( xliv ) 



natural selection." And again — " The colour and markings 

 of animals in general are primarily due to the action of light 

 and the colour of the environment or background. To sup- 

 pose that in the case of butterflies alone the colours of the 

 mimics are due to the attacks of bii'ds, whereas remarkably 

 few butterflies, as we have seen, are ever eaten by them, is a 

 cause so inadequate, so limited in its scope and so one-sided, 

 that it is no wonder the hypothesis has so many opponents." 

 In these observations, we seem again to have the idea of 

 climatic influence, and it appears to me to be desirable here 

 to quote the words of Professor Poulton in his paper read 

 before the Linnean Society in March 1898 — " There is 

 something attractive and plausible in the suggestion that the 

 strong mutual resemblance within a group of butterflies of 

 different genera and sub-families, inhabiting a single locality, 

 are due to the direct action of peculiar local physical or 

 chemical influences ; but the suggestion loses all its attract- 

 iveness when it is applied to the resemblance between a spider 

 and an ant or a moth and a wasp, and yet few could bring 

 themselves to believe that the resemblances which are here 

 contrasted have been built up by two entirely different sets 

 of forces." In another part of the same paper, Professor 

 Poulton points out the very difl^erent conditions under which 

 the larval stages of mimetic insects undergo their develop- 

 ment. Thus the larva of the drone fly lives on putrefying 

 animal matter, " a food as different as possible from that pro- 

 vided for the larval bee," and we are reminded that " when 

 the imago emerges from the pupa and its expanded wings 

 have dried, nothing that it will eat or endure henceforward 

 produces any further effect upon its colours or patterns. 

 Hence identity of food and condition during the final stage 

 cannot be of any assistance to the interpretation of mimicry." 

 Pi'ofessor Packard further proceeds to point out that bright 

 colours are not invariably associated with a nauseous taste or 

 smell, but surely a negative fact of this kind does not in any 

 way invalidate the theory of warning colours. The fact that 

 certain nauseous insects have not developed warning colours 

 does not appear to show that the bright colours of other 

 nauseous insects are not of a warning nature, neither does the 



