( xlii ) 



tion of that author and Professor Poulton, and published in 

 the Transactions of this Society for 1902. 



There is, however, one aspect of the matter which the 

 writer does not remember to have been previously put forward, 

 and that is the possible value of the force of example amongst 

 birds. It is well known that young birds are to some extent 

 trained by the example of their parents in many matters 

 which make for their ultimate success in the battle for 

 existence. They may and probably do indulge in much 

 experimental tasting, but the extent of such action may to a 

 great extent be curtailed by the example of the parents in 

 selecting more satisfactory food than butterflies in their 

 present condition aiford. Such a state of affairs would 

 minimize the present, or in some cases the local destruction 

 of butterflies by birds, and consequently also the number of 

 observed cases of such destruction, but it would not preclude 

 the possibility that a much more extensive process of selection 

 took place amongst butterflies in earlier times. In other 

 words the balance of nature is now maintained by other and 

 various agencies acting in concert with the now much-reduced 

 influence of insectivorous birds, and we are present now only 

 to see a state of affairs brought about by agencies which have 

 become considerably modified in their relative influences. 



It will probably be immediately pointed out that if the 

 selective influence of birds is not as great as ever it was the 

 well-known tendency of reversion to ancestral forms would 

 tend to do away with mimicry, and that only by constant 

 pressure of selective forces can mimetic resemblance be main- 

 tained. But is there any evidence that mimicry amongst 

 butterflies is at present more extensive and complete than it 

 has ever been before 1 Have we any right to say that all 

 cases of imperfect mimicry are in pi'ocess of being perfected 

 and not undergoing reversion? I have never been satisfied 

 that the white-winged form of Hi/j)olimnas misip2n(s is really 

 a mimic of the alcippii^s variety of Danais chrysipjjus, ex- 

 tremely attractive as the suggestion undoubtedly is. One 

 cannot forget that the male misippus has large white patches, 

 and the lack of geographical coincidence between alcippus and 

 misippus is much against the mimetic theory. The alcippus 



