374 sasaki ICAL GAZETTE [NOVEMBER 
done by him with the China bean (Stisolobium niveum var. ?), 
commonly considered as a variety of the Lyon bean. Perhaps it 
is not well to go into more detail, but rather to allow the reader 
the privilege of consulting the article for fuller details. His con- 
clusion is that, generally speaking, 50 per cent of the pollen of the 
F, generation loses its contents and becomes sterile. A similar 
degree of sterility seems also to hold for the embryo sacs. Where 
the experiment was carried to the F, and F, generations, it was 
found that the offspring of the fertile half all bred true to fertility, 
while those of the sterile half all persisted in having 50 per cent 
- sterile and 50 per cent fertile. The ratio here, as the investigator 
points out, is 1:1. As to the value of the results obtained, I am 
not in a position to judge. However, the fact that the investigator 
is dealing with the haploid generation instead of the usual diploid 
generation lends a new interest to the matter and opens up a possible 
new field of research. 
Thus far I have attempted to give a general idea of some of the 
work done heretofore upon sterility in pollen, and its relation to 
hybridization. The work of DEVriks and his “‘ Mutationstheorie” 
resulting therefrom have given a great impetus to a search for the 
true means of evolution in living matter. The Darwinian idea that 
species arise through a gradual loss or gain of characters and the 
survival of the fittest of these has been very seriously attacked. 
The “sporting” condition which Darwin recognized as occurring 
sometimes, but which he considered to be of far less importance 
than the changes brought about gradually, has been made to play 
a very much more important part by DeVries’ mutation idea. 
DeVries bases his theory primarily upon the much discussed 
species Oenothera Lamarckiana, which he had been growing for 
several generations in his garden, and of whose origin he is in no 
way certain. As I have already shown, sterile pollen, though not 
necessarily present in hybrids, is a common character. Recently 
Jerrrey (21) has shown that, so far as his examination has eX- 
tended, morphologically sterile pollen does not occur, except in 4 
percentage small enough to be accounted for by unfavorable con- 
ditions, in plants which are either monotypic or are isolated either 
in a geographical sense or through the time in which the flowers 
