaurs. The is the most distinctive element of the pelvis. central region of the medial iliac surface is In captorhinomorphs and pelycosaurs the long

slightly recessed and rugose below the me- dorsal margin of the ischium is strengthened dially directed shelf to receive the sacral ribs. by a rounded ridge. No such ischiadic ridge

Neither of these structural features of the in- is found here. Posterior to the acetebulum,

ternal surface of the ilium are found in capto- tlie thin dorsal margin of the ischium is rhinomorphs. The posterior half of the medial deeply notched, a feature unique to Petrola- surface of the blade shows a series of longi- co.^aurus. The posterior margin of the notch tudinal ridges, similar to those seen in the flares dorsolaterally. Posterior to the notch

primitive captorhinomorph Coelostegus (Car- the margin of tlie ischium remains thin and roll and Baird, 1972, Fig. 14) for the origin of cur\es medially to the posteromedial corner epaxial musculature. of the bone. The M. ischiotrochantericus ap- Pubis.—The pubis, well preserved in both parently passed from the posteromedial sur- KUVP 9951c and 9961 is massi\ely built. As face of the ischium laterally ventral to the in Araeoscelis, the anterolateral margin is ilioiscliiadic ligament to insert onto the proxi- formed by a thick ridge that is a continuation mal end of the femur. The dorsolateral flare of the anterior iliac ridg(\ A large tubercle of the ischiadic notch probably served as the projects dorsolaterally from this ridge. Its attachment area for a shortened and hence function, according to Vaughn (1955) was to stronger, ilio-ischiadic ligament. The notch of the iliopubic raise the anteroventral end itself likely served as a lateral passage for M. ligament in the absence of a well developed ischiotrochantericus. On the internal ischiadic anterior expansion of the iliac blade. This surface a ridge extending posteromedially tubercle is also found in Palcothyris and in from the notch, roughly parallel to the pos- primitive ophiacodonts, but in these genera terior margin, limits the area of origin of the is much smaller than in either Petrolacosaurus above nniscle. This posterior area of the or Araeoscelis. A second, smaller process ex- tends ventrolaterally from the ridge connect- ischium is slightly offset from the general ing the lateral pubic tubercle to the acetabu- interna] surface. No such specialization is lum. This process, also seen in Araeoscelis, seen in any other early reptile. —

1980 PENNSYLVANIAN DIAPSID REPTILE 43

i (111

Fic. 19. Petrohcosanrus kamensis hane. Reconstruction of tlie right humerus, X 1.5. (A-D) Proximal dor- sal, proximal ventral, distal dorsal and distal ventral surfaces, based mainly on KUVP 9957 and 33606a.

Limbs proximal plane is 60 degrees. Following Romer and Price's (1940) terminology, the A total of twenty-six limbs, belonging to humeral surfaces in Petrolacosaurus can be fourteen individuals have been recovered. divided into proximal dorsal, distal dorsal, Most of these were originally complete as proximal ventral and distal ventral surfaces. they lay in the matrix. Only eight were com- The proximal dorsal surface was probably pletely isolated from the body. Of the limbs gently convex along the long axis of the bone. associated with the girdles, only one was in Severe crushing flattened the proximal head. articulation (KUVP 9957a, Fig. 17). The surface shows a distinct ridge for the Humerus.—The reconstruction of the hu- M. latissinius dorsi. A long narrow ridge for merus (Fig. 19) is based mainly on two ma- the M. deltoideus insertion runs distally from ture but crushed bones, found in KUVP 9957a the delto-pectoral crest. and 33606a. The humerus is of the tetrahe- ventral surface probably dral type common to primitive reptiles The proximal was deeply concave below the articular surface. (Holmes, 1977) but the shaft is better de- coracobrachialis brcvis inserted in the veloped than in primitive captorhinomoiphs Tlie M. proximal area, and pelycosaurs. The width of the proximal concave region. The anterior delto-pectoral crest, is rounded and end of the humerus forms about 27 per cent above the rugosity. In lizards, the of the length of the bone, and the distal width lacks any modern M. supracoracoideus inserts on this area ( Holmes, is about 29 per cent of the length. The shaft 1977). Distally, a distinct delto-pectoral crest is exceedingly slender, only about 7 per cent anteroventrally the proximal of the length. These proportions are compara- juts out from the of this crest is a rela- ble to those found in Araeoscelis (Vaughn, head. At summit 1955). tively small circular area with a rugose sur- The estimated twist of the distal upon the face to which attached the powerful pec- — .

44 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

toralis inusculature. Distally, the crest slopes groove that widens distally. The moderately

rapidly into the shaft; its internal border sends developed ectepicondyle is a long ridge that

a delicate ridge distally to the entepicondyle. widens as it lains distally. Its end projects A well-developed entepicondylar foramen distally above and beyond the general con- and a large snpinator process distinguish the tours of the humerus. Its rugose distal head distal head of the humerus of Petrolacosaurus furnished a small area for the origin of much from that of most pelycosaurs (Romer and of the extensor musculatine of the lower arm

Price, 1940) as well as from Araeoscelis. The and foot. Anteriorly the ectepicondyle is radial nerve was only partially surrounded by bounded by the long ectepicondylar groove.

bone as it traversed the humerus, whereas In pelycosaurs the entepicondyle is usually

in Araeoscelis it was fully enclosed in bone larger and the ectepicondyle projects further

( ectepicondylar foramen ) dorsally than in Pctrolacosaiinis (Romer and The entepicondyle does not extend far Price, 1940). laterally. Its rugose lateral and distal margin Anterior to the deep ectepicondylar furnished the areas of origin of the flexor groove, which carried the radial nerve, the musculature of the lower arm and foot. The supinator process lies below the level of the inferior humeral nerve apparently ran along ectepicondyle and the general dorsal surface. the deep groove on the dorsal surface of the Unlike that in pelycosaurs, the supinator proc- entepicondyle, passed through the elongate ess in Petrolacosaurus does not project strong- entepicondylar foramen, and entered the fore- ly outward but extends far distally. Its rugose arm, as in S})Iien()don (Miner, 1925). On the tip lies close to the distal end of the bone

distal dorsal surface, the entepicondyle is and probably furnished a small area for the separated from the ectepicondyle by a long origin of the supinator muscles. The fully

I (III

Fic. 20. Petrolacosaurus kamcnsis Lano. Left radius (A) and ulna (B) in dorsal (prea.xial) view, KUVP 9957, X 2. 1980 PENNSYLVANIAN DIAPSID REPTILE 45 enclosed ectcpicondylar foramen in Araeo- of the element, whereas the distal end is scelis extended the possible area of origin of narrower—about 12 per cent of the length. the supinator musculature, as in modern liz- The shaft, as in the radius, narrows to 5 per ards (Romer, 1944). A ligament, running cent of the length of the bone. The ulna has from the supinator process to the ectepicon- a well developed olecranon and a concave dyle in Petrolacosaurus, may have served the sigmoid notch for articulation with the hu- same function. merus. The tip of the olecranon is capped The lateral edges of the distal ventral sur- by a ridge that curves over the apex of the face are formed by the margins of the entepi- bone. The main tendon of the triceps muscu- eondyle and the supinator process. Between lature probably attached here. The sigmoid these margins the ventral surface is relatively notch occupies the whole thickness of the flat but is pierced by the oval entepicondylar medial side of the head. The articular surface foramen. Much of the distal end of the hu- is higlily curved, turning through an arc of merus is occupied by the radial and ulnar about 140 degrees, while the distal end of the articulations. The radial articulation (capi- notch faces proximally, its proximal end faces tellum) is a slightly elongate swelling with medially and slightly distally. The anterior an unfinished surface. The capitellum not surface near the humeral articulation is ru- only projects ventrally but also slightly dis- gose. The anterior and posterior surfaces are tally, suggesting that the epipodials met the separated medially by a rounded ridge. The longitudinal humeral plane at an obtuse angle. distal end of the ulna is expanded and is

The capitellum is contiguous with the ulnar twisted slightly laterally. The autopodial sur- articulation posterior to it. The concave sur- face is convex. face of the latter faces essentially distally. Femur.—The reconstruction of the femur

Proximally from the articular area, the ventral ( Fig. 22 ) is based on the holotype KUVP surface is slightly depressed. 1424, as well as on KUVP 995ld and 33606. Radius.—The radius (Fig. 20) is a long, Dorsoventral crushing has flattened the fe- slender element with slightly expanded ends. mora and bent the internal trochanter. This In the articulated left forelimb of KUVP prevents consideration of the femur in an-

9957a (Fig. 17) the length of the radius is terior and posterior views. The femur is .92 per cent of the length of the humerus. closely comparable to that of primitive capto-

The width of the proximal end measures about rhinomorphs and pelycosaurs. It is more 14 per cent of the length of the bone and the massive than the femur in Araeoscelis distal width about 12 per cent of the length. (Vaughn, 1955).

The proximal head is cup-shaped for the The width of both the proximal and distal humeral articulation. Because of crushing, its ends of the bone is about 30 per cent of the outline cannot be determined. The lateral and length. The relatively massive shaft narrows medial edges of the bone, seen in anterior to about 13 per cent of the length. view, are partially formed by two longitudinal The proximal head of the femur is termi- ridges that probably separated the extensor nal as in all primitive reptiles and not in- and flexor muscles. The medial ridge is quite flected medially. The articular surface is prominent proximally and is slightly nigose. crescentic. The anterior portion of the artic-

The distal head of the radius is turned slightly ular area is the broadest. Distal to the head, laterally so that the autopodial surface can be a prominent anterior crest extends onto the seen in lateral view. ventral surface of the femur. The crest's Ulna.—The ulna (Fig. 20) is also quite proximal end probably forms the internal long and slender; it is slightly longer than the trochanter for the tendinous insertion of the humerus. The maximum width of the proxi- M. puboischiofemioralis externus. The distal mal end approaches 15 per cent of the length extension of the crest along the shaft forms a —

46 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

Fio. 21. Petrnlocasaurus kansrnsis Lane. Rctonstniction of the right manii.s in dorsal (A) and ventral (B) views, based mainly in KUVP 1423, 8,355 and 9957b, x 2.0. Abbreviation.s used in this FiKure: i, inter- medium; Ic, lateral centrale; mc, medial eentrale; pis, pisiform; r, radiale; ii, ulnare; 1-5, di.stal eaqials; I-V, metacarpals.

linea aspera for the insertion of the- addnctor parently for the probable insertion of the M. musculature. Tlicrc is no distinct fourth tro- ischiotrochantericus. The smooth, cur\ed, dor- chanter. The ventral surface of the proximal sal surface extends distally onto the broad head is occupied by the deeply hollowed shaft. Two prominent ridges on the dorsal intertrochanteric fossa for the M. puboischio- surface of the distal head run [Moximally from fcmioralis externus. the articular area of the anterior and posterior In dorsal view, the proximal head has a condyles respectively. The extensor musctda- smoothly curved, conical .surface broken pos- ture of the lower leg attached to these longi- teriorly by a well developed tuberosity ap- tudinal ridges. He\'ond these areas of muscle —

1980 PENNSYLVANIAN DIAPSID REPTILE 47

Fig. 22. Petrolacosaurus kansensis Lane. Reconstruction of the left femur in dorsal (A) and ventral (B) views, based mainly on KUVP 9951 and 33606, x 1.5.

attachment, the dorsal condylar surfaces are The tibia is immediately identifiable by concave. Between these ridges the femur is the conspicuous anterior concavity of the notched dorsally by a deep intercondylar shaft and by the large head. The articular fossa for the tendon of the quadriceps muscle. surface of the head is similar in outline to The surfaces for articulation with the tibia that of primitive captorhinomorphs (Carroll, and fibula extend onto the dorsal surface. 1969) and pelycosaurs (Romer and Price,

