CONSCIOUS PROTECTIVE RESEMBLANCE. 553 
and other di- or poly-morphism in the different individuals of the same 
species. A single instance will make this clear. There are certain 
genera of butterflies, such as Dismorphia (in the wide sense), Pseud- 
acrea, and Hypolimnas (also in the wide sense), of which almost the 
whole of the numerous species are mimetic. Within the limits of each 
genus the most divergent models have been followed, so that utterly 
different colours and patterns have been produced in forms which are 
still closely related, and in other structural features exhibit no corre- 
sponding differences. In the most extreme case known to me, immense 
differences occur in the different races of a form which systematists 
consider as a single species, viz. Hypolimnas bolina. If we compare 
the Indian form of female with those of the Malayan region, Australia, 
and Polynesia, including Fiji (in which the local race itself con- 
tains the most widely divergent forms), and remember that no corre- 
sponding differences exist which would justify us in conferring specific 
rank in any of the cases, we are driven to the conclusion that colour 
and pattern are the most superficial of all specific characters,—of all 
the least likely to persist unchanged when the models upon which they 
were founded have long since disappeared. 
In one special case which I have observed, there is evidence that 
changes in the nervous system have outlasted the markings which once 
gave a meaning to them. Some of the remarkable larve of the genus 
Ophideres have two eye-spots at the junction of the anterior and middle 
third of the body. They have the instinct of bending the anterior 
third so that it rests under the middle one, and thus the eye-spots are 
brought into an appropriate position apparently at the anterior end of 
a somewhat snake-like body. But a caterpillar of this genus which I 
found in Teneriffe assumed the attitude, on irritation, although the 
eye-spots were almost completely wanting. 
1 Tt is worth considering whether the Mullerian principle may have 
been operative in this case. 
2 Of course, no natural selectionist has ever been so unreasonable 
as to contend for absolute protection. In every species, whether 
defended by the most distasteful or dangerous qualities, or the most 
effective concealment, no more can be achieved than to keep up the 
average numbers under average conditions, and this means that an 
immense majority of individuals are doomed to failure. As regards 
concealment, success merely means that enemies have so far to work 
for their living that in the time at their disposal they cannot do more 
than reduce the number of individuals to the average. Warning 
colours and unpalatable or otherwise unpleasant qualities are more 
complex as a means of defence, depending as they do for their success 
upon the co-existence of other more desirable food. Their operation, 
