en CM Z LIBRARY OF THE UNIVERSITY OF ILLINOIS AT URBANA-CH/^PAIGN 550.5 n GEOLOGY UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPA/GM GEOLOGY u^a, FIELDIANA: GEOLOGY A Continuation of the GEOLOGICAL SERIES of FIELD MUSEUM OF NATURAL HISTORY VOLUME 15 FIELD MUSEUM OF NATURAL HISTORY CHICAGO, U.S.A. TABLE OF CONTENTS PAGE 1. The Non-Therapsid Reptiles of the Lufeng Basin, Yunnan, China. By David Jay Simmons 1 2. The Mammalian Genera Arctoryctes and Cryptoryctes from the Oligo- cene and Miocene of North America. By Charles A. Reed and William D. Turnbull 95 3. Catalog of the Collection of Meteorites in Chicago Natural History Museum. By Henry Horback and Edward J. Olsen 167 NOV 10 1965 LIBRAKY THE NON-THERAPSID REPTILES OF THE LUFENG BASIN, YUNNAN, CHINA DAVID JAY SIMMONS FIELDIANA: GEOLOGY VOLUME 15, NUMBER 1 Published by CHICAGO NATURAL HISTORY MUSEUM APRIL 29, 1965 GEOLOGY LIBRARY THE NON-THERAPSID REPTILES OF THE LUFENG BASIN, YUNNAN, CHINA DAVID JAY SIMMONS Department of Palaeozoology University of Chicago FIELDIANA: GEOLOGY VOLUME 15, NUMBER 1 Published by CHICAGO NATURAL HISTORY MUSEUM APRIL 29, 1965 Edited by Edward G. Nash Library of Congress Catalog Card Number: 65-19776 PRINTED IN THE UNITED STATES OF AMERICA BY CHICAGO NATURAL HISTORY MUSEUM PRESS CONTENTS PAGE Introduction 5 Geology of the Lufeng Basin 6 Preservation of Bone 11 Systematic Consideration 13 Order Thecodonta Suborder Pseudosuchia Dibothrosuchus elaphros gen. et sp. nov 13 Strigosuchus licinus gen. et sp. nov 31 Platyognathus hsui Young 34 Pseudosuchian 47 Order Crocodilia Suborder Protosuchia Microchampsa scutata Young 49 Order Saurischia Infraorder Coelurosauria Lukousaurus yini Young 54 Remains referrable to the Podokesauridae 55 Suborder Prosauropoda Gyposaurus sinensis Young 60 Lugengosaurus hueni Young 61 Lufengosaurus magni Young 62 Yunnanosaurus sp. Young 63 Infraorder Carnosauria Sinosaurus triassicus Young 64 Order Ornithischia Suborder Ornithopoda Tatisaurus oehleri sp. nov 65 Stratigraphic and Paleoecologic Considerations 73 zoogeographical and paleogeographic considerations 79 References 89 The Nori'Therapsid Reptiles of the Lufeng Basin, Yunnan, China INTRODUCTION The geological investigations of Bien (1941) some twenty years ago demonstrated that the sediments of the Lufeng Basin, Yunnan, China, held a richly varied vertebrate fauna. We are indebted to C. C. Young for our detailed knowledge of the fossil materials. The early taxonomic work progressed so rapidly that by 1940, the prin- cipal members of the assemblage were segregated and described briefly in several short preliminary notes (Young, 1939a, 1940b). Seven papers treating the individuals in detail were published in as many years (1941-1948), and in 1951 the fauna was assessed in mono- graphic fashion. The collection to be described was made in 1948 and 1949 by an expedition of the Fu Jen Catholic University of Peking led by Fr. E. Oehler, S.V.D. A re-examination of the fauna was deemed justifi- able on several counts. First, the field work conducted by Young continued sporadically for two seasons during the early war years; lack of funds, transportation and equipment provided such difficult circumstances that the collections were largely cursory and unsys- tematically procured. Second, new outcrops from which fossils might be obtained were certain to have been exposed in the years inter- vening between the early and late 1940's. Young's sample is domi- nated by large saurischians but the scanty and very imperfect remains of smaller forms were reported. The fragment of a jaw symphysis tentatively diagnosed as a small pseudosuchian was of particular interest. The fossils were lent from the S.V.D. Catholic Universities in 1957 by Fr. H. W. Rigney, S.V.D. The collection is known as the Fossil Collection of the S.V.D. Catholic University of Peking and will bear the provisional designation of CUP in the following report. I am indebted to Dr. E. C. Olson through whose efforts the ma- terials became available for study, and under whose guidance the 6 FIELDIANA: GEOLOGY, VOLUME 15 work was completed. I also wish to thank Dr. C. C. Young of the Institute of Vertebrate Paleontology, Peking; Dr. C. C. Mook and Dr. B. Brown of the American Museum of Natural History; Dr. R. Zangerl of Chicago Natural History Museum; Dr. A. S. Romer and Dr. B. Patterson of Harvard University, for their useful com- ments. Dr. Young kindly provided supplementary drawings used in the taxonomic determinations of many specimens. The bibli- ographic aid provided by Mrs. Rachel Nichols of the American Museum is gratefully acknowledged. The drawings were prepared by Mrs. Julia Child. GEOLOGY OF THE LUFENG BASIN The geology of the Lufeng Basin was described in a paper by Bien (1941) ; he arranged the list of vertebrate fossils from Young's publication (1939a) in stratigraphic sequence. The stratigraphy of the Basin was again reviewed generally by Young in his summary paper (1951), but the annotated list of fossils was incomplete rela- tive to the newly determined materials. Lufeng is situated about 60 km. northwest of Kunming, the capi- tal city of Yunnan Province. The tectonic history of Yunnan shows that the Lufeng sediments were deposited in a structural basin formed by the subsidence of the Central Yunnan Swell (Central Yunnan Block) which had oscillated throughout the Paleozoic and had last been revived in the Upper Permian. The structure is clearly recog- nizable since late middle Sinian time; Misch (1945) states that the structure underwent strong subsidence in the Triassic due to addi- tional sedimentation. Chang (1947) constructed a paleogeographic map (fig. 1) based upon the distribution of marine fossils which de- lineated the margins of the raised Swell and demonstrated that the Lufeng region was the site of terrestrial sedimentation. The uplift of the Swell in Permian times apparently served as a barrier to the transgressing Triassic sea that entered Yunnan from the south along the Sino-Burmese geosyncline so that it diverged into east and west waterways which continued in a northerly direction. During the Triassic, then, the basin received only sediments of continental origin typical of red bed development and flood plain environments. Inter- fingering of marine and continental deposits occurs along the borders of the Swell. In Lufeng, the Lufeng Series has an aggregate thickness of about 1000 meters divided into a Lower Series of alternating beds of dull purplish to dark red clays, siltstones, shales and sandstones, and into SIMMONS: NON-THERAPSID REPTILES 7 an Upper (Shihmen) Series of greenish-yellow banded bed sandy shales passing to variegated and wine red beds toward the top. The wine red beds contain calcareous and evaporitic matter. A thick green sandstone divides the red beds of the fossiliferous Lower Lufeng Series. The entire column is capped by a sandstone and rests on a basal breccia overlying the Sinian metamorphic rocks. PRELIMINARY MAP of YUNNAN I I Probable Distribution of Land ) during ESS! Probable Distribution of Sea ) t[me '°\ Fig. 1. A paleogeographic map of Yunnan Province showing the relative distribution of land and sea during Triassic time (redrawn from Chang, 1947). The Lufeng Series in toto is conformable with the Ipinglang Coal Series below and with the Shihmen sandstone above (Misch, 1945). Bien (1941) regarded the Shihmen as disconformable with the Lufeng Series in Lufeng, but Misch held that the sandstone is "one horizon in the upper part of the Upper Red Beds" of Yunnan and not a sep- arable unit. The apparent local discordance indicates that deposition of the Upper Lufeng Series was interrupted prior to the introduction of sand. The Shihmen sandstone is current cross-bedded and passes from a basal lenticular pebble to massive boulder conglomerate with shaley intercalations toward the top. The Upper Lufeng Series is also more highly variegated than most of the "Red Beds of Yun- 8 FIELDIANA: GEOLOGY, VOLUME 15 nan" and the calcareous clays and sandstones (with limestone bands) are apparently of a finer texture in Lufeng. The thick Lower Lufeng Series is characterized by saurian re- mains. The thinner Upper Series contains fish {Hybodus, Lepidotus), an indeterminate chelonian, fresh-water ostracods and minute pe- lecypods and gastropods. The abrupt shift in the pattern of sedi- mentation and faunas perhaps indicates that conditions favored the establishment of shallow water environments, small lakes or pools fed by streams bearing dissolved calcium salts that produced at times the strongly indurated limestone bands of the upper-middle part. The algal deposition of carbonate and evaporitic matter is also con- ceivable. The fossil content and high degree of variegation is attrib- utable to either a facies change, and /or to a change in the source area, or to the inclusion of organic materials resulting in the partial re- duction of iron. The second of these is indicated by the apparent regularity of the bedding and by the reported absence of plant remains to serve as reducing agents. The fauna indicates a facies shift; the beds pass upward conformably to unfossiliferous strata of the same general description as the bone-bearing beds of the Lower Lufeng Series. Bien (1941) assigned an Upper Triassic age to the whole of the Lower Lufeng Series where saurian remains are found and called the Shihmen Jura-Cretaceous. Misch correlated the very thick Upper Red Beds of Yunnan, of which the Lufeng beds were considered a part, to the red beds of the neighboring province of Szechuan and suggested that they comprised "much of the Jurassic if they do not even in some parts extend beyond it." Chang (1947) notes that de- finitive Rhaetic beds have not yet been discovered in Lufeng. Misch categorized the Lufeng sediments as T4, purple marly shale (from a section at Taoyuan, about 8 km. north of Kunming). Since beds of this description are only characteristic of the Upper Lufeng Series, the idealized traverse is less than satisfactory. The more massive Lower Lufeng Series, perhaps equal to T4 (purple sandy shale inter- bedded with thin bedded sandstone) or T6 (red shale), should have been included with the middle upper Triassic Noric Beds which ap- parently embrace the Ipinglang Coal Series. Young's taxonomic studies on the saurian elements indicate that the Lower Lufeng Series had to have been deposited not earlier than Noric or post-Noric time. The stratigraphic column and a map of the Lufeng Basin (figs. 2, 3) showing the leading collecting localities, have been redrawn from Bien (1941). The vertebrate remains have been gleaned from four SIMMONS: NON-THERAPSID REPTILES ~ fl LITHOLOGY NATURE of CONTACTS AGE SHIHMEN ..... a z uJt/} -"S ' a. a. zs SJ C9 Z 3 * o wo* Capping Escarpment Sandstone DISC ON POSTLIAS SERIES ' 1-"- •0* Wine Red Beds FOHMITY « tK>* Variegated Beds Greenish ■ Yellow Banded Beds L _« _« _ »_» 140* Dark Red Beds fOKMlfT UNCOMFOK GO < OE r- ae UJ Q. O- Massive Green Sandstone 10* " Dull Purplish Beds LUFENG J . Tati I . . * £-_T"- -Z Z 550* Dull Purplish Beds / Tactions - Dull Purplish Beds Basal Brecia no META- MORPHIC-^ SERIES i«§ Metamorphic Series MITY PRE- MBRIAN Fig. 2. Stratigraphic section of the Lufeng Series showing the principal collecting localities, Tachung, Hei Koa Peng and Ta Ti. (redrawn and modified from Young, 1951). rather narrow horizons, Fossiliferous Beds 1-4, ordered stratigraph- ically from the lowest to the highest levels. The saurian elements are known from the first and second zones traversing the Basin in a north-northwest direction. The third and fourth zones are less ex- tensively developed and are concentrated in the variegated beds of the Upper Series; these have been correlated across to the sediments of the nearby Anning Basin by the common occurrence of Hybodus. Stratigraphic Section of the Lufeng Basin Top (Young, 1951) Upper Lufeng Series 8. Wine Red Beds (30 meters) Alternating bright wine red clays, shales and sandstones with occasional purplish earthy limestone. Rock salt, brine and gypsum occurs at Yuanyungching, North of Ipinglang. 7. Variegated Beds (120 meters) Upper two-thirds — Limey beds of purple, mauve, greenish and pink sediments. Lower one-third — Purplish limey beds with honeycomb weathering. Lowest — Greenish mottled limey beds with fresh water unioids and ostracods, Hybodus, Lepidotus and an indeterminate Chelonian. 10 FIELDIANA: GEOLOGY, VOLUME 15 6. Green-yellow banded beds (70 meters) Dark red shales and ripple marked siltstones at the base passing upwards to regular greenish yellow shale bands. 5. Red Sandstone (5-10 meters) DlSCONFORMITY Lower Lufeng Series 4. Dark Red Beds (150 meters) Alternating beds of clays, shale, siltstone and sandstone with two sub- ordinate layers of dull purplish elements. Fossils occur 70M above Bed No. 3. 3. Massive Green Sandstone (20 meters) Sediments contain occasional angular fragments of rocks, a breccia at Yaochan. Quartzite and quartz derived from Sinian rocks. 2. Dull Purplish Beds (550 meters) Alternating clays, shales, silstones and sandstone. Green sandstone bands occur near the base. Calcareous nodular concretions are com- mon as knobs or tubular masses. The beds are penetrated by calcite Tachung-Shawan Area QUATERNARY | Qal | ALLUVIUM. Puo- pleistocene 1 I NP I I GRAVEL.SANDS&CLAY JURA- CRETACEOUS DULL PURPLISH- BROWN CROSS BEDDED SANDSTONES. ?Jk's. UPPER -TRIASSIC WINE-RED BEDS. VARIEGATED BEDS. GREENISH YELLOW BANDED BEDS. | DARKREDBEDS. fiuT MASSIVE GREEN SANDSTONE. I I DULL PURPLISH r H BEDS. I 1 BASAL BRECCIA. SINIAN | H | SUBMETAMORPHIC ROCK. GEOLOGIC BOUNDARY NORMAL FAULT. EXPOSED & CONCEALED WITH AMOUNT OF DIP Fig. 3. Map of the Lufeng Basin showing the distribution of the fossiliferous horizons and principle collecting localities (redrawn and slightly modified from Young, 1951). SIMMONS: NON-THERAPSID REPTLIES 11 streaks, veins or mottlings. Fossils occur 300M below Bed 3 and 200M above Bed 1. 1. Breccia (5-20 meters) Phyllite derived from the underlying rocks. Yaochan sediments con- tain a quartzite and quartz brecchia. PRESERVATION OF BONE With few exceptions, the bones occur as weathered -out nodular materials invested by a heavy scale of matrix and scattered about on the surface of the ground. The rocks are so heavily indurated with iron compounds that quarrying was difficult and articulated skeletons were obtainable in only a few instances. Some remains are suggestive of coprolitic material. The field notes did not define the precise area of each collecting site or the relative concentration and association of bone in every case. Many individual fragments bearing the same field number were either referred to several genera, or these contained specimens of the same species but of different ontogenetic development. The proper identification of other disarticulated but otherwise complete bones was difficult because of the unavailability of comparative materials and the morphologic similarity of the known Lufeng prosauropods and carnosaur. For these reasons, a number of specimens could not be treated effectively and were not considered in this paper. Table 1 lists the fauna recognized by Young and also that newly determined by this study. It will be noted that Oehler's collection sampled most of the previously recognized non-Therapsids (?para- suchian) and yielded many heretofore unrecognized forms. The new finds are principally pseudosuchians and coelurosaurs. Their dis- covery in effect, fulfills Young's first charge to future collectors (1951, p. 92) that "special attention should be paid to the less known forms such as parasuchians, pseudosuchians, possibly protosuchians. ..." Table 1.— FAUNAL LIST OF THE NON-THERAPSIDS OF THE LOWER LUFENG SERIES Young (1951) New Material thecodonts Pseudosuchia Platyognathus hsui Young: Platyognathus hsui Young: anterior part of the jaw symphysis. a nearly complete skeleton. CUP 2083 12 FIELDIANA: GEOLOGY, VOLUME 15 Young (1951) Parasuchia Pachysuchus imperfectus Young: jaws. Protosuchia Microchampsa scutata Young: vertebrae and dorsal armor. New Material Strigosuchus licinus: jaw. CUP 2082 Dibothrosuchus elaphros: a nearly complete skeleton. CUP 2081 ?Pseudosuchian: a fragment of an ilium and femur. CUP 2084 Microchampsa scutata Young: vertebrae and dorsal armor. CUP 2085, 2086 SAUROPODS Coelurosaurs Lukousaurus yini Young: anterior half of a skull and jaws; several limb fragments. Carnosaurs Sinosaurus triassicus Young: a nearly complete skeleton. Prosauropods Lufengosaurus hueni Young: several nearly complete skeletons. Lufengosaurus magnus Young: jaws and postcranial elements. Yunnanosaurus huangi Young: several nearly complete skeletons. Yunnanosaurus robustus Young: a nearly complete skeleton. Gyposaurus sinensis Young: many nearly complete skeletons. ORNITHISCHIA Lukousaurus yini Young: anterior half of a skull and jaws. CUP 2092 ?Coelurosaur: distal end of a femur and metacar- pals. CUP 2089 ?Coelurosaur: distal end of a femur. CUP 2091 ?Coelurosaur: metatarsals. CUP 2090 Sinosaurus triassicus Young: jaws, teeth and postcranial elements. CUP 2097 Lufengosaurus hueni Young: jaws and postcranial bones. Lufengosaurus magnus Young: jaws and postcranial elements. Yunnanosaurus huangi Young: a juvenile skull and postcranial ele- ments. Yunnanosaurus robustus Young: postcranial elements. Gyposaurus sinensis Young: skull and jaw fragments and post- cranial elements. Tatisaurus oehleri: a left jaw. CUP 2088 SIMMONS: NON-THERAPSID REPTILES 13 SYSTEMATIC CONSIDERATION OF THE NON-THERAPSID REPTILES OF THE LUFENG FAUNA Order Thecodonta Suborder Pseudosuchia Family Ornithosuchidae Genus Dibothrosuchus, gen. nov. Diagnosis.- — Skull elongate and extensively excavated by a large triangular preorbital fenestra, a round orbit and a long infratemporal vacuity; roofing bones depressed below the superior margin of the orbit. Lower jaw slender, possessing a long narrow lateral fenestra. Teeth laterally compressed, finely denticulated and somewhat re- curved. Neck short, cervical vertebrae not elongate; their centra sharply compressed and keeled, and the anterior zygapophyses very steep. The dorsal vertebrae less sharply compressed backward. Three sa- cral vertebrae possessing broadly triangular ribs. The presacral series show costal articulations oriented in an upward and backward direc- tion. The first seven caudals are in natural connection with the sacrum and are not especially elongate. The appendicular skeleton, consisting of the forelimb and several metacarpals, is slender and the bones are thin-walled and hollow. Dibothrosuchus elaphros, sp. nov. Holotype. — A partially complete skull and jaws, a series of articu- lated vertebrae from the cervical, dorsal, sacral and caudal regions of the column, and a fragment of the forelimb and hand. The bones are of a reddish color encrusted with a thick greenish matrix. Cata- log no. CUP 2081. Horizon and locality. — Ta Ti, Dark Red Beds, Lower Lufeng Series. Diagnosis. — Same as for genus. DESCRIPTION Skull (fig. 4,A): The skull is lightly built and broadly fenes- trated; the roofing elements are thin and crushed, and the paired bones overlap. The ventral surface bears grooves interpreted as im- pressions of the nasal tracts and cortical regions of the brain; the 14 FIELDIANA: GEOLOGY, VOLUME 15 nasal, prefrontal, and frontal bones must be represented in part but sutural contacts are obliterated. The supraorbital contours of the frontals and prefrontals form elevated ridges which pass forward and become confluent with the nasals above the preorbital fenestra. The median suture is slightly depressed and the frontal is plicated by an additional pair of furrows mediad to the supraorbital ridges. The supraorbital eminences project downward to sheath the dorsal orbits externally so that the eye muscles reside in deep pockets bordered in- ternally by the ventrolateral furrows delimiting the brain structures. Premaxilla (fig. 4,B2): The extreme tip of the snout is lacking and the relationship of the premaxilla to the external nares and maxil- lary is uncertain. The dentition indicates that the bone is present but probably exclusive of its suture with the nasal. The nasal tracts are seen on the median side by a concavity bounded from below by a palatine fragment; the external nares, then, lay laterally on the snout at the same level as or slightly lower than the antorbital fenestra. Maxilla (fig. 4,B2): The maxilla projects dorsally about the anterior corner of the triangular preorbital fenestra and terminates posteriorly under the anterior orbit. The bone is gradually replaced by the jugal and its inferior border is down turned. Jugal (fig. 4,B2): The jugal is a triradiate bone. It replaces the maxilla anteriorly and forms the two ascending bars delimiting the orbit from the antorbital and infraorbital fenestrae. The bone projects posteriorly and somewhat outwardly to bound the infra- temporal opening from below and to contact the quadra to jugal ele- ment. The inferior surface is flat. Quadrate (fig. 4,B2): The quadrate has been displaced during preservation, and lies horizontally above the surangular-angular com- plex of the jaw. It is a triangularly concave bone which bears a gently curved, broad articular facet, and a distinct vertical anterior supporting ridge on the lateral side. The inner surface is rugose and probably fragmented. Pterygoid (fig. 4,B2): An irregularly shaped flat element, per- haps the quadrate process of the pterygoid, lies in front of and touches the quadrate. The straighter anterior part has a finished border and may represent the incipient flange of the lateral process. Quadratojugal: A problematic quadratojugal fragment lies above the surangular-angular complex of the jaw and contacts the quadrate and pterygoid. Its features are quite indistinct. The skull is lightly built on the ornithosuchian pattern. The or- bital contours are similar to Ornithosuchus and Hesperosuchus agilis I 1 = 10 mm. Fig. 4. Dibolhrosuchus'elaphros (gen. et sp. nov.). A, Fragment of the skull roofing bones— dorsal (1) and ventral (2) views. B, Left lower (1) and upper (2) jaws. (CUP 2081). 