GAY PLUMES AND DXJLL 



ity. The swan is still more restricted in its range 

 and numbers than the goose, and, in our hemi- 

 sphere, is snow-white. The factor of protective col- 

 oration, so pronounced in the case of the goose, is 

 quite ignored in the swan. Neither the goose nor 

 the swan, so far as I know, has any winged enemies, 

 but their eggs and young are doubtless in danger 

 at times from foxes and wolves and water animals. 

 The duck must have more enemies, because it is 

 smaller, and is found in more diverse and sundry 

 places. Upon the principle that like begets like, 

 that variety breeds variety, one would expect the 

 ducks to be more brightly and variously colored 

 than their larger congeners, the geese and the 

 swans. 



The favorite notion of some writers on natural 

 history, that it is a protective device when animals 

 are rendered less conspicuous by being light be- 

 neath and dark above, seems to me a hasty con- 

 clusion. This gradation in shading is an inevitable 

 result of certain fixed principles. It applies to in- 

 animate objects also. The apples on the tree and 

 the melons in the garden are protectively shaded 

 in the same way; they are all lighter beneath 

 and deeper-colored above. The mushrooms on 

 the stumps and trees are brown above and white 

 beneath. Where the light is feeblest the color is 

 lightest, and vice versa. The under side of a bird's 

 wing is, as a rule, lighter than the top side. The 

 71 



