648 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



As compared with Palaeosyops the cervicals had 

 sHghtly longer and relatively smaller centra, shghtly 

 longer prezygapophyses and postzygapophyses and 

 perhaps larger pleurapophysial flanges on C. 1- C. 

 6; the spines are broken off but may have been 

 more slender at the base. The anterior dorsals had 

 relatively somewhat smaller centra with pronounced 

 inferior keels; the neural spines were broader antero- 

 posteriorly; the anterior zygapophyses of D. 2 faced 

 upward and inward, making a decided angle with the 

 neural spine as in Manteoceras, whereas in Palaeosyops 

 the zygapophyses were nearly continuous with the 

 spine. 



The neck, which measures only 320 millimeters in 

 length, is short, especially when compared with the 

 remarkably long skull, which is estimated at 540 

 millimeters. 



The remainder of the column all belongs to one 

 animal (Am. Mus. 1843). The axis (fig. 580) does not 

 exhibit so high a spine as in Palaeosyops or Manteoceras, 

 but we recall the fact that BolichorMnus has a low 

 occiput. There is little evidence of high spines on 

 C 3-C. 5; the powerful ligamentum nuchae was sup- 

 ported by the high and extensive spines of D. 1-D. 8. 

 Cei'vicals 1-6 are characterized by widely expanding 

 and actually overlapping inferior lamellae, distinct in 

 C 6 from the pleurapophysis above (thus unlike 

 Palaeosyops). The centra are quite deeply opistho- 

 coelous and laterally compressed. The zygapophyses 

 face vertically and obliquely outward and inward from 

 C 3 to the anterior face of D. 1 ; they face horizontally 

 downward and upward in D. 1 to D. 12; beginning 

 with the posterior face of D. 12 to L. 4 the zygapophyses 

 are vertically placed, facing outward and inward and 

 more or less sigmoid or revolute in curvature, as in 

 certain lumbars of Palaeosyops. The neural spines 

 from D. 1 to L. 4 are extended anteroposteriorly in 

 marked contrast to the feeble spines of Palaeosyops. 

 The metapophyses are unusually broad; that of L. 4 

 articulates with the front border of the ilium as in 

 Eguus. The centra throughout are relatively deep; 

 the depth equals the height in the posterior cervicals 

 and anterior dorsals, but in the lumbars the height 

 slightly exceeds the depth. The centra are com- 

 pressed, or keeled inferiorly. There are apparently 

 but three true sacrals in this specimen, but the number 

 can not be ascertained positively. 



The upward curvatm-e of the dorsolumbar region 

 of the column is greater than that represented in 

 Figure 580 but is correctly indicated in the restoration. 

 Figure 579. 



Arches and limb hones. — A scapula (fig. 582) is doubt- 

 fully associated with Dolichorhinus; it presents rather 

 high and narrow proportions and measures 320 milli- 

 meters vertically. 



The humerus is known from two specimens. It is 

 & highly characteristic and progressive bone, closely 



resembling that of the large lower Oligocene titano- 

 theres, such as Brontotherium leidyi. The most dis- 

 tinctive Oligocene titanothere character is the very 

 high, thin, and platelike great tuberosity (tub. maj.) 

 with an erect anterior process. The humerus asso- 

 ciated with Am. Mus. 1843 (fig. 583) is short; it 

 measures 285 millimeters to the tip of the great 

 tuberosity, while the length of the shaft is 255 mUh- 

 meters. The humerus of the other specimen (Am. 

 Mus. 13164) belongs to an animal of larger size; 

 length of shaft, head to distal extremity, 315 milli- 

 meters; extreme distal width 95. The disparity in 

 size of the two humeri as compared with the approx- 

 imate equality in size of the respective skulls is a 

 puzzling feature. 



The length of the forearm or ulna and radius can 

 only be estimated; it certainly is considerably less 

 than that of the humerus (estimated humeroradial 

 ratio 81). This indicates (see p. 733) that Bolicho- 

 rMnus was an animal capable of more speed than 

 Palaeosyops but of less speed than Mesafirhinus, in 

 which the humerus and radius are more subequal. 

 The most highly characteristic feature of the ulna 

 is the form of the olecranon process, which is obtuse, 

 highly rugose, and incurved, as shown in Figure 584, a 

 character which relates this animal to Mesatirhinus. 

 The manus of D. longiceps, figured by Peterson (Car- 

 negie Mus. Mem., vol. 9, pt. 4, pi. 54), is remark- 

 ably like that of Mesatirhinus but relatively some- 

 what shorter and more massive. 



Little is certainly known of the pelvis beyond the 

 outline indicated in Figure 580. 



The femur (Am. Mus. 13164, fig. 579) exhibits a 

 total length of 387 millimeters as compared with the 

 length of the humerus, 315, and with the total basUar 

 length of the skull, 540. The femur has the character- 

 istic straight shaft of the titanotheres generally and is 

 readily distinguished from that of the contemporary 

 Amynodon (fig. 518) by the more vertical position of 

 its patellar facet (in Amynodon these facets are placed 

 very obliquely) and the lesser development of the third 

 trochanter, which is very prominent in Amynodon, as 

 in aU other rhinoceroses. 



Mts III and IV (Am. Mus. 13164, fig. 585) are of 

 almost brachypodal proportions, being no longer than 

 those of Mesatirhinus petersoni but much broader 

 (dimensions are given above). They show syngenetic 

 resemblances to Mesatirhinus and Metarhinus. 



COMPARISON OF THE FOKE LIMB OF DOLICHORHINUS AND 

 AMYNODON 



Before it was learned that Dolichorhinus was brachy- 

 podal a finely preserved fore limb (Am. Mus. 1961, 

 from Uinta B 2, figs. 586, 587) was provisionally referred 

 to that genus. It differs from Dolichorhinus, however, 

 and agrees with Amynodon in the detailed characters 

 of the upper arm and forearm, especially in relative 



