riJOTKCTIVE COLORATION. 1 o!* 



prove this : after the eyes have been removed, the'action of 

 the colour of the environment on the colour of the skin ceases. 

 Thus, in many cases, chemical action and stimulation of the 

 optic nerves may be closely connected in the process by which 

 the colours of animals are affected by light." The use of the 

 term " photograph " is apt to prejudice the question ; nothing 

 of the kind occurs in the soft tissues of a caterpillar about 

 to change into a chrvsalis. Furthermore, Mr. Poulton has 



O f 



shown that the eyes have no share in the production of a 

 harmony in colour with the environment : he covered the eyes 

 with a varnish, so as to entirely exclude the light, and yet 

 there was no failure in the adaptability of the larva. 



The case is not to be compared with the change of colour in 

 frogs and fishes, where the pigment is contained in chromato- 

 phores connected with nerves, and subject, therefore, to reflex 

 action the stimulus coming through the eye. Nevertheless. 

 Mr. Poulton considers it probable that the change of colour is 

 due to nervous influence exercised upon the nerve terminations 

 in the skin. This influence, can, however, be hardly exactly the 

 same as that exercised upon the contractile chromatophores 

 in the skin of the frog : an actual bleaching possibly occurs in 

 a bright light, which probably also favours the production of 

 the gold colour ; this colour, it should be remarked, is not due 

 to gold-coloured particles deposited in the skin of the larva or 

 in the pupa case, it is a structural colour caused by unequal 

 refraction of the light through thin films of air (or some gas) 

 and chitin. Possibly, therefore, intense light may cause some 

 gas to be given off in greater abundance. 



In any case, the action of natural selection here must be 

 quite different from the action of natural selection in producing 

 fixed resemblances to the environment such as is seen, for 

 example, in the butterfly illustrated on Plate II. In Kfillima 



