Erica.] ERICACEE (Guthrie & Bolus). 297 
long) ; anthers either as in the type or crested-aristate; awns subulate, acumi- 
nate, slightly incised. 
Var. yy, laxa (Bolus); as in var. 8, but the anthers cristate; crests narrow- 
ovate, crenulate, about 2 the length of the cell. EF. lava, Andr. Heathery, t. 73, 
and Andr. Col. Heaths, t. 181; Benth. in DC. Prodr. vii. 687. E. rigescens, 
Baril. in Linnea, vii. 635, fide Benth. E. rigens, Benth. lc. 
SourH Arkica: without locality, Niven, Herb. Salisbury! and cultivated 
specimens ! Var. y: Niven! and cultivated specimen ! 
Coast ReGion: Clanwilliam Div.; Cederberg Range, 2600 ft., Marloth, 
2690! Leipoldt! Tulbagh Div. ; Winterhoek Mountain, Bolus, 6351! moun- 
tains near Saron, Schlechter, 10679! Cape Div.; Table Mountain, Pappe! 
Devils Peak, 1100-1500 ft., Guthrie, 1217! Bolus, 4470! 4590! Paarl Div. ; 
Sneeuw Kop, Marloth! French Hoek, Bolus, 6985! Caledon Div. ; Zwart Berg, 
Zeyher, 3278! Var. 8: Tulbagh Div.; Mitchells Pass, 1000 ft., Bolus, 5187! 
mountains near Tulbagh Waterfall, 800 ft., Bolus, 5464! 
The larger-flowered form of this has some resemblance to E. calycina, but the 
flowers are smaller. Our var. 8 is more lax in habit and is fewer-flowered, but 
otherwise seems to agree with the type. Niven’s specimen in Herb. Kew., 
marked E. lawa by Bentham, is the only one we have seen in which the anther- 
crest agrees with Andrews’ figure, and we have distinguished it accordingly. 
Taken as a whole the species differs from all the rest in the section by its corolla, 
which is not so uniformly spreading upwards, but has a tendency to become 
cyathiform or even globose-urceolate with a slight contraction at the throat; and 
_in so far it approaches to § Trigemma. 
_ 450. E. mucronata (Andr. Col. Heaths, t. 186, and Heathery, 
t. 225); erect, 11 ft. high ; branches erect or spreading, downy ; leaves 
spreading, incurved, narrow-lanceolate, setaceous-acuminate, mucro- 
nate, keeled, glabrous, 4-10 lin. long; flowers 3-nate (or 3-6-nate, 
Andrews) on short arrested branchlets, subcorolline ; pedicels downy, 
decurved, 3-4 lin. long; bracts, 2 approximate imbricating the 
sepals, 1 remote, lanceolate, setaceous-acuminate, scarious, keeled ; 
sepals like the bracts, but larger, including the bristle-like point, 
2} lin. long; corolla obconic-campanulate, rosy, 23 lin. long (in 
MacNab’s 233, a cult. sp. ?); segments slightly spreading, ovate, 
subacute, about 3 the length of the tube; filaments rather broad at 
the base, tapering upwards, a little longer than the anther; anthers 
included, lateral, narrow-oblong, subacute, minutely scaberulous, 
dark-coloured, 1 lin. Jong, 44 times longer than its greatest width, 
-muticous; pore small, about 4} times shorter than the cell and near 
its apex ; style slender, well-exserted ; stigma small, obconic ; ovary 
glabrous. Benth. in DC. Prodr. vii. 683. . eriopus, Benth. 
le. 650, 
SourH Arrica: without locality, Masson/ and cultivated specimens! 
Coast Region: Riversdale Divy.; mountains near the Kaffirkuils River, 
Niven, 103! summit of Kampsche Berg, Burchell, 7117! 
This species was placed in § Melastemon by Bentham and next to £. tetra- 
thecoides, which, in the external appearance of its flowers, it somewhat resembles. 
But it has not the anther of that section; and also differs in leaves, bracts and 
sepals. It also resembles in some respects H. acuta, but has larger flowers and a 
different corolla and anthers. Burchell’s specimens are not fully developed, and 
on this Bentham established Z. eriopus and placed it in § Platyspora. But 
Niven’s specimens, 103, afford mature seeds which are ellipsoidal and not at all 
compressed as in that section. 
