ENDOSPERM OF UTRICULARIA 5 
the fertile stalk is gradually submerged, either sinking slowly 
or tipping over bodily into the water. The seeds thus carry out 
most of their development in air, invested by the tough capsule 
_ of the ovary wall, but are soon submerged through the sinking 
of the entire inflorescence. 
FLOWERS AND POLLINATION 
The flowers are bright yellow in color and offer many special- 
izations for cross pollination. They are distinctly zygomorphic, 
with a strongly lipped corolla bearing the stamens, a tubular 
style, and a lipped stigma with numerous stigmatic hairs be- 
neath the upper lip. Insect visitors seem to be responsible for 
pollination, and the writers could find no evidence of cleisto- 
gamy in this species. Flowers opening in hooded bags which 
prevented the visits of insects, were uniformly sterile. 
While insects are numerous about the flowers of Utricularia 
most of them merely perch upon some part of the inflorescence. 
The damsel flies, for example, regularly use the flower stalks 
as places of temporary rest from flight. The only insect ob- 
served to enter the corolla was a small buff-colored fly, the 
species of which was not determined. These insects are numer- 
ous about the beds and were repeatedly observed to enter the 
closed throat of the corolla, disappearing within the flower and, 
after an interval of nearly a minute, to emerge and wing their 
way to other blossoms of the kind. Fertilization seems to be 
regularly accomplished and of the hundreds of ovaries sectioned 
few showed sterile ovules. Abortive embryos are not uncommon 
but this condition seems to be due mainly to crowding and 
possibly to loss of nutritive connection through partial dis- 
placement, rather than to failure of pollination. 
Much of the material for this study was killed in one per 
cent chrom-acetic acid, run through the alcohols and imbedded 
in paraffin. The sides of the ovaries were clipped so as to in- 
vite the free penetration of reagents into the cavity of the 
ovary. For critical study of fertilization and early post-fertili- 
zation stages, material was killed in Flemming’s stronger solu- 
tion. This was incomparably better, and practically all the 
figures were drawn from material fixed in this way. 
