xhibiting in both sets of measurements a
variation of more than one-sixth, combined with a very considerable
amount of independent variability.[24]

[Illustration: FIG. 15.--Variation of 8 skulls (Ursus labiatus).]

[Illustration: FIG. 16.]

The few facts now given, as to variations of the internal parts of
animals, might be multiplied indefinitely by a search through the
voluminous writings of comparative anatomists. But the evidence already
adduced, taken in conjunction with the much fuller evidence of variation
in all external organs, leads us to the conclusion that wherever
variations are looked for among a considerable number of individuals of
the more common species they are sure to be found; that they are
everywhere of considerable amount, often reaching 20 per cent of the
size of the part implicated; and that they are to a great extent
independent of each other, and thus afford almost any combination of
variations that may be needed.

It must be particularly noticed that the whole series of
variation-diagrams here given (except the three which illustrate the
number of varying individuals) in every case represent the actual amount
of the variation, not on any reduced or enlarged scale, but as it were
life-size. Whatever number of inches or decimals of an inch the species
varies in any of its parts is marked on the diagrams, so that with the
help of an ordinary divided rule or a pair of compasses the variation of
the different parts can be ascertained and compared just as if the
specimens themselves were before the reader, but with much greater ease.

In my lectures on the Darwinian theory in America and in this country I
used diagrams constructed on a different plan, equally illustrating the
large amount of independent variability, but less simple and less
intelligible. The present method is a modification of that used by Mr.
Francis Galton in his researches on the theory of variability, the upper
line (showing the variability of the body) in Diagrams 4, 5, 6, and 13,
being laid down on the method he has used in his experiments with
sweet-peas and in pedigree moth-breeding.[25] I believe, after much
consideration, and many tedious experiments in diagram-making, that no
better method can be adopted for bringing before the eye, both the
amount and the peculiar features of individual variability.

_Variations of the Habits of Animals._

Closely connected with those variations of internal and external
struc