756 



THE AMERICAN NATURALIST [Vol. XLVII 



In this comparison, all seeds which germinate at all 

 have been considered viable — although it is practically 

 certain that many of the abnormal ones would not have 

 been able to reach maturity. If one wishes to take the 

 other extreme, he may throw all the seeds producing ab- 

 normal seedlings with those which failed to germinate at 

 all and compare with those germinating in a perfectly 

 normal manner, A — (B + C). 



jii 



ssSF 



Si 

 31 



I now combine the differences (A + B) — (A+B + 

 C 4- • • • ) f or the 50 experiments of the present paper 



