SELECTION ON THE VARIABILITY AND CORRELATION OF ORGANS. 
35 
correlation than the French. We must accordingly conclude that by a leg selection 
from the French aimed at reproducing the proportions of the Aino leg, we should not 
obtain an arm equivalent to the Aino arm. The divergences are indicated in the 
accompanying table : — 
1 
Selection fiom the French. 
Aino. 
Unselected 
French. 
Mean humerus ... ... 
30-48 
29-50 
33-01 
,, radius. 
22-69 
21-55 
24-39 
Variability of humerus. 
1-39 
1-34 
1-54 
,, radius. 
1-11 
1-06 
1-17 
Correlation of humerus and radius . 
-799 
-776 
-845 
,, ,, femur and humerus . 
-819 
-858 
-842 
,, ,, femur and radius 
-694 
-789 
-744 
,, tibia and humerus . 
- 845 
- 745 
-860 
,, ,, tibia and radius . . 
-768 
-865 
-780 
There is, of course, no special reason for suj^posing that the French and Aino 
differ merely by an evolution which has acted by selection of femur and tibia. We 
might have obtained a race out of the French more nearly akin to the Aino by a 
selection of femur and humerus, but the ju’ocess would numerically be exactly 
similar. The particular illustration here chosen is taken merely as an instance, to 
indicate how the methods developed in tins memoir enable us to ascertain with 
quantitative certainty how far racial differences may be due to the more or less 
strino;ent selection of a limited number of organs in the one race. 
If we consider that local races have been differentiated from a parent stock by 
the selection of the chief or more markedly divergent oi’gans, then we have in 
processes such as that just illustrated a metliod of ascertaining, at least tentatively, 
whether two races are to be considered as merely local varieties, and further the 
particular organs through selection of which the differentiation has taken place. 
Illusteation II.— Injiucnce of a Selection of Femur and Humerus in Mod fying 
Stature. 
The following data have fjeen calculated for me by Miss Alick Lee from IIollet’s 
measurements on the French :— 
* They have been undertaking, with the view of determining more scientifically than appears to me 
yet to have been done, the mean stature of a race from a measurement of the long bones found in burial 
mounds, Ac. Rollet’s measurements are given in ‘ De la Mensuration des Os longs des Membres,’ 
Lyons, 1889. I hope shortly to publish a memoir on the siil)ject. [The memoir in cpiestion was published 
in ‘Phil. Trans.,’ A, vol. 192, pp. 1G9-24I, 1898.] 
