1910] SHULL—INHERITANCE OF SEX IN LYCHNIS IIg 
I am not unmindful of their possible importance. The number 
of offspring of this combination was too small to enable one to be 
sure that these are not mutants which bear no necessary relation to 
the fact that their mother was a hermaphrodite. It was found 
particularly difficult to secure offspring of crosses of this type, as the 
flowers were very often caused to drop off as a result of the process 
of castration. Very often also in the flowers of hermaphrodite 
plants the pistils are immature at the time the anthers are ready to 
open, and after castration these pistils frequently develop no farther, 
in which case there is no possibility of effecting fertilization. The 
occurrence of these two hermaphrodite plants in no. 08116 offers a 
further suggestion that the hermaphrodite character may be capable 
of independent movement, and that consequently it may be carried in 
some manner or to some degree by the female. This suggestion 
especially commends itself from the fact that the occurrence of the 
male and female organs on the same plant, as CoRRENS (4) has 
pointed out, constitutes in effect a mosaic, and it is well known that 
mosaic inheritance is often dependent upon the presence of a definite 
separate unit for the mosaic condition. If further investigation should 
demonstrate that the hermaphrodite character may be transmitted 
through the female, as is suggested by this one family, we will be forced 
to the conclusion that here also the existence of a unit for the mosaic 
condition is present. In whatever manner the male may be converted 
into a hermaphrodite, the results seem to demonstrate that in Lychnis, 
as in Bryonia, it is the male which is heterozygous and which carries 
both male- and female-producing genes as concluded by CorRreENs, 
and not the female as assumed by BATESON. - 
The demonstration that the hermaphrodite of Lychnis dioica 
is a modified male indicates that SrRASBURGER (II) was mistaken 
in his interpretation of the effects produced by Ustilago violacea 
upon this species. He believed that the infected plants were females 
in which the development of stamens was stimulated by the attack 
of the fungus. Instead of this it is probable that they were males 
in which the disease somewhat lessened or modified the dominance 
of the male character, thus allowing the female organs to develop; 
or, if the female isa positive homozygote, the disease may be assumed 
to have stimulated the single female gene or x element of the male 
