THE LAW WHICH UNDERLIES PROTECTIVE COLORATION. 479 



The reader, I think, must try these experiments for himself before he 

 can believe that in figures 3 and 9 I tinted the under surfaces exactly 

 as dark as the upper, and no darker. But I beg him to look at any 

 horizontal branch in the woods which is either on the level of his eye 

 or below it. He will see that, although it has exactly the color of its 

 surroundings, it is not in the least concealed, because, being of uniform 

 color above and below, like the birds after I had painted their under 

 sides, it wears that universal attribute of a solid, namely, a gradation 

 of shading from its light side to its dark side. 



I leave to the reader the pleasure of discovering for himself that this 

 principle of gradation in color is almost universal in the animal kingdom. 

 In certain classes of birds and of flying insects, however, the principle 

 gives place, more or less, to the device pointed out by Bates, namely, 

 the employment of strong arbitrary patterns of color which tend to 

 conceal the wearer by destroying his apparent continuity of surface. 

 This makes, for instance, the mallard's dark-green head tend to detach 

 itself from his body and to join the dark green of the shady sedge, or 

 the ruby of the humming bird to desert him and to appear to belong to 

 the glistening flower which he is searching. Yet many other cases of 

 color applied apparently at random conform essentially to the law 

 stated above. The dark patches are on top, the light ones beneath.^ 

 The dark breast mark, so widely used by nature on birds, usually has 

 the effect of putting out a conspicuous and shining rotundity of some 

 bright or light color, as in the meadow lark and the flicker, because it 

 comes just where the breast, in its usual position, rounds upward and 

 faces the sky. The dark collars of the males of most species of duck 

 are absolute counter shading to the light from the sky, when the birds 

 sit in their characteristic positions. For most female ducks nature 

 uses the complete gradation, like that of grouse and saudpiiiers. 

 Ground birds in general, such as grouse, sandpipers, and sparrows, are 

 usually clothed throughout in colors graded according to this principle. 

 But the males of many sj)ecies of pheasant are notable exceptions to 

 this last statement. 



'Now there is still one more very beautiful phenomenon to record. If 

 the animal itself is obliterated by this mechanism of nature, for what 

 useful purpose beyond considerations of sexual selections do his mark- 

 ings exist, since they are not obliterated? The answer is that the 

 markings on the animal become a picture of such background as one 

 might see if the animal were transparent. They help the animal to 

 coalesce, in appearance, with the background which is visible when the 

 observer looks past him. In many birds, for instance, those colors which 

 would be seen by an enemy looking down upon them are laid on by 

 nature in coarser and more blotchy patterns than are the colors on 



^I have proved, by experiments with painted decoys, that even brilliant top 

 colors, however strongly contrasted to surroundings, scarcely tend to betray the 

 wearer if his ensemble be a gradation from dark above to light below. 