The distal ventral surface of the femur is 1940). The proximal portion of the tibia is deeply concave and rugose. A prominent roughly triangular in cross-section as is the posterior ridge extends distally along the pos- outline of the articular surface. Two well terior condyle. Its continuation onto the ar- developed ridges pass distally along the lat- ticular area forms the boundary between the eral and anterior margins. The latter prob- areas of articulation with the tibia and fibula. ably afforded attachment for an interosseous Tibia. —The tibia and fibula are known ligament. As in Araeoscelis the distal articu- in at least two specimen;;—KUVP 9951b and lar surface of the tibia can be divided into parts 33606. The tibia (Fig. 23) is quite long, its two that form a fimi, locked joint with length equalling that of the femur. As with the astragalus (Vaughn, 1955). other distal limb bones, the tibia is slender. Fibula. —The fibula (Fig. 23) is very long

The width of the head in one specimen is and slender. The width of the head in a well about 20 per cent of the length of the bone, ossified specimen is about 10 per cent of the and the distal width is nearly 12 per cent of length of the bone; the distal width about 15 the length. The shaft has a minimum diam- per cent of the length. These proportions are eter of only 5.5 per cent of its length. These also comparable to those in Araeoscelis proportions are comparable to those in Araeo- (Vaughn, 1955). The fibula in Petrolaco- scelis (Vaughn, 1955). saurus is rather blade-like, with prominent —

48 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

Fic. 23. Petrolacosaurtis kansensis Lane. Tibia and fibula, X 1-5. (A-B) Dorsal and ventral views of the left tibia in KUVP 9951. (C) Distal articular surface of (B). (D-E) Dorsal and ventral views of the left fibula in KUVP 9951.

medial and lateral ridges. The medial ridge pactly together to form a well knit structure

for the interosseous ligament is blade-like. with little cartilage between articular sur- The medial margin of the fibula, formed by faces (Fig. 21).

the above ridge, is slightly concave. The In general proportions, the carpus in Pe-

proximal portion of the filiula is somewhat frolacosaurus is similar to that in Araeoscelis

twisted in relation to the distal one. The (Vaughn, 1955). It is slightly more elongate proximal articulatory surface for the femur than the carpus in primitive captorhinomoqihs

is key-hole shaped, due to crushing. The (Carroll and Baird, 1972). This elongation is

distal surface, on the other hand, is elongate probably related to the elongation of the

and divided into distinct astragalar and cal- radius and ulna. The ulnare is the largest

caneular areas. element of the carpus, is attached proximally

Mantis.—The manus is well represented to the ulna and the small pisiform, and is by eight specimens. Nearly all features of supported distally by the fourth and fifth the manus are seen in both dorsal and ven- distal carpals. It varies in outline from a tral views. The general number and arrange- discoid in KUVP 8355 to a highly elongate ment of bones resembles closely that of cap- structure in KUVP 33606, equal in length to

torhinomorphs. The preserved carpal bones an anterior dorsal vertebra. It is possible to are cither in perfect articulation, or only account for this variation by the apparent im- slightly disarticulated. All the elements are maturity of the former specimen. The ulnare well ossified with fully developed articulating of tlie mature individual, seen in KUVP 33607

surfaces. They appear to have fitted com- is a massive element, braced by a pair of 1980 PENNSYLVANIAN DIAPSID REPTILE 49 ridges near the medial edge. The more prom- strongly tilted dorsally as in many pelycosaurs inent of the two ridges extends on the dorsal (Romer and Price, 1940). The bulbous distal surface between the proximal and distal artic- articular surface fits into the angle formed by ular surfaces. A similar, but smaller, ridge the centralia. exposed in KUVP 9957b, extends proximodis- The lateral centrale in Petrolacosaurus is a tally on the ventral surface. Both the proxi- relatively short element which has contacts mal and distal articulatory surfaces are con- with most of the other carpal bones. In addi- tiguous with a pair of unusually massive tion to supporting the three proximal carpal medially and vcntromedially directed facets elements, it also articulates with the medial that articulate with the intermedium and centrale and the third and fourth distal car- medial centrale respectively. Between these pals. Between the ventrolaterally directed facets, the medial margin of the bone is process for articulation to the ulnare and the deeply notched to form the lateral surface of proximal articulation to the intermedium, the the perforating foramen. medial centrale contributes to the margin of The intermedium in Petrolacosaurus is the perforating foramen. The dorsal surface also an elongate element, although it is short- of the lateral centrale also bears a prominent er than the ulnare. In KUVP 33607 the inter- longitudinal ridge. Whereas in primitive cap- medium is two-thirds the length of the ulnare. torhinomorphs (Carroll, 1969) and pelyco- The surface of the proximal articulation with saurs (Romer and Price, 1940) the lateral the ulna is continuous with a massive medial centrale is elongate and the medial centrale articular surface with the ulnare. This sur- is short and broad, in Petrolacosaurus the two face, exposed in both KUVP 9957b and 33607, centralia are similar in shape. faces somewhat ventrally. The small, oval The medial centrale is similar in shape to distal articulation of the intermedium is re- that in Araeoscelis (Vaughn, 1955). It sup- stricted to the lateral centrale, as in Araeo- ports proximally much of the weight of the scelis (Vaughn, 1955) and pelycosaurs (Ro- radiale and articulates distally not only with mer and Price, 1940), rather than being the first and second distal carpals but also supported by both the lateral centrale and with a portion of the third. Its ventral sur- ulnare, as in Paleothyris (Carroll, 1969). A face, exposed in KUVP 9957b bears a promi- well developed ridge extends proximodistally nent finger-like process that extends towards along the lateral edge of the intermedium. the first distal carpal. Between the articular surfaces with the lunare The fourth distal carpal, as in other early and the lateral centrale, the intermedium reptiles, is the largest of the series, and bears forms the rounded medial surface of the per- most of the support of the ulnare. Medially forating foramen. The medial edge of the it does contribute to the support of the lateral intermedium is thin and is free of contact centrale. The ventral surface of the fourth with the radiale. distal carpal is considerably larger than the The radiale, best preserved in KUVP 8,3.55 dorsal and overlaps much of the head of the is a short, massive element supported by both fourth metacarpal and a portion of the small the medial and lateral centrale. As in Atqco- triangular fifth distal carpal. The third distal scelis and numerous other early reptiles the carpal supports most of the weight of the radiale in Petrolacosaurus bears a longitudi- lateral centrale and part of the medial cen-

nal ridge on its dorsal surface. On the ventral trale. It is only slightly smaller than the surface, this element lias a pronounced proxi- third, at least in dorsal view. In KUVP 9957b mal depression, which is interrupted by a the third distal carpal is exposed in both dor- proximally directed ridge. The radial surface sal and ventral views. The ventral surface of

is nearly transverse to the long axis, as in this clement is considerably smaller than its Araeoscelis (Vaughn, 1955) rather than being dorsal surface because the large surface of 50 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

articulation with tlii' fourth distal carpal faces saurus has not been found in any other Paleo- ventrolatcrally. In addition to this unusual zoic reptile. This may, howe\er. be due to feature, the third distal carpal also has a the general lack of information about the prominent finger-like process that extends dis- \entral aspect of the carpus in early reptiles.

tally beneath the head of the third meta- A definite joint plane is lacking within the carpal, well beyond the carpal-mctacarpal carpus, and even between the carpus and the articulation. The finger-like processes on the paired distal limb elements. Movement be- ventral surface of the third distal carpal and tween the carpus and the metacarpals appar- the medial centrale and the ridge on the ently was also slight because the proximal radiale may have provided attachment areas heads of the metatarsals overlap one another, for the flexor musculature of the forelimb. the line between the carpals and metacarpals

The first and second distal carpals support is curved, not straight, and the distal ends the medial centrale and are comparable in of the third and fourth distal carpals restricted

size to the fifth. the downward movement of the correspond- A complete set of metacarpals occurs in ing metacarpals considerably. It is probable, perfect articulation with the rest of the imma- therefore, that bending during locomotion

ture forelimb in KUVP 1423 ( Peabody, 1952, was achie\ed by discreet accommodation Fig. 6c). These metacarpals show a range in throughout the manus. length from 4 mm (first) to 14.5 mm Pes. —The pes is almost completely articu- (fourth). The proximal heads overlap one lated in the type specimen KUVP 1424 (Pea- another, as in Araeoscelis (Vaughn, 1955) body, 1952, Fig. 9), but most of the bone is and many diapsids (Carroll, 1975), restricting gone, leaving an impression on the matrix. considerably independent metacarpal move- The interpretation of the tarsus given here

ment. Such metacarpal overlap is not found ( Fig. 24 ) is based on well preser\ed left and in captorhinids (Holmes, 1977). As indicated right tarsi, exposed in both dorsal and ventral by Peabody (1952, pp. 28 and 29) both meta- views (KUVP 9951a and b) and a partial carpals and phalanges are relatively long and right tarsus (KUVP 33606) from mature indi- slender, as in other small early reptiles. The viduals. Peabody suggested that the first tar-

phalangeal count is 2, 3, 4, 5, 3. The terminal sal bone was lost in Petrolacosaurus. This

phalanges are developed into long, sharp element is, however, present in both the left claws, similar to those found in primitive cap- and right tarsi on KUVP 9951a and b. The torhinomorphs and early pelycosaurs. general arrangement of the tarsal elements One of the most striking features of the resembles that of Araeoscelis (\'aughn, 1955) manus in Petrolacosaurus relates to the orien- and to a lesser extent that of the primitive tation of the articular surfaces between the captorhinomorph Paleothyris (Carroll, 1969). ulnare and intermedium, between the ulnare The shape and level of ossification of the and lateral centrale, and between the third component elements indicates that the pes and fourth distals. The ventral edges of these was a compact structure, as was the mature

surfaces are widely separated if the carpus manus.

is reconstructed as a comiiletely flattened The astragalus is the most distinctive ele- structure (Fig. 21). This indicates that in ment in the tarsus. It is a long, relatively

life Petrolacosaurus had a strongly curved slender bone. As in other primitive reptiles carpus, with a medio-laterally convex dorsal it is essentially "L"-shaped, with a distinct surface. Most of this cur\'ature was formed neck region, the proximal end of which forms along the long axis of the carpus. The extent part of the tarso-fibular articulation. The of this curvature cannot be established with dorsal surface of the astragalus, exposed in certainty, ho\\(\er, because of crushing. The KUVP 9951b, is gently concave. The lateral type of cariiai cnr\ature seen in Petrolaco- border of the astragalus forms a massive 1980 PENNSYLVANIAN DIAPSID REPTILE 51 articulation with the calcaneiim along the by the small notch that forms the lateral length of the bone, interrupted only near the margin of the perforating foramen. The distal end by a small notch that forms the broad, rounded distal border of the calca- medial border of the perforating foramen. neum apposes the fourth and fifth distal tar- This longitudinal articular surface faces dor- sals. An unfinished area, marking the area of sally, in part, in KUVP 9951b suggesting that articulation with the distal tarsals, extends the pes, like the manus, is curved transversely far onto the ventral surface of the calcaneum. but with a concave dorsal surface. The latcro- Except for this extended area of articulation distal corner of the astragalus bears a condy- there is nothing to distinguish this calcaneum loid process, similar to that in Araeoscelis from those of pelycosaurs. (Vaughn, 1955), which articulates with the In contrast to the condition in primitive calcaneum laterally and to the fourth tarsal pelycosaurs, where there are both lateral and distally. The distal border of the astragalus medial centralia in the pes (Romer and Price, is notched between the condyloid corner and 1940), there is only a single large centrale in the articulation with the centrale. In con- Petrolacosaurus and most captorhinomorphs. trast to the condition in captorhinomorphs and It supported the astragalus. This element, pelycosaurus, where the tibial surface on the completely preserved in KUVP 9951a and b, astragalus is convex, that of Petrolacosaurus is much compressed proximodistally. It artic-

