be
more considerable than it appears, an inadequate idea is given of the
relative differences of size. The most instructive of the groups, I
think, is that of the Canaries; because the many and great changes in
different directions must in this case have been produced by artificial
selection in so comparatively short a time--the first mention of this
bird that I can find being by Gesner, in the sixteenth century.

* * * * *

Now, it is surely unquestionable that in these typical proofs of the
efficacy of artificial selection in the modification of specific types,
we have the strongest conceivable testimony to the power of natural
selection in the same direction. For it thus appears that wherever
mankind has had occasion to operate by selection for a sufficiently long
time--that is to say, on whatever species of plant or animal he chooses
thus to operate for the purpose of modifying the type in any required
direction,--the results are always more or less the same: he finds that
all specific types lend themselves to continuous deflection in any
particulars of structure, colour, &c., that he may desire to modify.

Nevertheless, to this parallel between the known effects of artificial
selection, and the inferred effects of natural selection, two objections
have been urged. The first is, that in the case of artificial selection
the selecting agent is a voluntary intelligence, while in the case of
natural selection the selecting agent is Nature herself; and whether or
not there is any counterpart of man's voluntary intelligence in nature
is a question with which Darwinism has nothing to do. Therefore, it is
alleged, the analogy between natural selection and artificial selection
fails _ab initio_, or at the fountain-head of the causes which are taken
by the analogy to be respectively involved.

The second objection to the analogy is, that the products of artificial
selection, closely as they may resemble natural species in all other
respects, nevertheless present one conspicuous and highly important
point of difference: they rarely, if ever, present the physiological
character of mutual infertility, which is a character of extremely
general occurrence in the case of natural species, even when these are
most nearly allied.

I will deal with these two objections in the next chapter, where I shall
be concerned with the meeting of all the objections which have ever been
urged against the theory