Pelomedusidae

GAFFNEY, EUGENE S., MOODY, RICHARD T. J. & WALKER, CYRIL A., 2001, Cearachelys, a New Side-Necked Turtle (Pelomedusoides: Bothremydidae) from the Early Cretaceous of Brazil, American Museum Novitates 3320, pp. 1-20 : 4-20

some Pelomedusidae , such as Pelusios subniger .

The prefrontal in Cearachelys has the usual pelomedusoid contacts: maxilla anterolaterally, frontal posteriorly. It borders the orbit laterally and the apertura narium externa anteriorly. The ventral surface of the prefrontal in Cearachelys is smooth, concave ventrally in its anterior portion, but bears the parasagittal ridge for the sulcus olfactorius more posteriorly. The prefrontal sends a process ventrally along the anterior margin of the fossa orbitalis comparable in size and extent to that in Recent Pelomedusidae . It contacts the maxilla for most of its length and does not reach other elements.

FRONTAL

The frontal of Cearachelys is preserved on both sides in TUTg 1798. It is similar in size, shape, and contacts to living Pelomedusidae . The frontal contacts are with the prefrontal anteriorly, orbit laterally, postorbital posterolaterally, parietal posteriorly, and the other frontal medially. Anteriorly the frontal projects on the midline to a considerable extent on the ventral surface, but only slightly on the dorsal surface.

The frontal forms the well­developed sulcus olfactorius on the ventral surface. In Cearachelys the sulcus is very similar to the living Pelomedusidae . The frontal in Cearachelys does not have a ventral process along the edge of In Cearachelys a very narrow process of the the processus inferior parietalis, although the postorbital runs ventrally along the orbital marfrontal does reach the processus as in recent gin to reach or nearly reach the maxilla and Pelomedusidae . prevent jugal exposure. On the left side of the

MPSC specimen there is some breakage along PARIETAL the orbital margin in this area, but it is clear

that the postorbital contacts the maxilla in a The parietal is preserved on both sides in

narrow area of sutural interdigitation. In TUTg TUTg 1798 and the MPSC specimen, although

1798 both sides show the contact. In Lapparent it is not entirely complete in either skull. The

de Broin and Werner (1998) an unnamed Morparietal of Cearachelys is nonetheless com­

occan skull is figured, which shows a broad pletely known through both specimens.

contact of postorbital and maxilla, and this The dorsal plate of the parietal in Cearach­

condition is similar to Cearachelys . No other elys contacts the frontal anteriorly and is ex­

bothremydid shows this retraction of the jugal cluded from the orbit as in other Bothremy­

from the orbit. Ventrally the jugal of Cearachdidae and the Pelomedusidae . Laterally the pa­

elys is exposed on the palate to a limited exrietal contacts the postorbital, but not the quad­

tent, not to as great an extent as seen in Bothratojugal, as in FR 4922. The temporal

remys and Rosasia .

emargination in Cearachelys is more extensive

The jugal is exposed in the floor and posthan in living Podocnemididae and Foxemys ,

terior wall of the orbit in Cearachelys as in but not as extensive as in Pelomedusidae . The

most pelomedusoids. In the orbital floor the otic chamber is completely uncovered, but

jugal contacts the postorbital dorsally and posthere is a relatively long suture between the

teriorly, the maxilla laterally and anteriorly, parietal and the postorbital relative to that seen

and the palatine medially. The limited postorin the Pelomedusidae .

bital­maxilla contact barely separates the or­ The ventral process of the parietal, the pro­

bital floor exposure of the jugal from the cheek cessus inferior parietalis, is well preserved in

exposure of the jugal, a condition unique to both skulls. It has relations similar to those in

Cearachelys . The postorbital wall of Ceara­ Pelomedusidae . Anteriorly it contacts the pter­

chelys shows the jugal forming the more venygoid and forms most of the side wall of the

tral portion, contacting the maxilla ventrolatbraincase. Posteriorly the parietal forms the an­

erally, the quadratojugal laterally, the postorterodorsal margin of the foramen nervi trigem­

bital dorsally, and the pterygoid posteromediini (the ventral margin is formed by the pter­

ally.

ygoid and the posterodorsal margin by the pro­

otic). Posterior to the foramen nervi trigemini,

the parietal contacts the prootic and then the QUADRATOJUGAL

supraoccipital in a suture that rises dorsally to

The quadratojugal is preserved on the left the skull roof.

side of the MPSC specimen and on both sides

of TUTg 1798.

JUGAL

The quadratojugal of Cearachelys forms the The jugal is preserved completely on the left posterior part of the cheek. There is only a side and in part on the right side of the MPSC slight dorsal curve to the lower margin of the specimen. In TUTg 1798 the left jugal is com­ quadratojugal to suggest a cheek emargination; plete, but only part of the right jugal is pre­ nothing like the emargination seen in Peloserved. medusidae is present. The quadratojugal of The jugal is a complex element that can be Cearachelys extends from the anterior limit of described as having three areas: lateral, ventral, the upper temporal emargination to the ventral and posterior (with two exposure surfaces, an­ margin of the cheek. In contrast to members terior and posterior). The lateral exposure of of the Pelomedusidae , which have a well­dethe jugal in Cearachelys is unusual among veloped cheek emargination and no jugal­maxbothremydids. The jugal lies just posterior to illa contact, Cearachelys has a broad maxillathe orbit as in all bothremydids; however, in quadratojugal contact, as in Foxemys, Nigeremost bothremydids the jugal enters the orbit. mys, Polysternon , an unnamed Moroccan form (Lapparent de Broin and Werner, 1998), and Rosasia . It is likely that this is a Bothremydidae synapomorphy.

