376 PROCEEDINGS OF THE NATIONAL MUSEUM vol. ii4 



subspecies described as incana by Godfrey will have to be considered 

 a sjmonyTQ of it. 



From the Athabaska River eastward, the population exhibits a 

 reddish-olive coloration dorsally, noticeably richer and warmer than in 

 the Alaskan and Rocky Mountain group. I have not been able to 

 separate specimens taken in the Atlantic maritime area (H. u. clare- 

 scens of Burleigh and Peters) from birds of the interior even on a 

 conservative 75 percent basis (Rand and Traylor, 1950). For that 

 reason I believe that H. u. clarescens Burleigh and Peters must be 

 considered a synonym of H. u. swainsoni (Tschudi). 



Topotypical material from the exact type locality of swainsoni is 

 unavailable, but specimens from Cypress Hills, Fort Chipewyan, and 

 the Athabaska River strongly indicate that the breeding population 

 from Carlton House, Saskatchewan, is not the grayish race of the 

 mountains and foothills to the west. 



Birds from California and southern Oregon northward along the 

 river bottoms, slopes, and ravines east of the Cascades to northern 

 Washington are definitely paler and less rufescent than the russet- 

 backed thrushes of the northwestern coastal forests; the name H. u. 

 oedica Oberholser, therefore, should be used for this population. 

 This usage will automatically restrict the nominate race H. u. usfulata 

 (Nuttall) with type locahty Fort Vancouver, Wash., to the area west 

 of the Cascades in northwestern Oregon and thence northward along 

 the coast to Juneau, Alaska. 



Specimens from the Warner Mountains of California are unavail- 

 able to me, but earlier authors (Grinnell and Miller, 1944) referred to 

 this section of the state as an area of intergradation between the 

 California race of Swainson's thrush and a darker or grayer race. 

 It seems reasonable to believe that the latter is H. u. almae. 



This paper is based in the main on study of 442 specimens in the 

 United States National Museum. Additional material was borrowed 

 from other museums. The sources and numbers of these additional 

 specimens and the abbreviations used in designating collections in 

 the text are: Cornell University (cu), 3; National Museum of Canada 

 (nmc), 37; Geological Survey Museum of Canada (gsmc), 12; Victoria 

 Memorial Musemn (vmm), 3; and the Ira N, Gabrielson collection 

 (ing), 16. All specimens not otherwise designated are from the U.S. 

 National Museiun. 



Prolonged study of the material showed that the only vaUd cri- 

 terion for separating the several races in question is the color of the 

 upperparts. Ventral coloration is mentioned in the hterature as 

 being of racial significance, but I have found that it is not reliable in 

 deUmiting the races. Wing, tarsal, and culmenal measurements taken 

 of samples throughout the population showed no significant differences. 



