other sections of the genus. In fasciculatum the cells are 
dissimilar in form, but of nearly equal size, and both fertile. 
In falcatum the posterior cell,,still more different in form, 
is also considerably reduced in width and entirely without _ 
ovula, nor is it, as I have stated in the specific character, 
fertile. Lastly, in adnatum, of which I am now disposed 
to consider propinquum a variety, the posterior cell is not 
only entirely destitute of oyula, but is reduced to a capil- 
lary tube, which continues to adhere to the fertile cell. 
** From this series of species, in which the upper or pos- 
terior cell is gradually obliterated, we are led to the true 
explanation of the structure of Lysrromia. This genus, 
belonging to Lobeliacee, is described by M. Kunru, as 
having a single parietal placenta; it may, however, be 
regarded ash ing a septum so closely pressed to the upper 
ide of the capsule.as to leave no manifest cavity, and 
Pte cntengid produces ovula on one:side only.” 
Sari isis eile. — F 
The genera Styuiprum, Levennooxra and Forstera form 
Mr. Brown’s natural order of Stylidee ; since adopted by 
Jussieu, in the 18th volume of the Annales du Muséum. 
By the separation of Lobeliacee, Goodenovie and Stylidec 
from the Campanulacee, the genera belonging to each, 
as Mr. Brown observes, are more readily defined, and the 
attention of Botanists better directed to the very singular 
structure of these tribes. = Te 
We have before given two species of this very curious 
genus. For the present one, a native of King George’s, 
Sound, on the South coast of New Holland, we are again 
indebted to our friend Roserr Barcray, Esq. who commu- 
nicated a flowering plant in June, in the present year. 
All the Sryzip1a require to be protected from frost, and 
some of them aawer tec better for the assistance of a little 
artificial heat. i ix oi 9 
