EFFECTS OF INBREEDING AND CROSSBREEDING. 45 
heterozygosis will remain true indefinitely. Summing up, a random- 
bred stock derived from n inbred families will have nth less superiority 
over its inbred ancestry than the first cross or a random-bred stock 
from which the inbred families might have been derived without 
selection. 
COMPARISON OF RESULTS WITH THEORY. 
In the foregoing theoretical considerations we have assumed that 
the characteristics depend on the heredity of the progeny produced 
by the mating in question. It will be recalled, however, that we 
found good reason for believing that most of the characteristics in 
the experiments with guinea pigs were really determined wholly or in 
part by the parents, in most cases the dam, as far as these character- 
istics were genetic at all. In a character determined wholly by the 
dam, Experiments CO and CA belong with the inbreds, CC, AC, and 
C1 represent the first cross, and C2 represents the first generation of 
Fic. 26.—The percentage of heterozygosis following matings among inbred families of the types PQ x RS 
PQ x QR,and PQ = PQ. Criginal families, x of type AA and y of type aa. 
renewed inbreeding. In characters determined partly by the parents 
and partly by the progeny we can easily find the relations which are 
to be expected by combining the expectations based on the breeding 
of the dam and sire with those based on the progeny. Table 13 
shows the expectation in certain cases, while Figure 27 presents some 
of the same conclusions graphically. 
