Evidences of Theory of Natural Selection. 321 
landscape. In all such cases, of course, there has 
been a deviation from the normal type in respect 
of colour, with the result that the inherited instinct 
is no longer in tune with the other endowments of 
the animal. Such a variation of colour, therefore, 
will tend to be suppressed by natural selection ; while 
any variations which may bring the animal still more 
closely to resemble its habitual surroundings will be 
preserved. Thus we can understand the truly 
wonderful extent to which this principle of protective 
colouring has been carried in many cases where the 
need of it has been most urgent. 
Not only colour, but structure, may be profoundly 
modified for the purposes of protective concealment. 
Thus, caterpillars which resemble twigs do so not 
only in respect of colour, but also of shape; and this 
even down to the most minute details in cases where 
the adaptation is most complete: certain butterflies 
and leaf-insects so precisely resemble the leaves upon 
which, or among which, they live, that it is almost 
impossible to detect them in the foliage—not only 
the colour, the shape, and the venation being all 
exactly imitated, but in some cases even the defects 
to which the leaves are liable, in the way of fungoid 
growths, &c. There are other insects which with 
similar exactness resemble moss, lichens, and so forth. 
A species of fish secures a complete resemblance to 
bunches of sea-weed by a frond-like modification 
of all its appendages, and so on through many other 
instances. Now, in all such cases where there is so 
precise an imitation, both in colour and _ structure, 
it seems impossible to suggest any other explana- 
tion of the facts than the one which is supplied by 
* Y 
