A 382 == 
Probably no form is absolutely immune; 1t should always be 
recognised that these matters are relative. But it seems to be 
clearly established by observation and experiment that some Birds 
at all events avoid some of these conspicuously marked Butterflies, 
and that there are various degrees of preference. Certain observers, 
it is true, have denied that Butterflies are fed upon by Birds at all, 
but there exists now a considerable body of evidence to the con- 
trary. This being so, we are led to the conclusion that the brilliant 
colours ¿of these) Inseets are, to "uses Prot. POuLLoNs aenm, 
« aposematic », that is to say that they are warning marks, which 
signify to insectivorous enemies such as Birds the presence of some 
quality whether of taste, or of odour, or of toughness, which makes 
their possessors unsuitable for food. If this conclusion is well 
grounded, we can find in the theories of BATES and of FRITZ 
MùLLER a sufficient explanation of the significance of mimicry. 
BATES pointed out, just upon fifty years ago, that a palatable 
Insect might escape attack by sailing under the false colours of an 
inedible species, and he was followed about twenty years later by 
FRITZ MÜLLER, who called attention to the fact that if Birds had 
to pass through an education in order to learn by trial what Insects 
to capture and what to avoid, the combination of unpalatable 
Insects into mimetic associations would protect each constituent 
of such an assemblage from a certain amount of experimental 
tasting. It has been shown, chiefly by Prof. LLOYD MORGAN, 
that this education of young Birds in what to eat and what to 
avoid is a reality and no mere assumption; and the theory of 
FRITZ MÜLLER may thus be said to rest on a substantial found- 
ation. The first of these theories, that of BATES, is the theory of 
what may be called true mimicry. That of FRITZ MÜLLER, as has 
been pointed out by Prof. POULTON, is more correctly desi- 
gnated as synaposematism, or the adoption by two or more forms 
of a common warning pattern. 
Opinions may legitimately differ as to the relative importance of 
these two theories; and until more data are at our disposal, it will 
be possible to doubt as to which of them is applicable to this or that 
given instance. But the theories are complementary to one another, 
and not mutually exclusive. And it is to be observed that both of 
them, equally with that of THAYER, imply the preservation, by 
natural selection, of appropriate variations. Let us now see how far 
these theories arein accordance with the facts with which we started. 
1. In the first place, it is obvious that the abundance of cases 
