fv. rary i -* ^.vcLutii uns DOOR^TI or DSrolWny Latest Date stamped below. A charge is made on all overdue books. U. of I. Library ' 22 104 FEE 5 A JAN 2 9 1984 '.] FIELD COLUMBIAN MUSEUM , PUBLICATION 13. ZOOLOGICAL SERIES. VOL. i, No. 5 ON THE SKELETON :" . OF '; TOXOCHELYS LATIREMIS BY O. P. HAY, PH.D. Assistant Curator of Zoology. D. G. ELLIOT, F.R. S.E. , Curator of Department. CHICAGO, U. S. A. June, 1896. PUBLICATIONS OF THE MUSEUM. For the convenience of scientific workers it has been deemed expedient to issue the publications of the Museum in separate series for each of the sciences represented. The following series have been established : Historical, Geological, Botanical, Zoological, Ornitho- logical and Anthropological. A consecutive number has been given the entire set of pub- lications to indicate the order of issue. Each departmental series, however, has its own volume number and individual consecutive pagination, making the literature of the science, or general subject, independent and complete for separate binding ; or they may be bound in order of publication without relation to subject. Proceed- ings, memoirs, monographs, bulletins, and hand-books and catalogues of collections, are included within the scope of the publications. Full lists of the publications of the Museum may be found in the Annual Report of the Director. Publications are sent to societies and institutions of a public character that reciprocate with their own literature, and to a limited number of scientists who are able to exchange. Four publications have been previously issued in the Zoological Series: HAY, " On the Structure and Development of the Vertebral Column of Amia "(Pub. 5); HAY, "On Certain Portions of the Skeleton of Protestega gigas" (Pub. 7); ELLIOT, "On Sundry Col- lections of Mammals" (Pub. n), and HAY, "On Some Collections of Fishes " (Pub. 12). FREDERICK J. V. SKIFF, Director. ON THE SKELETON OF TOXOCHELYS LATIREMIS. O. P. HAY. PH.D. The genus Toxochelys, having as its type T. latiremis, was origi- nally described by Professor E. D. Cope in the Proceedings of the Academy of Natural Sciences of Philadelphia for 1873, page 10. In the year 1875 the same author, in his Cretaceous Vertebrata, p. 98, pi. viii. figs, i and 2, vol. n, U. S. Geol. Survey, further described the original species and (op. cit. p. 299) a new one under the name of T. serrifer. Of T. latiremis Professor Cope possessed only a por- tion of a lower jaw and the coracoid, which he figured, and some of the phalanges. Of T. serrifer the materials were somewhat ampler, consisting of some skull bones and two marginals; but none of these were figured. In the Proceedings of the American Philosophical Society for 1877, p. 176, Professor Cope gave descriptions of two skulls of T. latiremis which had by this time come into his hands. From these, Professor Cope came to the conclusion that the general form of the skull was much like that of some of the Trionychida, but that from this family Toxochelys was separated by the possession of marginal bones and the character of the extremities. Before this, however, in his description of T. serrifer, Professor Cope had regarded the genus as having relationships with the Cheloniidce and the Chelydrida. In his description of T. latiremis in the same work he had spoken of the phalanges as being flattened, as in Protostega. As will be observed during the discussion of the skeleton of this genus, I conclude that it is related to Cheloniidce and Chelydridce, and not at all to the Triony- chidce. The materials in my hands consist of a single skull of T. latir- emis, whose length from the snout to the occipital condyle is close to 115 mm. The distance from the alveolar border to the condyle of the quadrate is about 105 mm. The length of the lower jaw described and figured by Professor Cope was 157 mm; so that it must have be- longed to a skull having a length from snout to condyle of about 170 mm. The skull in my hands has been broken across the middle, and some parts of the roof of the mouth are missing; but nothing essential for our understanding of the structure. The damage to the upper por- 101 102 FIELD COLUMBIAN MUSEUM ZOOLOGY, VOL. i. tion of the skull is more serious, since the frontals are wholly gone, as well as the postfrontals, the squamosals, the quadratojugals, and most of the parietals and of the jugals. A study of the parts remain- ing yields me the following results: 1 . As already observed by Professor Cope, the choanae are located far forward in the roof of the mouth, as they are in Chelydra; that is, immediately below the union of the prefrontals with the vomer. It will be seen from the figure on Plate xv that the descending processes of the prefrontals have been forced by pressure into the choanae. In the Cheloniidce, through the development of horizontal laminae from the vomer, the maxillaries, and the palatines, the choanae are thrown farther back in the roof of the mouth. 2. A vacuity, the palatine foramen, is present on each side of the roof of the mouth, lying in the point of meeting of the palatine, the maxillary and the pterygoid. This foramen is present in the Chelydridce and most other turtles, but is absent in the Cheloniidce and the Dermochelyidce. Dr. G. Baur (Zool. Anzeiger, vol. 12, p. 45) states that this foramen is placed between the palatine and the maxil- lary and is present in all turtles except the two families last named. However, in many, possibly all, of the Pleiirodira the foramen is located in the hinder border of the palatine, out of contact with the maxillary. 