PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, NY 10024 Number 3321, 16 pp., 4 figures February 27, 2001 Kurmademys, a New Side-Necked Turtle (Pelomedusoides: Bothremydidae) from the Late Cretaceous of India EUGENE S. GAFFNEY, 1 SANKAR CHATTERJEE, 2 AND DHIRAJ K. RUDRA 3 ABSTRACT The Maastrichtian Kallamedu Formation of southern India near the village of Kallamedu, Tamil Nadu, has yielded skulls and postcrania of a new genus of side-necked turtle. Kurma- demys kallamedensis, new genus and species, is based primarily on a single well-preserved skull. Kurmademys is a pelomedusoid pleurodire belonging to the family Bothremydidae Baur, 1891, with these bothremydid characters: (1) exoccipital-quadrate contact, (2) incisura colu- mellae auris closed by bone, and (3) eustachian tube and stapes separated by bone. Kurma- demys is unique among known bothremydids in having extensive temporal emargination, a small postorbital, a large precollumellar fossa, and a foramen posterius canalis carotici interni formed completely by the basisphenoid. 1 Curator, Division of Paleontology, American Museum of Natural History. 2 Paul Whitfield Horn Professor of Geology and Curator of Paleontology, Museum of Texas Tech University, Lubbock, Texas. 3 Former Professor and Head, Geology Unit, Indian Statistical Institute, Calcutta, India. INTRODUCTION Pleurodires, or side-necked turtles, are not part of the recent fauna of the Indian subcon- tinent, but they are known as fossils from the late Cretaceous into the Neogene. Their record in India, however, is very poor. Summaries of this record are in Wood (1970), de Broin (1987, 1988), and Jain (1986). Good skulls of the podocnemidid Shweboemys are known from the Neogene of Burma and Pakistan (see Wood, 1970). A skull of a pleurodire, Shwe- boemys pisdurensis, has been described by Jain (1977, 1986) and an unnamed bothremydid has been mentioned in Singh et al. (1998). Shweboemys pisdurensis appears to be a broad-jawed podocnemidid related to Shwe- boemys and Stereogenys (Wood, 1970). The bothremydid skull (SD S/VPL 1125) is pres-
Transcript
P U B L I S H E D B Y T H E A M E R I C A N M U S E U M O F N AT U R A L H I S T O RY
CENTRAL PARK WEST AT 79TH STREET, NEW YORK, NY 10024
Number 3321, 16 pp., 4 figures February 27, 2001
Kurmademys, a New Side-Necked Turtle(Pelomedusoides: Bothremydidae) from the Late
Cretaceous of India
EUGENE S. GAFFNEY,1 SANKAR CHATTERJEE,2 AND DHIRAJ K. RUDRA3
ABSTRACT
The Maastrichtian Kallamedu Formation of southern India near the village of Kallamedu,Tamil Nadu, has yielded skulls and postcrania of a new genus of side-necked turtle. Kurma-demys kallamedensis, new genus and species, is based primarily on a single well-preservedskull. Kurmademys is a pelomedusoid pleurodire belonging to the family Bothremydidae Baur,1891, with these bothremydid characters: (1) exoccipital-quadrate contact, (2) incisura colu-mellae auris closed by bone, and (3) eustachian tube and stapes separated by bone. Kurma-demys is unique among known bothremydids in having extensive temporal emargination, asmall postorbital, a large precollumellar fossa, and a foramen posterius canalis carotici interniformed completely by the basisphenoid.
1 Curator, Division of Paleontology, American Museum of Natural History.2 Paul Whitfield Horn Professor of Geology and Curator of Paleontology, Museum of Texas Tech University,
Lubbock, Texas.3 Former Professor and Head, Geology Unit, Indian Statistical Institute, Calcutta, India.
INTRODUCTION
Pleurodires, or side-necked turtles, are notpart of the recent fauna of the Indian subcon-tinent, but they are known as fossils from thelate Cretaceous into the Neogene. Their recordin India, however, is very poor. Summaries ofthis record are in Wood (1970), de Broin(1987, 1988), and Jain (1986). Good skulls ofthe podocnemidid Shweboemys are known
from the Neogene of Burma and Pakistan (seeWood, 1970). A skull of a pleurodire, Shwe-boemys pisdurensis, has been described by Jain(1977, 1986) and an unnamed bothremydidhas been mentioned in Singh et al. (1998).Shweboemys pisdurensis appears to be abroad-jawed podocnemidid related to Shwe-boemys and Stereogenys (Wood, 1970). Thebothremydid skull (SD S/VPL 1125) is pres-
2 NO. 3321AMERICAN MUSEUM NOVITATES
ently being studied by Gaffney, Sahni, Singh,and Schleich. This skull is a bothremydid, butdistinct from Kurmademys. The shell of pelo-medusoids is notoriously conservative and in-adequate to distinguish bothremydids from po-docnemidids even when completely known.Thus, the Indian pleurodire record rests on thetwo skulls indicated above and the new ma-terial from Kallamedu.
Shweboemys pisdurensis consists of askull, nearly complete shell material, andsome postcranial elements from the Maas-trichtian Lameta Formation at Pisdura andDongargaon, in Maharastra State, central In-dia. We have not seen this material, but theskull does appear to be similar to the de-scribed Shweboemys of Wood (1970). Theunnamed bothremydid skull of Singh et al.(1998) is from the Maastrichtian intertrap-pean Green Tuff bed of Amboli Quarry, Jo-geshwari, Bombay. Features of this specimenare noted in the comparisons below.
Although the Bothremydidae was namedas early as 1891 by George Baur, the termfell into disuse for most of this century andthe few included taxa, particularly Bothremysand Taphrosphys, were simply included inthe Pelomedusidae. Antunes and Broin(1988) and Broin (1988) revived Bothremy-didae, provided a new diagnosis, and addedtaxa, such as Rosasia, based on skulls andshells. Recent papers on fossil pleurodires,such as Meylan (1996), Broin de Lapparentand Werner (1998), and Tong et al. (1998),use the Antunes and Broin (1988) terminol-ogy, in which Bothremydidae, Podocnemi-didae, and Pelomedusidae (restricted to Pe-lusios and Pelomedusa) are contained in thePelomedusoides (which equals Pelomedusi-dae in the classic sense). Bothremydids arenow recognized as a more widespread anddiverse group than previously considered.
