MIMICRY IN CEYLON BUTTERFLIES. 5) 
Wallace came also an interpretation of this peculiar case in terms 
of the theory of’mimicry then just suggested by Bates. Briefly, 
that interpretation is as follows :— 
P. polytes is a palatable insect. The larva feeds on citronaceous 
shrubs and trees and, in its later stages, is inconspicuous upon its food 
plant (cf. Pl. I., fig. 7). The chrysalis may be regarded as protec- 
tively coloured (PI. I., figs. 8, 9).* Yet in this presumably palatable 
insect there exist two additional forms of female, which are character- 
ized not only by marked divergence from the normal type, but by the 
conspicuous form which that divergence takes. So far as can be 
seen there is no suggestion of ordinary protective colouration here. 
Now, living side by side with ‘this species are two other species of 
Papilionid butterflies, Papilio aristolochic (Pl. I., fig. 5) and Papiho 
hector (P1. 1., fig. 6), each of which bears a strong resemblance to one 
of the two aberrant forms of female of P. polytes.t Both of these 
forms have conspicuously coloured red and black larvee, which both 
feed upon the poisonous Aristolochia plants. Both are common 
species, and both consequently fulfil the conditions of abundance and 
distastefulness which the theory of mimicry exacts from qualified 
models. By their resemblance to these two unpalatable species 
the “ mimicking” forms of P. polytes have been enabled to cheat 
their enemies and to preserve their species. And the case is the 
more striking in that while P. hector and the hector form of P. 
polytes are confined to India and Ceylon, both P. aristolochie and 
the aristolochiz form of P. polytes have a wider range eastward. 
For the upholders of the mimicry interpretation the resemblance 
between the model and its mimic would appear to have been brought 
about by the piling up of minute variations in the required direction 
through a process of survival of those most like the model. Upon 
the adequacy of this conception I do not wish to dwell, until I have 
offered some criticisms derived from personal experience with 
reference to the resemblance obtaining between the “ mimics ” 
and their “ models.” 

* The colour is very variable, though whether this is in relation to the 
surfaces on which it pupates is at present unknown. The two specimens 
figured here were bred by me under conditions which in so far as could be 
seen were exactly similar, though the one became clear green and the other 
a darkish brown. 
+ In the account which follows I have used the terms ‘‘ male form, aristo- 
lochiz form,” and “* hector form,” respectively, for these three females, terms 
which indicate sufficiently which form is meant for the reader who is not 
familiar with this species. Technically these three forms are respectively the 
pammon, polytes, and romulus forms of the species P. polytes (cf. ** Fauna of 
British India, Butterflies,’ vol. II., pp. 61, 62). 
{ Though placed in the same genus as P. polytes these two species differ 
from it in many structural points, and will doubtless eventually find their way 
into another genus when the classification of the family has been placed 
on a more satisfactory basis. They are closely allied to each other and come 
into Haase’s group of Pharmacophagus or “‘ Poison-eaters.’’—(** Bibliotheca 
Zoologica,’ 1891.) 
29 66 