15 16 FIELDIANA: GEOLOGY, VOLUME 15 Table 2.— MEASUREMENTS OF DIBOTHROSUCHUS ELAPHROS (in mm.) Skull Length of the skull "cap" Preserved length along the inferior margin Ventral length of the preorbital fenestra Maximum breadth (anteroposterior) of the orbit. Ventral length of the infratemporal fenestra Length of the maxilla Jaw Preserved length Height: broken anterior end 20 mm. back of symphysis anterior margin of the lateral fenestra . constriction posterior to the fenestra . . Length of the mandibular fenestra Length of the tooth row left 157.0 11.2 10.2 20.0 14.3 29.0 Maxillary Teeth Fragment of the largest robust tooth basal breadth length (lacking tip) Basal breadth of the crown at the symphysis . Length of the crown at the jaw symphysis. . . Right Side 1 2 3 4 5 6 (replacement ?) 7 9. 10. Mandibular Teeth Length of the tooth at the symphysis (lacks the crown) 83.0 128.0 38.5 28.5 23.0 73.0 right 101.0 34.6 left right 5.0 5.0 7.3 2.5 10.0 length breadth 7.8 3.4 7.0 3.1 7.0 3.2 7.0 3.0 6.8 3.0 3.0 5.2 2.8 2.9 1.9 4.1 2.0 3.6 1.9 length breadth 7.0 2.3 Colbert (1952) and the ventral skull of the latter features nearly identical endocranial impressions. The position of the external nares high on the skull, either dorsally in Cerritosaurus binsfeldi Price (1946) or above the anterior angle of the preorbital fenestra in Salto- posuchus connectens and S. longiceps Huene (1921), excludes these forms from direct comparison. The preorbital fenestra in Dibothro- suchus is neither aetosaurid-like as in Erpetosuchus granti Newton (1894), nor crescent-shaped as in Prestosuchus chiniquensis Huene (1936, drawing by Price, 1946), nor as tiny as that of Cerritosaurus. Broadly oval orbits are known for Saltoposuchus connectens, Sclero- SIMMONS: NON-THERAPSID REPTILES 17 mochlus taylori (Huene, 1914-15) and Hesperosuchus, but the supra- orbital borders are "wulstig erholt" or padded in the latter two forms, and elevated above the roofing bones in Ornithosuchus and Hesperosuchus. The infratemporal region of the skull of Hesperosuchus was not preserved, but the associated fenestra lies slightly below and back of the orbit in Dibothrosuchus. Conversely, the orbit overlaps the in- teroanterior angle of the fenestra in Scleromochlus, or lies high above the fenestra in Prestosuchus, due to the depression of the inferior skull contours. The quadrate bones of Hesperosuchus and of the present form are very similar although the articular surface is broader and the sup- porting ridge is better developed and more anteriorly located in Dibo- throsuchus. The skulls of each in toto are equal in length to nine to ten dorsal vertebrae. Maxillary Dentition (fig. 4,B2) : The left maxilla holds 12 imper- fectly preserved thecodont teeth which become successively smaller posteriorly. The stoutest and largest lies 10.0 mm. in front of the preorbital and is succeeded by a gap accommodating two teeth. A less robust tooth lies below the anterior corner of the preorbital, fol- lowed by nine to ten teeth with alveolae for three to four others. The complete maxillary series numbers 16 to 18 teeth; the anterior- most slope backward but the others are erect. The teeth are broad laterally compressed, slightly recurved with sharp and finely den- ticulated (16/1.0 mm.) posterior edges. The right maxilla (fig. 5,A) holds well-preserved teeth which are laterally compressed but slightly convex, pointed and equipped with sharp trenchant edges. The crowns of two slender premaxillary teeth are embedded in the matrix of the symphyseal fragment. The left dentary (fig. 4,B1) holds the roots of 10-11 teeth of uniform size which are weaker than the maxillary series. The fourth root opposing the robust maxillary tooth is also the largest. The complement of the right jaw is masked by the opposed maxillary series. The development of canine-like teeth in the front of the maxilla is common among Pseudosuchia. The individual teeth compare more favorably with those of the coelurosaur Halticosaurus orbito- angulatus Huene (1932) than with the lanceolate type in Hespero- suchus, but the overall pattern — a needle-like premaxillary tooth passing to a more robust tooth posteriorly — is hesperosuchid. CO 1 a, 3 co to a: o CO O as o 05 fa O w < Pi PQ H H Pi fa > o Pi fa O H W fa O CO cH Z w w Pi to CO CO 9 ca H Ph d l« o pq ^ ■— co o.2 ■s *- coco fa.S eg io l-H ■* cc o oo o < m ^ H i u s a E- ^ W •< _3 p^ O ooot-to oo cr> io ■«* co U5 O CM © OO) ffi o w co © t- t- oo T-l OS lO CM CM 55 S CM SO "M CM J < Oi t- P 00 CO o o co i-h o CM 01 1—1 > i-H i-H o Q 1— C iH OS H co io rH CO a < 5 OS CM in i—i < jjij i—l i— ( o l-H r-\ T-H H H tf EH <*} oo • CC 00 o Oi IO lffl 56 CO lO PQ OS CM i—l CM CO -^ i-H i— 1 i— 1 « OS IO O CO CM O co ^ tj< io -^ co os o t}* CO © -^ io m CM CO CO CM i-H CM CO ^ CM 00 CO CO rj< co co co co © s S" K W u « o 5 H 35 ffiPQ ■81*5. S3 cs rt fl ** T3 T3 a> «_ rt CS — O CD CO^J fe £j CD CD Ph " as a O PS c O HH j3 u a I 0? ft .22 ft w 43 -ft^&c^ »• « & & ^£ N c & -l_> CO GO _ -u S3 e O ft .22 S3 tj C ft-p io rt - S3 « - >> - -j3 .b +* +* ft +» bo X CJC bo o ciO -S5 W S3 c a S3 ffiS Jh4 to O N N N + PhP-,Ph ^ •»a CCi CD CS C8 PQ-H • £ cd en 3 4. >> i J | & |ll | as bo 2* ® 'Si C b ® « PQ +i O1 . ■eS •4 *» £ as io o t- rj< cm os ■ •* Tj ^< Tf CO K5 lO P> • t-H i— I tH i-l i-l i— I i— I 00 <3<1 PhPh KHffiPQPQKPQ > «j < g £ ru HNMTjW(Ct-00a)OHNWTf 22 SIMMONS: NON-THERAPSID REPTILES 23 With the exception of the last or fourteenth dorsal, the costal articulations and the posterior zygapophyses of each vertebra are connected by a bony lamella directed upward and backward. The lamella of the fourteenth is relatively parallel to the long axis of the column. The relation of the parapophyses to the anterior zyga- pophyses is somewhat variable throughout. Parapophyses: The bony lamella essentially passes through the midplane of the parapophysis. The processes are rounded (4.0 mm. in diameter) with concave facets for the capitula of the ribs and they extend outward some 8.0 mm. In the first six dorsals, the para- pophyses lie just back of the anterior margin of the centra and under the posterior half of the anterior zygapophyses. In the seventh and eighth they migrate slightly backward. In the ninth dorsal it lies under the zygapophysis. The costal articulations fuse rather abruptly at the level of the tenth or eleventh vertebra; the distance between the parapophyses and diapophyses does not appear to decrease gradually in the latter part of the column, i.e., 1-9 (?10). Transverse Processes: The lateral projection and basal breadth of the processes are progressively reduced from the first-ninth verte- brae. They are supported by two rather vertical lamellae on the second dorsal. The processes are flattened above from the twelfth- fourteenth dorsal. Zygapophyses: In the first six dorsals, the anterior zygapophyses are low as in the posterior cervicals, but become progressively higher from the seventh-thirteenth. Early, their facets are oblique, but they rotate outwardly in the last half of the series. The posterior zygapophyses are longer than the anterior processes and project be- yond the end of the centra. The articulation between the thirteenth and fourteenth dorsals is nearly horizontal. Such an arrangement would tend to reduce the vertical motion of the posterior presacrals. Neural Spines: The neural spines are relatively long, low and broad. They project forward only slightly, but are more generally erect from the eleventh -fourteenth dorsal. Only a few are com- pletely preserved. Ribs: Several proximal fragments of broad flat ribs were fused during preservation to the neural spines of the thirteenth and four- teenth dorsal vertebrae. Summary: There is a general tendency for the vertebrae to in- crease in length throughout the dorsal column although the sixth 24 FIELDIANA: GEOLOGY, VOLUME 15 dorsal is the longest. The costal articulations become fused toward the sacrum and the zygapophysial articulations become more hori- zontal to restrict the vertical motion of the spine. Several broad rib fragments are associated with the last two presacral vertebrae, but it is questionable whether these can be considered to belong to this specimen. Sacral Series (Table 5; fig. 5,E): The sacrum is composed of three vertebrae with broadly rounded ventral surfaces. The first two are articulated but the third has been transversely broken at its mid-length. The total length of the articulated sacrum is 56.3 mm. (1st, 18 mm.; 2nd, 14.5 mm.; 3rd, 13.8 mm.) along their ventral lengths. The successive vertebrae decrease in length; the neural complex of the second and third are fused. The first and third centra are more strongly compressed than the second which is compara- tively massive. None of the neural spines has been preserved save that of the first which is supported by high lamellae most probably continuous with the posterior zygapophyses. The anterior zygapophyses lies at the level of the sacral ribs on the first vertebra but are raised high above the neurocentral suture in the second and third. The zygapophyses of the first vertebra ap- pear to be partially fused to the posterior zygapophyses of the last presacral. The neural articulations of the second and third sacrals also seem to be fused. The posterior zygapophyses of the first and second sacrals are elevated high above the neurocentral junction and are continuous with the lamellae supporting the neural spines. They originate at the posterior border of the sacral ribs. Ribs (fig. 5,E) : The first sacral lacks ribs. These, however, are represented by a scar at the anterior margin of the neurocentral suture (height, 11.3 mm.; breadth, 8.0 mm.). The second left sacral rib lies on the posterior half of the vertebra. It is dorsoventrally flat and irregularly triangular. The anterior border extends outward and slightly backward for half its length, and turns forward distally; the posterior border is straight and pro- longed backward. The facet receiving the ilium is expansive and lies at the anterior corner. Distally, the rib becomes rather flat but the facet receiving the ilium of the pelvis is expansive at the anterior- most corner. The third sacral bears a fragmented narrow transverse process opposing the posterior border of the second rib. SIMMONS: NON-THERAPSID REPTILES 25 Table 5.— MEASUREMENTS OF THE SACRAL SERIES OF DIBOTHROSUCHUS ELAPHROS (in mm.) Vertebral No. *V.L. I 18.0 II 14.5 III 13.8 I II III L.AB. 14.2 20.5 A.H. 14.9 11.0 13.0 L.PB. 19.0 A.B. 13.0 13.1 13.6 Vertebra M.B. 8.3 13.5 10.0 D.B. 21.7 Rib Pr.B. 13.3 Single Rib ?20.8 ?13.3 Ht. Az.+C. 22.0 ?21.0 ?19.0 D.F. 10.0 12.1 *V.L.- Ventral length L.AB. A.H.: Anterior height L.PB.: A.B.: Anterior breadth D.B.: M.B.: Median breadth Pr.B.. Ht. Az.+C. Height, ant. zygapophysis + centrum D.F. Ht. Ns.+C: Height, neural spine + centrum B.F.: Ht. Ns.+C. 28.6 B.F. 12.3 13.8 Length, ant. border Length, post, border Dorsal breadth Proximal breadth Dorsal facet Breadth of facet A fragment of a triangular rib similar to that of the second sacral is associated with these remains. Its articular facet for the ilium is considerably larger. Caudal Series (Table 6; fig. 5,E): The postsacral vertebrae are only partially represented. The first six caudals form an articu- lated series with the sacrals. The more distal vertebrae occur as multiple fragments. Centra: The first six caudals are quite uniform. The centra are squarely constructed and compressed with broadly rounded and fur- rowed ventral surfaces; haemapophyses are associated with each. The smaller distal centra are similar in form ar.d presumably amphy- coelous. Except for one very distal element, the articulated frag- ments are transversely broken at the mid-part of the terminal centra. Neural Spines: The neural spines, while incompletely preserved are clearly directed anteriorly. Transverse Processes: The processes lie at the center of the ver- tebrae and project somewhat outward and downward, on the first caudal. They gradually rotate upward and backward in the second - fifth to juxtapose the posterior zygapophyses. A lamella, similar to that noted for the dorsal series, unites the bases of the transverse processes and the posterior zygapophyses in the fourth -sixth caudals 26 FIELDIANA: GEOLOGY, VOLUME 15 and, in the last, the diapophysis lies under the zygapophysis. The terminal vertebrae lack diapophyses entirely. Zygapophyses: The anterior zygapophyses become progressively lower and more oblique posteriorly, but project at a low angle beyond the margin of the preceding centrum. From the fourth vertebra posteriorly, the anterior zygapophyses form a costal-neural "articular clump" with the posterior zygapophyses and transverse process of each antecedent vertebra. The posterior zygapophyses are short and widely divergent. Table 6.— MEASUREMENTS OF THE CAUDAL SERIES OF DIBOTHROSUCHUS ELAPHROS (in mm.) Position L. in Series *V.L. A.H. A.B. M.B. Az-Pz L.AC. L.Hen Anterior Caudal Series 1 13.3 14.6 8.7 18.5 12.7 2 12.5 11.6 8.0 17.8 12.5 3 10.0 10.6 7.3 17.2 11.3 4 10.0 10.9 7.0 17.0 10.2 ?18!6 5 12.0 9.7 7.4 17.0 9.6 ?21.4 6 12.5 9.0 18.0 8.7 Posterior Caudal Fragments 1 11.0 10.2 8.4 5.5 15.3 8.0 2 12.3 9.1 8.1 6.2 13.0 7.3 3 10.0 12.6 9.0 6.9 17.7 ?8.2 4 10.6 5.5 17.7 7.4 5 10.5 8.2 16.6 Posterior Caudal Fragments 1 ?9.0 7.2 6.0 7.5 13.5 2 10.0 7.5 6.0 7.7 15.0 3 10.0 7.5 5.7 3.7 ?12.0 *V.L.: Ventral length A.H.: Anterior height A.B.: Anterior breadth M.B.: Median breadth L.Az-Pz.: Length, tip ant. zygapophysis-tip post, zy- gapophysis L.AC: Length, articular 'clump' L.Haem.: Length, hemapophy sis Hemapophyses: Chevron bones are associated with all but the terminal centra; the rami are delicate and fuse distally. These are best preserved between the fourth and fifth, and fifth and sixth vertebrae. The anterior cervical column of Dibothrosuchus is best compared to those of Hesperosuchus and Parringtonia gracilis Huene (1939a), although these vary somewhat proportionately, and their centra are more compressed and narrowly rounded below. The anterior pre- SIMMONS: NON-THERAPSID REPTILES 27 sacrals of Parringtonia (Huene, 1939a, fig. 5) compare well to the seventh -eighth dorsals of the present form where the neural arch bears costal articulations connected by a common ascending lamella, and erect anterior zygapophyses. The anterior projection of the zygapophysial articulations of the imperfectly preserved dorsal col- umn in Hesperosuchus is of the dibothrosuchid-type; the parapophy- ses and diapophyses are similarly arranged in Stagonosuchus nyassicus Huene (1939b) but lack the common lamella. The rather horizontal orientation and length of the posterior zygapophyses of the posterior presacrals is similar to Parringtonia and Hesperosuchus, but the poorly preserved neural spines lack terminal expansions for the articulation of the dorsal armor in the former (Huene, 1939b, figs. 8, 10) and in Stagonosuchus (Ibid, fig. 11). The presacral column cannot be com- pared further to other ornithosuchia such as Erpetosuchus (which has weak costal articulations) which differ markedly also in their skull patterns. The sacral and caudal vertebrae are flatter and more broadly con- structed than those of Hesperosuchus, and the latter do not bear ventral longitudinal ridges evidenced in Parringtonia. Table 7.— MEASUREMENTS OF THE GIRDLE AND FORELIMBS OF DIBOTHROSUCHUS ELAPHROS (in mm.) CORACOID , .. . . . Length of: left n&ht base of glenoid fossa to posterior crest 29 . 6 dorsal crest 24 . 6 articular facet for scapula 14 .0 Breadth of articular facet for scapula 9.0 Length of free finished border of coracoid-dorsal 10.6 Breadth of glenoid region 11.0 Depth of glenoid region 9.6 Humerus Length of head to posterior tuberosity (maximum) 26.4 25.0 Dorsoventral breadth at crest 11.7 10 .2 Breadth (anteroposterior) above crest 16.3 15.0 Breadth, neck just above condyles (lateromedial) 12.0 Thickness, neck just above condyles (anteroposterior) 8.2 Breadth of condyles (maximum lateromedial) 18.3 Thickness, lateral condyle 7.6 Thickness, medial condyle 7.2 Length, intermediate protuberance 11.5 11.0 Length of humerus (allowing 6 mm. for sectioned ends) : proximal element 48.3 40 . 0 distal element 58 . 