464 
Miscellanea 
TABLE IV. 
Age and Eye Colour, British Schoolgirls. 
Eye Colour*. 
Light 
Medium 
Dark 
Totals 
Age 7—10 
73 
74 
41 
192 
10—13 
147 
180 
85 
412 
13—16 
120 
214 
96 
430 
16—19 
64 
99 
57 
220 
Totals 
404 
571 
279 
1254 
From these Tables by the method of mean square contingency we have : 
Correlation between Age and Hair Colour ='158. 
Correlation between Age and Eye Colour = '096. 
Thus by dealing with one sex only and going on to 19 years of age, we see that the corre- 
lations while still not large are yet sensibly increased. If we may compare the German and 
English material, we should say that there is little change in hair or eye colour with children 
under 14 ; by this we mean, not enough to influence the determination of pigmentation resem- 
blance of children of different ages. But after 14 there is even before 19 a more marked change, 
the correlations still, however, remaining low. This change is much more considerable in the 
case of hair than of eye colour, though sensible in both. Even thus the values of the collateral 
heredity coefl&cients, if we allowed for change of age, would hardly be influenced within the limits 
of their probable errors. 
It would be interesting to trace the change in the correlation for one sex and race for later 
ages, but the material does not appear to be available. There is a table " showing the colour of 
eyes and hair of both sexes at all ages of English and Welsh origin " in the Final Report of the 
Anthropometric Committee of the British Association (1883) t- But the classification is different 
from that used in our data. Moreover the results are classified first according to eye colour, and 
in the case of eyes of a ' neutral ' colour, the hair colour does not appear to be distinguished. 
Thus it is impossible to base a correlation table of hair colour at different ages on the B.A. data, 
and no satisfactory comparison can be made with our present results. 
The only other material which is known to us is embodied in the Sozialanthropologische 
Studien of the late Dr W. Pfitzner of Strasburg, whose researches are in many respects both 
valuable and suggestive. But even where he deals with fairly numerous data J, certain limitations 
soon appear. Thus for the period from 5 to 20 years he examined only 164 girls, and it seems 
impossible to determine a correlation from such numbers. Moreover his figures are exclusively 
based on observations made on the cor^sse. He ought therefore to have ascertained as a pre- 
* Dr Uchida has omitted the ages 4 to 7 as of very small frequency. Tables for much larger 
numbers will be eventually published. Dr Uchida has not included the ' blue ' schedules, prin- 
cipally infant schools, at all. Material is probably available for nearly 2,500 girls. K. P. 
t Report of the British Association, 1883, pp. 278-9. 
J Schwalbe's Zeitschrift fUr Morphologic, Bd. i. S. 372. 
