272 
AMERICAN JOURNAL OF BOTANY 
[Vol. io. 
dioica as an argument against a Mendelian explanation of sex inheritance, 
stating that “the Mendelian hypothesis of sex in itself can not account for 
the preponderance of one sex to the exclusion of the other.” The writer 
(7, p. 663), however, has shown that wherever in Shull’s work (3, 4) irregular 
sex ratios were obtained, they were the result of partial or complete elimina¬ 
tion of the male (and to a slight extent of the female) gametes which carried 
the mutant factor “narrow leaf” linked with femaleness (or with hermaph- 
roditeness in the case of Melandrium album). This explanation was borne 
out by the presence of large amounts of abortive pollen in the narrow¬ 
leaved Lychnis male. Dorsey (1), working with plums, has expanded the 
idea of gametic elimination due to unfavorable genetic combinations to 
include the elimination of embryos during various stages of their develop¬ 
ment because of unfavorable genetic combinations and has given consider¬ 
able evidence in support of this explanation as the cause of dropping of 
plums through the summer and of poor germination of apparently normal 
seeds from certain crosses. 
It appears from Yampolsky’s results with seed of selfed males that 
immaturity and lack of germination in the seeds from selfed males may be 
due to unfavorable genetic combinations which allowed development of 
certain embryos to proceed only to a certain point, when growth ceased. 
In his table 6 (8, p. 422) are reported a total of 156 flowers which produced 
283 seeds on male plants. Besides these there were produced on the same 
plants “ approximately 90 other female flowers . . . which failed to develop 
seeds and dropped off.” A total of 492 seeds might then have been expected 
if all had set and developed normally; 283 were actually produced, and of 
these, 31 were immature, 5 germinated weakly, 139 did not germinate, and 
26 were not planted. Certainly there is evidence of elimination in these 
results, and they can not therefore be taken as indicating that the genetic 
condition of the males of Mercurialis is pure male and not heterozygous 
for male and female determiners, as would be indicated by the results of 
crossing females with males. All possible genetic combinations should be 
thoroughly tested before conclusions as to the genetic conditions with 
respect to sex may be safely arrived at. The species in question makes 
such a study extremely difficult. 
In Fragaria, as in many other forms, the sex of all the flowers of any 
given plant is not necessarily the same. It is quite common in the perfect- 
flowered varieties for practically all the earliest flowers to be female. In 
some clones this condition has been exaggerated to a point at which only a 
few flowers produce stamens while the remainder are female. Male flowers 
on hermaphroditic varieties are common, and are practically always limited 
to the last flowers of a cluster to bloom. Female flowers may be found in 
the primary position of the inflorescence of wild males, and in such cases 
generally set fruit. Completely sterile flowers are not uncommon among 
the later flowers of a female inflorescence. 
