LIBRARY OF THE UNIVERSITY OF ILLINOIS AT URBANA-CH/^PAIGN GEOLOGY UNIVERSITY OF 'LLINOIS LIBRARY AT URBANA-CHAMPA/GM GEaOGY ^, V FIELDIANA . GEOLOGY Published by CHICAGO NATURAL HISTORY MUSEUM Volume 14 February 28, 1961 No. 4 The Triassic Reptile, Poposaurus Edwin H. Colbert Chairman, Department of Vertebrate Paleontology The American Museum of Natural History INTRODUCTION The genus Poposaurus has for many years been something of an enigma to students of fossil reptiles, a situation that is largely a result of the inadequacy of known materials. At the present time the litera- ture concerning this genus is based upon the single type specimen, which consists of some vertebrae, limb bones and part of a pelvis. The separate bones of the type seem, for the most part, to be fairly [ definitive, but the sum of the evidence from all of them is rather con- flicting. Therefore it is not surprising that published remarks as to the relationships of Poposaurus show considerable differences of opin- ions, with the result that the systematic position of the genus is at the present time undecided. Poposaurus, fragmentary as the remains may be, is important because it has been commonly regarded as an ornithischian dinosaur of undoubted Triassic age. Whatever may be the taxonomic rela- tionships of this fossil, there certainly is no doubt as to its strati- graphic position; the type was found in the Upper Triassic Popo Agie beds of Wyoming. The genus was named by Mehl (1915b) in a brief description. At that time Mehl did not attempt to make a definitive statement as to the relationships of Poposaurus, although he did say that "in some respects it resembles some of the early dino- saurs." He then suggested that Poposaurus might be related to Palaeodonus, a genus based upon very fragmentary remains from the Newark beds of eastern North America. It need only be said at this place that Palaeodonus, so far as can be determined upon the basis of known fossils, is probably a phytosaur. Mehl (1915b, p. 522) went on to say that "Everything in the structure of the form so far studied indicates a well-muscled creature light in weight, possibly bipedal in gait occasionally, and most assuredly swift in movement." Ur I l^iiSvi^ni^ Congress Catalog Card Number: 61-11257 MAR 8 1961 jj^O^OGY UBBAR* •VtlJIilTY OF IIUNOIS 60 FIELDIANA: GEOLOGY, VOLUME 14 During the years since Poposaurus was first described by Mehl various autliorities have given some attention to this genus and have tried to place it within the system of reptihan classification. The varying opinions expressed by these authors are summarized below. In 1921 Nopcsa published a short note devoted to the systematic position of Poposaurus. He reviewed the diagnostic characters of the genus as presented by Mehl, and came to the conclusion that Poposaurus should be regarded as a primitive orthopod, and should be placed taxonomically among the iguanodonts or the camptosaurs. Thus Nopcsa was the first author definitely to suggest dinosaurian relationships for Poposaurus. A few years later Nopcsa (1928) expressed this view as to the re- lationships of Poposaurus in a formal fashion. He made Poposaurus the representative of a new suborder, the Poposauroidea, one of three suborders that constituted his order Orthopoda. (The Orthopoda, as delineated by Nopcsa, is the equivalent of the order generally des- ignated in the literature as the Ornithischia.) Nopcsa defined his suborder Poposauroidea as follows: "Incompletely known; vertebrae biplane; anterior part of ilium expanded downward; bipedal." He recognized within this suborder one family having the same defini- tion as that for the larger taxonomic group. In Hay's (1930) bibliography of the fossil vertebrates, Poposaurus was classified as a theropod dinosaur and placed in the family Anchi- sauridae. Because of the nature of this publication, no reasons could be given for such an allocation of the genus. Oskar Kuhn (1937), in his classification of the fossil reptiles, fol- lowed the precedent set by Nopcsa ten years earlier by placing Popo- saurus in a family of its own, within a separate suborder which he called the Poposauria. For the suborder he gave a brief diagnosis which was essentially a paraphrased copy of Nopcsa's earlier diagnosis. In 1950 von Huene published a paper in which he considered the evidence for the presence of these dinosaurs in rocks of Triassic age. The paper is concerned primarily with tracks discovered in Argen- tina as indicative of the presence of ornithischians during Triassic times, but in passing he mentions Poposaurus as being one of the two genera of supposed ornithischians from the Triassic known from the bones. 1 In this paper he states that Poposaurus is "probably" a primitive stegosaurid, but he presents no evidence to uphold this opinion. 