54 



University of California Publications in Geology [Vol. 13 



that Alachtherium eretsii represents a surviving archaic type which 

 disappeared at the close of the Pliocene. 



A further advance in specialization may be illustrated by 

 Tricheeodon huxleyi, as figured by Rutten, 40 which differed from 

 Alachtherium antverpiensis, and possibly from Prorosmarus, in the 

 possession of a shorter and more upturned facial region. In addition 

 to this, the third upper incisor has shifted slightly from its original 

 position and is gradually taking its place in the molariform series. 

 The molariform series has already aligned itself more nearly in a 

 straight line than was the condition in Alachtherium. Thus in this 

 stage of progressive modification of the Odobenidae, many of the pecu- 

 liarities of the otarid skull are already lost. The mandible figured 

 by Van Beneden™ under the name of Tricheeodon koninckii exhibits a 

 further approach to the short-jawed type, with the extension and nar- 

 rowing of the bony lip beyond the canine. In this mandible there is 

 present an abrased surface in front of the molar series which has been 

 considered by some to have lodged an enormous upper canine. This 

 was the interpretation placed upon this cavity in the mandible by 

 Berry and Gregory. 51 It seems more probable, however, that it is 

 nothing more than the alveolus of the lower canine in which the 

 external wall has been broken away. The figures of Van Beneden 

 appear to favor this latter interpretation. The mandible figured is 

 incomplete. 



In Tricheeodon and likewise in Prorosmarus it is evident, accord- 

 ing to the figures of the original describers, that the alveoli are very 

 close together, the septa being very narrow. This same condition is 

 found in the existing genus Odobenus. Alachtherium eretsii, how- 

 ever, has the alveoli separated by somewhat thicker septa. With pro- 

 gressive evolution, as the premolars in turn specialized into a larger 

 and more effective crushing battery in order to avail themselves more 

 fully of the mollusks which abounded about them, there was a cor- 

 responding decrease in the distance between the molars themselves. 

 Of course some of the important types connecting the Odobenidae to 

 the Otariidae are so imperfectly known that the progressive stages 

 cannot be demonstrated from actual material. Yet we possess more 

 complete information on the dentition of the odobenids than we have 

 on the dentition of the otarids. 



*9 Eutten, L., op. cit., figs. 1, 3 and 5. 



so Van Beneden, P. J., op. cit., pi. fi, figs. 5-7. 



si Berry, E. W., and Gregory, W. K., loo. cit. 



