298 MR. BENTHAM ON TROPICAL LEGUMINOSA. 
branacea, utrinque viridia, pilis longiusculis conspersa, in petiolum 1-2-linearem subarticulata. 
Stipulse setaceze, patentes v. incurvæ, rigidulæ, petiolo sublongiores. Racemi graciles, fere filiformes, 
folio sæpius longiores, hirti. Flores secus rhachin dissiti, ad axillam bracteæ setacex v. subfoliaceæ 
solitarii, breviter pedicellati, vix 2 lin. longi. Calyx brevis, pilis longiusculis rigidis hirtus. Petala 
pallide coccinea, glabra. Legumen leviter falcatum, hirtum, perfectum 8-9 lin. longum, 1 lin. latum, 
sed sæpe abortu brevius. . 
Hab. West tropical Africa. Rocks on the Niger (Vogel, Barter). 3 
2. M. LATIFOLIA. (Plate XXXIII. B.) Foliis ovatis v. ovato-lanceolatis, stipulis late 
semicordatis subreniformibusve acuminatis. 
Folia (sive foliola solitaria) 2 poll. longa, 1 poll. lata. Stipulæ foliaceæ, 2-4 lin. late. Cætera M. tenelle,- 
cujus forte varietas est insignis. 
Hab. East tropical Africa. Near Luabo on the Zambesi (Kirk). 
II. GEISSASPIS, W. et Arn. ; Benth. et Hook. f. Gen. Pl. 516. 
(Tribe HepysARezæ. Tropical Asia and Africa.) 
I refer to this well-known small genus for the purpose of characterizing a remarkable 
West-African species, in which the large bracts have not the marginal cilia of the com- 
mon G. cristata. This plant has long been in our herbaria, having been gathered by 
Heudelot in Senegambia, and communicated by M. Delessert to Sir W. Hooker, as well 
as to myself, as early as 1838. In the Hookerian herbarium it has been correctly deter- 
mined to be a G'eissaspis by Dr. Planchon; but I do not find that it has ever been pub- 
lished as such. It is probable, however, that it is identical with the Sœmmeringia 
psittacorhyncha described by Webb, in his ‘ Spicilegia Gorgonea,’ from a Cape-de-Verd- 
Island specimen which I have been unable tosee. The character given is quite at variance 
with that of the Brazilian genus Semmeringia, and agrees, in all essential points, with our 
Senegambian Geissaspis. 
The following are the technical characters of the species :— 
G. LUPULINA, Planch. in herb. Hook. Stipulis latis basi oblique subcordatis vix pro- 
ductis, bracteisque integerrimis. 
Herba annua glaberrima. Foliola 3-6-juga, anguste cuneato-oblonga, obtusissima, pleraque semipollicaria, 
petiolo communi gracili 4—2 poll. longo. Stipulæ foliaceze, oblique obovatæ v. subreniformes, 3-6 lin. 
longæ, basi sessiles v. hinc in auriculam minimam obtusam product. Spies lupulinæ, demum 1-2- 
pollicares, pedunculo foliis longiore fultæ. Bracteæ ut in G. cristata ample, oblique reniformes, venosæ, 
imbricatæ, 6-9 lin. latæ, sed integerrimæ nec ciliatæ. Petala bracteas parum excedentia. Flores et 
legumina ceterum omnino G. eristate. 
Hab. West tropical Africa. Moist light soils on the banks of the Casamance, in Senegambia (Heudelot, 
nn. 555 and 664). 
IH. DESMODIUM, DC.; Benth. et Hook. f. Gen. Pl. 519. : 
(Tribe HEDysAREX. Tropical and Subtropical Asia, Africa, and America, North America, 
| Japan, Australia.) Ä 
This large offset from the old genus Hedysarum has been irregularly split up by 
various botanists, but chiefly by separating as distinet genera isolated species which pre- 