(as it is exposed in medial view in KUVP ulates with the rounded distomedial surface

33606) is strongly concave, with two massive of the astragalus. The proximal articular sur- ridges, separated by a trough. One of these face of the centrale is deeply concave, where- ridges forms the thickened distal two-thirds as the distal surface for articulation with the of the medial border. The second prominent first three distal tarsals and the lateral surface ridge located on the ventral surface runs for articulation with the fourth distal tarsal roughly parallel to the lateral border. This are strongly convex. Both dorsal and ventral arrangement of a pair of ridges separated by surfaces of the centrale show a central de- a deep trough is also seen in Araeoscelis and pression bound proximally by the ridged mar- in the eosuchian Kenyasaitrus (Harris and gin of the bone. Carroll, 1977). The distal articular surface of All the distal tarsals are well preserved in the tibia is divided into a lateral trough and KUVP 9951a and b and only slightly disartic- a medial ridge, which fits snugly onto the ulated from one another. As in other early astragalus to fomi a finn, locked joint. On reptiles, the fourth distal tarsal is the largest the \entral surface of the astragalus a strong of the series and bears much of the weight ridge connects the distal lip of a deep fossa, of both the astragalus and calcaneum. The which extends to the perforating foramen, dorsal and ventral surfaces are gently concave. with the proximal end of the tibial surface. The dorsal surface of this bone is contiguous A notch in the distal border, exposed in dor- distally with the broadly convex articular sur- sal view, extends onto the ventral surface as face for the fourth metatarsal. All other artic- a small fossa which recei\es the latero-proxi- ular surfaces of the fourth tarsal are sharply mal end of the centrale. offset from both the dorsal and ventral sur- The calcaneum in Petrolacosaurus is a face's. Medially the fourth distal tarsal artic- relatively simple oval stiiicture with a thick ulates with the centrale and the third distal medial portion that supports the proximal tarsal by a complicated system of concave fibular facet and the distal tarsal articulation. and convex surfaces (Fig. 24). The proximal

A somewhat thinner plate extends laterally. articular surface is strongly concave. Later- Both dorsal and ventral surfaces are gently ally the fourth distal tarsal articulates with concave. The straight medial margin, which the fifth distal by an extensive dorsolaterally articulates with the astragalus, is interrupted inclined surface. The first distal tarsal is also —

52 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

deeply concave pioximally in Petrohicosaurus second distal tarsals, like the third, have and matches the length of the fourth. The larger lateral and medial surfaces than cither lateral surface of the fifth distal tarsal is the dorsally or ventrally exposed areas. smoothly continuous with the narrow distal The articular surfaces between the proxi- surface, and is unfinished. This indicates that mal and distal elements of the tarsus extend both these surfaces articulated with the large in a straight line roughly peipendicular to the head of the fifth metatarsal. The fifth meta- long axis of the pes; in addition, the proximal tarsal was, therefore, widely divergent. surface, formed by the astragalus and calca- The third distal tarsal not only has rela- neum has a rounded, strongly convex outline tively small, simple, dorsal and ventral sur- and fits into the strongly concave distal artic- faces, hut also has extensive and complicated ular surface fomied by the centrale and the articular surfaces with the centrale and the fourth and fifth distal tarsals. The configura- second and fourth distal tarsals. The first and tion of the mesotarsal articulation indicates,

Fit;. 21 PetTolacosaunts kan.wmis Lane. Pes, all X l-S- (A) ReconstmcHon of the left pes, in dorsal view, based mainly on KUVP 1424 and 9951b. (B) Ventral view of the left tarsus, restored with the mesotarsal surfaces exposed, KUVP 9951a. (C) Medial view of the right astragalus, KUVP 33606. (D) Metatarsals in dorsal view, KUVP 99511). .abbreviations used in this Figure: as, a.stragalus; caj, caleaneum; cp, centralia pedis; fib, artieulation with fibula; tib, articulation witli tibia; 1-5, distal tarsals; I-V, metatarsals. 1980 PENNSYLVANIAN DIAPSID REPTILE 53 therefore, that hinge movement between the heads of both of these metatarsals have

proximal and distal elements is probable, strongly ridged medial portions and slender while the distal elements are solidly attached lateral portions which also extend proximally. to one another with little possible movement The proximal articular surfaces are not per- between them. Similarly, in primitive capto- pendicular to the long axis of the metatarsal. rhinomorphs and pelycosaurs, a functional The proximal ends engage each other in mesotarsal joint may have been present, but characteristic overlap. This overlap is, how- the corresponding articular surfaces are not ever, more accentuated in the pes than in the as well adapted to movement as in Petrolaco- manus. saurits. The phalangeal count, based upon the The metatarsals in general resemble the type specimen is 2, 3, 4, 5, 4. The terminal metacarpals, except that the fourth and fifth phalanges are modified into claws similar to metatarsals ha\e large proximal ends. The those in Paleothijris. 54 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO.

PHYLOGENETIC RELATIONSHIPS OF PETROLACQSAURUS

In the first detailed study of Petrolaco- Reisz, 1972; N'aughn, 1955) pro\ide a strong saurus Peabody (1952) tried to establish the basis for the critical evaluation of the phylo- probable relationships of this by eoni- genetic relationships of this unique early rep- paring it with reptiliomoiph "amphibians" tile. (diadectids and seymouriamorphs), capto- For the purposes of this discussion strict rhinomorph reptiles, pelycosaurian reptiles for the .\mmiota is assumed. It and diapsid reptiles. He separated those is further assumed that the following grovips morphological features of Petrolacosaiirus of late Paleozoic and early fossils that are also found in many amphibians can be included within the .\mmiota: pareia- (.shared primitive characters, or symplcsio- saurs, procolophonians, millerosanrs, meso- morphics of Hennig) from those he believed saurs, pelycosaurs, captorhinomoq^hs and were "progressive features" present in capto- "cosuchians." In order to test rigorously hy- rhinomorphs or pelycosaurs or cosuchians potheses of relationships of the Petrolaco- (shared derived characters, or synapomor- sauridae some well-tested equivalent level phies of Hennig). His conclusion that Petro- phylogenetic system is needed within the lacosaiirus is an eosuchian diapsid reptile Ammiota, but this is not available. No re- rather than a pelycosaur and that it is close stricted outgroups can be developed either; to the primary dichotomy of the reptilian only an unresolved multichotom\- containing radiation was generally rejected (Stovall, at least four of the above seven groups. It Price and Romer, 1966). The reasons for this was therefore felt that the following pro- rejection are two fold. The fossil material cedure should bi' pursued: Those characters availalilc to Peabody for study was meager shared by most of the above early reptiles will and crucial portions of the skull and post- be considered shared primiti\-e characters

crauial skeleton were unknown. In his analy- ( symplesiomoqDhs ) for .\mmiota and re-

sis of relationships Peabody included among moved from discussion when testing hypothe- his "progressive features" many morphologi- s(\s of relationships of Petrolacosaurus. These

cal characters tliat later proved to be charac- characters are listed below and separated teristics of early reptiliomorph amphibians or from later discussions because retention in

all early reptiles, rather than features restrict- Petrolucosatirus of many of these primitive ed to particular reptilian taxa. For example characters does not in itself indicate close the suture pattern of the palate in Pcfrolaco- lilnlogcnetic relationships within the Am-

saurus is close to that of Youngina, as indi- miota (Hennig, 1966). Hypotheses of rela-

cated by Peabody (1952) but is also similar tionships of Petrolacosaurus will be then to the pattern seen in primitive captorhino- presented, the basic taxa discussed and the morphs (Carroll and Baird, 1972), pelyco- hypotheses tested using shared derived char- saurs, millerosanrs (Gow, 1972) and even acters. such reptiliomorph amphibians as Gephyro- Phimitive Cn.\RACTEns stegus (Carroll, 1970) and (Heaton, in press). The discovery and study (1) While pelycosaurs show exponential of nearly complete, well preserved specimens increase in size during the Penns>l\anian, pro- of Petrolacosaurus and recent studies of other torothyridids, "cosuchians," millerosanrs, mes- Paleozoic reptiles (Carroll, 1964, 1969; Car- osaurs and procholophonids retain almost the

roll and liaird, 1972; Clark and Carroll, 1973; same body size throughout their history Gow, 1972, 1975; Heaton, 1979; Holmes, 1977; (Reisz, 1972). Among the protorothyridids 1980 PENNSYLVANIAN DIAPSID REPTILE 55 the skull also increases in length only slightly the cheeks, skull table, palate, occiput and (Table I). The skull of the largest known mandible remains relatively constant in mil- specimen of Petrolacosaurus is only slightly lerosaurs, protorothyridids and primitive pely- longer than the skull of the largest known cosaurs (Gow, 1972; Romer and Price, 1940; protorothyridid, Protorothyris archcri. Reisz, in press). Most features of this pattern The outline of the skull in dorsal view are retained in the skull of Petrolacosaurus in resembles an isosceles triangle with a rounded spite of the morphological advances associ- apex and with sides that bulge out slightly, ated with crania] fenestration. close to the base. In captorhinids, the skull The dorsal expansion of the maxilla re- becomes bulkier and wider than in protoro- mains moderate in most early reptiles. The thyridids. Among pelycosaurs a great variety long lacrimal extends from the external naris of skull outlines exists from the slender, elon- to the orbit, and fonns a large portion of the gate skulls seen in ophiacodonts to the bulky, anterior orbital margin. Although Petrolaco- wide configuration of edaphosaurs and case- saurus, most captorhinomorphs, millerosaurs ids. and primitive pelycosaurs retain this condi- The skull profile remains relatively low. tion, in more advanced pelycosaurs, "eosu-

The maximum skull width is greater than the chians" and mesosaurs the lacrimal fails to maximum lieight in any transverse section of reach the external naris because of the dorsal the skull roof. The ventral edge of the cheek expansion of the maxilla. In most early rep- is nearly straight from the snout to the sus- tiles the jugal forms part of the ventral mar- pensorium. The posterior edge of the skull gin of the cheek. The squamosal and the is tilted only slightly forward. In advanced quadratojugal form the slightly convex pos- pelycosaurs, on the other hand, the skull terior margin of the cheek. The postorbital profile deepens in the temporal region and extends to the supratemporal. The supratem- the ventral edge of the cheek is often strongly poral bone is large in parieasaurs, procolo- cur\'ed ventrally. phonians, millerosaurs and primitive pelyco- The outline of the skull in protorothyri- saurs. In ventral view, the occipital condyle dids, millerosaurs, "'eosuchians" and Petrola- extends posteriorly to the level of the jaw cosaurus, in occipital view, resembles a trape- suspension. In protorothyridids, millerosaurs, zoid with rounded upper angles. The nature pelycosaurs, "eosuchians" and Petrolacosaurus, of preservation of most primitive reptiles the long pterygoids are wedged anteriorly be- precludes precise determination of the angle tween the vomers. Three rows of well-devel- between the cheek and the skull table. It is oped palatal teeth radiate from the basiptery- only slightly greater than 100 degrees in the goid area in most early reptiles; the largest captorhinomorph Protorothyris, and similar palatal teeth are present on the transverse estimates have been computed tor Palcothyris flange of the pterygoid. The long, relatively on the basis of the width of the palate and narrow palatine abutts anteriorly against the skull table (Carroll and Baird, 1972). The plate-like vomer that forms most of the medial wealth of material available for the recon- margin of the internal naris. The ectoptery- struction of Petrolacosaurus .shows an angle goid is small. The parasphenoid is denticu- of 100 ± 5 degrees. A more precise determi- late. nation of this angle was not possible because The paired postparietal and tabular bones

of the curvature of both tlie cheek and skull are restricted to the occiput. The occipital

table. In pelycosaurs other than caseids and condyle is located far \entrally on the occiput,

eothyridids the occiput is much higher than close to the level of the jaw articulation. A in either protorothyridids, eosuchians, mil- large subvertical strip of the squamosal ex- lerosaurs, or Petrolacosaurus. tends onto the occiput. The posterior edge (2) The interrelationship of the bones of of the dorsal process of the quadrate is cov- —