Useful reviews of the literature on bothre-mydids can be found in Broin (1988) and An-tunes and Broin (1988). Previously describedbothremydid skulls are as follows: Bothremys(Gaffney and Zangerl, 1968; Gaffney, 1977),Taphrosphys (Gaffney, 1975), Rosasia (Antu-nes and Broin, 1988), Foxemys (Tong et al.,1998), Zolhafah (Lapparent de Broin and Wer-ner, 1998), Arenila (Lapparent de Broin andWerner, 1998), and Nigeremys (Bergouniouxand Crouzel, 1968; Lapparent de Broin and
Werner, 1998). Other skulls of pelomedusoidsare Araripemys (Meylan, 1996), and an un-named Santana genus (FR 4922) (Gaffney andMeylan, 1991). A general treatment and de-scription of pleurodire skulls, turtle skull mor-phology and terminology, and a literature re-view are in Gaffney (1979).
This paper is intended to name and brieflydescribe this new taxon. More detailed com-parative descriptions and a systematic anal-ysis are part of a larger project on pleurodires(Gaffney, Meylan, and Wood, 1997; Gaffney,Tong, Chatterjee, Moody and Hirayama,1998).
We use Lapparent de Broin and Werner’s(1998) informal reference to a BothremysGroup and a Nigeremys Group (which in-cludes Taphrosphys). These continue to formmonophyletic taxa in most of our currentanalyses. Contents of these groups are as fol-lows: Bothremys Group—Bothremys, Rosa-sia, Zolhafah, Foxemys; Nigeremys Group—Nigeremys, Arenila, Taphrosphys.
Institutional Abbreviations
ISI Indian Statistical Institute, Calcutta, In-dia
Fig. 1. Type locality of Kurmademys kallamedensis, ISI R152, near the village of Kallamedu, TamilNadu, southern India. The turtle material occurs here in a 6 in. thick pocket of fine grained sandstoneand clay in the upper Maastrichtian Kallamedu Formation.
SYSTEMATICS
ORDER TESTUDINES LINNAEUS, 1758
MEGAORDER PLEURODIRA COPE, 1864 (FIDEGAFFNEY AND MEYLAN, 1988)
HYPERFAMILY PELOMEDUSOIDES COPE, 1868
FAMILY BOTHREMYDIDAE BAUR, 1891
Kurmademys, new genus
TYPE SPECIES: Kurmademys kallamedensis,new genus and new species.
DISTRIBUTION: Maastrichtian of central In-dia.
ETYMOLOGY: Kurma, ‘‘turtle’’ in Sanskrit,in allusion to the second-stage incarnation ofLord Vishnu as a turtle in Hindu mythology.
DIAGNOSIS
A genus of bothremydid pleurodire withtriangular skull, orbits dorsolaterally placed,not dorsally as in Bothremys; extensive tem-poral and cheek emargination similar to Pe-lusios and Pelomedusa, but unique amongbothremydids; jugal enters orbit; postorbital
short in contrast to Foxemys and Bothremys;triturating surfaces triangular in contrast tonarrow in Nigeremys Group, but not greatlyexpanded as in Foxemys, Rosasia, Zolhafah,and Bothremys; no pit in triturating surfacein contrast to Bothremys and Rosasia; max-illa-quadrate contact absent; palatine exten-sively exposed on triturating surface as in allBothremys Group; antrum postoticum largein contrast to small in Bothremys; eustachiantube separated from stapes as in all otherbothremydids; incisura columellae aurisclosed as in Bothremys and Taphrosphys;precolumellar fossa large as in pelomedusids,but unique among bothremydids; pterygoi-deus depression moderate as in Rosasia, notdeep as in Foxemys; foramen posterius ca-nalis carotici interni completely enclosed inbasisphenoid unique among pelomedusoids;supraoccipital-quadrate contact present as inall Bothremys Group; foramen stapedio-tem-porale visible in dorsal view in contrast to allother bothremydids.
4 NO. 3321AMERICAN MUSEUM NOVITATES
Fig. 2. Kurmademys kallamedensis, n. gen. & sp., ISI R152. A, dorsal; B, ventral; C, right lateral;D, posterior; E, left lateral; F, anterior.
Kurmademys kallamedensis, new species
TYPE SPECIMEN: ISI R152, a nearly com-plete skull lacking the dorsal part of the pre-frontals, the posterior part of the crista su-
praoccipitalis, and part of the left quadrato-jugal.
TYPE LOCALITY: Near the village of Kal-lamedu, Tamil Nadu, southern India. Map oflocality is in Sastry et al. (1972).
2001 5GAFFNEY ET AL.: KURMADEMYS
Fig. 3. Key for fig. 2.
HORIZON: Kallamedu Formation of theAriyalur Group. Formation named and de-scribed by Sastry et al. (1972), who corre-lated it with the uppermost Maastrichtian; theCretaceous-Tertiary boundary is its upperlimit (Sastry et al.: 6). Dinosaurs have been
described from other exposures of the Kal-lamedu (Matley, 1929; Yadagiri and Ayyas-ami, 1987). The Kurmademys locality is asmall pocket of fine-grained sandstone andclay, about 6 in. thick. It also contained croc-odiles, gar scales, and freshwater gastropods
6 NO. 3321AMERICAN MUSEUM NOVITATES
Fig. 4. Kurmademys kallamedensis, n. gen. &sp. Partially restored ventral view. See fig. 3 forbone identifications.
and bivalves and is interpreted as a fresh-water pond deposit.
DIAGNOSIS: As for genus.ETYMOLOGY: For the Kallamedu Forma-
The Kurmademys kallamedensis typeskull, ISI R152 has a premaxilla-condyle me-dian length of 47.2 mm, a maximum widthof 49.4 mm, and a height from condylusmandibularis of the quadrate to the top of theskull roof of 28.2 mm.
The order of elements and the order oftopics within elements follows Gaffney(1979). More general features of the pleuro-dire skull and explanations of terminologycan also be found in Gaffney (1979) and arenot repeated here.
Although at present six skulls of Kurma-demys are known, the type specimen, ISIR152, forms the basis of this description,with a few contributions from the other spec-imens. The type skull is the best preservedand nearly complete. The other five skullsare not as well preserved and are not yet ful-ly prepared.
PREFRONTAL
Almost all of both prefrontals are missingin the type skull, ISI R152. The only frag-ments of prefrontal remaining are the ven-tralmost tips of right and left prefrontals ly-ing on the medial side of the maxillae in theanterior orbital walls. There is more of theright prefrontal than the left. The dorsal areaof the prefrontals has been restored arbitrari-ly on ISI R152. However, ISI R158 has pre-frontals preserved.
The shape and extent of prefrontal in Kur-mademys is consistent with that seen in mostother pelomedusoids, such as FR 4922. Theymeet for their entire length on the midline.There are no nasals.
