LIBRARY OF THE UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN % 550 GEOLOGY The person charging this material is re- sponsible for its return to the library from which it was withdrawn on or before the Latest Date stamped below. Theft, muHlafion, and underlining of books ore reasons for disciplinary action and may result In dismissal from the University. To renew call Telephone Center, 333-8400 UNIVERSITY OF IHINOIS LIBRARY AT URBANA-CHAMPAIGN FE| 2 8 1938 t vjAji^'^TggB L161 — O-1096 >^^ ^J>^.<^^ "n^ /, .^? THE VERTEBRATE FAUNA OF THE SELMA FORMATION OF ALABAMA PART VII THE MOSASAURS DALE A. RUSSELL PART VIII THE FISHES SHELTON P. APPLEGATE FIELDIANA: GEOLOGY MEMOIRS VOLUME 3, NUMBERS 7 AND 8 Published by FIELD MUSEUM OF NATURAL HISTORY FEBRUARY 12, 1970 fieoi '^OGi ^/6, Mfix THE VERTEBRATE FAUNA OF THE SELMA FORMATION OF ALABAMA PART VII. THE MOSASAURS THE VERTEBRATE FAUNA OF THE SELMA FORMATION OF ALABAMA PART VII THE MOSASAURS DALE A. RUSSELL Curator of Fossil Vertebrates National Museum of Canada FIELDIANA: GEOLOGY MEMOIRS VOLUME 3, NUMBER 7 Published by FIELD MUSEUM OF NATURAL HISTORY FEBRUARY 12, 1970 Library of Congress Catalog Card Number: 53-2305 PRINTED IN THE UNITED STATES OF AMERICA BV FIELD MUSEUM PRESS CONTENTS PAGE List of Illustrations 367 Introduction 369 Previous Work 369 Systematics 369 Subfamily Mosasaurinae 369 Halisaiirus sternbergi 369 Clidastes propython 371 Globidens alabamaensis 373 Subfamily Plioplatecarpinae 373 Platecarpiis sp 373 Prognathodon sp 373 Subfamily Tylosaurinae 374 Tylosaurus zangerli new species 374 Tylosaiirus sp 375 Subfamily Plioplatecarpinae or Tylosaurinae 376 Faunal Comparisons 377 Summary 379 References 380 365 LIST OF ILLUSTRATIONS PAGE 164. Halisauriis sternbergi, anterior cervical and posterior cervical or anterior thoracic vertebra 370 165. Halisaurus sternbergi, frontal 371 166. Clidastes, hind limb of C. liodontus and C. propython compared 372 167. Prognathodon sp., dentary 373 168. Prognathodon sp., atlas arch and jugal 374 169. Tylosaurus zangerli, humerus 375 170. Tylosaurus zangerli, femur 375 171. Tylosaurus sp., splenio-angular articulation 375 172. Mosasaur crania from the Mooreville Member of the Selma Chalk 376 173. Mosasaur crania from the Niobrara Chalk of Kansas 377 367 THE MOSASAURS INTRODUCTION In the course of preparing a revision of the North American mosasaurs I visited the Field Museum of Natural History briefly during the spring of 1963. Dr. Rainer Zangerl very generously presented me with the opportunity of describing the Selma mosasaurs, and sent the collection to me at Yale the summer of the following year. Unfortunately, there was insufficient time for me to study the collection at Yale and Dr. John Ostrom, of the Yale Peabody Museum, undertook to transport it to Ottawa before my arrival at the National Museum of Canada. I am very grateful to both of these gentlemen for their much appreciated assistance and courteous forbearance. The present paper was written after completion of the mosasaur monograph (Russell, 1967) and the reader is referred to this work for more complete generic and specific diagnoses of various American forms. The ma- terial under discussion was collected in 1945 and 1946 from the Mooreville Member of the Selma Chalk in central Alabama (see Zangerl 1948, pi. 3; 1953a, fig. 19 for locality maps) . Much of it seems to have been found as surface float. Numbers preceded by P or PR refer to specimens in the Field Museum of Natural History, those preceded by USNM and YPM refer to specimens in the U. S. National Museum and Yale Peabody Mu- seum respectively. PREVIOUS WORK The first mosasaur material known from the "Rotten Limestone of Alabama," and in all probability the Mooreville Member of the Selma Chalk, consisted of two badly weathered, otherwise unidentifiable vertebrae, three vertebrae of a plioplatecarpine or small tylosaur- ine, and a single tooth of a plioplatecarpine. These specimens were described and figured by Robert W. Gibbes (1850, 1851) as the name-bearers of three inde- terminate taxa of mosasaurs, respectively Amphorosteus brumbyi, Mosasaurus minor, and Holcodus acutidens. Cope (1869) firmly established the species Clidastes pro- python on an unusually complete skull and skeleton, and later (1869-1870) based "Liodon congrops" on a cervical vertebra probably belonging