Tasmaniosaurus triassicus is a small-sized basal archosauromorph (skull length approximately 16 cm; based on a linear regression between skull and dentary length for the South African proterosuchid Proterosuchus fergusi, n = 11, R 2 = 0.99) differentiated from other members of the clade by the following unique combination of characters (see below): premaxilla with posterodorsally oriented posterior process and ankylothecodont tooth implantation; maxilla with anteroposteriorly short anterior process; frontal with almost straight lateral margin; pterygoid with medial row of palatal teeth (modified from Camp & Banks [19] ); dorsal vertebrae with paradiapophyseal and prezygodiapophyseal laminae and without distinct distal expansion of the neural spine; probable absence of osteoderms (Camp & Banks [19] ); and interclavicle with a diamond-shaped anterior end and a gracile and slightly transversely expanded posterior process.

UTGD 54655, partial skeleton, mostly disarticulated, composed of the following elements: right premaxilla; left maxilla; probable right maxilla; right lacrimal; both frontals, postfrontals and parietals; interparietal; ?supraoccipital; right pterygoid; ?epipterygoid; both dentaries; left splenial; one cervico-dorsal and one anterior or middle dorsal vertebra; fourteen to sixteen caudal vertebrae; several ribs, gastralia and haemal arches; interclavicle; ?femur; both tibiae; and multiple metatarsals and pedal phalanges ( Table 1 ). The different blocks that composes the holotype of Tasmaniosaurus triassicus are currently assembled within a plaster slab mount; however, the arrangement of individual blocks in this slab does not necessary reflect their actual original arrangement ( Fig. 2 ). The artificial assembly of the different blocks in the slab mount is particularly evident in the case of the left maxilla. This bone is situated directly below its natural mould, and is thus artificially separated into two separate blocks in the slab mount but would actually have been a single block in situ. The vast majority of the bones possess the same kind of preservation, are of congruent size, there is no evidence of duplicate elements and all possess morphology congruent with that of a basal archosauromorph. Accordingly, these lines of evidence support the interpretation that almost all the elements included within UTGD 54655 belong to a single individual. However, the holotype of Tasmaniosaurus triassicus is mixed with, at least, an isolated maxilla of a considerably smaller animal [19] . Positive evidence could not be recognised for the presence of gut contents (contra Thulborn [20] ; see below).

A , right premaxilla in lateral view; B , left maxilla in probable medial view; C , partial right pterygoid and D , right lacrimal in medial views; E , skull roof elements in ventral view and possible supraoccipital and epipterygoid; F , anterior end of left dentary and G , left splenial in lateral views; and H , right dentary in medial view. Dotted areas are bone impressions, light grey areas are damaged bone, and dark grey areas are reconstructed bone. The light grey line in (A) is the reconstructed ventral mrgin of the posterior process of the premaxilla inferred from the slope of the anterodorsal margin of the left maxilla. Abbreviations: ?ep, possible epipterygoid; ?ms, possible medial shelf; ?po, probable postorbital; ?so, possible supraoccipital; abaf, anterior border of the antorbital fenestra; anp, anterior process; ap, ascending process; at, anterior tooth; ati, ankylothecodont tooth implantation; chi, cerebral hemisphere impression; cp, central posterior process of dentary; dvsi, dural venous sinus impression; f-n, fronto-nasal suture; hp, horizontal process; fo, fossa; ihfi, interhemispheral fissure impression; ip, interparietal; lff, laterosphenoid facet; li, lateral impression; mbsf, medial border of the supratemporal fenestra; Mc, Meckelian canal; nvf, neurovascular foramina; obi, olfactory bulb impression; od, orbital depression; oti, olfactory tract impression; pb, premaxillary body; pdf, posterodorsal flange; pdp, posterodorsal process of dentary; plp, posterolateral process of the parietal; pof, postfrontal; r, ridge; pp, posterior process; pt, premaxillary tooth; te, teeth; to, tooth; tu, tuberosity; sh, shelf; sy, symphysis; vt, ventral tuberosity. Scale bar equals 1 cm.

