s of
modification and variation had better leave the argument of recurrence
alone, or it may prove altogether too strong for them.

To sum up,--the evidence as far as we have gone is against the argument
as to any limit to divergences, so far as structure is concerned; and
in favour of a physiological limitation. By selective breeding we can
produce structural divergences as great as those of species, but we
cannot produce equal physiological divergences. For the present I leave
the question there.

Now, the next problem that lies before us--and it is an extremely
important one--is this: Does this selective breeding occur in nature?
Because, if there is no proof of it, all that I have been telling you
goes for nothing in accounting for the origin of species. Are natural
causes competent to play the part of selection in perpetuating
varieties? Here we labour under very great difficulties. In the last
lecture I had occasion to point out to you the extreme difficulty of
obtaining evidence even of the first origin of those varieties which we
know to have occurred in domesticated animals. I told you, that almost
always the origin of these varieties is overlooked, so that I could only
produce two of three cases, as that of Gratio Kelleia and of the Ancon
sheep. People forget, or do not take notice of them until they come to
have a prominence; and if that is true of artificial cases, under our
own eyes, and in animals in our own care, how much more difficult it
must be to have at first hand good evidence of the origin of varieties
in nature! Indeed, I do not know that it is possible by direct evidence
to prove the origin of a variety in nature, or to prove selective
breeding; but I will tell you what we can prove--and this comes to the
same thing--that varieties exist in nature within the limits of species,
and, what is more, that when a variety has come into existence in
nature, there are natural causes and conditions, which are amply
competent to play the part of a selective breeder; and although that
is not quite the evidence that one would like to have--though it is
not direct testimony--yet it is exceeding good and exceedingly powerful
evidence in its way.

As to the first point, of varieties existing among natural species, I
might appeal to the universal experience of every naturalist, and of any
person who has ever turned any attention at all to the characteristics
of plants and animals in a state of nature; but I ma