to this species. His (Ibid.) Tylosaurus"! perlatiis is known only from a very poorly preserved vertebra. Two inadequately known but highly interesting spe- cies of mosasaurs with inflated marginal dentitions have been discovered in the Mooreville Member. The iden- tity of the first of these to be described, "Platecarpus" intermedius (Leidy, 1870), is based largely on two den- tary fragments with closely spaced, bulbous teeth (see Leidy, 1873, pi. 34, figs. 1-5). The dentary terminates immediately in front of the first tooth as in Platecarpus, but new material will probably show the form belongs to a distinct genus. Globidens alabaniaensis with its characteristic hemispherical teeth was described by Gil- more in 1912. The few skeletal fragments in the type indicate the animal was more closely related to Clidastes than any other mosasaur (Russell, 1967). It is unfortu- nate that subsequent collecting has not increased our knowledge of the morphology of these two unusual Selma mosasaurs. More recently Renger (1935) has published a pop- ular description of a skull and skeleton of Tylosaurus, and Dowling (1941) noted the occurrence of a small but rather completely preserved skeleton of Platecarpus. Both specimens were taken from the Mooreville Mem- ber. Before the Field Museum of Natural History be- gan collecting in the Selma Chalk in 1945, the known mosasaur fauna consisted of: Clidastes propython, near Uniontown, Perry County Globidens alabamaensis, near Hamburg, Perry or Dallas County Platecarpus sp., near Eutaw, Greene County "Platecarpus" intermedius, Pickens County Tylosaurus sp., near West Greene, Greene County SYSTEMATICS Subfamily Mosasaurinae Halisaurus sternbergi Clidastes slernbergi Wiman 1920, Bull. Geol. Inst. Upsala, 18, p. 13, figs. 4-9, pis. 3-4. Referred material. — PR 186 frontal, pterygoid, and dentary fragments, one dorsal, and one pygal vertebrae. From West Greene, Greene County, Alabama. Col- lected by C. M. Barber. PR 195 frontal, basiocccipital, two cervical, 19 dorsal, and ten caudal vertebrae. From Harrell Station area, southeast of Marion Junction, Dallas County, Alabama. Collected by J. A. Robbins. Revised generic diagnosis. — Premaxilla "U"-shaped in horizontal cross-section, no rostrum present anterior to premaxillary teeth. Premaxillo-maxillary suture rises 369 370 FIELDIANA: GEOLOGY MEMOIRS, VOLUME 3 vertically from ventral jaw margin, makes abrupt turn and continues posteriorly in nearly straight line to posi- tion dorsal to eleventh maxillary tooth. Suture smoothly keeled and parallels longitudinal axis of maxilla. Su- tures for prefrontal and postorbitofrontal very widely separated on undersurface of frontal. Lateral margins humerus only slightly expanded, radial and ulnar tubei-- osities absent, well-developed spherical head present. Radius elongate, proximal end slightly expanded. Shaft of radius narrow. Distal end bears moderately well- developed anterodistal flange. Fig. 164. Lateral view of A, anterior cervical, and B, posterior cervical or anterior thoracic vertebra of Halisaurus sternbergi (PR 195, X 1). The outline of the anterior central articulation is shown beneath B. of frontal nearly straight, and converge only slightly anteriorly, median dorsal ridge rounded and large. Me- dian dorsal surface of parietal very broad, parietal fora- men moderately large. Foramen for cranial nerve VII enters prootic in center of prootic incisure. Supra- stapedial process of quadrate moderately large, tym- panic ala thick. Infrastapedial process absent on quadrate. Surangular does not extend behind mandib- ular cotylus. Vertebral formula: 31 presacral vertebrae, 4 pygals, 72 chevron bearing caudals. Articulating surfaces of cervical and anterior dorsal vertebral centra nearly twice as wide as deep, sub- rectangular in outline. Synapophysis located in center of lateral surface of cervical centra, occupies somewhat more posterior position in anterior thoracics. Ventral border of anteroventral extension of synapophysis weak and horizontal in anterior cervicals, becomes much en- larged in posterior cervicals and anterior thoracics, ex- tending far below flattened undersurface of centrum. Anterior zygapophysis of cervical and anterior thoracics connected by gently rounded, posteriorly descending crest to synapophysis. Zygosphene-zygantrum weak or absent. Hypapophyseal