MR. G. BENTHAM—REVISION OF THE GENUS CASSIA. 533 
Chamefistula Sophera, G. Don, Gen. Syst. ii. 452. 
Cassia sopheroides, Collad. Hist. Cass. 133. 
C. Canca, Cav. Descr. 132, ex char. 
C. ruscifolia, Jacq. Ic. Rar. i. t. 71, ex icone. 
E . lanceolata, Forsk. Fl. Æg. Arab. ex parte; Bisch.! in Bot. Zeit. 1850, 897, t. 10, non aliorum. 
C. ligustrina, Forsk. ! in herb. Mus. Brit. 
Senna esculenta, Roxb. ! Fl. Ind. ii. 346. 
Cassia esculenta, Roxb.! Hort. Beng. 31. 
€. @gyptiaca, Willd. Enum. Hort. Berol. 442, ex char. et fide Vog. 
E robinioides, Willd. l. c. 
C. Barclayaua, Sweet. Fl. Austral. t. 32. 
G schinifolia, A. DC.! Not. 7; Pl. Rar. Hort. Gen. 35. 
C. atropurpurea, Span. in Linnæa, xv. 201, ex char. 
Var. TORULOSA, foliis fere C. occidentalis, legumine tereti subtoruloso. 
C. chinensis, Jacq. Coll. i. 64; Ic. Rar. t. 73 (forma hortensis, speciosa, racemis fere in paniculam 
terminalem confertis). 
Chamefistula chinensis, G. Don, Gen. Syst. ii. 452. 
Cassia indica, Poir. Dict. Suppl. ii. 127. 
C. torosa, Cav. Descr. 131, ex char. 
C. torulosa, Poir. Dict. Suppl. ii. 126 (nomine Cavanillesco mutato). 
Chamefistula torosa, G. Don, Gen. Syst. ii. 452. 
Var. LIGUSTRINOIDES, inflorescentia magis corymbosa. 
C. coromandeliana, Jacq. Fragm. 67, t. 100, ex char. et ic. 
Chamafistula coromandeliana, G. Don, Gen. Syst. ii. 452. 
Cassia frutescens, Mill. ! Dict. n. 2. 
C. patula, Ait. ! Hort. Kew. ed. 1. ii. 51. 
Var. PURPUREA, ramis (fide Roxb.) purpurascentibus, foliolis sepius minoribus, ramu- 
lorum interdum obtusis. 
Senna purpurea, Roxb, ! Fl. Ind. ii. 342. 
Senna Sophera, var pupurascens, Roxb. ! l. c. 347. 
Cassia purpurea, Roxb.! Hort. Beng. 31. 
Hab. Tropical Asia, Africa, and Australia; more rare in tropical America, chiefly in 
Central America, the West-Indian Islands, and Columbia; in all continents most fre- 
quently only as cultivated or escaped from cultivation, with some evidence of being really 
indigenous in Australia, and more doubtfully so in Asia and America. 
I have in the former part of this paper alluded to the uncertainty which prevails as to the origin of this 
species and the degree of affinity which it bears to C. occidentalis, an uncertainty which there must 
always be great difficulty in clearing up. Almost the whole of the long list of supposed species enume- 
rated as synonyms have been described from garden specimens. The best observers describe C. Sophera 
as shrubby and C. occidentalis as annual; but from some collector's notes or gardener's observations it 
would appear that the former may flower the first year so as to appear annual, and that C. occidentalis, 
when large, may become more or less woody at the base. Roxburgh describes both his S. Sophera and 
S. purpurea as annual. The pod of the var. torulosa is described and figured with much more elevated 
transverse ridges than any I have seen. This appears to be a garden variety, with the foliage of C. occi- 
dentalis ; the only Chinese specimens I have scen are the true C. occidentalis with a flattened pod, which 
is a weed of waste places in S. China as in other parts of tropical Asia. ‘The var. ligustrinoides kas the 
