A slow-growing, creeping, evergreen shrub a few inches high on the level, but capable, according to Forrest, of climbing the faces of moist rocks 3 to 5 ft high, attaching itself, like ivy, by roots from the stems. Leaves broadly obovate to nearly orbicular, rounded at the apex, usually more or less tapered at the base, 1⁄2 to 11⁄2 in. long, 1⁄3 to 7⁄8 in. wide, dark glossy green and with conspicuously grooved veins above, pale beneath or more or less stained with purple; stalk 1⁄8 to 1⁄2 in. long, reddish, slightly downy and glandular. Flowers usually solitary, but sometimes in pairs or threes borne in April or May; pedicels up to 1 in. long, glandular. Calyx small, shallow, fringed with glands. Corolla narrowly bell-shaped, 13⁄8 in. long and wide, deep crimson, with five rounded deeply notched lobes. Stamens ten, white, glabrous; anthers dark brown. Ovary conical, clad with pale glands; style glabrous, longer than the stamens. Bot. Mag., t. 9186. (s. Neriiflorum ss. Forrestii)
Native of N.W. Yunnan, upper Burma, and also of S.E. Tibet, where it extends to the region of the Tsangpo bend and some way farther west; the altitudinal range is 10,000 to 14,000 ft. Forrest discovered this species in the midsummer of 1905 in Yunnan on the Mekong-Salween divide, but a few weeks later he narrowly escaped death at the hands of the rebel lamas and all the material was lost, with the rest of his collections, except for one small fragment sent home earlier, from which Diels described the species in 1912. More material was collected by Forrest in 1914 and 1918 in the same area, but in most of these specimens the undersides of the leaves were green or glaucous-green beneath, whereas in the type they are purple beneath, as they are in seedlings. The greenback forms were given specific status as R. repens, and it is by this name that the species was known for many years. But Dr Stapf of Kew considered that the peculiar coloration of the foliage in typical R. forrestii, even if fixed and regular, was too slight a distinction, and reduced R. repens to a synonym of R. forrestii. This judgement was accepted in previous editions, as it is here, but it should be noted that Cowan and Davidian, in the article cited below, prefer to maintain R. repens as a variety of R. forrestii.
Later collections have shown that R. forrestii, in its typical small-leaved, red-flowered, creeping form, is only one constituent of a puzzling complex of plants, all essentially alike but differing in habit, size of leaf, in the colour of the flowers, and the number in each inflorescence. Great variation is found even in a single wild locality – so much so that some batches of wild seed have produced plants of diverse character, some or all of which differ from the corresponding herbarium specimen. Cowan and Davidian, in their taxonomic treatment of R. forrestii and its immediate allies, recognise two varieties of R. forrestii (one of them var. repens), and a second species – R. chamaethomsonii – which is, however, linked to R. forrestii by intermediates (R.Y.B. 1951-2, pp. 66-71). The diagnostic characters and synonyms are as follows:
var. tumescens Cowan & Davidian R. repens var. chamaedoxa, nom. inedit., in part – Plant dome-shaped in the centre with the outer branches creeping. Leaves somewhat larger than in the type. Described from a cultivated plant raised from KW 5846. This variety is difficult to distinguish from R. chamaethomsonii var. chamaethauma (see below).
R. chamaethomsonii (Tagg & Forr.) Cowan & Davidian R. repens var. chamaethomsonii Tagg & Forr. – A dwarf but not creeping shrub, up to 3 ft high. Leaves oblong-obovate to obovate, up to 3 in. or even slightly more long, sometimes slightly hairy beneath. Flowers crimson up to four or even five in each truss. Calyx larger than in R. forrestii, fleshy. Described from specimens collected by Forrest and Rock in N.W. Yunnan and bordering parts of Tibet.
var. chamaethauma (Tagg) Cowan & Davidian R. repens var. chamaethauma Tagg; R. repens var. chamaedoron Tagg & Forr.; R. repens var. chamaedoxa, nom. inedit., in part – Leaves smaller than in the type of R. chamaethomsonii. Described from KW 5847, collected on the Doshong La at the eastern end of the Himalaya; the synonymous name R. repens var. chamaedoron is founded on specimens collected by Forrest and by Rock in the area where typical R. chamaethomsonii was first found, and links this species with R. forrestii through the latter's var. tumescens (Cowan and Davidian, op. cit., p. 71).
The plants of R. forrestii and R. chamaethomsonii in the Tower Court collection (most of which are now in the Species Collection in Windsor Great Park) were discussed by Roza Stevenson in R.Y.B. 1951-2, pp. 60-5. The following survey of the main introductions of R. forrestii and R. chamaethomsonii is based partly on this and on the note by Cowan and Davidian already referred to, partly on collectors' field notes, the writings of Kingdon Ward, and other sources.
F.13259 (1914) – The original introduction of R. forrestii (as R. repens). Plants were distributed by J. C. Williams from Caerhays.
F.19515 (1921) – Seed collected on the Londre La, Mekong-Salween divide. Some plants at Edinburgh have the leaves purple beneath as in typical R. forrestii, while in others they are green beneath (Cowan and Davidian, op. cit., p. 68).
