IN CEYLON. 
flowers of the same species, and it is only by actual obser¬ 
vation of fresh material in the forest that the true sex has 
been determined. It has been the common experience of 
myself and many others to regard trees of D. sylvatica and 
D. Gardneri as males, until they have been felled, when in 
addition to flowers possessing fertile stamens, abundance of 
ripe fruits have been obtained on the same branch. In order 
to further emphasize this condition I may say that the 
whole of the anatomy and development of seedlings of D. 
sylvatica, to be published elsewhere, was first worked out 
from seeds obtained from a tree which has for many years 
been labelled as a “ male,” and from which material for 
sketching the “ male ” inflorescence has been derived. 
From the following notes it will be seen that examination 
of fresh material in the forest has shown that there is a 
departure from the dioecious condition in ten out of our 
twenty species. Speaking generally, the sex appears to be 
very unstable, the dioecious, monoecious, polygamous, and 
hermaphrodite conditions having been found. The occur¬ 
rence of hermaphrodite flowers resulting in a polygamous 
condition is so frequent that there is every reason to doubt 
their non-existence in any of the Ceylon species, providing 
sufficient material can be examined outside the herbarium. 
It may be further stated that specimens of D. Thwaitesii 
have been repeatedly examined, which suggested that the 
sex of the flower might vary from time to time ; this has 
been described for other plants.* Certain it is that the sex 
of the flower often exhibits every variation in the same 
inflorescence of D. sylvatica, and the many stages observed 
readily allow one to form a series having staminate flowers 
at one end and hermaphrodite at the other. 
The types of flowers characteristic for the different Ceylon 
species may now be described under the following headings : 
( a ) dioecious only, ( b ) monoecious only, (e) dioecious and 
polygamous, and (d) dioecious, monoecious, and polygamous. 
* Willis, Gynodicecism, &c. (3rd paper); Proc. Camb. Phil. Soe., 1893. 
