COLORATION 95 



ment; a wood-thrush or hermit- thrush Is more difficult to 

 see (as is the case with a meadow-lark in the open) when 

 it sits with the nearly uniform dulness of its back toward 

 the observer — although there is then no effect of counter- 

 shading — than when it sits sideways, when there is an effect 

 of countershading but when this effect is more than offset 

 by the brightness of the white under-parts. As regards 

 these birds it is possible that natural selection has played 

 a certain part, not in producing a given pattern, but by 

 eliminating the birds that were brilliantly or advertisingly 

 colored. 



In the infinitely varied and brilliantly lighted landscape 

 of the tree tops, the coloration of the bird becomes so 

 much less important that selection ceases to eliminate the 

 brilliant hues. That selection does not in these circum- 

 stances produce the brilliant hues, as Mr. Thayer fancies, 

 is evident from the fact that many of the birds, especially 

 the females, which most need protection, remain dull-col- 

 ored; the females of the scarlet tanager and redbird are 

 no more conspicuous than the thrushes which dwell in the 

 comparatively sombre region beneath them. Of course, if 

 the brilliant hues of the males were developed by natural 

 selection for concealing purposes, the females would also 

 have developed them. The fact simply is that the bril- 

 liant colors have been developed in spite of their adver- 

 tising quality. 



What happens is analogous to what Mr. Reighard has 

 described in his "Experimental Study of Warning Colora- 

 tion in Coral Reef Fishes"; he shows that certain fishes, 

 whose habits and habitat render concealing coloration of 

 no consequence to them, develop brilliant, highly advertis- 



