( di 
belong to very different groups, e.g. sub-families, it is im- 
probable that such errors of judgment could lead to any 
important danger. It is very unlikely that a superficial 
resemblance would mislead the individuals of species 
belonging to different sub-families when they approached 
each other at all closely, and the impression made by each 
upon the whole of the sense-organs of the other became 
at all strong. But this would not apply to anything like the 
same extent when there was near relationship between the 
mimetic species—as in so many Jthomiine, Danaine, and 
Heliconine. When close resemblance obtains within the 
limits of such a sub-family as one of these,—and mimetic 
likeness of the kind is often extraordinarily exact,—it is not a 
far-fetched hypothesis to suggest that some special adaptation 
has arisen, enabling the females easily to discriminate between 
their own and the males of other closely similar species, and at 
once to repel those advances which are something of a danger 
and nothing of advantage to either species. Other facts, and 
especially the hard, cell-like structure secreted by the male 
upon the body of the female in Parnassius and in Acreinz, 
also support the conclusion that useless pairing and attempts 
to pair are an injury to the species. Colour and pattern being 
excluded ex hypothest, some special difference in scent is 
the most obvious means of discrimination. May not this be 
the meaning of the fact that the males of the Hupleini may 
be divided into groups (which have been given generic names) 
distinguished, and sometimes solely distinguished, by remark- 
able differences in the size, number, form and position of the 
areas presumed to be scent-producing? These Huplexas are 
remarkable for the number of their synaposematic associations 
and for the closeness of the resemblance between the con- 
stituent species. So far as my experience goes,—and further 
inquiry in the same direction will tend to supply confirmation 
or refutation of the hypothesis here put forward—these 
associations are made up of species belonging to groups with 
different forms of sexual brands and not by species with males 
bearing the same type of brand. And now Mr. Kaye has 
shown that the close synaposematic pairs within the Heliconine 
sub-family are made up of species of which one belongs to the 
