in, when she spied the caterpillar.
Immediately she began bobbing her head up and down in the odd way which
some small birds have, but was afraid to go nearer. Another joined her
and then another, until at last there was a little company of ten or
twelve birds all looking on in astonishment, but not one ventured into
the tray; while one bird, which lit in it unsuspectingly, beat a hasty
retreat in evident alarm as soon as she perceived the caterpillar. After
waiting for some time, Weismann removed it, when the birds soon attacked
the seeds. Other caterpillars also are probably protected by their
curious resemblance to spotted snakes. One of the large Indian
caterpillars has even acquired the power of hissing.

Among perfect insects many resemble closely the substances near which
they live. Some moths are mottled so as to mimic the bark of trees, or
moss, or the surface of stones. One beautiful tropical butterfly has a
dark wing on which are painted a series of green leaf tips, so that it
closely resembles the edge of a pinnate leaf projecting out of shade
into sunshine.

The argument is strengthened by those cases in which the protection, or
other advantage, is due not merely to colour, but partly also to form.
Such are the insects which resemble sticks or leaves. Again, there are
cases in which insects mimic others, which, for some reason or other,
are less liable to danger. So also many harmless animals mimic others
which are poisonous or otherwise well protected. Some butterflies, as
Mr. Bates has pointed out, mimic others which are nauseous in taste, and
therefore not attacked by birds. In these cases it is generally only the
females that are mimetic, and in some cases only a part of them, so that
there are two, or even three, kinds of females, the one retaining the
normal colouring of the group, the other mimicking another species. Some
spiders closely resemble Ants, and several other insects mimic Wasps or
Hornets.

Some reptiles and fish have actually the power of changing the colour of
their skin so as to adapt themselves to their surroundings.

Many cases in which the colouring does not at first sight appear to be
protective, will on consideration be found to be so. It has, for
instance, been objected that sheep are not coloured green; but every
mountaineer knows that sheep could not have had a colour more adapted to
render them inconspicuous, and that it is almost impossible to
distinguish them from the ro