1 In addition to Poposaurus, the other supposed Triassic ornithischian dinosaur that has been described is Geranosaurus, a South African form. Of the two genera, Poposaurus is based upon much the best and most definitive materials. O-S' i G- e^o \ oo \i COLBERT: TRIASSIC REPTILE 61 Finally, within the past few years three important works have appeared, in which the taxonomic relationships of Poposaurus are considered. In the first of these, the monumental Traite de Pale- ontologie, edited by Jean Piveteau, the authors of the chapter on dinosaurs, namely Messrs. de Lapparent and Lavocat (1955, p. 795), indicate that Poposaurus is a primitive member of the superfamily Iguanodontoidea within the Ornithischia, but they actually do not assign it to a definite position among any of the iguanodont families that they list, perhaps because they are seemingly doubtful as to the value of the evidence. They do say that "Poposaurus du Trias in- f^rieur, [sic] animal bipede, paraitrait, en effet, avoir r^alis^ d^jk certaines structures du bassin des Avipelviens." Farther on in this same work (pp. 829-830) they make the following remarks: "Ceux [the omithopods] qu'on aurait signal^s dans le Trias (Huene, 1950) paraissent douteux d'apres les documents actuels. lis ne reposent en effet que sur les 614ments suivants: "1. Pieces du squelette: une petite mandibule, Geranosaurus Broom 1911, du Trias sup^rieur d'Afrique du Sud; ilion, f^mur et tibia, Poposaurus Mehl 1915, du Trias sup^rieur du Wyoming. "2. Empreintes de pas: Anomoepus Hitchcock, du Trias de la Connecticut Valley; Rigalites Huene 1931, du 'Rh^tien' d'Argentine. L'attribution de ces empreintes a un Ornithopode demeure possible, mais non certaine." In the second of the three recent publications, von Huene (1956, p. 556) seemed rather definite as to the taxonomic position of Popo- saurus, for he regarded it as a member of the family Stegosauridae. In making this assignment, von Huene was reiterating his earlier opinion, expressed in 1950. In the third of these recent works Romer (1956) presents a very complete and well-considered classification of reptiles down to genera. He does not discuss Poposaurus, but in his classification he includes it among the phytosaurs as a synonym of the genus Lophoprosopus. Whatever Poposaurus may be, it almost certainly is not a phytosaur. To summarize this discussion, the various opinions as to the rela- tionships of Poposaurus may be listed as follows: Mehl, 1915. "Resembles some of the early dinosaurs." Nopcsa, 1921. Primitive orthopod. Nopcsa, 1928. Order Orthopoda; Suborder Poposauroidea; Family Poposau- ridae. Hay, 1930. Order Saurischia; Suborder Theropoda; Family Anchisauridae. Kuhn, 1937. Order Ornithischia; Suborder Poposauria; Family Poposauridae. von Huene, 1950. Probably a primitive stegosaurid. 62 FIELDIANA: GEOLOGY, VOLUME 14 de Lapparent and Lavocat, 1955. "Ordre des Avipelviens; Sous-ordre des Orthopodes; Super-famille des Iguanodontoides (Ornithopodes)." von Huene, 1956. Order Ornithischia; Suborder Thyreophora; Family Stego- sauridae. Romer, 1956. Order Thecodontia; Suborder Parasuchia; Family Phytosau- ridae. In view of these differing opinions as to the relationships of Popo- saurus, it is hereby proposed to make a new study of the original materials, to see if a more satisfactory basis can be reached for the classification of this genus. For permission to do this I am greatly indebted to the authorities of Chicago Natural History Museum, and particularly to Drs. Rainer Zangerl and Everett C. Olson, and to Mr. David Techter. Grateful acknowledgments should also be made to Dr. Claude Hibbard of the University of Michigan, for per- mission to study materials in the collection of the University Museum of Paleontology. A NEW DESCRIPTION AND AN ANALYSIS OF POPOSAURUS GRACILIS MEHL As it now exists, the type of Poposaurus gracilis Mehl consists of the following materials: Chicago Natural History Museum, no. UR 357, two dorsal vertebrae, one caudal vertebra, a left ilium, the proximal portion of a left femur, a right femur, the distal portion of the left tibia, part of the shaft of the right tibia. In addition to these bones, Mehl mentions in his original descrip- tion four or possibly five sacral vertebrae, which unfortunately can not be found at the present time. Also, he speaks of two caudal vertebrae, but in the material now available there is only one cau- dal. Finally, he figures a tibia (labeled in his drawing as the fibula) consisting of two parts, of which at the present time only the distal section is among the type materials. To confuse the situation still further, there is associated with the type the acetabular portion of a second right ilium, bearing the num- ber UR 358. This is the ilium figured by Lees (1907) in the descrip- tion of Paleorhinus bransoni. The skull of Paleorhinus hransoni is numbered CNHM-UC 632. All of the materials registered under these two numbers, UR 357 and UR 358, presumably were found at a single locality along the Little Popo Agie River, near Lander, in Fremont County, Wyoming. The horizon is the Popo Agie member of the Chugwater formation, of late Triassic age. Fig. 30. An original figure of Poposaurns