COOK—HABITS OF CACAO AND PATASHTE, 621 
One can only regret being unable to share the confidence of Dr. van 
Hall in the statement just quoted. Indeed, it is hard to believe that 
this account can relate to cacao. It suggests, rather, the question 
whether there is another tree in Java that could have been mistaken 
for cacao. As may be seen by reference to the photographs repro- 
duced in plates 51 and 53, the flower stalks of cacao are neither long 
nor ‘‘supple,’”’ nor do the flowers hang in a pendent position that 
would enable them to swing in the wind. The most drooping or 
nearly pendent flowers often have their petals in contact with the 
surface of the bark. No doubt some of the flowers might be made 
to flutter by a strong breeze, but there could be no ‘‘swinging to and 
fro” unless the pedicels were wilted. The stalks are strong enough 
to hold the flowers rather firmly in place. It may not be impossible 
that grains of pollen would reach the stigma if the flowers were sufli- 
ciently beaten about, but the preference of the tree for moist, sheltered 
situations and its peculiar habit of producing its flowers in the most 
sheltered positions are opposed to the idea that it is normally depend- 
ent on wind pollination. Even stronger reasons against this view are 
afforded by the structure of the flower itself, with its reduced number 
of anthers and the very small quantity of pollen. In several other 
respects the flowers are as highly specialized as those of orchids, 
asclepiads, or other groups that are known to be dependent on insects. 
It would be difficult to find adaptive functions for these specializa- 
tions under a theory of wind pollination. Indeed, this theory may 
be said to dismiss the problem of the cacao flower, instead of giving 
a solution. 
The theory that seems most worthy of consideration, on the basis 
of present information, is that proposed by Uzel, that the pollen may 
be carried by thrips, a suggestion dismissed by van Hall because 
thrips is injurious to cacao. Yet this hardly excludes the possi- 
bility of a beneficial function being performed by the same insects, or 
by other members of the same order. Some of the Thysanoptera 
are very active, free-flying insects, a fact that van Hall seems to 
overlook. The small size of the chambers that shelter the anthers 
may be taken to indicate that they are to be entered by small insects. 
The fact that the stamens are held in a notch of the petals naturally 
gives an impression that they are to be released or ‘‘tripped”’ by a 
large insect such as a butterfly, bee, wasp, or fly, but this may prove 
not to be the case. A small insect could enter the petal hood from 
the side and bring out pollen without disturbing the stamens. The 
staminodes also would appear more likely to have a function in rela- 
tion to small insects, since these organs form a complete ring around 
the pistil and are hirsute on their lateral faces as if to keep small 
insects from passing between. 
