AMERICAN MUSEUM Novitates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, NY 10024 Number 3395, 29 pp., 9 figures February 19, 2003 A New Specimen of Pinacosaurus grangeri (Dinosauria: Ornithischia) from the Late Cretaceous of Mongolia: Ontogeny and Phylogeny of Ankylosaurs ROBERT V. HILL,! LAWRENCE M. WITMER,? AND MARK A. NORELL? ABSTRACT Here we report the occurrence of a juvenile ankylosaur from the Upper Cretaceous locality Ukhaa Tolgod in southern Mongolia. The locality is well known for its exquisitely preserved theropods, mammals, and squamates, but until now has not yielded diagnostic ankylosaur material, although ankylosaur specimens are common at the site. The new specimen consists of a nearly complete skull with associated mandible and osteoderms that exhibits the following ankylosaurid synapomorphies: two pairs of osteodermal “‘horns”’ projecting from the quad- ratojugals and squamosals; a wide, triangular skull; a premaxillary beak edge that is not continuous with the maxillary tooth row; and absence of premaxillary teeth. We refer the specimen to Pinacosaurus grangeri based on the presence of a large premaxillary sinus, a quadrate not co-ossified with the paroccipital process, and several pairs of accessory openings in the narial region. The new specimen differs from the holotype and other specimens in having five pairs of openings in the narial region, indicating that extensive morphological variability exists in the narial anatomy of P. grangeri. The specimen is identified as a juvenile based on its small size and the incomplete fusion of secondary dermal ossifications to the skull roof, exposing sutural boundaries. Juvenile ankylosaur skulls are rare, but crucial for understanding the basic anatomy of the highly fused and apomorphic adult skull. Morphological data from the new skull and other specimens are ' Ph.D. Candidate, Department of Anatomical Sciences, State University of New York at Stony Brook, Stony Brook, NY 11794-8081. e-mail: [email protected] ? Associate Professor of Anatomy, Department of Biomedical Sciences and College of Osteopathic Medicine, Ohio University, Athens, OH 45701. e-mail: witmerl @ohio.edu 3 Curator and Chair, Division of Paleontology, American Museum of Natural History. e-mail: [email protected] Copyright © American Museum of Natural History 2003 ISSN 0003-0082 2 AMERICAN MUSEUM NOVITATES NO. 3395 added to existing phylogenetic analyses of the Ankylosauria in order to identify diagnostic characters that aid in resolving ankylosaur relationships. The specimen also provides data on the ontogenetic sequence of secondary dermal ossifi- cation. Presence of dermal ossifications covering only the narial region, quadratojugals, and Squamosals supports the hypothesis that osteoderms in these areas appear early in ontogeny. Furthermore, two well-developed osteoderms were found in close apposition but unfused to the ventrolateral edges of the mandible. All known mandibles from adult ankylosaurs exhibit fusion of these elements to the underlying bones. Thus, mandibular osteoderms also appear relatively early but do not become fused until much later. INTRODUCTION osteoderms, and several well-preserved skulls. Several of these specimens have been In the early 1920s, the American Museum found in associations (Jerzykiewicz et al. of Natural History began its Central Asiatic 1993; Norell, personal obs.) suggesting that Expeditions marking the advent of intensive these animals may have been social. Young paleontological research in the Gobi Desert (1935) named P. ninghsiensis from the Late (Andrews, 1932). The exceptionally well- Cretaceous of Ningxia Province, China. This preserved fossils discovered by these early taxon is now considered a junior synonym of expeditions included dinosaur eggs and the P. grangeri, as is Syrmosaurus viminicaudus, skeletal remains of Mesozoic mammals, the- which consists of fragmentary remains de- ropods, ceratopsians, and ankylosaurs. The scribed by Maleev (1952). Recently, Gode- first report of ankylosaurs in Asia was made froit et al. (1999) described a new taxon, Pin- by Gilmore (1933a), who described a partial acosaurus mephistocephalus, from