Several of the bones that comprise the holotype of Tasmaniosaurus triassicus suffered strong post-mortem compression and, in a number of cases, are currently covered by a dense layer of lacquer that prevents assessment of detailed anatomy and natural borders (e.g. right premaxilla). First hand study of the holotype of Tasmaniosaurus triassicus in August 2012 allowed a reassessment of several misinterpretations of bone identities made by Camp & Banks [19] and/or Thulborn [20] ( Table 1 ), as well as the recognition of some structures and features that were overlooked by previous researchers. Moulds were made from several of the bones and allowed the recognition of further anatomical details, such as details of the ventral surface of the skull roof (i.e. cranial endocast).

Cranium

Premaxilla. The preserved portion of the right premaxilla is exposed in lateral view and mostly congruent with the drawing of Thulborn ([20]: fig. 4a). That author corrected most of the original misinterpretations of Camp & Banks [19] (e.g. a reported premaxillary tooth count of 16) (Figs. 3A, 4A, B; Table 2). However, the anterior region of the premaxillary body is broken and anteromedially displaced from the rest of the bone. The drawing of Thulborn ([20]: fig. 4a) does not show this anteromedial displacement of the anterior end of the bone, and as a result, the premaxillary body appears artificially anteroposteriorly shorter in lateral view in his illustration than it would have been in life (Fig. 3A). Thus, the anteroposterior length of the premaxillary body of Tasmaniosaurus triassicus exceeded 2.21 times its dorsoventral height, resembling the condition present in Protorosaurus speneri (USNM 442453 cast of NMK S 180: ratio 2.59), Prolacerta broomi (BP/1/471: ratio 3.80), Archosaurus rossicus (PIN 1100/55: ratio 3.72), Proterosuchus africanus (RC 59: ratio 3.50; SAM-PK-11208: ratio 3.19; BP/1/3993: ratio 3.03; TM 201: ratio 3.03), Sarmatosuchus otschevi (PIN 2865/68: ratio 2.29) and Euparkeria capensis (UMZC T6921: ratio 2.61). By contrast, the premaxillary body of the erythrosuchids Erythrosuchus africanus (BP/1/5207 ratio 1.50; BP/1/4526: ratio 1.65), Shansisuchus shansisuchus ([46]: figs. 8, 9: ratio 1.07–1.33) and Garjainia prima (PIN 2394/5: ratio 1.82) is considerably anteroposteriorly shorter in comparison with its dorsoventral height. The anterior margin of the premaxillary body is rounded and does not form a distinct acute angle with the alveolar margin (contra Thulborn [19]: fig. 4a). The condition of Tasmaniosaurus triassicus resembles that of Prolacerta broomi (BP/1/471), Proterosuchus fergusi (RC 59, BP/1/3993), Erythrosuchus africanus (BP/1/4526, 5207; NHMUK R3592), Garjainia prima (PIN 2394/5), Euparkeria capensis (SAM-PK-5867) and proterochampsids (e.g. Chanaresuchus bonapartei: MCZ 4037; PULR 07; Gualosuchus reigi: PULR 05). The lateral surface of the premaxillary body is convex, but due to the presence of a thick layer of lacquer it is impossible to assess more details of its anatomy (e.g. presence of neurovascular foramina). The postnarial process ( = maxillary process or posterior process) of the premaxilla is partially preserved but is not in direct contact with the premaxillary body (Fig. 3A: pp; Fig. 4A: pp). However, it seems to be preserved in its original position with respect to the rest of the bone and its base was probably damaged during exposure of the fossil. This process is anteroposteriorly elongated and dorsoventrally tall, being subequal to the dorsoventral height of the preamxillary body, as also occurs in Proterosuchus fergusi (BP/1/3993; RC 59, 96; TM 201), “Chasmatosaurus” yuani (IVPP V90002, [47]), Archosaurus rossicus (PIN 1100/55), Sarmatosuchus otschevi (PIN 2865/68), Erythrosuchus africanus (BP/1/5207), Garjainia prima (PIN 2394/5) and Euparkeria capensis [48]. By contrast, the postnarial process is considerably lower than the height of the premaxillary body in Protorosaurus speneri [4], Prolacerta broomi (BP/1/471), Fugusuchus hejiapanensis ([49]: fig. 22), Shansisuchus shansisuchus (IVPP V2505, [46]) and proterochampsids (e.g. Chanaresuchus bonapartei: MCZ 4037; PULR 07; Gualosuchus reigi: PULR 05). The preserved portion of the premaxilla and the slope of the anterodorsal margin of the maxilla indicate that if the long axis of the main body of the premaxilla is placed in a horizontal orientation, the postnarial process of