0 .... Estimated total length 116.3 5 i 1 = 10 mm. 10 1 | = i0mm 28 SIMMONS: NON-THERAPSID REPTILES 29 PECTORAL GIRDLE (Table 7, Fig. 6,A) The pectoral girdle is known by a left proximal fragment of the coracoid. The bone must have been loosely connected to the scapula since the superior border is finished except for a roughly triangular area in back of the glenoid region, and it rotates inward posteriad (effect of pressure?) . The posterior surface is roughened to indicate a cartilaginous cap. The ventral border is finished. The lateral surface is concave and bears an extensive fossa for the M . supracoracoideus. The glenoid fossa formed by the scapula and coracoid is flared laterally and bears a lateral tubercle which probably represents an incipient acromion for the triceps muscle. The glenoid is set off below by a sharp notch and a marked fossa; the bone is narrow and may have been distally elongate. The inner surface is concave and perforated by a foramen (supra- coracoid) which enters at a superior post-glenoid depression and emerges at the level of the acromion in the forward wall of the supra- scapular fossa. APPENDICULAR SKELETON (Table 7) Fragments of the forelimb were segregated from a number of in- determinate limb bones associated with the skull and vertebrae on the basis of their common histologic structure and hesperosuchid-like configuration. Humerus (fig. 6,B): The humerus is long, slender and proxi- mally expansive; the shaft is bent medially. The head is rounded and positioned at the anterior inferior angle. A large, rather flat, semilunar internal tuberculum lies just posteriad to the head and projects over the crest to the dorsal side. More posteriorly and at approximately right angles, the dorsal crest emanates from the ven- tromediad wall as a strong ridge and passes distally to the neck of the bone. The ridge is acute above the apex which is swollen and roughened for muscle attachment. Fig. 6. Dibothrosuchus elaphros (gen. et sp. nov.). A, Coracoid, viewed from the medial (1), glenoid (2), and lateral (3) surfaces. B, Humerus, left (1-5) and right (6-10) proximal fragments viewed from the medial (1, 6), dorsal (2, 7), ventral (3, 8), and articular (5 ,10) surfaces. (CUP 2081) 30 FIELDIANA: GEOLOGY, VOLUME 15 The proximal end of the humerus is dorsoventrally convexocon- cave and the deltopectoral crest is set off from the dorsal surface by a narrow sulcus. The ventral surface becomes flatter distally. The distal segment of the left humerus (fig. 7,A) curves laterad to give the bone an overall sigmoid configuration. The condyles are subequal in size; the anterior, or radial, is the longer, rather triangu- lar in cross section, and its articular surface is delimited from above by a Y-system of ridges. The posterior ulnar condyle is broad and its articular surface is ventrally diposed. A transverse ridge con- nects the condyles on the dorsal surface of the bone. The trochlea is marked by a shallow intermediate depression. Ulna (fig. 7,B) : The ulna is known from a dorsoventrally com- pressed, moderately expansive proximal fragment which evidences a prominent olecranon process; the medial surface is irregular and asymmetrical. The surface articulating with the humerus is divided by a mod- erately elevated ridge occupying the trochlea and is plateaued. The depression receiving the lateral condyle is the broader and deeper of the two but the arrangement would probably not restrict the rotation of the forearm to any great degree since the elements were undoubt- ably capped by cartilage. Radius (fig. 7,B): Only the head of the radius is represented. It is broad and shows a semilunar depression for the articulation with the humerus. Its shaft is slender and dorsoventrally flattened. The girdle of the present form is distinguished from other ornitho- suchians in that (1) the coracoid and scapula were apparently loosely associated, (2) the glenoid is strongly rotated outward and (3) the subglenoid fossa is acute rather than broad. The glenoid fossa here has a more open face, but lacks both development in breadth and an accessory lateral process (?acromion) for the triceps muscle. The pseudosuchian humerus is generally characterized by broadly expansive ends (Stagonosuchus and Erpetosuchus) , but the slender- ness of the shaft, position and development of the deltopectoral crest, and condylar inequality in the present form is closest to Hespero- suchus. The coelurosaur Halticosaurus longotarsus Huene (1921, pi. IV, fig. 7) also shows rather similar condylar proportions. The proximal part of the ulna is constructed like that of Hespero- suchus, and the olecranon and dorsal oval depression are equally well developed. The sigmoid notch is, however, less strongly platformed and the humerus-ulnar articulation is simpler in Hesperosuchus. The SIMMONS: NON-THERAPSID REPTILES 31 peculiar nature of the joint would facilitate a fast running habit, allowing for a strong fore-aft motion with perhaps little rotation of the elements. The animal was probably quadrupedal since the humerus is elongate, and Romer (1956, p. 373) notes that the olec- ranon (origin of the triceps muscle), little developed in dinosaurs, appears in secondarily quadrupedal forms; its presence is presumably associated with a secondary return to a more flexed position of the forearm. The osteology of the skull and appendicular skeleton indicates a hesperosuchid-like animal, but the form is slightly larger. The dif- ferences in the jaw, skull and vertebral column are conspicuous (the latter approaching Parringtonia most closely). The limb structure is perhaps more highly specialized for a cursorial way of life. Order Thecodonta Suborder Pseudosuchia Family Ornithosuchidae Genus Strigosuchus, gen. nov. Diagnosis. — Jaw slender; lower margin straight but sharply up- turned at the symphysis. A flat splenial element sheaths the inner wall. The jaw bears only one preserved tooth. Strigosuchus licinus, sp. nov. Holotype. — A fragment of a left mandible, slender and upturned at the symphysis. Catalog number CUP 2082. Horizon and locality. — Ta Ti, from the Dark Red Beds of the Lower Lufeng Series. Diagnosis. — Same as for genus. DESCRIPTION Jaw (Table 8; fig. 7,C): The mandible is slender and upturned at the snout. Its outer wall is convex although flattened anteriorly; the inner or medial wall is flat in front becoming more concave pos- teriorly below the alveolar border. The preservation of the material was accompanied by distortion and checkering of the bone, but the sutures are clearly marked. The outer wall of the dentary has been broken away to expose the internal architecture of the jaw, and the articular bones are incomplete. "10mm. /L^g^Oi Fig. 7. Dibothrosuchus elaphros (gen. et sp. nov.). A, Left humerus, distal fragment viewed from the ventral (1), lateral (2), and articular (3) sides. B, Ulna and radius, proximal fragments in articular (1), medial (2), and lateral (3) aspect (CUP 2081). Strigosuchus licinus (gen .et sp. nov.). C, Jaw in lateral view. (CUP 2082). Sinosaurus triassicus Young. D, Jaw fragment posterior to the symphysis. (CUP 2097). Coelurosaur (CUP 2090). E, Fused metatarsals in dorsal view. 32 SIMMONS: NON-THERAPSID REPTILES 33 Table 8.— MEASUREMENTS OF THE JAW OF STRIGOSUCHUS LICINUS mm. Superior border from broken tip to end of fragment, i.e., lateral fenestration 49.6 Inferior border from broken tip to posterior fracture of angular .... 47.2 Anterior height of jaw at broken tip 7.0 Height of dentary just posterior to preserved tooth 8.3 Height of jaw at posterior fracture 12 .0 Basal length of tooth root 2.7 Height of preserved portion of tooth minus crown 2.7 Anteroposterior length of second preserved alveolus 2.5 The jaw appeared upon initial examination identical to Chasmato- saurus yuani Young (1936). Closer inspection revealed that the sur- face of the posterior outer fracture was in part finished — evidence that, unlike proterosuchians, the jaw was fenestrated. The man- dibles of Hesperosuchus and Dibothrosuchus are similarly formed, and the symphysis in each bears the greatest concentration of foramina. Further comparison is rather limited for reasons given below. In the present specimen, a flat spenial bone covers the inner sur- face of the jaw at the symphysis and contacts the dentary along the inferior margin of the ramus. The posterior portion has been dis- placed upward and outward by compression during preservation; it encroaches upon the dentigerous border to effect the postmortem loss of the teeth. Forward of the jaw articulation, the splenial wraps about the angular bone. The meckelian groove passing forward on the inside of the jaw from an oblique nubbin (7.0 mm. long) of the angular bone is limited by a long dorsal prearticular element which pinches out anteriad and disappears below the symphyseal portion of the splenial. The damage to the dentigerous border incurred by the displace- ment of the splenial does not allow us a precise knowledge of the dental complement. It is, however, likely that the jaw held 14 to 15 teeth. The single root of a large compressed, somewhat recurved tooth which is most chasmatosaurid-like, lies toward the end of the tooth row and its anteroposterior breadth is equal to the length of the first and second alveolae at the symphysis. Only the first four alveolae are sufficiently well-preserved to indicate the size of the teeth, but it is not known whether these were enlarged like those of Ornithosuchus and Pedeticosaurus leviseuri van Hoepen (1915; Haugh- ton, 1924). The teeth of Hesperosuchus are contrastingly lanceolate, and the jaw is three times as large as that of Strigosuchus. 34 FIELDIANA: GEOLOGY, VOLUME 15 Irrespective of the general chasmatosaurid configuration of the jaw and teeth, the presence of a lateral mandibular fenestra indicates that the present form should be referred to the ornithosuchian pseu- dosuchians. The specimen is regarded as a new form for which the name Strigosuchus licinus (gen. et sp. nov.) is proposed. Order Thecodonta Suborder Pseudosuchia Family Platyognathidae (fam. nov.) Platyognathus hsui Young (1944) Platyognathus hsui was named by Young (1944) from a jaw frag- ment, and was assigned to the pseudosuchia. Since then, its ultimate position in the scheme of classification has been speculative at best in view of its fragmentary nature. Young failed to relate it with the well-known primitive South African forms, and stated that it might represent an entirely new family. Huene (private communication; Young, 1951) suggested Platyognathus as a forerunner of the Orni- thischia since, following the contours of the jaw, the snout was inter- preted to be expansive. Romer, earlier (1955), concurred with Young's tentative assignment, but later (1956) referred it to the Notochampsidae (Protosuchidae) . Pivateau (1956) suggested a re- lationship to the pseudosuchian Stagonolepidae. The type specimen, No. V71, Cenozoic Research Laboratory, National Geologic Survey of China, consisted of the anterior part of a pair of jaws bearing alveolae for 11 and 12 teeth in the left and right sides respectively. The preserved dentition was restricted to a single damaged canine-like tooth on the right side. Additional Material. — The newly available material includes a partially complete jaw and skull, and armored postcranial elements. All but the sacrum of the vertebral column is represented in part, and the pectoral girdle is complete on both sides. The appendicular skeleton is not represented. Catalog No. CUP 2083. Horizon and Locality. — Ta Ti, Dark Red Beds, Lower Lufeng Series. Diagnosis.- — Skull V-shaped, low, flat and terminally expansive at the snout; external nares unpaired; symphyseal teeth anteriorly directed, polygonal in cross section. Lumbar and caudal regions of the body completely invested in armor; vertebrae procoelous. SIMMONS: NON-THERAPSID REPTILES 35 DESCRIPTION Skull (Table 9; figs. 8,A1-3,D): The skull is complete from the snout to the antorbital fenestra. The dentition is completely repre- sented. Skull low, broad and triangular; snout short, obtuse and laterally expansive; external nares terminal and confluent. The bone is extensively pitted with canals for vascular and nervous elements. Premaxilla: The premaxilla is expansive and forms the entire lat- eral wall of the snout and the ventral border of the external nares. The superior margin projects posteriorly about 1.0 cm. between the nasal bones and the anterior portion of the maxilla. There is no evi- dence of an internasal septum and the nares appear to be confluent; had the septum been cartilaginous, however, it would not have been preserved. The bone bears four teeth; the first two are anteriorly directed, polygonal, in cross section and subequal in size, the second is the smallest. The third and fourth teeth are recurved. The suture with the maxilla is edentulous where the bone may either have been broken, or was in fact emarginated. Nasals: The nasal bones are broad and flat, forming the dorsal roof of the skull and external nares. Their median suture is slightly depressed. The bones terminate abruptly at the snout, and it is probable that they projected a short way anteriad. The bones bear many fracture lines due to compression. Maxillary: While the preservation is generally good, it is difficult to distinguish cracks from sutures. The outer wall is convex, bounded anteriorly and /or dorsally by the posterior tip of the pre- maxilla and by the nasals. The contact with the premaxilla involves the edentulous region which accommodates a canine-like tooth of the mandible; the skull is constricted at this level. The bone extends about the rounded anterior angle of the antorbital fenestra. Seven- teen to 18 finely denticulated (9-10/1.0 mm.) and recurved teeth, the anterior seven of which are largest, lies on this bone. The bone reveals little about the conformation of the orbit. The remnant of the antorbital fenestra lies relatively high on the skull oriented more or less dorsally due to the convexity of the bone, and it is probable that the orbits were similarly placed. The inferior border of the maxilla is flared medially to form the lateral walls of the choanae. The internal and external borders of the tooth row are equal in height. Fig. 8. Platyognathus hsui Young. A, Skull and jaws from the right side(l), ventral (2), and dorsal (3) surfaces; (CUP 2083). B, Articulated posterior cervical vertebrae and scutes viewed from the dorsal surface (1), right lateral side showing the scapula (2), and coracoid (3) drawn from the unillustrated ventral surface (CUP 2083). C, Anterior caudal vertebrae viewed from the left side (CUP 2083). D, Jaw fragment showing the inner surface of the orbital region (CUP 2104). E, Posterior cervical vertebrae and scute complex viewed from the right side (CUP 2105). 36 SIMMONS: NON-THERAPSID REPTILES 37 The suborbital region is shown in a maxillary fragment of another specimen of comparable size (CUP 2104) ; the wall was depressed be- low the orbit where the bone is particularly thin and not very likely to have been preserved intact. The disposition of the bone might indicate a more forwardly oriented orbit, but it is impossible to say if it was directed more outward or more upward as in the case of later crocodilians. Internally, a deep fossa which may have held a lacry- mal gland is formed by an upward extension of the palate. This feature indicates that the antorbital and orbital openings were closely juxtaposed. Jugal: The jugal replaces the maxilla at the end of the tooth row, and it is probable that it limits the orbit from below. Palatal Complex (figs. 8,A3,D): The palate is poorly preserved and largely obscured by the long symphysis. The bones have been exposed in cross section at the level of the antorbital fenestra where the skull and jaws have been transversely fractured. Vomers: The vomers are median-paired, delicate bones (height, 5-6 mm.; thickness, 1-1.5 mm.) which lie just forward of the end of the symphysis and pass upward to divide the choanae. Above, the structure flares to form the palate proper (breadth, 14-15 mm.). Palate: The palate extends posteriorly and contacts the inner wall of the maxilla. If the terminal epipyterygoid element is included, it is as long as the tooth row; the posterior 14 mm. are free and project downward across the tooth row. As noted above, the palate forms a deep fossa (length, 13.8 mm.; depth, 3.7 mm.) below the orbit (fig. 8,D) . The mid-line structures are broken away behin the antor- bital fenestra and it is impossible to say whether there was any pterygoid participation. Nevertheless, the development of the vomer has effected at least the rudiments of a secondary palate which can best be described as mesosuchian. Ectopterygoid: This is a small, narrow element best preserved on the right side. The medial surface is entire and finished, indicating that the associated pterygoid vacuities may have been large. The snouts of Pseudosuchia are, as a rule, tapered and acute, and bear expanded premaxillae. The primitive crocodilian Protosuchus shows comparable expansions and a slight emargination of the suture with the maxilla marking an edentulous or diastemic area. All proto- suchians, however — with the exception of Pedeticosaurus which has been interpreted as a highly specialized pseudosuchian — lack ant- orbital fenestrae and the external nares open dorsally. The common suture of the nasal bones, nevertheless, is depressed. 38 FIELDIANA: GEOLOGY, VOLUME 15 Table 9.— MEASUREMENTS OF THE SKULL AND JAWS OF PLATYOGNATHUS HSUI (in mm.) left right Skull Length: from the external nares to the antorbital opening 35.7 dorsal margin of the premaxilla 21.0 20.0 ventral margin of the maxilla 45.0 50.0 Breadth: at the snout 15.0 of the external nares 7.0 premaxilla (maximum) 18 .0 of the nasal bones: anterior, posterior 8.0, 14.0 of the premaxilla-maxillary junction 16.0 Height of the maxilla at the antorbital fenestra 12.4 12.2 Jaw Length of the symphysis 30 . 5 Breadth: of the symphysis (maximum) 16.0 of the submaxillary constriction 13.0 Length of the rami (preserved length) 57.3 81 .4 Height of jaw at posterior end of maxilla 14 .0 Jaw Fragments CUP 2083 CUP 2104 left right Preserved length 34 .0 47 . 5 Height of maxilla: anterior, posterior 8.7, 3.0 6.0, ... . Height of mandible: anterior, posterior 9.0, 11.0 9.0, 14.0 The prolonged posterior extension of the palate is indicative of a sphenosuchid heritage; the choanae are abbreviated but lie on the anterior part of the skull like Erpetosuchus, and the formation of the primitive secondary palate (mesosuchian?) is apparently at the same level of development as that of Protosuchus. If the dorsal placement of the nares is a reliable feature, Protosuchus was prob- ably more aquatic than Platyognathus. Jaws (fig. 8,A2) : The shape of the mandible is particularly diag- nostic for Platyognathus. The jaw is blunt. Posteriorly, the sym- physis gradually expands and then constricts so that the prepre- maxilla overlaps the zone of increasing breadth and the diastemic area lies over the expansion. The snout is slightly broader than the jaw. The symphysis is long and deep. The rami are tightly co-os- sified and exhibit the shallow troughs noted by Young to either side of the dorsal mid-line. In the present example, the grooves appear at the level of the fifth maxillary tooth where the left and right rami diverge sharply at an angle of about 45°. The ventral surface of the symphysis is flattened and depressed just forward of the constriction SIMMONS: NON-THERAPSID REPTILES 39 by a shallow transverse groove. Young described a similar groove for the type which lay further forward and masked the individuality of the rami; here, the rami are distinctive throughout. The dentition is largely obscured by the close apposition of the skull. The jaw lacks a diastema and probably holds some 23-24 teeth which increase in size backwards in the series. The first two project anteriorly and are polygonal in cross section like the maxil- lary pattern. In contrast to the type, the others are more erect but still caniniform, while the more posterior are slightly recurved. The ventral surface bears round impressions of two sets of large roots, 8.8 mm. and 13.5 mm. from the tip respectively which may be the bases of the third and fourth teeth. Surangular and Angular: The bones diverge posteriorly and pinch out the dentary at the level of the orbit. The jaw becomes deeper and appears to contain a small lateral fenestra (CUP 2104). The angular displaces the splenial ventromedially which forms the lateral wall of a meckelian fossa. Splenial: The bone sheaths a medial wall of the dentary and their common suture is well marked. Young misinterpreted the meckelian groove as the point of insertion of a cartilaginous splenial and stated that the bone does not enter into the symphysis. It is here clearly seen to participate in the symphysis if only by a ventral tongue, and its anteriormost limit lies where the lateromedial dimensions of the diverging rami narrow abruptly. The mandibles are unlike those of any of the pseudosuchia or protosuchia. The disposition of the splenial is sphenosuchid-like, but the relationship of the dentary and splenial is unknown in Proto- suchus; this is, however, crocodilian in character, particularly in the longer-snouted forms. The symphysis of Erpetosuchus is elongate but proportionately shorter relative to the length of the head. Teeth, Upper Jaw: The premaxilla holds four subequal teeth. The first and second teeth point forward while the third and fourth are gently compressed, sharply pointed and distinctly recurved. The second is needle-like and the smallest. Young described the teeth of the type as octangular in cross-section, but since their geometry in fact varies with the level considered, it might be more precise to describe them as irregularly polygonal. Hence, distal to the tip, the teeth are rather hexagonal in cross-section and only the bases are octagonal. The post-diastemic teeth are immediately less robust and become progressively smaller backward and more sharply compressed. The 40 FIELDIANA: GEOLOGY, VOLUME 15 last three-four may be newly erupted. Fine denticulations (9- 10/1.0 mm.) appear on both cutting edges. Teeth, Lower Jaw: The maxillae mask all but the most anterior teeth, but these are patterned after the upper series. The first two teeth are uniformly small, anteriorly projecting, and are followed by larger and more erect teeth. Posteriorly, the teeth are smaller but indistinct. No pseudosuchian known to the author has been described with polygonal teeth. Moreover, the dental pattern of Protosuchus, where the reduction of the maxillary series is nearly immediate, and of Pediticosaurus are dissimilar to the present form. VERTEBRAE The vertebrae occur as disarticulated and multiple fragments of at least two animals. The cervicals and thoracics bear dorsal and ventral armor plates; the sacrals are missing; the caudals are com- pletely sheathed in armor. Cervical Vertebrae (Table 10; fig. 8,B,E): Two articulated series of vertebrae, comprising five cervicals (CUP 2105) and five cer- vicals plus three dorsals, respectively, are known, but they cannot be derived from the same animal since the combined vertebral count would be characteristic of saurischians rather than thecodonts. The latter series is believed to belong with CUP 2083. Centra short, moderately high, platycoelous-slightly amphicoe- lous, increasing in length only slightly posteriorly in the series. Ven- tral surfaces smooth but furrowed. Parapophyses lie anteromediad on the first centrum but move slightly back on the following verte- brae; the trend is accomplished by the typical thecodont rotation of the costal facet outward and upward toward the neurocentral suture. The neural complex is directed craniad in all but the first verte- bra where it is erect. The arches are very high and the spines are terminally expanded to receive the dorsal armor. The transverse processes are stout becoming progressively more obtuse relative to the spine posteriorly. The zygapophyses although poorly preserved, were presumably delicate, high anteriorly becoming more oblique posteriorly. Dorsal vertebrae: The first three dorsal are in articulation with the cervical series of CUP 2083; only the third is adequately exposed. An armored fragment from the mid-dorsal column shows three addi- tional centra. SIMMONS: NON-THERAPSID REPTILES 41 Table 10.