82 
WRIGHT : THE GENUS DIOSPYROS 
is very great, and though it is easy to connect the flowers of 
the third class in the polygamous group with those of the 
true male, it is not so easy to connect the hermaphrodite 
flowers of the polygamous group with the ordinary female 
flowers. 
We have again to realize that in the same species there is 
a wide variation in the number and orientation of the floral 
parts of male and female flowers. 
Dioecious , Monoecious , and Polygamous .-—This may occur 
in flowers of D. hirsuta, D. Thwaitesii, and D. Ebenum. 
In these species a very complicated sex relationship 
exists. If we take as our example D. Thwaitesii, we find 
that three types of trees exist. 
In the first place there are pure female trees. Here the 
flowers may occur solitary or in small sessile axillary 
clusters of two to seven flowers, each component having 
five staminodes disposed so as to alternate with the five 
lobes of the corolla and a four-celled ovary. 
In the second class there are pure male trees having the 
flowers in axillary groups of four to fifteen. Each flower 
has ten stamens arranged in five pairs alternating with the 
corolla lobes, or several pairs and the remainder single, 
arranged irregularly or as five series. In these flowers the 
pistil is always rudimentary, and consists of nothing more 
than a central mass of brown hairs. It is in the third class 
of trees that complicated relations exist. In this class we 
have the true male and true female flowers occurring on 
the same tree, sometimes in separate axillary groups of the 
same branch, and at other times in the same axillary group. 
Thwaites noted that in D. hirsuta the female flowers were 
sometimes mixed with the male. We may therefore speak 
of the monoecious condition in these species. But in 
addition to the monoecious condition we may have on the 
same tree a truly polygamous group or groups of flowers 
exactly analogous to the polygamous flowers described for 
D. Gardneri, D. sylvatica, and others. The polygamous 
