Erica.] ERICACEX (Guthrie & Bolus). 309 
siender ; stigma capitellate ; ovary glabrous. Benth. in DC. Prodr. 
vii. 649. KE. complanata, Nois. ex Spreng. Syst. Veg. ii. 196? £. 
crassifolia, Benth. l.c. 685, not of Andr. 
South AFRica: without locality, Niven / and cultivated specimens ! 
Coast ReaGion : Caledon Div., 3000-4000 ft.; southern slopes of the Houw 
Hoek Mountains, Bolus, 5452! Schlechter, 5453! mountains near Vogel Gat, 
Klein River, Schlechter, 9548 ! 
It is probable that Salisbury’s type is Niven’s specimen in Herb. Kew. This 
we have dissected, as also all the specimens cited above, besides one in Herb. 
MacNab at Kew, and one in the Berlin Herbarium, both marked E. crassi- 
folia, Andr., and are satisfied that all are identical. Bentham describes this 
species under the name of E. Corydalis, and also under E. crassifolia, with 
which it has a superficial resemblance: an easy mistake in so large and 
intricate a genus, from which we ourselves can hardly expect to have escaped. 
The species is very well marked, and quite different, upon dissection, from the 
last-named; the hemispherical “tube” is differentiated from the limb by an 
interior thin circular ridge and longitudinal nerves, and the anther is very 
different. We cite E. complanata, Nois., upon Bentham’s authority, and can 
only say that the description agrees, as far as it goes. ‘Ihe species connects this 
section both with § Pachysa and § Trigemma. 
469. E. flavisepala (Guthrie & Bolus) ; erect, slender, a foot or 
more high; branches subfiliform, glabrous or minutely puberulous 
at the apex; leaves 3-nate, erect, imbricate, linear or linear-lanceo- 
late, acute, the younger mucronate and ciliate, the floral somewhat 
dilated, 2-3 lin. long ; flowers 3-nate, clustered at the ends of short 
branchlets, subcorolline ; pedicels glabrous, red, 2-3 lin. long ; bracts 
remote, lanceolate, setaceous-acuminaté, concave, scarious, ciliate, 
the hairs often barbellate or forked, pale yellow, about 3 lin. long ; 
sepals like the bracts or sometimes a little broader and shorter ; 
corolla ovoid-urceolate, but little contracted at the throat, dry, 
glabrous, red, 4—44 lin. long; segments slightly spreading or erect, 
somewhat concave, ovate, acute, subcordate, about 4 of the tube in 
length; filaments capillary; anthers subcuneate-linear, or narrow- 
oblong; cells deeply parted, very slightly bilobed but scarcely 
prognathous at the base, about % lin. long, aristate; pore scarcely 
3 the length of the cell; awns inserted well above the base of the 
cell, subulate, projecting backwards, about 3 the length of the cell ; 
style straight, red; stigma capitellate, small; ovary subturbinate, 
shortly stipitate, glabrous. 
CENTRAL Region: Ceres Div.; sent with E. Uhunbergit and said to 
grow with it on the Cold Bokkeveld, Herb. Bolus, 6893! and Cape Govt. 
Herb, ! 
This species, of which we have seen only one (but a good) specimen, is a very 
well-marked one, but difficult to place satisfactorily in any of the sections. In 
general aspect it is strikingly similar to E. Thunbergii ; but has not the peculiar 
globose corolla-tube of either of the other species of this section. From 
§ Ceramus it recedes by its large loose bracts and sepals, by its corolla but little 
constricted at the throat, and in aspect is unlike any other species described 
there. Krom § Trigemma it differs by its lax and bright coloured bracts and 
sepals and its shortly stipitate ovary. 
