632 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



ever, that in a general way the vertebrae of Manteo- 

 ceras seem to be intermediate in form between those 

 of Palaeosyops and of DolicTiorhinus, sharing with 

 those of Palaeosyops the high spine of the axis and the 

 anteroposteriorly narrower spines of the anterior 

 dorsals and sharing with those of DolicTiorhinus the 

 larger pleurapophysal flange of C. 6, the marked angu- 

 lation between the prezygapophysis and the neural 

 spine in D. 2, the inferior keels on D. 1, D. 2. 



A3/ 



A2 



Figure 551. — Atlas of Manteoceras manteoceras 

 Am. Mus. 12204; upper Bridger. Ai, Posterior view; As, anterior 

 view; A3, ventral view; Ai, dorsal view. The arrows indicate the 

 course of the first spinal nerve and vertebral artery. One-third 

 natural size. 



The atlas of Manteoceras has a larger vertebrarterial 

 canal, the anterosuperior border of the cotylus is more 

 deeply concave, the superior openings for the spinal 

 nerve are nearer the lateral borders. The median 

 hypapophysis is acuminate rather than peg shaped. 

 The posterior root, or proximal portion of the pleura- 

 pophysis is thicker vertically. The atlas, on the 

 whole, is intermediate in form between those of 

 Palaeosyops and of DolicTiorhinus, it being broader 



than that of Dolichorhinus, but sharing with it several 

 of the characters already mentioned. 



The axis probably had a taller spine than in Palaeo- 

 syops; the crura of the neural arch appear shorter 

 anteroposteriorly; the postzygapophyses were larger 

 and less inclined downward; the neural canal was 

 more arched — that is, less angulate superiorly; the 

 posterior face of the centrum was not so wide. 



The third to seventh cervical vertebrae (see fig. 552) 

 differ from those of Palaeosyops (Am. Mus. 1580, 

 1562). The neural canal is smaller and more roundly 

 arched superiorly; the anterior faces of the centra 

 are more circular instead of being broadly flat- 

 tened at top; the posterior faces of the centra are 

 also rounder superiorly and shallower vertically; the 

 prezygapophyses and postzygapophyses are possibly 

 larger. The neural spines seem to be of about the 

 same relative size in the two genera, and lateral 

 flanges or pleurapophyses appear on C. 3 to C. 6 in 

 both genera, but the flange on C 6 seems to have 

 been larger in Manteoceras than in Palaeosyops. 



The first dorsal had a spine of nearly the same rela- 

 tive size and proportions as in Palaeosyops; the neural 

 canal is smaller and more arched superiorly, the front 

 face of the centrum is rounder at top, and the rear face 

 of the centrum, including the facets for the capitulum 

 of the second rib, is more transversely oval. The 

 oblique external buttress or column formed by the con- 

 joined pedicles of the prezygapophysis and pleura- 

 pophysis in Manteoceras is shorter; the prezygapoph- 

 ysis seems more horizontal and broader anteropos- 

 teriorly. In the second dorsal both the anterior and 

 posterior faces of the centrum are more transversely 

 oval, the pleurapophyses spring from the sides of the 

 centra at a lower level, the rib facets are smaller; 

 the prezygapophyses face upward and form an open 

 angle with the base of the neural spine, whereas in 

 Palaeosyops the prezygapophyses of this vertebra face 

 forward and upward and are nearly continuous with 

 the base of the spine. The remaining vertebrae are 

 too imperfectly known to warrant description. 



The scapula (Am. Mus. 1587) is too insufficiently 

 known to afford diagnostic characters. 



The humerus (Am. Mus. 12204, fig. 553, B) is of 

 intermediate length (290 mm.) between the long 

 humerus of Palaeosyops (325) and the relatively short 

 humerus of DolicTiorhinus (255 to 315); it agrees nearly 

 in length and general characters with the humerus of 

 the Limnohyops skeleton Am. Mus. 11689, from 

 Bridger B 2 (see above) ; it is massive proximally, with 

 a heavy deltoid ridge; it apparently had a platelike 

 crest for the infraspinatus muscle and stout tuber- 

 osities for the supraspinatus and deltoid muscles; the 

 supinator crest is relatively short. The radius and 

 ulna are not known. 



The manus (Am. Mus. 12204, 1587, 12216) repre- 

 sents a broad-footed phase of the same stock that gave 



