Rhododendron maddenii Hook. f.

---

R brachysiphon Hutch.

This species was described from a specimen collected by R. Cooper in Bhutan near Punakha, from bushes 8 ft high, growing on a steep, dry hillside. It is supposed to differ from R. maddenii in the much smaller corolla, with a tube only half the length of what it is in R. maddenii, and in the downy filaments of the stamens. However, the Japanese expedition to Bhutan in 1967 found plants in the Punakha region some of which agreed with R. maddenii, others with R. brachysiphon, while others again were intermediate between the two species. Dr Hara has accordingly placed R. brachysiphon under R. maddenii in synonymy.The plants cultivated as R. brachysiphon derive mainly or wholly from Kingdon Ward’s 6276, collected ‘blind’ in the Tsangpo gorge, Tibet, at Pemakochung, in 1924. Although the flowers are much smaller than in typical R. maddenii, they are quite pretty, borne just as late in the season, and just as fragrant. Also, the plants are perfectly hardy in a sheltered position south of London.

---

R manipurense Balf. f. & Watt

This species was discovered by Sir George Watt in 1882 in the Naga Hills, Assam, on Mt Japvo and neighbouring mountains, at 8,000 to 10,000 ft. A plant in the Temperate House at Kew, raised from the seeds which he sent, was named R. maddenii var. obtusifolium by Hutchinson in Bot. Mag., t. 8212, but nine years later this rhododendron was given specific rank by Sir Isaac Bayley Balfour, who made his description from the specimens collected by Watt.R. manipurense was reintroduced by Kingdon Ward, who visited Mt Japvo in 1927 and collected seeds of this and three other rhododendrons originally found by Sir George Watt (R. macabeanum, R. elliottii, and R. johnstoneanum). According to his field note under KW 7723, it makes a stout bush on razorback ridges, but becomes a fine tree in the forest, and also occurs as a tree 20 to 30 ft high on the face of the mountain, in company with R. macabeanum. Kingdon Ward also collected seeds in the Adung valley, N.W. upper Burma; in the Di Chu valley, on the borders between Assam, Tibet, and Burma; and in the Mishmi Hills, Assam.The key-character used in Species of Rhododendron to distinguish R. manipurense and R. polyandrum from R. maddenii is: stamen-filaments hairy in the lower part against glabrous in R. maddenii. But these hairs are not a very positive feature, being more in the nature of scattered, thick, watery excrescences, which can be seen on specimens otherwise agreeing with typical R. maddenii. The leaves of R. manipurense are commonly rounded at the apex, against acute in R. maddenii, but there are plants in the Himalaya, agreeing with R. maddenii in most respects, in which the leaves are obtuse. However, R. manipurense, at least in its type-locality, is a more robust plant than R. maddenii. Also, plants from KW 7723 appear to be quite hardy.

---

R polyandrum Hutch

In describing this species, from a specimen collected by Cooper in Bhutan, Hutchinson remarked that it is very near to R. manipurense, differing in its oblong or oblong-lanceolate leaves, rounded to obtuse at both ends, and in having flowers with twenty-five stamens. Cooper sent seeds to Edinburgh, but the first introduction to private gardens was by Kingdon Ward, who collected seed of what he took to be R. maddenii, growing in the Nyamjang Chu, Mönyul, just south of the border between Tibet and India where the two countries abut on north-west Bhutan (KW 6413, collected February 1925). When plants flowered at Nymans, Sussex, eight years later they were identified as R. polyandrum by Dr Hutchinson, and the species received an Award of Merit when shown by Lt-Col. Messel on May 9, 1933. In this form the flowers are creamy white, yellow in the throat, but three years later another form from Nymans received the same award, with rose-pink flowers. Both forms are hardy there.Kingdon Ward collected R. polyandrum in flower in 1938, some way south of the place where he collected the seeds thirteen years before, and reintroduced it. It may also be in cultivation from Ludlow and Sherriff 3164, collected in Bhutan.As noted above, R. manipurense, with which Dr Hutchinson compared R. polyandrum, is very near to R. maddenii, and this is also true of R. polyandrum. In this connection it is interesting to note that Ludlow and Sherriff’s 3164, col­lected in the Maru Chu valley at 8,000 ft, has twenty-three stamens with hairy filaments, and was therefore identified as R. polyandrum, while their 3147, from the same valley at 7,000 ft, has only twenty stamens to the flower and their filaments are glabrous, thus agreeing better with R. maddenii, though Dr Cowan noted that the two specimens were otherwise very similar (Notes R. B. G. Edin., Vol. 19, p. 321). Furthermore, the plants raised from Kingdon Ward 6413 vary considerably in their foliage. In one specimen preserved at Kew they are acute at the apex, as in R. maddenii, not rounded-obtuse as in the type of R. polyandrum. Also, some of these plants have glabrous stamens, as in R. maddenii, not hairy as in the type of R. polyandrum.

---

subsp. crassum (Franch.) Cullen

This is a very variable species as the synonymy quoted indicates. However, R. maddenii is consistently characterized by the large number of stamens and by the number of ovary chambers.

---

subsp. maddenii

R. brachysiphon is distinctive, with small flowers, 45–48mm long, but is no more than an extreme among a series of forms that do not have clear morphological boundaries.