62 



The range during summer of several of the forms is not so well un- 

 derstood. The exact breeding- ground is known only of L. amtralis 

 and L. griseinucha ; that of the latter being the Aleutian Islands 

 and neighboring mainland of TJnalashka, and of the former the alpine 

 summits of the Colorado peaks. Regarding the others, we can only con- 

 jecture where they breed from what is known of their distribution at 

 other seasons, and the time and direction of their migrations. We 

 can guess more nearly in the case of L. tephroootis var. tephrocotis 

 than any other. It was obtained by Mr. Aiken in Colorado in April, 

 during its northward passage, already in possession of the summer-attire. 

 In May, Richardson found it on the Saskatchewan, while Mr. J. K. Lord 

 observed in July what was probably this race on the summit of the Rocky 

 Mountains, near the Kootanie Pass, just south of the northern bound- 

 ary of the United States. At that locality several young-of-the-year were 

 shot. We therefore judge that the breeding-ground of var. tephrocotis 

 is the Rocky Mountain region of British America, and perhaps the ex- 

 treme northern Rockies of the United States. The var. littoralis occurs 

 in winter with var. tephrocotis throughout its southward range; its 

 abundance increasing to the westward, however. We have seen it in the 

 winter-plumage from Kadiak and Sitka of the Alaskan coast, from Fort 

 Simpson, British Columbia, Idaho, Wyoming, and Colorado ; while it is 

 known to extend in very severe winters as far southward as the paral- 

 lel of 39° in the Sierra Nevada and contiguous ranges to the eastward. 

 On the Cascade Mountains, Mr. J. K. Lord found them in October. In 

 the summer-plumage, this race has been obtained only in Colorado, where 

 Mr. Aiken shot it in flocks of tephrocotis as they were passing to the 

 northward. From the fact that it has been found in winter along the 

 northwest coast from Kadiak southward, it is extremely probable that 

 the breeding-ground of this race is the coast-system of mountain-ranges 

 of southern Alaska and British Columbia, possibly extending southward 

 to the alpine summits of the Cascades of Washington Territory and 

 Oregon. The L. atrata has been obtained at only two localities, the 

 Uintah Mountains, in September, and the mountains of Colorado, in 

 April ; it is therefore probable that the breeding-ground of this species 

 will be found in the northern Rocky Mountains of the United States, 

 or possibly an associate of L. tephrocotis at the southern limit of the lat- 

 ter, and of L. australis at its northern limit. It is not known how far to 

 the southward L. australis breeds, but probably on all the peaks of the 

 southern Rockies whose summits rise to above the timber-line. 



Mr. Aiken's notes pertaining to the distribution of these birds are of 

 interest, and are herewith appended: 



"I think that none of australis were seen before March, and then only 

 a few. In February, about 75 per cent, of all the Leucostictes were 

 tephrocotis, and the remainder the gray-cheeked varieties.* In the flock at 

 Canon City in April, I judged that about 1 per cent, were atrata; 30 per 

 cent, tephrocotis; 8 or 10 per cent, littoralis; and the remainder australis. 

 In this flock, the males and females of the last-named were about equal 

 in number; but of tephrocotis the number of females was double that of 

 the males. From these facts, and information derived from other sources, 

 I infer that the gray-cheeked variety {littoralis) is the most northern 

 race, and that many of them do not find their way so far south except 

 in severe winters. In this belief I am strengthened by the fact that, of 

 sixty birds killed in Wyoming in 1870, all but one or two were typical 

 tephrocotis; that tephrocotis occupies, during the breeding-season, a more 



* Littoralis a,ni '^ campestris" {— littoralis). 