the Late right ilium and caudal centrum too fragmen- Cretaceous of China. This species differs tary to name. Soon afterward, however, Gil- considerably from P. grangeri in characters more (1933b) named Pinacosaurus grangeri, of the narial anatomy and secondary dermal the first new taxon of ankylosaur from Asia, *‘horns’’. based on a skull, osteoderms, and anterior American Museum expeditions to the cervical vertebrae from specimens found at Gobi resumed in 1990, and by 1993 several Bayn Dzak. new fossil localities were discovered, includ- ing the richly fossiliferous Ukhaa Tolgod The last of the Central Asiatic Expeditions took place in 1930, after which date Ameri- (Dashzeveg et al., 1995). To date, Ukhaa Tol- god has produced approximately 1000 spec- can paleontologists would not return to the imens each of mammals and lizards, and Gobi for nearly 60 years (Novacek, 1996; abundant remains of theropods, ceratopsians, Gallenkamp, 2001). Swedish, Canadian, Pol- and ankylosaurs (Gao and Norell, 2000). ish, and Russian scientists, however, contin- Here we report the first diagnostic ankylo- ued to collaborate with their Mongolian and saur material to be described from Ukhaa Chinese colleagues, and their expeditions re- Tolgod. covered several new taxa of ankylosaurs Institutional Abbreviations: AMNH, such as Talarurus (Maleev, 1952), Saichania American Museum of Natural History, New and Tarchia (Maryanska, 1977), Amtosaurus York; DMNH, Denver Museum of Natural (Kurzanov and Tumanova, 1978), Shamo- History; IGM, Institute of Geology, Ulaan saurus (Tumanova, 1983), and Tsagantegia Baatar, Mongolia; IMM, Inner Mongolia (Tumanova, 1993). More recently described Museum, Hohhot; IVPP, Institute for Ver- ankylosaur material from China includes tebrate Paleontology and Paleoanthropology, Shanxia (Barrett et al., 1998), Tianzhenosau- Beijing; PIN, Paleontological Institute, Mos- rus (Pang and Cheng, 1998), and Gobisaurus cow; ZPAL, Institute of Paleobiology (Zak- (Vickaryous et al., 2001a). lad Paleobiolgii) of the Polish Academy of Despite discoveries of new taxa, Pinaco- Sciences, Warsaw. saurus grangeri remains the best known Asian ankylosaur, represented by over 15 GEOLOGICAL SETTING specimens (Coombs and Maryanska, 1990), For over 75 years (Andrews, 1932) ver- including a nearly complete skeleton, trunk tebrate fossils have been collected from Late 2003 HILL ET AL.: NEW SPECIMEN OF PINACOSAURUS GRANGERI S MONGOLIA © Ulaanbaatar Tugrugeen Shireh # =Flaming Cliffs Kheerman Tsav Khulsan @ g @Ukhaa Tolgod CHINA = Bayan Mandahu Beijing Fig. 1. Map of Mongolia showing major fossiliferous localities, including Ukhaa Tolgod, where the new specimen of Pinacosaurus grangeri was found. Modified from Gao and Norell (2000). Cretaceous localities throughout the Nemegt and Norell, 2000). The new specimen (IGM Basin of southern Mongolia. Most of these 1000/1014) was collected at the Camel’s localities can be assigned, based on faunal Humps locality in the floor of a large quarry composition and lithology, either to the Mid- on the eastern flank of this escarpment. dle Campanian Djadokhta Formation or to The precise stratigraphic and temporal po- the Late Campanian Barun Goyot Formation, sition of Ukhaa Tolgod is not known, as this although precise ages are unknown (see Lil- locality yields mammal specimens previous- legraven and McKenna, 1986 and Gao and ly thought to be unique to each of the two Norell, 2000). Among the most productive of major geological formations in the region these localities is Ukhaa Tolgod (“‘brown (Gradzinski et al., 1977). Gao and Norell hills’’; fig. 1). Discovered in 1993, the local- (2000) showed that the faunas cannot be dis- ity has yielded exceptionally well-preserved tinguished on the basis of fossil lizards, the fossils, including small mammals, lizards, most common fossil vertebrates recovered at and abundant dinosaur remains (Dashzeveg these localities. Thus the distinction between et al., 1995). The most spectacular finds in- the Djadokhta and Barun Goyot Formations clude fully articulated skeletons of ovirapto- is blurred. However, Dashzeveg et al. (1995) rids (Clark et al., 2001), some preserved in