288 ERIcAcEx (Guthrie & Bolus), [ Hrica. 
The type of Klotzsch was most probably Ecklon and Zeyher’s 30, and a careful 
comparison of this with Niven’s 95 (the type of E. polycodon, Benth.), appears 
to show that these can hurdly be separated. The corolla is a little different, but 
intermediate forms are present. In Galpin’s 3711 the anthers are subexserted, 
as in Alexander’s 9, and these approach H. decipiens in structure and appearance ; 
but in the latter the sepals are glabrous and more scarious and the corolla 
longer. All the forms (except Niven’s 95) appear to be characterized by similarly 
incurved sepals. In general appearance it resembles FE. floribunda, but may be 
distinguished by the simple hairs on its branches as well as by its capitellate, not 
obconic, stigma. 
436. E. floribunda (Lodd. Bot. Cab. t. 176); erect, 2 ft. or more 
in height; branches many, ascending, covered with a short pubes- 
eence of floccose shortly plumose ashy-grey hairs ; leaves suberect to 
subspreading, linear, obtuse, obscurely sulcate, glabrous, 1—1} lin. 
long ; flowers subcalyeine, mostly very numerous and crowded in 
dense masses on short branchlets; pedicels }—1 lin. long ; bracts 
remote or subapproximate, erect, ovate, small; sepals broad-ovate 
to orbiecular, keel-tipped, subacute, eoneave, subcartilaginous, coloured 
(mostly paler than the corolla) 1—} lin. long; corolla cyathiform to 
broad-cyathiform, mouth widened, pale rosy, 8—1 lin. long; segments 
broadly rounded, erect, about equal to the tube; filaments short, 
straight; anthers subineluded, generally manifest, oblong or ovate, 
obtuse, rough, nearly black, 1-2 lin. long, muticous; pore less than 
2 the length of the cell, style slender, well-exserted ; stigma obconic- 
cyathiform, or sometimes obconic (solid), small; ovary minutely 
hairy. Benth. in DC. Prodr. vii. 688. EH. sparsa, Lodd. Bot. Cab. 
t. 1467. EH. macrostoma, Klotzsch ex Benth. l.c. 688. 
SoutH ArFrica: without locality, cultivated specimens ! 
Coast Rea@ion : frequent, from George Div. eastwards to Albany Div., Niven, 
96! Burchell, 5523! MacOwan, 272! Bolus, 8293! 8673! Schiechter, 2244! 
5886! The following, which we have not been able to examine, probably also 
belong here: Niven, 98; Burchell, 5157, 53871, 5455! Oudtshoorn Div. ; Zwart- 
berg Pass, Tyson ! 
Loddiges’ figure is the type of this species; but for critical purposes is all but 
useless. Bentham identified with it E. galliiflora, Bartl., and we have a specimen 
so marked from the Cape Gov. Herb., most probably Zeyher’s. As far as it goes 
we cannot confirm from it Bentham’s identification. It has the stigma ‘‘ capitato- 
4-lobo”’ as described by Bartling, and in this respect differs from all our other 
specimens. The species much resembles E. paniculata and is easily confused with 
it. But it may generally be distinguished by its broader sepals (more calycine 
flowers), its longer, more slender style, its pale obconic stigma, and minutely hairy 
ovary. Both havea like floccose indumentum. ‘The stigma of this is somewhat 
variable; usually it is obconic, quite hollow or depressed within ; sometimes it 
is solid and has a truncate top. But it is always small, as compared, for 
instance, with EH. peltata, with which it has some affinity. 
437. E. rhodantha (Guthrie & Bolus) ; erect ; branches ascending, 
somewhat straight, with numerous short floriferous subdecurved 
branchlets of from 1-11 in. long, pubescent with simple hairs ; 
leaves erect or subspreading, crowded, imbricate, linear-trigonous, 
subobtuse, glabrous, 13—2 lin. long; flowers somewhat copious, sub- 
ealycine; pedicels softly floccose with minutely plumose hairs, 1} lin. 
