1922] Kellogg : Pinnipeds from Miocene and Pleistocene Deposits 33 



It is very hard from the material at hand, consisting' mainly of 

 imperfect premolars, to assign them to any particular place. The 

 illustration of the premolar (fig. 6) represents a much younger indi- 

 vidual than the ones in situ in the mandible. The second lower pre- 

 molar (fig. 7) was slightly larger than the first, with conical crown 

 and poorly defined cingulum. The premolars are all haplodont, no 

 doubt secondarily, without any trace, so far as could be ascertained 

 with the material available, of either an anterior or posterior accessory 

 cusp. There is a progressive increase in size of the premolars from 

 the first to the fourth, though the discrepancy in size of the first and 

 fourth is conspicuous. The roots of Pi% and Pm. ? are grooved both 

 externally and internally, though Pm T and Pi% appear to possess 

 the groove on the external side alone. The roots completely fill the 

 alveoli and approximate each other very closely. 



Fig. 6. Allodesmus Icernensis n. gen. and sp. Premolar, no. 275, Calif. Acad. 

 Sei. coll., X 1. Temblor formation near Bakersfield, California. 



Fig. 7. Allodesmus kernensus n. gen. and sp. Premolar, no. 275, Calif. Acad. 

 ScL coll., X 1. Temblor formation near Bakersfield, California. 



Figs. 8a, 8ft. ? Allodesmus kernensis n. gen. and sp. Milk teeth?, no. 275, 

 Calif. Acad. Sei. coll., X 1. Temblor formation near Bakersfield, California. 



The first lower molar as indicated by the alveolus (fig. 1&) is twice 

 the size of the second lower molar. It is characterized by an internal 

 as well as an external groove. There appears to be a progressive 

 lengthening of the alveoli from Pm T posteriad to M T . The broken 

 roots of M T in the mandible indicate that it was very well developed, 

 but was very small, tending to disappear, as we know it did during 

 late Miocene or Pliocene. There is no indication of M3, but when we 

 consider the size of M^, it seems probable that M3 had become vestigial 

 at an even earlier stage, and that now Mtt in turn had begun to 

 approach the same fate. 



6 



8a 



7 