In Cearachelys the anterior contacts of the quadratojugal are with the maxilla anteroventrally, the jugal anterodorsally, and the postorbital dorsally. Posteriorly there is a long curved contact with the quadrate. Posterodorsally a very narrow process of the quadratojugal contacts the squamosal.

SQUAMOSAL

The squamosal is preserved in both Cearachelys specimens.

The squamosal is a cone­shaped bone that antrum that is quite large. Cearachelys has an sits on the posterodorsal corner of the quadrate antrum postoticum intermediate in size be­ and contains much of the antrum postoticum. tween these, comparable to that in Foxemys .

The antrum of Cearachelys is moderately well The complete squamosal is not known for developed relative to other pelomedusoids. other bothremydids, such as Arenila , Zol­ Bothremys and Taphrosphys have an antrum hafah, Rosasia , or Bothremys . But in Foxthat is very small, merely an elongate canal, emys the posterior part of the squamosal is while members of the Pelomedusidae have an developed into a flat plate. In Taphrosphys there is a more vertical flange extending ventrally. Cearachelys lacks these posterior processes and ends in a curved edge with only a slight horizontal component continuous with the posterior extension of the opisthotic. It is similar to the condition in Pelomedusidae .

POSTORBITAL

Both postorbitals are preserved in TUTg 1798, but there is some damage and sutures are not entirely clear. However, by using information from both sides, the postorbital in this specimen can be fully restored. The MPSC specimen has only the left postorbital Anteroventrally the postorbital contacts the and that is missing part of its anterior edge. jugal, and posteroventrally it contacts the The postorbital in Cearachelys is a long quadratojugal. Anteromedially the postorbitelement extending from the orbit to the tem­ al contacts the frontal, and posteromedially poral skull roof margin, much as in Foxemys. the parietal.

The postorbital of Cearachelys is larger than

in Araripemys , Pelomedusidae , and Podoc­ PREMAXILLA

nemididae. A very narrow ventral process of

the postorbital in Cearachelys contacts the Both premaxillae are present, but dammaxilla in the orbital margin, barely pre­ aged, in TUTg 1798. They are absent in the venting jugal exposure in the orbital margin. MPSC specimen.

The premaxilla of Cearachelys bears a distinct labial ridge anteriorly with a flat plate posteriorly. The labial ridge, however, is relatively thin in Cearachelys in anterior view, in contrast to the deeper ridge of Foxemys . Posterior to the labial ridge on the palatal surface, the premaxilla rises in a shallow, smooth arch to the vomer contact. In Foxemys , Bothremys , Zolhafah , and Polysternon there is a step on the triturating surface between the labial ridge and vomer. The premaxilla of Cearachelys is triangular, being broadest anteriorly and narrowing posteriorly. The lateral limits, however, are represent­ ed by broken edges and the precise position of sutures is not clear.

The dorsal surface of the premaxilla forms the anterior floor of the fossa nasalis and the margin of the apertura narium externa. As on the ventral surface, the dorsal surface slopes posterodorsally to meet the vomer. The apertura ventral margin is nearly straight in Cearachelys , with only a slight dorsal curve on the midline, similar to Foxemys but in contrast to the apertura margin in Bothremys , which has a sharp dorsal curve at the midline. Nuances of shape of the apertura narium externa, however, are variable even among species of living Pelusios .

MAXILLA

The posterior portion of the left maxilla is preserved in the MPSC specimen. Most of both maxillae are present in TUTg 1798, but laterally into the maxilla, giving the sutural both are damaged to the extent that the an­ contact an interdigitating shape. teromedial edges and contacts are not known. Ventrally the maxilla of Cearachelys VOMER forms most of the upper triturating surfaces.

The vomer is preserved only in TUTg The labial ridge is relatively narrow, com­

1798. Its lateral margins are represented by parable to Pelomedusa , in contrast to the

broken edges. thick labial ridge in Bothremys and Rosasia .

The vomer of Cearachelys is roughly The labial ridge in Cearachelys , however, is

dumbbell shaped, the presumed primitive low, not deep as in Taphrosphys . The tritu­

condition for pleurodires and pelomedusoids. rating surface of Cearachelys , to the extent

The anterior end is swollen and contacts the preserved, is smooth, with complete absence

paired premaxilla, a contact preserved only of the pits seen in Bothremys , Rosasia , and

on the left side. The posterior end, also swol­ Zolhafah . The triturating surface, however,

len, contacts the palatine bones. The anterior agrees with these taxa and Polysternon and

end of the vomer lies ventral to the posterior Foxemys in being widened posteriorly to

end, and between them is a narrow bar sepform a triangular area. On the left side of

arating the paired apertura narium interna. TUTg 1798 the triangular triturating area is

The vomer of Cearachelys is quite similar in clearly visible. On the right side much of the

shape and proportions to that in FR 4922 and surface is damaged and the medial limits

lacks the larger anterior end of Bothremys or eroded.

