but turn them loose,
and they will all seek the barnyard, and soon your fancy breeds will
become corrupt. They "revert to type." By the exercise of intelligent
selection and training, man is able to emphasize certain points and to
produce new breeds, but not to change the essential structure nor to
alter the specific characteristics. The species are _fixed_. Huxley says:

"If you breed from the male and female of the same race, you of course
have offspring of the like kind, and if you make the offspring breed
together, you obtain the same result, and if you breed from these again,
you will still have the same kind of offspring; there is no check. But
if you take members of two distinct species, however similar they may be
to each other, and make them breed together, _you will find a check_. If
you cross two such species with each other, then--although you may get
offspring in the case of the first cross, yet, if you attempt to breed
from the products of that crossing, which are what are called hybrids--
that is, if you couple a male and a female hybrid--then the result is
that in ninety-nine cases out of a hundred you will get no offspring at
all; there will be no result whatsoever.

"The reason of this is quite obvious in some cases; the female hybrids,
although possessing all the external appearances and characteristics of
perfect animals, are physiologically imperfect and deficient in the
structural parts of the reproductive elements necessary to generation.
It is said to be invariably the case with the male mule, the cross
between the ass and the mare; and hence it is that although crossing
the horse with the ass is easy enough, and is constantly done as far as
I am aware, if you take two mules, a male and a female, and endeavor to
breed from them, you get no offspring whatever; no generation will take
place. This is what is called the sterility of the hybrids between two
distinct species." (Huxley, _"On the Origin of Species."_ p. 212.) He
continues:

"Thus you see that there is a great difference between 'mongrels,' which
are crosses between distinct races, and 'hybrids,' which are crosses
between distinct species. The mongrels are, so far as we know, fertile
with one another. But between species, in many cases, you cannot succeed
in obtaining even the first cross; at any rate it is quite certain
that the hybrids are often absolutely infertile one with another.

"Here is a feature, then, great or small as it m