og as dark as can be obtained, and bitches from this
mating breed to a seal brindle dog.

No. 8. Steel and tiger brindles I class together, as the process is the
same and results are easy. Breed first to a red brindle dog; bitches from
this union to a dark mahogany brindle, and then use seal brindle dog on
bitch from last mating.

No. 9. Red brindle. No skill is required here. Breed first to mahogany
brindles, and bitches from this union to seal brindles.

We have now enumerated practically all the less desirable shades, but let
me observe in passing, in the process of color breeding that the law of
atavism, or "throwing back," often asserts itself, and we shall see colors
belonging to a far-off ancestry occasionally presenting themselves in all
these matings. Once in a while a dog will be found that no matter what
color bitches he may be mated with, he will mark a certain number of the
litter with the peculiar color or markings of some remote ancestor. Just a
case apropos of this will suffice. We used in our kennels a dog of perfect
markings, coming from an immediate ancestry of perfectly marked dogs, and
mated him with quite a number of absolutely perfectly marked bitches that
we had bred for a great number of years that had before that had perfectly
marked pups, and every bitch, no matter how bred, had over fifty per cent.
of white headed pups. We saw the pups in other places sired by this dog,
no matter where bred, similarly marked. We found his grandmother was a
white headed dog, and this dog inherited this feature in his blood, and
passed it on to posterity. The minute a stud dog, perfect in himself, is
prepotent to impress upon his offspring a defect in his ancestry, discard
him at once. I have often been amused to see how frequently this law of
atavism is either misunderstood or ignored. Only recently I have seen a
number of letters in a leading dog magazine, in which several people who
apparently ought to know better, were accusing litters of bulldog pups as
being of impure blood because there were one or two black pups amongst
them. They must, of course, have been conversant with the fact that
bulldogs years ago frequently came of that color, and failed to reason
that in consequence of this, pups of that shade are liable once in a while
to occur. It is always a safe rule in color breeding to discard as a stud
a dog, no matter how brilliant his coat may be, who persistently sires
pups whose colors are indistin