THE MAMMALIA OF THE DEEP RIVER BEDS. 181 



dogs in North America, contrasted with their absence or unimportance in Europe, 

 renders it very probable that the family originated in the former continent. 



(4) The name Anchitherium has been improperly applied to American equines 

 from the White River and John Day, and should be replaced by Mesohippus and 

 Miohippus, the latter genus extending through the John Day and into the Loup Fork. 



(5) Desmatippus is a new genus of equines which nearly fills the gap between 

 Miohippus and ProtoJiippus, the molar teeth being intermediate in character between 

 the two, brachyodont, and yet with a thin deposit of cement in the valleys. 



(6) A quite unexpected discovery is that of a species of Anchitherium, of the 

 type of the European A. aurelianense. The genus is very probably of American 

 origin, and, as Schlosser and Mme. Pavlow have suggested, was almost certainly not 

 in the direct line of equine descent, but it has paralleled the true horses in many 

 interesting ways, such as the spout-shaped odontoid, etc. 



(7) Surveying the series of equine genera, which there is such good reason to 

 believe constitute an actual line of descent, we find a steady advance in differentia- 

 tion in the main, accompanied by alternating progression and regression in minor 

 details. It is also very probably true that a slight degree of specialization in a 

 direction away from that taken by the main line, is not incompatible with a place in 

 that line, as is exemplified by the peculiar character of the elbow joint in Mesohippus, 

 which is greatly diminished in Miohippus and dies out in succeeding genera. 



(8) Some of the accessory tubercles in both the American and European species 

 of Anchitherium appear to favor the view of " indeterminate variation." 



(9) The rhinoceroses of the Old World separated at a very early period from 

 those of the New and cannot well have any common ancestor nearer than the Acera- 

 theria of the Oligocene ; the American series has, however, run parallel to the 

 European in many important details of structure. 



(10) Mesoreodon, a new genus of oreodonts from the lower beds, agrees with 

 Eporeodon of the John Day in most characters of skull and dentition (though with 

 some resemblances to Merychyus) while the feet are altogether like those of the latter 

 genus. Very curious features of this genus are the presence of an ossified thyroid 

 cartilage of the larynx, a rudiment of the bony clavicle and a metacromial process 

 of the scapular spine. It is suggested that the large acromion of the artiodactyls, 

 and its absence in even the Eocene perissodactyls, may be correlated with the earlier 

 loss of the clavicle in the latter group. 



(11) The skeleton of the oreodont genus, Merycochoerus, is now almost com- 

 pletely known, which permits exact comparison with other members of the group. 



(12) Merychyus is probably to be derived from Oreodon through Eporeodon and 