and Loope et al. (1998) have suggested that birdlike brooding postures (Norell et al., Ukhaa Tolgod resembles most closely the 1995; Clark et al., 1999), or with embryos Flaming Cliffs locality (Bayn Dzak), the type and eggs (Norell et al., 1994). Other dinosaur section of the Djadokhta Formation, in its li- lineages are represented, including dromaeo- thology. Both localities are comprised of bas- saurids, alvarezsaurids, troodontids, ceratop- al massive sandstones, overlain by thick, sians, and ankylosaurs (Norell, 1997, Gao highly fossiliferous aeolian deposits, and fi- - AMERICAN MUSEUM NOVITATES NO. 3395 nally a sequence of fluvial facies consisting specimens; IGM 100/1014; PIN 614, uncat- of interbedded sandstones, mudstones, and aloged PIN specimens; ZPAL MegD-II/1, conglomerates. ZPAL MgD-II/2, ZPAL MgD-II/7, ZPAL MgD-II/9, ZPAL MgD-II/27, ZPAL MgD-II/ SYSTEMATIC PALEONTOLOGY 31, ZPAL MgD-II/32. DINOSAURIA OWEN, 1842 DIAGNosIs: As for genus; Pinacosaurus with squamosal dermal ossifications present ORNITHISCHIA SEELEY, 1887 as weakly developed pyramids, not long THYREOPHORA NOPCSA, 1915 spines as in P. mephistocephalus. (SENSU NORMAN, 1984) DISCUSSION: Abundant material collected ANKYLOSAURIA OSBORN, 1923 by the Sino-Swedish, Polish-Mongolian, Sino-Canadian, and Russian-Mongolian ex- ANKYLOSAURIDAE BROWN, 1908 peditions has been referred to P. grangeri or PINACOSAURUS Gilmore, 1933b Pinacosaurus sp. Most of the cranial material (= Syrmosaurus Maleev, 1952) collected by these teams has represented ju- venile or subadult individuals. All specimens TYPE SPECIES: Pinacosaurus grangeri Gil- of Pinacosaurus feature the distinctive set of more, 1933b. characters found in the holotype, including a INCLUDED SPECIES: Type species and P. highly specialized narial region with at least mephistocephalus Godefroit et al., 1999. three pairs of openings, prominent hornlike AGE AND DISTRIBUTION: ?Late Santonian— protuberances above the orbits, and lack of Late Campanian of Mongolia and People’s secondary dermal ossifications on the exter- Republic of China. nal rostroventral portion of the premaxillary REVISED DIAGNOSIS: Medium-sized anky- beak. The holotype skull, however, also dif- losaurid, attaining approximately 5 m in fers considerably from the numerous sub- length, with skull longer than wide in adult; adult crania that have been referred to the premaxillary beak edge not covered by sec- genus. The skull is longer than wide, with a ondary dermal ossifications; large, rostrally narrow premaxillary beak and palate. Mar- facing nares roofed by osteoderms; premax- yanska (1977) has attributed this condition to illary struts define at least two additional taphonomic processes, but the relatively mild openings in narial region, leading to an ex- distortion of the pterygoids and premaxillae tensive premaxillary sinus; prominent, horn- suggests that the postdepositional reworking like protuberances above orbit formed by la- of this specimen was limited to dorsoventral teralmost supraorbital bone; lacrimal incisure crushing. Nevertheless, skulls subsequently (marked pinching of snout in lacrimal region referred to P. grangeri consistently exhibit so that the orbits and narial region project the typical ankylosaurid condition of a skull laterally beyond the incisure); beak only and premaxillary palate that is wider than slightly wider than distance between caudal- long. most maxillary teeth; quadrate not co-ossi- A narrow skull can be found in more basal fied with the paroccipital process; quadrate thyreophorans such as Emausaurus ernsti, cotyla lying directly below caudal margin of Scelidosaurus harrisonii, and stegosaurs. orbit. The primitive ankylosaurids Tsagantegia longicranialis and Shamosaurus scutatus Pinacosaurus grangeri Gilmore, 1933b also exhibit this condition. The presence of (= P. ninghsiensis Young, 1935; a narrow skull in adult Pinacosaurus may be Syrmosaurus viminicaudus Maleev, 1952) just one indicator of a phylogenetic affinity HoLoTyPpE: AMNH 6523, dorsoventrally with these more primitive genera that has crushed skull and mandible with several as- previously been obscured, as many of the sociated