habit renders them valuable for covering the walls of 
stoves. In the present instance the plant was probably 
not sufficiently advanced to assume this habit, as no root- 
lets appeared on the specimen sent for figuring. 
I am indebted to Mr. Bull for the plant here figured, 
_ which flowered in his establishment in June, 1883. Unfor- aie 
tunately all record is lost of its origin; it, however, so far 
resembles some specimens of a Guiana species, that I think 
it may be a native of that country. 
Descr. Quite glabrous. Stem very stout, as thick as 
the little finger, obtusely four-angled, pale brown mottled 
with darker streaks. Leaves very large, full-grown nearly 
a foot long, broadly elliptic-ovate, subacute at both ends, 
crenulate, thick and almost fleshy, above bullate between 
the nerves, green mottled with light grey, beneath of a 
eK 
ight vinous purple with very prominent midrib and nerves; 
petiole very stout, two to four inches long, terete, chan- 
nelled down the front. Flowers fascicled in the leaf-axils ; 
peduncles one to four inches long, stout, ascending, rose- 
coloured spotted with grey. Sepals one to one and a 
quarter inch long, foliaceous, ovate, subacute, base cordate, 
strongly nerved, rose-purple. Corolla one and a half inch 
long, declinate, pale yellow slightly suffused with pink; 
tube half an inch in diameter; lobes rounded, concave, 
margins fimbriate. Stamens included, filaments hardly | 
united below, but disposed in two pairs, flattened, very 
strongly twisted; anthers linear-sagittate, obtuse. Disk- 
gland solitary, dorsal, erect, concave. Ovary quite glabrous ; 
style stout, stigma discoid, included.—J. D. H. 
Fig. 1, Base of corolla-tube and stamen ; 2 and 3, top of filaments and anthers; — 
4, ovary and disk :—aJ/ enlarged. 
