140 ANIMAL COLORATIOX. 



we may sujopose that there was originally no particular like- 

 ness to a dead leaf, except that the colours corresponded more 

 or less ; those individuals which were most like a withered leaf 

 were passed over for those which were less like ; and thus the 

 protective resemblance was heightened, until it attained to the 

 wonderful perfection which we now see. On the other hand, 

 the variable pupa may be supposed to have been, as many 

 pupte still are, originally iminfiueaced by light ; the acquire- 

 ment of such a suscejjtibility, being of manifest use, was favoured 

 by natural selection. In fact, we assume in this case not a 

 positive change of any kind, but the acquirement of the 

 capability of change when necessary. 



The whole matter, however, is very difficult of comprehension. 

 An investigation of the colouring substances which give the 

 colour to the pupa may show that they are sensitive to light 

 directly — that is, without the intervention of the nervous 

 system. In this case natural selection would hardly be 

 required, or would have had so roundabout an action as to 

 be incapable of being followed. At every step, in fact, in the 

 study of animal coloration we are met with closed doors, which 

 can only be unlocked by keys furnished by an intimate chemical 

 and physiological knowledge such as we do not at present 

 ])ossess. 



Variable Protective Coloration iu Vertebrates. 



Another kind of variable coloration is seen in certain animals 

 which possess the power of rapidly adjusting their colour to 

 that of the environment. 



The changes of colour in the Sole forms the subject of one 



of the most interesting chapters in Mr. J. T. Cunningham's 



work upon that fish,* lately published by the Marine Biological 



.Association. Altogether, five species of sole are found upon 



* " A Treatise on the Common Sole " (Plymouth, 1890). 



