A. M. Carr Saunders 365 
TABLE X. 
Incidence of Disease in Eye Colour Groups. Percentages. 
Boys and Girls. Age 3 — 7. 
Eye 
Colour 
Measles 
Whoopino 
Cough 
Scarlet 
Fever 
Diphtheria 
Mumps 
Chicken Pox 
Boys 
Girls 
Boys 
Girls 
Boys 
Girls 
Boys 
Girls 
Boys 
Girls 
Boys 
Girls 
Blue ... 
Light... 
Neutral 
Dark ... 
62-0 
72-4 
G5-2 
65-7 
58-4 
68-8 
57-8 
54-7 
37-0 
41-4 
39-3 
32-2 
38- 2 
52-1 
39- 1 
34-7 
8-2 
5-2 
4-5 
7-0 
5-8 
10-4 
10-9 
8-0 
1-5 
1-7 
1-1 
1-7 
1-2 
4-2 
1-6 
1-3 
14-9 
10-3 
19-1 
21-8 
17-3 
8-3 
20-3 
12-0 
31-7 
41-4 
34-8 
27-0 
27- 2 
33-3 
28- 1 
28-0 
alliance of one hair colour with a general excess of disease or liability to one 
disease. There is a greater variation and irregularity in these percentages, due to 
the smaller number of children, and to the consequently larger probable errors. 
TABLE XL 
Values of P. 
Measles 
Whooping 
Cough 
Scarlet 
Fever 
Diphtheria 
Mumps 
Chicken 
Pox 
Boys, Age 
13 
•298 
■075 
•506 
•346 
•611 
•098 
3- 
-7 
•171 
•612 
•661 
•969 
•522 
•264 
Hair ■ 
Girls, 
ii 
13 
•787 
•990 
•123 
•899 
•164 
•115 
J) 
ii 
3— 
7 
■097 
•756 
•416 
•770 
•082 
•019 
' Boys, 
ii 
13 
•304 
•909 
•596 
•238 
•039 
•215 
Eye • 
3- 
-7 
•079 
•825 
•963 
•651 
•356 
•679 
Girls, 
ii 
ii 
13 
•833 
•571 
•102 
•023 
•044 
•143 
» 
ii 
3- 
-7 
•477 
•256 
•513 
•543 
•248 
■858 
In the above table all the data used in calculating all the preceding tables 
of percentages are employed in a different way. Percentages alone give no 
clear or accurate measure of such a point as we are investigating; we want to 
obtain some definite indication of the value and importance of the fluctuations 
which the percentages showed. This can be obtained by using Elderton's tables 
to test for random sampling. It involves calculating %' 2 for the various distribu- 
tions ; this is done by the use of Pearson's formula given in a recent number of 
Biometrika*. The actual procedure is as follows: we are dealing with forty-eight 
groups and we may consider any one of them. We may, for instance, take the 
girls aged 13 arranged according to the eye colour classification, when the cases of 
measles among them are given. There are therefore two populations — an infected 
and a non-infected population ; and we have two independent distributions of the 
frequency of eye colour, namely in an infected and in a non-infected population. 
* Biometrika, 1911, Vol. vn. p. 186. 
