no. 8637 HERMIT THRUSH — ALDRICH 19 



Islands. In fresh plumage these have a distinct buffy wash on the 

 relatively dark gray flanks. The small size, particularly the bill, 

 which was stressed by Audubon (1839) in his description of nanus, 

 also would seem to eliminate all except the small western races, and 

 of these the only one with a buffy tinge to the flanks is the Queen 

 Charlotte Island population. Audubon said that the haunts of his 

 newly described nanus were the valleys of the Columbia River whence 

 he had obtained it through Dr. Townsend. This further indicates 

 the probability that the description of nanus was based on a western 

 specimen. Audubon mentioned a few very small specimens from 

 eastern localities that he referred to nanus. These were probably 

 also representatives of far western populations, although not neces- 

 sarily the Queen Charlotte Island one, which seems to stick more 

 closely to the Pacific Coast than some of the other small hermit 

 thrushes such as guttatus and oromelus. Specimens of both these 

 races from the Atlantic seaboard have been seen in connection with 

 the present study, but neither has any buffy tinge to the flanks and 

 so could not have been the basis for Audubon's plate. 



Since there is some question (see Phillips, 1962) about the authen- 

 ticity of the presumed type-specimen of nanus designated by Brewster 

 (1902) and located in the Museum of Comparative Zoology at Har- 

 vard, it seems prudent to rely most heavily on Audubon's original 

 description and the illustration that was mentioned in it. On the 

 basis of these, I believe the correct course is that set by Ridgway 

 (1907), Hellmayr (1934), and many others in assigning the name 

 nanus to the Queen Charlotte Island population on the basis of 

 Audubon's description of a migrant specimen taken near the lower 

 Columbia River. 



Ridgway (1907) seems to have correctly distinguished the dis- 

 tribution of nanus from that of guttatus in Alaska, with the former 

 restricted to the Queen Charlotte Islands and the outer islands of 

 the Alexander Archipelago. Specimens examined in the present study, 

 however, indicate that his postulated extension of the range of nanus 

 southward to include the coast of British Columbia and Washington 

 is without basis. 



Munro and Cowan (1947) noted that the population of the Queen 

 Charlotte Islands is typical nanus and, although extending its range 

 south to the mountains of the Vancouver region, noted that the 

 population of that area exhibits characters of the inland race, oro- 

 melus. This accords with findings in the present study but, unlike 

 Munro and Cowan, I consider the populations of southwestern 

 British Columbia too different from the Queen Charlotte birds 

 to include under the name nanus, and I have included them under 

 vaccinius. 



