EVOLUTION OF THE COLOES OP BIRDS. 139 



ment is a chemical composition thrown off from the 

 system of the bird, probably as a product of waste, and 

 lodged in the integument. The chemical substance thus 

 generated varies in different groups of birds, but is 

 probably generally constant in the species of one genus, 

 or frequently in an entire family. A certain genus 

 would thus be capable of generating only a given num- 

 ber of fundamental colors, but natural and sexual 

 selection by combining and rearranging this limited 

 assortment, can produce a great variety of effects. Be- 

 sides black and white, most North American genera 

 appear to contain but two fundamental colors, but in the 

 tropics, three frequently occur in a single genus. If 

 the pigment be regarded, to some extent at least, as a 

 product of waste due to superabundant vitality, as Wal- 

 lace maintains, it is not difficult to understand how 

 among some tropical birds a, third chemical product 

 might be generated in addition to the usual two. As an 

 illustration of this view, let color be compared to a 

 variety of skin diseases. Each genus of birds in gen- 

 eral, we may say, developes two kinds of skin disease 

 which are bound to be in the system of every species. 

 One of them may sometimes be suppressed or they may 

 be combined in such a way as to be no longer recogniz- 

 able, but they will still be present. 



This theory could not be demonstrated without further 

 study of the chemical properties of pigment. If it 

 could be shown that in closely allied species the color, 

 although the same, was produced by different pigments; 

 or that in the same or closely allied species, where one 

 color gave the appearance of being a combination of two 

 other colors which occurred in the same or allied species, 

 there was no real connection in the pigments, chemi- 

 cally; the theory would certainly become very weak. 

 Until such experiments have been made, however, it is 



