An evergreen shrub or small tree up to 30 ft, of spreading habit; young shoots hairy, becoming chocolate-brown and glabrous, finally nearly black, often developing spine-tipped spurs. Leaves ovate-lanceolate, often narrowly so, 11⁄2 to 3 in. long and up to 1 in. wide, almost entire, pointed, closely felted beneath at first; stalk 1⁄4 to 1⁄2 in. long, downy. Flowers stalkless, apple-like, 1 to 11⁄4 in. across, white, pink-tinted in bud, produced in umbels of two to four, hawthorn-scented; calyx felted; styles united. Fruit oval, 11⁄2 in. long, 1 in. wide, downy.
Native of Yunnan, China; introduced to France about 1890 and cultivated by Maurice de Vilmorin at Les Barres in 1901. Little was known of this species in Britain until a plant (probably from seed collected by Forrest in 1917-19) flowered at Wisley for the first time in 1938, when 14 ft high. An interesting account of this species by B. O. Mulligan will be found in Journ. R.H.S., Vol. 65, pp. 120-1 (1940). It is of botanical interest but has little value as an ornamental plant.
D. indica Decne. – This species is a native of the E. Himalaya, Assam, Upper Burma, N.W. and S.E. Yunnan, and possibly of Szechwan also. It is closely allied to D. delavayi but distinguished mainly by its leaves, which are ovate or oblong-ovate, mostly 21⁄5 to 31⁄5 in. long, 1 to 11⁄5 in. wide, acute or acuminate at the apex, rounded at the base, woolly-tomentose at first, later glabrous, margins finely toothed, sometimes almost entire; juvenile leaves lobed and toothed, hawthorn-like. See further under D. rufifolia.
D. rufifolia (Lévl.) Rehd. Pyrus rufifolia Lévl.; Malus docynioides Schneid.; Docynia docynioides (Schneid.) Rehd. – This species is founded on a specimen collected in E. Yunnan, China, and to it Rehder referred Schneider 1349 (the type of Malus docynioides) from S. Szechwan, and certain specimens collected by Henry in S.E. Yunnan and by Wilson in Szechwan, the Wilson specimens are Veitch Expedition No. 3493 and Arnold Expedition No. 2998, and seeds may also have been sent under these numbers and distributed as D. delavayi. However, Rehder himself was doubtful about the distinctness of this species from D. delavayi, and material in the Kew Herbarium which would, by his criteria, belong to D. rufifolia has by other workers been assigned to either D. indica or D. delavayi, mostly to the latter. The specimen under Wilson 3493 seems, however, to be intermediate between these two species.