256 University of California Publications in Zoology ["Vol. 24 



breeding grounds for song sparrows. None of the birds was seen at 

 Great Glacier but one was noted near the custom house just above the 

 boundary, August 16. 



At Sergief Island song sparrows were abundant. They were mostly 

 at the upper edge of the marsh land, just below the coniferous forest, 

 where a bordering strip of dense underbrush (alder, willow, devil's- 

 club, and other shrubbery) afforded shelter to many birds of similar 

 habits. The song sparrows ventured some distance out onto the 

 marshes, but always along narrow tidal channels where tall grass over- 

 hanging the sloughs made arched passageways beneath which the 

 birds found concealment when need arose. 



A large proportion of those seen at Sergief Island were young birds, 

 and young and old were mostly in various stages of molt. A young 

 male taken August 18 is in first winter plumage throughout. Most of 

 the young birds seen at that time were still in juvenal plumage; one 

 taken as late as September 1 had not yet begun the molt from that 

 stage. Two adults shot August 18 and 22, respectively, had not yet 

 begun to molt; another taken on September 1 had almost completed 

 the change. 



Forty-four specimens were collected (nos. 40020-40044, 40047- 

 40065), as follows: Telegraph Creek, three adult males, two adult 

 females, one juvenal female ; Doch-da-on Creek, three adult males, ten 

 juvenals; Flood Glacier, four juvenals; Sergief Island, two adult 

 males, one adult female, five immature males (first winter plumage), 

 thirteen juvenals. 



I had anticipated that the song sparrow of Telegraph Creek would 

 prove to be different from th# coast bird, and that it would probably 

 be Melospiza m. inexpectata, as a specimen ascribed to that race was 

 recorded from Telegraph Creek by Eiley (1911, p. 234) in his descrip- 

 tion of the subspecies. I am, however, unable to distinguish any 

 points of difference between birds from the upper Stikine Valley 

 and those from the coast. They seem to me to be all referable to rufina. 



Upon first consideration it seems strange that in two regions with 

 such diverse faunas in general, as is the case in the upper Stikine 

 and the coast regions, the song sparrow, usually so variable, should 

 be one of the few birds to remain the same in both places. This is by 

 no means inexplicable, however, upon consideration of some of the 

 factors involved. The species Melospiza melodia extends much farther 

 north along the coast than it does inland ; there is no subspecies of the 

 interior that ranges nearly as far north as does rufitia of the coast. 



