io8 
the stronger race to seize upon the supplies of food, to 
starve out the weaker species, and to kill them if they 
resisted. But there is another possibility. In the absence 
of other food, and perhaps sometimes on other occasions 
also, the rat is a cannibal ; feeding upon other members of 
its own species, just as in the scarcity of their normal 
diet caterpillars have been observed to do. But if this 
were so another result would follow. The experiments of 
Yung show that among frogs the relative number of males 
and females depends upon the food which the parents eat. 
The proportion of females to males increases when an 
animal diet is substituted for a vegetable diet, and when a 
cannibal diet is adopted no less than ninety per cent, are 
females. Now, if this result follows in the case of other 
kinds of animals, and if we consider the enormous fertility 
of the rat, in ordinary cases, and bear in mind how this 
would be enhanced by such an excessive amount of females, 
the victory of the brown rat over the black rat is accounted 
for without any difficulty whatever. 
The willow succeeds in displacing the English watercress 
from the rivers of New Zealand by a very simple and 
intelligible process. “The roots of these trees penetrate the 
bed of the stream in every direction, and the watercress, 
unable to obtain the requisite amount of nourishment, 
gradually disappears.”* 
In these cases there is no abstract survival of the fittest: 
both are equally fit apart from one another. 
But if the argument of the advocates of Natural 
Selection were amended by the substitution of undoubted 
cases, in which the more fit drove out the less fit species, 
this would not constitute Natural Selection. You may 
call it a selection of a favoured species—and it undoubtedly 
Darwinism, p. 24.