56 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

Fic. 23. Petrolacosaurus kansctuis Lane. Pattern of internal cranial ridges or thickenings in (A) dorsal, (B) ventral and (C) lateral \ie\vs, x 2. Composite. See tevt for further explanation. 1980 PENNSYLVANIAN DIAPSID REPTILE 57

ered by it, except in the region of the quad- trum is ventral to the crescentic atlantal cen- rate foramen. The dorsal process of the quad- trum and is loosely attached to it. Although rate is inchned shghtly anterodorsally. In these elements are disarticulated in Hylono- mesosaurs, early "eosuchians," captorhino- miis [BM(NH)R. 4168], their relative posi- niorphs, pelycosaurs, procolophonids there is tion was apparently similar. In all later no otic notch. The stapes is massive and has captorhinomorphs, however, the axial inter- both a dorsal process and a footplate. centrum and atlantal centrum are indistin-

The mandible is long and slender. The guishably fused. The axial neural arch of coronoid region is little expanded. There is Petrolacosaurus is strongly built, and has a no retroarticular process. large spine that extends anteriorly over the atlas, as in the captorhinomorphs, "eosuch- (3) In most early reptiles the marginal ians," millerosaurs and pelycosaurs. teeth are arranged in a single row. They are The cervical ribs flattened, slightly simple conical structures, slightly recurved, have with subthecodont implantation. An anterior expanded ends. The sacral and anterior cau- dal ribs are firmly attached to the vertebrae. food trap is partially associated with some There sacral caniniform tooth development. are two ribs. The primitive reptilian configuration of (4) Vertebrae are amphicoelous and noto- the girdles in early captorhinomorphs, miller- chordal. Relatively large intercentra are con- osaurs and some pelycosaurs is also seen in tinued in the tail as chevron bones. The Petrolacosaurus. atlantal intercentrum is very large, has a well- The limbs are well ossified. developed ventral median ridge, and a pair Both proximal and distal limb elements thick shafts of articular facets for the capitulum of the have and broad ends. The humerus has a well atlantal rib. In most pelycosaurs, in millero- developed entepicondylar foramen a su- saurs, "eosuchians" and captorhinomorphs and pinator process. The manus and pes are rela- the atlantal centrum is excluded from the tively large and broad, with a long-necked ventral border of the column by the axis astragalus and a plate-like intercentnmi. In Hylonomiis, the earliest rep- calcaneum. resentative of the captorhinomoi"phs, the Hypotheses of Relationships atlantal centrum is crescentic in outline and open ventrally. The axial intercentrum is Two of the hypotheses presented and longer than those in the rest of the column. tested here are illustrated in Fig. 26. The first

In Petrolacosaurus the large axial intercen- hypothesis is that Petrolacosaurus shares a

Pu/c'olhyri' PeltvL'dcmdiniis ) 'oim^ina

Fig. 26.—Cladogram illustrating hypotlieses of relationships of Petrolacosaurus. See text for further expla- nation. 58 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7 more recent ancestor with the captorhino- Eothyrididae. erected as a pro\isional group morph Paleothyris than witli the by Romer and Price (1940), also included Eothtjris. All could have been used the fragmentary remains of three pelycosaurs but Eothijris was selected as the representa- of large size because they are "Ophiacodont tive of this group because it is the most pelycosaurs, primitive in most known regards primitive pelycosaur and is most similar to but paralleling the higher sphenacodonts in Paleothyris and protorothyridids. This hy- the development of much enlarged canines pothesis is tested by characters 1-6 (see dis- and showing a tendency toward elongation of

cussion below ) . The only morphological char- the \ertebral column" (Romer and Price, acter that may falsify this hypothesis is the 1940, pp. 246-247). Both the presence of presence of the lower temporal fenestrac in caniniform teeth and long vertebrae represent both Eothijris and Petrolacosaurus. Other primitive conditions for , and the significant differences in the architecture of dental and vertebral patterns seen in the the skulls of these reptiles and the presence advanced ophiacodont Ophiacodon arc de- of six shared derived characters in the un- rived conditions. Both Stereophallodon and fenestrated Paleothyris and Petrolacosaurus Stereorhachis are ophiacodont pelycosaurs suggest that the lower temporal fenestrac in that have retained the primitive dental and the above taxa are not phylogenetically ho- vertebral pattern and ha\e been therefore mologous. included into the family Ophiacodontidae (Reisz, in press). Balduinonus trux appears The second hypothesis is that the primi- tive diapsid Youngina shares a more recent to present a more difficult taxonomic problem ancestor with Petrolacosaurus than with the because the type specimen includes fragments protorothyridid Paleothyris. All protorothyri- of large ophiacodont \ertebrae and the maxilla dids could have been used as the outgroup of a large sphenacodont pelycosaur (pers. for the tested characters but Paleothyris was obs.). The study of tiiis material is now in progress, but it is certain that none of the selected because it is the most primitive, well materials associated with this species is an preserved protorothyridid, is mo.st similar to eothyridid pelycosaur. Petrolacosaurus and is therefore best suited 196.5 for the search of characters unique to diap- Langston, in his description of small eoth\ridid Oedaleops campi, sids. Youngina was chosen for comparison another indicated that this species is remarkably simi- with Petrolacosaurus because it is the best known early "cosuchian" diapsid reptile lar to Eothyris. Oedaleops has smaller eanini- teeth than Eothyris, .similar in size to (Cow, 1972; Carroll, 1977) and is closest fomi but morphologically to the latter species. In addi- those found in early captorhinomorphs, Oedaleops is peculiar in that the supratcm- tion, there is little evidence to indicate that the posterior the "Eosuchia" are strictly monophyletic. poral projects slightly beyond edge of the cheek. The exact configuration of Although onl\' Youngina is shown in the clad- posterior of the supratemporal in ogram (Fig. 26), characters 7-14 (see di.scus- the end of overprepa- sion below) also test the hypothesis of a Eothyris is not known because monophyletic Diajisida including Petrolaco- ration. This posterior projection of the supra- shared derived saurus. No characters are known that falsify temporal may represent a for family, further materials this hypothesis. character the but are needed before this can be verified. Eothyridids, represented only by the spe- Basic Taxa cies Eothyris parkeyi and Oedaleops campi,

Eothyris is a small, remarkably priiniti\e, are closely related to the caseid pelycosaurs pelycosaur known from a single skull from as indicated by a series of shared derived the Lower Permian of . The familv characters (Reisz, in press). Significant 1980 PENNSYLVANIAN DIAPSID REPTILE 59

shared piimiti\e cliaractcis found in both was based on solely primitive ("ancestral") caseids and fiotliyridids that appear in a characters, and (3) the osteology of most spe-

derived form in all otiier pelycosaurs are cies is based on single partially preserved

worth noting: The width of the skull is great- sp(>cimens, or several poorly preserved scat- er than the height, e\en in the region of the tered skeletons. Of the better known protoro- snout: the frontal either does not extend to thyridids, Cephalerpeton ventriarmatum

the orbit (Eotlu/ris) or the orbital margin of (Carroll and Baird, 1972) is specialized in

the frontal is very short (Oedaleops and all having only 16 large maxillary teeth while

caseids ) ; the supratemporal bone is large and Anthracodwmeus longipes is specialized in broad. In eaptorhinomorphs, millcrosaurs and having long limbs. The most significant spe-

proeolophonids the skull retains the priniiti\e cialization of Coelostegus prothales is seen in

cotylosaur pattern of a low-profile skull. This the skull roof, where the posterior margin is primitive condition persists only in caseid and deeply embayed. The Lower Pennian Pro- eothyridid pelycosaurs. In other pelycosaurs, towthijris archeri (Clark and Carroll, 1973) the snout in the edaphosaurs, and the whole has an unusually large skull and tall neural skull in ophiacodonts, waranopsids and sphen- spines. These species probably represent sep- acodonts becomes narrow and deep. Only in arate monophyletic taxa rather than a single cotylosaurs ( Limnoscelis, Diadectcs, Sey- monophyletic group, but better preserved mouria) and the caseid and eothyridid pely- materials are needed before such an hypoth-

cosaurs is the primitive pattern of the frontal esis can be tested. retained. In other pelycosaurs, in millcrosaurs, A series of diapsids from the later Permian and in early eaptorhinomorphs one-third of and early Triassic are placed on the basis of

the dorsal orbital margin is fomicd by the primiti\'e characters, within the order "Eo- frontal. The broad supratemporal, a primi- suchia.' They arc small, and most are poorly

tive reptilian and cotylosaurian character is known. The assigned genera appear to con- modified in eaptorhinomorps, diapsids and stitute a number of distinct lineages. The

advanced pelycosaurs, but persists as a broad main stock is represented by Youngina from sheet in caseids and eothyridids. the Upper Permian Daptocephcdus zone of

Paleothyris, the best known protorothyri- South Africa. This species is known from sev- did captorhinomorph, was described by Car- eral skulls described by Broom (1914), Olson roll (1969). This small reptile is represented (1936), Gow (1975) and Carroll (1977) and by two nearly complete skeletons and numer- from postcranial skeletons described by ous disarticulated specimens, all recovered Broom (1924), Watson (1957) and Gow

from a single Middle Pennsylvanian Sigiillaria (1975). The skull of Youngina is in many stump from Florence, Nova Scotia. Paleothy- significant features similar to that of Petro- ris exhibits a combination of morphological lacosaunis, including large upper temporal,

characters that make it the least specialized lower temporal, post-temporal and suborbital

well preserved member of the family Protoro- fenestrae. There is no otic notch, but the

thyrididae. The group of small reptiles placed posterior margin of the (juadrate is not cov-

within this family is characterized by the ered by the squamosal. The \ertebrae are possession of features usually considered ((uite primitive and close to the pattern seen primiti\-e (plesiomorphic) for the Amniota. in eaptorhinomorphs, but they also ha\e It is in fact difficult to find any derived char- slightly swollen neural arches and j-jowerful acters that are unique to this group. The zygapophyses, with nearly horizontal articu- reasons for this are that ( 1 ) no attempt has lating surfaces, as seen in Pctrolacosaunis. been made to search for mor^^hological char- Heleosaurm (Broom, 1907; Carroll, 1976a)

acters that are unique to this group, (2) the from the Cisteccphahts zone, is known from placement of the species within the family a single partially preserved specimen appar- 60 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

ently combining primitive "eosuchian" fea- posterior end of the pectoral girdle to the tip

tures with a number of highly speciahzed, and of a long tail. Crushing makes interpretation

possibly archosaurian, adaptations. In this of the trunk vertebrae difficult, but it appears specimen only the palate, marginal bones of that, as in Petrolacosaiirus, the neural arches the cheek region, lower jaws, an articulated are massively built, the zygapophyses extend series of presacral vertebrae and ribs, dermal far laterally and the articular surfaces for the armour-plates along the vertebral column, in- ribs in the anterior dorsal region are elongate complete girdles and femora are preserved. with expanded dorsal and ventral ends. Al- The known parts of the skull of Heleosaurus though the preserved appendicular skeleton

are similar to the corresponding parts in of Kenijasaurus is in some features similar to

