1078 
Journal of Agricultural Research 
Vol. XXVIII, No. 11 
grown from seed matured in the dark-house. The classification of the plants as 
to sex when they first bloomed showed 164 carpellate : 162 staminate. In a few 
days after the first flowers opened several hermaphrodite buds were observed, 
the phenomenon being more common among the staminate individuals. At 
once the plants were separated into several lots in order to study, in detail, the 
effect of flower removal during the short days on sexual development. 
Early in the course of the experiments it became evident that the material used 
was sexually different from that used by Pritchard and Schaffner in their work. 
They state that the tendency is apparently from femaleness to maleness but in 
the material used in the writer’s experiments the tendency appears to be from 
maleness to femaleness. Only 12.8 per cent of the carpellate plants developed 
any staminate buds while 53.12 per cent of the staminate plants developed 
some carpellate flowers. The extent of the change in sex varied greatly in differ¬ 
ent plants but there can be no doubt that the changes were by far more com¬ 
mon among the staminate type. 
Although flower removal from staminate hemp plants which are grown in a 
short daily exposure to light does apparently cause some of them to develop 
carpellate flowers, there seems to be always a certain number which do not 
show any sexual changes. For instance, in the above experiment the flower 
buds were removed from 32 staminate plants in one plat, with the result that 
84.3 per cent of them later produced some carpellate flowers. Only 33.3 per 
cent of those in the control plat showed “sex reversal.” In this particular case, 
then, about 16 per cent of the plants showed no tendency to develop flowers 
of the opposite sex. Since all of the plants were treated alike, it may be as¬ 
sumed that they are not all genetically identical as regards the sexual state. 
A similar experiment on the effect of flower removal on the sexual develop¬ 
ment of carpellate plants showed less change among the test plants than among 
the controls. In the plat of 32 plants grown in the greenhouse during the winter, 
the flower buds were systematically removed as soon as they appeared with the 
result that 10.53 per cent of them later produced some stamjhate flowers. In 
the control plat 14.78 per cent produced some staminate flowers. It is very 
evident that flower removal from pistillate plants in this experiment has not 
resulted in significant sexual changes. 
Pruning may indirectly have some effect on development but experimental 
results fail to show any significant modification of sex by this means. When 
the top of a staminate plant is removed at the time of anthesis, or shortly before, 
branches arise from the leaf axils and grow to a considerable length. Some¬ 
times varying numbers of hermaphrodite flowers appear on such branches but 
the results of experiment show that this is probably only a matter of coincidence. 
In a plat of staminate plants from which the tops had been removed 50 per cent 
produced some hermaphrodite flowers. In the control plat 50 per cent also 
showed “sex reversal.” In a plat of carpellate plants from which the tops had 
been removed 18.1 per cent developed hermaphrodite or staminate flowers, 
while 20 per cent in the control plat produced sonle staminate flowers. It seems 
quite clear, therefore, that such drastic pruning as cutting off the whole top of 
the plant does not significantly affect the sex of the plant. 
DISCUSSION 
After having observed thousands of hemp plants growing in the greenhouse and 
in the open, and after having tried many different kinds of stimulation in attempts 
to change the sex, the writer can not fully agree with the statement that the 
control of sex in hemp has reached a stage where it is possible to take a sample 
of seed and produce, at will, a stand of plants of any desired degree of sexual 
expression. It is true that Schaffner ( 6 ) has obtained as high as 88 per cent 