gracilis (from Mehl, 1915b). It is here maintained that (a), the fem\ir, the tibia and the- bone designated in the original description as a fibula should be identified as of the right side, and (b), the ilium and the proximal end of a femur should be assigned to the left side. About one-third natural size. ' — ' - 63 64 FIELDIANA: GEOLOGY, VOLUME 14 Fig. 31. An original figure of Poposaurus gracilis (from Mehl, 1915b). In this figure the spine of the caudal vertebra, as originally preserved, is shown. Slightly smaller than natural size. THE VERTEBRAE Mehl's original description of the vertebrae is very complete and accurate. Indeed, some parts of his description constitute the only evidence we have now as to certain features of the structure of the vertebrae in Poposaurus, because there has obviously been some breakage of the specimens since the time he described them, with a resultant loss of parts. For example, the posterior zygapophyses of the dorsal vertebra figured by Mehl have since disappeared, and much of the spine of the caudal that he figured is now gone. For easy reference it is here proposed to redescribe the vertebrae, as well COLBERT: TRIASSIC REPTILE 65 as the other extant parts of the skeleton, thus facilitating the com- parisons with which this paper is particularly concerned. The two dorsal vertebrae (fig. 32) are somewhat elongated, the single remaining caudal somewhat less so. In all of these vertebrae the faces of the centra are considerably deeper than they are broad, somewhat concave, and in all of them the bodies of the centra are greatly constricted in their mid-portion. This is particularly appar- ent in the two dorsal vertebrae, which thereby somewhat resemble thecodont vertebrae. It is likely that the effect is exaggerated by lateral crushing. The neural arch is rather high, and in the dorsal vertebrae the diapophysis is a thin, horizontal plate. According to Mehl (1915b, p. 518), the diapophysis "suddenly thickens below [at its posterior extremity] for the tubercular facet which is separated from the pos- terior zygapophysis by a deep, rounded sinus." Because of the breakage, alluded to above, this point cannot now be verified. The diapophysis is supported, as Mehl pointed out, by two thin laminae, obliquely directed, diverging from the middle portion of its ventral surface toward the anterior and posterior faces of the centrum. The posteriorly directed lamina ends before reaching the edge of the artic- ular face, but the anterior one runs into a large, thickened protuber- ance, which carries the capitular facet for the rib. The zygapophyses are close together, the anterior ones facing up and inwardly, the pos- terior ones obviously down and outwardly. The two anterior zyga- pophyses of the dorsal vertebra are separated by a narrow gap, which Mehl thought might have accommodated a thin, vertical bony sup- porting plate, beneath the posterior zygapophyses of the preceding vertebra. In this supposition Mehl was probably mistaken. The narrow gap between the anterior zygapophyses of this vertebra can be compared with the generally broad space between these two facets in most vertebrae. In Poposaurus the narrowness of the gap is largely artificial, the result of the transverse crushing to which the vertebra has been subjected. The caudal vertebra (figs. 31, 32) is somewhat heavier and shorter than the two dorsal vertebrae. It has a rather short, posteriorly placed diapophysis that projects upwardly and laterally from the side of the neural arch. This vertebra originally had a high spine, most of which is now missing. The lower edges of the articular faces on the centrum show facets for chevrons. The vertebrae of Poposaurus show various features that are seen in the vertebrae of primitive dinosaurs. The rather elongated, spool- 66 FIELDIANA: GEOLOGY, VOLUME 14 Fig. 32. The three type vertebrae of Poposaurus gracilis as now preserved; CNHM no. UR 357. A and B, two dorsal vertebrae; C, caudal vertebra. Lateral views, approximately two-thirds natural size. shaped centra, biconcave at the ends and with constricted mid- portions, are very much like the centra of the early theropods found generally in beds of late Triassic age. In Camptosaurus and Hypsilophodon, primitive ornithopods, the dorsal vertebrae are elongated and have concave central faces, but in the middle portions of the centra they show no such constrictions as are seenin Poposaurus and in the early theropods. Another char£(.cter in which the vertebrae of Poposaurus may be compared closely with those of the early theropods is the presence in the dorsal vertebrae of oblique laminae beneath the diapophysis. This feature, so prominent in the Popo Agie specimen, is quite pro- COLBERT: TRIASSIC REPTILE 67 nounced in all but the most posterior dorsals of the central theropod type, Coelophysis, and in the prosauropod, Plateosaurus. In contrast, no such laminae or buttresses are to be seen in the primitive ornitho- pods, Camptosaurus and Hypsilophodon, or in any