488 University of California Publications in Zoology [Vol. 12 



bia is here characterized as a uew subspecies, Castor canadensis 

 belugae. 



(2) The race of beaver inhabiting the Pit River region of Cali- 

 fornia, east of the Sierra Nevada, is here characterized as a new 

 subspecies, Castor subauratus shastensis. 



(3) The race of beaver inhabiting the mainland of Washington 

 and Oregon is shown to be subspecifically distinct from the form 

 on Vancouver Island (Castor canadensis leucodonta Gray) requir- 

 ing the rehabilitation of Rhoads' name pacificus. 



6. The fragmentary palaeontologic history of the Castoridae 

 points to at least three intermigrations of beavers between the Old 

 World and the New. Steneofiber appears in the middle Oligocene 

 of Europe, and in the upper Oligocene of North America. The 

 Eucastor-Dipoides stock appears in the upper Miocene of North 

 America, and in the Pliocene of Asia. Castor first appears in the 

 Pontian upper Miocene of Europe, but in North America it does not 

 appear until later, its first known occurrence being in the Pliocene 

 of California. In the case of each of these genera, at least one 

 intercontinental migration is indicated. 



7. The beavers of North America (those of which material has 

 been available) are separable into two groups: the canadensis 

 group, with subspecies canadensis, michiganensis, belugae, leuco- 

 donta, pacificus, frondator, and texensis; and the subauratus group, 

 with subspecies subauratus and shastensis. The forms making up 

 the canadensis group are unequally related. 



8. On the whole, differentiation in the genus Castor is slight. 

 The California stock has undergone more divergence than any other. 



9. A consideration of its history and present status shows that 

 the widest diversity of opinion as to the isolation concept still pre- 

 vails, and emphasizes the necessity for more critical enquiries con- 

 cerning it. 



10. The evidence from beavers, and higher vertebrates gener- 

 ally, regarding Weismann's Amixia, Romanes' Independent Varia- 

 bility, and Gulick's Independent Generation, is inconclusive. In 

 the case of beavers it must be conceded, however, that the possi- 

 bility is by no means excluded that geographical isolation alone 

 has been the chief condition in speciation. 



11. The bulk of the evidence from the zoogeography of beavers 

 and other vertebrates seems to indicate that polytypic evolution in 

 these groups has been conditioned by the spatial partition of the 