osteoderms, atlas and axis. characters considered diagnostic for the ge- TyPE LocALity: Djadokhta Formation, nus (and indeed, for suprageneric ankylosau- Shabarakh Usu, Mongolia (e.g., ““The Flam- rian taxa) have been scored from juvenile ing Cliffs’’). skulls that do not exhibit all features of the REFERRED SPECIMENS: Uncataloged IVPP adult. In particular, the exceptionally com- 2003 HILL ET AL.: NEW SPECIMEN OF PINACOSAURUS GRANGERI 5 olele| mare pop— sop fm fole)e) Fig. 2. Pinacosaurus grangeri. IGM 100/1014. Dorsal view of skull. See appendix 4 for abbrevia- tions. Scale bar equals 5 cm. plete skull of P. grangeri described by Mar- hibits greater development of dermal sculp- yanska (1971; 1977), ZPAL MgD-II/1, is fre- turing over the nasals, which comprise the quently cited as displaying typical ankylo- rostral half of the skull roof. Fusion of sec- saurid morphology, despite the fact that it ondary dermal ossifications to the more cau- represents a juvenile individual (e.g., Sereno, dal portion of the skull roof has not occurred, 1986; Coombs and Maryanska, 1990). and this allows sutural boundaries to be read- ily identified. The skull was prepared me- DESCRIPTION chanically and the left mandible removed. It GENERAL DESCRIPTION: The new specimen was then subjected to computed tomography (IGM 100/1014) is a nearly complete skull (CT or CAT scanning), allowing assessment with both hemimandibles preserved intact of some structures still located within matrix, along with the predentary. The skull has been although crushing and radiodense mineral subjected to some dorsoventral crushing and deposits complicate interpretation. shearing, and the bones roofing the right an- SKULL ROooF (fig. 2): The frontals are sub- torbital region have been severely broken Square and comprise the central portion of and lost. The skull is slightly larger than the caudal half of the skull roof. They are in ZPAL MgD-II/1 in all dimensions, and it ex- contact with the nasals and prefrontals ros- 6 AMERICAN MUSEUM NOVITATES NO. 3395 Fig.3. Pinacosaurus grangeri. IGM 100/1014. Stereopairs of skull in occipital view. See appendix 4 for abbreviations. Scale bar equals 2 cm. trally, and the supraorbitals (= postfrontals dally to obscure the occiput in dorsal view, of Maryanska, 1971) and postorbitals later- although this character may be exaggerated ally. Unlike ZPAL MegD-II/1, the parietals by the dorsoventral crushing of the specimen. are fused, a condition that has been described OCCIPITAL REGION AND BRAINCASE (fig. 3): previously among ankylosaurs only in Minmi The basioccipital is badly crushed dorsoven- paravertebra (Molnar, 1996), Pinacosaurus trally, and much of its original morphology mephistocephalus (Godefroit et al., 1999), and that of the occipital condyle are difficult and Cedarpelta bilbeyhallorum (Carpenter et to interpret. A saddle-shaped depression on al., 2001); in other ankylosaurids, parietal fu- the underside of the basioccipital is present, sion cannot be assessed due to the sutural as in Nodocephalosaurus (Sullivan, 1999), boundaries being obscured by overlying os- but there is no evidence of a distinct “‘neck”’ teoderms. Because the other sutures have re- encircling the occipital condyle. The condyle mained unfused in this individual, fusion of is composed entirely of the basioccipital, and the parietals appears to be the first step in the the articular surface appears to have been an- extensive fusion of the entire skull roof gled ventrally from the plane of the maxil- known in adult ankylosaurs. The single, lary tooth rows. As in all ankylosaurs, the fused parietal element contacts the squamo- occiput of IGM 100/1014 is low and rect- sals and postorbitals laterally. A few small, angular, and wider transversely than it is unfused osteoderms are present at the rostro- high. The small, subrectangular supraoccipi- lateral corners of the parietal, extending lat- tal articulates suturally with the exoccipitals, erally over the postorbitals. There is no de- and is largely excluded from the foramen velopment of a “‘tabular’’ element as de- magnum. The exoccipitals are high, rostro- scribed by Maryanska (1971). caudally compressed, and