Petrolacosaiirus and Younghui despite the the- that of Petrolacosaurus, the forelimb is spe- codont marginal dentition. Heleosaurus also cialized in having a wide paddle-like humerus resembles Petrolacocaurus in a number of and the fifth tarsal appears to have fused to postcranial features including elongated cer- the fourth tarsal. vical vertebrae with long prezygapophyses, A number of primitive, but apparently dis- tilted centrosphenes, mammillary processes on tinct aquatic eosuchians from the Lower the neural spines and well-developed ventral Triassic of Tanzania and Madagascar are cur- rently keels. According to Carroll (1976a) the fe- being described by Mr. Philip J. Currie mur in Heleosaurus is far advanced from the at the Provincial Museum of Alberta, Edmon- primitive reptilian configuration, in correla- ton. The early descriptions by Haughton

tion with an upright posture. Careful exami- (1924) and Piveteau (1926) reveal little de- nation of the specimen reveals, however, that tail, but show general postcranial proportions the exposed posterior surface of the femur similar to those noted in Petrolacosaurus. was mistakenly identified by Carroll as the Prolacerta (Camp, 1945; Cow, 1975) from ventral surface. The exposed surfaces of both the Lower Triassic Lystrosaurus zone of South the proximal and distal heads and the curva- Africa represents yet another "eosuchian" ture of the femur correspond closely to those lineage. Although the skull of Prolacerta has

of primitive captorhinomoqihs and is not spe- an unusual combination of -like and

cialized in the manner suggested by Carroll. thecodont-like features, it also shares a large Another genus from the Cistecephalus number of primiti\x> characters with Young,ina

zone, Galesphyrus, was originally described and Petrolacosaurus. There is a well devel-

by Broom (1915) as a , but its oped upper temporal fenestra, but the lower

"eosuchian" nature has since been recognized temporal bar is absent. Although there is an by Romer (1966) and Carroll (1976b). Al- otic notch formed by the quadrate, the squa-

though this small reptile is known only from mosal extends far ventrally, preventing strep-

two incomplete postcranial skeletons it has tostyly. Well de\elopcd post-temporal and been associated with the "Eosuchia" on the suborbital fenestrae are also present. The

basis of general similarities in size and pro- dentition is apparently thecodont (Cow, portions to Youngina. and the particular simi- 1975), although the anterior and posterior larities of the vertebral, carpal and tarsal edges of the teeth are not serrated. The post-

structures. Many of these morphological cranial skeleton of Prolacerta is surprisingly characters are also seen in Petrohicosaurus. similar to that of Petrolacosaurus. The atlas- The postcranial skeleton of a new "eosuch- axis complex and the other cervical vertebrae ian" reptile, Kcntjasaurus, from the Triassic are nearly identical, the configuration of cervi- beds near Mombasa, Kenya has been de- cal, dorsal and sacral ribs and their pattern .scribed by Harris and Carroll (1977). Tlie of articulation are similar. In contrast to the type specimen of Kenijasaurus has a series of condition in Petrolacosaurus, the caudal re- articulated vertebrae that extend from the gions of Prolacerta and Youngina are short — —

1980 PENNSYLVANIAN DIAPSID REPTILE 61 and the laterally directed ribs appear to be condition is considered derived for capto- firmly fused to the centra. The pelvic girdle rhinomorphs and diapsids. In carnivorous

is similar to that of Petrolacosaurus. The limbs pelycosaurs the same advanced pattern is are lightly built, with the distal paired limb seen, but is probably arrived to independently elements equal in length or longer than the within the synapsids. proximal elements. The tarsus of Prolacerta 3. Reduction in size or loss of the tabular

is derived compared to Petrolacosaurus in bone. In Eothijris. and all other pelycosaurs lacking a discrete fifth tarsal bone and in the tabular bone is a large sheet located on having a hooked fifth metatarsal. the occiput. In Paleothyris, and other protoro- This group appears to have the equiva- thyridid captorhinomorphs, in Petrolacosaur- lent position among diapsid reptiles as the us, Younp,ina and some other diapsids the eothyridids have had among pelycosaurs tabular bone is greatly reduced in size, a con- (Romer and Price, 1940) —a provisional "gar- dition considered derived for these forms. In bage bag" assemblage of primitive forms. As captorhinid captorhinomorphs and in many

it now stands the "Eosuchia" has no taxonomic diapsids the tabular bone is lost.

validity because it cannot be defined in terms 4. Proximal and di.stal limb elements are of shared derived characters not found in any elongate and lightly built. In primitive pely-

other group. Until this group is thoroughly cosaurs (Reisz, in press) the limb elements reviewed, and proper diagnosis provided, no are relatively thick and massively built. Be-

new forms should be placed within it. cause the postcranial skeleton of Eothijris is unknown, no direct comparison is possible. Shared Derived Characters Testing In both Paleothyris and Petrolacosaurus the THE Hypothesis of Relationship proximal and distal limb elements the shaft is exceedingly slender, usually less than 10 per Between Petrolocosaurus and cent of the length; the extremities of the bones Paleothyris. are also reduced in size, with the areas of in-

The following characters correspond to sertion and origin of muscles important in the characters 1-6 in Fig. 26. power stroke and recovery concentrated near 1. Loss of contact ])etween the postorhi- the ends. Although this condition is consid- tal and supratemporal. In Eothijris, other ered advanced for these fonns, the heavy pelycosaurs, millerosaurs and other early am- captorhinid caiJtorhinomorphs and the aquatic niotes the postorbital has a long posterior diapsids appear to have secondarily acquired process that extends to the supratemporal. massive limbs. None of the captorhinomorphs, Petrolaco- 5. The cms and pes are narroiv and long saurus or diapsids show this primitive condi- uith overlapping metatarsals and metacar-

tion, and the loss of contact between these pals. In most pclycosaius ( Romer and Price,

bones is considered derived for captorhino- 1940) the pes and crus are broad and the morphs and diapsids. metatarsals and metacarpals do not overlap. 2. Reduction in the size of the supra- In both Paleothyris and Petrolacosaurus as temporal bone. In Eothijris and most pelyco- well as most protorothyridid captorhino- saurs, in millerosaurs and even in cotylo- morphs and primitive non-aquatic "eosuchian"

saurian "amphibians' the supratemporal is a diapsids, the cms and pes are narrow and large sheet of bone wedged between elements thus the proximal heads of the metacarpals of the skull table. In Paleothyris, captorhino- and metatarsals overlap extensively. Although

morphs, and in Petrolacosaurus and other this condition is considered derived for the

primitive diapsids this bone is modified into above forms, captorhinid captorhinomorphs a small, slender element generally located in appear to have either retained or secondarily a shallow trough of the parietal bone. This de\cloped the primitive condition. ——

62 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

6. Presence of a single centrale in the parietal. The ventral ramus extends down pes. In most pelycosaurs (Romcr and Price, onto the anterodorsal margin of the jugal and 1940) two ccMitralia pedis are present. In the posterior ramus of the postorbital comes Paleothtjris, as reconstructed by Carroll in contact with the squamosal forming the

(1969), there is a tiny lateral centrale, but anterior half of the intertemporal bar. In the medial centrale has the outline of a pea- Prolacerta, and even in the "eosuchian" de- nut. This suggests that the tiny lateral ele- rivatives, Ktdineosaurus (Robinson, 1962) and ment may have been an artifact and the much , where the skulls are known in larger imusually shaped centrale represents great detail, .specific relationships of the post- the fusion of the two centrale seen in pelyco- orbital to surrounding bones are the same saurs. In all other captorhinomorphs, in Petro- as in Petrolacosaurus, despite differences in lacosaiiriis and in diapsids a single large other cranial features. In all four genera, the centrale is present, considered to represent dorsal ramus of the postorbital sends a small the derived condition. In a few advanced wedge between the parietal and the post- carnivorous pelycosaurs the same derived frontal. The anteroventral ramus replaces condition has been observed, but this prob- much of the postfrontal and jugal in forming ably represents convergence. the posterior margin of the orbit, and extends far down onto the inside surface of the latter Shared Derived Characters Testing element, close to the jugal-ectopterygoid su- THE Hypothesis of Relationship ture. The lateral portion of each parietal is also Between Petrolacosaurus and Youngina modified as a result of fenestration. In capto- The following characters correspond to rhinomorphs the contact between the parietal characters 7-14 in Fig. 26. and the postorbital and squamosal is exten- 7. Presence of superior temporal fenestra. sive. In Petrolacosaurus, as in Youngina and The most significant departures in Petrolaco- all other diapsids, these contacts are greatly saurus, Yotmgina and other "eosuchian" diap- reduced and the lateral margin of the parietal sids from the captorhinomoqDh skull pattern is deeply emarginated to form much of the involve the development of fenestrae in the edge of the superior temporal fenestra. In skull roof. The upper and lower temporal addition, the parietal in Petrolacosaurus has fenestrae, typical of diapsid reptiles, have the a narrow lateral process that forms the antero- same position and orientation in Petrolaco- ventral corner of the upper temporal fenestra, saurus as in Youngina. The lower temporal a condition also seen in Prolacerta, Kuhneo- fenestra is, however, similar to that in Eothij- satirus and Euparkeria. ris. The derived characters listed in the pre- In Prolacerta, Euparkeria and Kuhneo- vious section shared by the unfencstrated saurus a well developed flange of the parietal

PaJeothyris and the diapsid Petrolacosaurus is directed ventrolaterally, roughly parallel to with Eothyris as the outgroup suggests that the direction of the adductor muscle action. the lower temporal fenestrae in the latter two The condition in Youngina, where the parietal species developed independently. has a very small ventrolateral process, prob- 8. The .shape and interrelationship of the ably represents an intiMiiK^diatc stage between circtnnfcnestral hones are modified as a rc.sidt Petrolacosaurus and the Triassic diapsid rep- of diapsidy. In Petrolacosaurus the postorbital tiles. A well developed ventrolateral process is triradiate in outline as in Youngina, Prola- of the parietal is al-so seen in modern lizards certa (Camp, 1945; Cow, 1975), Euparkeria and Sphcnodon. It is to the dorsal surface of

(Ewer, 1965) and numerous other diapsids. tliis llangt' that much of the adductor muscu-

The dorsal ramus of the postorbital is in con- lature is attached. This adaptation represents tact with a small lateral projection of the a secondary stage in the evolution of the —— —

1980 PENNSYLVANIAN DIAPSID REPTILE 63 upper temporal fenestra and probably devel- pterygoid and the cheek is greatly reduced, oped in response to the need for more secure as the jugal develops a small oval medial muscle attachment. The presence of the process for attachment to the lateral process temporal opening provided the opportunity of the ectopterygoid. There is no contact of attaching muscle fibers to the outside sur- between the ectopterygoid and the maxilla. face of the parietal. A similar development is The same arrangement is found in Youngina, seen in adxanced -like reptiles, where Heleosaurus, Euparkeria and even Kuhneo- the adductor musculature also invades the saurus. In Paleothijris and other protorothy- dorsal surface by passing through the lateral ridids the palato-maxillary suture extends to temporal opening. This occurs long after the posterior end of the palatine. In addition, initial fenestration and extensive adaptive it is probable that the ectopterygoid also radiation of early mammal-like reptiles—the attached suturally to the maxilla, since in pelycosaurs. these forms the ectopterygoid was a posterior The squamosal of Petrolacosaurus, al- extension of the plate-like palatine (Carroll, though retaining its plate-like configuration, 1969). In captorhinids the ectopterygoid is is modified anteriorly and dorsally to form lost, and the maxillary contact with the ecto- the margins of the temporal openings and the pterygoid is replaced by a strong contact intertemporal bar. As in Youngina and all between the jugal and the pterygoid (Heaton, other diapsids, the contact between the squa- 1979). In advanced pelycosaurs such as Di- mosal and the postorbital is much reduced metrodon, the jugal also comes to brace the over the condition in captorhinomorphs. The pterygoid, while the ectopterygoid-maxilla posterior part of the jugal is modified from a suture is reduced. plate-like sheet to two relatively narrow bands 11.—A pair of well developed post-tem- that extend dorsally and posteriorly. The poral fenestrae are bounded in Petrolaco- large suture between the jugal and squamosal saurus, by a narrow occipital flange of the of captorhinomorphs is replaced in Petrolaco- squamosal, the small tabular, the sitpraoccipi- saurus by the lower temporal opening. tal and the well developed paroccipital proc- 9. Presence of a well developed suborbi- ess of the opisthotic. In Youngina these fenes- tal fenestra. Elongate, well developed sub- trae are somewhat larger than in Petrolaco- orbital fenestrae are seen in Petrolacosaurus, saurus, but are of fundamentally similar con- Youngina and all other diapsids. No suborbi- figuration. In Paleothijris and primitive cap-

tal fenestrae are present in Paleothtjris, capto- torhinomorphs the occiput is poorly known, rhinomorphs or pelycosaurs. but the post-temporal fenestra appears to be 10. The maxilla, palatine, ectoptenjgoid much smaller than in Petrolacosaurus (Car- and jugal bones and their interrelationships roll, 1969). In pelycosaurs the post-temporal are modified as a residt of the presence of tJie fenestra is greatly reduced as a result of the suborbital fenestra. As in Youngina, Heleo- plate-like configuration of the tabular and of saurus and other diapsids the suture between the supraoccipital-opisthotic complex (Romer the cheek (maxilla) and the palate (palatine and Price, 1940, Fig. S).