peduncle located posteriorly on ventral surface of cervical centra, articulation for hypapophysis flat and lenticular, slightly inclined pos- teriorly. Neural spines of caudal vertebrae longest on postsacrals 30 -32, do not become vertical. Scapula much smaller than coracoid. Superior bor- der of scapula gently convex, posterior border emargi- nate. Coracoid expands broadly behind glenoid articulation. Distal and proximal ends of slender At least four ossified elements in carpus. Meta- carpal one equal to metacarpal two in length, antero- distal flange small or absent. Proximal ends of metacarpals, especially of two and three, not greatly expanded. Acetabular surfaces of pelvic elements convex, do not form solid smoothly-surfaced bowl. Obturator fora- men located near center of proximal end of pubis, dorso- anterior process rudimentary. Ischiadic tubercle sep- arated from acetabulum by short neck. Shaft of femur slender, distal end more expanded than proximal, well- developed spherical head present. Facets for tibia and fibula distally. Tibia and fibula slender, slightly ex- panded at distal and proximal ends. Discussion. — Halisaurus was founded by Marsh (1869) on two peculiar vertebrae from the latest Creta- ceous green sands of New Jersey. Merriam (1894) later recorded the presence of similar vertebrae in the Nio- brara Chalk of Kansas, but the general body form of the reptile remained unknown. The two Selma specimens mark an important range extension of the genus and provide evidence showing that Wiman's (1920) remark- ably complete skeleton of a clidastoid mosasaur from the Niobrara Chalk should be referred to Halisaurus. The vertebrae of the anterior portion of PR 195 so closely resemble the two type vertebrae of Halisaurus platyspondylus from the Maestrichtian of New Jersey that there can be little doubt about their generic assign- ment. The intervertebral articulating surfaces are nearly twice as wide as deep and the ventral border of the anteroventral extension of the synapophysis, ini- tially weak and horizontal, becomes heavily developed RUSSELL: VERTEBRATE FAUNA OF SELMA FORMATION 371 and projects far below the flattened undersurfaces of the anterior dorsal vertebrae. The hypapophyseal facets are somewhat more circular than in the type of H. platy- spondylus, although they are slightly dilated distally. Posteriorly, intervertebral articulations of the dorsal Fig. 165. Frontal of Halisauriis slernbergi, dorsal aspect (recon- structed after PR 195, X 2/3). Abbreviations: pf, posterior limit of prefrontal suture; pof, anterior limit of postorbitofrontal suture. vertebrae become more circular, as Merriam (1894, p. 36) pointed out, and the vertebrae are difficult to distinguish from slightly crushed Clidastes dorsals. How- ever, the zygosphene-zygantrum articulations are only about as well developed as they are in Platecarpus. The anterior caudal centra are indistinguishable from those of Clidastes. The frontals of PR 186 and PR 195 resemble those of Clidastes very closely. The broad anterior end of the bone gives it a more rectangular outline than in Cli- dastes liodontus or C. propython, and the large rounded median ridge anterior to the orbits is nearly absent in these species. The sutures for the prefrontal and post- orbitofrontal on the ventral surface of the frontal are much more widely spaced than in either of the above forms. However, these peculiarities are present in the frontal of "Cliastes" sternbergi (Wiman, 1920, fig. 4) and probably in another specimen from the Niobrara Chalk (USNM 3777) elsewhere referred to this species. On this basis it would seem likely that "Clidastes" stern- bergi should be referred to Halisaurus, and there are certainly sufficient characters in the type skeleton to justify its removal from Clidastes. The quadrate and posterior half of the lower jaw included by Merriam (1894) in his type of Halisaurus onchognathus, also from the Niobrara Chalk, probably belong to the type of Edenosaurus didastoides, and the peculiar nature of the retroarticular process in the latter element may be due to distortion, as I have never seen a similarly shaped element in any of the extensive Niobrara collections. The basioccipital of PR 195 is similar to that of Cli- dastes. Teeth preserved in a pterygoid fragment of PR 186 are rounded in cross-section and distinctly