INHERITANCE OF SEX IN MERCURIALIS ANNUA 
435 
The existence of these sporadic male and female flowers on plants predomi- 
nantly of the other sex shows, however, that the conception of sex as 
determined by hard and fast unit factors is probably radically wrong. A 
truer view of the nature of sex as a fluctuating variant is given by the 
recognition of the existence of these sex intergrades. 
My results tend to bear out the well known breeding law that "like 
tends to beget like." This is true of sex as of other character'stics. 
I have not yet pollinated female flowers on males with pollen from male 
flowers borne on females. It would be interesting to know the nature of 
the offspring from seed so secured. The difficulties of manipulation are 
very great. It is almost impossible to be certain that a female flower, 
appearing upon a male, is at any time free from some of the thousands of 
pollen grains that are continually being shed. I hope some time to be 
able to publish results on crosses between female intergrades and male in- 
tergrades, female intergrades and pure males, pure females and male in- 
tergrades, as well as on all possible crosses between hermaphrodites and 
graded males and females. 
Correns (191 3), in discussing Strasburger's results with Mercurialis 
annua, mentions work of his own with Valeriana dioica which yielded similar 
results. Females pollinated by males gave males and females, the latter 
in excess. Some of the males bore hermaphroditic flowers which were 
self-fertile and the seeds with one questionable exception produced only 
males. 
Tournois (1914) reports securing two female plants from seeds set on 
male hops {Humulus lupulus). The results, however, seem so incomplete 
and the data so few that they must be further investigated. 
The fluctuation in the proportion of males and females reported for 
dioecious plants may in part be explained on the basis that in a given popu- 
lation one or the other sex may be producing selfed seed. Heyer (/. c.) 
found a ratio of 100 males to 114.93 females in 40,000 hemp plants. This 
ratio may be explained on the assumption that among his plants there were 
female plants which produced male flowers and that selfed seeds resulted 
which in turn gave rise to female offspring. At the same time normal 
pollination by the males occurred and the progeny from that seed would 
occur in a one-to-one ratio. The addition of the selfed seed from female 
parents would raise the proportion of females in the population. When the 
proportion of males is greater than that of the females, it may be assumed 
that the male parents produced selfed seed in addition to the seed set by 
the females through normal pollination. Heyer (/. c.) in 21,000 Mercurialis 
annua plants reports a ratio of 106 males to 100 females. However, the 
above suggested assumptions do not entirely explain deviations from the 
expected one-to-one ratio as judged from an analysis of sex ratios in ex- 
perimental plants and animals. 
