Novitates PUBLISHED by the AMERICAN MUSEUM of NATURAL HISTORY CENTRAL PARK WEST at 79TH STREET, NEW YORK, N.Y

AMERICAN MUSEUM Novitates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 2639, pp. 1-12, figs. 1-8 November 3, 1977

An Endocranial Cast of the Side-Necked Turtle, Bothremys, With a New Reconstruction of the Palate

EUGENE S. GAFFNEY'

ABSTRACT A natural endocast of the skull of Bothremys, cavities as well as the brain cavity. The endocast a pelomedusid pleurodire from the Late Cre- provides new information on the basicranium, as taceous (Magothy Formation of Cliffwood Beach, well as the orbital and nasal cavities, and allows New Jersey) is unusual in preserving the orbital comparison with Recent pelomedusids. INTRODUCTION The Cretaceous pleurodire, Bothremys, is (see Edinger, 1929, for a review), there are very known from a series of shells and two skulls, few reported for turtles. Furthermore, this endo- described and figured in Gaffney and Zangerl cast preserves high quality representations of the (1968, see also references to previous work on orbital and nasal cavities as well as the basi- this genus). Subsequent to the publication of cranium. The only natural turtle endocast re- that paper another specimen referable to Both- ported in the literature is that of a presumed sea remys has come to my attention and I have turtle (Edinger, 1934). In this specimen, from developed a new reconstruction of the palate. the Cretaceous of Europe, the cavum cranii and The purpose of the present paper is to give the adductor chambers are preserved with some indi- description of the new Bothremys specimen. The cations of the orbits. Artificially made endocasts relationships of Bothremys lie with the pelo- of turtles are reported by Edinger (1929), Zan- medusid pleurodires (ibid.) and a study of this gerl (1960), and Gaffney and Zangerl (1968). group is deferred to the future. The new Bothremys endocast comes from the The specimen described here is unique and Cretaceous coastal plain sediments of New Jersey unusual in the manner, as well as the taxon, and is preserved in a manner that seems to be preserved. The fossil consists of the infilling of a typical of invertebrate fossils from certain litho- skull by limonitic sandstone forming a natural logic units in this region. Unfortunately, there is cast of the internal cranial morphology. Al- little available information on the process of though fossil "brains" are not particularly rare steinkern formation and I am unable to advance

'Associate Curator, Department of Vertebrate Paleontology, the American Museum of Natural History; Adjunct Assistant Professor, Department of Geological Sciences, Columbia University.

Copyright © The American Museum of Natural History 1977 ISSN 0003-0082 / Price $1.00 AMERICAN MUSEUM Novitates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 2639, pp. 1-12, figs. 1-8 November 3, 1977

An Endocranial Cast of the Side-Necked Turtle, Bothremys, With a New Reconstruction of the Palate

EUGENE S. GAFFNEY'

'Associate Curator, Department of Vertebrate Paleontology, the American Museum of Natural History; Adjunct Assistant Professor, Department of Geological Sciences, Columbia University.

An Endocranial Cast of the Side-Necked Turtle, Bothremys, With a New Reconstruction of the Palate

EUGENE S. GAFFNEY'

'Associate Curator, Department of Vertebrate Paleontology, the American Museum of Natural History; Adjunct Assistant Professor, Department of Geological Sciences, Columbia University.