F.21768 (1922) – From the divide between the Salween and the Kiuchiang (upper Irrawaddy). The herbarium specimen under this number is R. chamaethomsonii var. chamaethauma, but some of the plants raised from the seeds agree better with R. forrestii.
KW 5845 (1924) – This is Kingdon Ward's “Scarlet Runner”, from the Doshong La, at the eastern end of the Himalaya. The wild plants were prostrate, rising no more than 2 in. above the ground, and grew on bare gneissic rocks on the sunny side of the slope; flowers scarlet, solitary. The seed under this number and KW 5846 were distributed as R. repens var. chamaedoxa, an unpublished name. The plants at Tower Court raised from KW 5845 differed markedly from the wild plants described by Kingdon Ward in being bushy and up to 19 in. high, with carmine flowers borne three or four to each truss, and with large leaves. These plants, and others at Edinburgh from the same batch of seed, are referred by Cowan and Davidian to R. chamaethomsonii var. chamaethauma. The discrepancy between these cultivated plants and Kingdon Ward's description is so great as to suggest that at least part of the seed under KW 5845 did not come from the true “Scarlet Runner”. The plants were buried in snow when he returned to the Doshong La for the seed-harvest and it was only with great difficulty that he managed to get any seed at all (Gard. Chron., Vol. 78 (1925), p. 330).
KW 5846 (1924) – Also from the Doshong La. “Scarlet Pimpernel”, as Kingdon Ward dubbed it, grew on sheltered ledges with “Plum Warner” (R. campylogynum). 'At first sight it looked like a darker edition of Scarlet Runner, but on closer inspection it was seen to be a bigger plant, with larger leaves and darker flowers, borne two or three in a truss instead of singly.' One plant at Tower Court from this number was identified as R. chamaethomsonii var. chamaethauma, while a plant at Edinburgh from the same batch is the type of R. forrestii var. tumescens (see above). A plant from KW 5846 received an Award of Merit when shown from Bodnant in 1932.
KW 5847 (1924) – This is Kingdon Ward's “Carmelita” and the third of the trio from the Doshong La. 'It is bigger again, with still larger leaves, and flowers of luminous carmine, in threes. It grows socially, in foot-deep tangles, and is not really a creeping plant at all . . .' (Riddle of Tsangpo Gorges, p. 102). The specimen under this number is the type of R. chamaethomsonii var. chamaethauma, but it is not certain whether any plant from the seed of KW 5847 is in cultivation.
KW 6832 (1926) – From the Seinghku valley, in northernmost Burma, near the borders with Assam and Tibet, at 11,000 ft. It grew 'plastered on rocks and steep talus in very exposed situations', and was considered by Kingdon Ward to be the same as his “Scarlet Runner” (see above), but 'not such a brave sight'. Seed was distributed as R. repens and a plant raised from it at Tower Court received a First Class Certificate (as R. repens) in 1933. This plant has produced large crops of beautifully coloured flowers every year and is depicted in R.Y.B. 1951-2, fig. 56. It is, however, not a typical example of R. repens (R. forrestii as here understood), since, although low growing, it is not creeping, the leaves are larger, and the flowers are up to three in each truss. It is referred by Cowan and Davidian to R. chamaethomsonii var. chamaethauma.
Rock 59174 (1923) – From Kenichunpu on the divide between the Salween and the upper Irrawaddy (Kiuchiang) at 13,000 ft, making a shrub 1 to 2 ft high. Dr Rock's fruiting specimen was referred to R. repens var. chamaethomsonii (now treated as a separate species), but the plants raised from seed of Rock 59174 at Tower Court proved to be intermediate between R. forrestii var. tumescens and R. chamaethomsonii var. chamaethauma, and one of them received an Award of Merit under the former name in 1957. On the other hand, a plant at Nymans in Sussex, from the same seed-number, is bushy and about 21⁄2 ft high. It is interesting to note that seed collected by Dr Rock in the same area during his expedition for the American Rhododendron Society in 1948 (Rock 92) has produced plants varying from upright to prostrate and differing, too, in size of leaf and in number of flowers per inflorescence(Qtly Bull. Amer. Rhod. Soc., Vol. 11, p. 148).
R. forrestii and R. chamaethomsonii are both shy-flowering in some forms, but reliable clones are available in commerce. Both need a cool, moist soil and a position where they are exposed to the open sky but protected from the sun during the hottest part of the day. Their flowering time is April or May.
R. forrestii is the parent of several low-growing, mostly red-flowered hybrids, of which the most famous is the Elizabeth grex.
From the Supplement (Vol. V)
subsp. papillatum Chamberlain – The leading characters of this, to distinguish it from the typical subspecies, are the glaucous-papillate leaf-undersides, with numerous stalked glands, and the relatively narrower leaves, twice to almost three times as long as wide, against mostly less than twice as long as wide in subspecies forrestii. The type of this new subspecies is KW 5845, for which see page 666. The type of var. tumescens (see pages 665-6) is intermediate between this subspecies and R. chamaethomsonii (Rev. 2, pp. 407-8).
R. chamaethomsonii – The var. chamaedoron, given as a synonym of R. chamaethomsonii var. chamaethauma, is recognised in the Edinburgh revision. The type is F.21768, for which see page 667.