114 MR. BENTHAM ON MALVACEJE 
ception of Pterygota, characterized by its winged seeds. Brown 
has since (Pl. Jav. Rar. p. 224), with his usual perspicuity, pointed 
out the more important characters to be derived from the arrange- 
ment of the anthers and the structure of the seed. He reunites 
many of Schott’s genera with Stercülia, but still admits ten dis- 
tinet ones, including two not mentioned by Schott. Some of these 
are monotypic, and founded on the position of the radicle with re- 
lation to the hilum—next to it, at the opposite end, or between 
the two. Important, however, as similar characters are in most 
cases, they can yet be regarded only as artificially sectional when 
separating single species not otherwise distinct from the main 
group. In the case of Firmiana, the two species proposed to be 
generically united, as having in common the intermediate position 
of the radicle, are in habit and in their calyx as different from each 
other as any two species of the whole group. But we have no 
hesitation in adopting as a good genus the African Cola (including 
Courtenia) ; for there are several species at once distinguished from 
Sterculia by their anthers adnate in a single ring, and by their 
want of albumen, accompanied by other minor characters. Cour- 
tenia, with divaricate instead of parallel anther-cells, appears better 
considered as a section than as a genus; for neither here, nor in 
other Sterculiaceous genera where the same diversity occurs, does 
it entail any other tangible differences. ~Tetradia, Br., must also 
be provisionally admitted; for the fruit and seed are as yet un- 
known, and may present characters corroborative of those derived 
from the flower. But we would restore to Sterculia, as mere sec- 
tions, Firmiana, Brachychiton, Pterygota, Hildegardia, Scaphium, 
and Pterocymbium ; including in Brachychiton the Delabechea of 
Lindley, in which we find the radicle next the hilum as stated by 
Brown, not remote from it as described by Lindley. 
TannIETIA, Blume. 
This genus, allied in most respects to Sterculia, has the indu- 
mentum and inflorescence of Heritiera, with much smaller flowers 
and very peculiar samaroid carpels. It includes an Australian 
species published by F. Müller under the name of Argyrodendron. 
The leaves are digitately compound, with five leaflets according to 
Blume's figure and description, three only in the specimens dis- 
tributed by Miquel as Blume's species, as well as in the Australian 
species. 
Myropia, Schreb. 
We have already stated our reasons for excluding from Myrodia, 
