THE MAMMALIA OF THE DEEP KIVER BEDS. 133 



tion does not seem probable. (2) There is a longer and more curved suprasvlvian 

 fissure, which is connected anteriorly with the coronal fissure by means of a short 

 and faintly marked ansate sulcus. This connection is also found in some specimens 

 of Oreodon. (3) The coronal sulcus consists of two portions ; the anterior is longer 

 and curves downward and outward, while the posterior is shorter and more obscurely 

 marked and converges towards the middle line in a way that suggests the crucial 

 sulcus of the Carnivora. The lateral view shows, in addition to these fissures, a 

 short and nearly horizontal sylvian sulcus and a presylvian which has a more nearly 

 vertical course than in Oreodon. The sylvian fissure appears to be connected with 

 the fissura rhinalis, though in this region the sulci arc very obscure and diffi- 

 cult to interpret. Indications of a posterior suprasylvian sulcus are also to be 

 observed. 



The character of the cerebral sulci is, it is obvious, very much the same as that 

 which occurs among the smaller and more primitive forms of existing ruminants, and 

 these, as Krueg has shown, agree closely in fundamental plan with the Carnivora. 

 As there is every reason to believe that the Oreodontidue are connected with the 

 Pecora only through very ancient forms, in which the hemispheres were either smooth 

 or but very little convoluted, this resemblance must be chiefly ascribed to parallelism 

 of development. Still more obviously is this the case with regard to the likeness 

 between these artiodactyls and the Carnivora. 



In both of the brain-casts of Mesoreodon the olfactory lobes are broken away, 

 but it is plain that they were not at all overlapped by the cerebrum. The cerebellum 

 is very much as in the older White River type ; its posterior face rises nearly verti- 

 cally from the medulla ; the vermis is large and prominent and the lateral lobes are 

 broad. In neither of the specimens is the cerebellum sufficiently well preserved to 

 display the details of the convolutions. 



The Vertebral Column. The atlas is rather more like that of the true ruminants 

 than is that of Eporeodon. This is due principally to the more uniform width of the 

 transverse processes and their continuation into short spines behind the surfaces for 

 the axis, from which they are separated by decided notches. This prolongation of 

 the transverse processes is, however, much less marked than in the Pecora. On the 

 other hand, the processes are more widely expanded laterally than in JEporeodon, 

 which is a departure from the ruminant type. The anterior extension of the trans- 

 verse process has, as in the earlier genera of the family, converted the atlanteo-diapo- 

 physeal notch into a foramen, but there is no perforation for the vertebrarterial canal. 

 The anterior cotyli for the occipital condyles are deep and are more distinctly sepa- 

 rated at their inferior borders than in Eporeodon and the neural spine is larger and 



