336 KirKWoop: POLLEN-TUBE IN CUCURBITACEAE 
grew toward the pieces from the vicinity of the stigma and were 
most strongly attracted by ovules ready for fertilization, growing 
into the micropyle in each case. He concludes that this stimulant 
is in the nature of a fluid, and in another article has shown that 
pollen-tubes grow into the stomata of Zradescantia leaves, the in- 
tercellular spaces of which have been filled with sugar solution. 
That the pollen-tube is also directed mechanically in its course is 
maintained by Miyoshi, and Capus states that the mechanical role of 
the conducting-tissue consists in placing the pollen-tube in relation 
with the micropyle, and that its economy and simplicity depend on 
the structure of the ovary. 
That the stimulants which direct the pollen-tube are nutritive in 
character is quite evident. Dalmer® states that pollen-tubes are 
nourished by amyloid substances secreted by stigma, stylar canal, 
and the interior of the ovary, though he maintains that the entrance 
of the pollen-tube into the ovule is a purely mechanical phenome- 
non. Miyoshi points out that sugar as a stimulant can operate 
only extra-cellularly, and it is plain that where the course of the 
pollen-tube is intercellular or through the locular space of the ovary 
that its growth must be at the expense either of the reserve ma- 
terials within the tube or of substances elaborated by the tissues in 
its path. The cytological features of the conducting-tissues of the 
Cucurbitaceae here under consideration are similar to those of cer 
tain nectaries in the abundance of starch and quantity of cytoplasm, 
appearance of molecular structures, etc. The evidence here points 
to the secretion of carbohydrates (probably sugar) as the nutritive 
material. That the pollen-tubes are partial to the more glandular 
cells is indicated in Cyclanthera. In this case the ovary sometimes 
shows a considerable locular space at the time of the entrance of 
the pollen-tubes. The epidermal cells of the placenta are larger 
and more glandular in appearance than those of the ovary-wall 
opposite. Though their distance apart is not great, the pollen- 
tubes always cling to the placenta instead of to the opposite wall, 
though that likewise is supplied with starch. 
One noticeable feature of the relation between pollen-tube and 
conducting-tissue in Melothria, Micrampelis and Cyclanthera is the 
GPS, ORES en Dene go eaEe My j Fh 
fact hae a 
