296 A. FRANKLIN SHULL 
ever, as the food cultures were frequently changed, each of these 
recorded lines must have been fed from three or more cultures. 
The averages of these three or more cultures should, I believe, be 
nearly equal. The evidence, though not complete, is presented 
for what it is worth. Line A included 54.7 per cent of male- 
producers, the inbred line 55.7 per cent. 
Incidentally it may be pointed out that the average size of 
family in line A is 35.8, in the inbred line 37.4. 
On the whole it seems that inbreeding does not markedly alter 
the proportion of male-producers, though perhaps had the inbred 
lines been bred as long as were the original lines, the decrease in 
the proportion of male-producers which sometimes accompanies 
long-continued parthenogenesis might have shown aslight decrease 
in that proportion as a result of inbreeding. 
Effect of long duration of the fertilized egg stage on the proportion of 
male-producers 7 
DURATION OF THE FERTILIZED Eaa Stace. Whereas parthe- 
nogenetic eggs hatch pretty uniformly in twelve to fourteen hours 
after laying, great variability has been found in the length of time 
which fertilized eggs from the same source, or from different 
sources, spend in-the egg stage. Thus, in the cross between New 
York females and Baltimore males described in an earlier paper 
(Shull, ’11 a), 408 eggs were obtained from 38 matings made from 
May 14 to May 17, inclusive. On May 24, three of these eggs 
hatched; and each day thereafter, with four exceptions, up to 
June 10, one or more eggs hatched. In the seventeen days from 
May 24 to June 10, 53 eggs, laid by 19 out of the 38 females, 
hatched. The remaining 355 eggs were kept two months longer, 
to August 10, but no more of them hatched, and the lot was then 
discarded. In no case did all the eggs laid by one female hatch, 
the highest record being six out of seven; while 19 of the females 
laid eggs none of which hatched within twelve weeks. It is prob- 
able that many of these would never have hatched. 
About the same time, May 12 to May 15, 25 females of the Bal- 
timore line were inbred, that is, paired with males of the same line. 
