Duchesneodus uintensis ( Peterson, 1931 )

Mihlbachler, Matthew C., 2008, Species Taxonomy, Phylogeny, and Biogeography of the Brontotheriidae (Mammalia: Perissodactyla), Bulletin of the American Museum of Natural History 311 (1), pp. 1-475 : 316-329

Duchesneodus uintensis ( Peterson, 1931)

HOLOTYPE: CMNH 11809 , a mandible with right i2–i3, c (partial), p2–m3, left c (partial), and p2–m3.

TYPE LOCALITY: LaPoint Member of the Duchesne River Formation ‘‘ Duchesneodus quarry’’, 11 miles west of Vernal Utah.

AGE: Middle Eocene (Duchesnean land mammal ‘‘age’’).

REFERRED SPECIMENS: (From the LaPoint Member of the Duchesne River Formation ‘‘ Duchesneodus quarry’’, 11 miles west of Vernal Utah) AMNH 32604, a skull missing its right zygomatic arch with right P3–M1 (all partial), M2, M3, and left P4–M3; CMNH 9958, an anterior fragment of a skull with right I2–P3; CMNH 9960, a right maxillary fragment with P1–M1; CMNH 11754, a skull with right I2–M3 and left I1–M3; CMNH 11757, a skull with right P1–M3 and left P1– M3; CMNH 11758, a juvenile skull with right P2, P3, DP4, M1, left C, and P2–M2; CMNH 11759, a skull with right I1–M3 and left I2– M3; CMNH 11760, a juvenile skull with right and left DP2–DP4 and M1–M2; CMNH 11761, a mandible with right i2, p2–m3, left i1–c, and p2–m3; CMNH 11766, an anterior portion of a skull with P3–M1, C, and P3– M3; CMNH 11767, a skull with right I2, P1– M3, left I1, and P1–M3; CMNH 11809, a mandible with right i1, i2, c (partial), p2–m3, left c (partial), and p2–m3; CMNH 11815, a skull with right I1–C, P2, P4–M3, left I2, C, and P4–M3; CMNH 11816, a skull missing the right zygomatic arch with right P3–M3 and left P2–M3; CMNH 11821, a partial left mandibular ramus with m2–m3; CMNH 36294, a partial juvenile skull with no teeth; FMNH PM522, a skull with right C, P3–M3, left C, P2, P3, and M2–M3; LACM 128401, a right C, P3–M3, left C, and P3–M3; (from the LaPoint Member of the Duchesne River Formation) FMNH PM22410, a palate with right C–M3, left I2, and P1–M3; (from the Galisteo Formation, New Mexico) F:AM 108521, a partial skull with right P2–M3 and left M1–M3; F:AM 108529, a nasal bone and horns; (from the Vieja Formation, Presidio County, Texas) FMNH PM396, a skull with right P2–M3 and left P4–M3; FMNH PM136, a skull with complete dentition; FMNH PM401, a partial mandible with right i1–i2 and left i1–i3; FMNH PM163, a partial mandible with right p3–p4, left p1 (partial), and p2–m1.

DIAGNOSIS: Duchesneodus uintensis is a large brontothere with small- to medium-sized, relatively rounded frontonasal horns. The horns are positioned slightly in front of the orbits and are not elevated very high above the orbits. The nasal incision extends posteriorly to the anterior margin of the P4. The anterior rim of the orbit is usually anterior to M1. The nasal process is broad, unelevated, horizontal or slightly downturned, slightly tapered distally, strongly rounded anteriorly, and with shallow and thickened lateral walls. The premaxillomaxillary rostrum deepens posteriorly and it is sealed by a dorsal bony cover. Other cranial characteristics include a saddle-shaped cranium with a conspicuous bony dome on the dorsal surface, a dorsal cranial surface that is moderately constricted posteriorly by parasagittal ridges, laterally bowed zygomatic arches with large lateral swellings, unemarginated posterior nares, curved zygomatic arches, and a ventrally constricted and mediolaterally angled external auditory pseudomeatus. Ventral sphenoidal fossae and postzygomatic processes are absent.

