to limitations of the Milllerian Hypothesis of Mimicry. 95 



that young insectivorous animals are not born with an 

 instinctive knowledge of what insects are good to eat and 

 what are not. The slow flaunting flight and conspicuous 

 coloration of a Danaida or a Heliconius can have no 

 special significance for a young bird when it first starts 

 out in life to forage for itself. The knowledge that these 

 butterflies possess a pronouncedly unpleasant flavour can 

 only be acquired as the result of direct experiment, and 

 probably several butterflies would be destroyed by such a 

 bird before it succeeded in permanently associating the 

 conception of unpleasantness with any particular colour- 

 pattern. Thus each species with an independent type of 

 warning coloration would have to lose a certain per- 

 centage of its individuals before its true significance could 

 become a matter of common knowledge among all the 

 young and inexperienced insect-eaters in any particular 

 neighbourhood. From this Muller argued that if two 

 such species, having different colour-patterns and in- 

 habiting the same area, were to develop the necessary 

 variations leading up to a mimetic association, such a 

 combination would be of considerable utility in relation to 

 the experimental tasting of young birds; for then both 

 species together would only have to contribute the same 

 number of victims which each of them would have to 

 furnish if their colours were different. 



Given the initial variation, such a mimetic resemblance 

 can only be built up through the operation of some 

 eliminative or selective factors which shall result in an 

 advantage to the variation as compared with the typical 

 form from which it is derived ; thus leading up to the 

 gradual replacement of the latter by the former. And it 

 may be noted that, with the exception of resemblances in 

 which mere affinity has obviously played a great part, the 

 closer the mimicry, the greater must have been the 

 persecution of the mimic (whether Batesian or Mullerian) 

 and the higher the percentage of its elimination. In the 

 case of Mullerian mimicry, as we have seen, the selective 

 agent is provided by the destruction due to the experi- 

 mental tasting of unpalatable insects by inexperienced 

 insectivorous animals. This may be briefly denoted as 

 the " Miillerian factor." 



There are certain points in connection with the operation 

 of this factor which it is well to bear in mind. In the 

 first place, the differences in the relative intelligence of 



