92 THE MAMMALIA OF THE DEEP KIVER BEDS. 



surface for the ectocuneiform is much less developed and less completely separated 

 from the anterior portion by the transverse sulcus. On mt. ii the surface for the 

 entocuneiform is much less conspicuous. 



The type of the genus was found by Mr. Benet. 



The Systematic Position of Desmatipjms. 



Morphologically, there can be no doubt that this genus stands exactly interme- 

 diate between Miohippus of the John Day and Protohippus of the Loup Fork, and 

 fills up the gap which has hitherto existed between those genera. This intermediate 

 position is especially clear in the structure of the teeth. Desmatippus shares with 

 Miohippus the short-crowned molars, and with Protohippus the presence of cement 

 and the confluence of the posterior transverse crest with the outer wall in the upper 

 teeth, and in the lower teeth the extension of the inner cusps and narrowing of the 

 entrances to the internal valleys, though these features are less conspicuous. From 

 these molars to those of the relatively short-crowned species of Protoliippus the tran- 

 sition is an easy one. The same intermediate character is shown in the limbs and 

 feet, so far as they are known, save only the greater length and slenderness of the 

 metapodials, as compared with those of the more differentiated genus. These are, 

 however, but specific, as distinguished from generic, characters and have but little 

 importance. 



An apparently strong objection to the position which I have assigned to this 

 new equine genus may be drawn from the stratigraphical fact that it has as yet been 

 found only in association with Protoliippus. But, as we have already seen, there is a 

 marked break between the faunas of the lower Deep River beds (i. e., upper John 

 Day) and the upper beds of the same region (i. e., lower Loup Fork). With a few 

 possible exceptions, no species of mammal is common to the two horizons and the 

 great majority of the genera are different also. This abrupt change points with 

 great probability to a hiatus between the formations, and in this case we may well 

 believe that Desmatippus originated during the unrecorded period, and, after having 

 given rise to Pro'ohippus, persisted into the Loup Fork, just as Miohippus has done. 

 Were the John Day beds unknown, we should have precisely the same difficulty 

 with regard to the latter genus. Should this supposed unconformity prove not to 

 exist, we must then assume that the later fauna was developed in some other part of 

 the continent and reached the Montana valley by a migration. This assumption 

 would dispose of the difficulty equally well. 



So many cases of the apparent conflict between stratigraphical and morphologi- 

 cal facts have been removed by further investigation, that we may confidently expect 

 the same of this. 