FRONTAL
Both frontals are nearly complete in ISIR152, but their anteriormost margins are
breaks rather than sutures, so the contactwith the prefrontals is not determinable.However, the extent preserved is comparableto the frontal as seen in Pelomedusa and Pe-lusios, so it is not likely that much frontal ismissing.
As in other pelomedusoids, the frontal ofKurmademys contacts the postorbital later-ally and the parietal posteriorly. Its contactsand relative size are very similar to Pelusiosand Pelomedusa. On the ventral surface thefrontal of Kurmademys has a parasagittal
2001 7GAFFNEY ET AL.: KURMADEMYS
ridge separating fossa orbitalis from the sul-cus olfactorius. The sulcus is slightly narrow-er anteriorly than in Pelusios, but widensposteriorly as in Pelusios and most pelome-dusoids. The processus inferior parietalismeets the posterior edge of the frontal ridgedorsally as in FR 4922, without a ventralprocess as in Pelusios.
In other pelomedusoids, such as Baurue-mys, FR 4922, and Pelusios, the frontal isthickened lateral to the sulcus olfactoriusridge and forms a wall for the fossa orbitalisanteriorly and a dorsal margin of the ptery-go-palatine channel posteriorly. This wall isusually continuous with the posterior orbitalwall laterally and ventrally. However, in Kur-mademys the frontal is relatively thin andthere is no distinct connection between thesulcus olfactorius ridge and the posterior or-bital wall. The pterygo-palatine channel isrelatively open dorsally, at the anterior end,in contrast to most other pelomedusoids,which have a more restricted ridge of bonehere.
PARIETAL
Both right and left parietals are completein ISI R152, except for small cracks andbreaks.
The dorsal plate of the parietal is not ex-tensive in Kurmademys. Anteriorly there is atransverse contact with the frontal, and lat-erally a contact with the postorbital. There isno contact between the parietal and quadra-tojugal; the postorbital is widely exposedalong the edge of the temporal emargination.The temporal part of the parietal is about asextensive as in the living Pelomedusa, whichis slightly more emarginate than most Pelusios.
The ventral wall of the parietal, the pro-cessus inferior parietalis, can be seen in bothsides of ISI R152. The wall is similar to thatin other pelomedusoids. The anterior marginof the wall is formed completely by the pa-rietal and curves ventrally without the pos-terior indentation seen in some Pelusios.Ventrally the parietal contacts the pterygoid,which is low until it reaches the area anteriorto the foramen nervi trigemini, where thepterygoid/parietal contact is more dorsal. Theforamen nervi trigemini in Kurmademys isformed by the parietal, pterygoid, and pro-
otic as in most other pelomedusoids. Poste-riorly the parietal contacts the supraoccipitaland prootic as in other pleurodires. The pa-rietal of Kurmademys also has a short ventralprocess below its lateral margin that contactsthe dorsal process of the palatine in the lat-eral wall of the sulcus palatinopterygoideus(Antunes and Broin, 1988).
JUGAL
The jugal in ISI R152 is preserved on bothsides, but surrounding contacts are seen onlyon the right side.
The jugal is relatively small in Kurmade-mys, similar to Pelusios and Pelomedusa,and much smaller than in podocnemidids.The jugal is exposed in the posterior marginof the orbit, but its exposure is reduced by aposterodorsal process of the maxilla. Ven-trally the jugal contacts the maxilla, and pos-teriorly the quadratojugal. The jugal is notexposed on the dorsal edge of the cheekemargination due to a contact of the quad-ratojugal and maxilla. This is unusual be-cause Kurmademys has a cheek emarginationat least as extensive as in Pelomedusa andPelusios, and they have a large exposure ofthe jugal along this emargination.
The medial process of the jugal in ISIR152 is preserved and visible on both sides.It is basically similar to that area describedby Gaffney (1979: figs. 53, 130) for Pelusios.The medial process of Kurmademys contactsthe palatine medially and the maxilla ante-riorly. The jugal forms part of the anteriorwall of the adductor muscle chamber as inmost pelomedusoids. The jugal contacts thepostorbital medially and the pterygoid ven-tromedially. Ventrally the jugal contacts themaxilla. All of these contacts are quite sim-ilar to Pelusios.
QUADRATOJUGAL
The quadratojugal in ISI R152 is nearlycomplete on the right side. Some small partsof the dorsal and ventral margin are probablymissing, but based on the tapering of thebone and on surrounding bones, very little ofthe quadratojugal is missing. On the left side,however, only the anterior part and a smallfragment of the posterodorsal contact withthe quadrate are preserved.
8 NO. 3321AMERICAN MUSEUM NOVITATES
The quadratojugal of Kurmademys is un-usually small for pelomedusoids, because thetemporal and cheek emarginations are bothextensive. Kurmademys is unique amongbothremydids in having such a small quad-ratojugal and narrow temporal arch. Thequadratojugal contacts the quadrate posteri-orly, the postorbital anterodorsally, the jugalanteriorly, and the maxilla anteroventrally.The dorsal margin of the quadratojugal formsthe lateral edge of the temporal emarginationand the ventral margin forms the dorsal edgeof the cheek emargination.
SQUAMOSAL
Both right and left squamosals are presentin ISI R152 and both have some damage totheir posterolateral margins, but are other-wise complete.
As in most turtles, the squamosal of Kur-mademys is cone shaped, forming the pos-terolateral portion of the antrum postoticum.As in other turtles, the squamosal of Kur-mademys fits onto the circular posterior endof the quadrate and contacts the opisthoticmedially. Kurmademys has an extensive tem-poral emargination and the squamosal has nocontact with parietal or postorbital. The rightside of ISI R152 also shows that, althoughthe squamosal has a narrow anterior processlying on the quadrate, the process does notreach the quadratojugal. In Pelusios and Pe-lomedusa there is usually a contact betweenthe quadratojugal and squamosal, althoughthere is some individual variation with somespecimens having a very slight contact or nocontact. However, none of the specimens ofPelusios or Pelomedusa available to us showthe degree of reduction of quadratojugal andsquamosal nor the extent of dorsal exposureof the quadrate seen in Kurmademys. Thisfeature is unique to Kurmademys. The dor-solateral surface of the squamosal in Kur-mademys is rounded, with no parasagittalridge or wall seen in other Pelomedusoides.This ridge is a function of the degree of tem-poral emargination and is present to a vary-ing extent in all other Pelomedusoides.
