756 Tsabel McCracken 
to denote the potency or strength of its opponent bivoltinism, 
It looks as though by long selection univoltinism had become 
prepotent but that the possibly older bivoltinism, when once in- 
troduced is able gradually to reéstablish itself in the heredity. 
The question now arises, Is the condition of individuals taken 
as separate entities coincident with the condition of the average 
of the series? In 1907, and again in 1908, precaution wastaken 
to make a number of matings within each brood, to throw light 
on this query. It was found that all the tested females furnished 
by a few broods, particularly in 1908, were bivoltin-producing. 
Many of the broods furnished univoltin-producing females only 
and others furnished females part of whom were bivoltin-producing 
and part of whom were univoltin-producing. Of the latter group, 
sometimes univoltin producers predominated and sometimes 
bivoltin producers predominated. In each case these females 
were mated with males of similar ancestry. ‘This diversity of 
behavior is in general true not of any one series alone, but of 
all series. . 
For exact and graphic comparison, | have selected eight broods 
in Series E of 1907 and ten broods in Series H, I and J of 1908, 
from each of which ten mothers were chosen as parents. Diagram 
I shows the comparative number of univoltin and bivoltin broods 
in the former group (Series E). In this diagram the vertical 
distance in each case indicates the number of bivoltin-producing 
mothers (out of the total of ten mothers) furnished by each of 
the eight broods respectively: 
10 ee 
0 
ALB iC. Da Ey Grease 
Diagram1. Showing distribution of univoltin- and bivoltin-producing mothers in each of eight broods 
