J. F. Tocher 339 
XVIII. the relative values being given in all cases in the tables*. The percentages 
are given in Tables XXIII. and XXIV. of Supplement. 
II. Correlation of Hair and Eye Colour. Applying the contingency method 
to the data (see Table VIII. of Supplement) the following results were obtained. 
The author's results from the Aberdeenshire population and those of Pearson 
from Continental and British returns are also given, for the sake of comparison. 
TABLE XX. 
Correlation. — Hair and Eyes. 
Population 
Contingency Coefficient 
Returns by 
Male Asylum Inmates 
•3039 
J. F. Tocher 
Female Asylum Inmates 
> -2994 
)i 
Adult Scottish Population ... 
•3673 
•>■) 
Scottish Children t ... 
•3802 
i> 
Swedish Conscripts | 
•2495 
G. Retzius 
Prussian Children J ... 
•2714 
R. Virchow 
Italian Conscrijjts ... 
Jewish Children 1 
•3091 
R. Livi 
•3381 
R. Virchow 
Baden Conscripts J ... 
•3540 
0. Ammon 
British Schoolboys j ... 
•4203 
K. Pearson 
From this we see that there is no material difference between sane and 
insane populations in their degrees of correlation between hair and eye colours, 
although the result for the Scottish sane population is higher. The degree of 
correlation in the case of the Scottish children is slightly higher than that of the 
adult Scottish sane population. The continental results given above are not 
directly comparable, since while the children's data are available, those of the 
corresponding adult populations are wanting, and besides there are racial differences 
to consider. It would seem, however, from the foregoing that the correlation 
between hair and eye colours decreases in passing from a juvenile to an adult 
population. This is obviously due to a change of hair and eye colours in passing 
from childhood to manhood. The correlation between age and the colour of hair 
and eyes in man has been dealt with by Pearson§, who shows from Uchida's results 
on Prussian and British data that, with a range of 13 years (7-19), the correlation 
between age and hair colour amounts to -158, and between age and eye colour 
"096. From Pfitzner's hospital results the value •iol was obtained for hair colour 
and age, but it is pointed out that, owing to the positive correlation between 
fairness and disease in childhood, this value is too high ; probably "2 to •25 would 
* The foregoing is a short summary of the colour characteristics of the inmates. They will be 
dealt with in more detail in another paper when the results will be compared with the results of 
the Pigmentation Survey of Scottish School Children just carried out by the writer. 
I Not yet published. 
X Pearson: Bioinetrika, Vol. in. p. 461. § Biometrihci, Vol. iii. pp. 46'2 — 4fj(). 