the stout central bar of Nigeremys . The maxilla in Cearachelys contacts the premaxilla anteriorly and the palatine poster­

PALATINE omedially. Between the two contacts the maxilla limits are represented by broken edg­ Palatines are preserved on both sides of es, and the margin of the apertura narium both specimens. In the MPSC specimen the interna is not definite. However, by using the left palatine is nearly complete, while the bone edges as preserved on both sides, limits right one lacks part of its anterior edge. Presof the apertura are determinable. The aper­ ervation and sutures are clear. In TUTg 1798 tura narium interna of Cearachelys is placed the right palatine is nearly complete, but the relatively close to the midline. Posteriorly the left one lacks its posterolateral portion and is maxilla in Cearachelys contacts the jugal. poorly preserved, missing some of its ventral The more medial part of this contact is on surface. the palatal surface, and more laterally it On the ventral surface, the palatine in curves dorsally onto the posterior surface of Cearachelys contacts the maxilla anterolatthe postorbital wall. erally, the jugal laterally, the pterygoid pos­ In lateral view the vertical plate of the teriorly, the other palatine medially, and the maxilla in Cearachelys forms the posterior vomer anteromedially. In such forms as margin of the apertura narium externa, the Bothremys the palatine makes up a signifiventral margin of the orbit, and the anterior cant portion of the large triturating surface, portion of the cheek wall. The lateral margin but in Cearachelys this contribution is no of the apertura narium externa slopes poster­ more than in Pelomedusidae and FR 4922. odorsally as in nearly all other bothremydids. In forms such as Bothremys the palatine is The orbit is larger than in Bothremys and deeply curved to form a choanal trough leadsmaller than in Pelomedusa . The posterior ing to the apertura narium interna. In Cearcheek contacts of the maxilla are with the achelys the palatine surface leading into the quadratojugal more ventrally and with the ju­ apertura is much flatter and not strongly gal dorsally. A very thin ventral process of curved. the postorbital contacts the maxilla along the The dorsal surface of the palatine is well orbital margin. The exposure of the maxilla preserved and easily seen in the MPSC specin the orbital floor is bordered by the jugal imen. There is a low dorsal ridge that conposteriorly and the palatine posteromedially. tacts the processus inferior parietalis and The palatine sends a narrow process antero­ forms the lower margin of the foramen in­

(bottom) views of carapace. See fig. 9 .

terorbitale. This is in contrast to the condition in Emydura (and some Pelusios , i.e., USNM 42144), in which the palatine does not contact the parietal. However, in other Pelusios ( YPM 5429) there is a very similar condition. In the posterior suture with the pterygoid both bones form the foramen palatinum posterius. The palatine has an anterolateral extension meeting the jugal and the foramen lies in the suture along this extension.

QUADRATE

Both quadrates are well preserved in both specimens. The right quadrate of the MPSC specimen lacks its anterior edge along the cavum tympani.

The quadrate of Cearachelys has a long, C­shaped anterior contact with the quadratojugal. Posteriorly the cone­shaped squamosal fits onto the posterodorsal corner of the quadrate. The narrow quadratojugalsquamosal contact prevents exposure of the quadrate along the temporal emargination. Most of the quadrate is involved in the formation of the cavum tympani. Cearachelys lacks any precollumellar fossa, as is the case in nearly all bothremydids, but in contrast to most other pelomedusoids and chelids. The antrum postoticum of Cearachelys is not extremely small as in Bothremys and Taphrosphys , but it is significantly smaller than in pelomedusids and FR 4922.

The cavum tympani of most bothremydids is a hemispherical depression, with a canal in the center for the stapes, a small or absent

(bottom) views.

antrum postoticum, and a shallow groove for this difference may be due to TUTg 1798 the eustachian tube that is separated from the being younger and generally less ossified.

stapes by a wall of bone. This is the condi­ Cearachelys and Foxemys differ slightly in tion in Bothremys , Taphrosphys , Nigeremys , the shape of the incisura columellae auris. In Arenila , Rosasia , and Zolhafah . However, Cearachelys the incisura is oval in shape, Cearachelys , like Foxemys , has an open in­ much as in pelomedusids, so that, clearly, the cisura columellae auris, so that there is no eustachian tube was contained in it, and this bony separation posteriorly between the eus­ is demonstrated by examination of the living tachian tube and the stapes. In the MPSC species. In Foxemys the right incisura is specimen there is a small bar of bone on the complete in both available skulls, and it left side that closes the incisura at its most seems to show that the incisura widens latdistal portion. On the right side the incisura erally and is quite narrow for most of its is completely open but a broken surface sug­ length in contrast to Cearachelys . It is posgests that it was also probably closed origi­ sible that the eustachian tube is excluded nally. TUTg 1798, however, is open on both from the incisura in Foxemys . In Cearachelys sides and looks as if that was the original the oval­shaped incisura is continuous poscondition. The MPSC specimen is larger, and teriorly with a groove and lateral ridge on the

TABLE 1 Comparison of Santana Pleurodire Shells

posterior face of the ventral part of the quadrate. This ridge seems to form an outer limit for the continuing path of a eustachian tube. A similar groove and ridge occurs in some pelomedusids, and is associated with the eustachian tube. Foxemys also has this ridge, and it substantiates the presumption that the eustachian tube is included in the incisura of Foxemys .

