300 
be adduced that Natural Selection has acted in this way? 
Now we can apply two tests—i. The character of the selec 
tion which takes place in nature ; 2. The product of that 
selection as evidenced by the varieties which are actually 
to be found in any species, the stability of which has been 
long established. The analogy of artificial selection will 
be found useful in settling the first point. So far as 
type-producing selection is concerned, artificial selection 
results in an artificial product—“ the monstrosity of the 
fancier ” ; and it is obvious that this can be maintained 
only by the strictest selection. But the selection which 
is supposed to take place in nature results in an adaptation 
to the new circumstances which arise in nature, and, as we 
have said, when that adaptation is once accomplished, it 
will be maintained by the principle of regression to 
mediocrity. Even in those cases in which the transmuta 
tion effected is less abnormal, the highest possible ex 
cellence is only maintained by the strictest selection, as, 
for example, in the case of the Saxon flock-masters— 
where the lambs are thrice subjected to the most searching 
examination, and the best are put aside on each occasion. 
But we have no such selection as this in nature; for during 
the very period which elapses between birth and the adult 
condition, a great deal of the destruction is accidental— 
i.e., has no relation whatever to relative excellence. Then 
again, in artificial selection one type of excellence is 
preserved ; but in the destruction which occurs in nature 
the selection is first in favour of one variation and then in 
another, the effect of which must be to produce an average 
animal, and not one of especial endowments in one or 
more directions. Hence either in nature there is no 
selection, or a selection which produces general stability of 
species rather than an ideal perfection. Then again, if we 
apply the other test, and ask what is the product of this