418 



TITANOTHERES OF ANCIENT WYOMING, DAKOTA, AND NEBRASKA 



which thus appears to form a septum between two 

 elongate oval cavities, whose function is unknown. 

 It may be conjectured that they lodged diverticula 

 of the epithelial lining of the oral cavity or that the 

 "pits" were mere antra, or hollow spaces, the outer 



A4t 

 Figure 354. — Skull of Sphenocoelus uintensis 



One-fourth natural size. 



m. Mus. 1501 (type); Wagon Hound Bend, White Eiver, CTinta Basin, Utah, Uinta B I. 

 Ai, Top view; A2, basal view; A3, occipital view; A4, side view. 



wall of which became extremely thin or even partly 

 incomplete. Similar thin-walled or partly open cavi- 

 ties occur on either side of the basisphenoid in certain 

 specimens of the two-toed sloth {Chohej^us hoffmanni) . 

 When this peculiar skull was first described the 

 relationship of the animal to which it belonged was 



entirely uncertain. Now, after more detailed study 

 of the Eocene titanothere skull in general and especially 

 of the skull of Dolichorhinus, it appears that Spheno- 

 coelus is certainly an aberrant titanothere and not a 

 chalicothere (suborder Ancylopoda), as at first sug- 

 gested. The chief consid- 

 erations in favor of this 

 view are the following: (1) 

 The large alisphenoid canal 

 relates it to the Perissodac- 

 tyla; (2) the wide space 

 between the foramen ovale 

 and foramen lacerum medi- 

 um (40 mm.) removes it 

 from affinity with all other 

 perissodactyl families ex- 

 cept the Brontotheriidae 

 and Equidae; (3) the basi- 

 cranial region and especially 

 the elongate, oblique form 

 of the glenoid facets for the 

 lower jaw is strikingly sim- 

 ilar to that of Dolichorhinus 

 hyognathus, although the 

 postglenoid processes are 

 somewhat different in form ; 

 (4) the detailed relations of 

 the foramen rotundum, 

 alisphenoid canal, foramen 

 ovale, tympanic fossa, fora- 

 men condylare, as well as 

 the form of the petrosal 

 bone and of the paroccipital 

 processes, point in the same 

 direction; (5) the form of 

 the temporal, sagittal, and 

 occipital crests, the arrange- 

 ment of the venous postpa- 

 rietal foramina, the spacing 

 of the postglenoid and post- 

 tympanic processes (see top 

 view in fig. 354), are all 

 clearly foreshadowed in the 

 brain case Princeton Mus. 

 10041 (figs. 713, 716), which 

 is provisionally regarded as 

 a very progressive Mesati- 

 rhinus. 



Generic distinctions. — 

 Mingled with all these 

 points of resemblance to 

 Dolichorhinus are strildng 

 generic differences, such as the pair of pits already 

 mentioned, the short sagittal crest diverging into raised 

 temporal ridges, the excessively large and broad con- 

 dyles, and, as indicated in the widely separated postgle- 

 noid and post-tympanic processes, an extreme degree 

 of dolichocephaly. 



