A deciduous bush up to 10 ft high; shoots four-angled and, like the under-surface of the leaves, covered with a reddish-brown down. Leaves up to 9 or 12 in. long, by 3 in. wide; lanceolate, slender-pointed, tapered at the base, toothed; stalk 1⁄4 to 1⁄2 in. long. Inflorescence cylindrical, 2 to 8 in. long, about 1 in. wide, made up of shortly stalked, few-flowered, closely packed clusters; it is terminal, supplemented by others from the axils of the top pairs of leaves; all the stalks very woolly. Flowers fragrant, described by Forrest as 'grey and maroon', 'reddish maroon', 'soft lavender rose', 'pale mauve, almost white'; from which it would seem that, like B. davidii (variabilis), it has a considerable range of colouring in even a wild state. Corolla 1⁄3 in. long, 1⁄5 in. wide across the lobes, more or less woolly outside; stamens inserted immediately below the mouth. Calyx 1⁄/6 in. long, with four narrowly triangular lobes, more or less downy. Ovary downy, sometimes becoming smooth. Bot. Mag., n.s., t. 93.
Native of S.W. China; discovered by Forrest in 1903 and introduced by him. It has been confused in gardens with B. farreri, with which it has no close kinship; that species flowers in spring from the joints of the previous year's growth, whilst B. forrestii flowers in late summer and autumn at the end of the current season's growth, in the same way as B. davidii. It is not very hardy in the open ground at Kew; and I have not yet seen it in a sufficiently attractive condition to justify a recommendation of it.
The description given above is of the typical form, but in downiness as well as in flower colour it varies a good deal, some forms approaching an almost glabrous state in branchlet, inflorescence, and ovary. The insertion of the stamens close to the mouth of the corolla, and the corolla being about twice as long as wide are (in association) the most distinctive characters. In the example now at Kew, which is figured in Bot. Mag., n.s., t. 93, the lower surfaces of the leaves are covered with a thin, white indumentum and the stems are rounded, only rarely somewhat four-angled. It is grown on a wall of the Temperate House, where it is only cut in severe winters.
B. pterocaulis A. B. Jacks. B. longifolia Gagnep, not H. B. K. – Another ally of B. forrestii, differing in its narrower corolla-tubes, longer inflorescences and larger leaves; introduced by Farrer from Upper Burma. It is in cultivation at Highdown but Sir Frederick Stern recorded, 'The flowers are a poor colour; it was only kept because Bean when staying here said I must keep it as it was the only one he had ever seen.'
From the Supplement (Vol. V)
This was originally described from a specimen collected by Forrest on the eastern side of the Tali range of Yunnan (Cangshan), whence it has been reintroduced by the Sino-British expedition of 1981. Plants have already flowered and match the form of the species portrayed in Botanical Magazine, new series, t.93. In this, as in the new introduction, the flowers are slaty purple on the outside in bud but as they open the tubes become yellowish brown or orange. The species is variable in the indumentum of the lower leaf-surface, from thick and cinnamon-coloured as in the type to a thin coating (the leaf then appearing green beneath).
B. forrestii is allied to B. macrostachya Benth., described from the Himalaya. There is overlap in distribution between the two, B. forrestii ranging as far as Bhutan and B. macrostachya into Yunnan, and there has been some confusion between the two, though they are distinct enough in their typical states, B. forrestii having a short, relatively broad corolla-tube, which in the other species is narrowly cylindrical; they also differ in the indumentum of the calyx, ovary and, usually, the corolla, these being densely tomentose in B. macrostachya but much less so in B. forrestii. B. pterocaulis (longifolia), mentioned under B. forrestii, is included in that species by Dr Leeuwenberg, though in its narrow corolla-tubes it suggests B. macrostachya.
Another introduction by the Sino-British expedition (SBEC 360) is of uncertain identity. Of remarkable vigour, it has already attained a height of some 18 feet on a south wall by Roy Lancaster's garden (1986). The flowers are in dense, narrow panicles, white flushed with pink in bud, opening pale, dull red, making this buddleia of little ornamental value.