56 SPIDERS [CH. 



It is seldom difficult to see some such possibility 

 of gain in the numerous well-known cases of insect 

 mimicry. The wasp tribe formidable with their 

 stings are often "mimicked"; the unpalatable 

 Heliconid butterflies are "imitated ' by members of 

 edible families, and some insects are such exact 

 imitations of leaves that the all-devouring army ants 

 have been seen to run over them without discovering 

 the imposition. 



"Mimicry' is an unfortunate term inasmuch as 

 it seems to imply intentional imitation ; "protective 

 resemblance' is better. It is generally accounted 

 for by the action of "natural selection " upon random 

 variations. No two members of a brood are exactly 

 alike ; slight variations in form, size, colour etc., are 

 constantly occurring, and when the variation is a 

 useful one the animal possessing it has a slightly 

 better chance of surviving and rearing progeny, some 

 of whom will probably possess the same peculiarity, 

 perhaps even in a more marked degree, and will be 

 better equipped than their neighbours in the struggle 

 for life. The happy possessors of such favourable 

 variations are thus in a sense "selected' by nature, 

 and this selection, acting through countless genera- 

 tions, is thought to be the chief agent in bringing 

 about the remarkable phenomenon of protective 

 resemblance. 



The theory has, no doubt, been pushed too far; 