The medial contacts of the quadrate in Cearachelys on the dorsal surface of the otic chamber are with the prootic anteromedially, the supraoccipital medially, and the opisthotic posteromedially. The prootic and opisthotic contacts are found in all pleurodires, but the quadrate and supraoccipital contact occurs only within the Bothremys subgroup of bothremydids.

On the ventral surface the quadrate contacts the pterygoid anteromedially, the basisphenoid medially, and the exoccipital posteromedially. Between the basisphenoid and exoccipital contacts there is a narrow contact with the basioccipital. The basioccipital contact occurs in all bothremydids and podocnemidids. The basisphenoid­quadrate contact occurs in pelomedusids, podocnemidids, and all bothremydids.

Medially the foramen stapedio­temporale is formed in the quadrate­prootic suture. As in nearly all other bothremydids, the foramen is on the anterior surface of the otic chamber in Cearachelys . The canalis stapedio­temporale is well preserved and open on the left side of TUTg 1798. Although barely visible in dorsal view, the foramen in Cearachelys is not very close to the foramen nervi trigemini as in Bothremys .

PTERYGOID

In the MPSC specimen, both pterygoids are complete and well preserved, except for the distal margins of the pterygoid flange. In TUTg 1798, the left pterygoid is nearly complete, but the right one lacks a significant contact. Its position is similar to that in Rosasia portion anteriorly and is damaged by erosion and Foxemys . It differs from the pterygoidon its ventral surface. The dorsal structures quadrate formation of the foramen seen in of the pterygoid are visible and well pre­ Bothremys and the pterygoid­quadrate­basiserved in the MPSC specimen. sphenoid formation seen in Taphrosphys , Zol­

On the ventral surface, the pterygoid of hafah, and Arenila . The lateral margin and Cearachelys has a roughly transverse suture most of the ventral margin of the foramen poswith the palatine that trends anterolaterally to terius canalis carotici interni in Cearachelys meet the jugal. The foramen palatinum pos­ are formed by pterygoid. The medial edge is terius is formed in the palatine­pterygoid su­ formed by the basisphenoid. The dorsal roof ture as in most pleurodires. Medially the of the foramen has both the basisphenoid and pterygoids meet on the midline for about half pterygoid making it up. As in most of the other their length. The basisphenoid separates bothremydids the foramen leads into the cathem posteriorly. As in all pleurodires, there nalis carotici interni, lying in a nearly horizonis a laterally projecting processus trochlearis tal plane, so that the ventral margin of the fopterygoidei. In Cearachelys the processus ramen and canalis are quite thin. This is in lies at a less acute angle to the midline in contrast to the condition in pelomedusids, contrast to the more acute angle seen in pe­ where the canalis is much more vertical. lomedusids, chelids, FR 4922, and Araripe­ There is some variability in the structures mys. The flange of the pterygoid that extends around the foramen posterius canalis carotici ventrally from the base of the processus interni in Cearachelys . The MPSC specimen trochlearis pterygoidei along the quadrate shows very little damage of the thin bones process in all pleurodires is often broken in in this region, but TUTg 1798 is eroded to a fossils, because it is so thin and fragile. It is varying extent on both sides. In the MPSC partially preserved in both Cearachelys spec­ specimen the posterior flange of the pteryimens and seems to be developed to about goid just posterolateral to the foramen posthe extent seen in pelomedusids. terius canalis carotici interni has a foramen

The posterolaterally extended quadrate opening out of a canal that parallels the caprocess of the pterygoid in Cearachelys is nalis carotici interni and is just lateral to it. distinctly narrower and longer than in FR This foramen and canal is probably for the 4922, Araripemys , pelomedusids, or chelids. facial branch of VII and probably leads into In the latter groups the processus is relatively the prootic, where this nerve goes dorsally. flat and more horizontal, while in Ceara­ On the left side of the MPSC specimen a chelys it is narrower and more vertical. This small hole in the floor of the canal allows is related to the presence in Cearachelys of determination of its size and location, but the a ventrally concave depression in the pos­ opening into the prootic is not visible on eiterolateral part of the pterygoid. This depres­ ther side of this specimen. In TUTg 1798, sion has a sharp anterolateral margin that ex­ however, the prootic and the foramen nervi tends posterolaterally along the processus ar­ facialis are visible. Due to either erosion or ticularis of the quadrate, and has a barely less ossification in life (perhaps related to the perceptible margin medially on the basisphe­ smaller size of TUTg 1798), the pterygoid in noid. The depression is not as deep as in TUTg 1798 lacks the canal for the facial Foxemys nor much deeper as in Nigeremys , nerve seen in the MPSC specimen. On the but it is obviously present in contrast to right side of TUTg 1798, the prootic is ex­ Bothremys and Taphrosphys , which lack it. posed in a narrow band leading out of the This pterygoid depression differs from the canalis carotici interni and forming part of podocnemidid pterygoid chamber, seen in its the roof of the foramen posterius canalis carprimitive condition in Hamadachelys , by not otici interni, between the pterygoid and bahaving any development of an overhang by sisphenoid. In the anterior part of the prootic the basisphenoid or pterygoid. exposure is a dorsally extending foramen and

The foramen posterius canalis carotici inter­ canal, the foramen nervi facialis. On the left ni in Cearachelys lies in the basisphenoid­ side of TUTg 1798 more of the pterygoid is pterygoid suture, just anterior to the quadrate present, covering the foramen nervi facialis

ventrally. The foramen can still be seen by

looking along the length of the canalis carotici interni.

