A vigorous shrub of spreading or semi-scandent habit, 8 to 15 ft high; young shoots hairy or almost glabrous, armed with hooked spines, often purplish. Leaves 4 to 9 in. long, composed usually of five leaflets, reduced to three under the inflorescence; rachis prickly, slightly downy arid glandular. Leaflets elliptic-ovate to oblong-obovate, acute or shortly acuminate, 11⁄2 to 31⁄2 in. long (longer on sterile shoots), simply and sometimes deeply serrate, greyish and hairy beneath, rarely quite glabrous, sometimes strongly tinged with purple when young. Flowers white, about 11⁄2 in. across, with broad, overlapping petals, borne in late June, July or early August, in dense clusters. Pedicels short (barely 1 in. long) and moderately thick, they and the receptacle clad with hairs and stalked glands. Sepals not much longer than the ovoid flower-bud, downy and glandular on the back, entire or with a few slender appendages. Styles united in a shortly exserted downy column. Fruits globular, 3⁄8 to 1⁄2 in. wide, red.
Native of western and central China; discovered by Henry about 1886, but the type of the species was collected in Kweichow. Wilson introduced it from Hupeh under two numbers: W.473c, with leaflets velvety beneath, as in the type of R. rubus, and W.666a (R. ernestii f. nudescens Stapf), in which they are nearly glabrous. It was also raised from seeds collected by Farrer in Kansu (Farrer 786) but his no. 291, distributed as R. rubus, is R. filipes. Although hardy and vigorous, with attractively tinted young foliage, R. rubus has always been rare in gardens.
R. cerasocarpa Rolfe is very near to R. rubus and probably no more than a glabrous form of it.
R. mulliganii Boulenger – Near to R. rubus, but the leaflets sometimes seven and the flowers larger on more slender pedicels up to 11⁄2 in. long, in a laxer inflorescence. It was raised in the R.H.S. Garden at Wisley from seeds collected by Forrest during his 1917-9 expedition to Yunnan and was named by Boulenger in honour of Brian Mulligan, then Assistant to the Director at Wisley, who sent him specimens for identification.
A rose distributed as R. longicuspis is near to R. mulliganii and possibly from the same batch of seed.