39 
dense raceme, distinctly peduncled or subsessile, 6-17 cm. long, cylindric. Peduncle 
usually 1-3-5 cm. long, tomentose with spreading hairs. Pedicels 3 mm. long, tomentose 
with spreading hairs. Sepals 5-7 mm. long, 3 mm. broad, ovate, concave, acute or obtuse, 
tomentose with spreading hairs, two of the sepals united. Petals white 1-1-1 -5 cm. long, 
2 mm. broad, oblong, linear-oblong or linear, narrowing sometimes into a linear claw 
below, subostuse or acute, villous. Filaments 0 -8-1-2 cm. long, terete and glabrous in 
the upper half, expanded villous and monadelphous in the lower half; anthers 1-5 mm. 
long, oblong in outline. Ovary densely villous; style 8 mm. long, densely villous below, 
glabrous or almost so above; stigma subglobose or globose. Fruit 2:2 cm. long, about 
2-2 cm. in diameter, subglobose; the valves covered with woody protuberances. Seed 
1-2 em. long, about 6 mm. in diameter, more or less keeled on one face, convex on the other, 
reddish with a yellow arillus at the base. 
Komgha Div., Komgha, on Alan Page's farm “ Goldust,” May (in leaf only), Galpin, 
8470; Prospect, near Komgha, September, Flanagan, 309; Instsubani Forest, February 
(fruit), Leigh in Herb. Forest Dept., 1840; Wilo Forest, Mqanduli Dist., Kriel in Herb. 
Forest Dept., 1772; Mboleni Forest, Notineela, October, Fegen in Herb. Forest Dejt., 
1771; Manina Forest, Engcobo Dist., August, Zahn wm Herb. Forest Dept., 2047 (leaf); 
Cwebe Forest, January (fruit), Pretorvus in Herb. Forest Dept., 1854. 
Gerrard 1428 from Natal. Description of branches and leaves by Baker in Journ. 
Bot. Jan., 1907; p. 15. 
Arbor vel frutex? Ramulis cortice griseo tectis ; foliis saepissime 4-jugis cum impari 
petiolulato, foliolis petiolulatis oblongis apice apiculatis lateralibus aliquantulum inaegui- 
lateraeibus oppositis vel passim alternis bas! cuneatis vel interdum rotundatis coriaceis 
margine integris vel plus minus serratis costa superne impressa subtus subconspicua nerviis 
secundariis tenuibus inter se juxta marginem anastomosantibus foliolis terminalibus 
oblongo-oblanceolatis `foliolis proximis quam distalibus minoribus ; rachi omnino exalata. 
I have seen a specimen in leaf only (Zahn in Herb. Forest Dept., 2037) which agrees 
with the above description, except that the leaves are 6-jugate. The leaflets which appear 
to be young have a long hairy apiculus, and are either entire or serrated in the upper half. 
In Tyson, 6216, in the Bolus Herb. on which Oliver founded his B. tysoniana, I find that 
some of the leaflets have this hairy apiculus and some are also toothed, so that it would 
appear that Baker is correct in regarding Gerrard, 1428, as near B. tysoniama, and I am 
of opinion that it is this species. The distribution of B. tysomama suggests that it will 
probably be found in Natal. | 
