Eotitanops, Osborn 1907

Mader, Bryn J., 2010, A species-level revision of the North American brontotheres Eotitanops and Palaeosyops (Mammalia, Perissodactyla), Zootaxa 2339, pp. 1-43 : 4-14

Genus EOTITANOPS Osborn 1907

Age: Early Bridgerian.

Subage: Gardnerbuttean.

Type species: E. borealis ( Cope 1880) .

Included species: E. gregoryi Osborn 1913 (= E. minimus Osborn 1919 ).

Diagnosis: Small-sized (average length P2 to M3 is 91 mm in AMNH 14887) brontothere with a well developed upper canine; long upper and lower diastema; unmolarized premolars; relatively large paraconules; and hypocone or pseudohypocone variably present on M3. Eotitanops is probably distinguished from all other brontotheres by its relatively long face, a plesiomorphic character state that is similar to outgroup perissodactyls. There are no synapomorphies that distinguish Eotitanops from other brontotheres (although all other brontotheres have synapomorphies that distinguish them from Eotitanops ).

Discussion: In 1880 Cope described a maxilla fragment of a small brontothere (AMNH 4892) from the Wind River Basin as a new species of Palaeosyops , P. borealis . Osborn (1897) included P. borealis in the genus " Telmatotherium " ( T. boreale ) but later concluded that it was an entirely new genus, which he named Eotitanops . Osborn first used this new generic name in 1907 but did not formally diagnose the genus until 1908.

In 1929 Osborn gave the following diagnosis for Eotitanops borealis (comparisons are against E. gregoryi and E. brownianus , both discussed below):

“Of larger size, p2–m3 94–98 millimeters; premolar teeth more complicated, as shown in neotype and associated specimens; p2 with very rudimentary paraconid and metastylid; P2–4 with progressively developing tritocones [= metacones] and single internal deuterocones [= protocones] backwardly inclined, crowns subtriangular; M1–3 with distinct protoconules [= paraconules].”

The " neotype " mentioned by Osborn is a fragmentary skull (AMNH 14887, Figure 1 ) and jaws. Despite its poor condition the skull is the most complete cranial remains of Eotitanops known. Osborn designated this specimen as the neotype of E. borealis in 1929, but because the original type (AMNH 4892) is still preserved, Osborn's specimen has no nomenclatural significance (Article 75, International Code of Zoological Nomenclature, Ride et al. 1999). It is merely a referred specimen.

In 1881 Cope described a partial lower jaw (AMNH 4885) from the Wind River Basin, which he identified as a new species of Lambdotherium and named L. brownianum ( Cope 1881) . Osborn (1929), however, correctly recognized that this jaw is actually a specimen of Eotitanops . Osborn accepted E. brownianus (note the emendation of the trivial name) as a valid species and distinguished it from the other species of Eotitanops that he recognized by its relatively small size and simplicity of the premolars.

Expeditions from the American Museum of Natural History collected from the Wind River Basin in 1891 and between 1909 and 1911. In 1913 Osborn wrote a paper based on the material collected by these expeditions in which he named three new species of Eotitanops : E. gregoryi , E. princeps , and E. major .

Eotitanops gregoryi was based on an incomplete lower jaw with some of the right cheek dentition intact and fragments of the left maxilla containing M2 and M3 (AMNH 14889). Osborn (1913) gave the following diagnosis of the taxon:

"Of inferior size. p2 – m3 = [78.4 mm]; m1–3 = [49]; p2–3 with the internal cusps, paraconid and metaconid, consisting of rectigradations of most rudimentary stage; hypoconulid of m3 very small; M3 with a single internal cone, no hypocone".