pos- teromedially recurved. The enamel surface is smooth and there is a faint posteroe.xternal carina. Measurements: — PR 186 length median dorsal vertebra 41 mm. length pygal vertebra 34 interorbital width frontal 43 PR 195 length anterior cervical vertebra 39 length anterior dorsal vertebra 45 length posterior dorsal vertebra 54 length pygal vertebra 42 length anterior caudal vertebra 31 interorbital width frontal 47 Clidastes propython Clidastes propython Cope, 1869, Proc. Boston Soc. Nat. Hist., 12, p. 258. Referred material. — Harrell Station Area, Southeast of Marion Junction, Dallas County, Alabama: P 27324, disarticulated skull, presacral vertebrae. Collected by C. M. Barber. P 27428, pterygoid, six vertebrae. Col- lected by C. M. Barber. P 27439, parts of skull, atlas, cervical vertebrae. Collected by C. M. Barber. P 27440, scapula and coracoid fragments. Collected by R. Zan- gerl. P 27449, humerus. Collected by C. M. Barber. P 27478, tooth, metacarpal. P 27484, fragmentary disarticulated skull, presacral vertebrae. Collected by R. Zangerl and W. Turnbull. P 27545, humerus. Col- lected by R. Zangerl. P 27546, quadrate. Collected by C.M.Barber. PR 37, cranial elements, humerus. Col- lected by C. M. Barber. PR 38, disarticulated skull, vertebrae. Collected by A. Zangerl. PR 39, cranial elements, presacral vertebrae. Collected by A. Zan- gerl. PR 40, disarticulated skull, most of vertebral col- umn, appendicular elements. Collected by A. Zangerl. PR 41, cranial elements, vertebrae, hind limb. Col- lected by R. Zangerl. PR 42, vertebrae. Collected by A. Zangerl. PR 43, vertebrae. Collected by A. Zan- gerl. PR 45, cranial elements, presacral vertebrae, pec- toral elements. Collected by A. Zangerl. PR 46, verte- brae. Collected by R. Zangerl. PR 47, presacral ver- tebrae. Collected by C. M. Barber. PR 48, cranial elements, presacral vertebrae, humerus. Collected by R. Zangerl. PR 49, pygal vertebrae. Collected by CM. Barber. PR 50, dorsal vertebrae. Collected by R. Zangerl. PR 52, disarticulated skull, atlas. Col- lected by C. M. Barber. PR 53, caudal vertebrae. Col- lected by A. Zangerl. PR 138, quadrate, marginal tooth, caudal vertebra. PR 142, cranial elements, vertebrae. Collected by C. M. Barber. PR 192, pectoral elements, vertebrae. Collected by J. A. Bobbins. PR 200, pygal vertebra, phalanges. Collected by C. M. Barber. PR 21 1, postorbitofrontal. PR 238, dentary. 372 FIELDIANA: GEOLOGY MEMOIRS, VOLUME 3 Fig. 166. The hind limb of Clidastes. ventral aspect. A, C. liodontus (after Williston, 1898, pi. 34, X 2 3), B,[C. propython (PR 40, X 2/3). Abbreviations: as, a.stragulus; f, femur; fi, fibula; t, tibia. Cedarville, Hale County, Alabama: PR 51, verte- brae. Collected by A. Zangerl. Boligee, Greene County, Alabama: PR 63, humerus. Collected by W. Turnbull. Mount Hebron, Greene County, Alabama: PR 150, cranial elements, cervical vertebrae. Collected by J. A. Robbins. West Greene, Greene County, Alabama: PR 148, cranial elements, vertebrae. Collected by J. A. Rob- bins. PR 156, cranial elements, vertebrae, humerus. Collected by R. Zangerl. PR 162, quadrate, vertebrae. Collected by C. M. Barber. PR 164, disarticulated skull, most of vertebral column, appendicular elements. Collected by J. A. Robbins. PR 175, dorsal vertebrae. Collected by J. A. Robbins. PR 176, caudal vertebrae. Collected by J. A. Robbins. PR 183, radius. Collected by C. M. Barber. Locality unknown. — P 27355, dentary and splenial. Specific diagnosis. — Posteroventral portion of root of second premaxillary tooth exposed on sutural surface with maxilla. Premaxillo-maxillary suture rises poste- riorly in gentle curve to terminate at point above seventh maxillary tooth. Premaxillary suture of maxilla smoothly keeled and parallels longitudinal axis of maxilla. Six- teen to 18 teeth in maxilla. Median dorsal surface of parietal moderately broad. Infrastapedial process pres- ent on quadrate. Sixteen to 18 teeth in dentary. Discussion. — Clidastes propython was first described from the Selma Chalk of Alabama and has subsequently been found in the upper Niobrara and lower Pierre For- mations of Kansas and South Dakota. It is well repre- sented in the Chicago collection from the Selma, com- prising about three-quarters of all mosasaur specimens. Relative to individuals of the species from the Niobrara Chalk, those from