Mimetie Patterns to the Original Form. 73 
Pierine imitators. What is the meaning of this coinci- 
dence? The first answer that suggests itself is that it is 
simply an ordinary case of mimicry ; the red spots belong 
originally to the Heliconius, and the Pierine has acquired 
similar spots in order to complete the mimetic picture. 
‘wo facts, however, militate against this supposition. 
The first is that these red patches, so far from being 
confined to the mimicking Pierines, are found to have 
a very wide distribution throughout the whole Pierine 
subfamily, existing not only, as we have seen, in non- 
mimetic neotropical forms such as Pieris locusta and 
P. phaloe, but in numerous old-world genera as well, 
reaching a great development in the Indian and Austra- 
lian Delias, and having even left a relic in the common 
white butterflies of our own country. It would be extra- 
vagant to suppose that these widespread characters owe 
their origin simply to the necessity for mimicking certain 
South American Heliconii. Moreover, as I have else- 
where shown, such an origin for the old-world forms as 
this hypothesis would involve is at variance with what is 
known of Pierine phylogeny. The second fact is that 
although several Heliconii which are not the subjects of 
mimicry show marks of the kind, yet they are most con- 
stant, most distinct and most Pierine-like in species of 
Heliconius that serve as models. ‘There must, it would 
seem, be a relation between the two forms which is not 
entirely due to mimicry by the Pierine. Are we then to 
say that the Heliconius is the mimic and the Pierine the 
model? ‘This would appear to be going against all 
received ideas on the subject, and to be negatived by ali 
that is known of the inedible qualities of Heliconius and 
of the ancestral coloration of the Pierines ; nevertheless, 
with respect to the particular marks in question I believe 
that it comes near to the true expression of the fact, and 
I would suggest that the key to the difficulty is to be 
found in the following considerations. 
It has been well shown by Fritz Miiller,* whose con- 
clusions have been followed and amplified by Meldola and 
Poulton, that there exist two kinds of mimetic associa- 
tions—in one of which an edible form shelters itself by 
resemblance to another form well known to be inedible, 
this being the aspect of mimicry first detected and 
explained by Bates; while in the other a group is 
constituted all of whose members are inedible, and join 
* “ Kosmos,” 1879, p. 100. 
