36 



Life and Letters of Francis Galton 



percentage values on considerable numbers in the groups combined of Light 

 and Dark Blue, Grey, Blue Green: 



It will be seen that, while there is no significant change in the percentage 

 of light eyes in the women, there is really such a change in the light eyes 

 of the men ; the grandparental and great grandparental generations have 

 more bluish eyes. Were it not for the fact that there is no change in the 

 women, we might attribute this not to a racial change going on, but to men's 

 eyes growing lighter with extreme age. I have no statistical data to produce, 

 but my impression of the marked frequency of very light colour in old men's 

 eyes is strong. At the same time I know no physiological reason why men's 

 and not women's eyes should grow lighter with greater age. 



On the basis of his diagrams Galton considers that he may disregard "a 

 current popular belief in the existence of a gradual darkening of the popula- 

 tion, and can treat the eye-colours of those classes of the English race who 

 have contributed to the records, as statistically persistent during the period 

 under discussion" (p. 406). 



Galton next states that he considers that there are only two fundamental 

 types of eye-colour, the light and the dark, but under this supposition the 

 medium tints are troublesome. Such tints he has classified under "Dark 

 Grey and Hazel." In these cases the outer portion of the iris is usually of 

 a dark grey colour, and the inner of a hazel. The proportions of grey and 

 hazel vary, and the eye is called "dark grey" or "hazel" according to the 

 colour which happens most to arrest the attention of the observer. Galton's 

 attempt to deal with these medium eyes, of which there are in the popula- 

 tion about 12'7 / o , is to me unconvincing; yet the fact that he recognises 

 their existence is more satisfactory than the Mendel ian treatment which dis- 

 regards them entirely ! 



Galton for conciseness terms all these eyes "hazel." He defines a hazel- 

 eyed family to be one in which there is at least one hazel-eyed child, and he 

 proceeds to inquire into the constitution and ancestry of such "hazel-eyed" 

 families or sibships. He obtains the results tabulated on p. 37. 



Now it is clear from the table that when there is a hazel-eyed child in a 

 sibship, the percentage of dark eyes in the sibship is only very slightly 

 reduced, but the number of light-eyed brothers and sisters is 16% below 

 that of the general population. Again in the parental generation, there are 

 12 °/ o fewer light-eyed parents of hazel-eyed parents, and this 12°/ u is 

 transferred to the hazel-eyed group, the dark-eyed parents remaining at 



