100 Self-Incompatibility in Hermaphrodite Plants 
others. With only one exception the self-compatibility was feeble. It 
is clear that self-compatibility exhibited no tendency to develop as an 
“end season ” fertility. 
A most striking type of seed sterility was seen in the considerable , 
number of plants that produced pods which were empty. In the feebly 
self-fertile plants, as a rule, the few good seeds that were developed were 
located in the upper end of the capsules. The considerable number of 
poor seed present both in pods containing few seeds, and in empty pods, 
suggests that incompatibility may operate after fertilization, producing 
embryo abortion. The results here obtained also suggest that some 
seedless pods may develop as a direct result of the stimulus of pollen 
tube growth, and more obviously that the size of pods bears a relation 
to the number of ovules fertilized. 
Nicotiana Forgetiana Hyb. Hort. 
Series 1. In 1917, the writer, assisted by Miss Helene Boas (now 
Mrs. Cecil Yampolsky), tested the self-compatibility of 58 plants of 
N. Forgetiana Hyb. Hort. Controlled self-pollinations were made by 
hand on successive dates using the same method employed for Verbascum 
phoeniceum, except that bags were shifted from branch to branch. In 
all cases the pollinations were begun during the first days of bloom. : 
On the basis of their self-compatibility and incompatibility these 
plants may be grouped in three classes. - 
(1) Eleven plants were highly self-fertile from the first day of 
controlled self-pollination. There were only 3 flowers out of 95 selfed 
that failed to produce pods. The capsules were large and well filled 
with seeds. As soon as it was found that a plant was self-fertile no 
further controlled pollinations were made. 
(2) Eight plants were feebly self-compatible. Capsules were pro- 
duced in relatively few of the selfed flowers, and these were small an 
contained few seeds. For 6 plants some pods developed in flowers of 
the first date of selfing; in the other 2 plants pods were formed from 
flowers of the second date of selfing. All these plants, however, were 
partially self-compatible during the first days of bloom. : 
(3) Thirty-nine plants were completely self-incompatible to all 
controlled selfings made, which were extended to well past the climax — 
of bloom in all cases and for 10 plants the selfings were continued to 
the last date of bloom. 
The plants of this series were grown from commercial seed and there 
was considerable variation in the colour of the flowers. Several con- 