appear only loose- The skulls of ZPAL MegD-II/1 and IGM ly associated with the quadrate and squa- 100/1014 share a wide, equilateral triangular mosal. CT scans confirm that the paroccipital outline in dorsal view, with prominent horn- processes are not fused to the squamosal, as like supraorbital protuberances (postsupraor- in Saichania, nor are they fused to the quad- bitals of Maryanska, 1971) and much smaller rates, as in nodosaurids. The exoccipitals osteoderms extending caudolaterally from comprise the lateral borders and most of the the squamosal region. Thus, in both speci- dorsal border of the foramen magnum. Two mens, the widest point of the skull roof is small, rugose tuberosities, like those men- across the supraorbital elements. The dermal tioned by Maryanska (1977), lie on the lat- skull roof does not extend far enough cau- eral border of the foramen magnum, imme- 2003 HILL ET AL.: NEW SPECIMEN OF PINACOSAURUS GRANGERI a Fig. 4. Pinacosaurus grangeri. IGM 100/1014. Stereopairs of narial region in left anterolateral view showing narial apertures and recesses. See appendix 4 for abbreviations. Scale bar equals 2 cm. diately dorsal to the articulation of the ex- Edmontonia, Stegosaurus, and some other occipitals with the basioccipital. dinosaurs. As perhaps in ankylosaurs gener- The foramen for the hypoglossal nerve ally, the rostrolateral wall of the braincase is (C.N. XII) can be seen in the lateral edge of strongly ossified, with a stout laterosphenoid the exoccipital near its base. Immediately broadly contacting the parietal dorsally and rostral to this foramen is another foramen for with an orbitosphenoid in front that contacts the exit of the glossopharyngeal, vagus, and both the parietal and frontal. accessory nerves (C.N. [X—XI). Other details NARIAL REGION (fig. 4): The rostral surface of the braincase can be observed only in the and tomial (ventral) margin of the beak is CT scans. Most features are in agreement devoid of osteoderms, as is the case in most with Maryariska’s (1977) description of Asian ankylosaurids. Long caudolateral ZPAL MgD-I/1, but the intact (i.e., unpre- (maxillary) processes of the premaxillae ob- pared) nature of much of the braincase of scure the most rostral maxillary teeth in lat- IGM 100/1014 allows clarification of some eral view. The dorsal (nasal) processes form attributes with CT. For example, the otic ves- an internarial bar that is visible far caudally tibule opens broadly to the cranial cavity between adjacent osteoderms. CT scans con- suggesting that the medial wall of the inner firm that the internarial bar is prolonged ven- ear 1S poorly ossified and would have been cartilaginous in life; whether this reflects ju- trally as an internasal septum. The CTs reveal venile status is unknown, although an otic occasional gaps in the septum, but these are vestibule that is medially open to varying ex- best attributed to postmortem or postdeposi- tents has been observed by one of us tional damage to the septum, which is paper (L.M.W.) in CT scans of presumably adult thin in some sections. Contribution of the na- 8 AMERICAN MUSEUM NOVITATES NO. 3395 sal bone to this septum appears to be rela- lary) sinus system (Maryanska, 1971, 1977; tively slight. Tumanova, 1987; Coombs and Maryaniska, Comparison of available specimens of 1990; Witmer, 1997a). Godefroit et al. (1999: Pinacosaurus reveals variation in the number 24), however, noted that “‘this opening is... of openings (real or otherwise, see below) in completely closed”? in IMM 96BM3/1. One the narial region. AMNH 6523 and ZPAL of us queried P. Godefroit about the seeming MegD-II/1 each have three apparent apertures incongruity of a “closed opening’’, and he per side, which, for the sake of discussion, replied that indeed structure B is not an open- may be labeled A, B, and C (fig. 5). Godefroit ing at all, but is “‘a depression [rather] than et al. (1999) reported four pairs of openings a true opening” (personal comm. from P. in the holotype of P. mephistocephalus Godefroit to L.M.W., 30 August 2001). (IMM 96BM3/1). The new skull, IGM 100/ Moreover, Godefroit kindly reexamined 1014, clearly has five such