The dorsal surface of the pterygoid is visible in the MPSC specimen. The crista pterygoidea is quite low, there is a low dorsal area on the outside of the crista anteriorly, and it rises again to form the ventral margin of the foramen nervi trigemini. The anterodorsal margin of this foramen is formed by the parietal, and the posterodorsal margin by the prootic.

SUPRAOCCIPITAL

The supraoccipital is complete in both skulls. In the MPSC specimen a small part of the ventral edge of the crista supraoccipitalis is missing, and in TUTg 1798 part of the crista on the right side is gone near its base.

The supraoccipital of turtles is Y­shaped in posterior view, with paired lateral projections that contain the medial part of the cavum labyrinthicum, the recessus labyrinthicus supraoccipitalis. In cryptodires and most pleurodires the supraoccipital has a tripartite suture, with the prootic and opisthotic visible on the dorsal surface of the otic chamber. In Cearachelys the supraoccipital has a lateral projection that separates the prootic and opisthotic and contacts the quadrate. This contact is well preserved on both sides of the MPSC specimen and on the left side of TUTg 1798. On the right side of TUTg 1798 the area is damaged and not determinable.

The supraoccipital­quadrate contact also occurs in Bothremys , Rosasia , and Foxemys , but not in Taphrosphys . In Cearachelys the degree of contact is less than in the other bothremydids.

The crista supraoccipitalis is shorter in bothremydids than it is in most podocnemidids. In Cearachelys the crista extends posteriorly about as far as the posterior end of the squamosals. It is similar in length to most Pelomedusa and shorter than in most Pelusios . The crista in Cearachelys is slightly longer than in FR 4922.

EXOCCIPITAL

The exoccipitals are complete and well preserved in both the MPSC specimen and TUTg 1798.

The exoccipital in Cearachelys contacts

the supraoccipital dorsally (the exoccipitals do not meet in the midline above the foramen magnum as in some chelids), and the opisthotic laterally and anteriorly. Dorsomedially the exoccipital forms the lateral and ventral margin of the foramen magnum, a structure that varies little in pelomedusoids. Ventromedially the two exoccipitals form the condylus occipitalis, with no contribution from the basioccipital, a condition found in Bothremys , Zolhafah , Taphrosphys , and Arenila . Nonetheless, the basioccipital in Cearachelys does completely separate the exoccipitals in ventral view in both skulls. In Taphrosphys the exoccipitals are in contact ventrally as well and make up the neck of the condylus occipitalis.

Laterally the exoccipital in Cearachelys forms the medial portion of the foramen jugulare posterius, contacting the opisthotic dorsally and the quadrate ventrally. In Cearachelys the foramen jugulare posterius is open laterally as in Foxemys , not closed as in Bothremys , Taphrosphys , pelomedusids, and chelids. In Cearachelys , however, the foramen jugulare posterius is partially closed or restricted laterally in comparison to the more open condition in Araripemys and FR 4922. In Cearachelys there are two foramina nervi hypoglossi, as in all the other pelomedusoids. As in Taphrosphys and Bothremys , Cearachelys has an extensive quadrateexoccipital contact below the fenestra postotica. This contact is absent in chelids, podocnemidids, pelomedusids, and FR 4922.

BASIOCCIPITAL

The basioccipital is preserved intact in both TUTg 1798 and the MPSC skull.

The basioccipital of Cearachelys , as in other bothremydids, is relatively shorter than in chelids and pelomedusids, but the width is about the same. The basioccipital has a wide contact with the basisphenoid anteriorly and a narrow contact with the quadrate at its lateral limit. Posterolaterally and posteriorly the basioccipital contacts the exoccipital.

