A tree to about 20 ft high in the wild, or sometimes a shrub; young growths at first sparsely hairy, becoming light brown or grey-brown, furnished with a few large, pale lenticels; winter buds ovoid, 1⁄4 to 3⁄8 in. long, reddish brown, glabrous except for brown hairs at the tip. Leaves pinnate, with eight to eleven pairs of closely set leaflets; rachis deeply grooved, with the same indumentum beneath as the midribs of the leaflets. Lateral leaflets oblong, mostly 7⁄8 to 11⁄4 in. long, the middle pairs the longest, obtuse and finely aristate at the apex, rounded to truncate at the base, finely and closely serrated in the upper half, dark green, glabrous and finely reticulate above, glabrous beneath on the blade, but the midrib at first densely coated with a mixture of brown, spreading hairs and shorter more appressed white ones and retaining some hairs mostly of the latter kind, until autumn. Stipules deltoid, laciniated, seen only on strong branches. Flowers white, opening late May, about 3⁄8 in. wide. Inflorescence 4 to 5 in. wide, branched in the upper part, rusty-hairy at flowering time, almost glabrous by autumn, sparsely lenticellate. Receptacle soon glabrous. Styles four or five. Fruits at first green tinged with purple, pure white when mature except for a pink flush at the tip, about 3⁄8 in. wide.
Native of the Himalaya from the Simla area eastward and of southeast Tibet (Pome, Kongbo), with related but untypical forms in north Burma and Yunnan. The name Pyrus ursina was given in the catalogue to a set of specimens in the Wallich Herbarium, and was first validated by George Don in his General History of Dichlamydeous Plants in 1832. His description is probably based on the specimen collected by Wallich himself in Nepal, which was in unripe fruit, whence no doubt Don’s innaccurate statement that the fruits are red; the other specimens in the set were collected in Kumaon. It was apparently not introduced to Britain until Col. Donald Lowndes collected seeds in Nepal in 1950, from which plants were raised by Messrs Hillier and propagated by grafting (distributed as Sorbus sp. Lowndes). The above description is made from these trees, which are quite typical, but some wild plants have a denser and more persistent indumentum on the undersides of the leaves and more acute leaflets. More distinct is var. wenzigiana Schneid., in which the leaflets are not only more densely indumented with brown hairs but are unusually narrow and acutely tapered, toothed only near the tip.
Col. Lowndes’ introduction of S. ursina is perfectly hardy and vigorous, though far from being so ornamental as S. vilmorinii or S. hupehensis. The leaves do not colour in autumn, though they do so on some wild trees.
S. himalaica Gabrielian – This interesting and probably very ornamental species was described by the Russian authority on Sorbus Eleanora Gabrielian in 1971, the type being a specimen collected in Nepal by Stainton, Sykes and Williams in 1954. She groups it with S. ursina, from which it differs most obviously in its larger, pink-tinged flowers almost 1⁄2 in. wide, the relatively narrower leaflets and the red fruits, as well as in other more technical characters. The herbarium specimens cited by the author for this species range from just east of Kashmir, through Nepal, Sikkim and Bhutan to southeast Tibet. It is not known to be in cultivation (1979).
S. rehderiana Koehne – This species was described from specimens collected by Wilson in W. Szechwan in the vicinity of Kangting (Tatsien-lu). It is evidently closely allied to S. ursina, but more glabrous, with more pointed somewhat longer leaflets and relatively longer and narrower inflorescences. A remarkably constant character is the thick, dark young wood. It ranges as far south as Yunnan, perhaps to north Burma, and west into Tibet, where it was collected by Ludlow and Sherriff sixty miles north of Lhasa. In var. cupreonitens Hand.-Mazz., described from Yunnan, there are light brown hairs on the midribs of the leaflets, the underside of the rachis and in the inflorescence. It appears to be intermediate between S. rehderiana and S. ursina.
Wilson sent seeds of S. rehderiana in 1908 and again in 1910, but it is not known to be in cultivation. Plants distributed in the 1930s under this name do not even remotely resemble the true species, and appear to be a hybrid of S. aucuparia.