Mimetic Attraction. S27 
it will be sufficient to point out that there is more reason 
to suppose that the Heliconius has adopted certain 
features from the Pieris (for example, the whiteness of 
the ground colour, and the disposition, if not the exist- 
ence, of the basal red marks) than that the converse 
alone has taken place. 
Reeiprocal Mimicry and Convergence. ‘This fact of the 
reciprocal copying of two or more species by each 
other is perhaps implied, though not distinctly so, 
in the term ‘ convergence,” which has been used 
by many authors to express the phenomenon of Miil- 
lerian mimicry; but I am not aware that any writer 
who so employs the term has laid stress on the mutwal 
character of the changes involved, or has traced in 
any instance the actual modifications undergone by 
both species of a Millerian couple under the intluence 
of the attractive force existing between them. It seems 
hitherto to have been taken for granted that a dominant 
form will attract or retain other species within its own 
sphere of influence, without being itself attracted in 
return; whereas the fact is, as we have seen, that each 
member of an inedible association has more or less 
influence upon all the rest. The respective value of the 
attraction exercised or suffered by any member of a 
Miillerian group will depend on its numbers, its nauseous 
qualities, and its notoriety. The stronger any species is 
in these respects, the stronger will be its power of 
attraction, and the weaker in comparison will be any 
force tending to draw it in the direction of other mem- 
bers of the group. The actual mimetic path taken by 
any species will be the resultant of the various forces 
acting uponit. If the form happens to be a dominant one, 
these external forces will be insignificant in comparison 
with its own stability ; and it will therefore resist change 
to a large extent, or perhaps altogether. ‘The most com- 
plete intermingling of characters given and taken on 
both sides may be expected when two species meet on 
equal terms, neither being strong enough to predominate 
over the other. 
While there can be no doubt of the convenience of the 
term “convergence,” and its suitability to express 
relations of the kind just discussed, there would seem to 
be no sufficient reason for disallowing in their case the 
earlier term “ mimicry.” This latter word may be quite 