— MEASUREMENTS OF THE VERTEBRAL COLUMN, PLATYOGNATHUS HSUI (in mm.) Position Ventra Maximum Height of in Series Length 1 Breadth1 Cervical Series CUP 2105 Neural Complex 1 10.0 12.0 2 6io 10.0 3 6.4 11.0 4 6.0 ?13.3 5 6.5 ?13.0 6 6.4 ?13.0 7 9^5 Cervical Series CUP 2083 1 6.0 9.3 10.0 2 6.9 8.4 3 7.8 4 10.2 5 9.2 6 7 Dorsal Series CUP 2083 1 8.0 7.8 2 8.5 7.8 3 9.0 7.8 4 5 6 ?10.6 1 Caudal Series CUP 2083 ?ll!4 1 ?10.0 4.0 W 4.0 Armorless Caudals CUP 2083 1 6.0 7.0 9.0 2 6.6 j 11.0 3 5.9 7.0 11.0 1 Centrum Anterior centra are gently procoelous, compressed, and bear ven- tral longitudinal ridges which are lost posteriorly as they become deeply furrowed. These are less squat than the cervicals but not immediately more elongate. In the third vertebra (CUP 2083) the parapophyses have com- pleted their dorsal migration and lie just below the diapophysis. 42 FIELDIANA: GEOLOGY, VOLUME 15 Posteriorly, the parapophyses lie on or just above the neurocentral suture and subsequently appear to 'fuse' with the diapophyses form- ing narrow facets for single-headed ribs. The neural apparatus is high and directed craniad; the spine is elongate, flat, and terminally expansive like the cervicals. The zyga- pophyses are widely oblique to insure the maximum of lateral motion — a feature of distinct advantage to an aquatic animal. Caudal Vertebrae (Table 10; fig. 8,C): Anteriormost caudals are completely sheathed in armor. The posterior vertebrae are ex- posed to permit description. Centra procoelous, compressed, and anteriorly more elongate than the cervicals which they resemble in other respects. Ventral surfaces smooth, rounded, less deeply furrowed than the dorsals, and bear facets for hemapophyses. The caudal ribs articulate on the neuro- central suture, and above, the zygapophyses show widely oblique facets. The neural spines are low, co-ossified, and form a continuous ridge for the dorsal armor. Ribs (fig. 8,E) : The cervical ribs are plow shaped, approximately 11.0 mm. long and overlapping. The forward projection is longer. The anterior dorsal ribs are double-headed, acutely bent, and terminate on the sternum. The proximal and vertebral segments appear to be broken at their mid-lengths. The sternal segment bears a ventrolateral triangular flange at the level of the fracture which may have supported some lateral armor plates. The ribs of the mid- dorsal region are long, broadly arched, split sagitally and are appar- ently single-headed. The pattern of the vertebral column is protosuchid. The pro- gressive migration of the parapophyses toward the neural arch is essentially completed at the same level by Erpetosuchus (eighth - ninth presacral), Protosuchus (twelfth presacral) and by the present form (?ninth -tenth presacral). The ninth presacral of Erpetosuchus figured by Newton (1894) resembles the cervical of Platyognathus, but the protosuchid type is uniformly amphicoelous. The procoelous centra of Platyognathus is a much more advanced feature of croco- dilian evolution. PECTORAL GIRDLE The bones are nearly complete bilaterally, and lie on the cervical series (CUP 2083) . The dorsal blades of the scapulae are damaged in part. The left carocoid is complete, but the right retains only a proximal fragment. SIMMONS: NON-THERAPSID REPTILES 43 Scapula (fig. 8,B2): The bone is elongate and the dorsal and ventral blades are broadly expanded. The glenoid region is rotated slightly outward and bears a strong acromion tuberosity above. The dorsal portion tapers gradually to form a strong inferomedial constriction; the expansion of the ventral blade is immediate there- after and the anterior portion is prolonged. Coracoid (fig. 8,B3): The coracoid is elongate, expanded proxi- mally, and suturally united to the scapula where it participates in the glenoid. The dorsal surface is concave and pierced by an antero- medial foramen. Table 11.— MEASUREMENTS OF SCAPULA AND SCAPULOCORACOID OF PLATYOGNATHUS HSUI (in mm.) left right Scapula Preserved maximum length. ... 38.3 ?33.3 Level of: dorsal blade inferomedial constriction sutural union of scapula and scapulocoracoid acromion Scapulocoracoid Length of scapulocoracoid 21. 1 .... Dorsal blade (maximum) Medial constriction Ventral blade Sternum: A thin and unsegmented bowed plate of bone lying to the left of the ventral cervical scutes (21.5 mm. long) is interpreted as the sternum. The features of the pectoral girdle are those of primitive proto- crocodilians and of two relatively advanced suborders of the Croco- dilia. The scapula is reminiscent of Broom's reconstruction of the ventral blade of Sphenosuchus and of that of Notochampsa Broom (1904, Haughton, 1924), while the coracoid is modeled upon that of the eosuchian Alligator redrawn by Colbert and Mook (1951). The ratio, coracoid length /scapular length, is 0.52 for Protosuchus, circa 0.96 for Alligator, and 0.60(?) for Platyognathus; the latter figure should probably be lower since the dorsal blade of the scapula is in- complete and would then approximate that for Protosuchus although the configuration of the coracoid is quite different. Breadth Thickness left right left right 21.4 ?23.3 ... 1.0 .... 6.2 ... 3.0 8.2 ... 6.0 15.6 16.6 15.0 13.2 7.0 ... 5.1 ... 5.1 .... 11.1 44 FIELDIANA: GEOLOGY, VOLUME 15 SCUTES The postcranial skeleton is variably sheathed with ornamented scutes. Dorsal armor occurs uniformly on all regions of the body as a double row of transverse scutes. The postsacral region bears smaller and more numerous unpaired lateral scutes which form a transverse series with the ventral armor, paralleling the dorsal series in distribution. Rounded pits and a pattern of radial rugosities orna- ment the dorsal series; the lateral and ventral plates are randomly pitted. Dorsal Series (figs. 8,B,E; 9, A) : The plates form a broadly rectan- gular overlapping series corresponding in number and position to the underlying vertebrae. These are broader than long, the ratios (length/ breadth) being 0.63, 0.51 and 0.73 for the cervical, dorsal, and ante- rior caudal regions respectively. The width of the plates, then, in- creases posteriorly and the ratio may approach unity in the last postsacral plates. The lateral margins are slightly rounded, and each scute bears a lateral longitudinal keel becoming more prominent toward the posterior margin. The outermost surfaces are depressed below this axis. Lateral Series: The lateral plates form a tightly overlapping series which is apparently restricted to the caudal region, and are keeled in like-fashion to the dorsals. The length /breadth ratio increases posteriorly, i.e., 0.58-0.75 in the anterior caudal and posterior caudal regions. A single pair of convex scutes (fig. 8,E) oriented obliquely and outward to the dorsal series is associated with an anterior ver- tebra of cervical series (CUP 2105) ; these contact the cervical ribs of the centrum immediately in front and may indiacte that the entire neck was armored. Ventral Series: The ventral scutes are paired and distributed like the dorsals although only the right series has been preserved in cer- vical series (CUP 2083). The fragment of the dorsal vertebrae (fig. 9,B) is covered ventrally by eight paired rows of four scutes each, but there are only two scutes in each transverse row in the caudal region. The plates lack lateral keels, are irregularly quadrate and suturally united but overlap somewhat in the posteriormost caudal region. The length /breadth ratio is rather uniform in all regions of the body, i.e., cervical, 1.0; dorsal, 0.75; caudal, 1.0. The configuration of the scutes in Platyognathus is similar to Pro- tosuchus although the ventral deflection of the border lateral to the keel is less acute and the ornamentations are dissimilar. The cer- vical scutes in Erpetosuchus differ in that they are longer than broad, I 1 = 10 mm Fig. 9. Platyognathus hsui Young. A, Presacral vertebrae and scutes viewed from above (CUP 2083). P>, Ventral aspect of presacral scutes (CUP 2083). Microchampas scutata Young. C, Vertebrae and ribs viewed obliquely from below (1) and the ventral scutes (2); (CUP 2085). D, Hand with claws, metacar- pals and phalanges (CUP 2086). 45 46 FIELDIANA: GEOLOGY, VOLUME 15 but the dorsal transverse series is similarly ornamented. The ven- tral scutes are apparently developed here more extensively than in Protosuchus where armor is preserved only from the lumbar region posteriorly. In toto, the armor pattern is somewhat reminiscent of aetosaurus-like pseudosuchians. CHARACTERS INDICATING A THECODONT HERITAGE The short skull and elements about the terminal confluent nares are patterned after Sphenosuchus. The armor images the rather specialized pseudosuchians although Sphenosuchus is itself without armor. Antorbital fenestrae are almost exclusively limited to the thecodonts although the exceptional protosuchian Pedeticosaurus presents a similar structure. The centra of the vertebrae lack hema- pophyses. CHARACTERS INDICATING A CROCODILIAN AFFINITY A flat snout, an edentulous and emarginated premaxillary-maxil- lary suture, dorsally oriented orbits, a primitive 'mesosuchian' palate and ventro-lateral choanae are characters to be found in primitive crocodilian groups. Confluent external nares are common to many mesosuchians and eosuchians. The symphysis of the mandible is elongate and includes the splenial. Eosuchian-like procoelous vertebrae compose the presacral ver- tebral column. The parapophyses migrate upward onto the neural arch and fuse with the diapophyses to form wide lateral processes. The coracoid is elongate and intermediate in length between the protosuchian and crocodilian pattern with respect to the scapula. The features of the skull make it valid to assume that Platyog- nathus was most probably derived from a sphenosuchid-like pseudo- suchian. The line of descent of the crocodiles was evidently very plastic, and in three areas of the world, i.e., North America (Proto- suchus), South Africa (Sphenosuchus) and China, the pseudosuchia and protosuchia were developing along a quite definite crocodilian plan. F. von Huene (1925a) and Broom (1927) have outlined sev- eral schemes depicting the steps in the development of the crocodiles from a generalized sphenosuchid-like ancestor. Broom's scheme ap- pears to be the more reasonable interpretation, since he removes rather specialized forms such as Pedeticosaurus, Aetosaurus and Erpe- tosuchus from the line of direct descent toward Notochampsa (Erythro- SIMMONS: NON-THERAPSID REPTILES 47 champsa) ; in accordance, therefore, Platyognathus may be interpreted to have been independently developed in the same fashion as Pedeti- cosaurus which, as Colbert and Mook point out, is so different from other South African protosuchians that it had better be considered a pseudosuchian. The suite of characters which define Platyognathus bespeak a taxo- nomic position intermediate between the Pseudosuchia and Proto- suchia. If, however, the greater emphasis is placed upon the more conservative features of the skull rather than upon the progressive nature of the postcranial skeleton, and these are considered to be specialized responses to an aquatic habitat, Platyognathus will prop- erly be seen to be a pseudosuchian. It is recommended that the present form be placed in a new fam- ily of pseudosuchia, the Platyognathidae (fam. nov.) which has the following characters: Skull: triangular with a short, broad snout; confluent terminal nares; orbit upwardly placed; a small antorbital fenestra; a secondary palate of a mesosuchian type; a long symphy- sis. Vertebrae: dorsal and caudal series procoelous. Pectoral Girdle: coracoid very elongate. Armor: scutes broader than long. Order Thecodonta Suborder Pseudosuchia Family Incerta sedis Holotype. — A fragment of the right ilium and proximal end of the femur. Catalog No. CUP 2084. Horizon and Locality. — Ta Ti, Dark Red Beds, Lower Lufeng Series. description (Table 12, Fig. 10,Al-2) Ilium: The crest is straight, narrow, and prolonged anteriorly; the bone is fractured just back of the ischial process but if a posterior extension were also developed, it would most probably have been modest. The roof of the acetabulum is expansive anteriorly to pro- vide a broad, deep, articular surface for the head of the femur. The lateral surface bears an oblique ridge which passes antero- ventrally to the acute and somewhat upturned border of the supra- acetabular buttress. The buttress is strongly developed anteriorly, but gradually merges with the bone above the ischial process. The ischial process limits the acetabulum posteriorly (fig. 10, A), projects downward and outward, and bears a demilunar facet. 48 FIELDIANA: GEOLOGY, VOLUME 15 Table 12.— MEASUREMENTS OF THE PSEUDOSUCHIAN, CUP 2084 (in mm.) Ilium Length of the crest 37.0 Height of the crest above the mid-acetabulum 13 .0 Height of the ilium at the ischial process 30.7 Breadth of the acetabulum, from the tip of the ischial process to the lip of the supra-acetabular buttress 20.0 Height of the acetabulum above the obturator 15.0 Facet of the ischial process: length, breadth 11.0, 6.0 Femur Head: proximal dorso ventral breadth 14 .0 Anterior-posterior breadth of the proximal articular suface 22.3 Breadth of the proximal articular surface behind the head 10.6 Neck: anteroposterior, dorsoventral 11.0, 12.0 Ridge systems: proximal breadth of the posterior, ventral 11.4, 12.0 The medial surface features a pair of broad crescentic depressions above the ischial and pubic processes. These are separated by about 7.0 mm. above the obturator which is open and presents a finished surface. Femur (fig. 10,A3-5) : The femur was displaced during preserva- tion, and made contact with the posterior part of the iliac crest. The head is large and divided by a slight dorsal constriction 17-18 mm. backward, appearing then biconvex; the proximal anterior surface is gently concave. A pair of ridges, the posterior ridge and the prox- imal anterior branch of the ventral ridge are separated by a strong intertrochanteric fossa, and pass diagonally downward and anteriad across the ventral surface of the shaft. The bone is hollow and thin- walled. The taxonomic position of the fragments is indefinite. The con- figuration of the ilium does not approach that of either the Protero- suchia or of the ornithosuchian pseudosuchians where the primitive triradiate pelvis is a solid plate-like structure having a continuous ventral symphysis. The obturator in the present form is acute and finished indicating that the space was not filled with cartilage. The overall proportions of the ilium of the stagonolepid Typothorax meadei Sawin (1947) are comparable to the present form, and the bone shows a prominent supra-acetabular buttress, an incipient sub- acetabular foramen and downturned ischial processes. The develop- ment of the proximal biconvexity and of the ridge systems in the femur is intermediate between Typothorax and Episcoposaurus hor- ridus Cope (Huene, 1915), although the incompleteness of the remains limits comparison. A more definitive analysis of the relationships of SIMMONS: NON-THERAPSID REPTILES 49 CUP 2084 to the stagonolepid group might be possible if the distal shaft were known for, in these, the proximal and distal condyles are perpendicular to each other. It is probable that the remains are not those of the poorly known parasuchian Pachysuchus imperfectus Young (1951). Even though the ilia of other phytosaurs are low and prolonged anteriorly, the pubo-ischiatic suture is incompletely ossified and filled with cartilage. The nature of the femur affords a more likely comparison to the stagonolepid group. Order Crocodilia Suborder Protosuchia Microchampsa scutata Young (1951) Microchampsa scutata Young (1951) Holotype. — An articulated fragment of eleven vertebrae and ribs covered by three longitudinal rows of dermal armor. Horizon and Locality. — Dark Red Beds, Lower Lufeng Series, Huangchiatien, Lufeng, Yunnan. Diagnosis (Young, 1951, p. 31). — "Thecodontia with a single row of dorsal scutes closely arranged on each row of the lateral scutes. Presence of ventral scutes doubtful but probable. Dorsal vertebrae short and stout, becoming slightly elongated and constricted poste- riorly. Ribs are double headed, only slightly curved. Size very small, being about one-half that of Erythrochampsa longipes Br." DESCRIPTION The dorsals are amphiplatyan, stout and short (1.1-2.0 mm.) anteriad becoming slightly longer and constricted posteriad (2.1- 3.0 mm.). The parapophyses are small and become indistinct by the fifth centrum. Four of the seven ribs exposed on the left side had distinct capitula and tubercula which tended to lie closer to- gether backward along the column. The last centrum was cited as a lumbar. The ribs are short and only slightly curved. The dermal scutes become shorter posteriorly and correspond in number to the vertebrae. A single median dorsal row, which lacks pittings and other ornamental rugosities, is similarly constructed and adjoins broader unpaired rows of dorsolateral scutes at almost right angles. The fragment measures about 21.0 mm. long, 11.0 mm. high, and 7.0 mm. broad at the middle part. 50 FIELDIANA: GEOLOGY, VOLUME 15 ADDITIONAL MATERIAL Material. — A fragment of seven consecutive vertebrae and five ribs with dermal armor, Catalog No. CUP 2085. Bones and claws of the hand, Catalog No. CUP 2086. Horizon and Locality. — Ta Ti, Dark Red Beds, Lower Lufeng Series, Lufeng, Yunnan. Amended Diagnosis: An armored reptile. Three rows of scale-like dorsal scutes — a single median dorsal row adjoining unpaired bilat- eral series. Ventral scutes present. Thoracic vertebrae short, be- coming elongate and sharply compressed posteriorly; anteriormost ribs doubly headed, free and slightly curved, becoming fused to the lateral armor in the lumbar region and articulating with broad trans- verse processes. DESCRIPTION Vertebrae and Scutes, CUP 2085 (fig. 9,Cl-2): The fragment ap- pears to be from the lumbar region and is twice as large as the type, V87 (length, 39.0-40.0 mm.; height, 17.0-18.0 mm.; breadth, 18.0 mm. at midlength). Ventral armor plates arch over the left side of the vertebral column and cover the median dorsal and left lateral scutes. The specimen must have turned 180° during preservation. The right side was denuded of protective armor and covered with an in- durated, whitish waxy material of a color and texture approximating that of the underlying bone which was difficult to prepare — perhaps due to the surface solubility under pressure tending to fuse and there- by obscure the boundaries of the elements. The better preserved features of the left side were rather easily prepared from a soft red- dish matrix. Lumbar Vertebrae, CUP 2085 (Table 13) : The column is arched. The centra are amphiplatyan, ventrally furrowed and deeply incised posteriorly in the series. The costal articulations of the right side are fused, broad, dorsoventrally convexoconcave and lie on the neural arch. The neural spines are broad, flat and terminally expanded to ac- commodate the median dorsal row of scutes. Measurements of the height (Table 13) are only approximate due to crushing. The zyga- pophyseal articulations are obscured by the crowding of the vertebrae along the curvature of the column, but they are probably rather flat as in other proto-crocodilians. SIMMONS: NON-THERAPSID REPTILES 51 Table 13.