MIMICRY OF ANTS AND WASPS 281 



account for this series of Mimetic Resemblances by an 

 appeal to the operation of External or Internal Causes or 

 of Sexual Selection. There remains Natural Selection, 

 which at once offers a convincing interpretation. Ants 

 and wasps are known to be aggressive dominant insects 

 avoided by the majority of insect-eating animals, although 

 certain species are adapted to feed almost exclusively 

 upon them. It is in every way probable that a superficial 

 resemblance to ants and wasps would be beneficial in the 

 struggle for existence. There is, indeed, some experi- 

 mental evidence to prove that real advantage is conferred. 1 

 We find that species of many groups mimic ants and 

 wasps in a variety of entirely different ways. The results 

 are exactly what might have been predicted to occur if 

 Natural Selection be the efficient cause of Mimetic Resem- 

 blance. 



The attempt has been made, in recent years, to cut 

 away the foundation of an interpretation based on the 

 theory of Natural Selection, by calling in question the 

 conclusion that butterflies are, as a matter of fact, attacked 

 by insect-eating animals such as birds. I have recently 

 collected together a great mass of evidence bearing on 

 this point, most of it obtained in Mashonaland, South 

 Africa, by the admirable naturalist, Mr. Guy A. K. Mar- 

 shall. This material conclusively proves that the wings 

 of fresh unworn specimens of butterflies are constantly 

 notched, as if by the attacks of birds and lizards, and that 

 in a considerable proportion of the examples the notches 

 on opposite sides fit together, proving that the insect was 

 seized when its wings were in contact. The attacks are 

 most frequently directed to the posterior angle of the 

 hind-wing, less frequently to the tip of the fore-wing, still 

 less frequently to the intermediate borders and angles. 

 The points of attack are those where special marks and 

 structures, probably having a directive function, are com- 

 monly developed. Thus the tip of the fore-wing is often 

 rendered specially conspicuous and the posterior angle of 

 the hind-wing is continually produced into so-called 'tails ' 



1 Poulton, Colours of Animals, London, 1890, p. 247 : Lloyd Morgan, 

 Animal Behaviour, London, 1900, pp. 164-5. 



