Rosa longicuspis Bertol.

A large evergreen or semi-evergreen scrambling shrub, or a tall climber with a stout trunk, its branches armed with short, curved prickles. Leaves up to 8 in. long, dark green, they and the shoots reddish when young. Leaflets five or seven, of leathery texture, 2 to 4 in. long, elliptic or narrowly ovate, with slender acuminate tips (but more shortly acuminate on the flowering branchlets), rounded at the base, edged with numerous small teeth, glabrous on both sides except for occasional down on the midrib beneath, venation prominent on the undersurface. Stipules narrow, finely toothed at the edge. Flowers white, about 2 in. wide, up to fifteen or so in a lax panicle; flower-buds narrowly ovoid. Pedicels 1 in. or slightly more long, they and the receptacles glandular and often hairy, sometimes hairy and eglandular. Sepals hairy and glandular on the back, about 1 in. long, often slightly expanded at the apex, with a few lateral appendages. Petals densely silky on the back. Styles exserted, united into a hairy column. Fruits globular or broadly ellipsoid, up to 3⁄4 in. or slightly more long; sepals deciduous from the fully ripe fruit.

R. longicuspis was described in 1861 from a specimen collected by Hooker and Thomson in the Khasi Hills of Assam, and occurs in all the lower ranges of north-eastern India (including the outer Himalayas), probably extending into Burma and with close relatives in China. It is allied to R. brunonii, differing in the leathery, more acuminate leaflets almost glabrous beneath, fewer-flowered inflorescence, and the silkiness of the underside of the petals. But it is possible that intermediates occur in the eastern Himalaya.

There are no specimens in the Kew Herbarium of cultivated R. longicuspis, except for one from Headfort, Eire, apparently from a private introduction. But seeds were sent by Kingdon Ward from the Naga Hills of N.E. India under his KW 7740, collected in 1927, and plants may have been raised from these in private gardens. He may also have collected seed from the plants he found in the Assam Himalaya near Shergaon in 1935. A rose has been distributed as R. longicuspis, and is described and figured under that name in Graham Thomas, Climbing Roses, p. 35 and fig. 2. The author now accepts that it is not that species; as he remarks, it is near to R. mulliganii, for which see under R. rubus.

From the Supplement (Vol. V)

This species, in its broad sense, was reintroduced by Roy Lancaster from the Kunming area of Yunnan in 1980, and two plants from this seed-collection (L.694) have flowered in the University Botanic Garden, Cambridge (1986). The leaflets are up to nine in number, under 2 in. long, elliptic to broadly so, shortly acuminate, glabrous and glossy on both sides, with inconspicuous venation. The flowers are white, about 2 in. wide, opening in July, up to sixteen in each cluster, on slender, glandular pedicels 11⁄2 to 13⁄4 in. long. On one plant the petals become stained with red before falling. Fruits 1⁄2 in. or slightly more wide, globose, broad-ellipsoid or slightly obovoid, red, ripening in November, crowned by the persistent stylar column.

These plants disagree with typical R. longicuspis from north-east India in their smaller, less tapered leaflets, shorter sepals and in having the petals glabrous or almost so on the back. They represent very well a rose collected many times by Forrest and others in various parts of Yunnan; the most recent flowering specimens in the Kew Herbarium were collected by the Sino-British Expedition to the Cangshan (Tali Range) in 1981. Such specimens have usually been identified as R. longicuspis, but also, by some authorities, as R. lucens (mentioned on page 108) or as R. yunnanensis (Crépin) Boulenger, founded on a specimen collected by the French missionary Delavay in the region north of the Cangshan. Until the Chinese Synstylae have been further studied, it seems best to accept the Cambridge rose as R. longicuspis and to include in this species both R. lucens and R. yunnanensis.