238 
On Theories of Association 
recently have the original records come into possession of the Eugenics Laboratory 
and we now find they contain the important desideratum of age. It is proposed 
shortly to revise these tables of eye-colour, paying regard to the increasing pigmen- 
tation of the eye in extreme infancy and the decreasing pigmentation in extreme 
age. The inclusion of all cases, as no age was provided, has possibly something 
to do with the admitted irregularity of the tables. But does this irregularity 
invalidate the main conclusion drawn from the tables, i.e. that eye-colour pigment 
is inherited at the same rate as the measurable physical characters with a 
correlation lying somewhere between *46 and '50 ? The divisions made originally 
in the tables were selected purposely and with very definite ends, namely, 
(i) to give as good a physiological difference between the two groups as possible, 
(ii) to put into the same class the same eyes of both parents and offspring, and 
(iii) to get the least probable error by taking the divisions as near the median 
as was practically realisable. Mr Yule disregards the importance of (ii) and (iii) 
and gives the values of the tetrachoric r t found from a number of divisions, 
several of which have so few individuals in the quadrant that they are extremely 
untrustworthy. At the time the data were dealt with there was very little avail- 
able knowledge as to the distinction between a blue and grey eye. They were 
put into the same class, because it was considered that the total pigmentation of the 
iris of the grey eye was more akin to that of the blue eye, than to that of eyes like 
hazel with some macroscopic anterior pigment. The difference between the blue 
and the grey eye was considered to be one of structure rather than of pigment. 
With what we know of eyes now, we are not prepared to accept the Mendelian 
classification of eyes into those without and those with anterior pigmentation of 
the iris. If such a classification were absolutely legitimate, then grey eyes ought to 
be put with non-blue or with blue according to whether they possess such pigment. 
On such a classification as we have indicated the cj> for inheritance of eye-colour 
between parent and offspring ought to be ^. Accordingly to test the position for 
the grey eye we ought to consider whether the correlations come out more nearly 
^ with the greys put with the browns, or put with the blues, i.e. whether (f> for 
a fourfold table is nearer ^ for the division 1 + 2:3 + 4+5-1-6 + 7+ 8 or for 
1 + 2 + 3:4 + 5+ 6 + 7 + 8. There cannot be a moment's hesitation as to which 
is the more strictly Mendelian division. We find : 
Values of Boas-Yulean <jf>. 
Paii- 
Blue only, 
Blue and Grey, 
i.e. 1 + 2 
i.e. 1+2 + 3 
Father and Son 
■33 
•37 
Father and Daughter 
■22 
•28 
Mother and Son 
•28 
•32 
Mother and Daughter ... 
•24 
•34 
Mean 
•27 
•33 