The antrum postoticum is preserved onboth sides of ISI R152, and its internal extentis visible. The antrum is larger in Kurma-demys than in any other Pelomedusoides. Pe-
lusios, Pelomedusa, and FR 4992 have alarge antrum postoticum, but the antrum ofKurmademys is even larger. The size in FR4992, Pelusios, and Pelomedusa is interpret-ed as the primitive condition for Pelomedu-soides, because this is the condition in chel-ids. The Kurmademys condition is tentativelyinterpreted as a unique autapomorphy of thisgenus, because our current analyses showKurmademys deep within the Bothremydidae.
POSTORBITAL
The postorbital is preserved on both sidesof ISI R152. The posterior edge of the leftpostorbital is missing bone when comparedwith the more complete right postorbital.
The size and relations of the postorbital inKurmademys are very similar to the postor-bital in Pelusios and Pelomedusa. The post-orbital of Kurmademys lies between the orbitanteriorly and the temporal margin posteri-orly and forms part of the margins of thoseopenings. Medially the postorbital contactsthe frontal anteriorly and the parietal poste-riorly. Laterally the postorbital contacts thejugal anteriorly and the quadratojugal pos-teriorly. All of these contacts are as in Pe-lusios and Pelomedusa. FR 4922 differs inhaving a broad contact of the parietal andquadratojugal posterior to the postorbital.
The ventral process of the postorbital isalso similar to that in Pelusios and Pelome-dusa. As exposed in the anterior wall of thetemporal fossa, the postorbital contacts theparietal medially, the pterygoid ventrally,and the jugal laterally. On the right side,which is better preserved than the left side,there is a small contact between the parietaland pterygoid, preventing exposure of thepostorbital in the pterygo-palatine channel atthis point. In the posterior wall of the fossaorbitalis, the ventral process of the postor-bital contacts the dorsal process of the pala-tine medially in a sloping suture. Laterallythe postorbital contacts the jugal. The medialsurface of the ventral process of the postor-bital forms the lateral wall and the lateral partof the roof of a relatively short pterygo-pal-atine channel.
PREMAXILLA
Both premaxillae are present in ISI R152and are nearly complete.
2001 9GAFFNEY ET AL.: KURMADEMYS
Laterally the premaxilla contacts the max-illa in a parasagittal suture, and it contactsthe other premaxilla medially. The posteriormargin of the premaxilla forms at least partof the apertura narium interna, but has a bro-ken edge medially. The broken edge does notshow a sutural surface anywhere and there isno fragment of a vomer, but it is possible thatone was present. The premaxilla in FR 4922has a process of the maxilla lying behind it,but this is absent in ISI R152. The dorsalsurface of the premaxilla forms part of thefloor of the fossa nasalis. In Kurmademys thepremaxillae curve dorsally to form a sharplyrising median ridge in the fossa. This medianridge is present in other pelomedusoids, butit is lower and smaller than in Kurmademys.
The ventral surface of the premaxilla bearsthe continuation of the labial ridge. The la-bial ridge of Kurmademys is narrower thanin Pelusios and lacks the anterior projectionof the margin of the apertura narium externain that form. This area between the aperturanarium externa and labial ridge is very thinin Kurmademys, similar to FR 4922. Kur-mademys has a shallow median notch similarto Pelusios and wider than in FR 4922. Incontrast to Pelusios and Pelomedusa, Kur-mademys has a posterior extension to the pre-maxilla that bears a distinct, ventrally facingconcavity on the midline that is the ventralsurface of the dorsal ridge in the fossa na-salis. Lower jaws of Kurmademys show amarked symphyseal hook. The flat part of thetriturating surface narrows considerably fromthe maxilla to form a narrow shelf betweenthe concavity and the labial ridge. This mor-phology is also seen in Bothremys and Ro-sasia as well as Neochelys and some gener-alized cryptodires like baenids. There is nosign of a foramen praepalatinum.
MAXILLA
Both maxillae of ISI R152 are completeexcept for the distal ends of the dorsal pro-cesses, which are missing.
The vertical or alveolar plate of the max-illa is deep and fairly massive, not narrow asin FR 4922, but similar to Podocnemis. Themaxilla forms the apertura narium externaanteriorly and it is wider at its base than inFR 4922, similar to most other pelomedu-
soids. The apertura is not produced anteriorlyas in Pelusios and Podocnemis. The dorsalprocess of the maxilla in Kurmademys isthinner than in Pelusios and Pelomedusa andis similar to FR 4922. The ventral margin ofthe orbit, is formed by the maxilla, which hasa dorsal process along the posterior marginof the orbit reducing the contribution madeby the jugal to the orbit. Forms such as FR4922 and Podocnemis have a posterodorsalprocess, but it is separated from the orbit bythe jugal. Pelusios and Pelomedusa do nothave this process, but its expression is vari-able in pelomedusoids. The posterior edge ofthe maxilla forms the margin of the cheekemargination. Between the cheek emargina-tion and the jugal, the maxilla contacts thequadratojugal.
The horizontal plate of the maxilla is ex-posed in the orbital floor where the maxillacontacts the palatine medially and forms asmall part of the border of the large foramenorbito-nasale. Posteriorly the maxilla con-tacts the jugal.
The triturating surfaces of Kurmademysare narrow anteriorly and widely expandedposteriorly. The maxilla itself, however, ta-pers posteriorly, so that the palatine formsthe posterior and medial portion of the trit-urating surface. The triturating morphologyis similar to that in Foxemys, which is alsonarrow anteriorly and expanded posteriorly,with a significant contribution from the pal-atine. In both skulls the narrow, anterior parthas a raised medial edge along the lingualridge. They differ in that Foxemys has aslight pinching of the snout anteriorly, as inRosasia; this is absent in Kurmademys,where the snout is straight. Foxemys has twoaccessory ridges, absent in Kurmademys.
The triturating surface in Kurmademys israised anteriorly along the margin of the ap-ertura narium interna. The posterior expand-ed area is slightly concave. There are no ac-cessory ridges on the triturating surface,which ends posteriorly in a V-shaped margincompletely formed by the palatine.
VOMER
There is no vomer present nor are theresutural surfaces remaining for a vomer. How-ever, the bone edges in this area are not en-
10 NO. 3321AMERICAN MUSEUM NOVITATES
tirely complete, and the morphology sur-rounding the apertura narium interna is veryclose to that in Foxemys, which has a well-developed vomer. Thus it is quite possiblethat one was present in Kurmademys.
PALATINE
Both palatine bones are present in ISIR152, but are missing some of the anterioredges that form the margin of the aperturanarium interna and the possible vomerinecontact.
