284 FLORA NIGRITIANA. 
lata, in apicem obtusum angustata, basi in petiolum 1-2 lin. 
longum desinentia, opaca, subter pallidiora; venis incon- 
spieuis. Pedicelli j unc. longi. Flores lj lin. lati. Caly- 
cis lobi concavi, late rotundati. Petala brevissime unguicu- 
lata. Discus amplus, planus. Stamina filamentis brevissimis ; 
antheris transversee longatis, cylindraceis; polline trigono. 
The solitary specimen differs very much from any other 
species, though wanting any more striking character than 
the rufescent colour. Having seen mo fruit, the genus is 
perhaps doubtful. The form of the anther is nearly that of 
Hippocratea, but the inflorescence is very different. 
The only other described West African Salacia, the S. Afri- 
cana, DC. (Tonsella, Willd.) from Guinea, may possibly be the 
same as some one of the preceding species. 
XLIII. CELASTRINEX.* 
1. Celastrus (Catha) Senegalensis, Lam. DC. Prod. 2. p. 8.— 
On the Gambia, Don ; Senegal. 
The C. coriaceus, Guill. et Perr., also from Senegal, and 
the C. lancifolius, Schum. et Thonn., from Guinea, are the 
only other W. Tropical African species, and belong likewise to 
the section or genus Catha, well distinguished in most cases 
from the true Celastri by the axillary inflorescence, short style 
and thin incomplete arillus, although the numerous South 
African species have not yet been sufficiently examined to ascer- 
tain the real value of these characters. Presl has farther sepa- 
rated some species to form his two genera Encentrus and Poly- 
acanthus, to the former of which the C. coriaceus might be 
referred, were the chief character, the two-celled ovary and 
capsule, constant, but though the number of cells be indeed 
generally two, I have occasionally found three. So again, m 
Polyacanthus stenophyllus, neither the supposed quaternary 
parts of the flower, nor the unilocular capsule are by any means 
constant in one and the same individual, and both genera must 
* By G. Bentham. 