3. The cutting edge of the maxillary is very narrow, as we find it in Chelydra. In the Cheloniida and especially in Che Ionia the alveolar wall is proportionally much deeper, as shown by the distance of its lower border from the roof of the mouth and from the lower rim of the orbit. 4. The width of the horizontal alveolar surface, in proportion to the width of the roof of the mouth, much more closely resembles that of Chelydra than it does that of any of the Cheloniidce. The width of the effective crushing surface in Toxochelys is increased by the parti- cipation of the palatine. The outer edge of this bone extends from the outer anterior border of the palatine foramen to a point somewhat in front of the choana, nearly reaching the premaxillaries, an ar- rangement whereby the maxillary is excluded from the latter opening. Parallel with and close to the outer border of the palatine is a moder- ate, but distinct, roughened ridge, corresponding with which there was probably a rough triturating ridge on the horny beak. In Chelydra and Macrochelys the palato-maxillary suture passes directly between the choana and the palatine foramen; but mesiad of the suture there is a slight roughened ridge, as in Toxochelys. In the structure of these parts we have a combination of the characters found in both of the families referred to. 1896. SKELETON OF TOXOCHELYS HAY. 103 5. The structure and the connections of the anterior ends of the pterygoids are greatly like those of the Chelydridw. The ectopterygoid processes are prominent, their posterio-external angle directed back- ward, and their outer border broad for muscular attachment, just as in Chelydra. 6. The orbits are of moderate size, being intermediate between those of Chelydra and Chelonia. The plane of the rim has been di- rected, as in Chelydra, outward and upward. In this respect the orbits resemble those of Chelydra, but not those of Macrochelys. In this genus the orbits appear to have been given a perpendicular rim through the excessive narrowing of the snout, just as is also the case with Eretmo- chelys. 7. The external border of the pterygoid bone runs backward as a sharp keel terminating on the mesial side of the pedicel of the quad- rate. In this arrangement Toxochleys resembles Chelydra, Thalasso- chelys, Lepidochelys, and Eretmochelys, but not Chelonia. In the last genus the external border of the palatal portion of the pterygoid ceases abruptly on the under side of that portion of the bone which lies between the quadrates, thus leaving a broad notch in the ptery- goid. 8. That portion of each pterygoid which lies between the quad- ratic pedicels is furnished with a broad groove which terminates posteriorly behind the inner border of the pedicel. This groove does not appear to have had any considerable depth, but some allowance must be made for the effects of pressure. In Chelydra this region is flat and smooth; in Chelonia it is also nearly flat, while the border is notched. 9. Between the posterior ends of the pterygoidean grooves, on the basisphenoid and the basioccipital, is found an elevated triangu- lar rough surface for the attachment of muscles. A similar surface is found here in all the Cheloniidce, but not in the Chelydrida. 10. In Chelonia, there is found below and behind each jugular foramen a strong process into which enter the basioccipital, the exocci- pital, and the pterygoid. This process is wanting in the Chelydridcc and most other turtles, but is present in the Cheloniida. It was well developed in Toxychelys. In the Chelydrid(e\\\z exocipital contribution to this process is well developed and its outer extremity extending across the jugular foramen joins the paroccipital, thus cutting off a portion of the jugular foramen as a special pneumogastric foramen. In Toxochelys the arrangement of the parts is as in Chelonia. 11. The tympanic cavity closely resembles that of Chelonia. It has, however, been extended downward further on the pedicel of the quadrate, since this is distinctly concave nearly to the articulate 104 FIELD COLUMBIAN MUSEUM ZOOLOGY, VOL. i. surface for the lower jaw. Furthermore, there has been a deep notch in the upper and posterior part of the quadrate. This suggests that the tympanic cavity has been supplemented by a cavity in the squamosal. In Chelonia there is no such notch, but the quadrate is deeply scooped out in that direction. In Chelydra the squamosal is much enlarged backward and excavated to a thin shell of bone and forms a large bulla, into which the cavity in the quadrate freely opens. Most living turtles, outside of the Dermochelyidce and Chelo- niidce, possess more or less developed squamosal bullse. While that of Toxochelys was probably not greatly developed, there was a beginn- ing of the structure. 12. The quadrate bears an open stapedial notch, as it does in the Cheloniidce but not in the Chelydridce. 13. The parietals are mostly wanting in my specimen. On the left side there is a portion of the bone which entered into the side wall of the brain-case. From the upper outer edge of this piece there jutted out the plate which helped to roof over the temporal fossa. How far backward this roof extended, and whether or not there was produced a parieto-squamosal arch, cannot be determined; but, at any rate, at the hinder end of the piece which remains, the horizontal plate is quite thick. Professor Cope (Proc. Amer. Phil. Soc. 1877, P- X 7^) states that the temporal fossae, as shown in his skulls, were extensively roofed over. We are therefore justified in concluding that the arrangement was quite like that in the Cheloniidce. In front of the piece of the parietal referred to, there are seen on each side the lower ends of the descending processes of the parietals. 