96 AFRICAN GAME ANIMALS 



ing coloration patterns, unhampered by selection, in conse- 

 quence of some tendency or internal force having no refer- 

 ence to utility. Immense groups of birds, African bee-eaters 

 and sunbirds, for instance, American humming-birds, al- 

 most all kingfishers everjrwhere, gulls (in the adult form), 

 pelicans, most herons, darters, and cormorants (of all ages 

 and sizes), most adult male ducks, all swans, the whole 

 crow family, the whole grackle family, and multitudes of 

 others, have developed a highly advertising coloration with 

 no seeming relation to utility; and neither in these birds 

 nor in the great majority of the ordinary dull-colored birds 

 does countershading play any part in concealing them. 

 The effort to explain all or almost all patterns of coloration 

 in birds by natural selection working either for concealing 

 or directive markings represents nothing but the darken- 

 ing of wisdom; it serves merely to bemuddle an already 

 sufficiently difficult problem. What is needed is the kind 

 of study advocated in, and instanced by, Witmer Stone's 

 " Phylogenetic Value of Color Characters in Birds." 



Among mammals, squirrels are very apt to have an 

 advertising coloration. Countershading practically never 

 helps them. Among mice and shrews the species in which 

 countershading can have any survival effect in concealing 

 them are very, very few — and probably there are none. 



In sum, it appears certain that although the discovery 

 of the effect of countershading has a very important artistic 

 bearing, it is of practically no value as affecting the question 

 of concealing coloration in the higher vertebrates, the mam- 

 mals, and birds. Among mammals it can by no possibility 

 be of consequence save in a very few cases, and it is by no 

 means certain that it is of consequence in these. Among 