PROOTIC

The prootic in the MPSC specimen is present and well preserved on both sides, but in TUTg 1798 only the left prootic is well pre­ does not extend posterolaterally beyond the served, although both are present. squamosal as in pelomedusids, Araripemys , The prootic is exposed on the dorsal and and FR 4922. anterior surface of the otic chamber, contact­ Ventrally the opisthotic forms the roof and ing the parietal medially, the quadrate later­ some subdivisions of the fenestra postotica. ally, the supraoccipital posteriorly, and the In Cearachelys the foramen jugulare posterpterygoid ventrally. The foramen stapedio­ ius is open laterally and is continuous with temporale is formed in the prootic­quadrate the fenestra postotica. Laterally the fenestra suture and the foramen nervi trigemini is postotica is bordered by the quadrate, which formed between the parietal, pterygoid, and also forms most of the floor. The exoccipital prootic. contacts the quadrate and forms the more The foramen stapedio­temporale in Cear­ medial part of the floor. In Bothremys and achelys is situated on the anterior face of the Taphrosphys the fenestra postotica is comotic chamber as in other bothremydids, and pletely separated from the foramen jugulare in contrast to the primitive position in chelids posterius by a bar of bone formed by the and pelomedusids, where it is more posterior opisthotic and quadrate. This bar is absent in and faces dorsally. In Cearachelys and other Cearachelys . However, the fenestra postotica bothremydids the foramen stapedio­tempor­ in Cearachelys is nonetheless smaller and ale faces anteriorly and is barely visible in more restricted than in chelids, pelomedudorsal view. In chelids and pelomedusids the sids, and Araripemys . The medial part of the foramen lies on the dorsal surface of the otic fenestra postotica in Cearachelys is a narrow chamber and opens more dorsally. Although horizontal slit where the opisthotic and quad­ Cearachelys has the anterior­facing foramen rate nearly meet. In neither the TUTg 1798 stapedio­temporalis, the foramen is not in nor the MPSC skull is there any contact here. close proximity to the foramen nervi trigem­ The more lateral part of the fenestra postotica ini as in Taphrosphys , Bothremys , Rosasia , is figure­8­shaped, because it is partially di­ and Foxemys . As described under Supraoc­ vided by low ridges into a more lateral porcipital, the prootic does not contact the op­ tion for the lateral head vein and a more meisthotic in Cearachelys , due to a supraoccip­ dial portion for the stapedial artery. ital­quadrate contact. This contact is found in The processus interfenestralis of the opis­ Bothremys , Rosasia , and Foxemys . thotic is completely covered in Cearachelys The prootic in Cearachelys forms the pos­ as in all other bothremydids and podocnemterodorsal margin of the foramen nervi tri­ idids. The fenestra postotica is so small that gemini, similar to the condition in pelome­ the fenestra ovalis is only barely visible in dusids, and FR 4922. On the ventral surface TUTg 1798. a narrow band of the prootic is exposed in the roof of the foramen posterius canalis car­ BASISPHENOID otici interni in TUTg 1798. The prootic is exposed on the left side in the basisphenoid­ The basisphenoid is present and complete pterygoid­quadrate suture of the MPSC spec­ in both skulls, but in TUTg 1798 it is slightly imen. eroded. Most of the dorsal surface is visible

in TUTg 1798. OPISTHOTIC In ventral view the contacts of the basi­

sphenoid in Cearachelys are with the ptery­ Both opisthotics are present and complete goid anterolaterally, the quadrate posterolatin both the TUTg 1798 and MPSC speci­ erally, and the basioccipital posteriorly. The mens. basisphenoid is roughly triangular in shape, In dorsal view the opisthotic contacts the its apex separating the pterygoids for about supraoccipital anteromedially, the quadrate half their length. The basisphenoid forms the anterolaterally, the squamosal laterally, and medial margin of the foramen posterius cathe exoccipital posteromedially. The opis­ nalis carotici interni, but the foramen is close thotic in Cearachelys ends posteriorly at to the pterygoid­quadrate suture. The ventral about the same level as the squamosal; it surface of the basisphenoid in Cearachelys is broadly convex and does not participate in the pterygoideus muscle concavity.

The dorsal surface of the basisphenoid in Cearachelys shows the oval sella turcica, low dorsum sellae, and fused rostrum basisphenoidale as seen in Pelusios , Podocnemis , and Bothremys . The long, narrow rostrum and only barely overhanging dorsum are more similar to Bothremys . The paired foramen anterius canalis carotici interni lie close togeth­ er as in Podocnemis and Bothremys . There is no foramen caroticum laterale. The foramen nervi abducentis is posterior to the base of the processus clinoideus as in Pelusios and Podocnemis .

CARAPACE

The carapace of Cearachelys is nearly complete in TUTg 1798, but only fragments of the left and right bridge peripherals are present in the MPSC specimen.

The carapace of Cearachelys is moderately domed, much as in recent Pelomedusa . The carapace is composed of a nuchal, 8 neurals, 8 pairs of costals, 11 pairs of peripherals, a single suprapygal, and a single pygal. There are no fontanelles as in Araripemys , and all the bones are tightly sutured as in FR 4922 and most pleurodires. The principal distinguishing features of the carapace in Cearachelys lie in the neural bones. The first neural in most pelomedusoids is four sided and contacts only the nuchal, first costals, and second neural. In Cearachelys the first neural is six sided and has short, paired contacts with the second costals. The second neural is four sided, rather than six as in most pelomedusoids, and as a consequence does not contact the first costals. In Araripemys the second neural also does not contact the first costals. Neurals 3 to 6 in Cearachelys are the usual six­sided, coffinshaped bones, with short sides facing anteriorly. Neurals 7 and 8 are smaller and more irregular. Neural 7 is six sided, but the two lateral sides converge as in the other neurals. Neurals 7 and 8 occupy the area between costals 7 and 8. A complete neural series is not typical of Pelomedusoides. In most species the posterior costals intervene between the seventh or eighth neural and the suprapygal. The triangular suprapygal contacts the last neural.

The eight costals of Cearachelys are similar to those in Bothremys and Taphrosphys as well as those in FR 4922. The 11 peripherals are also similar to the other bothremydids, being wider posteriorly. The carapacial scales of Cearachelys are quite similar to the generalized condition for pelomedusoids seen in Bothremys , Taphrosphys , and Podocnemis . A cervical scale is absent as in all Pelomedusoides. Because of a complete neural series, the sulcus between vertebrals 4 and 5 falls on the seventh neural in TUTg 1798. As expected, the scales in Cearachelys differ from the unique condition in Araripemys , in which the first vertebral enters the nuchal emargination and the first two marginals are widely separated.

