522 MR. G. BENTHAM—REVISION OF THE GENUS CASSIA. 
Plate 62, Cassia bacillaris : fig. 1, two of the stamens, one long and one intermediate; fig. 2, one of 
the staminodia; fig. 3, ovary ; fig. 4, pod; fig. 5, the same when commencing to open; fig. 6, portion of 
the ripe pod. 
This species has been placed by Vogel with two others, C. Humboldtiana, DC. (C. spectabilis, DC.), 
and C. sclerocarpa, Vog., in a distinct subsection, Xylolobium, with the fruit of Fistula, E legumine lignoso 
indehiscente," and the flowers and foliage of Chamefistula, a combination which 1 had never met with ; 
and although flowering specimens of C. bacillaris are common in herbaria, I could never procure the pod, 
for all the fruiting specimens I should otherwise have referred to it were totally at variance with the 
descriptions given. Upon further investigation it appeared that no botanist since the younger Linnæus 
had been more fortunate ; all modern descriptions and figures of the fruit had been either copied from 
Linnæus or taken from that of C. Fistula, to which Linnæus says it is precisely similar. His words are:— 
“ Legumen omnino C. Fistule “ut distingui nequeat, subcurvum, cylindricum, glaberrimum, acumine 
filiformi terminatum, longitudine pedale, interius isthmis interceptum, quod externe tamen non apparet," 
Linn. fil. Suppl. 231. In this description the word subcurvum already indicates a character not found in 
C. Fistula ; and having thus ascertained that this was the sole foundation for the character generally given, 
I turned to Dahlberg's original specimen described by Linnæus, which is in Sir J. Smith's herbarium at 
the Linnæan Society. I there find the pod unripe and longitudinally cut with a knife. In that state it 
is not unlike the unripe pod of a small Fistula; but I have thus been enabled to identify it with that of 
other fruiting specimens in the herbaria before me, especially those from Wullschnagel's Surinam col- 
lection in Martius's herbarium, and to trace the pod to perfect maturity, when it splits along the inner 
suture and opens out into a kind of follicle, as in other species of the present series. 
Of the three representations ofthe species usually cited, that of C. fruticosa, Mill., in * Reliquiæ Hous- 
tonianze, t. 17, appears to have been taken from a specimen of the true plant; but to get it into the 
small compass allowed, the artist has so altered the proportions, shortening the leaflets and pod and inereas- 
ing the curvature of the latter, as to make it unrecognizable. Vogel, who had at first followed others in re- 
ferring C. fruticosa, Mill., to C. bacillaris, on seeing the original specimen in the Banksian herbarium found 
an expanded fruit so different from that of C. Fistula that he restored C. fruticosa as a distinct species 
(Linnza, xv. 67), which must now, however, again be united with C. bacillaris. Jacquin's figure (Frag- 
menta, t. 85. f. 4) is only conjecturally referred by the author to C. bacillaris ; he received the pod with 
the one he figures of C. grandis from the W. Indies as having been brought there from the American 
continent, without foliage or any other indication of the species they belonged to. As the pod in question 
was exactly like that of C. Fistula, and as C. Fistula is not American, Jacquin naturally concluded, relying 
on Linnæus’s description, that this was C. bacillaris. It must, however, have been in fact either a culti- 
vated C. Fistula or some American species such as C. moschata, which really has a similar pod. The third 
figure quoted, that of Cathartocarpus bacillus, Bot. Reg. t, 881, is certainly not the true plant, but most 
probably C. affinis; as, however, no specimen was preserved or described, Lindley's description having 
been taken from that of Linnæus, it is now impossible to identify it. The figures I now give are taken 
from specimens carefully compared with the original one, which is fortunately sufficiently perfect in flower 
and fruit to preclude all further mistake. 
Specimens from Nicaragua, Œrsted, n. 45, from Yucatan and Tabasco, E. P. Johnson, n. 33, and 
from Chachapoyas in Peru, Matthews, n. 1591, evidently nearly allied to C. oxyphylla, but with broader 
leaves and numerous flowers in a dense terminal panicle, may be considered perhaps as a broad-leaved 
pubescent variety of C. bacillaris. "The two species, however distinct they usually are, are sometimes 
difficult to determine when the specimens are imperfect. Others from Schlim, Ocaña, n, 6, have the leaves 
nearly of C. rugosa, but the paniculate flowers of C. bacillaris and oxyphylla. 
33. C. LATIFOLIA, G. F. W. Mey. Prim. Fl. Esseq. 160. Foliola ampla, obtusiuscula, 
nitida, subglabra; glandula inter inferiora. Stipulæ late, foliaccæ. 
C. ornata, Bernh. v Haye Surin. exs. n. 1408. 