of the other early ornithopods for that matter. Neither are they present in the phy- tosaurs nor generally in the pseudosuchian thecodonts. The single known caudal vertebra of Poposaurus (fig. 32), which is not as elongated from front to back as the dorsals, lacks the oblique laminae or buttresses that are so characteristic of the dorsal verte- brae. The centrum of this vertebra is heavy as compared to the centrum in the dorsals, and is less constricted. These are real differ- ences and are not to be attributed to differential crushing. In the light of the above comparisons it would appear that the vertebrae of Poposaurus, especially the dorsal vertebrae, show close affinities with comparable vertebrae in the primitive theropod dino- saurs. Whether this means that Poposaurus should be included among the theropods is not as definite as might be inferred from the anatomy of the vertebrae, because there is now good evidence (as yet unpublished) to indicate that in some of the late Triassic thecodont reptiles the vertebrae had already advanced very far toward the theropod condition, even though the limbs and girdles had retained characteristic thecodont features. The problem as to the relationships of Poposaurus must necessarily rest upon all of the known evidence, not only on that of the vertebrae. THE ILIUM The ilium of Poposaurus (fig. 33) is a very distinctive bone, and at first glance one would suppose that it might give an immediate clue as to the relationships of this Triassic reptilian genus. The trouble is that the ilium, as represented by the almost complete bone described by Mehl, and in addition by the partial specimen which was not included in the type description but which nevertheless would seem to have been found with or near the type, is rather dif- ferent from any other known reptilian ilium of Triassic age, or of later age for that matter. Therefore comparisons are difficult. It should be pointed out at the beginning that the interpretation of the ilium in Poposaurus has been complicated through the years by an initial misconception as to its position. Mehl described this bone as a right ilium, supposing the long iliac blade to be a preace- tabular portion. This in turn led subsequent authors to compare the blade of the ilium in Poposaurus with the elongated preacetabular 68 FIELDIANA: GEOLOGY, VOLUME 14 portion of the ilium characteristic of many ornithischian dinosaurs; indeed, the general tendency to place Poposaurus among the Orni- thischia may be attributed in large part to the identification of the ilium as belonging to the right side of the pelvis. It is here maintained that the ilium of Poposaurus belongs to the left side of the pelvis. According to this interpretation the long blade becomes postacetabular, homologous with the postacetabular part of the ilium as seen in various archosaurians, among them the theropod saurischians. The preacetabular part of the ilium, according to this view, is quite short, as is frequent in primitive archosaurians, includ- ing again the theropods. There are various reasons for thinking that the ilium of Popo- saurus comes from the left rather than the right side of the pelvis. For example, there is a definite rugosity along the medial surface of the long iliac blade, indicating an articulation with the sacrum. Such an attachment would be found only on a postacetabular iliac blade in the archosaurian reptiles; a preacetabular blade would be largely free, as indeed it is in the theropods, in which the preacetab- ular section is short, and as it is in the various ornithischians, in many of which this part of the pelvis is long. Again, of the two articulations on the dinosaurian ilium, one for the pubis and one for the ischium, the latter is commonly strongly convex, almost hemispherical in shape. In Poposaurus an articular surface of this type is beneath the long iliac blade, which is presumptive evidence that the blade is a posterior blade, and hence that this is the left ilium. Finally there is the evidence of the femur. The proximal portion of the femur that articulates with the type ilium is clearly that of a left one. Its form is somewhat obscured by crushing, but the oppo- site femur is in good shape and it is clearly a right femur not a left one as indicated by Mehl. This is shown by the configuration of the proximal portion of this latter bone, which corresponds in every way with the same section of a right femur of a phytosaur and of a thero- pod dinosaur. To return to the description, it may be helpful to consider the bone empirically, without reference to the ilia of other reptiles. It is an unusually elongated and rather low ilium, its anteroposterior measurement about 250 millimeters, its height approximately 80 millimeters. These proportions can be considered as approximately correct, for the bone appears to have suffered very little distortion during the process of fossilization. COLBERT: TRIASSIC REPTILE 69 That portion of the acetabulum contained within the ilium is large