404 EVOLUTION AND ANIMAL LIFE 



rangement of pigments or on the fine structure of superficial 

 parts, as feathers, scales, skin, etc., or on a combination of the 

 two color-producing conditions. In birds, for example, certain 

 fat pigments called lipochromes (which are either actual reserve 

 food products or are associated with such), are abundantly 

 present in the feathers, bill, feet, etc., producing reds, yellows, 

 browns, etc., and certain other dark melanin pigments are 

 distributed as minute amorphous granules in the cuticular 

 structures or epidermis, producing pjfain gray, brown, black, 

 and related tints. In addition, the feathers are so constructed 

 that they may, and do in some cases, produce the most brilliant 

 iridescent and metallic colors, as familiarly shown to us by the 

 humming birds, the grackles, etc. Most such metallic colors 

 in birds, however, are produced by a combination of pigment 

 and structure, and not by structure alone. The colors of 

 mammals, of reptiles, of amphibians and of fishes might also 

 be referred to, and as far as they have been studied or analyzed 

 according to their causes, we should find that in mammals the 

 pigmental colors are mostly produced by so-called melanins 

 which seem to be waste products. In the fishes, amphibians 

 and reptiles, the pigments are both lipochromes and melanins, 

 while in all the vertebrate classes there occur cases in which 

 vivid physical or optical colors are produced by cuticular 

 structure. The most extended study of color in animals, 

 however, has been devoted to insect colors. Here we have a 

 pretty clear understanding of all the color-producing agents, 

 and an analysis of all the colors more usually met with, into 

 their proper classes, that is, whether exclusively pigmental, 

 exclusively structural or mixed structural-pigmental. In a 

 valuable paper by Tower, a table of insect colors showing the 

 classification and mode of production of the various colors is 

 given, as follows (see next page) : 



The only hypothesis that gives to colors and markings a 

 value in the life of animals, at all comparable with the degree 

 of specialization reached by these colors and markings and by 

 the special structures developed to make them possible, is that 

 already referred to as the theory of protective and aggressive 

 resemblances, of warning and directive patterns, and of mimi- 

 cry. These various uses of color patterns are all concerned 

 with the relation of the animal to its environment: they are 

 means of protecting the animal from its enemies, or of enabling 