PLASTRON

The plastron in Cearachelys is known in both specimens. TUTg 1798 is nearly complete, with all sutures and sulci preserved. The MPSC specimen is missing some of the anterior edges of the plastron and the posterior margins of the xiphiplastra are broken off.

The plastron of Cearachelys has a broad, semicircular anterior lobe and a tapering posterior lobe with a shallow xiphiplastral notch. The anterior lobe in Cearachelys is much broader than in FR 4922 and Araripemys . It agrees with FR 4922 and most pleurodires in being rounded and differs strongly from Araripemys , which is pointed. The epiplastra in Cearachelys meet on the midline for a length that is much more than in Araripemys , but less than in FR 4922. The entoplastron in Cearachelys is trapezoidal, not V­shaped as in Araripemys , and it does not have a curved posterior margin as in FR 4922.

Paired, laterally placed mesoplastra are present in Cearachelys in contrast to Araripemys , which lacks them, and in common with FR 4922. The mesoplastra of Cearachelys are similar in size and shape to Podocnemis , Taphrosphys and Bothremys . The axillary and inguinal buttress attachments are not visible in any of the specimens at their current stage of preparation. The xiphiplastron has a moderate posterior projection and a shallow xiphiplastral notch, much as in ent de Broin and Werner, 1998) on the basis Bothremys , but in contrast to the pointed pro­ of these characters: (1) triangular triturating jections and C­shaped notch in Taphrosphys . surfaces and (2) supraoccipital­quadrate con­ The plastral scales in Cearachelys are tact.

much as in other pelomedusoids. The inter­ The other Bothremys Group taxa differ gular is elongate and roughly V­shaped. It from Cearachelys and are united by these extends onto the entoplastron, partially sep­ characters: (1) broad preorbital portion of arating the humerals as in Bothremys and Po­ skull, (2) triturating surfaces very wide, (3) docnemis. The intergular extends onto the palatine extensively exposed in triturating entoplastron slightly more than in FR 4922, surfaces, and (4) supraoccipital­quadrate but not as much as in Taphrosphys , in which contact extensive.

the intergular is large and completely sepa­ Some contradictory characters concerning rates the humerals. The pectoral­abdominal Cearachelys are the exposure of the jugal in sulcus crosses the anterior part of the meso­ the palate in Cearachelys , Rosasia , and Bothplastron as in Bothremys and pelomedusids, remys, but not in Foxemys , Polysternon , and but unlike Podocnemididae . The other plas­ Zolhafah ; the open foramen jugulare postertral scales are very similar to Bothremys . ius in Cearachelys , Foxemys , and Polysternon (presumed to be primitive), but the RELATIONSHIPS closed condition in all the other bothremydids, in which the character is determinable. Cearachelys is a pleurodire because it has

these synapomorphies of the group listed by ACKNOWLEDGMENTS

Gaffney and Meylan (1988) as diagnostic for

the Pleurodira : (1) processus trochlearis pter­ We wish to thank our associates in pleuygoidei present, (2) quadrate process below rodiran studies, Peter Meylan, Roger Wood, cranio­quadrate space, (3) epipterygoid ab­ and Haiyan Tong, for their support and counsent, (4) foramen palatinum posterius behind sel on this project. John Maisey shared his orbit, and (5) pelvis suturally attached to car­ large fund of Santana experience with us. apace and plastron. The sixth character, hyo­ Amy Davidson expertly prepared the very mandibular nerve in its own canal, cannot be fragile skulls. Ed Heck ( figs. 2 , 4 ) and Frank determined in the articulated skull. It is a Ippolito ( figs. 1 , 3 , 5 ) did the skull illustramember of the Pelomedusoides (sensu Broin, tions with their customary excellence. We are 1988; Lapparent de Broin and Werner, 1998; grateful to Lauren Redniss for the high qual­ Meylan, 1996; Tong et al., 1998), which is ity shell figures ( figs. 6–9 ). We particularly equivalent to the Pelomedusidae in the clas­ appreciate the help of Judy Galkin in the prosical sense (sensu Gaffney and Meylan, duction of this paper. Placido Cidade Nuvens 1988) because it has these characters: (1) cer­ greatly assisted in obtaining and studying the vical scale absent, (2) nasal absent, (3) pre­ type specimen.

frontals meeting on midline, and (4) splenial

absent (lower jaws present in TUTg 1798). REFERENCES

Cearachelys can be identified as a member Antunes, M. T., and F. de Broin

of the family Bothremydidae based on its 1988. Le Crétacé terminal de Beira Litoral, possession of the following characters: (1) Portugal: remarques stratigraphiques et precolumellar fossa absent, (2) occipital con­ écologiques, étude complémentaire de dyle consisting only of exoccipitals (reversal Rosasia soutoi (Chelonii, Bothremydiin Pelusios and Pelomedusa ), (3) foramen dae). Ciênc. Terra 9: 153–200.

stapedio­temporale facing anteriorly, (4) fe­ Baur, G.

nestra postotica reduced in size relative to 1891. Notes on some little known American fossil tortoises. Proc. Acad. Nat. Sci. pelomedusids, chelids, and podocnemidids,

Philadelphia 43: 411–430.

and (5) exoccipital­quadrate contact. Within Bergounioux, F.­M., and F. Crouzel

the Bothremydidae , Cearachelys can be al­ 1968. Deux tortues fossiles d’Afrique. Bull. lied with Bothremys , Rosasia , Foxemys , and Soc. Hist. Nat. Toulouse 104: 1–2, Zolhafah , (the Bothremys Group of Lappar­ 179–186, 3 figs.