— MEASUREMENTS OF MICROCHAMPSA SCUTATA (in mm.) Specimen No. CUP 2085 Specimen No. CUP 2087 Position in Series 1 2 3 4 5 6 7 Position in Series 1-2 2-3 3-4 Lumbar Vertebra Ventral Length 3.7 3.5 4.2 4.8 ?4.2 ?4.5 *A.H. 5.0 5.9 5.5 5.6 5.6 5.6 Right Side L.Tp. B.Tp. H. 4.0 4.0 4.6 4.6 4.0 ?3.6 5.5 5.2 5.0 5.5 6.4 5.1 17.0 16.7 16.0 16.7 ?15.0 ? H.Ns. 6!6 6.0 ?4.7 5.5 ?4.2 Caudal Vertebra Distance between the tips of the Transverse Processes ?2.0 2.5 * Anterior height Transverse process length Transverse process breadth Height of the vertebra Height of the neural spine Caudal Vertebrae, CUP 2087: The fragment consists of the neural arches and scutes of five vertebrae. The transverse processes are short and narrow, and contact the edges of the median dorsal scutes. The posterior zygapophyses are rather obliquely oriented (fourth vertebra) . Ribs, CUP 2085: The dorsal ribs of the type were double headed and free. The proximal shafts of the lumbar ribs apparently fuse to the anterior margin of the lateral scutes, and the heads persist as linear exostoses converging to form the shaft 2-3 mm. from the outer border of the armor; the tubercular relief is the higher. The features are best seen in the third and fourth ribs in the series. The shafts clearly extended beyond the end of the scutes but become structurally weaker posteriorly. Ventral Armor, CUP 2085 (Table 14) : Young's speculation that Microchampsa had ventral armor is confirmed. Four transverse rows of four scutes each overlap the mid-dorsal and lateral series, but are only unilaterally represented. The plates which would have con- tacted the midline are broken while the outermost plates have been displaced and overlap slightly. These are rather squarely fashioned (6.0 mm.2), but the lateral borders of the outermost plates are rounded 52 FIELDIANA: GEOLOGY, VOLUME 15 anteriorly and finished, indicating that the armor ended at this point and did not extend over the sides of the body. Dorsal Armor, CUP 2087 (Table 14) : Four scutes from the mid- dorsal and right lateral series are exposed on a caudal fragment. The laterals contact the under surface of the dorsals at an oblique angle, and are broader anteriorly than posteriorly in the thoracic column; in the caudal region, the dorsals are the broader (dorsals, 6.2 mm.; laterals, 3.0-4.0 mm.) and become shorter and more closely fitting posteriorly. The dermal armor is simply pitted and has orna- mental rugosities. Table 14.— MEASUREMENTS OF THE DERMAL ARMOR OF MICROCHAMPSA SCUT AT A (in mm.) Position Mid-Dorsal Lateral Series Specimen Along No. Column Series L.1 B.2 left L. B. right L. B. CUP 2087 1 ?2.0 ?5.4 2.9 3.3 2 2.2 5.6 2.7 5.4 3 2.5 6.1 2.0 3.5 4 2.4 1.3 XL.= Anteroposterior length 2B.=Lateromedial breadth APPENDICULAR SKELETON Hand, CUP 2085 (fig. 9,D) : The association of the specimen with the vertebral fragment and its small size figure importantly in the tentative assignment of the complex to Microchampsa. The fragment consists of two gently curving claws subequal in length, sharply tipped and grooved to receive the horn, and rather elongate, slender phalanges of equal length which have large condyles — probably the third and fourth or the fourth and fifth digits. The absence of skull materials limits the definition of Micro- champsa since the dermal armor pattern is most atypical with respect to the stem reptiles. The rectangular scutes are of a protosuchid type, but the dorsal plates of the Notochampsidae are paired and there is no indication that the microchampsid pattern originated by the simple fusion of plates to either side of the midline — absence of lateral keels. Moreover, the dorsal armor in Notochampsa istedana (Haughton, 1924) gradually becomes quadrangular posteriorly while the rectangular configuration in Microchampsa is retained in the caudal region and it is to be expected that the obscured lumbars were similarly fashioned. The pattern approaches most nearly the poste- SIMMONS: NON-THERAPSID REPTLIES 53 rior caudal scutes of Protosuchus and Pedeticosaurus which are quad- rate or become even slightly longer than broad ; the number of plates in each transverse ventral row also decrease toward the sacrum. Supernumerary ventral plates relative to the dorsals, however, dis- tinguish other protosuchians (2:1) from Microchampsa (1:1) although the caudal regions are similarly encased in dermal armor; the devel- opment, while not as great as that of the more specialized pseudo- suchians such as Aetosaurus, surpasses that of the more conservative thecodont group. The vertebral structure is most crocodilian; the centra resemble Protosuchus and related forms. The expansive spines are particu- larly like those of Pedeticosaurus and there is a tendency for the fusion and establishment of the costal articulations on broad, flat, transverse processes in the lumbar region of the body. The proto- suchian rib pattern is virtually unknown, however, the doubly flanged shafts of the thoracic ribs of Protosuchus are not found in either Microchampsa or in modern crocodilia where only anterior flanges persist. It is possible, nonetheless, that the lumbar ribs and scutes of Microchampsa were flanged and free in young animals, i.e., in the type, but became fused during ontogeny. The available structures are insufficient to establish the system- atic position with any certainty, but the most reasonable assignment appears to be to the Protosuchia, recognizing that it apparently rep- resents a new type of primitive crocodilian. Order Saurischia Suborder Theropoda Infraorder Coelurosauria Family Podokesauridae Lukousaurus yini Young (1948b, 1951) Original Diagnosis.- — "Skull rather small. Nasal opening small and anteriorly situated. Preorbital opening triangular and high. Orbits well rounded. Muzzle looks slender. Lower jaw slender with the lower margin straight. Premaxilla bears five teeth and maxilla about ten teeth. All the teeth are sharply compressed and pointed backward. Posterior side with fine and short serrations." Lukousaurus is the only carnosaur known from eastern Asia and the only small saurischian of the Lufeng assemblage. The original 54 FEILDIANA: GEOLOGY, VOLUME 15 diagnosis of the genus was based on materials from the Dark Red Beds at Huangchiatien — the anterior part of a skull and jaws (V23), a tooth (V263) and the proximal part of a right humerus (V261). Recently (Young, 1951), three additional fragments from the Dull Purplish Beds at Shawan (V77) were referred to this genus, the distal part of a right humerus and an upper middle part of a co-ossified tibia and fibula. The present collection includes an additional and similarly pre- served skull fragment with a more complete jaw than the type (V23) . Additional Material. — An anterior fragment of a skull and jaws with teeth. Catalog No. CUP 2092. Horizon and Locality. Hei Koa Peng, Dark Red Beds, Lower Lufeng Series. Diagnosis. — Same as for the Type V23 (Young, 1948b). description (Table 15) The skull is badly damaged. A fragment of the left premaxilla lies at the snout while the nasal bones are displaced to the right side. Posteriorly the roofing bones are limited to a small fragment of a sus- pect prefrontal. The left wall shows the anterior angle of the pre- orbital fenestra; the position of the external nares is only indicated since the borders are crushed and the orbits are ill-defined. The outer wall of the left maxilla is broken away. A short anterior fragment of the left jaw bears the roots of two teeth. The right mandible is transversely fractured and crushed pos- teriorly, but indicates the presence of a small, high, lateral fenestra; its internal architecture is exposed medially by the loss of the splenial which is intact only at the symphysis. The prearticular is seen to project anteriorly. The surangular-angular contact is indistinct due to the checkering of the bone. The angular is a large element which overlaps the smaller superior surangular and wraps about the infe- rior surface of the jaw. The tips of the second -fifth teeth of the left maxilla are imbedded in the matrix of the lower jaw; the fifth is completely exposed (11.0 mm.). The sixth and seventh exist as hemi-root fragments adhering to the inner wall of the maxilla. Roots of the eighth and ninth, with perhaps the crowns of their replacement teeth, are simi- larly contained while two large, badly checkered fragments — (?) lowers — lie above them. A series of four small teeth (terminals) occur just forward of the orbit. SIMMONS: NON-THERAPSID REPTILES 55 On the right side, a single premaxillary tooth (?fifth) lies internal to the space marking the suture with the maxilla, and is followed by the first maxillary tooth with its replacement. The first-third max- illary teeth lack crowns and the root of the second lies at the alveolar margin. The sixth-eighth and ?twelfth teeth become progressively smaller. As noted by Young, the jaw could not have supported more than 12 teeth. The tight apposition of the right maxilla masks the mandibular dentition. The short fragment of the left mandible exhibits the roots of three teeth mediad to the fourth maxillary tooth. Table 15.— MEASUREMENTS OF THE SKULL AND JAWS OF LUKOUSAURUS YINI (in mm.) left right Preserved length of the skull: upper 29.0 51 .6 lower 56.0 51.6 Preserved length of the jaw 63 .0 19 .4 Anterior height of the skull (excluding nasals) 10.6 13.0 Posterior height of the skull fragment 17.0 17.0 Length of the tooth row 37.0 56.6 Preserved length of the premaxilla 8.0 Remains Referrable to the Podokesauridae Catalog No. CUP 2089 Material. — An articulated fragment of a left femur, tibia and fibula with isolated fragments of metacarpals and phalanges. The bones were associated with the skeleton of Dibothrosuchus, but are quite distinct from other known pseudosuchia. Moreover, on the basis of the proportions of the appendicular skeleton among the Ornithosuchidae, where the acquisition of bipedalism parallels the reduction of the forelimb relative to the more massive hindlimb, the femur is as slender and equally as long as the humerus of Dibo- throsuchus. The histologic structure of these bones is similar and at apparently the same level of maturation, but morphologically identical fine structure is common among unrelated groups during their ontogeny. description (Fig. 10,B1^4; table 16) Femur: The shaft is long and curved; the dorsal and ventral sur- faces are normally angular in places, perhaps indicating the origins and insertions of the extensor and adductor muscles. The distal con- 56 FIELDIANA: GEOLOGY, VOLUME 15 dylar area is triangularly expansive and bent backward; the corre- spondingly formed intercondylar fossa is very deep and broad for the insertion of an apparently powerful M. quadriceps to extend the thigh, while the popliteal space is smooth, shallow and longitudinally cleft. The surface of the lateral condyle is smooth; the medial con- dyle is bifurcated. The bone is hollow and thin- walled. The rectangular articular surface and straight, slender shaft com- pare favorably to Coelophysis longicollis Cope (1887; Huene, 1906, 1915). The shortness of the condyles approaches the condition in Podokesaurus holyokensis Talbot (1911) but their ventral disposition is here less prominent. Small condyles are also reported for Haltico- saurus longotarsus Huene (1932), but this bone was not figured. Tibia: The proximal articular surface is roughly triangular and divided by a shallow groove. The convex, dorsal, extensor surface bears a triangularly broad and low cnemial crest for the insertion of the extensor muscle below the anterior femoral condyle. The shaft is broad anteroposteriorly and moderately concave ventrally. Table 16.— MEASUREMENTS OF THE APPENDICULAR SKELETON OF CUP 2089 (in mm.) Femur Preserved distal length (lateral, medial) 55.6, 56.4 Maximum condylar breadth (lateromedial) 21 .0 Maximum breadth above the condyles (lateromedial) 15.0 Maximum breadth at the most proximal end (lateromedial, antero- posterior) 7.4, 8.3 Depth of condyles (anteroposterior) : anterior condyle (fibula) 13 .3 posterior condyle (tibial) ?16 . 7 Depth above the condyles (anteroposterior) : anterior condyles (fibula) 13.0 posterior condyles (tibial) 9.0 Depth of the intercondylar fossa (lateral border, medial border) 6.6, 5.4 Proximal-distal height of the posterior condyle 10.0 Tibia Preserved proximodistal length (medial, lateral) 33.7, 31.6 Breadth of the proximal articular surface 18 .7 Maximum depth 13.0 Distal width (lateromedial) 8.0 Distal thickness (anteroposterior) 6.0 Fibula Preserved proximodistal length (medial) 30.0 Proximal breadth (lateromedial) 14 .0 Proximal depth (anteroposterior) 5.0 Breadth of the neck (lateromedial) 8.0 Depth of the neck 6.0 SIMMONS: NON-THERAPSID REPTILES 57 Tibias of Triassic coelurosaurs are poorly known. A straight, narrow shaft is reported for Dolichosuchus cristatus Huene (1932) and Podokesaurus; the first is twice as large as the present form and develops a strong pair of dorsal ridges (cnemial crest), but their mid- length cross sectional areas are comparable. While the tibia of Co- elophysis has a small head, its shaft bears a low, lateral tuberosity. Fibula: The fibula is very slender, subcircular in cross section and the head is sharply set-off from the shaft by a constriction. Fibulae of coelurosaurs, while little known and poorly preserved, are commonly thin, and may be straight (Podokesaurus) or curved (Saltopus elginensis Huene, 1910). The expansion of the head is a distinctive feature, but weak in Coelophysis. Hand (metacarpals and phalanges): The metacarpals are long and slender. The elements of the proximal and distal fragments are straight, round in cross section, and widely spaced although the con- dyles are contiguous. The phalanges are known from a small joint fragment. The hands of coelurosaurs are incompletely known as well. The bones of Procompsognathus (Huene, 1921) are a third smaller, lie closer together and are more heavily constructed. The apparent large size of the hand in the present specimen may be similar to Podo- kesaurus; the elongation and slenderness of the metacarpals and the reduced size of the phalanges (?claws) may have been of advantage to a form purportedly adapted for grasping like that of Coelophysis. Catalog No. CUP 2090 Material.- — The proximal fragments of two cojoined metatarsals. The bones were associated with the remains of Dibothrosuchus, but in view of their slenderness, a like-assignment is unwarranted. Moreover, these were not modeled upon the Hesperosuchid-type. Horizon and Locality. — Ta Ti, Dark Red Beds, Lower Lufeng Series. DESCRIPTION The fragments are entirely fused (fig. 7,E). The free condylar border of the larger bone (length, 42.4 mm.) is sharp. The breadth of the proximal and distal ends of the cojoined elements is 16.4 mm. and 11.0 mm., respectively. Metatarsals are rarely preserved in coelurosaurs. Avipes dill- stedtianus Huene (1932) is known solely from these bones which are I 1= 10 mm | (« 10mm. 58 SIMMONS: NON-THERAPSID REPTILES 59 proximally fused for at least half their length and bear acute con- crescent edges, but it is not known if they similarly diverged distally in the present example. Fusion of bones is doubtful in Podokesaurus and Saltopus, but entirely lacking in Procompsognathus and Hallopus. Catalog No. CUP 2091 Material. — The distal end of a right femur. Horizon and Locality. — Ta Ti, Dark Red Beds, Lower Lufeng Series. DESCRIPTION The shaft is rather triangular in cross section above the distal condyles and is preserved for a short distance (fig. 10,C). The condyles are subequal and strongly rotated to the ventral side; the lateral articular process for the fibula is upturned at its apex and is the longer and more massive; the shorter medial condyle for the tibia is down turned. The narrow superior crest of the lateral condyle is continued as a ridge on the shaft while the bone is broadly rounded above the median condyle. The ventral intercondylar space is tri- angular and sharply depressed, but viewed from below it appears narrow and shallow. The dorsal and inferior surfaces of the frag- ment are relatively flat and smooth and lie at right angles. The fragment appears closer in nearly every respect to the simi- larly preserved Walgettosuchus woodwardi Huene (1932) from the lower Cretaceous, New South Wales, Australia, than to any of the carnosaurs or prosauropods I know of where condylar inequality is less marked and the intercondylar sulcus is deeper, i.e., Gresslyo- saurus ingens, Plateosaurus gracilis, P. plieningeri Huene (1932). The proportions of Gresslyosaurus ingens are similar to the pres- ent form, but in toto, these forms are more massive. Conclusions.— The hind limb (CUP 2089), femur (CUP 2091) and foot bones (CUP 2090) rather clearly demonstrate coelurosaurian affinities, and particularly to the common upper Triassic family, the Podokesauridae. Generic identity has not been accorded these speci- FiG. 10. Pseudosuchian (CUP 2084). A, Ilium, medial (1) and lateral (2) sides. Femur, proximal fragment viewed from the ventral (3), dorsal (4) and articular surfaces (5). Coelurosaur (CUP 2089). B, Femur, distal fragment articulating with prox- imal fragments of the tibia and fibula (1, 2). 3, phalanges. 4, metacarpals. Coelurosaur (CUP 2091). C, Femur, distal condyles viewed from the medial (1), ventral (2), and articular (3) surfaces. 60 FIELDIANA: GEOLOGY, VOLUME 15 mens since it is felt that additional finds might be of more satisfac- tory diagnostic value. It is hoped that future collections from the Lufeng Basin will disclose more complete forms with which these fragments can be associated. Table 17.