Dentally, Duchesneodus uintensis has two small upper incisors that form a straight row, no postcanine diastema, a complex P1, a distinct P2 metacone, and distinct premolar hypocones on P3–P4. The molars of Duchesneodus uintensis have tall, lingually angled ectolophs with weak labial ribs, and thinned lingual ectoloph enamel with wedge-shaped paracones and metacones. A cingular parastyle shelf is absent. Distinct central molar fossae and anterolingual cingular cusps are present. Paraconules and metalophs are absent. The lower dentition of Duchesneodus uintensis includes three very small lower incisors, a postcanine diastema, a metaconid on p3 and p4 but not on p2, and a p2 talonid that is of similar length to the trigonid. The lower molars have shallow basins and m3 is slender.

Duchesneodus uintensis shares with Dianotitan and Megacerops (sensu Mihlbachler et al., 2004b) the unique combination of a saddle-shaped cranium, lack of a postcanine diastema, and lack of a horseshoe-shaped emargination surrounding the posterior nares. Duchesneodus uintensis has a conspicuous autapomorphic dome on the dorsal surface of its cranium. Megacerops specimens occasionally have a similar dome, but it is much less conspicuous.

DESCRIPTION

SKULL: According to Scott (1945) the Duchesneodus quarry sample includes 23 skulls, and three mandibles. Lucas and Schoch (1989b) reported 11 skulls and three mandibles. I encountered 15 skulls and three mandibles. Most of these are in the CMNH collection, although some have been distributed to other museums (AMNH, FMNH, LACM). The skulls represent individuals of different ontogenetic ages and they are all distorted to varying degrees. The most complete and least distorted of the Duchesneodus quarry skulls is CMNH 11759 (figs. 156a, b and 157), the same specimen that was originally described by Peterson (1931). CMNH 11759 is only slightly crushed dorsoventrally. Other well-preserved skulls from the Duchesneodus quarry skulls are CMNH 11767 (fig. 156c, d) and CMNH 11754 (fig. 158a). In addition to the Duchesneodus quarry sample, a completely undistorted skull from the Vieja Formation of Texas (FMNH PM136) is seen in figures 159–161. These specimens form the basis for the following description of the skull of Duchesneodus uintensis , although additional information on variation from other specimens is provided.

Overall, Duchesneodus uintensis is a large brontothere (table 11), but it is relatively small in comparison to Megacerops (sensu Mihlbachler et al., 2004b) . The frontonasal suture can barely be seen on the lateral side of the bony protuberance of CMNH 11759, but can be seen more clearly on other specimens (e.g., CMNH 11754, not shown). The nasal bone forms the proximal base and anterior side of the horn, while the frontal forms the superorbital pillar and the summit of the horn. The horn of CMNH 11759 is small. It is elliptical at its base, but it is more rounded distally. The summit of the horn is rugose. The horns are widely separated and are positioned immediately in front of the orbits, are not highly elevated above the orbits, and they project slightly laterally. The horns of D. uintensis vary in size, though they tend to be short and do not reach the extreme lengths seen in some specimens of Megacerops (sensu Mihlbachler et al., 2004b) . The horns of CMNH 11767 are the largest and most rugose of the Duchesneodus quarry sample. The horns of FMNH PM136 are similar in size to those of CMNH 11767, although they are taller. In more gracile specimens, such as CMNH 11759, the horns are not elevated high above the orbits, although the horns of the more robust specimens, such as FMNH PM136 and CMNH 11767, tend to be more highly elevated above the orbits.

The nasal incision is relatively shallow in all specimens. In specimens with undistorted faces the posterior margin of the nasal incision does not rise higher than the orbit. The posterior margin of the nasal incision varies in position from the anterior margin of P 4 in CMNH 11759 to above the anterior margin of M 1 in CMNH 11767, although this character in the later specimen may be distorted. FMNH PM136 and other skulls are more consistent with CMNH 11759 in this regard. The orbit of CMNH 11759 is positioned above M1, with the anterior rim of the orbit above the posterolateral root of P4. The orbital position of FMNH PM136 and most other specimens is similar with the anterior orbital rim extending anterior to M1. However, the position of the orbit fluctuates slightly. For instance, in CMNH 11767 the posterolateral root of P4 is anterior to the anterior rim of the orbit, although the face of this specimen is somewhat distorted.

