1910] Grinnell.—Birds: Alaska Expedition, 1908. 427 
swallows, hummingbirds, eagles, fish hawks, and some water- 
birds respond indifferently in this regard. Such ‘‘protective”’ 
coloration would have been acquired through gradual increments 
from individual variation, by the action of natural selection. 
But this theory is discountenanced by many on grounds it is 
not here necessary to discuss. That the light itself has no direct 
effect on the colors of birds at the time of pigment deposition 
in the forming feather is shown by the fact that the character- 
istic depth of coloring is acquired by young birds hatched and 
fledged within dark holes of trees, as woodpeckers and 
chickadees. 
In well-known regions there is often much difference in 
climate from year to year, one year showing a much ereater 
or lesser amount of cloudy weather than the next. There is 
nothing to show that relevant species of birds in the same region 
become darker during cloudy years. The characters are con- 
stant from year to year. The melanism is an inherited char- 
acter, not modified during the life of the individual, except 
because of age by means of molt, and by such adventitious 
agencies as abrasion and bleaching. 
The effect of temperature on color in birds is not indicated 
by any evidence that I know of. The conspicuous difference in 
temperature affecting sedentary birds in the Yukon and Sitkan 
districts, respectively, consists in the great annual range of 
temperature in the former (57° C. in one known ease) as com- 
pared with the latter (24°). This uniformity of temperature 
accompanies humidity everywhere, but I can not see in any case 
where it alone is accountable for any color characters. 
The matter of size is a less conspicuous differential feature 
of the birds of the interior and coastal regions under considera- 
tion, than melanism. Yet in many plastic groups, the repre- 
sentatives in the Yukon district and to a slightly less extent 
those in the Prince William Sound district, are decidedly 
larger throughout. The woodpeckers, kinglets, thrushes, and 
certain finches exhibit this peculiarity. No satisfactory explana- 
tion is known to me. The relatively shorter, less pointed wings 
and shorter tails of the migratory coastal birds appear to be 
correlated with shorter semi-annual movements. Many of the 