and deep, and the lower surface shows that it was broadly perforated. At the anterior and posterior extremities of the iliac portion of the acetabulum are the two articular facets, one between the ilium and the pubis in front, and one between the ilium and the ischium at the back. Both of these facets are strongly rugose, and the posterior one for the ischium is rather hemispherical in shape, as has been mentioned above. It is quite evident that the ilium was in contact with the other two bones of the pelvis only at these two points. From the thickened anteroventral section of the ilium which was in contact with the pubis, a shelf or ledge extends up and back, to form a broad roof over the upper portion of the acetabulum — an adaptation to receive the upward thrust of the femur. The strong development of this feature in the ilium indicates that Poposaurus was almost certainly a bipedal reptile in which the weight of the body was pivoted at the articulation between femur and pelvis. Midway between the front and the back the border of this dorsal shelf over the acetabulum is extended out and down to give it added width in this region. On its dorsal surface there is a large, expanded, rugose knob, very probably to accommodate the origin of the ilio-femoralis muscle, from which a strong ridge curves up and forward to the small preacetabular process of the ilium. The ridge is quite high, so that it forms a curved wall separating the forward part of the ilium from the much larger, posterior portion. There is a depression in the bone just in front of the ridge, so that the surface of the ilium above the anterior part of the acetabulum takes the form of a large and rather deep fossa. The structure of the supra-acetabular portion of the ilium in Poposaurus is not to be homologized in the least with the antitrochanter of the ornithischian dinosaurs, this latter feature be- ing a folding over of the superior iliac border. The postacetabular part of the ilium is much elongated. In its posterior part the surface is expanded by a downward growth of the bone in this region to form a sort of flat blade, and as a result the lower border of the ilium sweeps down in a decided curve, back of the acetabulum. The ilium terminates, however, in a sharp point. This postacetabular portion of the ilium carries on its surface a strong ridge that arises from the surface of the bone near the poste- rior border of the acetabulum and continues back in a flattened curve to the posterior part of the iliac blade. In its mid-portion this ridge is strong and rugose. Owing to the development of this ridge the 70 FIELDIANA: GEOLOGY, VOLUME 14 Fig. 33. Three ilia of Poposaurus gracilis. A, left ilium of type specimen, CNHM no. UR 357. B, fragment of right ilium from the Dockum beds of Texas, University of Michigan, Museum of Paleontology, no. 11748. C, fragment of large left ilium found near type specimen, CNHM no. UR 358. All external lateral views, approximately one-third natural size. upper surface of the iliac spine forms a flattened and moderately broad surface. What does this analysis and interpretation of the ilium of Popo- saurus show with regard to the relationships of the genus? In the first place, this is certainly an ilium with a perforate acetabulum, a short preacetabular portion, and a long posterior blade. Of the pos- sible Triassic orders that might be considered for the reception of the bone, the Protorosauria, the Thecodontia and the Therapsida may be immediately excluded. In none of these is the acetabulum per- COLBERT: TRIASSIC REPTILE 71 f orated; in all there is a conservative retention of the primitive rep- tilian structure of this portion of the pelvis. Moreover, in none of these orders is the posterior blade of the ilium produced into a long, low blade such as is characteristic of Poposaurus. This leaves the dinosaurs to be considered. If the identification of the ilium in question as coming from the left side be accepted, then this bone hardly can be attributed to the Ornithischia, in all mem- bers of which order there is considerable elongation of the iliac blade anteriorly. As for the Saurischia, the ilium is deep in the sauropods. Only in the primitive theropods is the ilium frequently long and shallow, with the preacetabular portion relatively short and the post- acetabular portion elongated as a blade-like structure. Thus, on the basis of general resemblances, the ilium in Poposaurus would seem to have various features in common with the same bone in some of the early theropod dinosaurs. Even here the comparisons are not particularly close. In the early coelurosaurian theropods, such as Procompsognathus or Co- elophysis, the ilium, though long and relatively low, is nevertheless expanded in its anterior portion much more than is the case in Popo- saurus. Indeed, the only close comparison with Poposaurus in this respect is to be seen in the strange pelvis from the Dockum forma- tion of Texas, described by Case as the pelvis of a phytosaur (which