EVOLUTION 



443 



ingly that a progressive change is taking place in the 

 stature of women. 



Again, take eye-colour in man, for which my statistics 

 are more numerous, running to 774 husbands and wives. 

 Here, if as on p. 427, we take the blue-green, gray range 

 as unity, and measure the fraction of this range that the 

 median tint is from the light end of it, we have : — 



Median Eye-Colours. 



Husbands . 

 Fathers of sons 



•4346 

 .5418 



Wives 



Mothers of daughters 



.7632 

 ■849s 



The differences are thus very substantial. We are 

 forced to the conclusion that the dark-eyed are under the 

 present environment more fertile than the light-eyed. In 

 other words, the type and the character of maximum 

 fertility are not coincident, or genetic selection tends to 

 change the type. In view of such statistics I think it 

 impossible to disregard this factor of evolution. Different 

 magnitudes of character are associated with different 

 grades of fertility, and the biologist cannot deal with 

 natural selection as if its effects were uncomplicated 

 by the action of differential fertility. So far we have 

 dealt with genetic selection as evidenced by changing 

 type ; we have yet to consider whether the maximum 

 fertility is associated with the mode in apparently stable 

 types. The data I am about to cite belong to material 

 which is as yet incomplete, but the evidence is, in my 

 opinion, very strong, and singularly suggestive for further 

 investigations. I was collecting the seed of various plants 

 for researches on the inheritance of plants, and found the 

 following frequency for the stigmatic bands on the seed- 

 capsules of 176 Shirley poppies growing in the garden of 

 Hampden Farm House : — 



Thus there were 4443 capsules in all. Not being at 

 the time sufficiently impressed with the importance of 



