MR. G. BENTHAM—REVISION OF THE GENUS CASSIA. 507 
only of the widely spread C. bicapsularis, its geographical and natural relations remain 
the same. 1 have, in the following enumeration, after much hesitation, elassed it as a 
variety ; but all that is thereby meant is, that, upon weighing all the evidence afforded by 
the materials at my disposal, it has appeared to me that its relationship to the other forms 
of C. bicapsularis is rather of that degree which botanists whose views I adopt call varie- 
ties than of the remoter degree which they term species. | 
Of the original seat or birthplace of the genus Cussia we have no evidence whatever, 
and I cannot even put forward a conjecture. The fossil Cassias mentioned by Heer and 
Saporta, even if any of them may with tolerable plausibility be referred to the genus, are 
far too recent to afford any clue. All that the consideration of the data at our command 
would suggest is, that at least the three great and well-defined types, Fistula, Senna, and 
Lasiorhegma were differentiated at a time when the configuration of land and water was 
very different from what it now is, and had been well fixed before the areas they occupied 
had become broken up by the intervention of apparently insurmountable obstacles; for ` 
all three of these great groups show the same combination of general characters in 
America, in Asia, in Africa, and in Australia. Going even a little further, we may ob- 
serve here and there a subordinate group, such as that of the Chamesenne Picte, bearing 
a close family resemblance in the New and the Old World, which seems to show a differ- 
entiation at an early period. Descending lower in the scale, if we observe any one organ 
similarly modified in several species in both hemispheres, the correlative changes in 
other organs are usually different in the two cases; and some groups, ranking almost as 
high as any of the subordinate sections, such as Psilorhegma and Xerocalya, are exclu- 
sively limited to one hemisphere, showing the probability of their formation since the 
interposition of impassable barriers. 
To illustrate these views I would pass in review the principal groups nearly in the 
order in which they are found in the detailed enumeration given below. 
Fistula comprises nineteen species, of which eight are American, five Asiatic, five 
Afriean, and one Australian, all retaining in each country those marked characters in 
flower and fruit which have induced several botanists to raise them to the rank of a 
genus; but each species confined to its own part of the world, and none of them forming 
natural combinations into intermediate groups, as if each species had been independently 
modified from the common stock without having passed through an intermediate stage 
of a smaller number of fixed species. It is true that the species of Fistula may be classed 
according to characters derived from foliage, bracts, or even from floral organs ; but such 
classification, as far as hitherto attempted, is purely artificial, the modification of one 
organ bearing no correlation to that of any other; and if some of these modifications of 
isolated organs are peculiar to one of the four continents, this may only show that cir- 
cumstances have acted in one case on one organ, in another on another. The large ovate 
leaves in few distant pairs of C. Fistula and others are chiefly characteristie of Asiatie 
and African species; and when they occur in American species ( C. Spruceana and C. 
Sagotiana) they are accompanied by a dilatation of the petiole never observed in Asia or 
in Africa. Numerous narrow tomentose leaflets are common in American species; and 
this divergence is accompanied in C. grandis by a marked difference + up pod and. — 
X 8 