The nasal process is usually about the same length as the premaxillomaxillary rostrum, although in some specimens it can be slightly longer or shorter. From a lateral view the nasal process projects from the skull in a horizontal (CMNH 11759) or slightly downward (CMNH 11767, FMNH PM136) direction. The nasal process also tends to curve slightly downward so that the dorsal surface is convex. The dorsal surface is convex while the ventral surface is concave. However, the nasal process of FMNH PM136 is straighter. A downturned thickened rim of bone forms the sides of the nasal process. Overall, the lateral walls are shallow and continue to become shallower distally. From a dorsal view the nasal processes of CMNH 11759 and CMNH 11767 are relatively long and broad, although that of FMNH PM136 is shorter and narrower. The nasal process tends to taper slightly distally and the distal end is strongly rounded. The anterior edge of the nasal process is slightly downfolded and can be roughened as in FMNH PM136. The nasal bones of CMNH 11759 are fully fused, although in CMNH 11767 and FMNH PM136 the nasal bones are separated at their distal ends, forming a median notch on the anterior margin of the nasal process.

In CMNH 11759 and CMNH 11767 the premaxillomaxillary rostrum curves upward distally and the dorsal margin of the rostrum slopes posterodorsally, but it does not rise higher than the midpoint of the orbit. The rostrum deepens posteriorly. In FMNH PM136, the rostrum is thicker, shorter, and so strongly curved upward that it almost contacts the nasal process, thus greatly constricting the nasal cavity and causing the dorsal surface of the rostrum to slope posteroventrally rather than posterodorsally. The premaxilla of Duchesneodus uintensis is very short and does not extend anterior to the canines. The premaxillary symphysis is fully ossified in all adult specimens. From the oblique view of CMNH 11759 (fig. 157), the symphysis is long and sloped upward. Behind the symphysis the dorsolateral margins of the rostrum are laterally divergent. I have not seen a specimen of Duchesneodus uintensis with a visible premaxillomaxillary suture.

From lateral views the dorsal surface is completely concave except for a bizarre, domelike convexity on the dorsal surface midway between the orbit and the occiput. In CMNH 11759 this convexity is conspicuous, but it is relatively low. It is extremely large in CMNH 117167. However, among most other specimens the size of this dorsal dome is intermediate (e.g., CMNH 11754; FMNH PM136). Scott (1945) thought that this cranial convexity was unique to D. uintensis , but he was incorrect. As Peterson (1931) had noted earlier, a similar but subtle cranial convexity is occasionally present on skulls of Megacerops (sensu Mihlbachler et al., 2004b) . A similar dome is also seen in Metatitan primus , although that species differs from D. uintensis in many other ways. In D. uintensis the size of the dorsal convexity seems to covary with the size of the horns and the thickness of the zygomatic arches (see next paragraph). The parasagittal ridges are prominent, remain widely separated, and weakly overhang the sides of the skull. The posterodorsal surface is only moderately constricted by the zygomatic arches.

The zygomatic arches of Duchesneodus uintensis skulls vary greatly in thickness, and the degree to which they are bowed laterally. Those of CMNH 11759 are relatively shallow and thin. From a lateral view the jugal process of the zygomatic is horizontal, while the squamosal zygomatic process slopes upward posteriorly at a shallow angle, thus giving the zygomatic a weak curvature. From the dorsal view the zygomatic arches are thin and moderately expanded laterally. The lateral swelling at the junction of the jugal and squamosal is distinct but rather small. In comparison, the zygomatic arches of CMNH 11767 and FMNH PM136 are deeper, strongly bowed laterally, have massive swellings, and are more strongly curved. The degree of variability in the size of the zygomatic swellings in Duchesneodus is similar to that seen in Megacerops . Scott (1945) and Lucas and Schoch (1989b) interpreted the notable variation in horn size, zygomatic swelling, and the dorsal dome as sexual dimorphism. This interpretation is consistent with that of Mihlbachler et al. (2004b) for Megacerops , in which the morphometric variation in the horns and zygomatic arches is similar in magnitude to the morphometric variability documented in the sexually dimorphic structures (horns, tusks) of extant ungulates (Mihlbachler, 2004).

