434 
CECIL YAMPOLSKY 
and a number were discarded. From the remainder he raised 36 plants to 
maturity all of which were males. 
The behavior of the male plants under my observation confirms Stras- 
burger's observations. The seeds from the male plants produced males. 
From 304 seeds sown 78 plants were secured, all of which were males. It 
is to be noted that only 35 plants out of almost 800 produced seed. Table 
6 shows the variation in seed production. 
In comparing table i with table 6, one is struck with the difference in 
seed production between females and males. It must be borne in mind, 
however, that a comparison on the basis of seed production is not justi- 
fiable. It was previously pointed out that an anther may contain 1000 
pollen grains. Assuming that all the grains are functional and that only 
one anther is produced upon a female plant, it is theoretically possible that 
500 of the two-celled ovaries may be pollinated and produce seeds. A 
vigorous female may at one time produce thousands of flowers. The maxi- 
mum number of seeds secured from a female flower upon a male plant is 
three when a three-celled ovary is present. Inasmuch as the number of 
female flowers upon a male plant is small, only a few seeds are secured. 
It would perhaps be more accurate to compare the total male flower produc- 
tion on females with the total female flower production on males. Here too 
one meets with the difficulty of finding all the male flowers on a female 
plant because they are so inconspicuous, whereas all the female flowers 
soon develop seed upon a male and are readily detected. The 26 female 
plants (table i) produced at least 95 male flowers (it is not claimed that 
this number inc udes all that were produced). The 29 male p'ants (table 
6) produced 156 female flowers, and to these should perhaps be added 90 
that dropped from plant no. 247. 
Nevertheless, it should be noted that fewer male than female plants 
produced seeds. The females may also, in some cases, have set seed from 
stray pollen from other female plants in the vicinity that produced sporadic 
male flowers. 
If we leave out of account the sporadic male flowers, the selfed females of 
Mercurialis annua may be said to have recorded their own "gametic con- 
stitution" by producing only prevailingly female offspring. The work of 
Kriiger, Bitter, and Strasburger, as well as the present results, do not con- 
firm Bateson's {I.e.) assumption that the female bears two kinds of eggs, 
male-producing and female-producing. On Bateson's assumption, selfed 
females should produce both male and female progeny. 
Leaving out of account the sporadic female flowers, the selfed males of 
Mercurialis annua may be said to record their own "gametic constitution" 
by producing only prevailingly male offspring. Strasburger's and the pres- . 
ent work do not confirm Correns' assumption that the male bears two kinds 
of pollen grains, male-producing and female-producing. On Correns' as- 
sumption male plants should produce both male and female progeny. 