Copyright © The American Museum of Natural History 1977 ISSN 0003-0082 / Price $1.00 2 AMERICAN MUSEUM NOVITATES NO. 2639 any useful hypotheses about the taphonomy of ABBREVIATIONS the endocast. The rock forming the endocast is a relatively coarse-grained, quartz sandstone with INSTITUTIONS limonitic cement. Limonite formation seems to AMNH[H], Department of Herpetology, the have been concentrated on the bone surfaces American Museum of Natural History resulting in a very high fidelity image of the bone FMNH, Field Museum of Natural History morphology. Sutures, foramina, and impressions PU, Princeton University of vessels are all preserved in the endocast. The ANATOMICAL sandstone has not been analyzed petrographically but it does seem to contain a high percentage of ang, angular wood and plant fragments. The dorsal surface of art, articular bo, basioccipital the orbital cavities and the cavum cranii shows bs, basisphenoid the impressions of elongated lumps, possibly cor, coronoid representing some sort of invertebrate fecal mat- dent, dentary ter. epi, epipterygoid The Bothremys endocast has had an interest- ex, exoccipital ing scientific history. It was "found loose on the fr, frontal beach at Cliffwood, New Jersey, by C. J. Adami, ju, jugal Princeton '26, on May 28th, 1926" (label). It mx, maxilla was identified as a brain cast of a gavial and a na, nasal sketch (possibly by W. J. Sinclair) accompanying op, opisthotic pa, panetal the specimen shows that the endocast was re- pal, palatine versed with the foramen magnum identified as pf, prefrontal the nasal area and the orbits as pterygoid bullae. pm, premaxilla Later, the endocast became part of the Princeton po, postorbital Geology Museum "What is a Fossil?" display and pr, prootic remained there for some time. Recently, Dr. pra, prearticular Donald Baird, Director of the Museum, reori- pt, pterygoid ented the specimen and discovered that the qi, quadratojugal "bullae" were orbital infillings. He also identified qu, quadrate the endocast as Bothremys. Dr. Baird also dis- so, supraoccipital covered in the files a sketch made some time ago sq, squamosal by the late Dr. Tilly Edinger, the foremost sur, surangular student of fossil "brains." In this sketch the vo, vomer major features (carotids, cerebrum, olfactory SYSTEMATICS tract, pituitary) are correctly identified and the specimen is oriented properly. The identification ORDER TESTUDINES "turtle?" is also on the sketch. As far as I am SUBORDER CASICHELYDIA aware, however, there is no reference to this specimen in Edinger's published works. INFRAORDER PLEURODIRA FAMILY PELOMEDUSIDAE ACKNOWLEDGMENTS Bothremys sp. I am very grateful to Dr. Donald Baird, Direc- tor of the Geology Museum at Princeton Univer- Locality. "Found loose on the beach at Cliff- sity, who brought this specimen to my attention wood, N. J." (label). and also provided me with the background mate- Horizon. "Supposedly from the ferruginous rial on its history. The photographs are the work sand layer at the top of the Magothy Clay. Slabs of Mr. Chester Tarka and the drawings are by Ms. of similar matrix with marine Cretaceous shells Lorraine Meeker, both of the Department of are found loose on the beach" (label). Late Vertebrate Paleontology. Cretaceous. 1977 GAFFNEY: BOTHREMYS 3

