404 ERICACER (Brown). | Salazis. 
Burchell notes on his label that the Knysna plant grows to 4 ft. high. The 
longitudinal ridges on the ovary, generally 8 or more in number, are very 
characteristic of this species, but they are only faintly marked on the fruit. 
5. S. octandra (Klotzsch in Linnea, ix. 353) ; apparently about 
1 ft. high; branchlets slender, erect, subparallel to very flexuose, 
distinctly or very minutely puberulous, sometimes becoming glabrous, 
pale brown or greyish ; leaves 3-nate, varying from adpressed and 
shorter than the internodes to erectly imbricate, $-3} lin. long, 
narrowly linear, acute, glabrous; flowers axillary and terminal, 
1-3 (rarely 4-5) together, subsessile; calyx (excluding the long 
lobe) 1—1 lin. long, shorter than the corolla-tube, unequally lobed to 
the middle or the long lobe nearly free, glabrous ; lobes oblong or 
ovate, acute or subobtuse or the long one acuminate, all minutely 
ciliate; corolla about 2 lin. long, eampanulate or subglobose-cam- 
panulate, glabrous; stamens 8; filaments about } as long as the 
anthers, rarely less, connate for —% of their length and sometimes 
up to (but not with) the base of the anthers, becoming more or less 
free in fruit and narrower at their apex than the base of the 3 to 
nearly 1 lin. long oblong anthers; ovary acutely 2—4-angled on the 
upper half, 2—4-celled, glabrous ; style very short to almost none ; 
stigma deeply funnel-shaped with erect or slightly inflexed sides to 
the cup, glabrous ; fruit distinetly 3-4- (oceasionally 2-) angled on 
the smooth hardened upper half, slightly reticulate-rugulose below. 
Klotesch in Linnea, xii. 212; Benth. in DC. Prodr. vii. 711; 
Dietr. Synop. Pl. ii. 1261. 
Var. B, artemisioides (N. E. Br.) ; flowers very shortly pedicellate ; calyx 
(excluding the long lobe) nearly or quite as long as the corolla tube; corolla not 
quite 4 lin. long, obconic-cup-shaped; staminal filaments at first connate 
almost up to the anthers; anthers } lin. long; otherwise as in the type. 8. 
artemisioides, Klotzsch in Linnea, xii. 212; Benth. in DC. Prodr. vii. 711; 
Dietr. Synop. Pl. ii. 1261. Erica artemisioides, E. Mey. ew Klotzsch lec. 
Sout Arica: without locality, Mund § Maire, ex Klotzsch. 
Coast RueGion: Worcester Div.; mountains of Bains Kloof, 4000 ft., 
Schlechter, 9127! Caledon Div.; mountain near Greitjes Gat, 2000-4000 ft., 
Ecklon § Zeyher ! Bredasdorp Div.; sand-dunes near the mouth of Ratel River, 
50 ft., Bolus, 6770! Var. 8: Worcester Div.; “ Dietris Kloof” an error for 
Dutoits Kloof, Drége, 1168! 
Drage has evidently mixed up three plants under the name of Erica artemisi- 
oides, for the two specimens at Kew, distributed under that name, are quite 
distinct from the type of Salawis artemisioides at Berlin, and appear to belong 
respectively to S. Sieberi and S. awillaris. Bentham in DO. Prodr. vii. 711 
referred these two specimens to S. hewandra (Coccosperma forbesianum), from 
which, however, they are entirely different. 
6. S. Sieberi (Benth. in DC. Prodr. vii. 711); 1 ft. or more 
high, much branched ; branchlets divergently aseending, minutely 
subtomentose ; leaves 3-nate, erectly imbricate to adpressed and 
shorter than the internodes, 1-21 lin. long, linear, acute, glabrous ; 
flowers axillary and terminal, 1-5 together, subsessile ; calyx (exclud- 
ing the larger lobe which is variably connate with the others or free 
almost to the base) 3—3 lin. long, shorter than the corolla-tube, 
