334 KIRKWoopD: POLLEN-TUBE IN CUCURBITACEAE 
course of the tube lies through the tissues, it appears to travel 
through the intercellular spaces. The destruction of cells in its 
path seems to be due mostly to mechanical action and not to 
digestive enzymes, except in the nucellus of Cyclanthera. Fresh 
ovaries when teased apart showed the pollen-tubes in the stylar 
canal and passing over various parts of the placental surface. 
FIGURES i and 3-5 represent the course of the tube. 
When the surfaces of the placentae are pressed together or 
against the ovary-wall, the space through which the tubes move is 
considerably restricted. Under such circumstances they crowd 
aside the cells of the conducting tissue (FIG. 11). Sometimes a 
tube may be seen to have passed under the conducting layer (FIG. 
15). Where the tubes traverse the ovarian space in Melothria 
they sometimes develop transverse walls or plugs. The manner 
of their development appears to be the formation of a ring trans- 
versely which gradually narrows the aperture until it is completely 
closed. An excessive development in thickness of this wall some- 
times follows, forming plugs as shown in FIGURE 2 @, 4, ¢. Similar 
plugs have already been observed by Osterwalder® in Aconitum 
Napellus, and by others. 
The pollen-tube proceeds by an approximately direct course to 
the micropyle, and passes down through the nucellus to the 
embryo-sac. In the apex of the nucellus it usually expands to 
several times its normal diameter in Cyclanthera (rics. 6, 13), and 
from this point sends down a straight and narrow tube to the 
embryo-sac. Sometimes the entire neck of the nucellus is destroyed 
by the dilation of the tube. This expansion of the pollen-tube 1m 
the apex of the nucellus occurs frequently in MJelothria, though it 
was never observed to reach the same proportions as in C; ‘yclanthera. 
In Micrampelis also such conditions were sometimes observed ; here, 
however, they are not the rule but the exception. In such cases, 
both in Melothria and Cyclanthera, the contiguous cells were 
destroyed. 
That such dilations of the pollen-tube are directly correlated 
with the occurrence of starch in the surrounding tissues is matn- 
tained by Longo, as above pointed out; but the behavior of the 
pollen-tube of Elodea, as describe 
a cribed by Wylie,® indicates that such 
igeranctesdicis a) always the cause of the enlargments of the tube, 