Measurements in the above diagnosis appearing in brackets were originally reported by Osborn in meters, but have been converted here to millimeters. The term "rectigradation" refers to newly evolved morphological characters upon which the taxon could be defined. In 1929 (p. 291) Osborn repeated the above diagnosis almost verbatim and went on to add that the “primitiveness” of the species is apparent when the p3 is compared to that of Eotitanops borealis and E. princeps . According to Osborn, the p3 of E. gregoryi is laterally compressed; the hypoconid is distinct; and the paraconid, metaconid, and entoconid are all in an extremely rudimentary state. Osborn also stated that the p2 is plesiomorphic, being short and compressed with a very rudimentary hypoconid, and that the molars are plesiomorphic as well. According to Osborn the metastylid and entostylid of the molars are extremely rudimentary and the hypoconulid of the m3 is small, subconic, and externally (= buccally) located.

Eotitanops princeps was based on a lower jaw and partial postcranial skeleton (AMNH 296) that had been originally referred to Palaeosyops borealis by Osborn and Wortman (1892). This specimen, although very incomplete, is the most complete skeleton of Eotitanops known. In 1913 Osborn concluded that this specimen represented a new species of Eotitanops and cited the following as diagnostic characters (all comparisons are against E. gregoryi , E. brownianus , and E. borealis ):

"Of still larger size, p2–m3 [105 mm (estimated)]. Inferior premolar teeth somewhat more complicated, as shown in the type specimen. p2 with elevated, distinct, but very rudimentary paraconid and metaconid; entoconid very rudimentary; talonid narrow. p3, paraconid quite distinct, elevated; metaconid small, distinct; entoconid rudimentary; talonid broad. p4, talonid broad; entoconid distinct. Hypoconulid of m3 rounded, more robust. Ramus larger and more robust. The more advanced development of the premolar rectigradations, the increased size of the teeth and of the jaw, the larger size of the hind feet in the referred specimen (Am. Mus. 4902) [= AMNH 4902] combine to distinguish this specimen as a mutation or subspecific stage between E. borealis and E. major ."

In 1929 Osborn stated that Eotitanops princeps is a valid specific stage of Eotitanops (p. 193) but elsewhere (p. 295) repeated his earlier statement (1913) that E. princeps is a subspecific stage between E. borealis and E. major . In 1929 Osborn repeated his previously published diagnosis for E. princeps (1913) almost word for word but made some important modifications. In the 1929 diagnosis Osborn stated that the metastylid of p2 and p3 is rudimentary or small rather than the metaconid, added that the p2 entoconid is rudimentary if present, stated that the entoconid shelf of p4 is distinct (rather than the entoconid itself), and added the observation that the p4 is submolariform.

Eotitanops major was based on a left median metatarsal (metatarsal III) and the distal end of a tibia (AMNH 14894). Osborn (1913) characterized the species as "ill-defined" and distinguished from other species of Eotitanops by its supposedly larger size. In 1929 Osborn again recognized E. major as a valid species.

To facilitate comparison of Osborn's diagnoses I have compiled in tabular form most of the characters that Osborn used to define the Wind River species of Eotitanops for which there was dental material available ( Table 1). All of the characters that Osborn used in the diagnosis of more than one species are included in the table, but I have omitted some of the characters that Osborn used in the diagnosis of only a single species.

Most of the characters listed in Table 1 show no appreciable differences between taxa. For example, the descriptions of the p2 paraconid as being "very rudimentary" ( Eotitanops princeps and E. borealis ), "very low on crown" ( E. brownianus ), and "extremely rudimentary" ( E. gregoryi ), do not imply any significant difference among the taxa. Similarly, descriptions of the p2 entoconid as being "very rudimentary" ( E. princeps and E. borealis ) or "invisible" (= absent?, E. brownianus ) do not suggest an important difference.

The only characters showing significant differences between two or more of the taxa are the morphology of the p2 hypoconid, p3 paraconid, and m3 hypoconulid. The p2 hypoconid is described as "distinct and elevated" in Eotitanops brownianus but "very rudimentary" in E. gregoryi . The p2 hypoconid on the type of E. brownianus is not particularly distinct or elevated, however, and the entire p2 talonid region is lacking in the type of E. gregoryi . Although Osborn's observation cannot be confirmed, the weak development of the p2 hypoconid in E. brownianus suggests that this difference is insignificant.

