EVOLUTION OF THE SKELETON OF EOCENE AND OLIGOCENE TITANOTHEEES 



657 



extended process of the articulation. I judge that the axis as 

 a whole, in the present form, is relatively shorter than in the 

 Princeton specimen. In more minute details the description of 

 Scott and Osborn (op. cit., p. 514) agrees well with the parts 

 preserved, in the specimen before me — that is, the heavy spine 



pared. The second, third, fourth, and fifth dorsals have their 

 spines very nearly complete. In proportion they agree quite 

 well with those of the Oligocene genus but are more strongly 

 inclined backward. As in Menodus the transverse processes are 

 not extremely heavy and the capitular facets for the ribs are of 

 large size, while the sides of the centra are deeply concave. The 

 latter are deeper than broad, and the inferior borders, especially 

 the posterior ones, are distinctly more keeled than in Menodus. 

 Back of the eighth dorsal there is a break in the vertebral 

 column and a number of bones are lost. A second block which 

 was found, together with the one just described, contains por- 

 tions of six poste- 

 rior dorsals and 

 three lumbar verte- 

 brae. [See fig. 599.] 

 The neural spines 

 of the dorsal series 

 are prominent and 

 quite lumbar -like 

 in tlieir general 

 character. The 

 zygapophyses are 

 also of the inter- 

 locking lumbar type, and there 

 are prominent metapophyses. 

 The centra are somewhat mu- 

 tilated, but enough is preserved 

 to indicate that they are deep 

 of comparatively small 

 transverse diameter. 



There are, as stated, three 

 lumbar vertebrae present in the 

 paratype (No. 2860). These 

 bones are fortunately found in 



position succeeding the last dorsal vertebra, and for the 

 first time apparently furnish data as to the correct num- 

 ber of the lumbar vertebrae of the titanotheres. That 



Figure 600. — Scapula of 

 Eotitanotherium osborni 



Carnegie Mus. 2859 (type); Uinta 

 B 2. After Peterson. One-sixth 

 natural size. 



Figure 599. — Vertebrae of Eotitanotherium osborni 



Carnegie Mus. 2860 (paratype); Uinta B 2. After Peterson. A, Last 

 cervioal and dorsal vertebrae, right side; B, posterior dorsal and lumbar 

 vertebrae, left side. One-sixth natural size. 



overhanging the postzygapophyses, the inner turn of the 

 transverse process, and a prominent inferior keel. 



The succeeding four cervical vertebrae in the paratype 

 (No. 2860) are represented only by fragments. They 

 appear to have short opisthocoelian centra, as in Dipla- 

 codon described by Marsh and Osborn, and a prominent 

 ventral keel. 



The seventh cervioal vertebra is completely worked out 

 in half relief and shows the chief characteristic features 

 [fig. 599]. The long and pointed spinous process is well 

 shown, as is also the neural arch and the centrum. The 

 pre- and post-zygapophyses are, as in the axis, located 

 quite laterally and face directly upward and downward as 

 in Menodus. The transverse process shows a tendency 

 to develop the broad round termination found in Brontops 

 validus of the Oligocene. 



There are eight dorsal vertebrae, which are worked 

 out in half relief and rest on the original block of sand- 

 stone on which they were found. The neural spine 

 of the first dorsal is broken off about 10 centimeters 

 above the neural arch, but judging from the size of the 

 fracture, the spinous'] process attained a length equal, 

 and perhaps even proportionally greater, than was the case 

 in B. validus with which the Uinta remains have been com- 



Ai A2 Bi B2 



Figure 601. — Humerus, radius, and ulna of Eotitanotherium osborni 



Carnegie Mus. 2860 (paratype); Uinta B 2. Ai, Anterior view of humerus; Ai, posterior view 

 of humerus; Bi, lateral view of radius and ulna; B2, anterior view. One-sixth natural size. 



the last one of this series is the last lumbar vertebra there 

 is but little or no doubt, inasmuch as the neural spine 



