368 THE FUR SEALS OF THE PEIBILOP ISLANDS. 



variations in the scattered noncommunicating members of each species because of 

 their different environments. That differentiation increased in time, but varied in 

 amount in the different groups for various reasons. Some of these divisions how 

 constitute good species, while others are only considered worthy of ranking as 

 subspecies, while of still others opinions differ as to the sufficiency of causes for 

 separation as species or subspecies or even for separation at all. This explains the 

 differences between such forms as Somateria mollissima and dresseri, and S. v-nigra, 

 Arenaria interpres and A. morinella, and Larus glaucus and L. barrovianus. This 

 differentiation is correlated by the time which has elapsed since the separation began 

 ^nd the differing environment, but the species or subspecies also vary in amount of 

 difference by the effects of another factor, the degree of communication possible during 

 the early or intermittent stages of the separation. The generally so-called circumpolar 

 species are really not circumpolar except generically, the amount of our ignorance 

 being more than sufficient to fill up the measure of our knowledge of what constitutes 

 a usually so-called circumpolar species. A good example is the case of the turnstoue. 



Thus it would appear probable that western Alaska and the Bering Sea islands 

 are the remnants of former land areas originally connected with or narrowly sepa- 

 rated from the Palaearctic continent, the differences now observable as to resident 

 faunal and floral life being due to greater or less volcanic action and greater or less 

 glacial influence in affecting and isolating that life. And the same is true of North 

 America as a whole. It is now essentially Nearctic, with a very strong double intrusion, 

 pre and post glacial (of Palaearctic derivation), from the northeast and northwest, 

 consequent on glacial and volcanic action, destruction, and dispersion. Where these 

 opposite types meet the modifying influences of the varied environments have effected 

 results tending to bridge over the gaps, thus producing transitional forms. Thus the 

 retreat of the ice permitted the reextension northward of Nearctic types, but con- 

 tact with boreal influences differentiated these frontier forms, so that we now find 

 them generically and specifically distinct from their nearest relatives. In less degree 

 have the southern outliers of northern forms differentiated. 



It would consequently seem from this discussion that from the elements of the 

 Bering Sea avifauna and vicinity are deducible several zoogeographical provinces and 

 subprovinces, and that the whole constitutes a division of the Holarctic region. 

 Professor Newton's term Alaskan can be retained for the region north and east of 

 the Aliaska Peninsula. This peninsula, with the islands adjacent on the south and the 

 mainland to British Columbia, may retain Mr. Nelson's name of Sitkan. It is properly 

 a true transitional subprovince of the Nearctic. The Aleutian Islands, the islands 

 of Bering Sea, and much, perhaps all, of the mainland coasts of Alaska and north- 

 eastern Asia to the Arctic Sea constitute a single subregion to which the name 

 Aleutican is more properly applicable. 



These views are necessarily somewhat suggestive and void of details, but seem 

 pertinent to a consideration of the avifauna of the Pribilof Islands. Unfortunately 

 little is known concerning the exact distribution of many of the forms, and the col- 

 lections that have been madein the region are so widely scattered that it is impossible 

 to bring them together for comparison. 