The p3 paraconid of Eotitanops princeps is described as "quite distinct and elevated" while that of E. gregoryi is "very rudimentary". Although the p3 paraconid of E. princeps is slightly more distinct than that of E. gregoryi , the p3 paraconid of E. princeps is actually rather poorly developed and, as Wallace (1980) pointed out, is hardly more prominent than the p3 paraconid of AMNH 14888, which Osborn (1929) referred to E. borealis . The small size of the p3 paraconid in the type of E. gregoryi might be diagnostic for the species, but the character variability cannot be assessed without additional material.

Finally, the m3 hypoconulid of Eotitanops princeps is described as "robust and rounded", that of E. borealis as "small and lophoid", and that of E. gregoryi as "very small and subconic". The m3 hypoconulid of E. borealis is further described as being "centrally positioned" while that of E. gregoryi is "positioned externally" (that is, buccally). The m3 hypoconulid on the type of E. gregoryi is actually very similar morphologically to that of AMNH 14888, which Osborn referred to E. borealis , and not more buccally positioned. Although the m3 hypoconulid on the type of E. princeps is very large and rounded, the significance of this character is unknown. As noted by Wallace (1980), the distal part of the brontothere tooth row is highly variable and the size and morphology of the M3 hypocone or m3 hypoconulid generally make poor diagnostic characters.

Character E. princeps E. brownianus E. borealis E. gregoryi p2 shape ––––––– Compressed ––––––– Short, compressed p2 paraconid Elevated and distinct Very low on crown Very rudimentary Extremely p2 metastylid Elevated and distinct Rudimentary Very rudimentary –––––––

p2 talonid narrow ––––––– ––––––– –––––––

p2 entoconid Very rudimentary Invisible Very rudimentary –––––––

p2 hypoconid ––––––– Distinct and elevated ––––––– Very rudimentary p3 paraconid Quite distinct and ––––––– ––––––– Extremely p3 metastylid Small but distinct ––––––– ––––––– –––––––

p3 talonid Broad ––––––– ––––––– –––––––

p3 entoconid Rudimentary ––––––– ––––––– Extremely p4 talonid Broad ––––––– ––––––– –––––––p4 entoconid Distinct ––––––– ––––––– –––––––

m3 hypoconulid Robust and rounded ––––––– Small, sublophoid, Very small, Thus, based on the morphology as described by Osborn there does not appear to be any significant differences between the taxa that he recognized. It is still possible, however, that size differences may allow for the delineation of different species.

Unfortunately, most available specimens of Eotitanops are fragmentary, and because I have refrained from measuring specimens with any but the slightest of imperfections, I have not been able to compile a database of sufficient size to analyze statistically. Accepting the imprecision introduced by partially estimated measurements, however, then a provisional statistical analysis may be performed using measurements provided by Osborn (1929, p. 290) for several specimens of Eotitanops ..

Table 2 presents the summary statistics for all of the specimens of Eotitanops for which Osborn (1929) provided data. Almost half of the individual values of V (rounded to the nearest whole number) are greater than 10, and the majority of the values of V between 4 and 10 are at the higher end of this range (>8). Finally, the average value of V for the sample is also very high (= 10.1). These results strongly suggest that the sample is heterogeneous and represents more than a single species. The types of three species ( E. brownianus , E. princeps , and E. gregoryi ) are included in the sample.

Cluster analysis of all of the variables listed in Table 2 produces a dendrogram ( Figure 2 ) in which the type of Eotitanops princeps groups with specimens referred to E. borealis and the types of E. gregoryi and E. brownianus group out separately. The type of E. brownianus is very incomplete and Osborn only reported measurements for the length of the lower cheek tooth series (exclusive of p1), length of the premolar series (exclusive of p1), and the length of the molar series. Cluster analysis using only these three variables produces a dendrogram in which the type of E. brownianus groups with referred specimens of E. borealis , while the types of E. princeps and E. gregoryi group out separately ( Figure 3 ).