the Selma tend to be larger, ranging from an estimated 73^ ft. in length (PR 148) to 20 H ft. (P 27324, PR 156, PR 162), the latter specimens being the largest known. The fact that the ventral surface of all the preserved parietals in the collection (PR 37, PR 40, PR 52, PR 164, P 27324) is not as deeply exca- vated behind the parietal foramen as is the type of C. propython or in referred specimens from the Niobrara Chalk is of unknown significance. It is of interest to note that one complete dentary (P 27355) has only 16 teeth, the least number recorded for the species. Complete skeletal material of Clidastes propython has never been described and it was provisionally as- sumed (Russell, 1967) that the body of this species closely resembled that of C. liodontus. The presacral vertebral count may have been very close in the two species, for in PR 164 there were at least 38 vertebrae from this region, while in a complete skeleton of C. lio- dontus there are 42 (Williston, 1898, p. 143). However, in certain poorly-preserved tails of C. propython (PR 40, PR 53, YPM 24901, see also Cope, 1869-1870, pi. 12, fig. 11) it seems likely that the posteriormost transverse processes are present beneath the longest neural spines in the dilated portion of the caudal fin. In C. liodontus they occur about 9 10 vertebrae in front of this region. Elements of the hind limb preserved in several speci- mens (PR 40, PR 41, PR 164) also show differences in detail from those of Clidastes liodontus (see Williston, 1898, p. 165, pi. 34), reflecting the specific distinctness of these two forms. On the distal end of the femur the facets for the tibia and fibula meet at a slightly higher angle in C. propython and the fibular facet occupies a more posterodistal position on the bone. Distal and proximal portions of the anterior flange on the tibia are separated by a relatively wide sulcus in C. liodontus, which has been reduced to a notch in C. propython. RUSSELL: VERTEBRATE FAUNA OF SELMA FORMATION 373 There seem to be facets on the tibia and astragulus for a centrale in the latter species, which are absent in Williston's descriptions and figures of C. liodonius. One specimen (PR 164) is of particular interest in that many vertebrae in the posterior dorsal and anterior caudal regions have been fused, and the vertebral centra are swollen, especially around the former location of the intercentral articulation. The resulting rigidity of the vertebral column must have seriously impaired the animal's ability to swim, and it very likely starved to death as a result of this disability. Globidens alabamaensis Globidens alabanuiensis Gilmore, 1912, Proc. U. S. Nat. Mus., 41, p. 479. Referred material. — PR 196, frontal fragment and pygal vertebra. From Harrell Station area, southeast of Marion Junction, Dallas County, Alabama. Col- lected by C. M. Barber and J. A. Robbins. Discussion. — A piece of the right side of a heavily constructed frontal may belong to Globidens. The supra- orbital border is rounded and, as in the type, the sutures for the prefrontal and postorbitofrontal do not meet on the undersurface. The pygal vertebra measures 65 mm. in length but is too poorly preserved to show other sig- nificant features. Subfamily Plioplatecarpinae Platecarpus sp. Referred material. — P 27399, disarticulated skull, nearly complete vertebral column, ribs, elements of pec- toral and pelvic appendages. From Harrell Station area, southeast of Marion Junction, Dallas County, Alabama. Collected by C. M. Barber. Discussion. — Platecarpus is represented by a well- preserved skeleton of a single individual. The animal was probably about 15 feet long, being of slightly less than average size. It could belong to P. ictericus or P. coryphaeus, but the jaws are not complete enough to assign it to one of these species. It most likely belongs to P. ictericus, for the Mooreville Member of the Selma Chalk is stratigraphically equivalent to the upper por- tion of the Niobrara Chalk in Kansas where P. cory- phaeus evidently does not occur. Prognathodon sp. Referred material. — Harrell Station Area, southeast of Marion Junction, Dallas County, Alabama: PR 143, tibia. PR 193, tibia, fragmentary ischium. Collected by Kirtley Brown. Eutaw, Greene County, Alabama: PR 146, surangu- lar fragment. Collected by C. M. Barber. West Greene, Greene County, Alabama: