217 



^yliuclric form; dull dark green, indistinctly or tlie inner dis- 

 tinctly banded with paler green; margins acute, with a red- 

 brown line inside a membranous white edge; epidermis rather 

 TOUgh- Leaves of adult growths 3-5, in two ranks, straight, or 

 nearly so, rigid, erect, but gradually diverging from base to apex, 

 so that the inner are about ^-l ft. apart at the tips; smooth, 

 2^-3 ft. long, about 1^1^^ in. thick from front to back and 

 1^-1^ in. thick from side to side at the base, cylindric, slightly 

 compressed, gradually tapering from the base to a very acute, 

 rigid, terete-subulate (not compressed) whitish point -3-I in. long, 

 4\nd with a channel much narrower than the breadth of the leaf 

 on the inner face and extending nearly to the apex, concave at 

 the basal part, acutely triangular at the upper part, slightly 

 glaucous-green or bluish-green, wdien young somewhat faintly 

 banded with paler green, rather inconspicuously marked with 



eenish-white 



numerous very interrupted darker lines, and 9-12 continuous 

 slightly impressed lines, forming slight longitudinal grooves; 

 margins of the channel acute from base to apex, greeni 

 in young leaves, becoming red-brown with a whitish edge. 

 Floicer-steni imknown ; only one flower-cluster in fruit was found, 

 with 4 pedicels h in. long, iointed at about \ in. above the base. 

 IS ernes orange. 



Thopical Africa. South Angola : among rocks in open 

 forest near Km. lOS'S on the Mo^samedes EaiTway at 1700 ft.; 

 Pearson, 2073! Also occurs between Gambos and Houmbe, ex 

 Pearson. Damaraland : Grootfontcin, in TJpingtonia, Schinz^ 15! 



Described from a living plant sent to Kew by Prof. Pearson, 

 iind from seedlings raised from the fruit that accompanied it. 

 I have not seen an infloTesrence, and Prof, Pearson informs me 

 that when be discovered the plant the flowering season was over, 

 and, although he examined many jdants, only some remains of a 

 single inflorescence was found, consisting of the base of the flower- 

 stem still attached to the plant, a single cluster of fruit and some 

 loose berries laying on the ground, from which it could not be 

 determined if the inflorescence was a panicle or a spike-like 

 raceme. In all probability S. Pearsonii is the same species as 

 that mentioned under the name of S. cylindrica by Baum 

 (Kunene-Sambesi Expedition, pp. 25, 28, 78, 197, 460, 470 and 

 510), and (on p, 197) stated to have a narrow channel down the 

 face of the leaf, found growing under trees and shrubs on termite- 

 anthills along the Chitanda River, between Kakele and Goudkopje 

 at 4000 ft. elevation, and in the region of the Cunene and 

 Ivubango Rivers ; also by Schinz (Deutsch-Sudwest-Africa, 

 j)p. 238, 472) as occurring near the Cunene River in Amboland, 

 which is about 150 miles soufh-east of where Prof. Pearson col- 

 lected the j»lant here described. The true *S. cylindrica, Bojer, is 

 a native of Tioanda in Xorthern Angola, and its leaves are de-^ti- 

 tute of the acute channel down the inner face whicli characterises 

 S. Pearsonii; ihev are also of a darker ffreen. 



20. S. cylindrica, 7?o;>r, Hort, Maurlt. p. 349 (Fig. 5, c).- 



Stemlessy with a stout creeping rootstock 1-li in. thick. Leav 

 3-4 to a growth, besides some basal sheatlis, 2-rankedj stiffly erect, 



C 



