Lopez stands entirely alone as the most primitive mem- 
ber of the genus. Examination of ample material has 
disclosed the most extraordinary lack of differentiation 
which, combined with other factors of morphology and 
phytogeography, I can interpret to mean only that we 
are confronted with the most primitive species yet known 
in the group. 
Usually, in Micrandra, the large terminal flower is 
pistillate, whereas the smaller laterally placed flowers are 
staminate. Furthermore, any condition suggestive of a 
perfect flower in this group of plants has been completely 
unknown. 
The curious structure of the densely glomerate inflo- 
rescence in Micrandra Lopezt is matched in significance 
only by the unexpected lack of differentiation of the 
flowers. he large terminal flower is usually pistillate, 
but may be staminate; rather frequently, it is found to 
be perfect, with anthers full of pollen grains, which would 
suggest that the flower is functionally as well as mor- 
phologically perfect. From six to nine stamens may be 
present. There are, likewise, transitions from perfect to 
purely pistillate terminal flowers. The anomalous forma- 
tion of the stamen-filaments is probably significant, em- 
phasizing in a striking manner the lack of differentiation 
of floral parts. Some of the filaments seem to arise from 
the disk, while a few appear to grow from within the 
disk and a few from without this structure. In some of 
the transitions from perfect to pistillate flowers, these 
‘‘filaments’’ lack anthers and, were their position on the 
disk constant, could be taken for long, subulate lobes 
of this structure. 
The large terminal flower is flanked by several (usu- 
ally two or three) smaller flowers. As in the terminal 
flower, these laterals may be pistillate or even perfect, 
although they usually are staminate. There may even be 
[ 207 ] 