From a lateral view of the skull the occiput is strongly tilted backward. From a dorsal view the nuchal crest is strongly concave. From the posterior view (shown for FMNH PM136: fig. 160b), the dorsal margin of the occiput is strongly arched, although it is notched medially. The dorsal portion of the occiput is somewhat wider than the ventral portion. On the surface of the occiput there are prominent occipital pillars and a shallow central depression. Overall, the occiput of Duchesneodus uintensis is not as massive as that of Megacerops .

The anterior rim of the posterior nares is situated slightly anterior to the M3 protocones in CMNH 11759, although the position of the anterior margin of the posterior nares fluctuates among specimens from the Duchesneodus quarry from between the M2 hypocones to between the M3 protocones. There is no horseshoe-shaped rim of bone emarginating the anterior and lateral sides of the posterior nares.

In CMNH 11759 and FMNH PM136, the thin, elongate vomer can clearly be seen bisecting the elongate posterior narial canal. The posterior narial canal extends slightly into the region of the sphenoid, although large ventral sphenoidal fossae are not present in D. uintensis . The foramen ovale and foramen lacerum are widely separate. The mastoid process is short and arches forward, making contact with the posterior surface of the postglenoid process, forming a tube-shaped or ventrally constricted external auditory pseudomeatus.

UPPER DENTITION: Several specimens from the Duchesneodus quarry have well-preserved and lightly worn teeth. Pictured in close-up are the molars and premolars of CMNH 11754 (fig. 158b, c) and the premolars of CMNH 11767 (fig. 158d), CMNH 9960 (fig. 158e), and FMNH PM136 (fig. 161c). The more complete incisors and canines of FMNH PM136 are shown in close-up (fig. 161b).

There are two pairs of upper incisors. The incisors are very small, essentially featureless and globular, and of similar size. They form a straight row between the anterior margins of the canines. The canines are extremely small in FMNH PM136. They are somewhat larger in CMNH 9960, but the canines of all D. uintensis specimens can be characterized as small. With the exception of a median gap between the central incisors (which tends to occur in most brontotheres), there are no diastemata in the upper dentition.

The P1 is small but complex. It is nearly rounded in outline. The ectoloph has two convexities, indicating the presence of both a paracone and metacone. On the small lingual shelf, a lophlike protocone arches around the lingual side of the crown and is connected to a preprotocrista. The P2–P4 of most specimens are essentially rectangular, although the shape of P2 can be somewhat oblique as in CMNH 9960. The parastyle of P2 is straight or slightly lingually arched. The P3 parastyle is straight or somewhat labially directed. The parastyle of P4 is always strongly labially directed. The metastyles of P2 and P3 are straight, while the P4 metastyle can be directed somewhat labially. The labial paracone ribs are weak and become progressively thinner in more posterior premolars. Mesostyles are absent on the premolars of CMNH 11754, but a poorly developed mesostyle is occasionally present on the P4 (e.g., CMNH 9960, CMNH 11757, FMNH PM136).

The lingual morphology of the P2–P4 is highly variable. A protocone and hypocone are always present on these teeth, in addition to very small but distinct preprotocrista. The preprotocrista is largest on P2 and smallest on P4. However, in some specimens, such as CMNH 9960, the preprotocristae are well developed on P2, P3, and P4. On the P2 of CMNH 11754 the protocone is essentially absorbed by a large crest that arches around the anterolingual side of the crown. The hypocone of that tooth is distinct, but it is connected to the lingual crest. The protocones of P3 and P4 of that specimen form more discrete cusps. The hypocones of these teeth are smaller than the protocones and completely separated. The lingual premolar