Collector. C. J. Adami, May 28, 1926. walls of the fossa nasalis are represented in the The Bothremys endocast is conspicuously dif- endocast. The morphology is quite consistent ferent from most other natural endocasts in with that area of AMNH 2521 and little new preserving the orbital cavities and portions of the information can be obtained. The paired for- palate as well as the cavum cranii proper. The amina praepalatinum lie in the floor of the fossa identification of the endocast as Bothremys is and are prominent in the endocast. The central first of all based on its identification as a pleu- part of the endocast forming the fossa nasalis is rodire. The position of the foramen palatinum hollow, presumably due to weathering. Very posterius and the development of a postorbital little of the apertura narium externa is preserved wall is diagnostic for pleurodires (Gaffney, 1975) although the anterodorsal margin of it is buried and the endocast has the pleurodiran condition. in matrix. The posterior portion of the fossa Among pleurodires (actually, among all turtles nasalis, the apertura narium interna, the foramen that I am familiar with) the extensive enclosure orbito-nasale, and most of the dorsal surface of of the orbital cavities by bone and the dorso- the palate are not preserved in the endocast, lateral orientation of the orbital openings (Gaff- which has a large, irregular area missing in this ney and Zangerl, 1968, fig. 14) are unique to region. Bothremys. This combined with the skull width, The fossa orbitalis in PU 1295 1 is nearly internal morphology of the fossa nasalis, and the spherical and broadly connected to the cavum basicranial structure strongly substantiate the cranii region medially and posteromedially. The identification of the specimen as Bothremys. orbital openings are obscured by matrix antero- The Princeton specimen, however, does show medially but otherwise can be seen to agree with some differences from the other two known the openings in AMNH 2521. Bone remains Bothremys skulls. The Princeton endocast is deeply buried in the matrix in the area between nearly twice the size of the type specimen of the apertura narium externa and the orbits. The Bothremys cooki (AMNHi 2521) and about the visible interorbital cross-section shows a thick same size as the Selma skull B. barbeni (Gaffney sheet of bone comparable in depth to FMNH PR and Zangerl, 1968; FMNH PR 247). The width 247 and relatively thinner than in AMNH 2521. of the cavum cranii at the level of the cerebral What appears to be the foramen alveolare superi- expansions is somewhat greater in PU 12951 us lies in the area between the fossa nasalis and than in FMNH PR 247 even though the skulls are the fossa orbitalis and can be seen on the left side about the same size. The cavum cranii, then, is of the specimen. A foramen supramaxillare was very similar in the two New Jersey specimens not identified by Gaffney and Zangerl (1968) in (AMNH 2521 and PU 12951), which differ from Bothremys, and the endocast supports this. Al- the Selma specimen (FMNH PR 247). This fact brecht (1976) has reported that the foramen seems to refute the contention (Gaffney and supramaxillare is absent in Pelusios and Pelome- Zangerl, 1968, p. 228) that the differences in the dusa but present in Podocnemis. The presence of endocranium of B. cooki and B. barbedi are due this foramen in nearly all other turtles leads me solely to relative skull size. Other comparisons to conclude that its absence is derived and is between PU 12951 and FMNH PR 247 are consistent with monophyly of Bothremys, Pelo- difficult to make because of the poor preserva- medusa, and Pelusios. Further comparative work, tion and fragmentary nature of the latter speci- however, will be required. There is a foramen at men. A subjective comparison of the orbital the anterolateral margin of the fossa orbitalis cavities (fossa orbitalis) in AMNH 2521 and PU that occurs in Podocnemis, Pelomedusa, and Pelu- 12951 seems to indicate that the cavities are sios as well as Bothremys. Albrecht (1976) iden- relatively smaller in the former, which is the tified this in the living pelomedusids as the smaller skull. Precise measurements were not foramen alveolare superius. However, the fora- attempted and the fragile natture of the American men alveolare superius usually opens inside the Museum's Bothremys skull precludes making a fossa nasalis and I have found such a foramen in complete internal cast for comparison. the living pelomedusids and in Bothremys. In The anteroventral and portions of the lateral Podocnemis and Pelomedusa this foramen com- 4 AMERICAN MUSEUM NOVITATES NO. 2639

i M .111 .1

FIG. 1. Bothremys sp. (PU 12951, 93 mm. maximum length), Late Cretaceous, Magothy Forma- tion, New Jersey. Dorsal (left), ventral (right), and left lateral (lower) views of natural endocast. See figure 2. 1977 GAFFNEY: BOTHREMYS 5

foramen rbit foramen ralveolare superius praepalatinum pm~~. ~~unnamed' foramen fornau foramen orbit cartilaginous"ridrmx

p~~~~~~~~~~~~~~~~~~~~~~a

pforamen fosa n s sfae olforamnctoulfactusforiumnervic_ aneru.ant anauvidian foramnum procepalausrstrum\polatriumu,-e ' inferiorparietalisbaspeode expansions. crista pterygoidea suicus

foramen anterius I canalis carotici interni foramen nervi abducentis foram ~~~~~~~~~foramen nervi facialias- foramen nervi foaaa acustico-facialis b rgmn cartilaginous "rider" hiatua acuaticus bo dosmosmelaela

foramen magnum foramen magnyiumi orbit cartilaginous "rider" fr foramen nervi trigemini

I po pa ~~~~~~~~~~~~~~~~hiatusacusticus ~~ -I- t~~~~oramen magnum "Unnamed" toramen pau

foramen foramen foramen cavernosum praepalatinum palatinum

FIG. 2. Key to figure 1. Irregular stipple is broken matrix, whereas the reticulate pattern shows areas still retaining bone. 6 AMERICAN MUSEUM NOVITATES NO. 2639

