252 ANIMAL COLOEATIOX. 



ptera and Hymenoptera) are less rare, but not common ; re- 

 semblances between species, belonging to different families or 

 genera of the same order (e.//., among Lepidoptera) , are very 

 common: their number bears a certain proportion to the variety 

 of forms in the order ; thus, they are most common among- 

 insects, less common among birds and reptiles, rare among 

 mammals. This fact is against the hypothesis that the re- 

 semblances have been produced by the action of natural 

 selection, since protection of this kind would be just as effica- 

 cious in the higher group as in the lower ; and we find among 

 mammals, as among insects, a large number of cases of pro- 

 tective coloration. 



Nevertheless, cases of mimicry that do occur — particularly 

 among Lepidoptera — are often so striking that no other explana- 

 tion than that offered by Mr. Bates seems to account for the 

 finishing touches, at least, of the resemblance. A considerable 

 initial resemblance may be fairly set down to other causes ; 

 because it is impossible to believe that a slight move in the 

 required direction would be of sufficient importance to serve as 

 material for the action of natural elimination. At the same 

 time, instances of mimicry, which are to be appreciated only 

 by insects {e.g., Volucella and bee), must, in the present state 

 of our knowledge of insect vision, be removed from the 

 category ; so also are certain other cases such as Brephos- 

 nothum and Ploseria dirersata, and the annelid mentioned by 

 Semper. Seeing, then, that resemblances may occur between 

 animals which either cannot be, or are probably not, advan- 

 tageous to either, it is at least necessary to wait for more 

 convincing proofs before it can be more than provisionally 

 assumed that natural selection is responsible for these re- 

 semblances in other cases where they appear to us to be 

 useful. 