— MEASUREMENTS OF THE FEMUR OF CUP 2091 (in mm.) Lateromedial breadth at the condylar surfaces 21 .0 Transverse cross section length at the break 13. 7 Sagittal cross sectional length at the break 12 .4 Sagittal thickness of the lateral condyle 23.4 Transverse thickness of the lateral condyle 8.2 Sagittal thickness of the medial condyle 15.0 Maximum separation between the condyles 12.0 Lateral length of the fragment 34 . 3 Medial length of the fragment 33 .4 Suborder Prosauropoda Gyposaurus sinensis Young (1941a, 1948a, 1951) Original Diagnosis. — "Small Prosauropoda, about the size of Gy- posaurus capensis. Teeth of general Thecodontosaurus-type with anterior and posterior coarse serrations. Vertebrae number prob- ably: neck 10, dorsal 15, sacrum 3, and tail ?. Anterior limb ex- tremely short, scapula vigorously bending at the base of the shaft. Pelvic girdle rather slender with the ilium high. Both hand and foot with distinct lateral reduction." The collection made by Young represents about 14 individuals of which the majority are unquestionably identified. V45 and V46 from Huangchiatien were perhaps considered immature individuals of Lukousaurus or Gyposaurus since they demonstrated some ana- tomical differences from known comparative materials. Remains Referred to G. sinensis Young Element Third-fourth cervical vertebral centrum. Neural apparatus, anterior dorsal vertebra. Cojoined mandible and maxillary. Scapula, distal end. Maxillary fragment, left. Mandibular symphysis and ramus including teeth; a premaxillary fragment with teeth; cojoined upper and lower jaw fragment with teeth. Catalog Locality No. CUP Unknown 2021 Hei Koa Peng 2027 Hei Koa Peng 2028 Hei Koa Peng 2029 Hei Koa Peng 2026 TaTi 2006 SIMMONS: NON-THERAPSID REPTLIES 61 TaTi 2007 TaTi 2008 TaTi 2009 TaTi 2010 TaTi 2011 TaTi 2012 TaTi 2013 TaTi 2014 TaTi 2015 TaTi 2016 TaTi 2017 TaTi 2018 TaTi 2019 2020 TaTi and 2023 TaTi 2024 TaTi 2025 Neural spine of a cervical vertebra. Fourth metatarsal. Left ulna and radius, proximal end Vertebral centra, sacrals I and II. Left pubis, proximal end. Right humerus, distal end. Left femur, distal end of a juvenile. Right radius, proximal end. Right pubis, distal end. ?Pubis, distal end of a juvenile. Fourth metatarsal, proximal end. Right ulna, proximal end. Left ulna, proximal end. Neural apparatus, cervical vertebra. Vertebral centra, sacral ?I. Left fibula, proximal end. Right fibula, distal end. Lufengosaurus hueni Young (1941b, 1947b, 1951) Original Diagnosis. — "From large to gigantic size Plateosauridae, Prosauropoda. Skull small. Teeth with short crown and long, strong root. They are weakly spatulated and with coarse anterior and posterior serrations. Vertebrae number: neck 10, dorsal 14?; sacrum 3, and caudal 45 or more. Pectoral girdle with well-ossified sternum. Pubes weakly constructed. Tibia comparatively short. Mt. Ill is rather long. Anterior limbs are very short. Hand and foot with the first claw are especially strong." Locality Yang Tsao Ti (Loc. No. 3) Hei Koa Peng Hei Koa Peng Hei Koa Peng Hei Koa Peng Hei Koa Peng Hei Koa Peng Hei Koa Peng Hei Koa Peng TaTi TaTi TaTi Remains Referred to L. hueni Young Catalog No. CUP Element 2071 2062 2063 2064 2065 2066 2067 2068 2069 2070 2059 2060 2061 Claw Claw, proximal end. Claw, proximal end. Right and left maxillary fragments, with the ?premaxilla. Tibia and Tibiale, distal end. Haemal arches. Basisphenoid. Right anterior maxillary fragment with teeth. Right mandibular ramus with teeth. Jaw fragment with teeth, left. Tibia and tibiale, distal end. Neural apparatus, dorsal vertebra. 62 FIELDIANA: GEOLOGY, VOLUME 15 Lufengosaurus magnus Young (1947b, 1951) Original Diagnosis. — "Size one third larger than L. hueni, Ver- tebrae robust, Centrum large and short. Scapula strongly ben ted. Humerus-radius extremely short and massive. Hand short. Pelvic girdle massive, ilium long, Ischium comparatively weakly built. Pubis long and slender with thickened distal end. Tibia relatively long as compared with the femur. Mt. Ill one half of the length of the tibia. Both the hand and the foot are built in the same way as in the case of L. hueni. Anterior leg is much shorter as compared with the whole posterior leg." Remains Referred to L. magnus Young Element Jaw fragment with teeth. Jaw fragment with teeth. Left fibula, proximal end. Jaw fragment with teeth. Proximal end of right ulna. Crown of tooth. Neural spine, caudal vertebra. Jaw fragment. Catalog Locality No. CUP Unknown 2052 Unknown 2053 Hei Koa Peng 2054 Hei Koa Peng 2055 Hei Koa Peng 2107 TaTi 2056 TaTi 2057 TaTi 2058 Remains Referrable to Lufengosaurus sp. Locality No. CUP Element Yang Tsao Ti 2072 Haemal arch. Hei Koa Peng 2073 Tibiale. Hei Koa Peng 2074 Jaw fragment with teeth. Hei Koa Peng 2075 Jaw fragment lacking teeth. Hei Koa Peng 2076 Premaxillary fragment, right TaTi 2077 Premaxilla, right. TaTi 2078 Haemal arch. TaTi 2079 Centrum, caudal vertebra. TaTi 2080 Two metatarsals. Yunnanosaurus sp. Young (1942, 1951) Original Diagnosis. — "From the middle to rather large sized Pro- sauropoda with the skull proportionately rather long, more than 3.5 the length of the anterior caudal vertebrae, ca. 4. Nasal opening small, anterior orbital opening rather large. Orbit rounded and rep- SIMMONS: NON-THERAPSID REPTILES 63 resents the largest opening of the skull. Upper temporal opening is bean-shaped and can be seen mostly from the lateral aspect of the skull. Lower jaw slender with small dental foramen. Teeth with long pointed crown, distinctly spatulated in primitive Sauropod fash- ion. Serrations are poorly developed. Vertebrae robust. Scapula is strongly bending above the proximal expansion. Sternum is ossi- fied and elongated. Humerus robust and shorter than the scapula. Hand with strong lateral reduction. Ilium small with the iliac border low. Ischium slender. Pubis rather thin. Femur and tibia massive, the latter is only somewhat shorter than the former. Fibula rather straight. Foot similarly built as that of Gyposaurus." Remains Referrable to Y. huangi Young (Young, 1942, 1951) Element Six or seventh cervical vertebra. First metacarpal, left. left femur, proximal end. Left fibula, proximal end. Left femur, distal end. Third and fifth metacarpals. Right fibula, proximal end. Distal ends of two fibulae. Two centra, caudal vertebrae. Skull and jaws, nearly complete, juvenile. Maxillary and mandibular fragments. Remains Referrable to Y. robustus Young (1951) Catalog Locality No. CUP Unknown 2039 Hei Koa Peng 2030 Hei Koa Peng 2031 Hei Koa Peng 2032 TaTi 2033 TaTi 2034 TaTi 2035 TaTi 2101 TaTi 2036 TaTi 2037 TaTi 2038 Catalog Locality No. CUP Element Unknown 2040 ?18th caudal vertebral centrum. TaTi 2041 Two dorsal vertebrae (seven and eight?) TaTi 2042 Left maxillary. TaTi 2043 Right maxillary. TaTi 2102 Ilium. Hei Koa Peng 2046 Vertebral centrum. Hei Koa Peng 2047 Tooth fragments. Hei Koa Peng 2048 Claw fragment. Hei Koa Peng 2049 2044 j Centrum, sixth or seventh dorsal vertebra. Hei Koa Peng and | 2045 Caudal vertebrae (2). 64 FIELDIANA: GEOLOGY, VOLUME 15 Remains Referrable to Yunnanosaurus sp. Hei Koa Peng 2050 Tooth fragment. TaTi 2099 Ilium fragment. TaTi 2100 Ilium fragment. Unknown 2051 Two teeth. Infraorder Carnosauria Family Teratosauridae Sinosaurus triassicus Young (1948b, 1951) Original Diagnosis. — "Carnosauria of gigantic size, Maxilla high. Teeth with long massive root, compressed sharply pointed and curv- ing backwards. Both the anterior and posterior edges are marked by fine serrations." A number of articulated neck, dorsal, sacral and caudal vertebrae strongly built but comparatively slender to Lufengosaurus magnus; part of the anterior and posterior limbs massively constructed but the proportions of the appendicular skeleton are similar to Yunnano- saurus sp. Hand short and broad. Additional Remains Referred to Sinosaurus triassicus Young Catalog No. CUP 2097 Material. — A fragment of the anterior part of the jaws with teeth ; two isolated teeth. Horizon and Locality. — Ta Ti, Dark Red Beds, Lower Lufeng Series. DESCRIPTION The rami have been transversely fractured behind the symphysis and disoriented during preservation (fig. 7,D). The alveolar bor- der is acute and forms a narrow ledge above a rather thick but flat, lateral wall. The right jaw fragment is the shorter. The splenial sheaths the inner surface of the jaw at the symphysis and the dentigerous border, and limits the meckelian groove which passes forward below and internal to it. The left jaw holds eight-nine teeth which lack crowns. The right, in addition, possesses two large and recurved, narrowly compressed broad teeth which are finely denticulated on their back edges. The isolated teeth are flattened with longitudinal median depressions (ref. V35, V48- Young, 1948b). SIMMONS: NON-THERAPSID REPTILES 65 The referral of the jaw and teeth to S. triassicus is problematical since similar materials are unknown for the type. The teeth vary from known specimens in their greater compression and lack of ser- rations on the anterior trenchant edges. Remains Referrable to S. triassicus Catalog Locality No. CUP Element Unknown 2095 ?Centrum; ?third dorsal vertebra. Unknown 2096 Tooth fragments (3). Hei Koa Peng 2001 Teeth (24). Hei Koa Peng 2002 Teeth (2). Hei Koa Peng 2003 ?Centrum; ?third dorsal. TaTi 2004 Teeth (4). TaTi 2005 Teeth (3). TaTi 2098 Centra; one cervical, two dorsals. Order Ornithischia Suborder Ornithopoda Infraorder Hypsilophodontidae Diagnosis. — An ornithischian of small size. Mandible low ante- riorly, slender and tapered; anterior ventral border bends medially toward the symphysis; jaw higher and more convex posteriorly. Teeth thecodont, overlapping, relatively simple and increasing in size from front to rear. Dentary-predentary junction edentulous. Tatisaurus oehleri sp. nov. Holotype. — A fragment of a left mandible with teeth. Catalog No. CUP 2088. Horizon and Locality. — Ta Ti (Locality No. 8) from the Dark Red Beds of the Lower Lufeng Series. Diagnosis. — Same as for genus. DESCRIPTION (fig. 11,A,B,C) The ramus is low anteriorly, of slender proportions and broken in front and in back of the tooth row. Viewed from above, it is slightly sigmoid; from below, the inferior border is acute and bends antero- medially toward the symphysis. All sutural contacts have been obliterated by the extensive checkering of the bone. l = IOmm. Fig. 11. Tatisaurus oehleri (gen. et ep. nov.). Jaw viewed from the lateral (A), medial (B) and dorsal (C) surfaces in association with the head of a rib (B). (CUP 2088). 66 SIMMONS: NON-THERAPSID REPTILES 67 The jaw is biconvex, but the early oblique displacement of the dental series from the midline exaggerates the lateral curvature, while the alveolar border follows the conformation of the median wall, i.e., moderately S-shaped and laterally concave. The outer wall is weakly constricted at the level of the eighth-ninth teeth below which the curvature of the bone becomes sharper. Posteriorly, the curvature is greatly exaggerated and effects a plateau or shelf lateral to the teeth. The apex of curvature appears as a well-defined but low ridge which then becomes confluent with the shelf. The greatest breadth of the "shelf" is about 7.0 mm. and it perhaps indicates the life-presence of a cheek pouch. The dentary becomes progressively deeper posteriad, but the superior and inferior borders are parallel from the eighth or ninth- thirteenth teeth. The profile of the superior border is sharply ele- vated behind the tooth row but it is doubtful that it terminated in a coronoid process since the lateral ridge (definitive in all ornithischia possessing the element) is reduced and disappears at the cheek pouch. It is not known whether the postcoronoidal region is as short as in the more advanced ornithischia. It is difficult to distinguish the tiny foramina for nervous and /or vascular elements which are scattered over the extensively fractured and pitted outer surface. Foramina oppose the fifth and twelfth- fourteenth teeth, and with the exception of the first, penetrate the bone at right angles. A group of at least five foramina lies above the "ridge" at the lateral constriction and the fifth is the largest. The splenial sheaths the inner wall, but its sutural union with the more posterior jaw elements, if present, cannot be discerned. It covers the anteriormost limit of the large meckelian fenestra and groove which extends to the predentary. Like the saurischian dinosaurs, the dentigerous border is higher externally, but has a sigmoid configuration. The dentary holds 18 thecodont teeth which become larger posteriorly and are arranged in a uniformly overlapping series. Anteriorly, the dentary-predentary junction is edentulous. The functional teeth are worn and lack crowns. Unworn replacement teeth (five) lie on the inner surface. The dentition exhibits a slight degree of heterodonty. The roots are generally subcircular or slightly constricted lateromedially, but may manifest an inner swelling suggestive of a triangular cross sec- tion. The crown-root junction of the smaller anterior teeth is vari- ably constricted or expansive. 68 FIELDIANA: GEOLOGY, VOLUME 15 The crowns of the functional teeth are low, rather symetrically triangular with gently curved or evenly bevelled, smooth polished surfaces. The trenchant edges of the replacement teeth are coarsely denticulated, but unlike most ornithischians, the lateromedial sur- faces are not fluted. The anterior edge is the steeper; the apical denticle is pointed slightly backward and bordered by three-four marginal denticles (some with accessory cusps) . An anterior replace- ment tooth appears adenticulate, perhaps due to incipient wear since the mature teeth are small. The crown-root junctions are frequently swollen, probably in response to wear but none of the replacement teeth have definitive cingula. Table 18.— MEASUREMENTS OF THE JAW OF TATISAURUS OEHLERI (in mm.) height breadth Estimated length of the jaw (postcoronoid= one- third total length) 78 .6 Length of the alveolar border 52 . 4 Anterior end of the dentary 4.0 3.6 Level of: 8th-9th tooth 13.0 12.4 constricted lateral wall 8.1 13th tooth 13.4 12.5 18th tooth ?18.6 11.4 ?coronoid process ?15.6 .... Maximum breadth of the cheek pouch (lateromedial) 14 .4 Discussion and Relationships.- — Since Tatisaurus is the temporal equivalent of the earliest known ornithischian derived from upper Triassic Rhaetic Beds in South Africa, it is considered that the rela- tionships of the form can be drawn to the best advantage by a sys- tematic survey of the groups which appear to be related to it. HYPSILOPHODONTIDAE The fragmentary Geranosaurus atavus Broom (1908; Haughton, 1924) from the Stormberg Series is the earliest representative, but differs importantly from Tatisaurus in details of size and large, chisel- like teeth. The jaws of Hypsilophodon foxii Huxley (1870; Hulke 1882; Nopsca, 1905) and Thescelosaurus edmontonensis Sternberg (1940) are fashioned like Tatisaurus but taper forward only slightly and decrease markedly in height at the dentary-predentary contact. The coronoid process, indistinct in Tatisaurus is prominent in T. ed- montonensis and blunt in H. foxii. While their tooth structures are dissimilar, they are arranged in an overlapping series which arches inward, but the pattern is reversed from the fifth-eighth tooth in SIMMONS: NON-THERAPSID REPTILES 69 H. foxii. Similar to the progressive Nodosaurids, however, a splenial element occludes much of the meckelian groove. IGUANODONTIDAE The jaw construction, while hypsilophodontid (spout-like preden- tary-dentary contact and a short symphysis) shows a more prominent coronoid process and a flatter symphysis than Tatisaurus. The meckelian groove passes forward to approximately the same level, but an obtuse prominent ridge marks the outer surface of the jaw of Iguanodon foxii Owen (1874) which is correspondingly thickest where it is penetrated by a series of five foramina. The overlapping tooth row of Iguanodon is nevertheless not sigmoid but simply curved to effect the lateral concavity and the teeth are elaborately flutted. ANKYLOSAURIA The diagnostic features of Tatisaurus were initially interpreted as ankylosaurian. These forms, however, are strictly upper Creta- ceous and of relatively larger proportions. The recognized Chinese ornithosuchians are Ornithopoda (Young, 1944) and include no anky- losaurs. The Asian ankylosaurs (from Mongolia) are all referrable to the Nodosauridae.1 NODOSAURIDAE Unlike Tatisaurus, the jaws of nodosaurs are generally massive and invested in dermal bone which obscures the elements. Those of Edmontonia rugosidens Gilmore (Russell, 1940) and E. longiceps are more or less rectangular. Except for Pinacosaurus grangeri Gilmore (1933) where the dermal bone is reduced, a splenial element covers the meckelian groove (as in the Hypsilophodontiae) and the dentiger- ous border lies at the posterior two-thirds of the dentary after its 1 Since this paper went to press, Crompton and Charig (1962, Nature, 196, pp. 1074-1077) have published a preliminary description of a newly discovered ornithischian, Heterodontosaurus tucki, from the transitional beds of the Upper Triassic Cave Sandstones, South Africa. Compared with Tatisaurus, the jaw of H. tucki has a well-defined coronoid process and the dentition includes an anterior recurved caniniform tooth. While the general morphology of the cheek teeth (concave crowns) is quite different from that of Tatisaurus, the teeth in both speci- mens importantly lack cingula. Heterodontosaurus appears to be more closely related to the Hypsilophodontidae than Tatisaurus, but both seem to show inter- esting similarities to Upper Cretaceous forms. This further indicates that the main radiation of Ornithischia probably occurred prior to the Upper Triassic (Huene, 1948). 