PR 165, fragmentary skull, 17 presacral and 38 caudal vertebrae, elements of pelvic appendage. Description. — A fragmentary skull and skeleton (PR 165) and a few isolated bones indicate the presence of a very large plioplatecarpine in the Selma Chalk. The abbreviated premaxilla, massive jaws and teeth, and relatively slender splenial and angular suggest the specimen should be referred to Prognatkodon. The re- mains indicate an animal somewhat larger than P. over- toni and P. rapax, and approximately twice as large as P. solvayi. The premaxilla is short and broad, and is very sim- ilar to that of Platecarpus. Sutural surfaces on the an- terior end of the frontal show that the internarial bar terminates in two small tongues, one on either side of the midline of the skull, in front of the posterior end of the external nares. The preserved anterior portion of the frontal seems to be narrower transversely than that of Prognatkodon overtoni. The posterior end of the ex- ternal naris is contained in a notch between the anterior edges of the frontal and prefrontal, and does not lie en- tirely within the frontal as it does in Platecarpus. The squamosal is indistinguishable from that of Platecarpus. There is an extremely large postero- inferior process on the jugal, very unlike Prognatkodon overtoni and P. sol- vayi where the process is nearly absent. A poorly pre- served left quadrate appears to closely resemble that of Platecarpus, as the tympanic ala was thin, the supra- stapedial process is unconstricted dorsally, and the supra- and infrastapedial processes may not have been fused. There is a bilobate channel through the basioc- cipital for the basilar arteries. The anterior end of the dentary is more massive than in both Prognatkodon overtoni and P. solvayi, but other- wise bears a closer resemblance to that of the latter spe- cies. Its alveolar margin is similarly more concave upward than in P. overtoni, and the anteriormost teeth Fig. 167. Dentary of Prognalliodon sp. (PR 165, X 1/3), lateral aspect. 374 FIELDIANA: GEOLOGY MEMOIRS, VOLUME 3 in the dentary are procumbent, as are those of the pre- maxilla. The enamel surfaces of the marginal teeth in the Selma Prognathodon and in P. solvayi are divided into vertical prisms, while in P. overtoni, P. rapax, and P. giganteus the enamel surfaces are smooth. The dor- FlG. 168. A, Lateral view of left atlas arch, and B, lateral view of left jugal of Prognathodon sp. (PR 165, X 1/2). Abbreviation: syn, synapophyseal proces.s. sal edge of the surangular is wide and rounded in cross- section in front of the mandibular cotylus in the present form. Although the vertebral column is poorly preserved in PR 165, a few vertebrae are complete enough to show several interesting characters. The body of the atlantal neural arch is a relatively thick and heavily built piece of bone, with a small synapophyseal process located slightly above the center of its posterior border. In Clidastes, Mosasaurus, Platecarpus, Ectenosaurus, and Tylosaurus the synapophyseal process is usually larger and much more ventral in position. A small, longitudi- nally compressed ridge near the posterior base of the spinous process of the atlas marks the site of a tendonous attachment of much of the cervical longissimus muscles. On the axis centrum the facet for the hypapophysis is located close to the suture for the axis intercentrum, as in Tylosaurus but not in Platecarpus and Clidastes, where it occupies a more posterior position. In the dorsal series the zygosphene-zygantral articulations are about as well developed as in Platecarpus, being for the most part rudimentary. The haemal arches articulate with the caudal centra as in all species of Prognathodon except P. overtoni. There are several elements of the posterior append- ages in PR 165 which were previously unknown in Prognathodon . Facets on the ilium for the ischium and pubis are more distinctly developed than is usual in Platecarpus and the central portion of the ilial shaft seems to have been more anteroposteriorly expanded than in this genus. However, the proximal portion of the ischium is very much like that of Platecarpus. Only the distal halves of both femora are preserved and these conform in shape to that of Prognathodon crassartus fig- ured by Cope (1875, pi. 26, fig. 10). The femur of Prog- nathodon differs