Broin, F. de

1988. Les tortues et le Gondwana. Examen des rapports entre le fractionnement du Gondwana au Crétacé et la dispersion géographique des tortues pleurodires à partir du Crétace´. Stud. Palaeocheloniol. 2(5): 103–142.

Cope, E. D.

1864. On the limits and relations of the raniformes. Proc. Acad. Nat. Sci. Philadelphia 16: 181–183.

1868. On the origin of genera. Proc. Acad. Nat. Sci. Philadelphia 20: 242–300.

Gaffney, E. S.

1975. A revision of the side­necked turtle Taphrosphys sulcatus (Leidy) from the Cretaceous of New Jersey. Am. Mus. Novitates 2571: 24 pp .

1977. An endocranial cast of the side­necked turtle, Bothremys , with a new reconstruction of the palate. Am. Mus. Novitates 2639: 12 pp.

1979. Comparative cranial morphology of recent and fossil turtles. Bull. Am. Mus. Nat. Hist. 164(2): 65–376.

Gaffney, E. S., and P. A. Meylan

1988. A phylogeny of turtles. In M. J. Benton (ed.), The phylogeny and classification of the tetrapods. Vol. 1, Amphibians, reptiles, birds. Syst. Assoc. Spec. Vol. 35A: 157–219.

1991. Primitive pelomedusid turtle. In J. G. Maisey (ed.), Santana fossils: an illustrated atlas: 335–339. Neptune NJ: T.F.H. Publications.

Gaffney, E. S., and R. Zangerl

1968. A revision of the chelonian genus Bothremys ( Pleurodira : Pelomedusidae ). Fieldiana Geol. 16: 193–239.

Hirayama, R.

1998. Oldest known sea turtle. Nature 392: 705–708.

Lapparent de Broin, F. de, and C. Werner

1998. New late Cretaceous turtles from the western desert, Egypt. Ann. Paléontol. 84(2): 131–214.

Leidy, J.

1865. Memoir on the extinct reptiles of the Cretaceous formations of the United States. Smithson. Contrib. Knowledge 14(6): 1–135.

Linnaeus, C.

1758. Systema naturae. 10th ed., vol. 1. Stockholm, 824 pp.

Maisey, J. G. (ed.)

1991. Santana fossils: an illustrated atlas. Neptune NJ: T.F.H. Publications.

Meylan, P. A.

1996. Skeletal morphology and relationships of the Early Cretaceous side­necked turtle, Araripemys barretoi ( Testudines: Pelomedusoides: Araripemydidae ), from the Santana Formation of Brazil. J. Vertebr. Paleontol. 16(1): 20–33.

Tong, H., and E. Buffetaut

1996. A new genus and species of pleurodiran turtle from the Cretaceous of southern Morocco. Neues Jahrb. Geol. Paläontol. Abh. 199(1): 133–150.

Tong, H., E. S. Gaffney, and E. Buffetaut

1998. Foxemys , a new side­necked turtle ( Bothremydidae : Pelomedusoides) from the Late Cretaceous of France. Am. Mus. Novitates 3251: 19 pp.

Recent issues of the Novitates may be purchased from the Museum. Lists of back issues of the Novitates and Bulletin published during the last five years are available at World Wide Web site http://nimidi.amnh.org. Or address mail orders to: American Museum of Natural History Library, Central Park West at 79th St., New York, NY 10024. TEL: (212) 769­5545. FAX: (212) 769­ 5009. E­MAIL: [e-mail]

Figures

Fig. 9. Cearachelys placidoi, n. gen. & sp., TUTg 1798, dorsal (left), ventral (right), and lateral

Fig. 2. Cearachelys placidoi, n. gen. & sp., MPSC specimen, holotype, Santana Formation, Santana do Cariri, Brazil. A, dorsal; B, ventral; C, right lateral; D, anterior; E, left lateral; F, posterior. See fig. 3 for key.

Fig. 4. Cearachelys placidoi, n. gen. & sp., TUTg 1798, Santana Formation, Santana do Cariri, Brazil. A, dorsal; B, ventral; C, right lateral; D, anterior; E, left lateral; F, posterior. See fig. 5 for key.

Fig. 1. Cearachelys placidoi, n. gen. & sp., Santana Formation, early Cretaceous, Ceará State, Brazil. Restoration of skull based on MPSC specimen and TUTg 1798. Dorsal (left), ventral (right), and lateral (center) views.

Fig. 3. Key to fig. 2.

Fig. 5. Key to fig. 4.

Fig. 6. Cearachelys placidoi, n. gen. & sp., MPSC specimen, holotype, shell in ventral (upper) and dorsal (lower) views. (See fig. 7).

Fig. 7. Cearachelys placidoi, n. gen. & sp., MPSC specimen, holotype, shell in ventral view.

Fig. 8. Cearachelys placidoi, n. gen. & sp., TUTg 1798, dorsal (left), ventral (right), and lateral