Erica.| ERICACE® (Guthrie & Bolus). 167 
Coast REGION: Uitenhage Div.; without exact locality, Zeyher, 75 (ex 
Sprengel), but in Herb, Berlin without a number, stated to be the type! Var. 8: 
Uitenhage Div. ; between Galgebosch and Melk River, Burchell, 4783, probably, 
and without precise locality, Alevander, 9? Miss Kensit in Herb. Bolus, 6467 | 
Port Elizabeth Div. ; near Port Elizabeth, Cooper, 1480! Bolus, 1675! 
CentRaL Rxeaqion: Somerset Div.; on the Bosch Berg, 3000 ft., MacOwan, 
1231! Div.? Drége, 7799! Var. y: Humansdorp Div.; Clarkson, 900 ft., 
Galpin, 3711! 
The type of this species (of which a portion bas been kindly sent us by Prof. 
Engler, from Berlin), exhibits an anther so unusually different from the speci- 
mens hitherto associated with it, that we have hesitated whether to separate 
them or not. Besides, the leaves upon the specimen are clearly 3-nate, though 
Sprengel describes them as 4-nate. Finally, the specimen standing alone, and 
our material being a fragment merely, we have decided to leave them as they 
were, for the present. Further, our var. y (Galpin, 3711), does not well agree 
with either the type or var. 8, and may hereafter be separated. x 
The species is closely related to E. paniculata, but the inflorescence is usually 
less dense, the flowers larger, the anthers more generally muticous; also to E. 
demissa. é 
223. E. demissa (Klotzsch ex Benth. in DC. Prodr. vii. 666) ; 
erect, much branched, probably under 1 ft. high ; branches puberu- 
lous or shortly floccose, glabrescent ; leaves 3-nate, erect to spread- 
ing, or sometimes recurved, somewhat crowded, linear to oblong, 
subobtuse, sulcate, glabrous, glossy, ciliate, 11-2} lin. long ; flowers 
terminal, or by partial abortion of floriferous branchlets sometimes 
apparently lateral ; pedicels 1-2 lin. long; bracts two approximate, 
or nearly so, third remote and longer, or all subapproximate ; sepals 
lanceolate, acute, keeled, rigid and somewhat scarious, glabrous, 
about % lin. long ; corolla narrow-cyathiform to subovoid, the mouth 
in the latter case contracted, minutely hispidulous, rarely (perhaps 
only in age) glabrous, 1-1} lin. long; anthers exserted, lateral, 
oblong, subobtuse or acute, a little under or over 2 lin. long, 
muticous; pore less than } the length of the cell; style exserted, 
stoutish ; stigma peltate-capitate ; ovary puberulous, rarely sub- 
glabrous. E. globuliflora, Klotesch ex Benth. in DC. Prodr. vii. 
667. 
: nsdorp Div.; near the Gamtoos River, Masson, 58! 
bbe Rrcnie Miiste Burokell 4838! Uitenhage Div. ; Van Stadens Moun- 
tains, Zeyher, 780! 785! 3224! Burchell, 4709! Albany Div.; near Grahams- 
town, Burchell, 3576! Guthrie, 2370! 
There are two forms, or perhaps three, of this species so far as we know it. 
The two first differ chiefly in ae leaves, the one being broader and somewhat 
recurved, the other narrower and suberect. These are connected by Masson's 
specimens. The third form is from Grahamstown; the habit is somewhat 
different, the branches ternate and more slender, the flowers paler in colour, the 
anthers a little longer, and more acute. Guthrie's specimens gathered a the 
same month, 78 years after Burchell’s, almost exactly agree with the — 
In floral characters very much like E. decipiens, but looks different ; w “4 
brauches more rigid; leaves usually longer, stouter, and more glossy : pedice 
twice as long. 
224, E. paniculata (Linn. Sp. Pl. ed. 2, 508, not of Wendl, nor 
Thunb. nor Lodd.); erect, 1-1} ft. high; branches numerous, 