Eotitanops brownianus is probably not a valid species. The type is morphologically similar to specimens of E. borealis and none of the characters cited by Osborn convincingly distinguish them. The three variables for which Osborn provided measurements for the type specimen (length p2 to m3, length p2 to p4, and length m1 to m3) are close to those of his referred specimens of E. borealis and it is likely that E. brownianus is a junior synonym of this taxon. Although Osborn did not indicate it, his measurements for E. brownianus must be estimated values because the type of E. brownianus is missing the crowns on all but the left second premolar, and, thus, the minor size differences cited between the two species cannot be accorded great significance.

Eotitanops princeps is also probably invalid, based on morphology and size considerations (see below), but E. gregoryi appears to be valid. In Figure 2 and especially in Figure 3 , the type of Eotitanops gregoryi is well-separated from most other specimens of Eotitanops in the sample. This separation is accounted for by the extremely small size of the specimen. If the type of E. princeps (a rather large individual) is dropped from the statistical analysis so that the remaining specimens in the sample are closer in size to the type of E. gregoryi , the individual and average values of V for the sample remain high ( Table 3 ). This result suggests that the type of E. gregoryi is largely responsible for the high individual and average values of V in the original statistical analysis ( Table 2 ).

If the type of Eotitanops gregoryi is excluded from the statistical analysis, however, (and the type of E. princeps is retained) then all but three of the individual values of V (rounded to the nearest whole number) fall within the range of 4 to 10 ( Table 4 ). Two out of the three values of V that are outside of this range are below 4, which suggests that the samples from which these values were calculated may be too small to show all of the variation that is actually present. The average value of V for the sample exclusive of E. gregoryi is 6.36, which is within the ideal range suggested by Simpson et al. (1960). Thus, if the type of E. gregoryi is excluded from the analysis, the individual and average values of V for the specimens remaining in the sample are within the range accepted for a single species ( Table 4 ). If E. gregoryi is included in the sample, however, then the individual and average values of V generally exceed the range established for a single species ( Tables 2 and 3 ).

Eotitanops gregoryi is therefore considered a valid species of Eotitanops distinguished from E. borealis primarily by its much smaller size. This conclusion is further supported by t -tests performed by Wallace (1980) in which the type of E. gregoryi separated out from other specimens of Wind River Eotitanops with a probability of greater that 95 % for most dental measurements.

The preceding statistical analyses suggest that Eotitanops princeps is a junior synonym of E. borealis . Although the type of E. princeps is the largest specimen of Eotitanops described to date, it is not so large that it must be recognized as a distinct species. Even though E. princeps groups out separately from most specimens of Eotitanops in a cluster analysis comparing the lengths of p2–m3, p2–p4, and m1–m3, E. princeps groups with referred specimens of E. borealis when all variables are considered. Analysis of the coefficient of variation demonstrates that the type of E. princeps and specimens referred to E. borealis all fall within the size range of a single extant mammalian species. While the large hypoconulid on the m3 of the type of E. princeps might be considered diagnostic, the documented variability of the distal brontothere tooth row renders this character suspect for diagnosis.

The fourth Wind River species of Eotitanops recognized by Osborn, E. major , was based upon a left median metatarsal (metatarsal III) and the distal end of a tibia. Osborn's main justification for the recognition of E. major as a valid species was its larger size compared to E. princeps . Although the type specimen of E. princeps does not have any postcranial material that can be directly compared to the type of E. major, Osborn referred a partial pes (AMNH 4902) to E. princeps that probably was the basis for his size comparison. Although the median metatarsal of this referred specimen is considerably smaller than that of the type of E.

major , neither size ranges nor the significance of size differences have been established for this element in Eotitanops . Accordingly, E. major is regarded as a junior synonym of E. borealis . If it is later be determined that E. princeps is a valid species, then E. princeps and E. major may be synonyms.