it almost certainly is not). In this pelvis the preacetabular portion of the iliac blade is in the form of a very short, anteriorly projecting spine, as was seemingly the case in Poposaurus. Otherwise the ilium of this pelvis is in most respects quite different from the same bone in Poposaurus. The presence of the strong, forwardly curving ridge extending from the mid-portion of the acetabulum to the anterior tip of the iliac blade, the long, posterior iliac blade, expanded in its posterior portion and terminating in a point, and the development on this posterior iliac blade of a dorsal, flattened shelf, are all characters that cannot be closely matched among the ilia of any other thero- pods, or of any other known Triassic or later Mesozoic reptiles, for that matter. In certain respects the configuration of the back portion of the iliac blade in Poposaurus shows some resemblances to the same region in Coelophysis, but comparisons are only of the most general nature. The one close comparison between the ilium of Poposaurus and this bone in other early theropods is in the form of the acetabulum itself. Here, in the configuration of the lower acetabular Jborder, in 72 FEILDIANA: GEOLOGY, VOLUME 14 the shape of the pubic and ischial facets, and in the lateral extension of the ilium to form a deep socket for the reception of the head of the femur, the resemblances to the primitive theropods, such as Coelophy- sis, are notable. On the basis of these considerations it would seem that perhaps the most valid interpretation of the ilium in Poposaurus is to regard it as a specialized theropod type, derived from a very early theropod in which the preacetabular portion of the iliac blade had not ad- vanced appreciably beyond the thecodont condition. The acetabulum itself is rather characteristically theropod in form. In the postace- tabular portion of the pelvis there has been strong specialization, with the development of the long, blade-like structure of rather com- plex form, very possibly in part as an adaptation for the origin of strong ilio-fibularis and ilio-tibialis muscles. THE FEMUR If the proximal and distal ends of one femur were truly associated, as Mehl indicates in his figure of Poposaurus, then this reptile had a rather long straight femur. The probabilities are strongly in favor of such an association, but there is always the faint possibility that there has been a mixture of bones in this Popo Agie material — a pos- sibility that must be kept in mind. It is here assumed, however, that the materials all belong to a single animal, and the description will be made upon the basis of this assumption. As mentioned before, Mehl indicated the proximal fragment asso- ciated with the ilium as from the right side, and the two pieces com- prising an essentially complete femur as from the left side. And as also mentioned above, this is almost certainly wrong; the relation- ships of right and left should be reversed. The head and the proximal end of the femur in Poposaurus are rather unspecialized (fig. 34). The head, although directed medially, is not set off from the shaft of the bone by a decided neck; rather, there is a gradual curve from the shaft to the head of the femur, much as is the case in the thecodont archosaurians. Moreover, there is no well-defined great trochanter at the proximal end of the femur in this reptile. Consequently, the upper end of the femur has a de- cided thecodont appearance, being not unlike the same portion of the phytosaurian femur in this respect. On the other hand, the middle and lower portions of the femur are definitely specialized. The lower segment of the right femur as COLBERT: TRIASSIC REPTILE 78 preserved in the type indicates that in this reptile the shaft of the femur was fairly straight. Perhaps this relatively straight shaft in- fluenced some students favorably toward the idea that Poposaurus shows ornithischian characters, for it must be admitted that there is a certain resemblance to the hadrosaurian dinosaurs in this respect. This resemblance to the ornithischian dinosaurs, and particularly to the hadrosaurs, is strengthened by the shape of the distal condyles of the femur in Poposaurus, for in this reptile from the Triassic of Wyoming the posterior curve of the internal condyle swings up to ter- minate in a sort of "hook." A hook-shaped condyle like this is seen in many ornithischians, particularly the hadrosaurs, but it should be noted that in these dinosaurs the condyle having this shape is the outer one. The inner condyle in the hadrosaurs does not show such extreme curvature of its back portion. By way of contrast, some of the saurischians, particularly the more advanced camosaurs, show a somewhat hook-shaped inner condyle of the femur. Another interesting feature of the femur of Poposaurus is the presence of what appears to be a trochanter on the inner side of the shaft and in about its mid-portion. If this feature is real, and not the result of crushing, it is almost surely a fourth