1907] Beebe: Geographic Variation in Birds. 13 



haps being supplanted by the white type.* In other words, if 

 Chen caerulescens and the dark phase of Gallinago gallinago 

 were bred in captivity from C. hyperboreus and from normally 

 colored Gallinago, we should be inclined, in the present state of 

 our knowledge, to consider the former an atavism, the latter a 

 recent variation. 



I have taken this theoretical point up in some detail to show 

 the possibilities of species-formation from distributional varia- 

 tion — on the one hand (Gallinago) a humid phase evolving in a 

 restricted locality from a widely spread typically colored species, 

 and in the other case (Chen) a form, local at least in its present 

 distribution, perhaps immediately ancestral to a lighter type of 

 bird, which, in one or the other of its two intergrading subspecies 

 (C. hyperboreus and C. h. nivalis), is circumpolar. 



Examples which have apparently attained the final step in 

 species demarcation, but which are separated specifically by only 

 the single character of a difference in color, are the white and 

 the scarlet ibises Guara alba and G. rubra, and the great white 

 and the Ward heron Ardea occidentalis and A. herodias wardi. 

 These are said to be specifically identical in all characters except 

 the pigmentation of their feathers. 



Ardea rufescens, the chestnut and bluish colored reddish 

 egret and Ardea pealei, a pure white bird, were thought to be dif- 

 ferent species until the discovery of their interbreeding proved 

 them to be only two phases of the same species. In this case, 

 intermediate, particolored birds are not uncommon. 



We shall pass over many other interesting cases of dichro- 

 matism in birds, and mention only Falco sparverioides, the 

 Cuban sparrow hawk, which exhibits two very distinct color 

 phases. I have had two living adult birds, one in the light and 

 one in the dark phase, under observation for two years, having 

 received them when they had not yet shed all their nestling 

 down. With each succeeding moult there has been a more and 

 more sharp demarcation between the coloration of the two types. 

 The crown is the same color in both birds, but the pure cinnamon 

 back of the light phase is so encroached upon by dark blue 

 in the other bird, that the cinnamon is reduced to irregular and 

 broken bars, present on only some of the feathers. In the dark 

 bird the black area on all the wing feathers is increased and 

 the parts of these feathers which are white in the light bird are 



* The white color of the snow goose, homologous as it is with the same phenomenon in 

 most other terrestrial Arctic birds and mammals, must not be confused with albinism to 

 which it is in no way analogous. With this abnormal condition we have nothing to do in this 

 paper. 



