arked variety or into a distinct species; and there
would evidently be little chance of this modification being checked by
intercrossing with the parent form which remained in the forest.

Another mode of isolation is brought about by the variety--either owing
to habits, climate, or constitutional change--breeding at a slightly
different time from the parent species. This is known to produce
complete isolation in the case of many varieties of plants. Yet another
mode of isolation is brought about by changes of colour, and by the fact
that in a wild state animals of similar colours prefer to keep together
and refuse to pair with individuals of another colour. The probable
reason and utility of this habit will be explained in another chapter,
but the fact is well illustrated by the cattle which have run wild in
the Falkland Islands. These are of several different colours, but each
colour keeps in a separate herd, often restricted to one part of the
island; and one of these varieties--the mouse-coloured--is said to breed
a month earlier than the others; so that if this variety inhabited a
larger area it might very soon be established as a distinct race or
species.[48] Of course where the change of habits or of station is still
greater, as when a terrestrial animal becomes sub-aquatic, or when
aquatic animals come to live in tree-tops, as with the frogs and
Crustacea described at p. 118, the danger of intercrossing is reduced to
a minimum.

Several writers, however, not content with the indirect effects of
isolation here indicated, maintain that it is in itself a cause of
modification, and ultimately of the origination of new species. This
was the keynote of Mr. Vernon Wollaston's essay on "Variation of
Species," published in 1856, and it is adopted by the Rev. J.G. Gulick
in his paper on "Diversity of Evolution under one Set of External
Conditions" (_Journ. Linn. Soc. Zool._, vol. xi. p. 496). The idea seems
to be that there is an inherent tendency to variation in certain
divergent lines, and that when one portion of a species is isolated,
even though under identical conditions, that tendency sets up a
divergence which carries that portion farther and farther away from the
original species. This view is held to be supported by the case of the
land shells of the Sandwich Islands, which certainly present some very
remarkable phenomena. In this comparatively small area there are about
300 species of land shells, almost all o