The anterolateral part of the palatine inKurmademys contacts the maxilla and formsthe posteromedial part of the triturating sur-face. The triturating surface is a low platformthat ends in a V-shaped margin completelyformed by the palatine. Foxemys is similar toKurmademys in this area. The palatine formsthe posterior margin of the apertura nariuminterna, but most of this margin is missing inISI R152. The choanal grooves are barelydiscernible in Kurmademys; they are betterdefined in Foxemys. The foramen palatinumposterius is formed in the palatine-pterygoidsuture by both bones (fig. 4), as in Foxemysand in contrast to FR 4922, Pelusios, Pelo-medusa, and Podocnemis, in which most ofthe foramen is in the palatine. As in otherpleurodires, there is a median contact withthe other palatine and a transverse, posteriorcontact with the pterygoid. On the dorsal sur-face both right and left palatines are visibleand free of matrix. The palatine forms themedial part of the orbital floor and the lateralmargin of the large foramen orbito-nasale.
Posteriorly the palatine of Kurmademyshas a large dorsal process forming the lateralwall of the pterygo-palatine channel. Thisprocess contacts the jugal laterally, the post-orbital dorsolaterally, and the parietal dor-somedially. The process tapers dorsally, sothat its medial edge is higher than its lateraledge. The medial edge forms that lateralmargin of the adductor channel. This dorsalprocess of the palatine reaches the parietal inthe postorbital wall, which is quite unusualand has not been found so far in other pe-lomedusoids. The posterior wall of the orbitis complex in pelomedusoids and particularlyso in Kurmademys. Behind the dorsal processof the palatine is the ventromedial process of
the postorbital and a short ventral process ofthe parietal, all visible on the posterior sur-face of the postorbital wall.
QUADRATE
Most of both quadrates are complete andfree of matrix. Part of the medial area of theleft quadrate is broken and partially restoredwith something awful.
The squamosal lies at the posterolateralcorner of the quadrate and its relations andcontacts with the quadrate in Kurmademysare similar to those in Pelusios and Pelo-medusa. The quadrate exposure along the lat-eral edge of the temporal emargination pre-vents contact of squamosal and postorbital.Anteriorly the quadrate contacts the quadra-tojugal, but the contact is relatively small be-cause of the extensive temporal and cheekemargination. The quadratojugal contact issmaller in Kurmademys than in Pelusios, Pe-lomedusa, or FR 4922.
Most of the quadrate is involved in the for-mation of the cavum tympani and its twospaces, the antrum postoticum and the pre-collumellar fossa. The antrum postoticum ofKurmademys is unusually large for pelome-dusoids; it is as large as the antrum in Emy-dura, the presumed primitive condition forpelomedusoids. The antrum of Kurmademysis swollen to completely fill the area insidethe space formed by squamosal and quadrate.The precollumellar fossa is also deep andvery large in Kurmademys, and also com-parable in size to primitive chelids. However,other pelomedusoids, such as Pelusios andPelomedusa, also have a large precollumellarfossa. In many features, the cavum tympaniof Kurmademys is more primitive than inother bothremydids.
The other feature of interest in the cavumtympani is the incisura columellae auris,which still has the stapes present in the rightquadrate of ISI R152. The incisura is reducedto a completely closed, small foramen con-taining only the stapes in Kurmademys, incontrast to the open incisura of Foxemys. Ar-enila, Zolhafah, Bothremys, and Taphros-phys have the closed incisura as in Kurma-demys but in Araripemys, FR 4922, and chel-ids, it is open. The combination of a com-pletely closed incisura columellae auris with
2001 11GAFFNEY ET AL.: KURMADEMYS
a gigantic antrum postoticum is a combina-tion unknown so far in pleurodires. Thequadrate of Kurmademys has a kidney-shaped cavum tympani as in Bothremys. Al-though the incisura columellae auris is com-pletely closed by bone, behind it is a grooveformed in the quadrate for the eustachiantube. Dorsally the groove slopes up to a hor-izontal, straight-edged ridge that separatesthe eustachian tube surface from the openingfor the stapedial artery in the fenestra pos-totica.
The medial contact of the quadrate is withthe prootic anteriorly. As in nearly all otherturtles, the prootic and quadrate form the fo-ramen stapedio-temporale. The foramen sta-pedio- temporale is certainly placed more an-teriorly in Kurmademys than in Emydura, butnot any more anteriorly than in FR 4922 orPelusios and Pelomedusa. The canalis sta-pedio-temporalis can be followed posteriorlyon the right side to the aditus canalis stape-dio-temporalis, which is partially dividedfrom the rest of the fenestra postotica, show-ing the entry of the stapedial artery into theskull. A separate canal for the stapes itself ispresent from the incisura columellae auris tothe aditus canalis stapedio-temporalis as inBothremys and Taphrosphys.
Behind the prootic there is a contact withthe supraoccipital in ISI R152 (fig. 3A), thatintervenes between the usual opisthotic-pro-otic contact. This contact also occurs in allother Bothremys Group taxa. The opisthoticcontact appears to be reduced in favor of ex-pansion of the supraoccipital contact whencompared with a more generalized pleurodirelike Pelusios. The quadrate-opisthotic con-tact is anteromedial to posterolateral in Kur-mademys, between the supraoccipital andsquamosal.
On the ventral surface the quadrate in Kur-mademys contacts the pterygoid anterome-dially from the base of the condylus mandi-bularis along the anterolateral edge of theprocessus articularis, as in most pleurodires,such as Emydura and Pelusios. Kurmademyshas a medial process of the quadrate as inother pleurodires, that contacts a narrowlyexposed prootic, and broadly contacts the ba-sisphenoid. Behind the basisphenoid, thequadrate has a broad contact with the basi-occipital. Dorsal to that the quadrate contacts
the exoccipital. The basioccipital contact ofthe quadrate characterizes the Bothremydi-dae plus Podocnemididae, but the exoccipitalcontact is more restricted, found so far onlyin the Bothremydidae.
The medial contacts of the quadrate in-clude the opisthotic and prootic as in otherturtles. The roof of the fenestra postotica (ad-itus canalis stapedio-temporalis) has a lowparasagittal ridge showing the passage of themore lateral stapedial artery from the moremedial (and ventral) lateral head vein. Thisridge is largely formed by the quadrate, withthe opisthotic contributing medially. The fe-nestra postotica of Kurmademys is subdivid-ed by bony partitions, so that the lateral headvein and stapedial artery are separated fromthe more medial parts of the fenestra posto-tica. The opisthotic forms this wall dorsallyand the quadrate forms it ventrally. On bothsides of ISI R152 the contact of these twobones is broken, and a small amount of com-pression is visible on the left side. The quad-rate also forms the ventral margin of a smallopening medial to the one just described, butlateral to the foramen jugulare posterius,which seems to be a remnant of a more openfenestra postotica. In the Bothremydidae ingeneral the fenestra postotica is strongly sub-divided and separated by bony partitions. InKurmademys the bony partitions are thinnerand some are probably represented by carti-lage or thin bone, allowing the collapse andbreakage of foramen edges during fossiliza-tion. There is a well-developed quadrate-ex-occipital contact medially as in Bothremys;this is considerably more extensive than thesmall contact in FR 4922.
