400 - University of California Publications in Zoology. (Vou. 
collectors from degree of ossification shown by the skull, so that 
lacking this criterion I am unable to distinguish first winter 
from subsequent winter plumages. Suffice it to remark that 
there is considerable range of variation regularly graded from 
a relatively grayish bird (as in no. 1503) to a dark brownish 
bird (as in no. 1492). The latter, a male taken August 30 on 
Chenega Island, has the back with black streaked feathers 
broadly margined with deep hazel and laterally margined with 
clay color; the breast, flanks, and the whole sides and top of 
head are strongly pervaded with clay color. In another ex- 
ample (no. 1507) the anterior part of the superciliary stripe is 
distinetly yellow, though in most of the birds this marking is 
faint. Since these extreme dark birds are in the minority and 
appear not to be peculiar to any particular locality, I cannot 
but consider them co-subspeecifie with the rest of the series from 
Prince William Sound, and to be accounted for as. representing 
age, sex, and individual variants combined. Individual varia- 
tion appears sometimes to be widest in intermediate regions. 
The characters of individuals occupying an intermediate region 
are more unstable. It is more remotely possible that these dark 
individuals may really be examples of savanna, that is, invaders 
from the Sitkan district. 
Only two out of the series are in juvenal plumage: no. 1502, 
from Port Nell Juan, August 16; and no. 1509, from Thompson’s 
Pass, September 2. The latter must have been hatched very 
late, as the juvenal plumage in Passerculus is of short duration. 
However, the change was imminent, as it shows many first 
winter feathers appearing in the wing-coverts, dorsum and 
flanks. Finally it should be remarked that most of the points 
visited, both insular and on the mainland, are represented in 
the series of thirty-eight specimens; and that I can see little 
evidence of variation in this material to indicate the existence of 
subsidiary differentiation centers. 
The species was recorded from the following localities: Cor- 
dova, Hawkins, Hinchinbrook, Montague, Green, Hoodoo, Knight, 
Chenega, Dise, Eleanor, and Naked islands; head of Port Nell 
Juan; Valdez Narrows; and Thompson’s Pass, on the Eagle 
Government Trail, being abundant at the last three poimts. The 
