AND STERCULIACEJ. 101 
. A detailed monograph of several of the above genera, especially 
of those which, like Sida, Abutilon, Pavonia, Hibiscus, &c., con- 
tain numerous widely spread species, is much wanted, but would 
lead me too far on the present occasion; nor can I stay to inves- 
tigate or deseribe many apparently unpublished forms which we 
possess in our herbaria. There are, however, a few genera on 
whieh I should wish to add some observations, or to characterize 
some of the more remarkable new species. 
PALAVA, Cav. 
The P. rhombifolia, Grah., from Lima, is probably the same 
species as the .P. malvifolia, Cav., of which the latter author had 
probably only examined undeveloped flowers, and thus described 
the petals as of the length of the calyx. P. moschata, Cav., is a 
very distinct species; and the following one, with the habit of a 
Cristaria, appears to have been hitherto overlooked. 
P. DISSECTA, sp. n., tomentosa, foliis profunde bipinnatifidis dissectisve, 
lobis cuneato-oblongis obtusis integris v. 3-5-lobis, pedunculis calyci- 
busque hispidis. 
Hab. Peru, Cuming, n. 945; near S. Lorenzo, Maclean. 
Marva, Linn. 
This genus, stripped as it has been by A. Gray of its American 
and South-African species, becomes at once more natural and better 
characterized. Amongst European ones, the M. Sherardiana, Linn., 
notwithstanding the almost constant presence of two small bracts, 
must be referred to Sida, of which it has the styles and the seeds. 
M. Behriana, Schlecht., Linnea, xx. p. 633, from Australia, is Lava- 
tera plebeia, Br. 
PraAarANTHUS, Forst. 
Notwithstanding the close proximity of this Australian genus 
to Sida, most of the species have, under various names, been pub- 
lished as Sterculiaceous genera; for the longitudinal partition in 
the anthers (much more prominent than in the generality of Mal- 
vacesm) has usually suggested the idea of their being really bilocular. 
In some species also the ovary is reduced to three, two, or even a 
single carpel, so as, at least in the latter case, to give readily a 
false idea of its structure ; and one such species, very nearly allied 
to the original one of Forster, has even been considered as the 
type of a distinct natural order under the name of Philippodendree, 
the affinities of which have much puzzled those who only knew 
