290 A. FRANKLIN SHULL 
=>? 
€ 
Oo 
re 
=) 
os 
oO 
=} 
3 
2 Ww 
72} (9-2) 
£ 
is 
Ze 
wo 
Saiktor 11 to 15 21 to 25 31 to 35 41 to 46 
Number of Generation 
Fig. 6 
be compared with fig. 2, which shows the proportion of male- 
producers for the same lines. Fig. 6 represents line 5 of table 1, 
and is to be compared with fig. 3. Both curves show a decrease 
in size of family, which is especially marked in fig. 6. It does not 
follow, however, that there is any relation between the two phe- 
nomena, namely, the decrease in the size of family and the 
decrease in the proportion of male-producers with long-continued 
parthenogenesis. 
To discover whether the great fluctuations in the proportion of 
male-producers has any relation to size of family, several lines 
that showed the greatest fluctuations have been examined in 
detail. While in some cases the families, or groups of families, 
showing a great increase in the proportion of male-producers 
over the preceding generations also showed a great increase in 
the size of the family, this was not true in a number of other cases. 
One need not conclude from this that there is no relation between 
size of family (vigor of line) and proportion of male-producers. 
There are many accidents which might happen to a female whose 
family includes many male-producers. I have known a number 
of females to die as a result of accidentally (?) starting to devour 
a large Paramecium or a fibér in the water. The small families 
produced by such females may have included mostly male- 
producers, so that a study of individual families can hardly be 
