1916J Taylor: Beavers of Western North America 463 



of beaver distribution, with special reference to the problem of specia- 

 tion in the group. 



It should be remembered that the beavers of Eurasia and America, 

 while not distantly related, are still very clearly marked off spe- 

 cifically. It should further be called to mind that there are, on the 

 North American continent, two distinct groups of beavers, the cana- 

 th nsis group, including the species canadensis with at least its sub- 

 species canadensis, belugae, leucodonta, pacific us, frondator and tex- 

 ensis, and the subauratus group, including the species subauratus 

 with its subspecies subauratus and shastensis. 



In the case of each one of these subspecies of beavers, geographic 

 rauge seems to be as characteristic as any physical attribute. It 

 should be emphasized that differentiation in beavers has progressed 

 only to a slight degree as compared with, for example, the Geomyidae, 

 and that all the forms of beavers, even the two groups, are relatively 

 closely related. No case of overlapping of subspeeific or specific 

 ranges is known. Knowledge of the details of the relations of dis- 

 tribution of beavers to topography is not sufficient to warrant cate- 

 gorical assertions without some qualification, but available evidence 

 points to the truth of the following general statements. 



Castor c michiganensis, a small, dark form, is very closely related 

 to C. c. canadensis, its neighbor on the north. The type locality of 

 canadensis is Hudson Bay, and the subspecies probably ranges south 

 to the chain of Great Lakes, which would seem to be a barrier separat- 

 ing it from michiganensis. 



The type locality of Castor c. canadensis (Hudson Bay) is con- 

 nected by a nearly continuous series of streams and lakes with extreme 

 western Canada. Distance is practically the only deterrence to the 

 migration of individuals from eastern Canada to the base of the 

 Canadian Rockies and the Alaskan Mountains. Subspecies belugae, 

 of canadensis, occupying the Pacific territory from Cook Inlet to 

 British Columbia, is closely related to canadensis on the one hand and 

 to leucodonta and pacificus on the other. Subspecies phaeus of Ad- 

 miralty Island, southeastern Alaska, is comparatively sharply marked 

 off. Frondator, type locality San Pedro River, Mexico, which is sup- 

 posed to range some distance to the north, is also closely related to 

 canadensis. On the other hand, subauratus and its subspecies shas- 

 ti nsis are much less closely related to canadensis than any of the other 

 forms, and there do occur, bounding the range of the subauratus 

 group, masses of high mountains on the north, east and south, as 



