MR. G. BENTHAM—REVISION OF THE GENUS CASSIA. 505 
ones about which I feel some doubt. This total number might be readily raised to at least 
360, by adopting as species several which T have considered rather as marked varieties, 
or might perhaps equally well be reduced to 300, by uniting those which are not 
marked off by any definite limits. 
Of the general character and affinities of the genus I need say nothing; for the territory 
assigned to it in the system has long since been well fixed, and no ambiguous or con- 
necting species has hitherto come forward to disturb the frontier. The distribution of 
the numerous species into subordinate groups, however, has presented some difficulties. 
De Candolle's seven sections are mostly natural, but very unequal in value; and the cha- 
racters he gives, which he could only verify in a small number of species, will not stand 
the test of a closer investigation. Vogel’s seven sections, partly the same as, partly 
different from De Candolle's, are much more accurately defined, and some of them excellent; 
but the characters being taken sometimes from the anthers alone, sometimes from the 
pods or the seeds alone, some of his groups are too technical, of unequal value, or so far 
from natural, that, for instance, between the great sections of Chamefistula with 28 
species, Chamesenna with 51, and Senna with 12 species, there are no less than 45 species 
well known by good flowering specimens, which he is obliged to leave as ambiguous 
between the three. Following up, however, the indications so well broached by Vogel, 
it appears to me that the genus may be divided into three large groups, so well defined in 
flower and fruit as to have a right to the designation of subgenera. These are :— 
1. FISTULA, in which the three lower anthers, on long incurved filaments, open more 
or less from the top downwards on the inner face in longitudinal slits; four or more of 
the others, on shorter filaments, open by basal pores, one to three of the uppermost being 
usually smaller and indehiscent. The pod is long, cylindrical or slightly compressed, very 
hard, and usually, if not always, indehiscent ; the transverse septa between the seeds are 
very perfect, and the seeds flattened parallel to those septa, usually inclosed in a juicy or 
dry and pith-like pulp. The species of this subgenus are all trees or tall shrubs; the flowers 
large and showy, in racemes, produced usually, but not quite always, on the old wood ; 
and there are no petiolar glands. 
2. SENNA, in which the perfect anthers (either seven only, the three upper ones being 
small and imperfect, or the whole ten) open in single or double terminal pores, or in 
terminal short slits, occasionally continued a short way down the sides. The pod opens 
along one or both sutures without elasticity, or in a few species remains indehiscent with 
a thin or soft pericarp. The seeds, varying much in the direction in which they lie in 
the pod, are always attached by filiform funicles. The habit and inflorescence are very 
varied. 
3. LASIORHEGMA, in which the anthers, either ten or sometimes fewer by abortion, 
open in short terminal slits, occasionally continued down the sides as in a few species of 
Senna, but usually with a marked woolly line down the sutures. The pod opens elas- 
tically in two valves; and the funicle of the seed is reduced to a small adnate or triangular 
protuberance, never filiform. The habit and inflorescence of each of the three sections 
comprised in the subgenus is peculiar; but, in the case of two at least, corresponding 
- forms are to be found in the other subgenera, 
| VOL. XXVII. — 23x 
