258 ERICACEE (Guthrie & Bolus). [ Hrica. 
anthers oblong, obtuse, ascending at the base, + lin. long, with the 
pore less than } the length of the cell, or longitudinally semi- 
elliptical, tapering to base and apex, 2 lin. long, with the pore 2 the 
length of the cell; cells deeply parted, crested; crests subovate, 
acute in outline, incised, naked or ciliolate, about 1 the length of 
the cell; style included, angular; stigma 4-fid, cruciform; ovary 
pallid, lobed, glabrous. Benth, in DC. Prodr. vii. 656. E. gnap- 
halodes, Berg. Descr. Pl. Cap. 119; Wendl. Eric. Ic. fase. 19, 
109, #.42. HE. gnaphaliiflora, Salisb. in Trans. Linn. Soc. vi. 337. 
E. calycina, var. minor, Andr. Heathery, t. 108, and Col. Heaths, t. 
156. E. paniculata, Lodd. Bot. Cab. t. 1194? not of Linn. £. 
tetrastigmata, Bolus in Journ. Linn, Soc. xxiv. 178. E. lilacina, 
Klotzsch ex Benth. l.c. 657. 
Coast REGION, ascending to 2500 ft.: Paarl Div.; French Hoek, Bolus, 
6984! Cape Div. ; Cape Flats, near Rondebosch, &c., Burchell, 170! Zeyher, 1111! 
Bolus, 3747! Guthrie, 88! 381! Steen Berg, Wolley Dod, 3481! Muizen Berg, 
Bolus, 7300! Stellenbosch Div.; Lowrys Pass, Schlechter, 5897! Caledon Div. ; 
Houw Hoek, Bolus, 5451! mountains near Hemel-en-Aarde, Zeyher, 3276! 
Zwart Berg, Schlechter, 5577! Bredasdorp Div.; near Potte Berg, Mund, 28! 
Riversdale Div. ; Garcias Pass, Burchell, 6943 ! 
This species has been much confused, and is often marked “ E. articularis”’ in 
herbaria. The 4-fid cruciform stigma is generally a diagnostic character, and 
was noted by Linnzeus. Drége’s 7743, while it exhibits this character, has flowers 
much more like those of E. chlamydiflora. We regard it as a hybrid, and merely 
note it. We do not cite Thunberg’s E. gnaphalodes, Diss. Erica, 45, although 
Rach’s determination of this as £. articularis, L., is probably wrong. Our 
plants agree well with Wendland’s figure cited above. There are two forms, one 
represented by the last-named figure, slender and virgate; and a_ stouter 
and more branching one, with more numerous and crowded flowers, resembling 
some species of § Trigemma. But the two forms are by no means distinct, 
and are connected by intermediates. 
390. E. articularis (Linn. Mant. 65); mostly erect, 1-12 ft. 
high; branehes usually virgate, not numerous, slender, glabrous, 
rarely puberulous; leaves 3-nate, erect, mostly incurved and 
adpressed, generally about equal to the internodes, linear or narrow- 
lanceolate, keeled or suleate, mostly 14-3 lin. long; flowers 3-nate, 
more or less clustered, often so crowded on short branchlets as to 
form a dense long pseudo-spike; pedicels slender, 1-12 lin. long; 
bracts 2, subremote or subapproximate, linear-subulate to oblanceo- 
late, third subbasal, longer or wanting ; sepals obovate, keel-tipped, 
acute or obtuse, concave, spreading at full maturity, generally reach- 
ing from 3—} the length of the corolla (including the limb), 3-14 
lin. long; corolla suburceolate, only slightly contracted at the 
throat, sometimes subtetragonous, 11-2 lin. long; segments erect or 
spreading, subovate, deep flesh-eolour (Andrews), concolorous or 
dark-coloured, 4—} the tube in length; anthers sublateral, almost 
basifixed ; cells deeply parted, semi-ovate or elliptical, 12 lin. long, 
crested ; pore about 3 the length of the cell; crests subovate in out- 
line, more or less deeply incised ; style included; stigma capitate ; 
ovary lobed. Benth, in DC. Prodr. vii, 653; E. teretiuscula, 