the premaxilla would be posterodorsally oriented relative to the main body, resembling the condition of Prolacerta broomi (BP/1/471, [50]), Erythrosuchus africanus (BP/1/4526, 5207) and Garjainia prima (PIN 2394/5). By contrast, the postnarial process of small-sized Proterosuchus fergusi specimens (RC 59) is orientated parallel to the alveolar margin, and would be directed posteriorly if the main body of the premaxilla is held horizontally. Finally, in medium to large-sized Proterosuchus fergusi specimens (BP/1/3993; SAM-PK-11208; TM 201), Archosaurus rossicus (PIN 1100/55), “Chasmatosaurus” yuani (IVPP V90002, V4067) and Sarmatosuchus otschevi (PIN 2865/68) the postnarial process is directed posteroventrally if the main body of the premaxilla is held horizontally. The orientation of the postnarial process and the slope of the anterodorsal margin of the maxilla seem to indicate that the premaxilla of Tasmaniosaurus triassicus did not possess the extreme downturning observed in the above mentioned species (Fig. 5). Conversely, those features suggest that the premaxilla of Tasmaniosaurus triassicus would have been only slightly downturned, probably resembling the condition of Prolacerta broomi [50], Garjainia prima (PIN 2394/5) and Erythrosuchus africanus (BP/1/5207). The prenarial process ( = nasal process, ascending process) of the premaxilla is completely missing and the ventral border of the external naris could not be distinguished. The palatal process of the premaxilla is either not preserved or not exposed. The right premaxilla preserves three teeth in situ, one situated at the anterior end (in the anteromedially displaced anterior portion of the bone) (Fig. 3A: pt; Fig. 4A) and two positioned at the mid-length of the premaxillary body, at the posterior end of the alveolar margin. The most anterior tooth only preserves the base of its crown, whereas the two more posterior teeth possess completely preserved crowns (Fig. 4B). The anteromedially displaced anterior portion of the premaxillary body bears the partial crown and has room for another tooth position. In the main fragment of the premaxillary body, the probable presence of four or five tooth positions is estimated. Accordingly, the premaxilla of Tasmaniosaurus triassicus may have possessed a total of six or seven alveoli (Fig. 5A), resembling the condition of medium to large-sized Proterosuchus fergusi specimens (BP/1/3993; SAM-PK-K140; TM 201), Sarmatosuchus otschevi [51] and Chanaresuchus bonapartei [52]. By contrast, in small Proterosuchus fergusi specimens (RC 59), Prolacerta broomi [50] Shansisuchus shansisuchus [46], Garjainia prima (PIN 2394/5) and Erythrosuchus africanus [53] the premaxilla possesses five tooth positions. In the most posteriorly preserved tooth of UTGD 54655, the crown is fused to the alveolar margin of the premaxilla via thin bony ridges (Fig. 3A:ati; Fig. 4B: ati), indicating the presence of an ankylothecodont tooth implantation, as also occurs in Teraterpeton hrynewichorum [54], Prolacerta broomi [50], Proterosuchus fergusi (BSPG 1934-VIII-514; RC 59; SAM-PK-11208; TM 201), and some teeth of Azendohsaurus madagaskariensis (UA 8-7-98-284) and Garjainia triplicostata (PIN 951/63). This crown also possesses denticles on at least its distal margin (Fig. 4B: dd), but evidence for denticles on the mesial margin could not be recognised. However, the mesial denticles of the premaxillary teeth of several basal archosauriforms are very small and restricted to the apical half of the crown (e.g. Sarmatosuchus otschevi: PIN 2865/68) and, as a result, the presence or absence of mesial denticles cannot be confidently assessed in Tasmaniosaurus triassicus because of the presence of a thick layer of lacquer covering the tooth and a poor state of preservation. The denticles are subrectangular in labial view and perpendicular to the main axis of the crown (Fig. 4B), as usually occurs in carnivorous archosauriforms. By contrast, the teeth of non-archosauriform diapsids are completely devoid of mesial or distal denticles (e.g. Youngina capensis: GHG K106; Protorosaurus speneri: [4]; Macrocnemus bassanii: PIMUZ T4822; Prolacerta broomi: BP/1/471). Both complete crowns are labiolingually compressed and slightly distally curved (with a convex mesial margin and a concave distal margin of the crown in labial view) without evidence of enamel ornamentation or ridges on their labial surfaces.

