Y 
aY. 
POLLINATION Li 
The mature embryo-sac is of the typical monocotyledonous 
form. It is covered with a cap of nucellar tissue rich in starch, 
and rests upon a considerable basal mass of the same kind of 
tissue. Four to eight days may pass between the maturing of 
the embryo-sac and the withering of the stigma and consequent 
impossibility of fertilization. During this time the embryo-sac 
may almost double in size. The greatest change in surrounding 
tissue to be noted during this time is in the lateral portions of the 
nucellus. In some cases sacs just mature with fertilized eggs have 
been found completely invested with a layer of nucellus. In 
others the last vestiges of lateral nucellus have disappeared before 
fertilization. 
POLLINATION 
The pollination of A. triphyllum presents a problem which has 
either escaped the notice of investigators or has baffled attempts 
at solution. In the case of bisexual spikes where securing pollina- 
tion would seem to be a simple matter, the staminate flowers mature 
so long before the pistillate that their pollen is probably inactive 
when the ovaries are mature. The dioecious character of most 
of the flowers makes cross pollination necessary. The pollen is 
slightly adhesive, and being borne deep in the hooded spathe, has 
practically no chance to be carried by the wind. There are no 
nectaries or similar structures connected with either staminate 
or pistillate clusters. In a few spathes insects have been observed 
eating the stigmatic hairs of pistillate flowers; but this has been 
- observed so few times that the idea of insects coming to the spathes 
to feed on the stigmatic hairs cannot be entertained. As men- 
tioned in the description of the pistillate flower some of the hairs 
formed inside the ovary produce a gum which is later reduced to a 
slimy mass filling the ovary cavity. Whether or not this produces 
an odor attractive to insects can only be conjectured. 
A brief reference to the structure of the two forms of inflor- 
escence will make clearer the observations on insect visitation. 
The space between the pistillate spike and its spathe (TEXT 
FIG. 11) is much less than in the case of the staminate spike 
(TEXT FIG. 12). Such a difference is quite general, although it is 
not always as great as shown in the figures. In fact, while 
the staminate spathe may be entered and left again by 