The presumed foramen chorda tympani in-ferius is present on the posterior surface ofthe processus articularis, roughly similar inposition to Podocnemis. Because of glue andcrud on the right quadrate, this is only visibleon the left.
PTERYGOID
Both pterygoids are preserved in Kurma-demys and both are nearly complete. Most ofthe thin pterygoid flange extending ventrallyfrom the quadrate process is missing fromboth pterygoids; the right one is more pre-served than the left.
12 NO. 3321AMERICAN MUSEUM NOVITATES
On the ventral surface the pterygoid con-tacts the palatine in a roughly transverse su-ture that trends slightly anterolaterally. Theforamen palatinum posterius is formed in thepalatine-pterygoid suture as in nearly allbothremydids. Medially the pterygoids meeton the midline for a bit less than half theirlength. They are separated posteriorly by thetriangular basisphenoid.
As in all pleurodires, there is a laterallyprojecting processus trochlearis pterygoidei.In Kurmademys the processus does not ex-tend at a sharp right angle as in the Santanabothremydids (Gaffney et al., in press), butis only slightly less than a right angle, muchas in Foxemys. It is not as acute as in chelidsand Araripemys. The flange or web that ex-tends ventrally from the base of the proces-sus trochlearis pterygoidei along the quadrateprocess in all pleurodires is mostly missingin Kurmademys. The portion preserved isconsistent with that seen in other bothremy-dids.
The posterolaterally extended quadrateprocessus in Kurmademys is narrower andlonger than in FR 4922, Araripemys, pelo-medusids, and chelids. In these groups theprocess is relatively flat and more horizontal,while in Kurmademys and some other both-remydids, such as Foxemys, it is narrowerand more vertical. This condition seems tobe related to the presence of a ventrally con-cave depression in the posterolateral part ofthe pterygoid in these forms (fig. 4). Thisdepression is presumably the pterygoideusmuscle attachment site. In Kurmademys thedepression is shallower than in Nigeremysand Foxemys, but it covers a larger area. Itsmargins are not as well defined in Kurma-demys as it is in those taxa. There is no de-velopment of an overhang of this depressionby the pterygoid as in the Podocnemididae.
The foramen posterius canalis carotici in-terni in Kurmademys lies entirely within thebasisphenoid (fig. 3B); the pterygoid doesnot participate in its formation as it does inmany other bothremydids. The Kurmademyscondition is unique within pelomedusoids.The posterior margin of the pterygoid con-tacts the narrowly exposed prootic betweenthe basisphenoid and quadrate contacts.
Most of the dorsal surface of the pterygoidis visible in Kurmademys, although the re-
gion inside the cavum cranii is variably ob-scured by pieces of matrix. The crista pter-ygoidea is relatively low. The pterygoidforms the ventral margin of the foramen ner-vi trigemini as in other bothremydids, but theforamen is not placed very close to the fo-ramen stapedio-temporale as it is in manyother bothremydids, such as Bothremys andFoxemys. The pterygoid forms the floor ofthe sulcus palatinopterygoideus, which liesbetween the side wall of the cavum craniiand the processus trochlearis pterygoidei.
The anterior contacts of the pterygoid atthe base of the processus trochlearis ptery-goidei are visible on both sides of ISI R152.The pterygoid plus the palatine and jugalform the postorbital wall, as exposed poste-riorly in the adductor muscle chamber. Thepterygoid has a very narrow contact with theparietal medially, broader contacts with thepostorbital more laterally, and with the jugalmost laterally.
SUPRAOCCIPITAL
The supraoccipital in ISI R152 is completeventrally and anteriorly, but is missing theposterior part of the crista supraoccipitalis.
The supraoccipital in turtles is Y-shaped,with paired lateral projections forming themedial part of the cavum labyrinthicum oneach side. In cryptodires and most pleurodi-res the supraoccipital has a tripartite suture,with the prootic and opisthotic visible on thedorsal surface of the otic chamber. It is un-usual to find that in a group of bothremy-dids—the Bothremys Group of Lapparent deBroin and Werner (1998)—the supraoccipitalcontacts the quadrate and separates the pro-otic from the opisthotic. In Kurmademys thisunusual condition is present. The supraoccip-ital on the right side has a broad contact withthe quadrate laterally and separates the pro-otic from the opisthotic. On the left side thesupraoccipital is complete and separates theprootic and opisthotic, but the quadrate isdamaged. This degree of quadrate contact bythe supraoccipital is similar in Kurmademys,Bothremys, Foxemys, and Rosasia. The con-tact is absent in Taphrosphys and indeter-minate in Zolhafah, Nigeremys, and Arenila.
The crista supraoccipitalis is usually rela-tively short in bothremydids. In Kurmademys
2001 13GAFFNEY ET AL.: KURMADEMYS
it is broken posteriorly and its length is in-determinate.
EXOCCIPITAL
Both exoccipitals are preserved and onlylack the condylus occipitalis, although thereis some breakage around the posterior foram-ina.
The exoccipital contacts the supraoccipitaldorsally, the opisthotic dorsolaterally, thequadrate ventrolaterally, and the basioccipitalventrally. The quadrate-exoccipital contactoccurs in all Bothremydidae and is absent inother pleurodires.
Dorsomedially the exoccipital forms thelateral and ventral margin of the foramenmagnum. Ventromedially the exoccipital pre-sumably participates in the formation of thecondylus occipitalis, but this structure is bro-ken off at its base in ISI R152, so the pres-ence or absence of the basioccipital in thecondylus is not determinable. The foramenjugulare posterius in Kurmademys is formedentirely by the exoccipital. The bone sur-rounds most of the foramen, but on each sidethe foramen is open laterally due to the pres-ence of a narrow fissure. This fissure is dif-ferent in shape on both sides and may be dueto breakage, in which case the original con-dition of the foramen would be closed. Kur-mademys has two foramina nervi hypoglossias in all other pelomedusoids; they lie nearthe base of the condylus occipitalis, ventro-lateral to the foramen magnum. The moredorsal foramen is formed entirely within theexoccipital, but the more ventral one isformed in the exoccipital-basioccipital su-ture.