TABLE 11 Summary statistics for selected morphometric variables of the ‘‘ Duchesneodus uintensis quarry’’ sample See Methods for measurement definitions morphology of CMNH 11767 is similar to CMNH 11754, although the hypocones and protocones are of similar size and the P2 hypocone is more strongly connected to the protocone. Other specimens differ more dramatically. For instance, on CMNH 9960, the protocone and hypocone of P2 and P3 are completely absorbed by the prominent lingual crest. In the same specimen, the lingual cusps of P4 are strongly connected by a tall crest. The degree to which the lingual cusps are connected is intermediate between CMNH 11754 and CMNH 9960 on several specimens from the Duchesneodus quarry (e.g., AMNH 32604); specimens from the quarry do not exhibit distinct anterolingual cingular cusps on the premolars, although some specimens from elsewhere do. For instance, the P3 and P4 of FMNH PM136 display prominent anterolingual cingular cusps. Anterolingual cingular premolar cusps are also frequent among the Duchesneodus californicus (a nomen dubium) material from the Sespe Formation. Occasional anterolingual cingular premolar cusps are also seen among specimens Megacerops (sensu Mihlbachler et al., 2004b) and (more prominently) in Sivatitanops birmanicum (a nomen dubium). The labial cingula of the premolars of D. uintensis are absent to very weak. The lingual premolar cingula are variably continuous and discontinuous around the lingual base of the protocone.

The upper molars of Duchesneodus uintensis have tall, lingually angled ectolophs, very weak labial ribs, thin lingual ectoloph enamel, and wedge-shaped lingual margins of the paracone and metacone. The anterior cingulum is thin and passes proximally to the distal peak of the parastyle. Distinct central molar fossae and small anterolingual cingular cusps are present on the molars. On CMNH 11754 there is a small bead of enamel positioned distally from the anterolingual cingular cusp on all the molars, but generally Duchesneodus uintensis upper molars lack any structures that could be interpreted as paraconules. Metalophs are also absent on M1 and M2. The M3 hypocone morphology varies in size from a small bead of enamel (CMNH 11754) to nearly as large as those of the M1 and M2 (CMNH 11759). The labial molar cingula are always weak or absent and always discontinuous around the mesostyle. Distinct lingual cingula are lacking on the M1 and M2 of CMNH 11754, but a thin, distinct lingual cingulum can be seen on the M3 that is discontinuous around the base of the protocone.

MANDIBLE AND LOWER DENTITION: There are only three mandibles in the Duchesneodus quarry sample and none of these is associated with a skull. One of these, CMNH 11821 (not shown), is poorly preserved and does not contribute a significant amount of morphological information. The holotype mandible (CMNH 11809) appears not to be significantly distorted from a lateral view (fig. 162). However, from a dorsal view it is clear that either the left ramus has been artificially shortened or the right ramus has been artificially elongated. Additionally, the mandible has been slightly crushed transversely. The ascending rami of the third mandible are significantly distorted, but the shape of the symphysis is essentially intact (fig. 163).

The inferior margin of the symphysis of CMNH 11809 is straight and steeply angled (.45 °), while in CMNH 11761 the inferior margin is concave and with a shallower overall angle (, 45 °). In these respects, both specimens contrast with the more normally convex inferior margin of Megacerops and more closely resemble Eubrontotherium clarnoensis . The symphysis of CMNH 11761 is broad and it extends to the anterior margin of m1.

The lower incisors are small and essentially featureless, although this is partially related to wear. The i2 is somewhat larger than i3. The incisor rows of both specimens are incomplete, but the incisor alveolar surface is most complete in CMNH 11761. The incisors form a straight row between the anterior margins of the canines. Although the number of upper incisor is reduced to two pairs, the remaining lower incisors of CMNH 11761 plus the unoccupied incisor alveoli indicate three pairs of lower incisors.

The canines of both specimens are small. The canines of CMNH 11761 are distinctly separated from the cheek teeth by a postcanine diastema that is slightly shorter than the p2. Even if a small p1 were present in that specimen there is still enough space for a short postcanine diastema. The width from canine to canine is much narrower than the width across the p2s. A small alveolus anterior to the p2 indicates the presence of a single-rooted p1.

The heavy wear of the remaining cheek teeth prohibits a thorough description; however, the following details are revealed, particularly from the left cheek-tooth row of CMNH 11809, which seems to be the least worn available cheek-tooth row of Duchesneodus uintensis . The trigonid and talonid of p2 are about the same length. The trigonids of p3 and p4, on the other hand, are shorter than the talonids. The trigonids of p2–p4 are somewhat narrower than the talonids. The paralophid of p2 is arched slightly lingually, while the p2 protolophid appears to have been positioned lingually and directed posteriorly. The p2 trigonid lingual notch is small. The trigonid of p3 is very worn, but its dimensions resemble the p4 trigonid. The paralophid and protolophid of p4 arch fully lingually, creating a deep (and probably broad when unworn) lingual trigonid notch. A metaconid is not seen on p2, and although worn, the shapes of the p3 and p4 trigonids are consistent with those of other species that have a large lingually positioned metaconid. The talonids of p2 and p3 are heavily worn, but their shape suggests there were well-developed cristids obliqua and hypolophids, with nearly molariform basins, like p4. There are no distinct entoconids on the premolars of Duchesneodus uintensis (contra Lucas and Schoch, 1982, 1989b).