and ectopterygoid ) in Petrolacosaurus is in- 12. The relative size of the skull, as indi- terrupted by the suborbital fenestra. The cated by the ratio betiveen the length of tlie

suture between the palatine and maxilla is skidl and trunk, is considerably smaller in reduced in length and part of the alveolar Petrolacosaurus than in Paleothyris and cap-

shelf of the maxilla is smooth. The ectoptery- torhinomorphs, and is similar to that in

goid in Pctrolaco.murus is smaller than in Youngina and Heleosaurus. The postorbital

Paleothijris and other protorothyridids (Car- region of the skull is much reduced in Petro- roll and Baird, 1972) and has a robust lateral lacosaurus and Youngina. The orbits in projection. The contact between the ecto- Petrolacosaurus, Youngina and other early —

64 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

diapsids are, on the other hand, relatively ( 1969 ) felt, however, that Araeoscelis and a larger than in captorhinomorphs. closely related form, Kadaliosaurus from the

13. The limhs are long. The humerus Lower Permian of , should be placed and femur in Petrolacosaurus are equal in within the captorhinomorphs on the basis of length to about seven mid-dorsal vertebrae. their similarities to the primitive captorhino- In Youngina and Kenijasaunis the femora re- morphs.

tain the same relative length, but the humeri Vaughn ( 1955 ) has also recognized the are relatively shorter, equal in length to about possibility that Araeoscelis and Petrolaco- five trimk vertebrae. In Paleothyris and most saurus are closely related and suggested that primitive captorhinomorphs the femora and they may belong within the same family. In humeri are equal in length to from 4.5 to 5 fact, these genera resemble each other so dorsal vertebrae. In "eosuchians" the radius closely that there is no doubt about their and ulna, tibia and fibula are nearly equal to close relationship in spite of some apparently (Youngina Galesphijrus, Kenijasaunis) or significant differences in their skulls. These even longer than the himierus and femur re- two reptiles differ only in a few osteological spectively (Prolacerta). The distal end of features.

the tibia of the above genera is relatively The dentition in Araeoscelis is robust and much larger than that of the captorhinomoriDh shows some cusp development in contrast to Paleothyris, suggesting a firmer joint with the the delicate, piercing type of dentition seen astragalus. in Petrolacosaurus. In association with this adaptation to a different type of diet, the 14. Locked tibio-astragaJar joint. The massive prefrontal extends down to the max- pes of Petrolacosaurus is generally similar to illa and excludes the thin lacrimal from the that of Galesphijrus, Prolacerta, Kenyasaurus, orbit. The postorbital has a long posterior and (the pes of Youngina is projection that reaches the supratemporal. A poorly known) except for the relatively large recently uncovered specimen of Araeoscelis size of the proximal elements (Carroll, 1977). confirms Vaughn's interpretation of only up- The tibial facet of articulation on the astra- per temporal fenestrae in the postorbital por- galus of Petrolacosaurus, formed by two lon- tion of the skull roof. The postorbital region gitudinal ridges separated by a deep trough of the skull is very deep with the articulation fits snugly onto the distal articular surface of between the jaws far below the level of the the tibia to fonn a firm, locked joint. This maxilla. The occipital plate as restored by configuration can be detected only if the Vaughn (1955), is somewhat more extensive tarsus is exposed in ventral or medial views. than in Petrolacosaurus and the post-temporal Thus, only Petrolacosaurus and Kenyasaurus fenestra arc therefore small. This inteq^reta- are now known to have this character, but it tion is based on a single, poorly preserved is probable that other "eosuchians" have the occiput in an immature specimen. The lower same pattern. No pclycosaur, millerosaur or jaw is also relatively massi\e. All these dif- captorhinomorph has this advanced condition. ferences in cranial morphology may be adap- tations of Araeoscelis to a different diet from Co.mpared Petrolacosmims With the insectivorous Petrolacosaurus. Aracoscelis The atlantal centrum in Araeoscelis is in- The reptile Araeoscelis from the Lower distinguishahly fused to the axial intercen-

Permian of Texas has been often allied by trum, a condition which is somewhat ad- Romer (1956, 1966) and VViUi.ston (1925) vanced over that in Petrolacosaurus. There with a group of poorly known late Permian are nine elongate cervical vertebrae in Araeo- and early Triassic genera, referred to as pro- scelis, according to Vaughn's reconstruction torosaurs. Roth Vaughn (1955) and Carroll versus six in Petrolacosaurus. The neural 1980 PENNSYLVANIAN DIAPSID REPTILE 65 arches of the two genera differ in that those possible ways of interpreting the morphologi- of Araeoscelis have lower neural spines and cal characteristics of these two genera: (1) the diapophyses and parapophyses are widely they evolved separately from an unfenestrated separated in the dorsal region of the column, ancestor which already achieved some of the with no bony web to connect them. The postcaptorhinomorph skeletal features shared girdles are very similar; the only distinguish- by both genera; (2) they evolved from a rep- ing features are the wider scapular blade in tile with an upper temporal fenestra, by the pectoral girdle and the ischiadic notch of strengthening the skull in the line leading to the pelvis in Petrolacosaiirus. The limbs in Araeoscelis, and by lightening the skull and Araeoscelis are slightly more specialized than also developing a lower temporal fenestra in in Petrolacosaiirus in the presence of a fully the line leading to Petrolacosaurus; (3) Arae- enclosed ectepicondylar foramen on the hu- oscelis evolved from a diapsid condition like merus versus a supinator process, in having that of Petrolacosaurus by closure of the lower a more slender femur, a cnemial process on temporal fenestra during adaptation to a dif- the tibia and in occasional fusion of the fourth ferent diet. and fifth tarsals. These few osteological dif- Similarities in the configuration of the can be ferences in the postcranial skeleton upper temporal fenestra and of the suborbital readily accounted for by the fact that Petro- fenestra suggests that one of the latter two lacosaurus is the larger (therefore requires possibilities may be valid. A more precise relatively more bony support) and geologi- interpretation of the taxonomic position of cally older animal (therefore more primitive Araeoscelis and the evolutionary relations of in structure). the must await the results of a Clearly Petrolacosaurus and Araeoscelis two genera are closely related, with the younger genus study (now in progress) of two recently dis- specialized to a diet that necessitates a covered and well preserved skeletons of stronger, more massive skull. There are three Araeoscelis. SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

HEARING IN PETROLACOSAURUS AND OTHER EARLY REPTILES

Advanced cranial osteological features of within the oval window (Werner and Wever, late Permian and early Triassic "eosuchians" 1972).

include modifications of the posterior margin In lizards the quadrate, freed from its of the cheek, the stapes and the posterior end squamosal cover, forms the strongly notched of the mandible. The posterior border of the posterior border of the cheek. In addition, squamosal in Youngina does not completely the lower jaw extends backward, beyond the cover the dorsal process of the quadrate, so articulation between the quadrate and articu- that this bone is visible in lateral view. The lar. In most extant lizards the strongly

posterior border is slightly concave. The notched quadrate and the retroarticular proc-

shaft of the stapes is slimmer than in capto- ess of the lower jaw support a large t)'mpanic rhinomorphs and PetTolacosaurus. This ar- membrane. The posterior border of the tym- rangement of a light stapes and a slightly con- panum is in part supported by the M. de-

cave quadrate is somewhat similar to that pressor mandibularis which originates from seen in modern lacertilians, where the strong- the top of the occipital plate of the skull and ly notched quadrate supports a tympanic inserts on the posterior end of the retroarticu- membrane. It has therefore been assumed lar process. Because similar osteological fea- that Youngina had a tympanum. The condi- tures are seen in such Triassic diapsids as tion in Euparkeria and Prolacerta is even Prolacerta, Euparkeria, Kuhneosaurus, Paleo-

closer to that seen in living lizards. The con- gama and , it is possible that in cavity of the posterior edge of the quadrate these forms the quadrate and the retroarticu- is much stronger than in Youngina, there is lar process supported a relatively large tym- a well developed retroarticular process and panum. None of these structural features are, the stapes is reduced to a slender rod. however, seen in Petrolacosaurus, captorhino-

In PetTolacosaurus, on the other hand, the morphs or even pelycosaurs. There is neither squamosal forms the slightly convex posterior a notch at the posterior edge of the cheek, border of the skull and the dorsal process of nor a posteriorly directed retroarticular proc-

the quadrate is completely covered. The stapes ess. Under these conditions, the M. depressor

is massive (Figs. 5 and 13). mandibularis had to hug the posterior border

Consideration of the functional implica- of the skull as it extended down to the lower tions of structures observed in early reptiles jaw. There was therefore, no place for the is impossible without comparisons witli living tympanum. There is no evidence for any forms. Altliough generalizations about the bony support for a tympanum. This support middle ears in living reptiles are tenuous would be needed to maintain high membrane because of tlie morphological differcTices seen, elasticity in the hydraulic and curved-mem- even within families, certain overall charac- lirane level action of the middle ear mechan- teristics can be noted in forms which are ism (Wever and Werner, 1970; Manley, sensitive to airborne sound. Reptiles sensitive 1972). to airborne sound have a well-developed tym- All these features indicate that there could panum held taut by extensive bony support. be no tympanum in Petrolacosaurus, capto- The areal ratio between the tympanum and rhinomorphs or pelycosaurs. Vaughn (1955), stapedial footplate is large (at least 30 to 1) Hotton (1959) and Parrington (1955) have and the mass of the stapes is low. The stapes suggested that the t>'mpanum in primitive

is tv^pically a slender osseous rod with a reptiles were located more medially and were slightly expanded footplate that is suspended connected to the outside by an external audi- 1980 PENNSYLVANIAN DIAPSID REPTILE 67 tory meatus. These reconstructions did not and most of the aerial sound impinging upon take into account that the tympanum had to the inner ear through \arious pathways would be located at the distal end of the stapes not be transmitted, but absorbed and re- and that the suggested locations of the tym- flected (Wever and Lawrence, 1954). We panum were occupied by extensive cranial would, therefore, expect the ear of these and neck musculature. primitive reptiles to be insensitive to airborne In captorhinomorphs, pelycosaurs and soiuid.