BASIOCCIPITAL
The basioccipital in ISI R152 is nearlycomplete. A small amount of breakage is vis-ible on each tuberculum basioccipitale andthe broken condylus occipitalis does notclearly show basioccipital sutures.
The basioccipital of Kurmademys has abroad, transverse contact with the basisphe-noid anteriorly. Posterolaterally the basioc-cipital contacts the quadrate, and posterodor-sally there is a broad contact with the exoc-cipitals. The tuberculum basioccipitale isformed about equally by the quadrate and ex-
occipital. A shallow, median concavity liesbetween the paired tuberculae and is formedalmost entirely by the basioccipital.
PROOTIC
Both prootics in ISI R152 are preserved;the right one is nearly complete and the leftone has a partially eroded dorsal surface.
The prootic is exposed on the dorsal andanterior surface of the otic chamber with thefollowing contacts: the parietal medially, thequadrate laterally, the supraoccipital posteri-orly, and the pterygoid ventrally. There is noprootic-opisthotic contact in Kurmademys.The foramen stapedio-temporale is formed inthe prootic-quadrate suture. In contrast to allother bothremydids, the foramen opens an-terodorsally rather than anteriorly. It is visi-ble in dorsal view in Kurmademys, but inother bothremydids it is barely or not visiblein dorsal view. The position of the foramenstapedio-temporale in Kurmademys is verysimilar to that in pelomedusids and chelids.The foramen nervi trigemini is formed by theprootic dorsolaterally, the parietal dorsome-dially, and the pterygoid ventrally. The fo-ramen is best preserved on the right side; theleft foramen nervi trigemini is larger and hasbroken edges.
On the ventral surface of ISI R152, theprootic is exposed where the three bones (thepterygoid, basisphenoid, and quadrate) meet(fig. 4). This is in the deepest part of theconcavity formed by these bones. The con-cavity is presumed to be for the pterygoideusmuscle attachment. The prootic is exposedand has a very similar shape on both sides,so this is not interpreted as a consequence ofpreservation or an artifact. The prootic has adistinct foramen, here interpreted as the fo-ramen nervi facialis (VII), for the facialnerve, always closely associated with theprootic ossification. The form of the prooticexposure in Kurmademys is not like that inany other pleurodire. The primitive conditionof the ventral prootic exposure occurs inchelids, pelomedusids, and Araripemys, allof which have a large prootic exposure withthe foramen posterius canalis carotici interniin the prootic. In all bothremydids, podoc-nemidids, and FR 4922 the prootic does notcontain the internal carotid as it does in chel-
14 NO. 3321AMERICAN MUSEUM NOVITATES
ids and pelomedusids. In FR 4922 the prooticis partially exposed in a narrow space be-tween the basisphenoid and quadrate, similarto the primitive position found in chelids andpelomedusids, but is distinctly posterior tothe prootic, as exposed in Kurmademys. It islikely that the prootic exposure in Kurma-demys is not a retention of a primitive state,but is an autapomorphy of this genus, prob-ably related to the development of the pter-ygoideus concavity that removed coveringelements. Small areas of the prootic are var-iably exposed in other bothremydid skullsdue to erosion or the development of the con-cavity.
OPISTHOTIC
Both opisthotics are preserved in ISIR152; the right one is nearly complete andthe left one is missing a small part anteriorly.
In dorsal view, the opisthotic has thesecontacts: supraoccipital anteromedially, thesquamosal laterally, and the exoccipital pos-teromedially. There is no prootic-opisthoticcontact. The opisthotic in Kurmademys endsposteriorly at about the same level as thesquamosal; it does not extend posteriorly be-yond the squamosal as in pelomedusids, Ar-aripemys, and FR 4922. Ventrally the opis-thotic forms the roof and some of the sub-divisions of the fenestra postotica. In ISIR152 the opisthotic divides the fenestra pos-totica into two portions: a more lateral onethat seems to have contained the stapedialartery and lateral head vein, and a more me-dial one with unknown contents. The medialforamen has a finished dorsal margin, al-though as preserved, the ventral margin isbroken. Presumably the more medial fora-men just held cartilage as in living turtleswith an open fenestra postotica (Gaffney,1979).
BASISPHENOID
The basisphenoid is complete and wellpreserved in ISI R152. The cavum cranii islargely free of matrix and some of the dorsalsurface of the basisphenoid is visible.
The basisphenoid in Kurmademys is notstrongly triangular as in other bothremydidsbut is more pentagonal. It is a relatively largeelement, wider than long. The anterior con-
tact with the pterygoids trends posterolater-ally and anteromedially, and the angle thissuture makes with the midline is similar tothat in Foxemys and Zolhafah. At the antero-lateral corner of the basisphenoid, betweenthe pterygoid and quadrate contacts, is ashort contact with the prootic. In nearly allbothremydids the prootic is covered, so thiscontact is unusual. The lateral margin of thebasisphenoid is a long, parasagittal contactwith the quadrate. This contact in Kurma-demys is longer than in any other bothre-mydid; Foxemys and Polysternon most close-ly approach it. Posteriorly the basisphenoidhas a transverse contact with the basioccipital.
In contrast to all other bothremydids, inKurmademys the foramen posterius canaliscarotici interni is formed completely by thebasisphenoid, without participation of thepterygoid. However, the foramen is veryclose to the pterygoid suture, particularly onthe right side. The foramen posterius canaliscarotici interni in Kurmademys is also placedfarther anteromedially than in any otherbothremydid. This could be explained mor-phologically by a reduced ossification of thecanalis caroticus internus posteriorly. The ca-nalis in all bothremydids travels anterome-dially and slightly dorsally to enter the sellaturcica. If the canalis in a form like Bothre-mys (which has the foramen posterius canaliscarotici interni placed far posterolaterally)were to be exposed by the removal of boneventrally, the foramen would appear to mi-grate anteromedially along the path of the ca-nalis caroticus internus. It is possible that thiscondition could result from the developmentof a deep pterygoideus muscle concavity,which is formed directly ventral to the ca-nalis caroticus internus. Although Kurmade-mys has a distinct pterygoideus concavity, itis relatively shallow compared to such formsas Foxemys, and Foxemys does not have theforamen posterius canalis carotici interniplaced anteromedially.