The molars are typical with shallow basins, thin lingual enamel, and an elongate m3. The labial cingulids of the premolars and molars are weak, as noted by Lucas and Schoch (1989b) in their diagnosis of D. uintensis , although their weakness may have resulted from heavy wear on the teeth.

REMARKS

In 1929–1930 a Carnegie Museum field party led by J. L. Kay collected a large assemblage of a new species from the ‘‘Titanothere Quarry’’ (‘‘ Duchesneodus Quarry’’ of Lucas and Schoch [1989b]) of the Duchesne River Formation near Vernal, Utah. This new species, Duchesneodus uintensis ( Peterson, 1931) , was described from the limited material that was initially available (two jaws and a skull). It is both surprising and unfortunate that Peterson (1931) assigned holotype status to a mandible (CMNH 11809) that does not facilitate a good diagnosis of this species. A complete skull (CMNH 11759) that does facilitate a clear diagnosis of Duchesneodus uintensis was available to Peterson (1931) but was assigned only paratype status. The holotype jaw cannot be readily differentiated from the holotypes of Protitanops curryi , Notiotitanops mississippiensis , or from Megacerops . Therefore, from a strictly morphological perspective, the taxonomic identity of the holotype jaw is unclear. The skulls of the quarry, on the other hand, clearly indicate a unique species of brontothere. The validity of Duchesneodus uintensis relies, therefore, on the assumption that the quarry sample to which the holotype belongs is a monospecific assemblage.

Peterson’s (1931) initial description of the quarry sample was intended to be preliminary and was not based on the entire assemblage. Scott (1945) subsequently provided a thorough description of the quarry assemblage and documented the morphological variation found within that sample. Rather than splitting the sample into multiple species based on minute differences (some related to taphonomic distortion) that Osborn (1929a) and other prior authors (e.g., Granger and Gregory, 1943) would have regarded as evolutionary stages, Scott (1945) recognized taphonomic variation and differentiated it from biologically relevant variation. Additionally, he readily accepted the real morphological variation in the sample as intraspecific variation. For this reason, Scott’s description of Duchesneodus uintensis was a potentially significant advancement in brontothere taxonomy, although those who practiced brontothere taxonomy in the following decades seem not to have recognized the significance of Scott’s recognition of intraspecific variation in this sample. Duchesneodus uintensis is similar to Megacerops (sensu Mihlbachler et al., 2004b) , but differs most conspicuously in the retention of three lower incisors and the possession of a conspicuous dome on the dorsal surface of the skull.

Peterson (1931) originally assigned Duchesneodus uintensis to the genus Teleodus because it possessed six globular lower incisors like the type species of Teleodus , T. avus Marsh. However, Lucas and Schoch (1982) and Lucas (2004) convincingly argued that the holotype of T. avus Marsh (YPM 10321) has more incisors added to it during the preparation of the specimen than actually belonged with the specimen. For this reason Teleodus was deemed an invalid genus by Lucas and Schoch (1982) and Teleodus uintensis was assigned a new genus, Duchesneodus .

Lucas and Schoch (1982) also transferred other species that had been previously assigned to Teleodus to Duchesneodus : D. primitivus ( Lambe, 1908) , D. californicus ( Stock, 1935) , and D. thyboi ( Bjork, 1967) . In a subsequent publication, Lucas and Schoch (1989b) further revised the genus Duchesneodus . Duchesneodus californicus and D. thyboi were considered junior synonyms of D. uintensis . However, D. primitivus was recognized as valid species. In addition to these taxonomic revisions, Lucas and Schoch (1989b) referred a large number of brontothere specimens to D. uintensis , including material from the Duchesneodus quarry as well as more fragmentary material that has previously been described by Bjork (1967), Nelson et al. (1980), Stock (1935; 1938), Lucas (1982, 1983b), and other material. With the exception of the Duchesneodus quarry material, much of the additional material is fragmentary and lacks diagnostic features. Lucas and Schoch (1989b) did not recognize that much of the very fragmentary material they considered to be D. uintensis is too incomplete to identify to a species level. Much of the material in their hypodigm could belong to other brontotheres such Protitanops curryi or Notiotitanops mississippiensis . Among these materials are the holotypes of D. primitivus , D. californicus , and D. thyboi , which are now considered nomina dubia.

