of selection, we ought not to expect to find great and
constant differences in the skeleton; for fanciers can neither see, nor do
they care for, modifications of structure in the internal framework. Nor
ought we to expect changes in the skeletons from hanged habits of life; as
every facility is given to the most distinct breeds to follow the same
habits, and the much modified races are never allowed to wander abroad and
procure their own food in various ways. Moreover, I find, on comparing the
skeletons of _Columba livia_, _oenas_, _palumbus_, and _turtur_, which are
ranked by all systematists in two or three distinct though allied genera,
that the differences are extremely slight, certainly less than between the
skeletons of some of the most distinct domestic breeds. How far the
skeleton of the wild rock-pigeon is constant I have no means of judging, as
I have examined only two.

_Skull._--The individual bones, especially those at the base, do not
differ in shape. But the whole skull, in its proportions, outline, and
relative direction of the bones, differs greatly in some of the breeds,
as may be seen by comparing the figures of (A) the wild rock-pigeon,
(B) the {164} shortfaced tumbler, (C) the English carrier, and (D) the
Bagadotten carrier (of Neumeister), all drawn of the natural size and
viewed laterally. In the carrier, besides the elongation of the bones
of the face, the space between the orbits is proportionally a little
narrower than in the rock-pigeon. In the Bagadotten the upper mandible
is remarkably arched, and the premaxillary bones are proportionally
broader. In the short-faced tumbler the skull is more globular; all the
bones of the face are much shortened, and the front of the skull and
descending nasal bones are almost perpendicular; the maxillo-jugal arch
and premaxillary bones form an almost straight line; the space between
the prominent edges of the eye-orbits is depressed. In the barb the
premaxillary bones are much shortened, and their anterior portion is
thicker than in the rock-pigeon, as is the lower part of the nasal
bone. In two nuns the ascending branches of the premaxillaries, near
their tips, were somewhat attenuated, and in these birds, as well as in
some others, for instance in the spot, the occipital crest over the
foramen was considerably more prominent than in the rock-pigeon.

[Illustration: Fig. 25