INHERITANCE OF SEX IN MERCURIALIS ANNUA 
so-called perfect — hermaphroditic^ — flowers are produced, and, in contrast 
with the animal kingdom, these include the great bulk of the seed plants. 
I shall discuss this aspect of the question later. 
Cytological evidence for sex chromosomes in phanerogamic plants is so 
far negative. Sykes (1909) failed to find any difference in the nuclei of 
male and female plants of Hydrocharis, Morus, Bryonia dioica, Lychnis 
dioica, Mercurialis perennis, Sagittaria, and Cucurhita Pepo. Strasburger 
(1910) and Malte (1910) found no evidence of chromosomal differences in 
male and female cells of Mercurialis annua. 
The evidence that a phanerogamic plant may produce two kinds of 
eggs or two kinds of sperms is only of an indirect character. It has been 
based mainly on the work of Correns (1907) and Bateson (1909) , who worked 
with crosses between Bryonia dioica and B. alba, to which I have referred 
in an earlier paper (191 6). Correns assumes that there are two kinds of 
pollen grains; Bateson assumes that there are two kinds of eggs. According 
to both assumptions one of the sexes is heteroyzgous for the sex factor. 
Shull (1910), working on crosses of Lychnis dioica with "hemaphrodite 
mutants," claims to verify Correns' assumption that the male is the hetero- 
zygous sex. 
Sex ratios in dioecious forms have been reported in the main as conform- 
ing to the Mendelian hypothesis of sex. In mass populations the expected 
one-to-one ratio is secured with slight fluctuations in one or the other direc- 
tion. Heyer (1884) found a ratio of 100 males to 114.93 females in 40,000 
hemp plants and 106 males to 100 females in 21,000 Mercurialis annua 
plants. Fisch (1887) found a ratio of 100 males to 154.23 females in 66,327 
hemp plants. Strasburger (1900) found a ratio of 100 males to 128.16 
females in 10,662 plants of Melandrium album. But in the offspring of 
certain matings in both the animal and plant kingdoms there are striking ex- 
ceptions to the one-to-one ratio. Certain matings will give exclusively 
female or almost all female offspring, whereas other matings will give the 
opposite result. Doncaster (19 16) reports for the gall-fly, Neuroterus 
lenticularis , that galls from six females produced 4,235 males and 83 females, 
and six other galls produced 5,139 females and 117 males. Montgomery 
(1908) secured an overwhelming number of males as contrasted with females 
in a culture of the Aranead Lutroductus nactans, namely, 37,210 males to 
4,539 females. Doncaster (1913) found in his cultures of the moth Abraxas 
^rossulariata lines of females that tended to produce only female offspring. 
He also had bisexual lines in which the proportion of males to females was 
about equal. However, in mating some of the unisexual lines he secured 
offspring that were all female. 
Riddle (19 17), on the basis of a number of sex ratios reported by various 
authors showing that the degree of closeness of relationship between the 
parents affects the sex ratio, concludes that family crosses yield only male 
offspring, 20 males to o females; generic crosses a ratio of 4.9 males to i 
