CONTRIBUTIONS TO BOTANY. 
75 
only after the ovarium has attained a considerable increment 
and advance towards maturity. I have met with fertile female 
flowers only in the New Holland species^ and there three very 
distinct and equal styles exist ; the ovarium is deeply grooved 
into three corresponding lobes, each lobe being distinctly 1 -celled, 
as I have found to occur in Stemonurus. Only one of the ovules 
in one of these cells becomes perfected, so that the fruit is only 
1 -celled and 1 -seeded, but partial deviations from this rule some- 
times occur. The ripe fruit of this same species is an oval 
drupe, surmounted by a sessile 3-lobed diseoid plate, similar to 
that observed in the two other species of Pennantia, and in every 
species of Stemonurus, in Sarcostigma and in Discophora. As I 
have traced in Pennantia Cunninghami the existence of several 
distinct styles or stigmata, and the subsequent conversion of these 
organs into sessile discoid processes, we may safely infer that it 
is common to all the species of the several genera above men- 
tioned ; that after impregnation the styles become flattened and 
expanded, until they form a sessile fleshy disc, more or less 
iobed, upon the summit of the fruit : I feel convineed that Bauer, 
when he made his drawing of P. Endlicheri, sketched the figure 
of the ovarium after this transformation had taken place. The 
fact of the existence of three distinct styles and three separate 
cells in the ovarium, does not militate against the ordinary cha- 
racter of the Icacinacece, as demonstrated by the structure already 
shoum to exist in the genera previously described ; for though the 
ovarium is there uniformly 1 -celled, I have all along endeavoured 
to prove that such is the case only by abortion, and that 
normally, in all those instances, it is in reality 3- or 5-celled. My 
object in constantly maintaining that fact, has been to show the 
fundamental difference in normal structure that exists between 
the Icacinacere and the Olacaceee, In regard to the circumstance 
of the presence of three distinct styles in Pennantia, we must 
remember the very analogous structure in several genera of the 
section last described ; for instance in Apodytes, Rhaphiostylis 
and Leretia -, but in those cases, two of these styles are nearly 
obsolete, as are also two of the corresponding cells of the ovarium : 
in all these instances, the normal axis is in the centre of these 
cells, although two of them be only rudimentary ; for where the 
ovarium is apparently only 1 -celled, the ovules are invariably 
suspended near the apex of the cavity, upon the side of the 
abortive cells and of the two rudimentary styles. It is desirable 
to keep these facts in view when we come to speak of Stemonurus 
and Sarcostigma. 
Among the numerous specimens in Sir Wm. Hooker’s herba- 
rium of Pennantia corymbosa from New Zealand, I have not 
been able to meet with a single ovarium that shows any trace of 
L 2 
