CHAP. IV Shifts for a Living 59 



poisonous species. Thus, the very poisonous coral-snakes 

 (Elaps), which have very characteristic markings, are 

 mimicked in different localities by several harmless forms. 

 Similarly in regard to birds, Mr. Wallace notices that the 

 powerful "friar-birds'' {Tropidorhynchus) of Malaya are 

 mimicked by the weak and timid orioles. " In each of the 

 great islands of the Austro-Malayan region there is a dis- 

 tinct species of Tropidorhynchus, and there is always along 

 with it an oriole that exactly mimics it." 



That there may be mimetic resemblance between distinct 

 forms there can be no doubt, and the value of the resem- 

 blance has been verified ; but there is sometimes a tendency 

 to weaken the case by citing instances or using terms which 

 have been insufficiently criticised. Thus the facts hardly 

 justify us in saying that the larvae of the Elephant Hawk 

 yii:>\!b. (ChcBrocampa) " terrify their enemies by the sugges- 

 tion of a cobra-like serpent ; " or that the cobra, which " in- 

 spires akrai by the large eye-like ' spectacles ' upon the 

 dilated ffooa, offers an appropriate model for the swollen 

 anterior end of the caterpillar, with its terrifying markings." 



There is only one theory of mimicry, namely, that among 

 the mimicking animals varieties occurred which prospered 

 by being somewhat like the mimicked, and that in the 

 course of natural selection this resemblance was gradually 

 increased until it became dominant and, in many cases, 

 remarkably exact. 



As to the primary factors giving rise to the variation, we 

 can only speculate. To begin with, indeed, there must 

 have been a general resemblance between the ancestors of 

 the mimicking animal and those of the mimicked, for cases 

 like the Humming-Bird and its Doppel-Ganger moth are very 

 rare. , But this does not take us very far. The beginning 

 of the mimetic change is usually referred to one of those 

 "indefinite," "fortuitous," "spontaneous" variations which 

 are believed to be common among animals. It is logically 

 possible that this may have been the case, and that there 

 was at the very beginning no relation between the variation 

 of the mimicker and the existence of the mimicked. But 

 as illustrations of mimicry accumulate — and they are already 



