1912] STEVENS—HETEROSTYLOUS PLANTS 287 
for experimental purposes in the greenhouse, which showed such 
unusual floral variations as to be worthy of record. The plant pro- 
duced 10 blossoms, and 8 of these had stamens and pistil both long; 
while 2, the third and eighth, were normal short-styled flowers. 
The earlier blossoms were removed in order to secure continued 
flowering. Later, however, two of the abnormal flowers were pol- 
linated, one with pollen from a short-styled and the other with 
pollen from a long-styled flower, but neither developed seed. One 
of the normal short-styled flowers on this plant, however, produced 
a good seed when self-pollinated. 
It will be noted that in both the cases cited above, in which the 
stamens and pistil are of the same length, it is the pistil which has 
varied from its normal length. That is, if a flower on a short-styled 
plant shows pistils and stamens of equal length, both are long; 
while in such a flower on a long-styled plant, both pistil and stamens 
are short. . A curious case which shows a similar variation in the 
length of the pistil has been noted by BATESON and GREGORY 
(2, p. 583) in Primula sinensis. There is a variety with a very 
large yellow ‘“‘eye”’ extending up over the limb of the corolla, quite 
distinct from the small yellow pentagon characteristic of the normal 
flower. This variety has the anthers in the position normal for 
long-styled plants, but the style is short and the stigma just reaches 
the anther level, a condition which BATESON and GREGORY des- 
ignate as “‘equal-styled.” In investigating the inheritance of these 
two unusual characters, they find that the “equal-style” is the 
form which the long-styled type assumes when the plant is homo- 
zygous in the large eye character. In this case, as in the others 
mentioned, the pistil alone varies from the normal length. 
DaRWIN experimented in a rather imperfect manner on the 
relative fertility of the:two forms in the buckwheat, and showed 
that illegitimate fertilization is less successful than legitimate, 
yielding fewer and smaller seeds. In order to gain some knowledge 
of the relative ease of legitimate and illegitimate fertilization, the 
following experiment was made. Plants were grown from seed in 
a greenhouse, where there were no insects which might bring about 
pollination. The flowers which were to be experimented upon 
were examined with a lens in order to make sure that the pistils 