The dorsal surface of the basisphenoid inKurmademys has the dorsum sellae and sellaturcica visible. The dorsum sellae overhangsthe sella turcica and has the foramen anteriuscanalis carotici interni also hidden in dorsalview and lying at the posterolateral corner ofthe sella turcica. There is a small processusclinoideus on the left side; the right one is
2001 15GAFFNEY ET AL.: KURMADEMYS
broken. The shape and general proportions ofthe dorsum sellae and sella turcica are similarto those in Pelusios. The degree of overhangof the dorsum sellae, however, is greater inKurmademys than it is in Pelusios. The ros-trum basisphenoidale is fused into a singlestructure, but its anterior end shows the twoossified trabeculae rather than the single ros-trum seen in Pelusios. There is no sign of aforamen nervi vidiani in the left sulcus cav-ernosus. The right side still has some matrix.
RELATIONSHIPS
Kurmademys is a pleurodire because it hasthese synapomorphies of the group listed byGaffney and Meylan (1988) as diagnostic forthe Pleurodira: (1) processus trochlearis pter-ygoidei present, (2) quadrate process belowcranio-quadrate space, (3) epipterygoid ab-sent, (4) foramen palatinum posterius behindorbit, and (5) pelvis suturally attached to car-apace and plastron (based on shell materialfrom same locality ascribed to Kurmademys).It is a member of the Pelomedusoides (sensuBroin, 1988; Meylan, 1996; Lapparent deBroin and Werner, 1998; Tong et al., 1998),which is equivalent to the Pelomedusidae inthe classical sense (sensu Gaffney and Mey-lan, 1988) because it has these characters: (1)nasals absent, (2) prefrontals meeting onmidline, and (3) splenial absent (based onlower jaws from same locality ascribed toKurmademys). Kurmademys can be identifiedas a member of the family Bothremydidaebased on its possession of the following char-acters: (1) exoccipital-quadrate contact, (2)incisura columellae auris closed by bone, and(3) eustachian tube and stapes separated bybone. Within the Bothremydidae, Kurmade-mys can be allied with Bothremys, Rosasia,Foxemys, and Zolhafah (the BothremysGroup of Lapparent de Broin and Werner,1998) on the basis of these characters: (1)triangular triturating surfaces, (2) supraoccip-ital-quadrate contact, (3) maxilla-quadrato-jugal contact, and (4) palatine widely ex-posed on triturating surface.
Kurmademys is unique among knownbothremydids in having extensive temporalemargination, a small postorbital, a large pre-columellar fossa, and a foramen posterius ca-nalis carotici interni formed completely by
the basisphenoid. Foxemys, Polysternon, Ro-sasia, Zolhafah, and Bothremys can be unitedby having a foramen stapedio-temporale veryclose to the foramen nervi trigemini and abroad preorbital part of the skull, features ab-sent in Kurmademys.
Kurmademys has characters in commonwith the Bothremys Group, but the state ofthe current analysis shows only a few stepsfrom Kurmademys as the sister group to theNigeremys Group plus Bothremys Group.The addition of as yet undescribed taxa willhelp to resolve its phylogenetic position.
ACKNOWLEDGMENTS
We wish to thank our associates in pleu-rodiran studies, P. Meylan, H. Tong, and R.Wood, for their support and counsel on thisongoing project. Ed Heck did the photogra-phy and illustrations with his customary ex-cellence. We particularly appreciate the helpof Judy Galkin in the preparation of the pa-per.
S. L. Jain and A. Sahni provided infor-mation and access to specimens, which wegreatly appreciate. We thank the Indian Sta-tistical Institute for field logistics and sup-port, permission to study the specimens, andthe National Geographic Society for fundingthis project.
REFERENCES
Antunes, M. T., and F. de Broin1988. Le Cretace terminal de Beira Litoral,
Gaffney, E. S., P. A. Meylan, and R. C. Wood1997. Changing interpretations of the history
of the side-necked turtles. JVP Abstr.17(3): 47A–48A.
Gaffney, E. S., K. E. Campbell, and R. C. Wood1998. Pelomedusoid side-necked turtles from
Late Miocene sediments in southwest-ern Amazonia. Am. Mus. Novitates3245: 12 pp.
Gaffney, E. S., H. Tong, S. Chatterjee, R. T. J.Moody, and R. Hirayama
1998. Evolution of the bothremydid turtles.JVP Abstr. 18(3): 44A.
Gaffney, E. S., D. de A. Campos, and R. Hira-yama
2001. Cearachelys, a new side-necked turtle(Bothremydidae: Pelomedusoides) from
the Early Creatceous of Brazil. Am. Mus.Novitates 3319: 20 pp.
Jain, S. L.1977. A new fossil pelomedusid turtle from
the Upper Cretaceous Pisdura sedi-ments, Central India. J. Palaeontol.Soc. India 20: 360–365.
1986. New pelomedusid turtle (Pleurodira:Chelonia) remains from Lameta For-mation (Maastrichtian) at Dongargaon,central India, and a review of Pelome-dusids from India. J. Palaeontol. Soc.India 31: 63–75.
Lapparent de Broin, F. de, and C. Werner1998. New late Cretaceous turtles from the
western desert, Egypt. Ann. Paleontol.84(2): 131–214.
Linnaeus, C.1758. Systema naturae. 10th ed., vol. 1.
Stockholm, 824 pp.Matley, U. A.
1929. The Cretaceous dinosaurs of the Tri-chinopoly district and the rocks asso-ciated with them. Rec. Geol. Surv. In-dia 61: 337–349.
Meylan, P. A.1996. Skeletal morphology and relationships
of the Early Cretaceous side-neckedturtle, Araripemys barretoi (Testudines:Pelomedusoides: Araripemydidae),from the Santana Formation of Brazil.J. Vertebr. Paleontol. 16(1): 20–33.
Sastry, M. V. A., V. D. Mamgain, and B. R. J.Rao
1972. Ostracod fauna of the Ariyalur group(Upper Cretaceous), Tiruchirapalli dis-trict, Tamil Nadu. Mem. Geol. Surv. In-dia, n. ser. 40: 1–59.
Singh, S. D., A. Sahni, E. Gaffney, and H. H.Schleich
1998. Reptilia from the Intertrappean Beds ofBombay (India). Veroff. Fuhlrott-Mus.4: 307–320.
Tong, H., E. S. Gaffney, and E. Buffetaut1998. Foxemys, a new side-necked turtle
(Bothremydidae: Pelomedusoides)from the Late Cretaceous of France.Am. Mus. Novitates 3251: 19 pp.
Wood, R. C.1970. A review of the fossil Pelomedusidae
(Testudines, Pleurodira) of Asia. Bre-viora 357: 1–24.
Yadagiri, P., and K. Ayyasami1987. A carnosaurian dinosaur from the Kal-
