1922] Kellogg: Pinnipeds from Miocene and Pleistocene Deposits 53 



has been currently believed. The differences between Prorosmarus 

 and the Temblor otarid are almost as great as those between modern 

 Otariidae and Odobenidae. 



It seems reasonable to assume from what is known of Prorosmarus 

 that it possessed a skull somewhat similar to that of an otarid, with 

 upper canines or tusks much enlarged. The arrangement of the teeth 

 is very similar to that of the Otariidae excepting that the true molars 

 are absent. The worn anterointernal face of the lower canine leads 

 one to believe that Prorosmarus must have possessed the third upper 

 incisor. This also relates it with the Otariidae. Another point of 

 interest is that the canine still retains its caniniform shape and has 

 not as yet been taken over into the molariform series to form a dental 

 battery. The possession of a long sloping chin, the persistent separa- 

 tion of the rami at the symphysis, and the slender coronoid, together 

 with the facts enumerated above, indicate that this form had not as 

 yet progressed very far in adapting itself to feed upon mollusks. 



The next stage as revealed by fossil material is comparable to 

 Alachtherium antverpiensis, which marks an intermediate stage in the 

 evolution of this family. The most noticeable modification in this 

 form of the otarid type of skull lies in the foreshortening of the 

 rostrum accompanied by the development of a very large crista 

 lambdoidea and a big mastoid process. The arrangement of the 

 molariform series in a curved line and the possession of three upper 

 incisors in the young and two in the adult Alachtherium reveals how 

 closely the walruses had up to this time retained many of the features 

 of the otarid skull. A marked increase in the size of the tusks was 

 undoubtedly one of the earliest steps in the modification of the otarid 

 skull to the odobenid type. 



Alachtherium cretsii resembles Prorosmarus in the retention of the 

 second and third lower incisors, but the ramus as a whole is much 

 heavier. The worn anterointernal surface of the lower canine indicates 

 that this species also possessed a third upper incisor. Thus the reduc- 

 tion of the upper incisors was brought about in some more advanced 

 type during the evolution of the Odobenidae. The mandibular ramus 

 of Alachtherium cretsii as figured by Van Beneden 48 belonged to a 

 very large animal, too large to be considered in the main line of 

 descent. It may either belong to a very old individual or may repre- 

 sent an aberrant offshoot of some earlier form. It is highly probable 



*8 Van Beneden, P. J., op. cit., Atlas, pi. 1, figs. 2, 3 ; pi. 2, figs. 1, 2, 4. 1877. 



