Erica.] ERICACER (Guthrie & Bolus). 183 
ultimate very slender; leaves erect or spreading, linear, obtuse, 
glabrous, slender, often curved, 3}—1 lin. long; flowers 4-nate, some- 
times by abortion of branchlets pseudo-axillary ; pedicels slender, 
3-] lin. long; bracts remote, very smali; sepals ovate, obtuse, 
ciliate, shining, less than 2 lin. long; corolla subglobose-campanu- 
late or broad-urceolate-campanulate, not (or hardly) constricted at 
the throat, somewhat widened at the mouth, $-1 lin. long; 
segments about 2 as long as the tube; anthers included, sub- 
triangular, with broadly rounded angles, subscabrid, ciliate at the 
base, 1 lin. long, aristate; pore about 3 the length of the cell; 
awns spreading, distant, naked on the margins, shorter than the 
eell; style included, short; stigma subpeltate-capitate; ovary 
glabrous or puberulous. . levis, Andr. Heathery, t. 221 and Col. 
Heaths, t. 182; Spreng. Pugill. i. 30? Lodd. Bot. Cab, t. 1393. 
E. caffra, Lodd. lc. t. 196, not of Linn. nor of others. EH. panicu- 
lata, var. alba, Wendl. Eric. Ic. fasc. 14, 25% £. paniculata, Wendl. 
ex Steud. Nom. ed. 2,1. 577. E. persoluta, var. major, Wendl. Lc. 
fasc. 24,187, ¢.71? E. persoluta, var. levis, Benth. in DC. Prodr. 
vii. 679. Z stenophylla, Benth. lc. 679. E. odorata, Spreng. 
Neue Entdeck. i. 271. 
Sourn Arrica: without locality, Herb. Salisbury! and cultivated speci- 
mens ! 
Coast Rr@ion: Worcester Div.; banks of streams in Hex River Valley, 
1600 ft., Tyson, 637! Caledon Div.; mountains near Grabouw, Bolus, 4176! 
and without exact locality, Thom, 978! Tulbagh or Worcester Div.; Bolus, 
5191! 
There has been some confusion about this species. Our specimens from several 
localities agree well with each other, with the specimen in the Berlin Herbarium 
marked E. levis, Spreng., and with Bentham’s type of E. stenophylla. Salisbury 
himself saw the resemblance of his species to EL. viridipurpurea, L., though he 
observes : “Closely allied to [it], the anthers being almost exactly the same ; 
yet on account of the difference in the interstices of the corolla-limb, I can 
scarcely regard it as a variety of that.” To this we can only add that the leaves 
appear to be constantly smaller,—about half the size, It is much nearer to 
E. viridipurpurea than to E. subdivaricata, under which it was also placed by 
Bentham. It is also allied to E. quadrangularis, but the habit and general 
aspect are different, leaves constantly smaller, and the corolla less wide at the 
mouth. 
252. E. hemastoma (Wendl. Eric. Icon. fase. 27, 18); branches 
few, erect, slender, glabrous; leaves erect to spreading, linear, 
sulcate, the younger pubescent and ciliate like the bracts and sepals 
with minutely plumose hairs, glabrescent, terminating In a sharp awn, 
3-6 lin. long; flowers sub-4-nate ; pedicels 2-3 lin. long; bracts 
subremote, two very small, the third longer, foliaceous ; sepals 
lanceolate, acuminate, awned, coloured, 1-13 lin. long ; corolla 
ovoid to globose-urceolate, mouth much contracted, 2}—3 lin. long or 
more ; segments ereet, rounded ; anthers included, subeuneate or 
longitudinally semiovate, dorsally curved, 2 lin. long, crested at the 
base; pore £ as long as the cell; crests more than } the length of 
the cell, lobed, acute, erosulate ; style included or just manifest ; 
