An evergreen shrub 2 to 5 ft high, of stiff close growth; young shoots at first white with a close down, grey the second season. Leaves obovate to oblanceolate, tapered abruptly at the apex to a short mucro, always more or less wedge-shaped at the base, 2 to 4 in. long, 3⁄4 to 11⁄2 in. wide, dark green and glabrous above, covered beneath with a glaucous white, thin, scurfy down; stalk 1⁄4 to 1⁄2 in. long. Flowers opening in May, six to eight in a terminal truss; flower-stalk 1⁄2 to 1 in. long, white with down. Calyx very variable in the size and shape of the lobes; 1⁄/6 to 1 in. long and awl-shaped to roundish ovate, approximating the corolla in colour. Corolla fleshy, elongated-bell-shaped, 13⁄4 in. long, scarcely so wide; the colour is variable in different plants and may be of various shades of orange, often more or less suffused with pink. Stamens ten, scarcely as long as the corolla, slightly downy at the base. Ovary clothed with white down and fascicled hairs; style glabrous. Bot. Mag., t. 8815. (s. Neriiflorum ss. Sanguineum)
Native of the Tali range, Yunnan, China; discovered by Forrest in 1906 and introduced to cultivation by him. His F.11597, with deep orange flowers, is perhaps the most desirable form, but others are very attractive. Forrest describes some shades as 'creamy rose' or 'yellowish rose'. In its best forms this is one of the most attractive members of the series and quite hardy.
subsp. apodectum (Balf. f. & W. W. Sm.) Cowan R. apodectum Balf. f. & W. W. Sm.; R. jangtzowense Balf. f. & Forr.; R. liratum Balf. f. & Forr. – Distinguished from typical R. dichroanthum by its very leathery, round-based leaves. It appears to be confined to the Shweli-Salween divide, and was introduced by Forrest in 1913. Bot. Mag., t. 9014.
subsp. herpesticum (Balf. f. & Ward) Cowan R. herpesticum Balf. f. & Ward – -This is easily distinguished from typical R. dichroanthum and subsp. apodectum by the flower-stalks, calyces, and ovaries being furnished with glands. Also the leaves are more like those of R. sanguineum in size and shape, being oblanceolate or oblong-elliptic, up to 3 in. long and 11⁄8 in wide. The shoots are sometimes bristly. This subspecies was discovered by Kingdon Ward in July 1914 in upper Burma, on the northern slope of Imaw Bum, a mountain just over 13,000 ft high on the divide between the Salween and the upper Irrawaddy (Nmai Hka), forming tanglements about 1 ft high. It is fairly widely distributed in the borderland between Yunnan and Burma, and was sent many times by Forrest, who introduced it.
subsp. scyphocalyx (Balf. f. & Forr.) Cowan R. scyphocalyx Balf. f. & Forr. – This does not really differ from the subsp. herpesticum in any character of genuine taxonomic importance, but typically it is four or five times as tall and the leaves are larger and thinner. But intermediate forms occur. It was discovered by Forrest in 1919 about twenty miles to the west of the type-locality of subsp. herpesticum and has the same distribution. He sent seeds many times from 1919 onward and it is also in cultivation from Farrer 1024, this too collected in 1919. The flowering-time of both subspecies is mostly May. The colour range of wild plants is extraordinarily wide – rose-orange, yellowish crimson, coppery yellow, fiery bronze, flame, orange, cinnabar-scarlet, very dark yellowish crimson, etc. Unfortunately, cultivated plants often bear flowers the colour of marmalade or mustard-pickle.
subsp. septentrionale Cowan R. scyphocalyx var. septentrionale Tagg MS – Leaves as in subsp. herpesticum, but ovary without glands. Flowers yellow or yellow flushed rose. Introduced by Forrest. He found it originally about 50 miles to the north of the type-locality of subsp. scyphocalyx, whence the epithet septentrionale.
From the Supplement (Vol. V)
Dr Cowan's treatment of this species is accepted in the Edinburgh revision, except that subsp. herpesticum is submerged in subsp. scyphocalyx (Rev. 2, p. 394).