410 THE CAUSES OF THE XI 
breed of sheep, the wool of which was inferior to 
that of the Merino, was gradually allowed to 
die out. 
You see that these facts illustrate perfectly well 
what may be done if you take care to breed from 
stocks that are similar to each other. After having 
got a variation, if, by crossing a variation with the 
original stock, you multiply that variation,and then 
take care to keep that variation distinct from the 
original stock, and make them breed together,— 
then you may almost certainly produce a race whose 
tendency to continue the variation is exceedingly 
strong. 
This is what is called “selection” ; and it is by 
exactly the same process as that by which Seth 
Wright bred his Ancon sheep, that our breeds of 
cattle, dogs, and fowls are obtained. There are 
some possibilities of exception, but still, speaking 
broadly, I may say that this is the way in which 
all our varied races of domestic animals have arisen; 
and you must understand that it is not one 
peculiarity or one characteristic alone im which 
animals may vary. There is not a single peculiarity 
or characteristic of any kind, bodily or mental, in 
which offspring may not vary to a certain extent 
from the parent and other animals. 
Among ourselves this is well known. The sim- 
plest physical peculiarity is mostly reproduced. I 
know a case of a woman who has the lobe of one 
of her ears a little flattened. An ordinary obser- 
