1922] Swarth: Birds and Mammals of the SUkine Region 251 



most part these birds are essentially like the Stikine . specimens. One 

 male (Prov. Mus., no. 3683) is much paler colored than the others, 

 more of the slaty gray of eastern hyemdlis, a variation that may be 

 indicative of intergradation with hyemmlis in Alaska, a little to the 

 northward. One female of the lot (Prov. Mus. no. 3698, Wilson Creek, 

 Atlin, B. C, June 29, 1914) is so very different in appearance from 

 the rest of the series as to suggest its belonging to another species. 

 It is appreciably larger than the others, actually so in length of wing 

 and tail, and to all appearances in bulk also, as well as can be judged 

 from prepared skins. It is faintly brownish on the back, and with a 

 suggestion of pink on the sides. This bird may, perhaps, be an 

 example of m.ontanus, which has been recorded as occurring at Alaskan 

 points not very far distant. 



Altogether, the scanty material from Alaska, together with the 

 British Columbian series, points to the restriction of the race connec- 

 tens to a region mostly south of Alaska, and probably almost entirely in 

 British Columbia. In this connection, attention should be drawn to 

 a record by Bishop (1900a, p. 86) of the occurrence of "Junco liye- 

 malis connectens, Shufeldt Junco," at "White Pass City and Glacier, 

 Alaska, a record that may, perhaps, pertain to connectens as here 

 regarded. Osgood (1909&, p. 41) records Junco hyemaUs montanus 

 from points on the upper Yukon: "An adult female having vinaceous 

 sides and pronounced of this form by Mr. Ridgway was collected by 



HoUister at Circle July 7 Also a specimen collected by myself 



near Charlie Creek in 1899 appears to be of this form." 



When we arrived at Telegraph Creek, May 23, juncos were abun- 

 dant and in pairs; evidently nesting was well under way. The male 

 birds at that time were fond of perching on some elevated position, a 

 telegraph pole or a house in town, a dead tree top out in the woods, 

 where, over and over again, they gave utterance to a brief, monotonous 

 trill, hardly long enough or loud enough to merit being termed a song. 

 The first nest was found on June 4, about midway between Telegraph 

 Creek and the Summit, a region in which there was still much snow. 

 This nest was placed in an overhung crevice in the dirt wall bordering 

 the trail, and was discovered through the brooding bird's sudden de- 

 parture. It contained four eggs with incubation just begun (see 

 fig. Z). From then on, nests were found at intervals up to July 5. 

 On that date two were discovered at Glenora, one with four fresh eggs, 

 one with four young about five days old. 



