Inheritance of Rest Period of Seeds in the Peanut
with very little in the means of greater parents. Less dormant
parents have quite significantly negative average breeding values
as measured by means of crosses with more dormant types.
4. Transgressive segregation above the greater parent is evi-
dent in most crosses. Particularly, F2 family No. 77 listed at
the bottom of Table 5 has a mean gain over its greater parent
of 111 days with standard error of 15 days. Standard error of
gain is calculated from two near normal distributions. The gain
is seven times its standard error and equal to 72 percent of the
greater parent. Genetic variance in F2 family No. 77 is evident
from greater variance of its F4 seeds in comparison with the
pure strains of Table 2 and in analysis of its variance in Table
7. Decline of variance due to selfing of F3 plants is shown in
the ratio of variance of F4 seeds between F3 families to the vari-
ance within F3 families to be highly significant. Considerably
longer rest periods than the mean of No. 77 are easily possible
in its progeny. Yet the less dormant parent is strain No. 1
which produced more than 95 percent of seeds with zero rest
period and is the least dormant pure strain tested. It may be
said further that the mean of No. 77 is a highly significant gain
over that of any parent strain in the cultures and that it gives
unmistakable evidence of Spanish ancestry. No possibility of
admixture or random outcrossing can abrogate its evidence that
the less dormant parent strains possess considerable genic ma-
terial whose effect is in the direction of increasing rest period.
5. Transgressive segregation below the lesser parent is not
easily demonstrated. The best evidence is in the two F3 families
at the top of Table 5. With 792 seeds tested from both families
the latest seed emerged 48 days after planting. In the lesser
parent, top of Table 1, with 193 seeds tested one emerged at
68 and one at 132 days after planting.
6. Retaining attention on the same three strains, it is noted
that their variances are very small. Also, in Table 4, 30 percent
of the 944 F3 families averaged less than 17 days from planting
to emergence of their F4 seeds. Variances of such families are
necessarily small. It is not only necessary to explain how posi-
tive experimental errors become so drastically restricted in
these samples but also the failure of genetic variance to appear
in so many hybrid families. The extreme plus F2 family of the
same cross has been discussed and its large proportion of genetic
variance noted.