r offspring always reproducing the same
types; so that at the end of many generations the men would remain pure
white, and the women of the same well-marked races as at the
commencement.

The distinctive character therefore of dimorphism is this, that the
union of these distinct forms does not produce intermediate varieties,
but reproduces the distinct forms unchanged. In simple varieties, on the
other hand, as well as when distinct local forms or distinct species are
crossed, the offspring never resembles either parent exactly, but is
more or less intermediate between them. Dimorphism is thus seen to be a
specialized result of variation, by which new physiological phenomena
have been developed; the two should therefore, whenever possible, be
kept separate.

3. _Local form, or variety._--This is the first step in the transition
from variety to species. It occurs in species of wide range, when groups
of individuals have become partially isolated in several points of its
area of distribution, in each of which a characteristic form has become
more or less completely segregated. Such forms are very common in all
parts of the world, and have often been classed by one author as
varieties, by another as species. I restrict the term to those cases
where the difference of the forms is very slight, or where the
segregation is more or less imperfect. The best example in the present
group is Papilio Agamemnon, a species which ranges over the greater part
of tropical Asia, the whole of the Malay archipelago, and a portion of
the Australian and Pacific regions. The modifications are principally of
size and form, and, though slight, are tolerably constant in each
locality. The steps, however, are so numerous and gradual that it would
be impossible to define many of them, though the extreme forms are
sufficiently distinct. Papilio Sarpedon presents somewhat similar but
less numerous variations.

4. _Co-existing Variety._--This is a somewhat doubtful case. It is when
a slight but permanent and hereditary modification of form exists in
company with the parent or typical form, without presenting those
intermediate gradations which would constitute it a case of simple
variability. It is evidently only by direct evidence of the two forms
breeding separately that this can be distinguished from dimorphism. The
difficulty occurs in Papilio Jason, and P. Evemon, which inhabit the
same localities, and are almost exactly alike in form, size, a