In 1919 Osborn described the lower cheek teeth (p2–m3, AMNH 17439) of a diminutive brontothere from the upper part of the Huerfano Formation (Huerfano B) as a new species of Eotitanops , E. minimus distinguished from E. gregoryi by its smaller size. In 1929 Osborn again recognized E. minimus as a valid species of Eotitanops , but did not provide a more complete diagnosis of the taxon.

Osborn (1929) also identified specimens of E. gregoryi (AMNH 17418) and E. brownianus (AMNH 17441) from the upper part of the Huerfano Formation, but Robinson (1966, p. 67) referred the specimen of E. gregoryi to E. minimus (apparently because it is smaller than E. gregoryi from the Wind River Basin) and identified the specimen of E. brownianus as having come from his locality VII, which is in the lower faunal zone (the part of the Garcia Canyon Local Fauna that is equivalent to the Eotitanops borealis Assemblage Zone in the Wind River Basin). In addition to AMNH 17418 and the type, Robinson identified two other specimens of E. minimus : AMNH 56539 and YPM 16439. According to Robinson, all four of these specimens are from his locality II, which is in the upper faunal zone of the Huerfano Formation (Gardnerbutte Local Fauna).

According to Osborn (1929), the discovery of a "dwarf titanothere" together with brontotheres of the same size as Eotitanops gregoryi and E. brownianus in the same geologic horizon "reveals the existence of what is probably a distinct phylum of diminutive titanotheres separable from the Eotitanopinae ." Osborn was not willing, however, to formally recognize such a group until more complete skeletal material was available.

Wallace (1980) concluded that Eotitanops minimus represents an entirely new genus and provided the following diagnosis of the taxon:

“Small brontotheriid with relatively low, bunodont lower molar and premolar cusps; molar talonid basin broad and shallow and trigonid basin shallow, but relatively short anteroposteriorly; molar metalophid poorly developed and molar hypolophid absent; third lower molar hypoconulid reduced relative to other (earlier and later) brontotheriids, with cusp closely appressed to entoconid and having no lingual basin whatsoever; third upper molar with rudimentary or no metastylar ridge and no posterior cingulum; fourth lower premolar entoconid small, low, but distinct.”

Of these characters, the presence of a distinct entoconid on p4, and possibly the size and morphology of the hypoconulid of m3, are likely synapomorphies.

Eotitanops minimus is most probably a junior synonym of E. gregoryi . Both E. gregoryi and E. minimus possess an entoconid on p4 (see Figure 4 ), and this may be a derived character as Wallace suggested. The presence of an entoconid on the p4 of Phenacodus (regarded as a sister taxon to perissodactyls), however, and on some specimens of Hyracotherium (basal Equidae ?), raises some uncertainty about the polarity of this character. A limited statistical analysis also strongly suggests that E. gregoryi and E. minimus are similar in size (within the range of a single extant mammalian species), but are distinctly smaller than specimens of E. borealis .

Although I do not have numeric data for specimens of Eotitanops minimus, Robinson (1966, p. 67) provided measurements of the second lower molar for four specimens of Eotitanops from the Huerfano Formation (including the type of E. minimus ) and four specimens of Eotitanops from the Wind River Formation. The Wind River sample consisted of three specimens of E. borealis and the type of E. gregoryi .

The coefficients of variation for the combined Wind River and Huerfano sample ( Table 5 ) are very high (12 to 13, rounded to the nearest whole number), suggesting that the combined sample is not homogeneous and probably consists of more than a single taxon. Cluster analysis of all three variables in Table 5 results in a dendrogram ( Figure 5 ) in which two size groups are clearly delineated. Specimens of Eotitanops borealis from the Wind River Formation form one size group and specimens of E. minimus from the Huerfano Formation form the other. Interestingly, although the type of E. gregoryi is from the Wind River Formation, it groups among the specimens of E. minimus .

If the specimens of Eotitanops borealis are dropped from the analysis, and the coefficient of variation recalculated for the remaining individuals (the specimens of E. minimus and the type of E. gregoryi ), then the resulting values of V are well within the parameters established for a single mammalian species ( Table 6 ). Although this statistical analysis was based on size variation in a single tooth, it indicates that E. gregoryi and E. minimus are probable synonyms.

