2 BOTANICAL GAZETTE [JANUARY 
stamens are commonly found in the pistillate flower, and six have 
been reported. to occur in rarer instances. These staminodia 
are composed of uniform parenchyma and show no evidence of 
the suppressed sporangia. Much more unusual is the develop- 
ment of stigmas upon the staminate flower. Eichler (5) refers to 
these as being possibly rudimentary stamens; the writer, how- 
ever, has found well-developed stigmas above the nine stamens 
usually borne by the staminate flower. In one instance the 
stigma was clothed with papillae and covered with pollen grains 
which had sent their tubes into its tissue. Insuch acase it would 
appear that the usual stigmatic secretions are present and that it 
differs in no essential respect from a normal stigma. Such con- 
ditions recall Chamberlain's (2) discussion of the teratology of 
Salix,in which monosporangiate, dioecious genus he reports almost 
all possible combinations in the production of the sporangia. 
The occurrence of these rudiments in Elodea must point back 
to an ancestry with perfect flowers, and their specialization has 
no doubt been correlated with the changed conditions of the sub- 
mersed habitat. The flowers are now dioecious and are pecul- 
iarly adapted to the combined influences of wind and water for 
pollination. The pistillate flower is quite complex and presents 
one of the most striking cases of epigyny known; the staminate 
flower is simpler and has acquired the habit of breaking loose 
from the stem at maturity. 
The development of the pistillate flower was studied by 
Horn (3), who also investigated the vegetative plant body; but 
since his work was based on the external aspects or rough sec- 
tions only, the account, while accurate ina very general way, lacks 
most of the details secured through modern methods. The pistil- 
late flower begins as a protuberance from the side of the stem 
near the growing point (fig. r). This swelling pushes out rapidly, 
soon equaling the stem tip in length and giving it the appear- 
ance of having bifurcated (jig. 2) , but the main axis soon reasserts 
its dominance, leaving the flower as a distinctly lateral member 
(fig. 3). At this time ridges near the base of the flower mark 
the origin of the spathe (fig. 2), which pushes out rapidly and 
for a long time envelops the developing flower (jigs. 2-12). 