Lacrimal. Camp & Banks [19] originally interpreted this bone as a left quadratojugal. Subsequently, Thulborn [20] noted clear discrepancies between this bone and archosauromorph quadratojugals and interpreted it as a probable composite element formed by parts of two or more gastralia. However, the element is composed of a single bone (contra Thulborn [20]). This bone is alternatively identified here as a right lacrimal exposed in medial view (Figs. 3D, 5; Table 2). The overall shape of the lacrimal closely resembles that of small to medium-sized specimens of Proterosuchus fergusi (BP/1/4016; SAM-PK-11208, K10603), with an angle between the anterior and ventral processes slightly higher than 90° and a similar shape of the concavity formed by the posterodorsal border of the antorbital fenestra. Furthermore, the size of the bone with respect to the premaxilla, maxillae, skull roof and mandibular bones is completely congruent with this interpretation (Figs. 3, 5). In particular, the lacrimal of Tasmaniosaurus triassicus is interpreted to be exposed in medial view because of the presence of a deep, well-defined fossa along the posterodorsal border of the antorbital fenestra, which is principally developed along the proximal half of the anterior process (Fig. 3D: fo; Fig. 6: fo), closely resembling the condition observed on the medial surface of the lacrimal of Proterosuchus fergusi (BP/1/4016; SAM-PK-11208). By contrast, the depression on the lateral surface of the lacrimal of Proterosuchus fergusi covers a proportionally larger area of the bone (BSPG 1934-VIII-514; GHG 231; RC 96), contrasting with the surface exposed on the lacrimal of Tasmaniosaurus triassicus. The lacrimal lacks the distal end of the anterior process and most of the distal end of the ventral process. The ventral process has a damaged posterior margin ( = orbital margin). The lacrimal of Tasmaniosaurus triassicus is an inverted L-shaped bone, resembling the condition of Proterosuchus fergusi (BSPG 1934-VIII-514; SAM-PK-11208, K10603), Fugusuchus hejiapanensis ([49]: fig. 22), Garjainia prima (PIN 2394/5) and Erythrosuchus africanus (BP/1/5207). By contrast, in the lacrimal of Euparkeria capensis (SAM-PK-5867) the anterior and ventral processes merge more smoothly into one another, and in non-archosauriform archosauromorphs the lacrimal is a sub-triangular bone (e.g. Protorosaurus speneri: [4]; Trilophosaurus buettneri: [55]; Youngina capensis: [59]; Prolacerta broomi: BP/1/471). The preserved portions of the anterior and ventral processes of the lacrimal are subequal in length, but because of the damaged ends it is not possible to determinate if the anterior process was longer than the ventral one as in Proterosuchus fergusi (BSPG 1934-VIII-514; BP/1/4016; SAM-PK 11208) and Euparkeria capensis (SAM-PK-5867). The anterior process is straight, transversely thin and tapers slightly towards its distal end (Fig. 6: ap), resembling the condition of some specimens of Proterosuchus fergusi (e.g. BSPG 1934-VIII-514). The posterodorsal corner of the lacrimal is sub-quadrangular and possesses a transversely thin, posteriorly extended flange that probably contacted the ventral process of the prefrontal (Fig. 3D: pdf; Fig. 6: pdf). The poor preservation of the bone surface prevents the identification of an articular facet for the prefrontal in this area. A similar flange, which is variably developed, is also present in some specimens of Proterosuchus fergusi (RC 96; SAM-PK-11208). The main body of the lacrimal (i.e. the portion at which the anterior and ventral processes converge) possesses a centrally placed, medially inflated tuberosity immediately next to the margin of the medial depression (Fig. 3D: tu; Fig. 6: t). This tuberosity disappears at the base of the anterior process but is well developed ventrally on the ventral process, delimiting the posterior border of the medial depression. The tuberosity merges gradually ventrally with the rest of the bone and disappears close to the mid-length of the ventral process. The medial surface of the ventral process posterior to the tuberosity is almost planar. The ventral process (Fig. 6: vp) tapers gradually distally, and due to its damaged posterior and distal margins it cannot be assessed if it was posteriorly curved or distally expanded. The ventral process forms the posterior border of the antorbital fenestra and it is slightly anteriorly concave (Fig. 6: pbaf).

