1922] Swarth: Birds and Mammals of the Stikine Region 257 



It follows that any song sparrow colony of an interior point closely 

 adjacent to the coast (such as the Telegraph Creek region) must have 

 been derived from the coast region. 



Song spai^rows in other sections are known to be influenced in their 

 distribution by extent of riparian surroundings suited to their needs. 

 They will follow a favorable water course through otherwise unsuitable 

 surroundings, less regardful of faunal or zonal limitations than of 

 the associations of plants and water. It therefore seems apparent that 

 the occurrence of the song sparrow at the headwaters of the Stikine 

 River represents an outpost of rufina, an overflow of birds that have 

 penetrated to that point by ascending the river from the coast; not 

 part of another race that is of general distribution throughout the 

 interior, tielospiza m. inexpectata may be a recognizable race in the 

 region of the type locality (Moose Lake, British Columbia), in south- 

 eastern British Columbia, and southwestern Alberta (cf. Oberholser, 

 1918a, p. 187) , but the song sparrow of northwestern British Columbia 

 is unmistakably rufina. 



Apparently the song sparrows of the inland regions are not com- 

 pletely enough isolated from the parent stock, or have not been for a 

 sufSeient length of time, to produce distinguishable differences. As 

 to the length of time that they have been there we have no means of 

 knowing. As to the completeness of isolation in the upper Stikine 

 Valley, my impression, derived from the season's observations, is that 

 there is practically continuous distribution of song sparrows along 

 the river. Breeding colonies are doubtless scattered, and perhaps at 

 rather wide intervals, but apparently migrating birds occur at any 

 point. 



There is an interesting point involved in the migration of the song 

 sparrows at the upper end of the Stikine Valley. Do these birds 

 travel north and south to the eastward of the coast range, as is the 

 case with so many other species of that region, or do they follow the 

 Stikine to and from the coast? Our own observations shed no light 

 upon this question. It can hardly be solved save by some person M'ho 

 is permanently residing in the country. 



In this connection it may be noted that Brooks (1912, p. 253) 

 definitely cites the subspecies rufina as the form occurring at Okana- 

 gan, British Columbia, where the fauna otherwise is mostly of the 

 interior. Also, Taverner (1919, p. 84) records rufina as the song 

 sparrow at Hazelton, in the interior of British Columbia, some two 

 hundred miles south of Telegraph Creek. I have seen some of Taver- 

 ner 's Hazelton specimens and agree with him in his determination. 