As noted above, Osborn (1929) suggested and Wallace (1980) concluded that the species Eotitanops minimus represents a new genus distinct from Eotitanops . Both writers, however, recognized E. gregoryi as a valid species of Eotitanops . Little is known about the morphology of E. gregoryi (= E. minimus ), however, and it is difficult to demonstrate that it can be referred to Eotitanops , or conversely that is should be assigned to a new genus. Given that the lower dentition of " Eotitanops " gregoryi is similar to that of E. borealis , and that previous authors have recognized " Eotitanops " gregoryi as a species of Eotitanops , I provisionally accept this generic identification as correct. Alternatively Eotitanops gregoryi could be classified simply as a "primitive brontothere, incertae sedis " rather than attempting to make a definite generic assignment.

The conclusions regarding valid species of Eotitanops presented above are largely the same as those reached by Gunnell and Yarborough (2000). Both the present paper and Gunnell and Yarborough recognize a large form and a small form of Eotitanops , and both papers refer the large form to E. borealis . The small form is here referred to E. gregoryi , in contrast to Gunnell and Yarborough who referred it to E. minimus , which is regarded here as a junior synonym of E. gregoryi . Gunnell and Yarborough, however, consider E. gregoryi a junior synonym of E. borealis .

The validity of Eotitanops gregoryi is supported by its small size and the presence of a p4 entoconid in the type, which distinguishes it from specimens of E. borealis , but makes it similar to specimens referred to E. minimus . Furthermore, Gunnell and Yarborough distinguished E. borealis from the smaller species (my E. gregoryi ) by its large, elongated hypoconulid, whereas in the smaller species the hypoconulid is proportionately smaller. Although the distal part of the brontothere tooth row is highly variable and the size of the hypoconulid is probably a poor diagnostic character, the hypoconulid in the type of E. gregoryi is rather small in size and, thus, is similar to specimens of E. minimus , but different from specimens referred to E. borealis .

As stated above, Osborn (1929) designated a relatively complete skull (AMNH 14887) as the " neotype " for the type species of Eotitanops ( E. borealis ), but because the original type is still preserved, Osborn's specimen is nothing more than a referred specimen. Osborn (1929) argued that the facial region of this " neotype " skull is longer than the cranial region while in all other brontotheres the face is shorter than the cranium. This proportional difference was Osborn's main justification for separating Eotitanops from middle Eocene brontotheres such as Palaeosyops . Wallace (1980), however, pointed out that Osborn's interpretation of the facial and cranial proportions of the " neotype " skull is questionable because the specimen is very fragmentary and the position of key morphological landmarks is uncertain. Wallace noted that Eotitanops is dentally similar to Palaeosyops and argued that until a more complete skull of Eotitanops is known, cranial proportions cannot be used to separate it from Palaeosyops . Wallace, therefore, regarded Eotitanops as a junior synonym of Palaeosyops .

Despite Wallace's valid objections, Mader (1989) argued that there are enough morphological differences between Eotitanops and Palaeosyops to justify separation of the two at the generic level. Mader noted that the left zygomatic arch of Osborn's " neotype " skull of E. borealis is complete and shows that this structure was thin and probably relatively straight. The zygomatic arch of Eotitanops was thus probably quite similar to the zygomatic arch of Mesatirhinus . In Palaeosyops , however, the zygomatic arch is very robust and sharply curved. Mader further argued that although much of the skull is lacking, the portions of the zygomatic arches and palate that are preserved suggest that the skull was dolichocephalic (as stated by Osborn 1929) or mesaticephalic. In contrast, all skulls of Palaeosyops are strongly brachycephalic, and this is a synapomorphic character of that genus.

