288 University of California Publications in Zoology ['Voh. 24 



selections can be made of closely similar specimens in the two series. 

 The variation in dorsal coloration is, I believe, at least partly due to 

 different degrees of wear and fading of the feathers. Specimens of 

 the two series that were taken at approximately the same dates are 

 practically indistinguishable. Measurements of the three lots are 

 given below. They are of little diagnostic importance. 



In my opinion the Stikine River birds should not be classed with 

 rubiginosa. They present certain differences in appearance; and in 

 habitat and migration they are as sharply set off from the coastal 

 birds as is the case with practically all the other bird species of the 

 region where they were taken. On our trip down the river I watched 

 for yellow warblers carefully and purposely, and they are rather 

 more conspicuous in migration than most small birds. The frequently 

 uttered zip-zip, as they fly from bush to bush, attracts attention, and 

 the yellow color is easily seen. At Doch-da-on Creek, at Flood Glacier, 

 and at Great Glacier, yellow warblers were in evidence, obviously 

 migrating, but below Great Glacier I could not see that there were 

 any traveling down stream. My belief is that they find an outlet 

 through the mountains to the southward, and that they do not reach 

 the coast. When we reached Sergief Island, in the habitat of the 

 coastal race {Dendroica aestiva rubiginosa), yellow warblers abruptly 

 became scarce. "We saw very few, whereas this island was directly in 

 the line of travel down the river. During the first two weeks of our 

 stay, the latter part of August, none was seen. On September 1 sev- 

 eral were noted in company with other migrating warblers, and on 

 September 2 a single bird wa* observed. 



On both scores, of physical characters and of distribution, I regard 

 the yellow warblers of the upper Stikine River as best referred to 

 Dendroioa ■aestiva aestiva, showing some variation toward rubigi- 

 nosa. Females and immature birds do not show the subspecific char- 

 acters as do the adult males, and such of the former as were collected 

 at Flood Glacier and Great Glacier are ascribed to a&stiva on the 

 grounds of probability. 



TABLE XI 



Measurements in millimeters (average, minimum and maximum) of Dendroica 



aestiva aestiva and Dendroica aestiva rubiginosa 



Wing Tail Culmen 



Dendroica a. aestiva 10 males' 62.3 (60.0-64.0) 43.9 (42.0-46.0) 9.6 (9.0-10.0) 



Dendroica a. aestiva 9 males^ 62.4 (60.5-64.0) 44.5 (42.5-46.5) 9.3 (8.5-10.0) 



Dendroica a. rubiginosa 10 males' 61.5(59.5-64.0) 43.6(42.0-46.0) 9.7(9.0-10.0) 



^Ontario, Canada, 4; Wisconsin, 4; Michigan, 2. 



^Telegraph Creek, 7; Glenora, 2. 



'Prince William Sound, Alaska, 2; Sitka, 1; Chickamin River, s.e. Alaska, 3; Vancouver Island, 4. 



