i8o 
sufficiently disguised so as not to be observed, how is it that 
the enemy can acquire the greater keenness of vision ? All 
that the theory requires is that it should not observe those 
variants which are concealed by their colour or by their 
form. This oversight is absolutely necessary to the theory, 
but how can the repetition of such oversight—howsoever 
often it may be repeated—produce the necessary increase 
of discernment, which the theory absolutely demands. 
If, on the other hand, it be maintained that the change 
takes place gradually, then the slight differences in colour 
would not make the animal unobserved, and it would 
continue to be caught ; even if the whole race became 
gradually more accommodated to the colour of their 
surrounding, or to the resemblance to some other object, 
this would only quicken the vision of the enemy, and 
it would be quite educated up to the point of seeing what 
was invisible to those less interested by the time the great 
change was consummated. But how in such circumstances 
could the insect-eater become the agent of Natural 
Selection ? 
So that in this sphere of nature, either Natural Selec 
tion comes very late upon the scene and acts through 
the blindness of the enemy, or the enemy, educated by the 
observation of slight variations, continues to devour the 
species indiscriminately, and thus does not fulfil the office 
assigned to it by the theory. But this choice is not per 
mitted to us. The first step taken before Natural Selection 
comes into play, leads to the favoured individuals escaping 
observation. How can a repetition of this process increase 
the discernment of the enemy ? 
In connection with protective colouring in that sphere 
in which it is most usually manifested, we find a realm 
of nature in which accidental death is normal ; in which 
animals owe their survival to escape from many different