COLORATION 85 



certain circumstances is very valuable indeed as a contri- 

 bution to coloristic law. But his application of it as being 

 the prime and well-nigh all-important factor in concealing 

 coloration, produced through the "omnipotence" of natu- 

 ral selection, is wholly untenable. Nor is there warrant 

 for his assertion that the brilliant white bellies of so many 

 animals are the result of natural selection working toward 

 concealing coloration. Most, although by no means all, 

 mammals and birds are countershaded, in the sense of being 

 of lighter color below than above. But this is also gener- 

 ally true of the most plentiful of all vegetable productions, 

 leaves, and of many fruits. There is no reason to suppose 

 that concealment has been the reason for the development 

 in one case more than in the other. A pure white belly 

 cannot aid in concealing an animal when that belly is itself 

 concealed. There are snakes with white or yellow or light- 

 colored belHes; but as these bellies are themselves con- 

 cealed they have no effect in concealing the wearer, and 

 cannot have been developed by selection or otherwise for 

 such a purpose. The same is true of the myriads of shrews 

 and meadow mice with legs so short that the bodies, or at 

 least the hairs on the bodies, practically touch the ground at 

 all times when the animal is at rest. In these cases the light 

 belly can have no effect whatever in concealing the ani- 

 mal; yet when the sides are lighter than the back the 

 belly is almost always lighter than the sides. If concealing 

 coloration was the object, and natural selection the agent, 

 the belly, which is not seen and the coloration of which is 

 of no concealing service, would not have been lighter than 

 the sides. When in motion a chipmunk or rabbit is visible 

 anyhow, and the glimpses occasionally obtained of the color- 