structures, which IMM 96BMs3/1 in China and confirmed ‘“‘that are better seen on the left side than on the our [Godefroit et al., 1999] ‘gland opening’ badly damaged right side. Given that most is completely closed, [forming] some kind of authors (e.g., Godefroit et al., 1999: 29) re- cup-shaped depression’’ (personal comm. gard such narial differences as ‘“‘indubitable from P. Godefroit to L.M.W., 7 November specific characters’’, further discussion is 2001). Close inspection of ZPAL MgD-II/1 warranted. failed to reveal conclusive evidence that the Opening A is clearly the entrance to the margins of the “‘opening’’ were not broken, airway, and represents the “‘true’’ bony nos- and the possibility must be entertained that a tril, as recognized by all modern authors thin bony floor may have been present in (Maryanska, 1971, 1977; Tumanova, 1987; structure B and subsequently lost. CT scans Coombs and Maryanska, 1990; Godefroit et of IGM 100/1014 are not as conclusive as al., 1999). Gilmore (1933b) regarded C as might be hoped due to crushing and damage, the airway entrance, but we can confirm but some slices may support the fossa hy- Maryanska’s (1971, 1977) suggestion that pothesis in that a thin wall of bone remains AMNH 6523 preserves A. Opening A in embedded in matrix 6—19 mm medial to the IGM _ 100/1014 is very similar to those in external margin of B. The significance here ZPAL MgD-II/1 and AMNH 6523 in being is that a foramen and a fossa are fundamen- bounded dorsally by two osteoderms (or os- tally different. If B is a foramen, as typically teodermal masses), with A forming an inci- conceived, then it represents an additional sure between the two. Similarly, in all these opening into the premaxillary sinus (with C skulls A is separated from C by a transverse- being the other, see below). If B is a fossa, ly thin, laminar bar of bone that CT scans however, then it implies a soft-tissue struc- show is concave medially. ture residing externally within the nasal ves- Structure B is located caudolaterally fully tibule that may have nothing to do with any within the premaxilla. This is the “gland sinuses. Clearly, CT scans of a well-pre- opening” of Maryarska (1977) and Gode- served but generally unprepared specimen froit et al. (1999). In all known skulls, this are needed. structure has a well-marked fossa around it. Determination of the soft-tissue relations The apparent shape of the structure in ZPAL and hence function of structure B is perhaps MgD-I/1 and IMM 96BM3/1 is more or less moot pending conclusive evidence corrobo- round, whereas in the holotype and IGM rating or refuting its status as a fossa. Al- 100/1014 it is larger and more rostrocaudally though Maryanska (1971, 1977) regarded it elongate. Structure B in IGM 100/1014 re- as an aperture, her proposed function for it— sembles AMNH 6523 and ZPAL MgD-I/1 housing a nasal salt gland—is actually more in being located caudally near the contact consistent with the fossa hypothesis. Osm6ol- with the maxilla, whereas IMM 96BM3/I1 is ska (1979; see also Tumanova, 1987) ac- unique in being more rostrally located. cepted this gland hypothesis and extended it Previous workers have regarded B as an beyond Pinacosaurus to many other dino- aperture, that is, an opening leading inter- saurs. Extant archosaurs and other sauropsids nally into the premaxillary (or even maxil- indeed possess a nasal gland that often is in- 2003 HILL ET AL.: NEW SPECIMEN OF PINACOSAURUS GRANGERI 9 Fig. 5. Pinacosaurus grangeri. Narial regions in left anterodorsolateral view showing the diversity and proposed homologies of the narial apertures and recesses. A, AMNH 6523. B, ZPAL MgD-II/1. C, IGM 100/1014. See appendix 4 for abbreviations. Scale bars equal 2 cm. 10 AMERICAN MUSEUM NOVITATES NO. 3395 volved in electrolyte balance and salt excre- opens broadly into the sinus. The variation tion; according to Witmer (1995, 1997a), the in the number of C openings among the spec- nasal glands of extant archosaurs (birds and imens is curious and may represent species- crocodilians) are homologous, as is their po- level differences. However, given their open- sition well caudal to the nasal vestibule