inition will apply to aggressive (pseud-episematic) resemblance.
The conditions under which mimicry occurs have been stated by
Wallace:--"(1) that the imitative species occur in the same area and
occupy the same station as the imitated; (2) that the imitators are
always the more defenceless; (3) that the imitators are always less
numerous in individuals; (4) that the imitators differ from the bulk of
their allies; (5) that the imitation, however minute, is _external_ and
_visible_ only, never extending to internal characters or to such as do
not affect the external appearance." It is obvious that conditions 2 and
3 do not hold in the case of Mullerian mimicry. Mimicry has been
explained, independently of natural selection, by the supposition that
it is the common expression of the direct action of common causes, such
as climate, food, &c.; also by the supposition of independent lines of
evolution leading to the same result without any selective action in
consequence of advantage in the struggle; also by the operation of
sexual selection.

It is proposed, in conclusion, to give an account of the broad aspects
of mimicry, and attempt a brief discussion of the theories of origin of
each class of facts (see Poulton, _Linn. Soc. Journ. Zool._, 1898, p.
558). It will be found that in many cases the argument here made use of
applies equally to the origin of cryptic and sematic colours. The
relationship between these classes has been explained: mimicry is, as
Wallace has stated (_Darwinism_, London, 1889), merely "an exceptional
form of protective resemblance. "Now, protective (cryptic) resemblance
cannot be explained on any of the lines suggested above, except natural
selection; even sexual selection fails, because cryptic resemblance is
especially common in the immature stages of insect life. But it would be
unreasonable to explain mimetic resemblance by one set of principles and
cryptic by another and totally different set. Again, it may be plausible
to explain the mimicry of one butterfly for another on one of the
suggested lines, but the resemblance of a fly or moth to a wasp is by no
means so easy, and here selection would be generally conceded; yet the
appeal to antagonistic principles to explain such closely related cases
would only be justified by much direct evidence. Furthermore, the
mimetic resemblances between butterflies are not haphazard, but the
models almost invariably belong only to certain sub-families, the
_