150 ERICACEEX (Guthrie & Bolus). [ Erica. 
terminal, 3-nate or in clusters of 3-6 ; pedicels slender, about 1 lin. 
long; bracts closely approximate, small; sepals linear-subulate or 
narrow-laneeolate, acute, foliaceous, hirsute, shortly gland-ciliate, 
about 1 lin. long; corolla suburceolate, tubular-urceolate or ovoid- 
urceolate, more or less contracted at the throat, usually tetragonous, 
viscid, glabrous, 11-14 lin. long; segments more or less spreading or 
suberect, semiovate, rounded, from 4-3 the length of the tube; 
stamens usually 8, but often 4-7; filaments narrow, a little dilated 
and dark-coloured below the anther; anthers exserted or subexserted, 
lateral, sublateral or rarely subterminal, narrow-oblong, 3—2 lin. 
long, muticous; pore about $ the length of the cell; style slender, 
exserted ; stigma small, subsimple ; ovary globose, glabrous. Benth. 
in DC. Prodr. vii. 670. E. humilis, Benth. l.c. 615, not of Neck. 
nor Salisb. E. divergens, Wendl., E. flavida, Klotasch, and E. 
connivens, Klotzsch, ex Benth, l.c. 670. 
VaR. 8, maritima (Bolus) ; leaves somewhat longer and more distant than 
usual ; flowers somewhat smaller, 1}—1} lin. long; stamens 4; anthers narrow- 
elliptic, subcuneate at the base; cells approximate, pallid, 4 lin. long; pore 4-4 
the length of the cell. 
Var, y, longibracteata (Bolus) ; bracts slender, but longer, one foliaceous and 
reaching to the top of the corolla ; ovary pubescent. 
SoutnH AFRICA: without locality, Roxburgh! Herb. Salisbury! Drége! 
Coast Rxreion, from 800 to 2000 ft.: Caledon Div. ; Klein Houw Hoek, 
Zeyher, 3217! Nieuw Berg, near Palmiet River, Zeyher, 3332! hills near 
Grabouw, Bolus, 4177! 4178! Guthrie, 4169! Houw Hoek, Bolus, 7369! 
6958! Schlechter, 9425! Var. 8: Bredasdorp Div.; bills near Cape Agulhas, 
250 ft., Schlechter, 10559! Var. +: Stellenbosch Div.; Lowrys Pass, 1500 ft., 
Schlechter, 7247 ! 
This is a curiously variable species in respect of the number of its stamens. 
We have found them 4 to 8. Those specimens with 4 stamens (Bolus, 4178, and 
Zeyher, 3332) are so in most, if not all, the flowers. They are then technically 
Bleria and not Erica. Even if these stood alone it would seem a forcing of 
nature to separate them from E. filiformis, with which they agree in all other 
respects. But intermediate forms appear to indicate at once an unstable 
condition, which induces us to abandon (for systematic purposes in this 
particular case) a character based upon stability in the number of the stamens. 
194. E. Tysoni (Bolus in Journ. Linn, Soc. xxiv. 181); diffuse, 
decumbent, a few in. high ; branches numerous, straggling, slender 
but woody and rigid, scabrid-hispid or glabrescent, 6-12 in. long ; 
leaves 3-nate, erect-spreading, linear or narrow-lanceolate, suleate, 
viscidulous, scabrid-hispidulous, sometimes glabrescent, about 1 lin. 
long; fiowers axillary, 3-nate; pedicels + lin. long ; bracts approxi- 
mate, oblong or obovate, foliaceous; sepals oblong or lanceolate, 
obtuse, foliaceous, gland-ciliate, about 1 lin. long; corolla tubular- 
campanulate, slightly widened to the mouth, generally more or less 
(sometimes strongly) tetragonous, viseidulous, purple (Hvans), 14-2 
lin. long; segments broad, erect-spreading, from 2—% the length of 
the tube; anthers subexserted, lateral, dorsifixed shortly above the 
base, oblong, scaberulous, ciliolate, about + lin. long, aristate ; pore 
3-4 the length of the cell; awns ciliolate, eurved, about 4-2 the 
length of the cell; style exserted; stigma capitellate; ovary 