Although the anterior dentition of Osborn's skull is incomplete, Mader also noted some differences between it and Palaeosyops . The left P1 is lacking entirely on Osborn's skull and the place on the maxilla to which it had been attached has been covered by plaster. The root of the left upper canine and all of the left P2 are, however, preserved. There is a long space between the left canine and left P2, so that no matter where the left P1 was originally placed there must have been a considerable diastema present. Skulls of Palaeosyops have either an extremely short diastema or no diastema at all.

Mader (1989) was unsure whether facial proportions could be used to distinguish Eotitanops from other brontotheres, but later (1991, 1998) concluded that Osborn was probably correct in asserting that the facial region of the “ neotype ” skull of E. borealis is proportionally much longer than in other brontothere genera. Mihlbachler (2008, p.375), however, cited a skull, unfortunately not illustrated, that he referred to Eotitanops (UCMP 132049) with a short face as in other brontotheres.

Until about twenty five years ago, classic rock units yielding specimens of Eotitanops (such as the Lostcabin Member of the Wind River Formation and upper part of Huerfano “A”) were regarded as being Wasatchian in age (e.g., Robinson 1966; Savage & Russell 1983). Indeed, the original Wood Committee report ( Wood et al. 1941) cited the first appearance of Eotitanops as one of the characteristics of the then newly named Wasatchian Land Mammal Age. In recent decades, however, strata yielding the earliest occurrence of Eotitanops have been regarded as being early Bridgerian (Gardnerbuttean) in age ( Stucky 1984; Gunnell & Yarborough 2000; Zonneveld et al. 2000; and Robinson et al. 2004).

Figures

FIGURE 1. Skull of Eotitanops borealis (AMNH 14887, Osborn’s “ neotype ”) in A, ventral, and B, lateral view. After Osborn (1929).

FIGURE 2. Cluster dendrogram for specimens of Eotitanops from the Wind River Basin resulting from a cluster analysis of all variables listed in Table 2 (based on data from Osborn 1929). a, type of Eotitanops princeps; b, type of E. gregoryi; c, type of E. brownianus.

FIGURE 3. Cluster dendrogram for specimens of Eotitanops from the Wind River Basin resulting from a cluster analysis of length of the cheek tooth series (exclusive of p 1), length of the premolar series (exclusive of p 1), and length of the molar series (based on data from Osborn 1929). a, type of Eotitanops brownianus; b, type of E. princeps; c, type of E. gregoryi.

FIGURE 4. Stereophotographs of lower dentitions of Eotitanops gregoryi. A, AMNH 17439 (type of E. minimus, cast of p 4 – m 3) and B, AMNH 11489 (type of E. gregoryi, cast of p 4). Arrows point to entoconid on p 4.

FIGURE 5. Cluster dendrogram for specimens of Eotitanops from the Wind River and Huerfano Basins resulting from a cluster analysis of all three variables listed in Table 5 (based on data from Robinson 1966). a, type of Eotitanops gregoryi; b, type of E. minimus.

Tables

TABLE 2. Summary statistics for specimens of Eotitanops from the Wind River Basin (data from Osborn 1929; all measurements in millimeters).

	n Range	M	s V
Length p2 to m3	6 78.0 – 105.0	93.5	±9.1 9.7
Length p2 to p4	6 29.4 – 39.0	35.6	±3.4 9.5
Length Lower Molar Series	6 49.0 – 66.0	57.8	±5.6 9.8
Length m3	5 19.5 – 25.0	22.5	±2.0 8.9
Width m3	3 10.7 – 14.0	12.1	±1.7 14.3
Length m2	4 15.5 – 21.0	18.4	±2.3 12.5
Width m2	4 10.5 – 14.0	12.3	±1.5 12.4
Length m1	4 14.5 – 18.3	16.2	±1.6 10.0
Width m1	4 8.5 – 12.0	10.4	±1.5 14.4
Length p4	3 12.0 – 13.0	12.5	±0.5 4.0
Width p4	3 7.5 – 8.0	7.8	±0.3 3.7
Length p3	4 9.5 – 12.5	11.5	±1.4 11.8
Width p3	4 5.2 – 7.0	6.3	±0.8 12.1
Length p2	4 8.8 – 13.0	11.3	±1.8 15.9
Width p2	4 6.0 – 6.3	6.1	±0.2 2.5
AVERAGE V			10.1