14. The frontals of my specimen are wholly missing. Pro- fessor Cope in describing T. serrifer (Cret. Vert., p. 300) indicates that they entered into the rim of the orbits. This is most probable, judging from the shortness of the prefrontals of my specimen. In the Chelydridce the frontals are widely excluded from the orbital rims. 15. A portion of the jugal is still present, bounding the left orbit. It has been deeply excavated on the lower border, while the bone between this excavation and the rim of the orbit is very narrow. 1 6. From the above examination of the skull of Toxochelys it is seen that in general the anterior portion resembles more or less closely that of Chelydra, the posterior portion that of the Cheloniidce. In the anterior portion the most important deviation from the Chelydrine pattern consists in the extension of the palatine around the outside of the choanae. In the hinder portion there is no essential deviation from the Chelonian type, unless it shall prove to be in the JUNE, 1896. SKELETON OF TOXOCHELYS HAY. 105 lack of a parieto-squamosal arch and in the occurrence of a well developed squamosal bulla, neither of which things is probable. In its general form the skull of Toxochelys is more like that of the Cheloniidce than it is like that of the Chelydridce. As to the hinder part, there is no question. Anteriorly the snout is broad and rounded, not acuminated, as it is in the Chelydridce. The large orbits also remind us of the sea-turtles. The skull appears to have been rather depressed, as in Thalassochclys; not high, as in Chelonia. The lower jaw figured by Professor Cope resembles that of Chclydra; but is not hooked, and has no decided cutting edge. Professor Cope has figured the coracoid of his specimen. In form and proportions it closely resembles that of Chelonia. Its length was 250 mm. The coracoid of a specimen of Chelonia, whose shell is i metre long, is 300 mm. in length and the ramus of the lower jaw is 150 mm. Since the proportional lengths of the coracoids in Chelydra and Chelonia are about the same, the shell which belonged to Professor Cope's specimen was probably about 825 mm. long, and the head was somewhat larger than that of Chelonia. ' Dr. Joseph Leidy, in his "Contributions to the Extinct Vertebrate Fauna of the Western Territories," page 269, plate 36, 1873, has de- scribed and figured some turtle bones. These he refers with some doubt to Professor Cope's genus Cynoccrcus. A comparison of Dr. Leidy 's figures 20 and 21 with Professor Cope's figures of the coracoid of Toxochelys convinces me that they belong to the same animal. Be- side the portions of the two coracoids, Dr. Leidy possessed a portion of* the scapula, a complete femur, and the proximal end of the humerus. Notwithstanding that these bones are somewhat fragmen- tary and more or less flattened by pressure, we may learn something from a comparison of them with the corresponding bones of the Chelydridce and the Cheloniidce. One character which as much as any other distinguishes the sea- turtles from other turtles is the great development of the anterior limbs. In most turtles the humerus has about the same length as the femur. This is true of the Chelydridce and such habitual swimmers as the Trionychidce. But in Chelonia the femur is only two-thirds the length of the humerus. Now, it is not wholly certain that the humerus and the femur in Dr. Leidy's hands belonged to the same skeleton, but he supposed that they did. If so, then the humerus of that animal must, on proper estimate, have been nearly as much longer than the femur as that bone is in Chelonia. This deduction, taken in connection with Professor Cope's statement that the phalanges are flattened, compels us to regard Toxochelys as a true sea-turtle, as does also the fact that it is found so abundantly in the Cretaceous deposits of Kansas. 106 FIELD COLUMBIAN MUSEUM ZOOLOGY, VOL, i. Professor Cope has himself suspected that his genus Cynocercus is identical with Toxochelys. If this is true the latter name becomes a synonym of Cynocercus. This genus was founded on a metapodial bone and some caudal vertebrae. As Professor Cope has observed these vertebrae differ from those of the Chelydridce in being procoelous. If the two genera referred to are identical, we have in these caudal vertebrae an additional evidence of the relationship of our fossil to the Cheloniidtz. In short, I conclude that Toxochelys is related both to the Chely- dridce and the Cheloniidce, but that the relationship is much closer to the last named family, and with the carnivorous division of this family, Thalassochelys, etc. As to its systematic position, I agree with Dr. G. Baur's recent statement that it should form the type of a new family. As to its phylogenetic relations, Toxochelys is probably an off- shoot from the line which led to the Cheloniidx, an offshoot after the latter family had disengaged itself from forms like the Chelydridce. The skull of Toxochelys is too much modified in the direction of the sea-turtles ever to be transformed into that of the modern snappers; and the same remark will apply to the flattened phalanges, the prob- ably greatly developed anterior limbs, and proccelous caudal vertebrae. While there appears to be no special reason why Toxochelys might not develop into a modern sea-turtle, we must remember that true Cheloniidce had already made their appearance at the time when Tox- ochelys left its remains in the upper Cretaceous deposits of Kansas. 3 ^>*V. UN1VE 1 11 1 11 HHI ''' p > ct'l7 m ^^Rdi^^HI PyBoy d^"