526 MR. G. BENTHAM—REVISION OF THE GENUS CASSIA. 
Var. ?cHILENSsIS, foliolis ssepius minoribus rigidioribus, floribus minoribus, legumine 
angustiore. 
C. flexuosa, A. DC.! Not. 7, Pl. Rar. Hort. Gen. 34, t. 4, non Linn. 
C. Candolleana, Vog.! Syn. Cass. 42. 
C. frondosa, Hook. et Arn.! in Hook. Bot. Misc. iii. 210, non Ait. 
C. quebracho, Steud. 1833, ex ejusd. Nom. Bot. ed. 2. 
C. obtusa, Closs !, in C. Gay, Fl. Chil. ii. 235. 
C. emarginata, Closs !, in C. Gay, Fl. Chil. ii. 237. 
Hab. Tropical and subtropical South America, very common, extending from South 
Brazil to the W. Indies and Central America. Frequently sent also in collections from 
tropical Asia, but there only in cultivation. 
The var. pubescens occurs both in Columbia and Brazil, differing more or less from the common, more 
glabrous form. The West-Indian and Central-American specimens have generally the leaflets rounder 
and thinner than the southern ones. The var. Zenuifolia, with larger and thinner leaflets, has a very limited 
Brazilian range. The var. chilensis appears to be common in Chile, and being there tolerably constant, 
I have felt much hesitation ás to whether it should be regarded as a distinct species or not. The pod is 
sometimes scarcely above ¿ in. in diameter, with the seeds not crowded, alternate, and mostly solitary ` 
between each two transverse septa, without any longitudinal partitions; whilst in luxuriant, especially 
cultivated specimens of the typical C. bicapsularis, the pod is turgid, above 4 in. in diameter, and, the 
crowded seeds being in nearly opposite pairs, the septa separating them become so displaced as to assume 
the appearance of a transverse septum between each pair, with a vertical one between the two seeds of the 
pair; and when this is regularly continued throughout the pod, it has the appearance of being vertically 
divided by a continuous longitudinal septum, whence the specific name. This difference, if constant, 
together with that in the foliage, would fully have justified the maintenance of the two species; but I 
have seen so many specimens where the foliage is intermediate, or where the arrangement of the seed is 
irregular, that, unless more positive characters be detected by further observation, I feel unable to recog- 
nize them as sufficiently distinct. 
47. C. CORYMBOSA, Lam. Dict.i. 644. Foliola 3-juga (rarius subinde 2-juga), oblonga 
v. oblongo-lanceolata, obtusa v. acutiuscula, glabra; glandula inter inferiora v. omnia 
paria. Racemi axillares v. in paniculam corymbosam dispositi.—77. Bras. 107. 
C. corymbosa, Jacq. Fragm. t. 101. f. 1; Bot. Mag. t. 633. 
C. crassifolia, Ort. Dec. x. 122, ex DC. 
C. falcata, Dum. Cours. ex DC. et Hort. plur. non Linn. 
Chamefistula corymbosa, G. Don, Gen. Syst. ii. 451. 
Hab. Extratropical and subtropical South America: Buenos Ayres, Soati Brazil, and 
Bolivia. 
C. cubensis, Hoffmans. Verz. 209, ex DC. Prod. ii. 506, is probably some garden form of C. corymbosa 
or a ‚levigata, with an erroneous origin. It is not mentioned either by A. Richard or by Grisebach as 
a Cuban plant. 
Var. ? PLURIJUGA, foliolis ssepius 4-5-jugis, racemis omnibus in paniculam corymbosam 
dispositis. 
C. floribunda, Cav. ex Collad.! Hist. Cass. 88, t. 11, non Cav. Descr. 
E corymbosa, Ort. Dec. 124, ex Collad. 
I have seen no wild specimens corresponding with the original one from the garden of Madrid, figured 
by Colladon and preserved in herb. DC. Cavanilles’s description (Descr. 132) applies evidently to the 
common C. levigata, under which name was arcs in Germany the plant generally known at the 
