465 
1911-12.] On Inheritance of Hair and Eye Colour. 
As an example, the figures Dr Beddoe obtained by observation in the 
town of Caen are given. The numbers are as follows : — 
Light, Medium 
and Red Hair. 
Dark Hair. 
Jet-black Hair. 
Total. 
Light eyes .... 
Mixed and dark eyes 
/ 149-5* 
\ (149-5) 
51-5* 
27 
(27-15) 
93-5* 
1-5* \ 
(1-11) / 
16 
178 
161 
Total . 
/ 201 
l (201) 
120-5* 
(120-0) 
17-5* \ 
(17-9) / 
339 
In this case a 2 = 149'5 and (<x + 6) 2 = l78. This gives on solution a = 1223 
and 6 = 111, so that we obtain 2a6 = 27T5 as against 27 found and 6 2 = T23 
as against T5 found. Whether we regard the T5 as really one individual or 
two, the fit is exceedingly good. The same process applied to the total gives 
(a + c) 2 = 201 and (a + 6 + c + c£) 2 = 339, so that (a+c) = 14T8 and (b + d) = 
4*23, which give 2(<x + c) (b + d)=120 as against 120*5 and (6 + c6) 2 = 17*9 as 
against 17*5 found. 
This example illustrates how the race mixture can be analysed and the 
closeness with which the numbers accord with such distribution of the 
population as is given by the Mendelian theory. Such complete correspond- 
ence is of course rare. Another example almost equally good is that of 
Bradford. Here the numbers are even larger, the sample of the population 
observed numbering 1400 persons. In this case the theoretical numbers are 
printed in brackets above the actual : — 
Light, Medium 
and Red Hair. 
Dark Hair. 
Jet-black Hair. 
Total. 
Light eyes .... 
/ (663) 
( 663 
(117-8) 
117 
(5-24) | 
786 
Total (all eyes) 
f (968) 
\ 968 
(392-4) 
387 
(39-6) 1 
45 / : 
1400 
The method of testing the suitability of such fitting is that given by 
Professor Pearson (4). The differences are taken between each theoretical and 
actual number ; these are squared, divided by the corresponding theoretical 
(5*4) 2 (5*4) 2 
number, and summed. In the case of the totals this is equal to ^ ^ 
or *81. 
* Where -5 occurs, the indications were so nearly equal that the individual was recorded 
half in one class and half in another. 
VOL. XXXII. 
30 
