perianthium supplies: this is so abundant, that nearly a 
tea-spoonful may be obtained from every flower. 
Our drawing was made last October, from a plant in 
Mr. Colvill s Nursery, which had been brought from the 
interior of the Cape of Good Hope by Mr. Synnet. 
If the pollen of this plant is examined under a high 
magnifying power, the grains will be found to be united 
occasionally by little filaments like those which are so 
conspicuous in Onagrariae,—a remarkable circumstance, 
which we are not aware has been noticed in any other 
neighbouring plant. When the granules are dry, they have 
an irregular, oblong, depressed figure, strongly marked 
with a little pit, or hole, on one side; but as soon as 
moisture is applied to them, they dilate, and become of a 
regular oblong figure, with a perfectly smooth surface, 
which, after a time, if the granule does not burst, assumes 
a shrivelled appearance. 
In M. longifolia Mr. Ker noticed the presence of three 
pores between the angles of the ovarium at the base of the 
stigma, whence he found the honey to be secreted. In this 
species, in which the honey is exceedingly abundant, we 
did not notice these pores; but we are disposed, neverthe¬ 
less, to suspect that they were only overlooked, and that 
they do m fact exist. At least there is a curious conforma- 
tl ^ n l ln the ovarium, which seems to indicate the probability 
? r • ^ ex * s tence, and which in the foregoing description 
is briefly alluded to, by the terms inter loculos lacunosum. 
11 the ovarium is cut across, it will be seen to present six 
apparent cells, of which three are ovuliferous, and opposite 
t le sides of the ovarium, and three are empty, and opposite 
i s angles. The latter must be considered as cavities 
occasioned by the imperfect union of the sides of the cells 
ol the ovarium, and as pointing out, in a distinct manner, 
the accuracy of Richards’s truly excellent notion, that 
tie dissepiments of a pericarpium are occasioned by the cohe- 
sion or conf<errumination of the two sides of a given number 
oS simple unilocular ovaria in juxtaposition round a common 
aais. e have no hesitation in declaring our opinion that 
1 Is m such correct and philosophical views of the 
f P an * s as *his is, and in such alone, that 
len i c otany can be said to consist; although we 
