MR. G. BENTHAM—REVISION OF THE GENUS CASSIA. 511 
perhaps in their requirements, may have either never spread into Asia, or not been able 
to maintain themselves there. 
From the Picte have evidently diverged the true Sennas, differing chiefly in the short- 
ness of the pod, with a few minor characters. The divergence arose probably since the 
first disseverence of the general area, as it exists only in the Old World, and is rather 
greater in the four African or Asiatic species than in the single Australian one. 
The last and most distinct group of the subgenus Senna is the section Psilorhegma, 
strictly limited to the Old World, and chiefly Australia, where it disports itself after the 
truly Australian fashion, the foliage varying from the ordinary flat membranous leaflets 
to rigid vertical or terete ones, or the whole leaf reduced to a phyllodium, the pods 
straight or variously twisted, and all with so little ‘regard to specific distinctness that, 
whilst T considered that I had gone as far as I fairly could in reducing the published 
species to a more normal standard, F. Müller is of opinion that six or seven of those I 
have adopted are really variable forms of a single one. "There is one in Cunningham's 
collection, C, gonioides, which I had overlooked when working up the ‘Flora Austra- 
liensis;” this carries the number of „species, as there proposed, to sixteen, one of which 
extends over a great part of tropical Asia. These, with one from the Eastern Archi- 
pelago, one from New Caledonià, and one from the Sandwich Islands, make a total of 
nineteen species. The section characterized chiefly by having the ten stamens all 
perfect, with the anthers approaching those of Lasiorhegma, and the pod and seeds of 
Chamesenna, passes into the latter section through the E. Indian C. auriculata and 
C. divaricata. The similarity of the fruit is indeed so great that the Indian species 
C. glauca has always been placed in Chamæsenna, even by Vogel, who, under the impres- 
sion that Psilorhegmas were all Australian, never examined the flowers. Miquel, how- 
ever, examined a specimen of Horsfield's, and published it as a new species of Psilo- 
rhegma ; but then he also retained the identical C. glauca in Chamesenna by the side of 
Er divaricata, Nees, which is also a Psilorhegma, although with almost precisely the 
aspect of the American C. biflora, a Chamesenna, and indeed approaching that species in 
the inequality of its anthers, which, however, being all perfect, places the plant, with its 
Asiatic countrymen, in Psilorhegma, and not with the American Chamesenne. 
The third great subgenus, Zasiorhegma, distinguished by its anthers as well as by the 
` elastically dehiscent pod and exceedingly short funicle, has three very distinct sections, 
although the most prominent character of each is in the inflorescence. 
The first and most marked, Apoucouita, consists of only three South-American trees, 
two of which have but recently been discovered by Mr. Spruce. With the general 
characters of Lasiorhegma, but rather peculiar anthers, and the habit rather of Fistula, 
they suggest nothing further as to their origin or connexions. 
Absus, distinguished from Chamecrista chiefly by the inflorescence and perhaps by 
more regular flowers, has, when taken as a whole, a singular distribution. Seventy Ame- 
rican species, all shrubby or perennial and decandrous, are mostly very local, chiefly 
Brazilian, a few limited to Guiana or Columbia, and some of the few common ones 
extending to Central America and Mexico. None appear to cross the Southern Andes, 
and there is no trace of them in the Old World. The seventy-first species, on the other 
