n it. The ultimate reason why one and
the same insect should occur in green and in brown, as often happens
in caterpillars and locusts, lies in the fact that variations towards
brown presented themselves, and so also did variations towards green:
_the kernel of the riddle lies in the varying_, and for the present we
can only say, that small variations in different directions present
themselves in every species. Otherwise so many different kinds of
variations could not have arisen. I have endeavoured to explain this
remarkable fact by means of the intimate processes that must take
place within the germ-plasm, and I shall return to the problem when
dealing with "germinal selection."

We have, however, to make still greater demands on variation, for it
is not enough that the necessary variation should occur in isolated
individuals, because in that case there would be small prospect of its
being preserved, notwithstanding its utility. Darwin at first
believed, that even single variations might lead to transformation of
the species, but later he became convinced that this was impossible,
at least without the cooeperation of other factors, such as isolation
and sexual selection.

In the case of the _green caterpillars with bright longitudinal
stripes_, numerous individuals exhibiting this useful variation must
have been produced to start with. In all higher, that is,
multicellular organisms, the germ-substance is the source of all
transmissible variations, and this germ-plasm is not a simple
substance but is made up of many primary constituents. The question
can therefore be more precisely stated thus: How does it come about
that in so many cases the useful variations present themselves in
numbers just where they are required, the white oblique lines in the
leaf-caterpillar on the under surface of the body, the accompanying
coloured stripes just above them? And, further, how has it come about
that in grass caterpillars, not oblique but longitudinal stripes,
which are more effective for concealment among grass and plants, have
been evolved? And finally, how is it that the same Hawk-moth
caterpillars, which to-day show oblique stripes, possessed
longitudinal stripes in Tertiary times? We can read this fact from the
history of their development, and I have before attempted to show the
biological significance of this change of colour.[38]

For the present I need only draw the conclusion that one and the same
caterpillar may e