IX.] RETROGRESSIVE DEVELOPMENT IN NATURE. 1 9 



original colouring ? Clearly because colour became of little or 

 no importance to them as soon as they were sheltered under 

 the protection of man, while in a wild state it was a great 

 safeguard against detection by their enemies. 



Similarly the hair}^ covering has ceased to be of importance 

 to certain of the Mammalia — and has disappeared. Thus whales 

 and dolphins have a naked skin for the most part entirely 

 devoid of hair, although they are unquestionably descended 

 from hairy ancestors, and even now rudimentary hairs may be 

 detected in certain parts of the body by the aid of the micro- 

 scope. Obviously, the disappearance of the hairy covering 

 cannot be a direct consequence of disuse, for hair will grow as 

 well, whether its protective warmth be useful or of no import- 

 ance to the animal. But its disappearance as an indirect 

 consequence of disuse is plain ; for as soon as an immense 

 thickness of blubber was developed beneath the skin of the 

 whale, the warmth of an additional covering was unnecessary : 

 the hair becoming superfluous, natural selection ceased to 

 affect it, and degeneration at once set in. If anyone is inclined 

 to doubt whether the direct action of sea-water may not have 

 caused the disappearance of the hair, it is only necessary to 

 point to the group of seals, in which all the smaller species 

 possess a thick coat of fur, while, among the larger kinds, the 

 walrus has but a scanty covering of bristles, because, like the 

 whale, it has developed a layer of blubber, which is amply 

 sufficient to protect its huge body from cold. 



Examples of an entirely different kind are afforded by those 

 animals which hide themselves in cases or houses. The 

 hermit-crab partly conceals itself in empty shells, the aquatic 

 larvae of caddis-flies {Phryganidae) build cases within which 

 their cylindrical bodies are enclosed^ and the larvae of certain 

 small moths [Psychidae) do the same. Whenever the body of any 

 such animal is thus partially enclosed in a case, the protected 

 parts are soft and whitish, i. e. more or less colourless, while 

 the exposed parts retain the ordinary hard integument of the 

 Arthropoda and are variously and strongly coloured. Now we 

 may maintain that, in a certain sense, the hard integument of 

 crabs and insects fulfils the ' function ' of protecting the soft 

 parts of the animal from injury, but, correctl}^ speaking, this 

 defence is not a real function at all, because the exercise of 



c 2 



