62 Studies in the Theory of Descent. 



unable to define these differences precisely, we may 

 nevertheless safely conclude from such observa- 

 tions that they exist. 



I have given special prominence to this subject 

 because, in my idea, Darwin ascribes too much 

 power to sexual selection when he attributes the 

 formation of secondary sexual characters to the 

 sole action of this agency. The case of Bryonice 

 teaches us that such characters may arise from 

 purely innate causes ; and until experiments have 

 decided how far the influence of sexual selection 

 extends, we are justified in believing that the sexual 

 dimorphism of butterflies is due in great part to 

 the differences of physical constitution between 

 the sexes. It is quite different with such sexual 

 characters as the stridulating organs of male Or- 

 thoptera which are of undoubted importance to 

 that sex. These can certainly be attributed with 

 great probability to sexual selection. 



It may perhaps not be superfluous to adduce 

 one more similar case, in which, however, the male 

 and not the female is the most affected by climate. 

 In our latitudes, as also in the extreme north, 

 Polyommaius Phl&as, already so often mentioned, 

 is perfectly similar in both sexes in colour and 

 marking ; and the same holds good for the winter 

 generation of the south. The summer generation of 

 the latter, however, exhibits a slight sexual dimor- 

 phism, the red of the fore wings of the female being 

 less completely covered with black than in the male. 