FIG. 3. Bothremys sp. (PU 12951), Late Cretaceous, Magothy Formation, New Jersey. Latex cast of basicranium. See figure 2 (right) for identification of structures. municates with the canalis alveolaris superior goidei (laterally). Although the region is present (Albrecht, 1976, but again he identified it as the in cryptodires, the development of a lateral bony foramen alveolare superius). If Albrecht has, as I margin is a distinctive pleurodire feature. In suggest here, missed the "true" foramen alveolare living pleurodires portions of the M. pterygoideus superius, then this foramen is unnamed. In any occupy this region (Schumacher, 1973). Some case, it is not the foramen supramaxillare because of the bone making up the processus inferior both the foramen supramaxillare and the "un- parietalis persist and are particularly visible as named" foramen are present in Podocnemis. The spicules of bone in the ventral view (fig. 2). "unnamed" foramen is double in many Podo- Posteriorly, this portion of the endocast is con- cnemis specimens and seems to be double in fluent with the matrix extending out of the Bothremys. In both taxa it is distinct and promi- foramen nervi trigemini. The foramen palatinum nent in contrast to Pelusios in which it is quite posterius lies behind the posterior orbital wall as small. in other pleurodires and is prominent in the Posteromedially, the fossa orbitalis is con- endocast. fluent with a block of matrix that represents the The dorsal portion of the fossa nasalis is space between the processus inferior parietalis buried in matrix and the roof of most of the (medially) and the processus trochlearis ptery- cavum cranii is not preserved. The outline of the 1 977 GAFFNEY: BOTHREMYS 7 sulcus olfactorius and the cerebral expansions is a synapomorphy between Bothremys and can be seen, however. The cerebral expansions Podocnemis, as an elongated rostrum does not are relatively well developed, as in AMNH 2521, occur in chelids or living pelomedusids. in contrast to FMNH PR 247. The cartilaginous As stated by Gaffney and Zangerl (1968), extension of the supraoccipital is partly pre- Bothremys lacks the hypertrophied canalis ca- served and the cavum cranii roof is preserved roticus internus ofPodocnemis and the Princeton posterior to the extension or "rider." The general endocast substantiates this statement. The fora- position of the foramen magnum is apparent but men anterius canalis carotici interni open into its margins are not preserved nor are indications the sella turcica beneath the dorsum sellae over- of the exoccipitals or condyle. hang, and may be seen in the endocast figures The lateral walls of the cavum cranii show the (fig. 1) more clearly than in the latex peel. A positions of the foramen nervi trigemini, foramen paired structure, the processus clinoideus, usually cavernosum, and (on the right side) the fossa occurs on each side of the dorsum sellae but each acustico facialis and hiatus acusticus. Ventrally, one appears to be broken off in the Princeton most of the features of the basicranium are specimen. The broken base of each processus preserved. A latex peel (fig. 3) was made of this clinoideus reveals a groove, the foramen nervi area to facilitate comparison with living pelo- abducentis. medusids (figs. 4-7) and the other specimens of Albrecht (1976) and Gaffney (1975) have Bothremys (Gaffney and Zangerl, 1968, fig. 19). pointed out the utility of basicranial structures in The following description is primarily of the elucidating pleurodire phylogeny. I hope that latex peel, for ease of comparison with Recent this description of the basicranium in one of the specimens, and with references to the natural oldest known pleurodires will eventually aid endocast in parentheses. The most lateral troughs phylogenetic work on this group. are the spaces defined by the processus trochlearis pterygoidei laterally, and the processus RECONSTRUCTION THE PALATE inferior parietalis - crista pterygoidea medially. OF The wall formed by the processus inferior pari- IN BOTHREMYS etalis and crista pterygoidei is relatively much thicker in Bothremys than in any living pelo- The type skull of Bothremys cooki (AMNH medusid. Seemingly, this is related to the appar- 2521) was figured in Gaffney and Zangerl (1968) ent crushing adaptations seen in the palate. but this specimen is lacking many of the poste- Medial to this wall is a trough (ridge), the sulcus rior regions of the skull. Another specimen cavernosus. Some small foramina are visible in (FMNH PR 247) identified as Bothremys barberi the sulcus but I think they are only nutritive in by Gaffney and Zangerl has the otic regions nature. Posteriorly, the sulcus is confluent with preserved. A reconstruction (fig. 8) of Bothremys the foramen nervi trigemini rather than separated is presented here using information from both from it by a ridge as in living pelomedusids. The skulls. The AMNH skull is indicated in the figure endocast stops roughly at the position of the by the shaded area, whereas the outline and foramen cavernosum. The rostrum basisphenoi- sutures of the FMNH skull are added. The FMNH dale in Bothremys is long and narrow as in skull, besides being identified as a separate Podocnemis rather than short and blunt as in species, is also about twice the size of the AMNH Pelusios. The outline of the sella turcica of specimen, so the reconstruction must be con- Bothremys also resembles Podocnemis more than sidered with these qualifications. it does Pelusios. At present I cannot hypothesize The identification of bones and structures can the significance of these similarities but it ap- be readily obtained from the figures in Gaffney pears that the rostrum basisphenoidale character and Zangerl (1968) and are not repeated here. 8 AMERICAN MUSEUM NOVITATES NO. 2639