trochanter, com- parable to the fourth trochanter as developed on the shaft of the femur in various theropods, but in them generally somewhat higher on the bone than it is in Poposaurus. Also, this process may be com- pared with the fourth trochanter in the ornithischian dinosaurs. Generally speaking, the trochanter is larger and more prominent in the ornithischians than it is in the theropods, and than seems to be the case in Poposaurus. Moreover, the fourth trochanter in the bi- pedal ornithischians is commonly "pendent," with a downwardly directed point which serves as an origin for a tendon to the lower leg. What is the evidence of the femur as to the relationships of Popo- saurus? Although the evidence is somewhat mixed, showing a prim- itive proximal region combined with various specializations in the rest of the bone, on the whole it indicates relationships with the theropods. BONES OF THE LOWER LEG Associated with the right femur of Poposaurus are two bones which Mehl designated as the left tibia and fibula, respectively. The end of the so-called fibula nearest to the right femur has since disappeared, but the remainder of the bone, the major portion, is still among the type materials. The bone that Mehl labeled the tibia is only a portion of the shaft. 74 FIELDIANA: GEOLOGY, VOLUME 14 Both of these bones are large. It seems quite probable that the bone designated as the tibia by Mehl is the right tibia, associated with the femur. The other bone is almost certainly too robust to have been the fibula, as labeled by Mehl. More probably this is the other tibia, perhaps, as indicated in Mehl's figure, with its proximal end away from the femur and its rounded distal end next to the femur. The end of the large section now at hand is badly crushed and flattened, so nothing definitive can be said with regard to its characters. It should be noted that this bone, as originally figured by Mehl, would seem to have been appreciably shorter than the femur. Because of the crushed and fragmentary nature of these two bones they are of little value in helping to determine the relationships of Poposaurus. Measurements of Poposaurus gracilis Mehl (Type, CNHM-UR 357) MM. Dorsal vertebra, length of centrum 57 Dorsal vertebra, ant. articular surface, transverse 32 Dorsal vertebra, ant. articular surface, vertical 40 Caudal vertebra, length of centrum 50 Caudal vertebra, post, articular surface, transverse 38 Caudal vertebra, post, articular surface, vertical 48 Ilium, length 251 Ilium, height 80 Femur, approximate length 480 CONCLUSIONS A careful assessment of the bones that comprise the type of Popo- saurus gracilis shows that this Triassic reptile combines various ana- tomical characters that, taken together, make it a particularly baffling problem in taxonomy. In some characters, such as the features of the dorsal vertebrae, of the acetabular region, and to a large degree of the femur, Poposaurus shows definite theropod resemblances. In others, such as the general aspects of the ilium, Poposaurus is quite unlike any other known reptile. Other characters, such as the fea- tures of the lower leg bones, are indeterminate. After consideration of the various characters that are involved, it is here suggested that Poposaurus is a theropod dinosaur. As to its larger relationships within the suborder, the known char- acters would seem to indicate that Poposaurus is a carnosaur. It is moderately large for a Triassic reptile and the bones are thick-walled, as one would expect in the carnosaurs. The form of the vertebrae is ^ Fig. 34. Type limb bones of Poposaurus gracilis. CNHM no. UR 357. A, proximal end of left femur, external lateral view. B, proximal and distal sections of right femur, internal lateral view. C, distal section of right tibia. All approxi- mately one-third natural size. 75 76 FIELDIANA: GEOLOGY, VOLUME 14 what might be expected in a primitive carnosaur. The iHum seem- ingly has a small anterior process, which would be characteristic for a primitive carnosaur but not for a coelurosaur. There is a reasonably large acetabular foramen. The straight femur, longer than the tibia, is a carnosaur feature that is in contrast to the coelurosaurs. Where then should the genus be placed within the carnosaurs? Its relationships are probably closest to such Triassic carnosaurs as the Paleosauridae and the Zanclodontidae, but the form of the ilium is so unlike that of any known archosaurian ilium, and the ilium is such a basic feature in the interpretation of these reptiles, that the assignment of Poposaurus to a separate family is here suggested as the best solution to the problem under the present circumstances. Of course Poposaurus has already been placed within its own family, the Poposauridae, by Nopcsa, but as a member of the Orni- thischia, not the Saurischia. It is here proposed that the family, as based upon its single type genus, be lifted bodily from the one dinosaurian order and be transferred to the other. Thus its taxo- nomic relationships might be