Petrolacosaunis the stapes is very massive and In spite of the supposed relative in- is firmly attached both to the quadrate and sensitivity of the primitive reptilian type of the braincase. Its higher inertia makes initia- ear to airborne sound, there is little doubt tion of movement much more difficult than that with reduction of stapes mass and the in modern forms. The footi^late of the stapes development of bony support for a tympa- is also very large. This makes high areal ra- num, the ability of the ear to perceive low tios between a t\'mpanic membrane and frequency airborne sounds would greatly in- stapedial footplate impossible. To achieve crease. The presence in Euparkeria and Pro- comparable areal ratios to those seen in mod- lacerta of a light stapes and a large area for ern reptiles, the tympanum would have to be a bone supported tympanum, suggests that too large to fit anywhere behind the skull. these genera have already achieved higher

It is even more significant that the large sensitivity to airborne sounds than captorhino- footplate was fitted within a bony socket of morphs, or Petrolacosaunis. the braincase, the fenestra ovalis externa It is relatively easy to derive the structural (Hcaton, 1979), and that the distal end of condition of the ear region of Prolacerta and the stapes was probably continued in a short Euparkeria from that in captorhinomoqohs or cartilaginous process that fitted in the stape- Petrolacosaunis. The lightening of the stapes, dial recess of the quadrate. All of this indi- the loss of the squamosal cover of the pos- cates that the stapes could not fomi an ef- terior end of the quadrate and the associated fective link within a mechanical transducer notching in the quadrate are related with system sensitive to airborne sounds. Although the middle ear's increasing efficiency as a Hotton attempted to show that the massive mechanical transducer. The development of stapes of Dimctrodon can be activated by a a retroarticular process may, however, have relatively small membrane, his poorly cali- the dual advantage of (a) increasing the brated experiments (1959, Fig. IB) provide mechanical advantage of the M. depressor no useful evidence in this regard. No attempt mandibularis, and (b) increasing the effi- was made during the experiments to establish ciency of the middle ear by both increasing the intensity of airborne sound. In addition, the available surface area between the quad- the model was not intended to approach the rate and the M. depressor mandibularis, and structural pattern of the middle ear of any by providing better support for the ventral reptile. part of the tympanum. What conditions The middle ear in captorhinomorphs, pely- would lead to selection of a middle ear sensi- cosaurs and Petrolacosaunis could not func- tive to airborne sound? Vocal communica- tion as in most modern reptiles. In modern tion has been cited as a possible factor ( Man- reptiles a highly effecti\e mechanical trans- ley, 1973). The fossil evidence indicates that ducer amplifies small sound pressures im- middle ears with efficient transducer mechan- pinging upon a large tympanic membrane isms developed from the captorhinomorph into strong pressures exerted at the fenestra condition at least three times (, ovalis. Without an eflFectivc tiansducer mech- "eosuchians" and chelonians). anism the great difference in impedances of Studies now in progress ( Heaton, pers. air and inner ear could not be compensated, comm.) indicate that the stapes in primitive 68 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

reptiles braced the braincase against the quad- region are taking place. The type specimen rate in a metakinetic skull. In the lineage of Youngina capensis shows that the squa- giving rise to lizards anterior rotation of the mosal no longer covers the posterior margin braincase relative to the skull roof (metakine- of the quadrate, and the posterior margin of sis) was limited by the development of ac- the quadrate is slightly concave, but not cessory structures on the supraoccipital and notched. The type of Youngopsis rubidgei parietal and, to a lesser extent, by the full has a stapes with a relatively slender shaft ossification of the paroccipital process, there- and the paroccipital process extends dorso- by freeing the stapes from its supportive laterally to a notch between the squamosal fimction. Once freed, the stapes become a and the quadrate. Both Youngina and Heleo- hearing ossicle. The condition seen in Petro- saurus have very short retroarticular proc-

lacosaurus may represent the first stage in the esses. It is unlikely, however, that a tympa- sequence, where the stapes is still massive, num was present in these "eosuchians." The but in contrast to the condition seen in primi- condition seen in Prolacerta, Palaeagama and tive captorhinomorphs, the paroccipital proc- Euparkeria represent the final stages in this ess is fully ossified and extends far laterally. sequence, where the osteological modifica- Youngina would represent an intermediate tions necessary for having a middle ear sensi- stage, where several modifications of the ear tive to airborne sound have been completed. 1980 PENNSYLVANIAN DIAPSID REPTILE

CONCLUSIONS

The detailed study of the osteology of the greater agility suggest that this reptile adapt- Pennsylvanian reptile Petrolacosaurus kan- ed to a somewhat different food supply than sensis adds to our understanding of the anat- captorhinomorphs and pelycosaurs. It is dif- omy of early reptiles. A good knowledge of ficult to specify what food was selected, but the osteology of this form also permits de- it is probable that Petrolacosaurus fed mainly tailed comparisons with other Paleozoic rep- on terrestrial arthropod invertebrates. tiles. Close phylogenetic relationships of Petro- Several authors have suggested that Petro- lacosaurus to early "eosuchians" are indicated lacosaurus should be allied with pelycosaurs. by derived characters in common with

This association is invalid for the following Youngina, Galesphyrus, Heleosaurus, Kenya- reasons: As already indicated, the morpho- sauras, and Prolacerta. Most derived petrola- logical features in which this reptile is de- cosaurid features, not seen in either capto- rived (the synapomorphies of Hennig) over rhinomorphs or pelycosaurs, are found in the captorhinomorph pattern are not at all Youngina. The similar pattern of fenestration similar to the derived features seen in pely- in the skulls of Petrolacosaurus and Youngina cosaurs. The similarities between Petrolaco- is the most significant of these advanced fea- saurus and early pelycosaurs are mostly tures. Youngina and other diapsids are more primitive characters, also shared with the advanced than Petrolacosaurus in some cra- captorhinomorphs. In addition it has been nial features, but the latter genus is geologi- argued (Reisz, 1972) that the lower temporal cally much older, and understandably more opening seen in the pelycosaurs developed primitive. The most significant of these spe- in response to selection for more efficient use cialized features, not found in Petrolaco- of the jaw musculature in forms of increasing saurus, involve osteological modifications to- body size. The skulls of the Middle and Up- wards development of a middle ear system per Pennsylvanian pelycosaurs arc larger and sensitive to airborne sound. There is evidence considerably more massive than the skulls to indicate that such modifications in the re- of cither Petrolacosaurus or early captorhino- gion of the ear may have been initiated by morphs. Early pelycosaurs were, therefore, the adaptations already seen in Petrolaco- able to include into their food supply saurus. Similarities in function, shape and of greater size, and were probably able to orientation of the temporal openings and the feed upon captorhinomorphs and other small interrelationship of the circumfenestral bones contemporary tctrapods. In the postcranial are strong indications that the upper and skeleton, pelycosaurs show advances over the lower temporal fenestrae in Petrolacosaurus, primitive captorhinomorph pattern that can Youngina and other diapsids are phylogeneti- be readily attributed to accommodation for cally related. the greater body size. Petrolacosaurus, on the The erection of the family Petrolacosauri- other hand, evolved in a completely different dae by Peabody (1952) is valid because there direction. Its diapsid skull configuration prob- are several derived characters that can be ably developed in response to the selective used to define this taxon. These derived char- advantage of maintaining a relatively small, acters are: an unusually slender, lightly built light-weight skull at the end of a long neck. premaxilla, unusually thin walled marginal The postcranial features indicate that Petro- dentition, sLx elongate cervical vertebrae, lacosaurus was more agile than the capto- mammillary processes on the neural arch of rhinomorphs or pelycosaurs. The combination the first sacral vertebra, a well developed of a relatively small, lightly-built skull and ischiadic notch and forelimbs equal in length 70 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

to hindlimbs. Placement of the family Pctro- nature of the fossils at Gamett, Kansas sug- lacosauridae at the base of the order "Eo- gest that Petrolacosaurus and other terrestrial

suchia" (Peabody, 1952; Reisz, 1977) is how- , invertebrates and plants were ever inappropriate because the latter taxon washed into a quiet lagoon from dry ground

is undefined. The inclusion of early diapsid (Peabody, 19.52). The gap in the fossil rec- reptiles into the order "Eosuchia" solely on ord may indicate that the early diapsids lived the basis of characters that are primitive for in environmental conditions which were not,

diapsids is invalid. Both the best known under normal conditions, conductive to pres- "cosuchian" family, the Younginidae and the ervation—areas away from standing water, order "Eosuchia" therefore need to be re- or that areas where they may have been viewed and testable hypotheses of relation- buried have either been eroded away or not ships constructed before their phylogenies can exposed by erosion. be reevaluated. Recent studies of Upper Permian and Acceptance of Petrolacosaurus as the old- Lower Triassic diapsid reptiles from South est known diapsid reptile indicates a long Africa ( Carroll, 1975, 1976, 1977; Cow, 1975; fossil record of the diapsid hiatus in the Harris and Carroll, 1977) and a preliminary radiation between its first evidence in the late examination of Permo-Triassic diapsids from Pennsylvanian and the subsequent differentia- Madagascar and Tanzania, show that this tion in the late Permian. The extent of the stage of the diapsid evolution is far more early Triassic diapsid radiation confirms that complicated and extensive than formerly be- the origin of the group should be sought lieved. A more complete picture of the early much earlier, probably well within the Paleo- is order zoic, as Petrolacosaurus appears to substan- diapsid adaptive radiation required in

tiate, but the intervening gap is puzzling. to understand fully how Petrolacosaurus re- The conditions of preservation and the lates to later diapsids. 1980 PENNSYLVANIAN DIAPSID REPTILE 71

SUMMARY

Fossil evidence indicates that Petrohco- small, and the limbs are lightly built. Com- saurus kaiisensis, a primitive diapsid reptile parison of this species with captorhinomorphs from the Upper Pennsyhanian of Garnett, and other early reptiles indicates that the Kansas presents a combination of morpho- diapsid skull configuration may have devel-

logical features that place it near the base oped in response to the selective advantage of subsequent diapsid lineages, while also of maintaining a relatively small, light weight evidencing close relationships to capto- skull at the end of a long neck. The con- rhinomorph reptiles. Shared derived char- figuration of the ear region in both primitive acters (synapomorphies) of diapsids and captorhinomorphs and Petrolacosaurus indi- Petrolacosaurus include the presence of well cates that these animals were insensitive to developed superior and inferior temporal, airborne sounds. There is, however, evidence occipital and suborbital fenestrac. Close to suggest that the selective forces for modi- phylogenetic relations between Petrolaco- fications towards development of a middle saurus and protorothyridid captorhinomorph ear system sensitive to airborne sounds may

reptiles is indicated by the following shared have been initiated in Petrolacosaurus.

derived characters: the supratemporal is Petrolacosaurus is the only described reduced to a small narrow sliver of bone member of a distinct family of diapsid rep- that lies in a shallow groove formed by the tiles whose exact position is difficult to ascer- parietal and squamosal, the postorbital does tain until the younger diapsids, commonly

not reach the supratemporal, the tabular is called the "eosuchians," become better known. .

72 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

LITERATURE CITED

Broom, R. Soc. Lond., ser. B, 248:379-435, figs. 1-19, pis. 1907. On some new fossil reptiles from the 31-34. Karroo beds of Victoria West, South Africa: Cow, C. E. Trans. S, Afr. Phil. Soc, 18:31-42, figs. 7-11. 1972. The osteology and relationships of the

Broom, R. (Reptilia: Cotylosauria ) . Jour. 1914. A new thecodont reptile: Proc. Zool. Soc. Zool., Lond., 167:219-264, figs. 1-25. London, 1914: 1072-77. Cow, C. E. Broom, R. 1975. The morphology and relationships of 1915. On the origin of mammals: Philos. Trans. Yutingina capensis Broom and Prolacerta broomi Roy. Soc. Lond., ser. B, 206:1-48. Parrington. Paleont. afr., 18:89-131, figs. 1-37.