Among the four species assigned to Duchesneodus by Lucas and Schoch (1982, 1989b) only D. uintensis is known from complete fossil material and is clearly valid. The list of referred specimens given here only includes specimens that can be unambiguously assigned to D. uintensis based on morphological characters and/or inclusion in the presumably monospecific quarry sample. Because Duchesneodus (sensu Lucas and Schoch, 1989b) is considered an index taxon of the Duchesnean land mammal ‘‘age’’, it will be important to reconsider the identification of brontothere specimens (mostly fragmentary) attributed to Duchesneodus by Lucas and Schoch (1989b), and the significance of this to the age assignments of various faunas considered by Lucas (1992) to be Duchesnean.

MISCELLANEOUS DUBIOUS, PROBLEMATIC, AND UNNAMED POTENTIAL TAXA

Because fossil species are often based on fragmentary material, it is inevitable that many species names will eventually be determined to be invalid. Although many species are found to be synonyms, others are dubious altogether. Many of the earliest names for brontotheres that were based on very fragmentary materials were found to be dubious by Osborn (1929a). In all of these cases, where Osborn considered a taxon to be dubious, my determinations are in agreement. Therefore those taxa considered dubious by Osborn (1929a) are not reconsidered here. In addition to the valid species and their synonyms that are discussed above, an additional 32 species (or potential species) are discussed below. These 32 taxa fall into three categories: (1) nomina dubia, (2) problematic, and (3) unnamed potential species.

The majority of these taxa (N 5 23) were found to be nomina dubia. An additional eight dubious species were discussed above in prior sections dealing with Metarhinus and Rhinotitan , thus yielding a total of 31 species determined to be dubious in this study. Species considered to be nomina dubia could not be diagnosed as unique species nor could they be synonomized with other taxa. Species were considered dubious for two reasons. Most dubious species are represented by fossils that lack autapomorphies or unique combinations of character states. A smaller number of nomina dubia, including Sivatitanops birmanicum , Hyotitan thomsoni , ‘‘ Protitan ?’’ cingulatus , and Epimanteoceras amplus , exhibit autapomorphies or unique combinations of character states, although synonymy with other fragmentary taxa could not be ruled out due to nonoverlapping parts (e.g., a skull from one taxon and a jaw from another), which prevents comparison.

Three other previously named species, ‘‘ Eotitanops ?’’ dayi , Pakotitanops latidentatus , and Mulkrajanops moghliensis , were considered to be problematic, but not necessarily of dubious taxonomic status. Problematic taxa are seemingly valid according to the methods of taxon delimitation used in this study, but the fossil materials (or available figures and descriptions) are so incomplete, poorly preserved, or (in cases where I could not examine original material or casts) insufficiently described in the literature that the available character data is extremely limited, thus prohibiting a thorough description or taxon diagnosis.

Finally, six potentially new species are considered below. These possibly new species are described, but for a variety of reasons, a formal name and diagnosis was not applied, pending the discovery of more appropriate material. Primarily these taxa are known from very fragmentary materials.

The dubious, problematic, and potentially new species were excluded from the phylogenetic analysis. Nonetheless, many of the specimens representing these taxa are highly significant. For instance, many of the fossils currently assigned to nomina dubia contain valuable character data if future discoveries allow them to be linked with valid species known only from highly fragmentary specimens. Additionally, many of the specimens upon which these dubious, problematic, and potential taxa are based occur in regions where brontothere fossils are rare (e.g., Europe, southeast Asia, Pakistan, Korea, etc.), and thus significantly expand our understanding of brontothere paleobiogeography, and are thus deserving of attention.