TABLE 3. Summary statistics for specimens of Eotitanops from the Wind River Basin exclusive of the type of E. princeps (data from Osborn 1929; all measurements in millimeters).

	n Range	M	s V
Length p2 to m3 Length p2 to p4	5 78.0 – 98.0 5 29.4 – 38.0	91.2 34.9	±8.0 8.7 ±3.3 9.3
Length Lower Molar Series	5 49.0 – 60.0	56.2	±4.4 7.9
Length m3 Width m3	4 19.5 – 23.2 2 10.7 – 11.5	21.9 11.1	±1.7 7.8 ±0.6 5.1
Length m2	3 15.5 – 19.0	17.5	±1.8 10.3
Width m2 Length m1	3 10.5 – 13.0 3 14.5 – 16.3	11.7 15.4	±1.3 10.7 ±0.9 5.9
Width m1	3 8.5 – 11.0	9.8	±1.3 12.8
Length p4 Width p4	3 12.0 – 13.0 3 7.5 – 8.0	12.5 7.8	±0.5 4.0 ±0.3 3.7
Length p3	3 9.5 – 12.0	11.2	±1.4 12.9
Width p3 Length p2	3 5.2 – 7.0 3 8.8 – 12.0	6.2 10.8	±0.9 14.9 ±1.7 16.0
Width p2	3 6.0 – 6.2	6.1	±0.1 1.9
AVERAGE V			8.8

TABLE 4. Summary statistics for specimens of Eotitanops from the Wind River Basin exclusive of E. gregoryi (data from Osborn 1929; all measurements in millimeters).

Length p2 to m3 Length p2 to p4	n 5 5	Range 90.0 – 105.0 35.0 – 39.0	M 96.6 36.8	s ±5.6 ±1.6	V 5.8 4.5
Length Lower Molar Series	5	55.0 – 66.0	59.6	±4.0	6.8
Length m3 Width m3	4 2	22.0 – 25.0 11.5 – 14.0	23.3 12.8	±1.3 ±1.8	5.4 13.9
Length m2	3	18.0 – 21.0	19.3	±1.5	7.9
Width m2 Length m1	3 3	11.7 – 14.0 15.5 – 18.3	12.9 16.7	±1.2 ±1.4	8.9 8.6
Width m1	3	10.0 – 12.0	11.0	±1.0	9.1
Length p4 Width p4	3 3	12.0 – 13.0 7.5 – 8.0	12.5 7.8	±0.5 ±0.3	4.0 3.7
Length p3	3	12.0 – 12.5	12.2	±0.3	2.4
Width p3 Length p2	3 3	6.3 – 7.0 11.5 – 13.0	6.6 12.2	±0.4 ±0.8	5.5 6.3
Width p2	3	6.0 – 6.3	6.1	±0.2	2.8
AVERAGE V					6.4

TABLE 5. Summary statistics for specimens of Eotitanops from the Wind River and Huerfano Basins (data from Robinson 1966; all measurements in millimeters).

	n Range	M	s V
Length m2	7 14.5 – 20.0	17.2	±2.1 12.2
Width m2 Trigonid Width m2 Talonid	8 9.1 – 12.5 8 9.3 – 12.5	10.7 10.9	±1.4 12.7 ±1.3 11.6

TABLE 6. Summary statistics for specimens of Eotitanops from the Wind River and Huerfano Basins exclusive of specimens of E. borealis (data from Robinson 1966; all measurements in millimeters).

	n Range	M	s V
Length m2 Width m2 Trigonid	4 14.5 – 17.0 5 9.1 – 10.6	15.7 9.8	±1.0 6.6 ±0.6 6.3
Width m2 Talonid	5 9.3 – 10.9	10.1	±0.6 5.8