XV THE RELATION OF FACTS TO THEORIES 317 



The author then proceeds to discuss various cases, 

 of which the first is an example rather of protective 

 coloration than of mimicry proper. It is the case of 

 the colour-change of the caterpillar in the Sphingides 

 from green to brown just before pupation. The 

 caterpillar feeds among green leaves and is then 

 green, but it forms a chrysalis in earth, and the 

 brown colour has been held to be of protective 

 importance during the period when the caterpillar is 

 in search of a retreat. Piepers, however, observes 

 that the period which elapses between the cessation 

 of feeding and the formation of the chrysalis is 

 exceedingly short, in some cases not more than some 

 minutes ; that the brown is constant, while the tint of 

 the earth varies and is often quite different ; and that 

 as the caterpillar is necessarily moving all the time, 

 a protective colour can hardly be of much avail. 

 Finally, the colour-change is exceedingly common in 

 the larvae of Lepidoptera at this stage, and is probably 

 due to a discoloration resulting from the drying up 

 of the skin preparatory to its being shed. It occurs 

 also in Sphinx larvae which are brown to begin with 

 and not green, though here the change is so slight as 

 to be little noticed. 



The next case taken up is that difficult one of 

 the occurrence of several different forms among the 

 females of certain Papiliones of India and the Malay, 

 which was discovered and discussed by Mr. Wallace 

 {Trans. Linn. Soc. xxv. ; see also Contributions to the 

 Theory of Natural Selection, London, 1 8 7 1 ). The case 

 is a somewhat difficult one, in part because the names 

 used involve in themselves an interpretation of the 

 facts. So far as it is possible for one who is not an 



