148 MIMICRY IN N. AMERICAN BUTTERFLIES 



one of the best and safest means perhaps the 

 very best of forming a judgement between this 

 revived opinion and Darwin's conclusion that, 

 although the rate of transformation varied greatly 

 and might slow down to nothing for long periods, 

 the steps of change were small, forming a gradual 

 and ' continuous ' transition between the successive 

 forms in the same evolutionary history. 1 



The study of the causes of Mimicry is more 

 difficult than that of the history of Mimicry, the 

 conclusions far less * certain. Nevertheless the 

 evidence at present available yields much support 

 to the theory of Natural Selectk 

 cause of evolutiojLJ^Efee facts certainly do not 

 any other interpretation. T3jgv_iifiga 1 tive 

 that mimetic resemblances have 



beea-prodttood by 4Jie_-dIrectIatiQHlgf -external 

 forces (Hypothesis of External Causes) or -by 

 variatioB-tmguidod-bjL selection (Hypothesis of In- 

 ternal Causes). Nor do they support Fritz Miiller's 

 earlier and daring speculation (see pp. 127-8) 

 that female preferences were influenced by the 

 sight of the patterns displayed by the* models 

 (Hypothesis of Sexual Selection). The only 

 hypotheses whiehr-ara ia^ny way consistent with 

 the body of facts^ considered as a whole, are those 

 whicH assume that the_.resemblances in question 

 have been built up by the selection of variations 

 beneficial in the struggle ibxjife. 



In its concentration on a minute fraction of the 



1 See pp. 42-51; also Appendix B, p. 254. 