and ing into a presumably pneumatic sinus and premaxilla. Thus, there is ample evidence the inherent variability of pneumatic aper- that dinosaurs such as Pinacosaurus had a tures in general (Witmer, 1997a, 1997b), it well-developed nasal gland but that it was may be more likely that these are individual not located in the premaxilla. Of course, it is variants of little systematic significance. A possible that some unknown gland was pre- larger sample is necessary to resolve this is- sent, but this hypothesis is probably untest- sue. able (Witmer, 1997a). A vascular function for As mentioned above, the bone separating structure B is perhaps more likely given that, aperture A from aperture C (or C,) was re- based on CT studies of ZPAL MgD-II/1 and garded by Maryanska (1971) and especially IGM 100/1014, branches of the maxillary ar- Tumanova (1987) as the septomaxilla. The tery open into the structure’s caudal margin vomeronasal (VNO, Jacobson’s) organ (see below). Masses of highly vascular narial would have been ventral to this bony lamina, cavernous (erectile) tissue have been inferred according to Tumanova (1987). As additional for other extinct archosaurs (Witmer, 2001), evidence, both authors cited rostrally placed and thus B may have housed a particularly foramina in the premaxillary palate, which large mass. they attributed to the VNO. We join Gode- Whereas all known skulls referred to Pin- froit et al. (1999) in rejecting this assessment. acosaurus have a single A and B on each First, the septomaxilla is an element that has side, the apertures labeled C in figures 4 and never been reliably identified in any crown- 5 are the most variable in number. AMNH group archosaur. Moreover, our CT scans 6523 (fig. SA) and ZPAL MgD-II/1 (fig. 5B) show complete continuity of the definitive have just a single opening C per side, where- premaxilla with the bone under question, as IMM 96BM3/1 has two and IGM 100/ and, again, the bone is a wall between two 1014 (fig. 5C) has three. This variation ap- spaces, not a bridge. The palatal foramina are pears to be real. Based on CT scans of ZPAL almost certainly nothing more than the ap- MgD-II/1 and IGM 100/1014, we can cor- ertures transmitting the vascular anastomosis roborate Maryanska’s (1977) and Tumano- between the palatal and narial vessels; such va’s (1987) observation that aperture C opens anastomoses (with attendant foramina) are a into a relatively large sinus system within the typical feature of amniotes in general and di- premaxilla. CT scans of IGM 100/1014 re- nosaurs in particular (Witmer, 2001). veals that the “bony bridge’? (Maryanska, Based on CT scans of IGM 100/1014 and 1971: 48) dividing A and C in ZPAL MgD- ZPAL MgD-I/1, we can identify at least two II/1 is actually a thin wall of bone separating major premaxillary spaces on each side, the the nasal airway (into which A opens) from vestibulum nasi (nasal vestibule) and the pre- the sinus (into which C opens); some work- maxillary sinus. Scans of IGM 100/1014 (fig. ers have regarded this “‘bony bridge”’ as the 6) suggest that these two spaces did not com- septomaxilla (but see below). C; in IGM 100/ municate broadly, if at all (except to transmit 1014 is the smallest of the three C apertures nerves and vessels). There are some gaps in and is separated from its contralateral fellow the partitions between the vestibule and si- by a thin median septum of premaxilla. This nus, but it cannot be ruled out that these are septum is incomplete in some CT slices but artifacts of preservation. The nasal vestibule probably was intact in life. C, overlies the comprises the main airway, conducting in- premaxillary sinus rostrally and opens into it spired air from the fleshy nostril through ap- caudally. The aperture labeled C, in figure erture A (the bony nostril) into the main nasal 5C (GM 100/1014) is most comparable to cavity. As mentioned above, two osteoder- aperture C in the other specimens and ap- mal masses cover the nasal vestibule, and pears to be the principal opening into the si- each osteoderm bears a small recess or sinus nus. Aperture C; is more horizontal and also associated with the vestibule. The medial of