70 FIELDIANA: GEOLOGY, VOLUME 15 maximum depth has been attained. The slenderest jaws obtain in the Canadian Dyoplosaurus acutosquameous Parks (1924) and in Ana- dontosaurus lambei Sternberg (1928b), however, commensurate with the ornithopods, the decrease in height at the predentary-dentary contact is precipitous rather than gradual. Considerable intergroup variability exists both with respect to (1) the inferior border of the mandible — keeled (Edmontonia longiceps Sternberg, 1928a), acute (Panoplosaurus mirus Lambe, 1919) or flanged (Europlocephalus tutus Lambe Gilmore, 1923)— and (2) the predentary relationships- forms a yoke between the rami (P. mirus), merely caps the symphy- sis (Edmontonia spp., Paleoscincus sp.), or makes a minimal ingression between the cojoined rami (P. grangeri). The inward curvature of the dentary anteriorly is as strong in the cited forms as in Tatisaurus, but the incompleteness of preservation vitiates comparison. The coronoid element and postero-lateral ridge when present are pro- nounced in all but Ankylosaurus magniventris Brown (1908) and Panoplosaurus, while these features are only suggested in Tarisaurus. The dentigerous border of nodosaurs is rather simply curved and not sigmoid, but the displacement from the midline may be so ex- treme (E. longiceps) that it overhangs the infero-internal border at its midlength. While the number of teeth may vary from 11 (E. rugosidens) to 17-21 (E. tutus), they occupy a space roughly com- parable to that of the plateaued ninth -eighteenth teeth of Tatisaurus. The teeth are grossly similar in outline, but they are elaborately fluted as in the ornithopods and bear more or less well-developed median ridges; those of P. grangeri alone lack the features of foldings and definitive cingula which are conspicuous in its Asian relative P. ninghsiensis Young (1935) . ACANTHOPHOLIDAE Tatisaurus is favorably compared to this group which shows a sigmoid alveolar border; this feature, while not exclusively ankylo- saurian, occurs in the more massive stegosaurs and ceratopsians (Microceratops gobiensis Bohlin, 1953), but it delimits the present form from the groups already considered. Two European genera have been selected for comparison, Acanthopholis horridus Huxley (1867), Struthiosaurus transylvanicus Seeley (1881) and S. austria- cus Nopsca (1929). The nature of the teeth and jaw fragments of Hylaeosaurus Owen (1872) preclude any affinity to Tatisaurus. Acanthopholis horridus shows a sigmoid alveolar border and the absence of an outer ridge, i.e., no coronoid process, irrespective of SIMMONS: NON-THERAPSID REPTILES 71 the folded enamel surfaces of its teeth (Nopsca, 1923) . To my knowl- edge, the mandible has never been figured and the original description was largely based on post-cranial materials. Drawings of the fragments of the mandibular rami and teeth of Struthiosaurus sp. show that the alveolar border is sigmoid but that it approaches the extreme mediad displacement observed for the nodosaur Edmontonia. Accordingly, the internal wall of the man- dible is concave in height while that of Tatisaurus is moderately con- vex, but in length, the convexity of the jaws is quite comparable. The meckelian groove is considerably larger in Struthiosaurus, but it terminates anteriorly at approximately the same level. S. austri- acus apparently has a weak predentary element forming a yoke be- tween the rami, but the preservation of Tatisaurus does not permit comparison on this point. While Struthiosaurus lacks a coronoid ele- ment, the lateral side of the jaw is convex and sharply ridged. The morphology of Tatisaurus is, then, somewhat intermediate between it and Acanthopholis— the ridge is low and restricted largely to the midlength of the ramus. A series of four large foramina occupy an essentially similar position on the outer wall of Struthiosaurus and might have resulted from the coalescence of several smaller perfora- tions. The jaw of S. trannsylvanicus is less tapered than Tatisaurus. Similarities in dental architecture, including patterns of wear (Nopsca, 1929, fig. 5) form the greatest point of comparison. The teeth of S. austriacus are graded from small to large anteroposteriorly. In early use, the functional teeth first lose their denticulations, and additional impairment produces a low median ridge and reduces the cingulum (S. trannsylvanicus) to a cinguloid swelling. The ultimate picture is a symmetrical, compressed crown with adenticulate trench- ant edges and a low median rib. It is not known if the teeth were oriented into an overlapping row. In S. trannsylvanicus, the dental series numbers at least 20, but only 13 occupy a length commen- surate with the total complement (18) of Tatisaurus. Summary.— The specimen is significant in that it represents the third find of ornithischia in the Triassic; Geranosaurus of the South African upper Triassic-lower Jurassic was the first recorded and placed into the primitive hypsilophodont family. In Tatisaurus, the precoronoid part of the dentary and the relationship of its elements is morphologically like the Jurassic-Cretaceous iguanodonts, partic- ularly Iguanodon foxii, rather than the most primitive ornithopods such as Hypsilophodon and Thescelosaurus where the splenial element masks the forward course of the meckelian groove. It is, moreover, 72 FIELDIANA: GEOLOGY, VOLUME 15 a good deal more slender than any of the cited forms and tapers more gradually toward the symphysis. The placement of the foramina, however, is entirely comparable to these early appearing ornithopods. The detailed structure of the jaw differs importantly from the forms cited above; the apparent reduction or complete lack of a coro- noid element and the distinctly sigmoid character of the dentigerous border and tooth row resembles most closely the Upper Cretaceous ankylosaur group, particularly the Acanthopholidae. The form of the teeth and the pattern of wear especially is remarkably similar to that of Struthiosaurus austriacus. Since, however, this particular character occurs only in this form, it is likely that we are dealing with a particularly conservative rather than a progressive feature and Struthiosaurus may therefore retain the traces of its heritage. Evolutionary Considerations of Tatisaurus By their expansive premaxillae and armor, it seems feasible to seek the ancestors of the Ornithischia among the Pseudosuchia (Romer, 1956). Yet, no pseudosuchian is known which expresses the orni- thischian tetraradiate pelvis, toothless premaxilla or specialized den- tition. Huene's suggestion (Young, 1951) that Platyognathus, with its polygonal teeth and anterior jaw expansion, might serve as the forerunner is vitiated by the knowledge of its crocodilian-like post- cranial skeleton. Nopsca (1923) notes that certain camptosaurs (ex. Rhabdodon) have teeth with tapered and elongated roots with feebly polygonal cross sections. This suggests that the iguanodonts, to which Tatisaurus appears most closely related on the basis of the precoronoid portion of the jaw, as well as Platyognathus, perhaps arose from a generalized sphenosuchid line which was sufficiently plastic to be the predecessor of other evolutionary lineages such as the crocodiles. When ornithischians first appear in the Upper Triassic, they seem already diversified. The jaw of the South African Geranosaurus from the Stormberg Series has a concave upper surface which indi- cates that the premaxilla was toothed. A toothed anterior jaw in Tatisaurus is probably obviated by the narrowness of the bone, but the dentition was specialized to conform to one of the more advanced ornithischian families. The Triassic materials are heterogeneous, and incorporated both primitive and advanced characters. It is considered that the osteo- logic affinities of Tatisaurus are to these primitive hypsilophodonts, although the dentition is particularized as ankylosaurid, and that it SIMMONS: NON-THERAPSID REPTILES 73 represents an early offshoot from the line of which Scelidosaurus Owen (1861) is the base. So little is known about the pre-Cretaceous evolution of the group that the suite of characters which seem to ally the jaw with the advanced ankylosaurs may really be primitive rather than diagnostic (Romer, 1959). STRATIGRAPHIC AND PALEOECOLOGIC CONSIDERATIONS Continental Triassic beds are widespread in northern China, but occur less well developed in the south where the facies are chiefly marine. Depositional basins in southwest Sinkiang, southeast and southwest Shansi and in Yunnan — the Lufeng Basin — have yielded vertebrate fossils, but the temporal sequence of these Triassic sedi- ments is discontinuous. In order to determine the age of these beds more precisely, Young (1946) compared the vertebrate assemblages from these localities to other faunas in the world during the Triassic and noted their relative developments. Preliminary to a considera- tion of the fauna from Lufeng which is rather homogeneous in all the four fossiliferous horizons, it may be useful to visualize a chart com- paring the Triassic continental stratigraphic divisions pertinent to the following discussion based upon the work of Young, Huene (1940) and Reeside et al (1957). The physical character of the rocks is dis- cussed in this section, but the vertebrates have proven to be the more sensitive indices of age. Lower Triassic. — In Sinkiang (Chitai), Lower Triassic sediments are well developed by the Tunghunshan and Shoufangkou Series. The older has yielded three genera which appear in the South Afri- can Lower Triassic Lystrosaurus Zone. Although Lystrosaurus and Dicynodon are the principal anomodonts of the South African Kar- roo as well, the Chinese area is relatively impoverished in terms of diversity. No. of species Genera S. Africa Sinkiang Dicyndon Chasmatosaurus Lystrosaurus 3 1 19 1 (Young, 1939a) 1 (Young, 1936) 1 (Yuan and Young, 1934) At the type section in Sinkiang, taken at Fuyuan, the Triassic lies conformably on the Upper Permian; the Tunghunshan fauna in- cludes a fragmentary specimen of a probable eosuchian, Santaisaurus 74 FIELDIANA: GEOLOGY, VOLUME 15 yuani Koh (1940) which is apparently related (Romer, 1956) to the Paliguanidae from the Karroo, Cistacephalus Zone. An additional eosuchian fragment (Efremov, cited by Huene, 1940) has also been described from the Upper Permian Rhophalodon Horizon (Zone 1) of Russia. The anomodont character of the Chinese assemblage and the diagnostic genus Lystrosaurus are, nevertheless, strong arguments for a lower Triassic age. The genus is also recognized from the Trias- sic Panchet Beds of India and from the Capitosaurus Level (Zone VI) in Russia. Eosuchians are known as well from the Karroo Lystro- saurus Zone. Middle Triassic. — Three determinable genera have been recorded from the Shansi Beds: Chasmatosaurus and Lystrosaurus (Young, 1958) from the lower levels (Lutzeyu, Wuhsiang) and Sinokanne- meyeria pearsoni Young (1937) — osteologically very similar to Kan- nemeyeria spp. from the Karroo Upper Beaufort Cynognathus Zone — in the middle level. The Cynognathus Zone has variously been regarded as Lower Triassic (Watson, 1942), Middle Triassic (Huene, 1940), or Upper Triassic (Broom, 1932). In Shansi, as in Sinkiang, the Permian and Triassic sediments are entirely conformable and relatively unfossiliferous, but Young (1946) reasoned that owing to the high level at which Sinokannemeyeria was found and since the horizon is apparently overlain elsewhere by a coal-bearing series of Rhaetic age, the beds were deposited not earlier than the uppermost part of the Middle Triassic and not later than the lowermost part of the Upper Triassic. The finds of Chasmatosaurus and Lystrosaurus connect the Lower-Middle (Chitai) and Middle (Wuhsiang) Triassic faunas in China. Upper Triassic. — In Sinkiang, the Upper Triassic is questionably developed as the Kankou Series, and in Shansi the Sinokannemeyeria Beds have been divided between the upper part of the Middle Trias- sic and the lower part of the Upper Triassic. Upper Triassic -Lower Jurassic vertebrate-bearing beds are represented in a structural basin in Lufeng, Yunnan, where the age determinations were similarly based upon the character of the representative vertebrates. The sediments everywhere appear to have undergone typical red bed development: (1) alternating red sandstone and shales — Tung- hunshan Series, Sinkiang; (2) dark red shales, clays and sandstones for the Lower Triassic and green sandstones with red sandy clays for the Upper Triassic-Shansi. The character of the Lower Lufeng Series is entirely like the Upper Triassic beds in Shansi. While physical correspondence alone is inconclusive evidence of contemporaneity SIMMONS: NON-THERAPSID REPTILES 75 since the vertebrates are individually exclusive, it may nevertheless demonstrate somewhat equivalent conditions of deposition. The predominately silty deposits of the Upper Triassic Moenkopi and Chinle Formations of North America are fine textured like the fossiliferous clays, shales and sandstones of the Lower Lufeng Series; but otherwise, they feature lenticular and conglomeratic members, etc. — particularly evident in the Chinle, lower Red Member — and CHIN A WESTERNSOUTHERN EUROPE ASIA A F R 1 C A WESTERN NORTH SOUTH NORTH SCOTLAND AMERICA AMERICA RUSSIA EASTERN US SINKING SHANSI YUNNAN GERMANY INDIA S AFRICA RUHUHU CHIWETA ELGIN BRAZIL WYOMING PENNSYLVANIA LUFENG SERIES SAUHSCHLANS RHAETIC CAVE SANDSTONE RED REDS CAMANDAS DE SANTA MARIA g£ U r.EUPER CHINLE FMl OOCKUM REDS CHUGWATER FM KANKOU SERIES ■ ~ PL* I KEUPCR MALERI SCHICHTEN MANOA SCHICHTEN STACONOLEPi: SANDSTONE RIODORASTO REDS PHOENIIVIILE TUNNEL «y SINOKANNE - MEVEAIA UMUSCHCl ! ;VMOGNATHUi ZONE ? L. HUMMEL ■ KALK 1 shcwngkhi SEMES ZONE VI CAPITO- SAURUS ZONE LVSTROSAU RUS ZONE IUHTSAND STEIN PANCHET SCHICHTEN WGHUNSHW SEStlES ZECHSTEIN KUPFER SCMIETER CISTACEPH- ALUS ZONE ZONE IV PAREIO SAUR ZONE LOWER RONE RED UPPER ■ ONE RED ENOOTHIO DM ZONE 2 < AOTLIEGENDES OS H RHOPHA. LOOON a. Fig. 12. A chart showing the world distribution of vertebrate bearing horizons during the late Paleozoic and early Mesozoic comparable in time to the Chinese section. the gypsiferous and ripple marked, cross-bedded sands wanting in the Lower Lufeng Series (Akers et al, 1958). The Triassic-Jurassic Kayenta Formation, from which Lewis secured a Tritylodon-like rep- tile (Reeside et al, 1957), is typically silty, grading to sandstone to the south-east; the sediments from southwest Utah and that part of northern Arizona demonstrate a reddish siltstone and mudstones with ripple marks and mudstone pellets characteristic of shallow 76 FIELDIANA: GEOLOGY, VOLUME 15 water deposition which are somewhat grossly similar to the Upper Lufeng Series, but the beds at Lufeng become somewhat more cal- careous toward the top of the section. The German Upper Keuper, from which sauropods comparable to those at Lufeng have been found, has been characterized as a lagoonal, saliferous formation containing variegated marls, anhy- drates and plant bearing sandstones and porous dolomitic limestones (Gignoux, 1955). To the east in Upper Silesia and Poland, unioid- bearing fresh water limestones occur which perhaps are attributable to a facies change concomitant with isolation from the Alpine Sea. While marly layers are developed much more extensively elsewhere than in Lufeng during the Triassic, some poorly preserved plant re- mains have been found (Chang, 1947), but it is doubtful that these are Equisitales like those from the German Keuper (Gignoux, 1955). Although the sediments of the Lufeng Series were deposited under terrestrial conditions in an area isolated from the Triassic Sea enter- ing Yunnan from the south, unioid-bearing marl-like rocks do occur in the Upper Series. In Tanganyika (German E. Africa), the Manda-Schichten upper bone beds, K8 (Krenkel, 1939), consist of rose feldspathic sandstone and marl, and iron containing calcareous concretions. A prosauro- pod, Thecodontosaurus, comparable with the Lufeng Yunnanosaurus is derived from these strata which are, as in Lufeng, disconformably separated from the lower bone beds (K6) by a sandstone. The sec- tions are, however, not entirely comparable in detail. The mineral- ogy of the Lufeng sediments is unknown, but iron-bearing rocks with calcareous concretions are common to both. The lower bone beds contain Dicynodon species like the lower Triassic Sinkiang sediments. Perhaps the depositional history of the upper Triassic Red Beds of the South African Stormberg Series parallels that of the Lufeng section most closely. These consist of well-layered friable, red violet clay marls penetrated by carbonate and red and green sandstones. In many places, silicified woody stems of over yH a < Ph o o w o o o N 8 8 2 e §1 2 s i c CO Oh I it E-. &3 s 3 e e ec CO CO O o e 5 •g c ^ El Oh So© ^ ^£ 0 s PQ < Eh a. a c a> "3 c» fa 2 c O w^ O 00 s o 2> Ph I Oh Q V3 O Ph CO g 'S » 2 9 CO 03 5? S o M ai iH oo as o o (M (M Ph Ph •G £ o .. a © § * 2 2 is O C £ S i f s* Ok) kj e e S .2 s S g s a -5 . c 'C a; eg a 3 £ _> o '5 u cr a> -•-> o a> cS h c C 3 Ed En >h >-, C5 c E« 80 SIMMONS: NON-THERAPSID REPTLIES 81 doubt, pseudosuchia are limited to Upper Triassic sediments. The Lufeng fauna as described by Young included one barely determin- able animal, Platyognathus hsui from Huangchiatien (second fossil- iferous horizon). The present collection has yielded a diverse array of forms derived essentially from the same horizon (Ta Ti and Hei Koa Peng) which have been compared to South African and North American types; none has yet been found above or below the sedi- ments just bordering the Upper Series. The levels at which related thecodonts are found have been tabulated in Table 20. All but Dibothrosuchus and CUP 2084 resemble elements in the Stormberg Series. Dibothrosuchus is similar to Ornithosuchus in the shape of its skull, but resembles Hesperosuchus and Parringtonia in most other respects. The jaw of Strigosuchus is hesperosuchid-like. The forms, then, can- not be older than the lowest part of the Upper Triassic, but their upper limit in time is ill-defined as Upper Triassic-Lower Jurassic. The sphenosuchid-like appearance of the skull of Platyognathus indicates an age not later than the Upper Triassic, but the complete investiture of the body by scutes in primitive crocodilian fashion is like the North American Protosuchus (Orange Red Sand, Moenhave Formation, Dinosaur Canyon sandstone member) which, like Hes- perosuchus, enjoys a questionable stratigraphic position. Reeside et al (1957) considers the Moenhave to be Upper Triassic; Harshbarger et al (1958) considers the orange red sand the lower Member of the Moenhave (the basal formation of the Glen Canyon Group) not younger than the Middle Jurassic (Carmel Formation), and inter- prets it like Colbert and Mook (1951) as both Triassic and Lower Jurassic. The decision for Platyognathus depends on whether em- phasis is placed upon the more conservative general character of the skull or upon the suite of apparently independently evolved, highly adaptive characters of the postcranial skeleton. Pelvic girdle and femoral fragments of the indeterminate pseudo- suchian CUP 2984 bear some resemblance to Typothorax from the Texas Dockum Group and the New Mexican Chinle Formation. The Stagonolepidae as a whole are of Middle Triassic age in Europe and of Upper Triassic age in North America. Parasuchia (Phytosaurs) .• — Incomplete skull and jaw fragments of Pachysuchus imperfecta Young (1951) derive from the upper fos- siliferous horizons of the Lower Lufeng Series. Phytosaurs are major components of European and North American assemblages and ap- pear to be an exclusive development of the Triassic. In America, 9 d 3 o JS 3 os « -e s 3 «j ^; S? 3 w 05 O 3 c^ a, a* ft* Bq -ess s 5 5 5 o> 6 £ <4> © g 3 d S 4) k ^5 © *> (5 &2 tt: s s -=: •e v> ■»* 3 oo 00 o o « s> PM © P CO H^ o in H A O Q O o P O w p w p Eh o CO 525 O i— i M o o m m < w Eb'g ojPh o tf « a a P S CO .2 cx 'G a H P a> o •3 +j 73 h 03 § a 1 +-> u CO CD 9 r* & o o HH J £ 2 fl 5 PQ ^ .a o u _G O IS u PQ o oo H CO 03 3 .1 a .2 3 03 § s rt ^ cS "Si ^o c 03 bo 2 B s 03 Eh -3 c l-H < a a, P a CO OJ CO bo CD o o C 03 CO bo S CO be to G 0) 5h M "3 ►J 0> 3J pq "3 ►J t-. cd Qj ss £ •sS T3 & * 03 o Jm^"1 0) o C o hH o P* HH J .2 CD s .° .2 .2 °w 'C 'co "oq < CJ a 3 a CO c c a) O >. a co s 82 SIMMONS: NON-THERAPSID REPTILES 83 they are known from the Dockum, Chugwater and Cumnock For- mations. The Indian genus Brachysuchus is reported from the Ma- leri Schichten with two other poorly known forms, and perhaps from North America as well; it emphasizes the paucity of like forms from Asia. That the South African group is devoid of phytosaurs is of particular interest to this discussion. Pachysuchus has been compared to Mystriosuchus planirostris (Germany) and to Rutiodon carolinensis (North America). SAURISCHIANS The saurischian remains are of limited time diagnostic value. Prosauropods, known mostly from isolated fragments, are the dom- inant forms and appear to be typically Upper Triassic forms; Lufen- saurus hueni is the most complete specimen. "Structurally, the posi- tion of the skull to the vertebral column, the primitive features of the teeth, presence of abdominal ribs, the absence of the foot-like expansion of the pubis and the relative length of the femur, exceed- ing that of the tibia and fibula and five digits of the hand and foot are all primitive characters which are rarely or never found in the Saurischia of later age (Young, 1939a)." Coelurosaurs. — These are known from the second fossiliferous ho- rizon of the Lower Lufeng Series. They are so imperfectly realized that it was considered expedient to refer them (Lukousaurus, CUP 2089, 2090, 2091) to the general Upper Triassic family, the Podoke- sauridae. Only Lukousaurus is known by cranial materials. Lukousaurus and other Podokesauridae occur in Europe and North America in the Upper Triassic sediments. CUP 2089 has been related to the North American Coelophysis. The only compari- son afforded the femur CUP 2091 was to an Australian Upper Cre- taceous form. No coelurosaurs are recognized from the African assemblage, but Young listed Gryponyx as the faunal equivalent of Lukousaurus. Carnosaurs.