Danainae_ and _Acraeinae_ in all the warmer parts of the world, and, in
tropical America, the _Ithomiinae_ and _Heliconinae_ as well. These
groups have the characteristics of aposematic species, and no theory but
natural selection explains their invariable occurrence as models
wherever they exist. It is impossible to suggest, except by natural
selection, any explanation of the fact that mimetic resemblances are
confined to changes which produce or strengthen a superficial likeness.
Very deep-seated changes are generally involved, inasmuch as the
appropriate instincts as to attitude, &c., are as important as colour
and marking. The same conclusion is reached when we analyse the nature
of mimetic resemblance and realize how complex it really is, being made
up of _colours_, both pigmentary and structural, _pattern_, _form_,
_attitude_ and _movement_. A plausible interpretation of colour may be
wildly improbable when applied to some other element, and there is _no_
explanation except natural selection which can explain all these
elements. The appeal to the direct action of local conditions in common
often breaks down upon the slightest investigation, the difference in
habits between mimic and model in the same locality causing the most
complete divergence in their conditions of life. Thus many insects
produced from burrowing larvae mimic those whose larvae live in the
open. Mimetic resemblance is far commoner in the female than in the
male, a fact readily explicable by selection, as suggested by Wallace,
for the female is compelled to fly more slowly and to expose itself
while laying eggs, and hence a resemblance to the slow-flying freely
exposed models is especially advantageous. The facts that mimetic
species occur in the same locality, fly at the same time of the year as
their models, and are day-flying species even though they may belong to
nocturnal groups, are also more or less difficult to explain except on
the theory of natural selection, and so also is the fact that mimetic
resemblance is produced in the most varied manner. A spider resembles
its model, an ant, by a modification of its body-form into a superficial
resemblance, and by holding one pair of legs to represent antennae;
certain bugs (Hemiptera) and beetles have also gained a shape unusual in
their respective groups, a shape which superficially resembles an ant; a
Locustid (_Myrmecophana_) has the shape of an ant painted, as it were,
on its body, all other