ing direction and amount of his attention
in different cases; finally, it is said that the changes found in nature
are within the limits to which the variation of domestic animals
extends,--it being the case that when changes of a certain amount have
occurred to a species under nature, it becomes _another species_, or
sometimes _two or more other species_ by divergent variations, each of
these species being able again to vary and diverge in any useful direction.

But the fact of the rapidly increasing difficulty found in producing by
ever such careful selection, any further extreme in some charge already
carried very far (such as the tail of the "fan-tailed pigeon" or the crop
of the "pouter"), is certainly, so far as it goes, on the side of the {117}
existence of definite limits to variability. It is asserted in reply, that
physiological conditions of health and life may bar any such further
development. Thus, Mr. Wallace says[109] of these developments: "Variation
seems to have reached its limits in these birds. But so it has in nature.
The fantail has not only more tail-feathers than any of the three hundred
and forty existing species of pigeons, but more than any of the eight
thousand known species of birds. There is, of course, some limit to the
number of feathers of which a tail useful for flight can consist, and in
the fantail we have probably reached that limit. Many birds have the
oesophagus or the skin of the neck more or less dilatable, but in no known
bird is it so dilatable as in the pouter pigeon. Here again the possible
limit, compatible with a healthy existence, has probably been reached. In
like manner, the differences in the size and form of the beak in the
various breeds of the domestic pigeon, is greater than that between the
extreme forms of beak in the various genera and sub-families of the whole
pigeon tribe. From these facts, and many others of the same nature, we may
fairly infer, that if rigid selection were applied to any organ, we could
in a comparatively short time produce a much greater amount of change than
that which occurs between species and species in a state of nature, since
the differences which we do produce are often comparable with those which
exist between distinct genera or distinct families."

But in a domestic bird like the fantail where Natural Selection does not
come into play, the tail-feathers could hardly be limited by "utility for
flight," yet two more tail-feathers could