- — Remains of the sole Lufeng carnosaur, Sinosaurus triassicus are ubiquitous in the saurischian facies, and Young (1946) compared it to the German Teratosaurus suevicus. It has also been cited as the faunal counterpart of the South African Upper Triassic Orosaurus; while these are not closely compared, it is striking that both assemblages are characterized by individual specimens of the Teratosauridae. Elsewhere, the family ranges from the Middle- Upper Triassic of Europe (3 genera) and North America (1 genus). 84 FIELDIANA: GEOLOGY, VOLUME 15 The jaw (CUP 2097) tentatively identified as that of S. triassicus may actually represent a second carnosaur. Among the carnosaurs, the range of generically identical forms is very broad. An advanced megalosaurid, Antrodemus, has been ten- tatively identified from east Asia, North America and east Africa. Similarly, the North American tyrannosaurid, Aublysodon, is reported from east Asia. It may be considered that carnosaurs, therefore, as indicators of faunal relationships, need relate only to familial asso- ciations. The contribution of Sinosaurus to the faunal complex is in this way directly comparable to that of the South African genus. Prosauropods. — Saurischia comparable to the Lufeng prosauro- pods occur in Triassic beds world-wide in distribution. Age Geographic position Site Lower-Upper Triassic North America Phoenixville Tunnel, Pa. Upper Triassic Tanganyika Manda-Schichten, upper bone bed. India Maleri-Schichten. Germany Keuper. South Africa Red Beds. "Rhaetic (?)" South Africa Cave Sandstone. The massive Lufengosaurus resembles the Knollenmergel Plateo- saurus fraasianus of Germany more closely than the correspondent South African species; Yunnanosaurus spp. however have been com- pared to Massospondylus of the Red Beds and Cave Sandstones, although the genus is also known from India and east Africa. The referral of the most common and rather completely known animal, Gypsosaurus sinensis, to the South African genus from the Cave Sandstone (G. capensis) is particularly significant. ORNITHISCHIA The new find of Tatisaurus oehleri in the second fossiliferous hori- zon evidencing both primitive and relatively advanced characters is particularly pertinent both to the considerations of the time element, and to the faunal relationships. The geographic distribution of the first recorded ornithischians is scattered discontinuously — in South Africa and North America. Although incomplete, Tatisaurus is more favorably compared to younger European forms which are not older than Upper Jurassic rather than to Geranosaurus but it never- theless contributes to the Lufeng assemblage in like proportions. SIMMONS: NON-THERAPSID REPTILES 85 Ornithischians are rare in Upper Triassic sediments. Gerano- saurus is derived from the Cave Sandstone, South Africa and is Rhaetic in age. The progressive ornithischian characters of Tatisau- rus which resemble the Upper Cretaceous ankylosaur Struthiosaurus have been considered conservative features which are not indicative of a direct lineage. It is interesting that the earliest ornithischians are a heterogeneous group, but are mutually associated with forms of a more or less late Triassic age. THERIODONTS Although outside the scope of this paper, it should be mentioned that the Lufeng sediments have yielded mammal-like reptiles, the cynodont Kunminia minina Young (1947a) and the ictiodosaur Bi- enotherium Young (1940a). Cynodonts appear generally throughout the Upper Permian to the Upper Middle-Lower Upper Triassic Kar- roo sediments and their range elsewhere in the world is essentially the same. Time Permian Upper-Middle Triassic Lower-Upper Triassic South Africa Time equivalent Geographic (zone) strata location Endothiodon- Ruhuhu-Gebiet: Tanganyika Cistacephalus lower Bone Bed. Zone Cistacephalus Chiweta: Nyassaland Zone upper Bone Bed. Pareiosaur Zone, Russia Zone IV Upper Rio do Rasto South America Cynognathus Manda-Schichten : Tanganyika Zone upper Bone Bed. Kunminia is too poorly known to afford direct comparison with any of the forms representative of the localities cited above; it oc- curs at Huangchiatien just below the purported Triassic-Jurassic boundary. Ictidosaurs range into the red beds and Cave Sandstones of the Karroo Upper Triassic to the Jurassic boundary and are known (Archaeodon) from the Stormberg Series in southwest Africa and from the Kayenta Formation of Arizona (Averett er al, 1955) . The group is evidenced in Jurassic sediments, Trityodon being closely compared to Bienotherium. 86 FIELDIANA: GEOLOGY, VOLUME 15 Oehler's collection of therapsids, now under study by Drs. E. C. Olson and J. Hopson from the University of Chicago, have yielded a wealth of new Bienotherium materials which demonstrate a more complex structure than hitherto recognized for the genus. The form ranges throughout the Lower Lufeng Series.1 SUMMARY The fauna of the Lower Lufeng Series is best described as cosmo- politan. Relative to the standard section of the Triassic (Germany), the sediments should probably be classified Upper Triassic, equiva- lent in time to the Upper Keuper, Knollenmergel facies. Support for this correlation is based upon the generalized structure of the saurischian fauna, but there are many instances of forms morpho- logically related to animals derived from Rhaetic, Jurassic and even younger beds. For example, Platyognathus is similar to Protosuchus (Upper Triassic-Lower Jurassic) and to Sphenosuchus (Upper Tri- assic Red Beds, Stormberg Series) . Prosauropods are typically Up- per Triassic forms, yet it is possible to infer a somewhat younger age from the numerous examples of Gyposaurus (and the theriodonts) . In reference to the North American section, a tritylodont has been found in the uppermost level of the Kayenta Formation bearing likeness to Bienotherium and to Tritylodon from the Red Beds and Cave Sandstones (?Rhaetic) of South Africa. The opinion has been expressed (Reeside, et al, 1957) that the Kayenta may be Upper Tri- assic in time, but the dinosaur Megalosaurus weatherilli, more com- parable to the Middle-Late Jurassic megalosaurids of Europe than to the Late Triassic forms in its proportions, derives from a lower level than Tritylodon. The theriodont evidence speaks for a Rhaetic or Lower Jurassic age at the earliest. The ictidosaurs, while a highly advanced spe- cialized group may also be a relatively small and possibly conserva- tive aggregation among themselves and one might not expect them to be a particularly definitive index of stratigraphic contemporaneity. Bienotherium has been compared (Watson, 1942) to the Upper- Middle Triassic cynodont Diademodon (Compsognathus Zone). For the present, it is probably safer to place the greatest reliance upon the more profuse suarian elements of the Lufeng fauna and to de- 1 Patterson and Olson (1961, Internat. Evol. Mammals, Kon. Vlaamse Acad. Wetensch. Lett. Sch. Kunsten Belgie, Brussels, pp. 129-191) have recently pub- lished a description of a triconodontid mammal from Hei Koa Peng of Upper Triassic age. SIMMONS: NON-THERAPSID REPTILES 87 emphasize the importance of the mammal-like reptiles where strati- graphic problems are the principal concern. The relationships of the Lufeng non-therapsids is predominately to European, South African and North American forms. The Pseu- dosuchia relate directly to North American and South African ele- ments. Coelurosaurs, with which the Lufeng fauna now appears to be richly endowed, have no South African components. There is, however, an essential correspondence in the contribution of carno- saurs, prosauropods, and ornithischians to the faunas of Lufeng and South Africa during this same approximate geologic time, notwith- standing the more specific relations of some of the prosauropods to European forms. The non-therapsids have an old world flavor despite the apparent overlap in some groups to the North American elements. Impor- tantly, the prosauropod contribution is large and for the most part lacking in Triassic new world assemblages (other than Paleosaurus) . Instances in which the vertebrates as a group or as individuals fail to correspond to South African forms are rather exceptional. Young's interpretation of the cosmopolitanism of the fauna and their essentially African flavor is supported by the therapsid evidence and by the important discovery of the early ornithischian. Equal emphasis has not been accorded to all parts of the Lufeng Basin where the fauna is known almost entirely from the Lower Series. However, even here, the general mode of collecting, i.e., from surface nodules, by two individuals with different habits has yielded different results. The present collection has brought to light several new forms and it is most probable that the diversity of life of the Lufeng Basin is greater than now known. Re-exploration of the beds is decidedly warranted and additional fossil evidence will clarify the existing uncertainties as to the time factor. REFERENCES Akers, J. P., M. E. Cooley and C. A. Repenning 1958. Moenkopi and Chinle formations of Black Mesa and adjacent areas. In New Mex. Geol. Soc. Guidebook for 9th Field Conf., edited by R. Y. Anderson and J. W. Harshberger. Averitt, P. J., J. S. Detterman, J. W. Harshberger, C. A. Repenning and R. F. Wilson 1955. Revisions in Corrleation and Nomenclature of Triassic and Jurassic Formations in Southwestern Utah and Northern Arizona. Bull. Amer. Ass. Petrol. Geol., 39, pp. 2515-2524. Bien, M. N. 1941. "Red Beds" of Yunnan. Geol Soc. China 21, pp. 159-198. Bohlin, B. 1953. Fossil Reptiles from Mongolia and Kansu. Sino-Swedish Expedition Publication 37, 6, pp. 1-113. Broom, R. 1904. On a new Crocodilian Genus (Notochampsa) from the upper Stormberg beds of South Africa. Geol. Mag., 1, pp. 582-584. 1908. On the Dinosaurs of the Stormberg, South Africa. Ann. S. Afr. Mus., 12, pp. 291-308. 1927. On Sphenosuches, and the Origin of the Crocodiles. Proc Zool. Soc. London, 1927, pp. 395-370. 1932. On some South African Pseudosuchians. Ann. Natal. Mus., 7,, pp. 55-59. Brown, B. 1908. The Ankylosauridae, the new family of Armored Dinosaurs from the upper Cretaceous. Bull. Amer. Mus. nat .Hist., 24, pp. 187-201. Bystrow, A. P. 1947. Hydrophilous and Xerophilous Labyrinthodonts. Acta Zool., 28, pp. 137-164. Chang, C. H. 1947. The Triassic Paleogeography and Stratigraphy of Yunnan. Bull. Geol. Soc. China, 27, pp. 193-204. Colbert, E. H. 1952. A pseudosuchian reptile from Arizona. Bull. Amer. Mus. Nat. Hist., 99, pp. 565-592. Colbert, E. H. and C. C. Mook 1951. The Ancestral Crocodilian Protosuchus. Bull. Amer. Mus. Nat. Hist., 97, pp. 147-182. Cope, E. D. 1887. A contribution to the History of the Vertebrata of the Trias of North America. Proc. Amer. Phil. Soc. 24, pp. 209-228. 89 90 FIELDIANA: GEOLOGY, VOLUME 15 GlGNOUX, M. 1955. Stratigraphic Geology. (English translation from 4th ed., 1950, by G. G. Woodford) San Francisco. GlLMORE, C. W. 1923. A new species of Corythosaurus with notes on other Belly River Dino- sauria. Canad. Field Nat.,37, pp. 47-52. 1933. Two new Dinosaurian reptiles from Mongolia with notes on some frag- mentary specimens. Amer. Mus. Nov. no. 679. Harshberger, J. W., C. A. Repenning and J. H. Irwin. 1958. Stratigraphy of the uppermost Triassic and the Jurassic rocks of the Navajo country. In New Mex. Geol. Soc Guidebook for 9th Field Conf., edited by R. Y. Anderson and J. W. Harshberger., pp. 98-114. Haughton, S. H. 1915. Investigations in South African Fossil Reptiles and Amphibians. Ann. S. Afr. Mus., 12, part 9, pp. 98-105. 1924. Fauna and Stratigraphy of the Stormberg Series. Ann. S. Afr. Mus., 12, pp. 323-437. Hulke, J. W. 1882. An attempt at a complete osteology of Hypsilophodon foxii: a British Wealdon Dinosaur. Trans. Roy. Soc. London, 173, pp. 1035-1062. Huxley, T. H. 1867. On Acanthopholis horridus, a new reptile from the chalk-marl. Geol. Mag. 4, pp. 65-67. 1870. On Hypsilophodon foxii; a new dinosaurian from the Wealdon of the Isle of Wight. Quart. Journ. Geol. Soc, 26, pp. 3-12. Koh, T. 1940. Santaisaurus yuani gen. et sp. nov., ein neues Reptil aus den unteren Trias von China. Bull. Geol. Soc. China, 20, pp. 73-92. Krenkel, E. 1939. Geologie der Deutschen Kolonien in Afrika. Berlin. Lambe, L. M. 1919. Description of a new genus and species (Panoplosaurus mirus) of an armored dinosaur from the Belly River Beds of Alberta. Trans. Roy. Soc. Canada, sect. IV, ser. Ill, pp. 39-50. Leidy, J. 1960. Extinct vertebrata from the Judith River and great lignite formations of Nebraska. Trans. Amer. Phil. Soc, 11, pp. 139-154. Misch, P. 1945. Remarks on the Tectonic History of Yunnan, with special reference to its relations to the Type of the young orogenic deformation. Bull. Geol. Soc. China., 25, pp. 47-153. Mook, C. C. 1933. A skull of Crocodilus clavis Cope, in the United States National Museum. Amer. Mus. Nov., no. 678, pp. 1-7. Newton, E. T. 1894. Reptiles from the Elgin Sandstone. Description of two new genera. Phil. Trans. Roy. Soc. London, ser. B, 185, pp. 573-607. SIMMONS: NON-THERAPSID REPTILES 91 Nopsca, F. Baron 1905. Notes on British Dinosaurs. Part I. Hypsilophodon. Geol. Mag. 2, pp. 203-208. 1923. Notes on British Dinosaurs. Part VI. Acanthopholis. Geol. Mag., 60, pp. 193-199. 1929. Dinosaurerreste aus Siebenburgen. V. Geol. Hung. Tomus I, facs. 4, pp. 1-76. Owen, R. 1861. A monograph of the Fossil Reptilia of the Liassic Formations. Part I. Scelidosaurus harrisonii. Palaeontogr. Soc. (Monog.), le, pp. 1-13. 1872. Suppl. no. IV. to the Monograph on the Fossil Reptilia of the Wealdon Formation. Palaeontogr. Soc. (Monog. ),pp. 1-15. 1874. Suppl. no. V. to the Monograph on the Fossil Reptilia of the Wealdon and Purbeck Formations (Iguanodon). Palaeontogr. Soc. (Monog.), pp. 3-16. Parks, W. A. 1924. Dyoplosaurus acutosquameus, a new genus and species of Armored dino- saur. Notes on a skeleton of Prosaurolophus maximus. Uni. of Toronto Studies, Geol. Ser., Ill, 18, pp. 1-35. PlVATEAU, J. 1956. Traite de Paleontologie, 5, Paris. Price, L. I. 1946. Sobre Um Novo Pseudosuquio Do Triassico Superior Do Rio Grande Do Sul. Departmento Nacional de Producao Mineral, Divisao Di Geologie Mineralogie. pp. 7-39. Reeside, J. B., chairman., et al 1957. Correlation of the Triassic Formations of North America exclusive of Canada. Bull. Geol. Soc. Amer., 68, pp. 1451-1541. Romer, A. S. 1955. Vertebrate Paleontology. Chicago 1956. The Osteology of the Reptiles. Chicago. Russell, L. S. 1940. Edmontonia rugosidens (Gilmore) an armored dinosaur from the Belly River Series of Alberta. Univ. of Toronto Studies, Geol. Ser., no. 43, pp. 3-28. Sawin, H. J. 1947. The pseudosuchian reptile Typothorax meadei. Journ. Paleontol. 21, pp. 201-238. Seeley, H. G. 1881. The reptile fauna of the Gasau formation preserved in the geological museum of the University of Vienna. Quart. Journ. Geol. Soc, 37, pp. 620-707. Sternberg, CM. 1928a. A new armored dinosaur from the Edmonton Formation of Alberta. Trans. Roy. Soc. Canad. (ser. 3), 22, sec, pp. 93-106. 1928b. A toothless armored dinosaur from the upper Cretaceous of Alberta Nat. Mus. Canada, Bull. 54, pp. 28-33. 1940. Thescelosaurus edmontonensis, n. sp., and classification of the Hypsilo- phodonotidae. Journ. Paleontol., 14, pp. 481-494. Talbot, M. 1911. Podokesaurus holyokensis a new dinosaur from the Triassic of the Con- necticut Valley. Amer. Journ. Sci., 31, pp. 469-479. 92 FIELDIANA: GEOLOGY, VOLUME 15 VAN HOEPEN, E. C. N. 1915. Contributions to the knowledge of the reptiles of the Karroo Formation 4, A new pseudosuchian from the Orange Free State. Ann. Transvaal Mus., 5, pp. 83-87. van Houten, F. B. 1948. Origin of Red-Banded Early Cenozoic Deposits in Rocky Mountain Region. Bull. Amer. Ass. Petrol. Geol., 32, pp. 2083-2126. von Huene, F. 1906. Uber die Dinosaurier der Aussereuropaischen Trias. Geol. palaont. Abh., 8, pp. 19-22. 1910. Ein primitiver Dinosaurier aus der mittleren Trias von Elgin. Geol. palaont. Abh., 8, pp. 317-322. 1914-1915. Beitrage zur Kenntnis und Verteilung der Pseudosuchier. Geol. palaont. Abh., 13, pp. 1-82. 1915. On Reptiles of the New Mexican Triassic in the Cope Collection. Bull. Amer. Mus. Nat. Hist., 34, pp. 485-507. 1921. Neue Pseudosuchier und Coelurosaurier aus dem Wurttembergischen Keuper. Acta. Zool., 2, pp. 329-403. 1925a. Die Bedeutung der Sphenosuchus-Gruppen fur den Ursprung der Krok- odile. Z. indukt. Abstamm.-u. VererbLehre, 38, pp. 307-320 1925b. Die sudafrikanische Karroo-Formation als geologisches und faunist- isches Lebensbild. Berlin. 1932. Die Fossils Reptil-Ordnung Saurischia, ihre Entwicklung und Ge- schlichte. Leipzig. 1936. Die Altersbeziehungen der Sudamerikanischen Gondwana-Fauna. Physis (Rev. Soc. Argentina Cienc. Nat.), 14, pp. 289-497. 1939a. Ein kleiner Pseudosuchier und ein Saurischier aus den ostafrikanischen Mandashichten. Neus. Jb. Miner., Mh., Abt. B., pp. 61-69. 1939b. Ein grosser Stagonolepide aus der jungeren Trias Ostafrikas. Neues Jb. Miner., Mh., Beilage-Band Abt. B. 80, pp. 264-278. 1940. Die Saurier de Karroo-. Gondwan-, und verwandten Ablagerungen in faunistischer, biologischer, und phylogenetischer Hinsicht. Neues Jb. Miner., Mh., Beilage-Band Abt. B, 83, pp. 246-437. 1948. Short Review of the Lower Tetrapods. Roy. Soc. S. Afr. Spec. Publ., Robert Broom Comm. Vol., pp. 65-106. 1950 Die Entstehung der Ornithischia schon friih in der Trias. Neus. Jb. Geol., 0, pp. 53-58. Watson, D. M. S. 1942. On Permian and Triassic Tetrapods. Geol. Mag., 79, pp. 81-116. Welles, S. P. 1954. New Jurassic Dinosaur from the Kayenta Formation of Arizona. Bull. Geol. Soc. Amer., 65, pp. 591-598. Young, C. C. 1935. On a new Nodosaurid from Ninghsia. Palaeont. sinica, (ser. C), fasc. 1, 11, pp. 1-34. 1936. On a new Chasmatosaurus from Sinkiang. Bull. Geol. Soc. China, 15, pp. 291-320. 1937. On the Triassic Dicynodonts from Shansi. Bull. Geol. Soc. China, 17, pp. 393-412. 1939a. Preliminary notes on the Lugeng saurischian remains. 40th Anniversary Paper. Nat. Univ. Peking, pp. 111-114. 1939b. Additional Dicynodontia remains from Sinkiang. Bull. Geol. Soc. China, 19, pp. 279-315. SIMMONS: NON-THERAPSID REPTILES 93 1940a. Preliminary note on the Mesozoic mammals of Lufeng, Yunnan, China, Bull. Geol. Soc. China, 20, pp. 93-111. 1940b. Preliminary note on the Lufeng vertebrate fossils. Bull. Geol. Soc. China, 20, pp. 235-239. 1941a. Gyposaurus sinensis Young (gen. et sp. nov.), a new Prosauropoda from the Red Beds at Lufeng, Yunnan. Bull. Geol. China, 21, pp. 205-252. 1941b. A complete osteology of Lufengosaurns hueni Young (gen. et sp. nov.), from Lufeng, Yunnan, China. Palaeont. sinica, (ser. C), 7, pp. 1-53. 1942. Yunnanosaurus huangi Young (gen. et st. nov.), a new Prosauropoda from the Red Beds at Lufeng, Yunnan. Bull. Geol. Soc. China, 22, pp. 63-104. 1944. On a supposed new pseudosuchian from Upper Triassic Saurischian- bearing beds of Lufeng, Yunnan, China. Amer. Mus. Nov., no. 1264, pp. 1-4. 1946. The Triassic vertebrate remains of China. Amer. Mus. Nov., no. 1234, pp. 1-14. 1947a. Mammal-like reptiles from Lufeng, Yunnan, China. Proc. Zool. Soc. London, 117, pp. 537-597. 1947b. On Lufengosaurus magnus Young (sp. nov.) and additional finds of Lufengosaurus hueni Young. Palaeont. sinica, 12, pp. 1-53. 1948a. Further notes on Gyposaurus sinensis Young. Bull. Geol. Soc. China, 28, pp. 91-103. 1948b. On two new Saurischians from Lufeng, Yunnan. Bull. Geol. Soc. China, 29, pp. 75-90. 1951. The Lufeng saurischian fauna in China. Palaeont. sinica, (ser. C), 13, pp. 19-96. 1958 On the occurrance of Chasmatosaurus from Wuhsiang, Shansi. Verte- brata Palasiatica, 2, pp. 251-262. Yuan, P. L. and C. C. Young 1934. On the occurrance of Lystrosaurus in Sinkiang. Bull. Geol. Soc. China, 13, no. 4, pp. 575-580. Publication 990