THE INHERITANCE OF PRODUCTIVITY IN FARM LIVE STOCK.       289

The real limitation of selection lies in the fact that, owing to the apparent
similarity of production of the homozygous and heterozygous gene groupings,
it is impossible to forecast with accuracy the productive capacity of all offspring
from any mating. To illustrate this with regard to one particular quality, say
persistency, making use of genetical knowledge, let us suppose that only females
showing persistency were selected as breeders ; they would have either the genetic
constitution PP or Pp. There could be no evidence from the male himself as
to his genetic constitution, and so, from matings with persistent females, there
are three groups, each of two classes, into which his resultant progeny might
fall (Table III).

                                               TABLE III

       Parents Offspring

Group 1

      PP♂×PP♀—PP . . .

all persistent and true-breeding.

      PP×Pp—1 PP : 1 Pp . . .

all persistent, but only half true-breeding.

Group 2

      Pp♂ × PP♀—1 PP: 1 Pp . .

all persistent, half true-breeding.

      Pp × Pp—1 PP : 2 PP : 1 pp . .

three persistent to one non-persistent, but

two in three of persistent not true-breeding.

Group 3

      pp♂ × PP♀—Pp . . . .

all persistent, but not true-breeding.

      pp × Pp—1 Pp : 1 pp . . .

one-half persistent, but not true-breeding.

The fact that females exist in the stock which exhibit persistency, and yet
cannot transmit this quality to all their offspring, makes it extremely unlikely
that, using this as a basis of selection, a strain homozygous for the character would
ever be developed, even though care was taken to breed only from birds all of
whose progeny were persistent. Such a procedure could only tend to limit the
selection of breeders to the first three classes, and since only the first is of fixed
constitution, a certain number of undesirable, i.e., recessive individuals would
almost inevitably occur among the succeeding progenies, which might belong
to any of the first four classes.

Genetical work has given us at least one clear conception, namely, that a
bird may exhibit a desirable character, but only be able to transmit it to part
of its offspring. The immediate application of this knowledge could only be
made if further experiment were carried out in sufficient detail to demonstrate
whether measurable differences between the hetero- and homo-zygous forms
exist. If this could be done, then the development of homozygous strains carrying
all the desirable characters, could be rapidly undertaken. Only by this means
can the control of heredity by breeders be readily visualized.