P. arbuscula is allied to P. fruticosa, but both in cultivation and in the herbarium it is quite distinct. The stipules are a more prominent feature of the plant than they are in P. fruticosa, being darker in colour, opaque, conspicuously veined and not so closely wrapped round the stem. Although pinnate and with five leaflets in the type of P. arbuscula, the leaves are frequently trifoliolate and in some plants wholly so; there is even a form in which the leaves consist of a single leaflet (var. unifoliolata Ludlow). The leaflets are relatively broader than in P. fruticosa, especially on trifoliolate leaves, where they are broad-elliptic and scarcely more than twice as long as wide. In the typical state the leaflets are coated above with appressed, silky hairs but almost glabrous beneath. A character stressed by the Austrian botanist Handel-Mazzetti is that the venation of the leaflets is more closely meshed than in P. fruticosa and the lateral veins less prominent. In cultivated plants of P. arbuscula (and often in P. davurica also), the hairs are swollen at the base, so that the midribs of the leaflets and the young stems appear to the eye to be covered with ‘goose-pimples’. The significance of this peculiarity is uncertain. It has been noted on herbarium specimens of P. fruticosa from Asiatic Turkey and Mongolia but does not seem to be a normal feature of that species. In his original description of P. arbuscula, Don said that his species differed from P. fruticosa in having an outer calyx composed of ten oval-rotund segments instead of five linear ones. This is perhaps not a wholly reliable difference, since the outer calyx segments in P. fruticosa are sometimes split, and in P. arbuscula they may be single and entire, but the difference would probably hold good for the majority of specimens. The flowers are usually larger than in P. fruticosa, sometimes remarkably large, and in cultivated plants the petals are more overlapping and often with crinkled edges. They appear to be hermaphrodite.
P. arbuscula is a native of the Himalaya, S.E. Tibet, W. and N. China. It was described by Don in 1825 from a specimen in Lambert’s herbarium, sent by Wallich from Nepal. The synonymous name P. rigida was given to specimens in Wallich’s main collection of Indian plants, catalogued under No. 1009, of which duplicates were distributed to several continental herbaria. Adopting the catalogue-name, Lehmann published a description in 1831, which happened to be based on a specimen which bore trifoliolate leaves only, which explains why the name P. rigida (or P. fruticosa var. rigida (Lehm.) Wolf) has come to be associated with this form, rather than with plants bearing leaves with five leaflets, or a mixture of the two kinds. In the following year Wallich himself published another description made from a plant with both sorts of leaf, but otherwise very similar (Pl. Asiat. Rar., Vol. Ill, t. 228). The specimens named P. rigida represent the form of P. arbuscula commonest in Nepal and Kumaon, which makes a robust plant with rather stiff stems (whence, no doubt, the epithet rigida). Don’s type looks as if it came from an intricately branched plant, and has rather smaller leaves. But only one species is involved.
P. arbuscula is mainly represented in gardens by plants from Kingdon Ward’s 5774, which was found by him in 1924 on the Temo La and Nam La, making thickets about 2 ft high in the company of dwarf rhododendrons. It was accepted by the Austrian botanist Handel-Mazzetti as a pure P. arbuscula, though it disagrees with it in that the venation is not particularly dense and the lateral veins are raised on the underside. But it agrees well enough in other respects, especially the large, conspicuous brown stipules, which on the twigs conceal the internode on the side of their insertion. It is impossible to agree that it belongs to P. parvifolia, as has been suggested. It has also been placed under P. fruticosa, first as var. grandiflora Marquand (not Schlecht.) and later as f. wardii Rehd. It could, if thought necessary, be distinguished from typical P. arbuscula by transferring Rehder’s name to that species. The Kingdon Ward form has been found many times since in S.E. Tibet and extends as far west as Nepal, where specimens with remarkably large flowers have been collected. A similar form occurs in Yunnan. Plants from KW 5774, which have been in commerce since the 1930s, are perfectly hardy and quite free flowering from midsummer, growing to about 3 ft high. The flowers are golden-yellow, about 13⁄4 in. wide. Award of Merit 1965, after trial at Wisley.
A trifoliolate form of P. arbuscula is in cultivation from Stainton, Sykes and Williams 8221, collected in Nepal in 1954. Plants from these seeds bear large, golden or pale yellow flowers rather late in the summer; the leaflets are of firm texture, 3⁄8 to 5⁄8 in. long, 1⁄4 to 3⁄8 in. wide, bright green under the silky hairs. In the trifoliolate form distributed by Messrs Hillier, the leaflets are silvery green above from the dense coating of hairs and of less firm texture. Both forms are worth growing for their foliage alone.
var. albicans (Rehd. & Wils.) Hand.-Mazz. P. fruticosa var. albicans Rehd. & Wils. – Described as having the leaves laxly appressed hairy above, white and silky-tomentose beneath. The type was a cultivated plant raised from seed sent by Wilson to the Arnold Arboretum from the Tatsien-lu (Kangting) area of W. Szechwan under W.1213a. The foliage was said to resemble that of ‘Vilmoriniana’ (q.v. in the section on garden varieties). The wild-collected specimen in the Kew Herbarium under W. 1213a is typical P. arbuscula, but there are specimens from the same region collected by Père Soulié and by Pratt which have the leaves silky on both sides and probably represent var. albicans. It has also been collected in Bhutan. The garden clone ‘Beesii’ (q.v.) is near to var. albicans.
var. bulleyana Balf. f. ex Fletcher – Described as having leaves glabrous above and both white-tomentose and silky beneath. Found by Forrest in Yunnan and introduced by him, during his first expedition for A. K. Bulley. The plant in cultivation under this name has remarkable foliage: the leaves have five leaflets of firm texture, which are broad-elliptic and up to 3⁄4 in. long, half as wide; they are densely silky and white beneath, but the upper surface is fairly densely appressed silky – sufficiently to give the foliage a silvery aspect in dry weather. It could just as well be referred to var. albicans. The flowers are deep yellow, about 11⁄2 in. wide.