Broom, R. Hahrls, J. M. & Carroll, R. L. 1924. Further evidence on the structure of the 1977. Kenyasaurus, a new eosuchian reptile from Eosuchia: Amer. Mas. Nat. Hist. Bull., 51:67- the Early Triassic of Kenya. Jour. Paleont., 51: 76, figs. 1-4. 139-149, figs. 1-6. Camp, C. L. Haughton, S. H. 1945. Prolacerta and the protorosaurian reptiles: 1924. On reptilian remains from the Karroo beds Am. Jour. Sc, 243:17-32, 84-101, figs. 1-6. of East Africa: Quart. Jour. Geol. Soc, 80:1-11. Carroll, R. L. Heaton, M. ]. 1964. The earliest reptiles: Jour. Linn. Soc. 1979. Cranial anatomy of primitive captorhinid (Zool.), 45:61-83, figs. 1-14. reptiles from the Late Pennsyhanian and Early Carroll, R. L. Permian, Oklahoma and Texas. Bull. Okla. Geol. 1968. A Diapsid (Reptilia) parietal from the Surv., 127:1-84.

Lower Permian of Oklahoma: Postilla, 117:1-7, Heaton, M. J. figs. 1-3. (in press). The Cotylosauria: .\ reconsideration Carroll, R, L. of a group of archaic tetrapods. In: Panchen, 1969. A Middle Pennsylvanian captorhinomorph, A. L. (ed. ), The terrestrial environment and the and the interrelation.ships of primitive reptiles: origin of land . Systematics Assoc, Jour. Paleont., 43:151-170, figs. 1-12. Spec. Symp. Carroll, R. L. Hennic, W. 1970. Quantitative aspects of the - 1966. Phylogenetic Systematics. Transl. by D. D. reptilian transition: Forma et Functio, 3:165-178, Da\is and R. Zangerl. Illinois Univ. Press, 263 figs. 1-5. pp. Urbana, 111. Carroll, R. L. Holmes, R. 1975. Permo-Triassic "lizards" from the Karroo. 1977. The osteology and musculature of the pec- Paleont. afr., 18:71-87, figs. 1-8. toral limb of small captorhinids. Jour. Morph., Carroll, R. L. 152:101-140, figs. 1-20. 1976a. Eosuchians and the origin of Archosaurs. HOTTO-V, N. In: Churcher, C. S. (ed.) Athlon: Essays on 1959. The Pelycosaur tympanum and early evo- Paleontology in Honour of Loris Shano Russell. lution of the middle ear. Evolution, 13:99-121, Misc. Publ. R. Out. Mus.:58-79, figs. 1-11. figs. 1-6. Carboll, R. L. KUHN-SCIINYDER, E. 1976b. Galesjjhtjrus caitemis, a younginid eo- 1963. I Sauri del Monte San Giorgio: Comm. suchian from the Cistecephalus zone of South Instit. Paleont. Univ. Zurigo, no. 16:811-854, figs. Africa. Ann. S. Afr. Mus„ 72 {4):54-68, figs. 1-3. 1-31. Carroll, R. L. Lane, H. H. 1977. The origin of lizards. In: Andrews, S. M., 1945. New Mid-Pennsylvanian reptiles from Kan- Miles, R. S. and Walker, A. D., (eds.): Problems sas. Trans. Kansas Acad. Sci., 47:381-390, figs. in Vertebrate Evolution. Linn. Soc. Symp. Series. 1-3. No. 4:3.59-396, figs. 1-15. Lane, H. H. Carroll, R. L. & BAmi), I). 1946. A sur\'ey of fossil vertebrates Kansas: Part 1972. stem reptiles of the Family 3, the reptiles: Trans. Kansas Acad. Sci., 49:289- Romeriidae; Harvard, Mus. Comp. Zool., Bull., 332, figs. 1-7. 143(5):321-364, figs. 1-14. Lancston, W. 1965. (Reptilia: Pelycosauria ) Clark, J. & Carroll, R. L. Oedaleops campi 1973. Romeriid reptiles from the Lower Permian. A new genus and species from the Lower Permian Harvard, Mus. Comp. Zool., Bull., 144(5):353- of New Mexico and tlie family Eothyrididae. Bull. 407, figs. 1-25. Texas mem. Mus. 9, 1-46. Ewer, R. F. Manley, G. a. 1965. The anatomy of the thecodont reptile 1972. The middle ear of the Tokay Gecko: Jour. Euparkeria capensis Broom. Philos. Trans. Roy. Comp. Physiol., 81:239-250, figs. 1-5. 1980 PENNSYLVANIAN DIAPSID REPTILE 73

Manley, G.A. Reisz, R. R. 1973. A review of some current concepts of the (in press) The Pelycosauria: A review of phylo- functional evolution of the ear in terrestrial genetic relationships. In: Panchen, A. L. (ed.). vertebrates. Evohition, 26(4):608-621, figs. 1-6. The terrestrial environment and the origin of land Miner, R. W. vertebrates. Systematics Assoc, Spec. Symp. 1925. The pectoral limb of Ertjops and other Robinson, P. L. primitive tetrapods. .\nier. Mus. Nat. Hist. Bull., 1962. Gliding lizards from the Upper Keuper of 51:145-312, figs. 1-104. Great Britain. Proc. Geol. Soc, no. 1601:137- Oelrich, T. M. 146, figs. 1-2. 1956. The anatomy of the head of Cteno.saura ROMEB, A. S. pectinata (Iguanidae): Misc. Publ. Mus. Zool. 1944. The development of limb muscu- Univ. Michigan, no. 94:1-122, figs. 1-59. lature—the shoulder region of Lacerta. Jour. Olso.v, E. C. Morph., 74:1-41. 1936. Notes on the skull of Yomigiiia capensis Romeb. a. S. Broom. Jour. Geol., 44:523-533, fig. 1. 1947. The relationships of the Permian reptile Olson, E. C. Protorosaunis. .i^mer. Jour. Sci., 235:19-30. 1961. Jaw mechanisms: rhipidistians, amphib- Romer, a. S. ians and reptiles. Amer. Zool., 1:205-215, figs. 1956. Osteology of the Reptiles. University of 1-7. Chicago Press, Chicago: i-.xxi, 1-772, figs. 1-248. Parrington, F. R. Romer, A. S. 1949. Remarks on a theory of the evolution of 1966. . University of Chi- the tetrapod middle ear. Jour. Laryng. Otol., cago Press, Chicago: i-v, 1-468, figs. 1-443. 63:.589-595, figs. 1-4. Romeb, A. S. Parrington, F. R. 1967. Early reptilian evolution reviewed. Evolu-

1955. On the cranial anatomy of some gorgonop- tion, 21(4 ) :831-833. figs. 1-3. sids and the synapsid middle ear. Proc. Zool. Soc. Romer, A. S. Lond., 125:1-40, figs. 1-14. 1971. Unorthodo,\ies in reptilian phylogeny. Evo-

Parrington, F. R. lution, 25( 1 ) :103-1 12, figs. 1-3. 1958. The problem of the classification of Rep- Romer, A. S. & Price, L. I. tiles: Jour. Linn. Soc. (Zool.), 44(295) :99-115, 1940. Review of the Pelycosauria. Geol. Soc. figs. 1-7. America, Special Paper no. 28:i-x, 1-588, figs. Peabody, F. E. 1-71, pis. 1-46.

1949. Mid-Pennsylvanian pelycosaurs from Kan- Stoval, J. W., Price, L. I. & Romer, A. S. sas: Geol. Soc. America, Bull., 60:1913 (abs.). 1966. The postcranial skeleton of the giant Per- Peabody, F. E. mian pelycosaur Cotiilorbynchus romeri. Bull. 1952. Pctrolacosaurns kansensis Lane, a Pennsyl- Mus. Comp. Zool., 135:1-30. figs. 1-17. vanian reptile from Kansas. Paleont. Contr. Univ. Tatarinov, L. P. Kansas, Vertebrata: 1-41, figs. 1-11, pis. 1-3. 1964. Order : Subclass Lepidosauria. Peabody, F. E. In: Osnovy Paleontologii, 12:332-3.38, 439-493. 1957. Pennsylvanian reptiles of Gamett, Kansas: Vaughn, P. P. Edaphosaurus. Jour. Paleont., 31:947-949, fig. 1. 1955. The Permian reptile Aracoscelis restudied. Bull. Mus. Comp. Zool., 113:305, 467, figs. 1-15, PiVETEAU, J. 1926. Paleontologie de Madagascar, XIII. Am- pis. 1-2. phibiens et reptiles permiens. Ann. Pal. (Paris), Watson, D. M. S. XV: 1-128, pis. 12. 1951. Paleontology and modem biology. Yale

Price, L. I. Univ. Press, New Haven: i-xii, 1-216, figs. 1-77. 1935. Notes on the brain case of Captorhinus. Watson, D. M. S. Boston Soc. Nat. Hist., Pr., 40:377-386. 1953. Evolution of the mammalian ear. Evolu- Reic, C. a. tion, 7(2): 159-177, figs. 1-7. 1967. Archosaurian reptiles: a new hypothesis on Watson, D. M. S. their origins. Science, 157:565-568, fig. 1. 1954. On Bolosaunis and the origin and classi- Reic, C. A. fication of reptiles. Bull. Mus. Comp. Zool., Ill: 1970, The Proterosuchia and the early evolution 297-449, figs. 1-37, pi. 1. of the archosaurs; an essay about the origin of a Watson, D. M. S. major taxon. Bull Mus. Gomp. Zool., 139(5): 1957. On Millerosaurus and the early history of 229-292, figs. 1-16. the sauropsid reptiles. Philos. Trans. Roy. Soc. Reisz, R. R. London, ser. B, 240:325-400, figs. 1-23. 1972. Pelycosaurian Reptiles from the Middle Werner, Y. L. & Weveb, E. G. Pennsylvanian of North America. Bull. Mus. 1972. The function of the middle ear in lizards: Comp. Zool., 144(2) :27-62, figs. 1-20. Gekko gecko and Euhlcpharis macularius (Gek- Reisz, R. R. konoidea). Jour. Exp. Zool., 179:1-16, figs. 1-12. 1977. Petrolacosaunis, the oldest known diapsid Wever, E.G. reptile. Science, 196:1091-1093. 1973. The function of the middle ear in lizard's- 74 SPECIAL PUBLICATION MUSEUM OF NATURAL HISTORY NO. 7

Divergent types. Jour. Exp. Zool., 184( 1 ) :97-125, Crotapliylus collaria (Iguanidae). Jour. E.vp. figs. 1-32. Zool., 175:327-342, figs. 1-9. Wever, E. G. & Lawrence, M. WiLLISTON, S. W. 1954. Physiological acoustics. Princeton Univer- 1925. The osteology of the reptiles. Harvard sity Press: i-xii, 1-354, figs. 1-142. Univ. Press, Cambridge, Mass.: 1-300. Wever, E. G. & Werner, Y. L. 1970. The function of the middle ear in lizards:

3 2044 072 228 8 AVAILABLE SPECIAL PUBLICATIONS MUSEUM OF NATURAL HISTORY, UNIVERSITY OF KANSAS

1. Catalogue of publications in published by the University of Kansas Museum of Natural History. By Linda Trueb. Pp. 1-15. December 1976. $0.25 postpaid.

2. Catalogue of publications in mammalogy published by the University of Kansas Museum of Natural History. By Robert S. Hoffmann. Pp. 1-19. 15 February 1977. $0.25 postpaid. 3. Maintenance of rattlesnakes in captivity. By James B. Murphy and Barry L. Annstrong. Pp. 1-40. 29 December 1978. $3.00 postpaid. 5. The natural history of Mexican rattlesnakes. By Barry L. Armstrong and James B. Murphy. Pp. 1-88. 14 December 1979. $6.00 postpaid. 6. 1980 Catalog of publications of the University of Kansas, Museum of Natural History. Pp. 1-28. September 1980. $1.00 postpaid or free with orders.