CONTRIBUTIONS TO BOTANY. 
81 
lid<E : in such case, the latter name, on account of its priority, 
ought to claim the preference. As however it is contrary to the 
rules of science to form a compound generic term from both 
Greek and Latin roots, the name would necessarily require to be 
modified into Lasiandra, one that has long been preoccupied. 
Besides this, we have to consider the confusion likely to arise 
from increasing a list of consimilar names, already too numerous, 
as Lasiandra, LasiantJuea, Lasiantha,Lasianthus, mdi Lasianthera, 
and also, that in reality the latter name is untenable, because of 
the incorrectness of its signification, for in the present case it 
will be seen, that it is not the anther, but the filament which is 
villous. For all these reasons, 1 strongly recommend the pre- 
ference to be given to Stemonurus, the next in priority, as the 
most appropriate designation of this genus. 
Its most marked peculiarity consists in the character of its 
stamens ; the filaments sometimes shorter, often longer than the 
petals, are generally very broad, extremely thick and fleshy, 
obtuse at their summit with a small apical point, to which the 
anthers are attached, and they have a somewhat prominent in- 
ternal keel down the middle : the margins of their broad summit 
and the upper part of the keel are fringed with long transparent 
white hairs, clavate at their extremity and bent, so as to form a 
crest over the anthers : from this character both the names of 
Beauvois and Blume originated. The flowers, sometimes herma- 
phrodite, are frequently polygamous in the same plant, that is 
to say, either the anthers are void of pollen, or the ovarium is 
deficient of any ovules, or both these imperfections occur at the 
same time : it does not appear to me that they are constantly 
unisexual, as generally stated. The analysis of the structure of 
this genus has been attended ivith much difficulty, because of 
the frequent abortion of some of its parts, especially the ovarium, 
which is often deficient of cells or ovules ; and even when the 
ovules exist, it is not easy to detect their presence, on account 
of their extreme minuteness, in an early stage of the flower. I 
was for a long while unable to solve the anomalies of its struc- 
tm'e, and almost gave up the matter in despair, but patient exa- 
mination at length overcame the difficulties ; not one in twenty 
instances exhibits the smallest trace of an existing ovule, nothing 
but a fleshy mass appearing to constitute the ovarium, which is 
always comparatively small : indications of the existence of more 
cells than one are sometimes observable, but these are not large 
enough to be well defined ; and even in the case where a single 
distinct cell exists with two suspended ovules, these are so mi- 
nute that they might readily be overlooked. After the period of 
fecundation, however, the petals and stamens fall away, when the 
ovarium attains a rapid growth, and soon displays itself as an 
VOL. I. M 
