on the floor of the forest in great abundance. An inves- 
tigation showed that the tree was of medium size, with 
a perfectly columnar, unbuttressed trunk which was cov- 
ered with a smooth bark of a yellowish hue—all charac- 
ters unusual indeed in Cunuria and Micrandra as the 
genera have hitherto been understood. Flowers and 
ripened fruit were found on the same tree. A hasty ex- 
amination in the field indicated that, although there were 
definite characters of the leaves and fruit which would 
throw it into Micrandra, the external structure of the 
flowers and the overall appearance of the inflorescence 
strongly suggested Cunuria. Until arecent opportunity 
of examining the flowers microscopically, | was puzzled 
as to which genus it rightly belonged. Indeed, even 
during my field work, I felt that the intermediate posi- 
tion of this concept might perhaps necessitate our reduc- 
tion of Cunuria to synonymy under Micrandra. 
Detailed microscopic studies of Micrandra Rossiana 
have convinced me that the characters used to separate 
Cunuria from Micrandra do not hold. It was this dis- 
covery, supported by additional evidence from other 
species, which led to the reduction proposed in this paper. 
The leaf of Micrandra Rossiana suggests that of Mi- 
crandra siphonioides in size, texture and shape. The pres- 
ence on the under surface of the leaf of dense tufts of 
hair in the axils of the secondary veins with the central 
nerve is a character possessed by several species of Mi- 
crandra, but unknown in Cunuria. The capsule, especially 
its very thin epicarp, would fall into what has been con- 
sidered Micrandra and not into Cunuria. The caruncu- 
late seed suggests Micrandra. The gross structure of 
the flowers is very definitely ‘‘cunurioid,’’ with a deep 
cup-shaped calyx split into lobes for only about one third 
of its length, as opposed to the ‘‘micrandroid’’ calyx 
which is split nearly to the base with the lobes expanded 
[ 216 | 