Postfrontal. The postfrontal delimits the posterodorsal border of the orbit (Fig. 3E: pof; Fig. 7A: pof), but the bone is not as anteriorly extended onto the ventral surface of the skull roof as in Prolacerta broomi (BP/1/2675). By contrast, the development of the postfrontal of Tasmaniosaurus triassicus is very similar to that of a South African proterosuchid (NM QR 880) and Archosaurus rossicus (PIN 1100/48). The suture between the right postfrontal and the frontal is clear, but its suture with the postorbital is not discernable because the area is damaged. On the left side of the skull roof the sutural contacts of the postfrontal are not preserved. The ventral surface of the postfrontal is concave and contributes to the posterolateral end of the orbital depression. It cannot be confidently determined whether or not in Tasmaniosaurus triassicus the parietal was excluded from contact with the postfrontal, as is the case in a South African proterosuchid (NM QR 880), or if these bones contacted each other, as occurs in Prolacerta broomi (BP/1/2675) and Erythrosuchus africanus (NM QR 1473).

Interparietal. The interparietal of Tasmaniosaurus triassicus is relatively large and firmly sutured to both parietals (Fig. 3E: ip; Fig. 7A: ip), contrasting with the non-archosauriform archosauromorphs Prolacerta broomi, Trilophosaurus buettneri and Mesosuchus broomi, in which the interparietal is absent [1], [50]. The interparietal is a semilunate bone in ventral view, as a result of a posteriorly concave suture with the parietals. The lateral tip of the interparietal contacts the base of the posterolateral process of the parietal, as also occurs in Proterosuchus fergusi (SAM-PK-K10603), Archosaurus rossicus (PIN 1100/48) and Fugusuchus hejiapanensis ([49]: fig. 22). By contrast, in Erythrosuchus africanus the interparietal is more reduced in extent transversely (NM QR 1473). The posterior margin of the interparietal possesses a robust, rounded posterior projection, resembling the condition observed in Proterosuchus fergusi (RC 96; SAM-PK-K10603), Erythrosuchus africanus (BP/1/5207) and Euparkeria capensis (SAM-PK-5867).

Possible supraoccipital. Camp & Banks [19] identified a partial, thin bone situated a few millimetres away from the interparietal as a supraoccipital (Fig. 3E: ?so; Fig. 7A: ?so). Subsequently, Thulborn [20] reinterpreted this bone as the probable end of the partial rib shaft that lies next to the right parietal (Fig. 6A: dr). However, the shape of the bone does not match that of a rib head (capitulum or tuberculum) because it is too wide and planar (Fig. 6A). The size (width of 10.7 mm) of the bone closely resembles that expected for a supraoccipital. Furthermore, the position of the bone is strongly suggestive of a supraoccipital detached from the skull roof during burial. Accordingly, the original interpretation of Camp & Banks [19] is cautiously followed here. The partial supraoccipital is not very informative and no further details can be provided.

Possible epipterygoid. Camp & Banks [19] identified a slit-like bone preserved next to the right postfrontal as a partial epipterygoid (Fig. 3E: ?ep; Fig. 7A: ?ep). Thulborn [20] questioned this assignment, stating that there is no evidence to support the proposed identification. This bone, with a maximum preserved length of 13.4 mm, is too thin to represent a partial cervical rib shaft and too straight and gracile to be a fragment of hyoid. As a result, a possible explanation is that it represents an anteriorly displaced epipterygoid lacking its ventral end (cf. Camp & Banks [19]). However, as pointed out by Thulborn [20], the evidence supporting this interpretation is weak.