from that of Platecarpus in that the edges of the central portion of the shaft parallel each other between the expanded distal and proximal ends. This gives the l)one a dumbbell-shaped outline unlike the more smoothly hourglass-shaped outline of the fe- mur of Platecarpus. The tibia is a columnar bone. Its equally expanded ends were probably finished in thick cartilage in life. A weathered astragulus appears to be generally similar to that of Platecarpus. It is apparent from the foregoing description that the Selma Prognathodon may be distinguished from all the known species of the genus with the exception of P. crassartus, from the lower Pierre Shale of Kansas. In view of the fact that the humerus of the Selma form is unknown and that it is from a horizon equivalent to the lower Pierre in stratigraphic position, a proper eval- uation of the affinities of both forms should await the discovery of new material. Measurements: — PR 165 height of quadrate 139 mm. length between 1-6 tooth, right dentary 252 length between 1-6 tooth, left dentary 246 length median cervical centrum 68 Subfamily Tylosaurinae Tylosaurus zangerii new species' Type. — P 27443, humerus and femur. Collected by C. M. Barber. Horizon and locality. — Lower, marly member of the Selma Formation (Mooreville Chalk), Late Cretaceous. High river exposures, Moore farm, southeast of Marion Junction, Dallas County, Alabama. Description and diagnosis. — A small but excellently preserved humerus and femur mark the occurrence of a new and relatively primitive species of Tylosaurus in the Mooreville Chalk. The humerus is more slender than in Niobrara species of the genus and the ends are much more abruptly terminated. It may at once be distin- guished from the similarly proportioned humerus of Halisaurus by the absence of a spherical head. The pectoral and deltoid crests are separated from each other by a deep sulcus, typical of Tylosmirus. The pec- toral crest is nearly as anteriorly located on the shaft of the humerus as is that of Clidastes, and it extends proximally to the glenoid surface, while in Niobrara tylosaurs these two regions are partly or completely separated by a transverse groove. The deltoid crest is less well developed than in Niobrara specimens, al- though it occupies a similar position. On the distal end of the bone a I'adial process is completely lacking, which is normal for Tylosaurus, and a well-marked ectepicon- dylar foramen penetrates the anterodorsal edge. The ulnar process is exceptionally well developed for the genus. As may be seen from Figure 6, the areas of mus- cle attachment are very clearly defined. ' The species is named for Dr. Rainer Zangerl of the Field Museum of Natural History. RUSSELL: VERTEBRATE FAUNA OF SELMA FORMATION 375 Fig. 169. Humerus of Tylosaurus zangerli (P 27443, X 1). A, ventral view, B, dorsal view, C, proximal view. Abbreviations: dc, deltoid crest; ef, ectepicondylar foramen; pc, pectoral crest; up, ulnar process. Muscle scars: del, M. deltoides, undivided; hr, M. humeroradialis; Id, M. latissimus dorsi; sea, M. scapulohumeralis anterior; scp, M. scapulohumeralis posterior; ssc, M. subscapulocoracoideus. The femur is also slender, relative to its proportions in Niobrara specimens. The internal trochanter is nar- rower and slightly more posteriorly located. The distal end is squarely terminated in dorsal or ventral aspect and lacks the bell-shaped profile characteristic of the femur in Niobrara tylosaurs. Measurements: — femoral length 82.2 humeral length 82.2 Tylosaurus sp. Referred specimens. — Harrell Station Area, southeast of Marion Junction, Dallas County, Alabama: PR 204, cranial fragments, one dorsal and 28 caudal vertebrae, elements of pelvic appendage. Collected by C. M. Barber. Eutaw, Greene County, Alabama: P 27474, inter- narial bar of premaxilla. Fig. 170. Femur of Tylosaurus zangerli (P 27442, X 1), ventral aspect. Abbreviation: it, internal trochanter. Fig. 171. Splenio-angular articulation of Tylosaurus up. (PJl 204, X l/2\ medial aspect. Abbreviations: an, angular; p, proces.s of splenial; sp, splenial. Discussiou. — Remains of a medium-sized Tylosaurus (PR 204), which may have originally measured 20 ft. in length, are present in the Field Museum collections. There is a small conical process on the posterior rim of 376 FIELDIANA: GEOLOGY MEMOIRS, VOLUME 3 a o c