I02 Bulletbi American Museum of Natural History. [Vol. X, 



Wasatch. The primitive condition of this oldest type of the 

 Artiodactyla was two buniform external cusps, two buniform 

 intermediates, together with one large more or less lunate internal 

 cusp, flanked by a rudimental postero-internal cusp which is 

 clearly shown to be an outgrowth from the cingulum. There is 

 no evidence that this postero-internal cusp was ever developed on 

 the last molar, because in all the forms from Pantolestcs to Buno- 

 meryx it is persistently absent. The next step consisted in the reduc- 

 tion of the large lunate internal cusp and the full development 

 of a well-marked postero-internal cusp, or hypocone, on the first 

 and second molars. This condition is seen in, and is character- 

 istic of, Homacodon. The third step consisted in the disappear- 

 ance of the true hypocone and the gradual usurpation of its place 

 and function by the posterior intermediate in the crown of the 

 second true molar, a condition seen in Buno/iieryx. As a fourth 

 step in this development one can readily imagine this process 

 extended to the first true molar, when it would be complete. 



This hypothesis may be objected to on the ground that Buno- 

 meryx cannot stand as the direct ancestor of any of the Seleno- 

 donts at present known, on account of its reduced premolar 

 dentition in the lower jaw, but if we are to regard the type of 

 superior molar exhibited by Githtx Homacodon, Dichobune or Helo- 

 hyus as the one which preceded, and from which was derived the 

 tetraselenodont or four-crescented crown, then this hypothesis 

 must be accepted as true. 



The only case so far known wherein the true hypocone has 

 been preserved and has become crescentic, is in Cienotherium and 

 Plesiomeryx^Awd here we have three well-developed crescents upon 

 the posterior moiety of the crown, of which the inner one repre- 

 sents the hypocone and the middle one the posterior interme- 

 diate. It is possible that the cusps of the two anterior superior 

 molars of Xiphodon are to be interpreted in the same way, and 

 that the posterior inner crescent is composed solely of the posterior 

 intermediate, the true hypocone having come to occupy a more 

 anterior and median position. In this case the anterior internal 

 crescent would be made up of protocone and the anterior inter- 

 mediate. Future discovery will no doubt reveal considerable 

 variety in the formation of the internal crescents in the various 



