2 5 2 THE VARIATION OF ANIMALS IN NATURE 



Mtillerian groups are very varied. We find every stage, from 

 cases where a single abundant species is resembled by a single 

 rarer species occurring in the same neighbourhood (Alcidis 

 agathyrsus, Moth : Papilio laglaizei, Butterfly — New Guinea) to 

 those in which an enormous number of species of supposedly 

 varying degrees of distastefulness are all more or less similarly 

 coloured, as in the great African complex of species resembling 

 Lycid beetles (partly illustrated by Marshall, 1902, pp. 575-8, 

 plate xviii). In some cases the colour resemblance is rein- 

 forced by convergence in behaviour, as in the bee-flies Eristalis 

 and Volucella, which when disturbed often lift one hind-leg, 

 just like a sleepy bee. 



We can only summarise here the arguments for and against 

 the theory that such resemblances are due to the selective 

 action of insectivorous enemies, principally birds. The 

 following appear to be the chief points in the arguments : 



(1) The extent to which the supposed methods of protection 



prevent the attacks of insectivorous animals. 



(2) The limits of the phenomenon of parallel evolution — 



i.e. the production, in forms not closely allied, of 

 similar colour-patterns, probably owing to certain 

 fundamental similarities in genetic constitution. 



(3) The possibility of alternative factors (probably edaphic) 



determining colour convergence. 



(4) If we admit that the mimicry theory provides a true 



explanation of some of the facts, to what extent does 

 it fail in particular cases ? 



(5) How far are the characters involved in mimetic 



resemblances analogous to specific characters ? 



(1) As a preliminary to discussing the origin of mimetic 

 resemblances, some evidence is required that the mimics belong 

 to groups with numerous predacious enemies. It has been 

 established that insectivorous insects discriminate very little 

 in their attacks and often eat protected forms, so they are 

 little likely to be concerned in any selection of warning colour- 

 patterns. It is therefore amongst birds (possibly also lizards 

 and mammals to a minor extent) that the significant enemies 

 must be found. It has always been stated by opponents of 

 the mimicry theory that birds very rarely eat butterflies, and 

 Heikertinger still maintains this opinion. However, there is 