FIG. 4. Pelusios sp. (AMNH[H] 10062, 52 mm. in length), Recent, Zaire. Dorsal view of horizontally sectioned skull. Hatching indicates cut surfaces. 1 977 GAFFNEY: BOTHREMYS 9

foramen alveolare superius .-hlearis mx -,pptf 1 foramen orbito-nasale rygoidei rostrum 3S Sbasisphenoidale inferior palp ~~~~processusparietalis Po < , g 4 foramen palatinum \/, posterius foramen nervi vidiani pt sellaturcica processus clinoideus nervi trigemini pr foramen anterius bs I canalis carotici interni foramen nervi abducentis bot cavum labyrinthicum semicircularis qu 6 ~~~~~~~~canalis\foramen horizontalis foramen internum nervi glossopharyngei foramen jugulare anterius foramen nervi hypoglossi antrum postoticum

FIG. 5. Key to figure 4. 10 AMERICAN MUSEUM NOVITATES NO. 2639

FIG. 6. Podocnemis expansa (AMNH[H] 97124, 126 mm. in length), Recent, Bolivia. Dorsal view of horizontally sectioned skull. Hatching indicates cut surfaces. 1977 GAFFNEY: BOTHREMYS 11

foramen praepalatinum

foramen alveolare superius (?) pm | t < unnamed foramen foramen orbito-nasale t1'\g foramen supramaxillare

rostrum basisphenoidale foramen palatinum posterius foramen anterius canalis nervi vidiani crista pterygoidea sella turcica foramen caroticum laterale foramen anterius canalis carotici interni canalis cavernosus bs foramen nervi abducentis bo foramen nervi facialis foramen stapedio-temporale Li.\ processus \>interfenestralis

foramen nervi hypoglossi

FIG. 7. Key to figure 6. .&P I..ki

FIG. 8. Bothremys cooki (AMNH 2521, skull length as preserved 69 mm.), Late Cretaceous (?), Hornerstown Formation, New Jersey. Palatal views with posterior restored from Bothremys barberi (FMNH PR 247), Late Cretaceous, Selma Formation, Alabama. Limits of processus trochlearis pterygoidei hypothetical.

LITERATURE CITED Albrecht, P. W. Gaffney, Eugene S., and Rainer Zangerl 1976. The cranial arteries of turtles and their 1968. A revision of the chelonian genus Both- evolutionary significance. Jour. Morph., remys (Pleurodira: Pelomedusidae). vol. 149, no. 2, pp. 159-182. Fieldiana:Geol., vol. 16, pp. 193-239. Edinger, Tilly Schumacher, G.-H. 1929. Die fossilen Gehirne. Ergeb. Anat. und 1973. The head muscles and hyolaryngeal Entwicklungsgesch., Band 28. Berlin, skeleton of turtles and crocodilians. In Julius Springer Verlag, 249 pp. Gans, C., and T. S. Parsons (eds.), 1934. Anton Fritsch's "Grosshirn von Po- Biology of the Reptilia. London, New lyptychodon" ist der Steinkern eines York, Academic Press, vol. 4, pp. Schildkr6tenschadels. Psych. Neurol. 101-199. Bladen, nos. 3-4, pp. 396404. Zangerl, Rainer Gaffney, Eugene S. 1960. The vertebrate fauna of the Selma For- 1975. A phylogeny and classification of the mation of Alabama. V. An advanced higher category of turtles. Bull. Amer. cheloniid sea turtle. Fieldiana:Geol. Mus. Nat. Hist., vol. 155, art. 5, pp. Mem., vol. 3, no. 5, pp. 281-312. 387-436. 12 ff 0> 3-s'0<>*,+4-SD xW ww rriwp {winwinw"Da AMERICAN MUSEUM Novitates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 2639, pp. 1-12, figs. 1-8 November 3, 1977

An Endocranial Cast of the Side-Necked Turtle, Bothremys, With a New Reconstruction of the Palate

EUGENE S. GAFFNEY'

'Associate Curator, Department of Vertebrate Paleontology, the American Museum of Natural History; Adjunct Assistant Professor, Department of Geological Sciences, Columbia University.