Pterygoid. A thin bone bearing some small teeth is preserved next to the possible partial left dentary (the left dentary was identified as the left maxilla by Camp & Banks [19] and Thulborn [20]) (Figs. 3C, 4E, F). Camp & Banks [19] interpreted this tooth-bearing bone as a probable right ectopterygoid, or less likely a vomer or fragment of pterygoid, and Thulborn [20] considered it as an indeterminate element. Camp & Banks [19] described the presence of five tiny teeth along one of the edges of the bone, but Thulborn [20] considered these projections to be misinterpreted needle-marks produced during preparation of the specimen. First hand observation confirmed that the bone does in fact have palatal teeth based on the following lines of evidence: i) the teeth are regularly spaced; ii) the teeth possess almost exactly the same shape and size along the preserved series; iii) the teeth possess a mustard-like to black colour, suggesting an enamel covering, as in the premaxillary, maxillary and dentary tooth crowns; and iv) the teeth are continuous with the bone surface and they are not well-defined depressions in the matrix as would be expected for needle-marks (Fig. 4F). Accordingly, the evidence clearly supports the original interpretation of Camp & Banks [19] instead of the re-interpretation of Thulborn [20]. The preserved portion of the bone can be tracked along a long extension (cf. Camp & Banks [19]: fig. 5m) and seems to be interrupted by an overlying natural mould of a long, curved bone. Thus, the morphology of the bone does not match that of an archosauromorph ectopterygoid because it is too long to represent a medial ( = pterygoid) process and clearly differs from the strongly posteriorly bowed lateral process of an ectopterygoid. Furthermore, the ectopterygoid of archosauromorphs does not bear teeth (e.g. Mesosuchus browni: SAM-PK-6536; Proterosuchus fergusi: RC 59; Euparkeria capensis: [48]). In addition, the morphology of the bone does not match with that of a vomer or palatine because the orientation of the palatal teeth would result in a bone that is too dorsoventrally deep. By contrast, the morphology of the preserved portion of the bone is almost identical to that of the pterygoid of Proterosuchus fergusi (RC 59) and Prolacerta broomi (BP/1/2675). The preserved portion of bone seems not to belong to the lateral process of the pterygoid because in this region the palatal teeth are arranged perpendicular to the main plane of the process ([58]; e.g. Proterosuchus fergusi: SAM-PK-11208). Instead, the palatal teeth are oriented parallel to the main plane of the bone, as occurs in the anterior process of the pterygoid of Prolacerta broomi (BP/1/2675) and Proterosuchus fergusi (RC 59). Furthermore, this bone is laminar and possesses an upraised shelf immediately next to the dentigerous margin (Fig. 3C: sh). This shelf is observed on the medial surface of the anterior process of the pterygoid of Prolacerta broomi (BP/1/2675) and, as a result, the palatal bone of Tasmaniosaurus triassicus likely represents the anterior process of a pterygoid exposed in medial view (Tables 1, 2). The curvature of the palatal teeth indicates that the element represents a right pterygoid if the interpretation that it is exposed in medial view is correct. The dentigerous margin possesses a series of six or, more probably, seven compressed and blade-like palatal teeth, which are interpreted to belong to the T4 row of pterygoid teeth (sensu Welman [58]). The row of palatal teeth should have continued along the non-preserved dentigerous margins of the bone, as is the case in Prolacerta broomi (BP/1/2675) and Proterosuchus fergusi (RC 59). Each tooth has an apicobasal height of 0.7–0.8 mm and the best-preserved teeth are slightly distally curved, as is the case in the T4 teeth of Prolacerta broomi (BP/1/2675) and Proterosuchus fergusi (RC 59). No clear distinction is evident between the teeth and the tooth-bearing bone, implying that the teeth were probably ankylosed to the bone. However, due to the poor state of preservation of the element this interpretation should be considerate tentative. Similarly, due to preservation it could not be ascertained whether additional tooth rows are present on the rest of the anterior process of the pterygoid, such as the T3 row that lies immediately lateral to the T4 row in Proterosuchus fergusi [58]. The bone identified by Camp & Banks ([19]: fig. 5l) as a pterygoid is here interpreted as an indeterminate element, in agreement with Thulborn [20]. Camp & Banks [19] described the presence of at least four, sharply pointed small teeth along the medial and posterior borders of the supposed right pterygoid. However, these palatal teeth seem to be misidentified and instead represent irregular needle-marks produced during preparation. Although Thulborn [20] erroneously suggested the same interpretation for the “ectopterygoid” teeth identified by Camp & Banks [19] (see above), in the case of the supposed pterygoid teeth he did not raise a similar objection and merely stated that the teeth could not be identified, probably because they were concealed by lacquer. The rest of the bone is planar, long, and is, at least partially, still covered by matrix, with a maximum exposed linear dimension of 56.7 mm. Two needle-marks expose part of the covered surface of the bone and artificially appear like tiny, black palatal teeth. The identification of the bone remains elusive.