additional buds at the base of these smaller, lateral flow- 
ers, but they seem rarely to develop. It is obvious from 
this condition that, however densely glomerate it may 
appear to be, the inflorescence of Micrandra Lopezii is 
of indeterminate growth. These lateral flanking flowers 
so tightly enclasp the larger terminal flower that they 
are flattened and distorted. They may be of unequal size 
und are enclosed in large, ovate, caducous bracts. 
The staminate flowers are variable both in size and in 
number of stamens. The usual number seems to be eight, 
but a variation from six to nine exists. he staminate 
flowers usually have a hairy, lobate disk. 
This extraordinary display of protean versatility on the 
part of the flowers of Micrandra Lopezu may be of much 
deeper significance than we can now understand. The 
genera of the Huphorbiaceae are usually well defined and 
stable. The discovery of such an apparent lack of differ- 
entiation in generic characters is indeed unexpected. 
On a basis of most of its characters, the concept here 
described as Micrandra Lopezti would have been accom- 
modated by the genus Cunuria. The species, however, 
has two characters which would have thrown it into M- 
crandra as formerly delineated. ‘The presence of a disk 
in the staminate flower and the low and inconstant num- 
ber of stamens are ‘‘micrandroid’’ characters; whereas 
Cunuria was, in great measure, based upon an absence 
of a staminate disk and a constant presence of ten sta- 
mens. Therefore, since Micrandra Lopezii is intermediate 
between Cunuria and Micrandra as formerly understood, 
it can be taken as a partial basis for the reduction to syn- 
onymy of Cunuria herein proposed. 
This most unusual tree [ am dedicating to the late 
Francisco Lopez who first discovered it and recognized 
it as a ‘‘different™” cunuri. ‘‘Pacho’’ Lopez was born in 
La Pedrera in the Colombian Amazonia in 1927 and died 
[ 208 | 
