HAZEN: TRIMORPHISM AND INSECT VISITORS OF PONTEDERIA 483 
In the field one is often inclined to think that one of the three 
forms of the species predominates in a particular locality. This 
might occur quite as readily as a result of vegetative propagation 
by rapid growth of the rhizomes as by greater fruitfulness of one 
form. But the pickerel-weed stem regularly becomes geniculate 
after flowering, usually just below the insertion of the spathe-like 
floral bract (the ‘knee’ is already indicated in PLATE 15) and bend- 
ing downward, the inflorescence is lowered into the water for ma- 
turing the fruits, and they are most apt to fall to the bottom in 
the near vicinity of the parent plant; in this way also extensive 
patches of one plant form may be established. МіШег reported 
in 1883 ( op. cit. 299) that all the Pontederiaceae known to him, 
including Heteranthera reniformis and H. zosteraefolia, the two 
species of Eichhornia, and the Pontederia from Curitibanos, have 
this habit of bending the flower-stalk down to the marshy ground 
or water in which they grow. In 1912 Hauman-Merck* reported, 
as a peculiarity which he thought had escaped previous observers, 
this habit of maturing the fruit under water in Pontederia rotundi- 
folia, and stated that P. cordata growing in abundance in the same 
places in pools of the banks of the Rio de la Plata matures its 
fruits out of the water. This statement is rather surprising, since 
our plants are so fixed in this habit of bending down after flowering 
that plants kept in the greenhouse with little water develop in a 
manner precisely similar to those left in the field. Pontederia 
montevideensis shows practically no such tendency when grown here 
and it would appear possible that Hauman-Merck was really 
dealing with this species rather than with P. cordata. 
The final judgment reached at the Arcola station was that al] 
three forms of the plant were about equally numerous there. 
During the dry August (in the region of New York) of 1917, one 
visit to the station revealed such a desolate and discouraging 
group of plants that no attempt was made to do anything further 
during that season. Perhaps clumps of the plant growing in the 
borders of the Hackensack River, if they could have been reached, 
might have proved more rewarding in such a season, for, contrary 
to the usual statement that this family comprises only fresh water 
* Hauman-Merck, Lucie Sur cas de géotropisme hydrocarpique chez 
Pontederia rotundifolia L. kai Inst. ва. Léo Errera 9: 28—32. 1912 
