hotographic or chemical
action on the skin or by a reflex action through vision. The other is,
that the white colour is chiefly beneficial as a means of checking
radiation and so preserving animal heat during the severity of an arctic
winter. The first is part of the general theory that colour is the
effect of coloured light on the objects--a pure hypothesis which has, I
believe, no facts whatever to support it. The second suggestion is also
an hypothesis merely, since it has not been proved by experiment that a
white colour, _per se_, independently of the fur or feathers which is so
coloured, has any effect whatever in checking the radiation of low-grade
heat like that of the animal body. But both alike are sufficiently
disproved by the interesting exceptions to the rule of white coloration
in the arctic regions, which exceptions are, nevertheless, quite in
harmony with the theory of protection.

Whenever we find arctic animals which, from whatever cause, do not
require protection by the white colour, then neither the cold nor the
snow-glare has any effect upon their coloration. The sable retains its
rich brown fur throughout the Siberian winter; but it frequents trees at
that season and not only feeds partially on fruits or seeds, but is able
to catch birds among the branches of the fir-trees, with the bark of
which its colour assimilates. Then we have that thoroughly arctic
animal, the musk-sheep, which is brown and conspicuous; but this animal
is gregarious, and its safety depends on its association in small herds.
It is, therefore, of more importance for it to be able to recognise its
kind at a distance than to be concealed from its enemies, against which
it can well protect itself so long as it keeps together in a compact
body. But the most striking example is that of the common raven, which
is a true arctic bird, and is found even in mid-winter as far north as
any known bird or mammal. Yet it always retains its black coat, and the
reason, from our point of view, is obvious. The raven is a powerful bird
and fears no enemy, while, being a carrion-feeder, it has no need for
concealment in order to approach its prey. The colour of the raven and
of the musk-sheep are, therefore, both inconsistent with any other
theory than that the white colour of arctic animals has been acquired
for concealment, and to that theory both afford a strong support. Here
we have a striking example of the exception proving the rule.

In th