ding to retain an
approved type. An outcross is introduced when the breed operated upon
is declining in stamina or is in danger of extinction, or when some
new physical or mental quality is desired. New types and eccentricities
are hardly wanted, however, and the extreme requirements of an
outcross may nowadays be achieved by the simple process of selecting
individuals from differing strains of the same breed, mating a bitch
which lacks the required points with a dog in whose family they are
prominently and consistently present.

Inbreeding is the reverse of outcrossing. It is the practice of mating
animals closely related to each other, and it is, within limits, an
entirely justifiable means of preserving and intensifying family
characteristics. It is a law in zoology that an animal cannot transmit
a quality which it does not itself innately possess, or which none of
its progenitors has ever possessed. By mating a dog and a bitch of the
same family, therefore, you concentrate and enhance the uniform
inheritable qualities into one line instead of two, and you reduce the
number of possibly heterogeneous ancestors by exactly a half right
back to the very beginning. There is no surer way of maintaining
uniformity of type, and an examination of the extended pedigree of
almost any famous dog will show how commonly inbreeding is practised.
Inbreeding is certainly advantageous when managed with judgment and
discreet selection, but it has its disadvantages also, for it is to be
remembered that faults and blemishes are inherited as well as merits,
and that the faults have a way of asserting themselves with annoying
persistency. Furthermore, breeding between animals closely allied in
parentage is prone to lead to degeneracy, physical weakness, and
mental stupidity, while impotence and sterility are frequent
concomitants, and none but experienced breeders should attempt so
hazardous an experiment. Observation has proved that the union of
father with daughter and mother with son is preferable to an alliance
between brother and sister. Perhaps the best union is that between
cousins. For the preservation of general type, however, it ought to be
sufficient to keep to one strain and to select from that strain
members who, while exhibiting similar characteristics, are not
actually too closely allied in consanguinity. To move perpetually from
one strain to another is only to court an undesirable confusion of
type.

In founding a kenne