438 
CECIL YAMPOLSKY 
females. Plant no. 124&, female with small anther rudiments (1906 showed 
hermaphroditic tendency, 1907 and 1908 female), when pollinated by her- 
maphroditic plants nos. 149 and 118 respectively, gave 73.8 percent and 
81 percent females, the remaining offspring being distributed among forms 
between hermaphrodite and female. Plant no. 128, which showed both 
female and hermaphroditic tendencies in 1906 and 1907, and which appeared 
to be female in 1908, when crossed with the hermaphroditic plants nos. 
149 and 118 gave 18.5 percent and 58.8 percent females respectively. 
From the above noted results, Correns concludes that the egg is re- 
sponsible fbr the nature of the sex of the offspring. Plants nos. 122, 1245, 
and 128 when pollinated by plant no. 118 gave varying percentages of 
female offspring. The pollen also has an influence on the nature of the 
offspring. When no. 118 was used as pollen parent more pure females 
were produced than with no. 149 as pollen parent. The more pronounced 
the production of sex cells with a female tendency is in a plant, the less 
influence the pollen parent has on the offspring. Plants nos. 122 and 124& 
which are strong females gave with plants nos. 149 and 118, 89 percent and 
97 percent females. Plant no. 128, which was only partly a true female, 
gave only 58.8 percent female offspring when pollinated with plant no. 
118. It may be conceived that if the female plants were capable of pro- 
ducing functional pollen capable of self-pollination, the offspring therefrom 
would be prevailingly female and would thus behave like the selfed females of 
Mercurialis annua. 
Shull (1910) describes the appearance of hermaphroditic mutants among 
his pure bred families of Lychnis dioica L. The hermaphrodites were of 
two kinds. When his so-designated "^4" and "B" hermaphrodites were 
used as pollen parents, there resulted 398 females, 305 hermaphrodites, and 
2 males. When his so-designated "C" and "D" hermaphrodites were 
used as pollen parents, there resulted 65 females and 73 males. Hermaphro- 
dite A as female parent was pollinated by a normal male and there resulted 
21 hermaphrodites, 2 females, and ii males. Shull assumes that plant A 
is' a modified male and thus explains his results. When A was self-polli- 
nated it produced 23 females and 25 hermaphrodites. B upon self-polli- 
nation gave no females and 95 hermaphrodites. The author harmonizes 
his results with those of Correns on the crosses between Bryonia dioica and 
B. alba. The male is the heterozygous sex. The hermaphrodite must have 
eggs of one kind and pollen grains of two kinds. 
The existence of hermaphrodites in Lychnis dioica, although described 
as a pathological phenomenon, the result of the invasion of the smut Ustilago 
violacea in the female of the species, has been first described by Shull as a 
normal condition. An analysis of the behavior of hermaphroditic plants 
A and B makes it appear that these two plants were graded females. When 
they were used as pollen parents and when they were selfed they produced 
hermaphroditic and female offspring, which would suggest such a condition. 