expressed as follows: Order Saurischia Suborder Theropoda Infraorder Carnosauria Family Poposauridae Genus Poposaurus Species Poposaurus gracilis Mehl By removing Poposaurus from the Ornithischia to the Saurischia, the evidence for ornithischian dinosaurs in the Triassic period, al- ready weak, becomes even weaker. Only Geranosaurus remains as a supposed Triassic ornithischian known from bones, and as a speci- men Geranosaurus is even more unsatisfactory than Poposaurus. The only other evidence for ornithischians within the Triassic is that of the footprints — and how reliable is such evidence? THE GEOLOGIC OCCURRENCE OF POPOSAURUS It is stated above that more materials are needed to put our knowledge of Poposaurus on a firmer basis than that on which it now rests. If such materials are to be found, it seems likely that they may be discovered in the Dockum beds of Texas as readily or per- haps even more readily than in the Popo Agie sediments of Wyoming. COLBERT: TRIASSIC REPTILE 77 In the collections of the University of Michigan there is a portion of a right ilium of Poposaurus (fig. 33, B) that was found in the Dockum formation near Otis Chalk, in Howard County, Texas. As may be seen from the figure it is very closely comparable to the type ilium, and it certainly indicates an identity of genera if not of spe- cies between the Triassic sediments of Wyoming and Texas. When one considers how much more productive the Dockum has been in past years than the Popo Agie, it seems likely that at some time in the future the Texas Triassic may yield the relatively complete skele- ton of Poposaurus that is so much to be desired. Certainly this fossil from Texas indicates a close correlation be- tween that portion of the Dockum in which it was found and the Popo Agie. Dr. J. T. Gregory, on the basis of his detailed study of the phytosaurs, feels that the Popo Agie, in which there occur the relatively primitive phytosaurs, Angistorhinus and Palaeorhinus, is somewhat earlier in age than most other North American Triassic horizons such as the Chinle or the Newark series. Angistorhinus is also found in the lower part of the Dockum, and the presence of Poposaurus in the Texas sediments broadens the evidence as to the relatively early position within the upper Triassic sequence of this particular horizon. REFERENCES Case, Ermine Cowles 1922. New reptiles and stegocephalians from the upper Triassic of western Texas. Publ. Carnegie Inst. Washington, no. 321, pp. 1-84, figs. 1-33, pis. 1-14. 1943. A new form of phytosaur pelvis. Amer. Jour. Sci., 241 , pp. 201-203, 1 fig. Colbert, Edwin Harris, and Gregory, Joseph T. 1957. Correlation of continental Triassic sediments by vertebrate fossils. Bull. Geol. Soc. Amer., 68, pp. 1456-1467. Hay, Oliver Perry 1929. Second bibliography and catalogue of the fossil Vertebra ta of North America, vol. I. Publ. Carnegie Inst. Washington, no. 390, viii+916 pp. 1930. Second bibliography and catalogue of the fossil Vertebrata of North America, vol. II. Ibid., xiv-t-1074 pp. Huene, Friedrich R. von 1906. tJber die Dinosaurier der aussereuropaischen Trias. Geol. Paleont. Abhandl., Jena, Neue Folge, 8, Heft 2, pp. 1-60, 102 figs., pis. 1-16. 1926. Notes on the age of the continental Triassic beds in North America, with remarks on some fossil vertebrates. Proc. U. S. Nat. Mus., 69, art. 18, pp. 1-10, 6 figs. 1932. Die Reptil-Ordnung Saurischia, ihre Entwicklung und Geschichte. Monogr. Geol. Paleont., (1), 4, parts 1 and 2, viii+361 pp., 41 figs., 56 pis. 78 FIELDIANA: GEOLOGY, VOLUME 14 1939. Zur Deutung der Reptilreste vom Popo Agie River. Zentralbl. Min. Geol. Paleont., Abt. B, 1939, pp. 397-399, 2 figs. 1950. Die Entstehung der Ornithischia schon friih in der Trias. Neues Jahrb. Geol. Palaont., Monatsh., Abt. B, part 2, pp. 53-58. 1956. Palaontologie und Phylogenie der Niederen Tetrapoden. Jena, Gustav Fischer, xii+716 pp. KUHN, OSKAR 1937. Die fossilen Reptilen. Berlin, Gebruder Borntraeger, viii+121 pp., 92 figs. Lapparent, Albert F. de, and Lavocat, RENfi 1955. Dinosauriens. In Piveteau, Jean, Traite de Paleontologie. Paris, Masson et Cie, tome 5, Amphibiens, Reptiles, Oiseaux, pp. 785-962, figs. 1-156. Lees, J. H. 1907. The skull of Paleorhinus, a Wyoming phytosaur. Jour. Geol., 15, pp. 121-151, figs. 1-8. Lull, Richard Swann 1953. Triassic life of the Connecticut Valley. Revised. State of Connecticut, State Geol. Nat. Hist. Surv., Bull. no. 81, pp. 1-331, 168 figs., 12 pis. Mehl, M. G. 1915a. The Phytosauria of the Trias. Jour. Geol., 23, pp. 129-165, figs. 1-20. 1915b. Poposaurus gracilis, a new reptile from the Triassic of Wyoming. Jour. Geol., 23, pp. 516-522, figs. 1-2. NOPCSA, F. VON 1921. Zur systematischen Stellung von Poposaurus (Mehl). Zentralbl. Min. Geol. Paleont., p. 348. 1928. The genera of reptiles. Palaeobiologica, 1, pp. 163-188. Piveteau, Jean 1955. See Lapparent, Albert F. de, and Lavocat, RenS RoMER, Alfred Sherwood 1956. Osteology of the reptiles. University of Chicago Press, xxi+772 pp., 248 figs. m" W