98 MR. BENTHAM ON MALVACEJE 
the arborescent Hibiscee than with any true Sterculiaceous ge- 
nera. It is true that the smooth pollen-grains have been adduced 
as a positive character connecting them rather with Sterculiaceæ 
than with Malvaceæ, but its constancy is very far from being 
proved. The pollen has only been described in a very few genera. 
As far as my observation goes, it is always tuberculate or muricate 
in what are considered as true Malvacee ; but so I have found it 
also in Hampea, in several Helicteroid genera, &c.; so that, in 
the present state of our knowledge, the pollen cannot be taken as 
furnishing an ordinal character. 
Although the one-celled anthers are thus taken by common 
consent as the essential characters of Malvaceæ, on account of its 
remarkable constancy in genera otherwise related, yet it is of very 
little organic importance. It is not occasioned by the constant 
abortion of one cell, but by the two cells, placed end to end as in 
many distinctly two-celled genera, but confluent from a very early 
period. In many genera no trace of any transverse partition or 
even contraction can, I believe, be traced at any age; in others, 
in the young bud, there is a distinct contraction in the middle of 
the anther, showing its normal structure. Helicteres, which on 
many accounts belongs undoubtedly to Sterculiaces, shows, in 
regard to the anthers, the passage from the Malvaceous confluent 
anthers of H. pentandra, &c., to the Sterculiaceous distinctly two- 
celled ones of H. angustifolia, &c.; but here the contraction in the 
young anthers may, I believe, be found in all the species. 
MALvVACEX have been distributed into tribes, and the genera 
circumscribed with so much tact and ability by A. Gray, that I 
have little to propose in modification of his arrangement, except 
the addition of one or two genera which had not come under his 
observation, and the annexation of Bombaceæ,—in consequence of 
which I should propose reducing some of his tribes to the rank of 
subtribes, as will appear in the following enumeration of tribes 
and genera. 
Tribus I. Marvzz. Columna staminea apice v. usque ad apicem 
antherifera. Styli rami tot quot ovarii loculi v. carpella. 
Carpella matura ab axi v. receptaculo secedentia (exceptis 
Bastardia et Howittia). 
Subtribus 1. MaroPrz. Carpella inordinate congesta. Ovula so- 
litaria, adscendentia. 
* Styli rami longitudinaliter stigmatosi. 
Genus :—1. Malope, Lina. 
