The New Plantsman 1(1): 10-13 (March 1994)
Rosa roxburghii: the species, its forms and hybrids
Graham Stuart Thomas

Here is a rose distinct from all others, so much so that it is placed in its own subgenus — subgen. Platyrhodon (Hurst) Rehder, which is characterized by spiny heps, leaves with many pairs of narrow leaflets and having the achenes borne on a slightly elevated torus. It happened that a double-flowered form of the rose was first grown in our gardens in 1824, having been introduced from Calcutta Botanic Garden to whom it had been sent from China by Dr Roxburgh.

Rosa roxburghii was first described from a double-flowered plant in 1823 by Trattinnick, and thus the type form of the species is forma roxburghii (syn. R. roxburghii var. plena Rehder; R. microphylla Roxb. ex Lindl.; R. 'Plena'; R. 'Roxburghii'). The single-flowered form, also from China is forma normalis (syn. R. microphylla var. hirtula Regel; Rosa roxburghii forma normalis. R. hirtula (Regel) Nakai; R. roxburghii var. hirtula (Regel) Rehder & E.H. Wilson) which was described by Rehder & E. H. Wilson in 1915, having been introduced by Wilson from Sichuan in 1908. Var. hirtula was the name applied to the Japanese wild variety, the leaves of which are pubescent beneath. The double-flowered form is here reproduced from a water-colour of the early nineteenth century in the collection acquired in Canton for the Horticultural Society by John Reeves.

It should be noted that, as grown today, the flowers usually have wider outer petals. Forma roxburghii is much less commonly grown in gardens than the single-flowered forma normalis. The wild single-flowered form makes a large shrub, up to 3 m or so even on poor sandy soils, bushy, with stout, angular-branching stems. If seen in winter, bare of leaves, the papery, peeling bark and pairs of straight prickles at each node are noteworthy. In summer the leaves usually have 9-15 small, narrow leaflets, though sometimes there are as few as 7 or as many as 19. They tend to obscure the floral display of the wide, clear pink flowers which have creamy yellow stamens and a good scent. These are followed in early autumn by large, rounded, distinctly spiny heps, which are also fragrant but remain greenish even when ripe. The spines extend to the stalks and calyx-lobes, and the latter persist on the heps. It is the spines which are so distinct and which, because the heps are reminiscent of a sweet chestnut (Castanea) fruit, have caused it to be known as the chestnut rose or rose châtaigne.

As long ago as 1863 the double form had been used as a parent in hybridizing by Messrs. J. B. Guillot (Guillot Fils) at Lyon-Monplaisir in France, and the result, probably crossed with a Tea rose, was named 'Triomphe de la Guillotiere'. I remember seeing this in a wall at the Roseraie de l'Hay, near Paris some years ago, but I suspect that neither it nor 'Jardin de la Croix', raised at the same garden in 1901, still exist. They were both good doubles and inherited the style of foliage and growth of the parent species. It is sad if they have gone for there are all too few hybrids with obscure species. Rosa roxburghii is reputed to be a parent (with 'Baby Chateau' — a crimson double-flowered Floribunda raised by Kordes in 1936) of three Floribunda roses, namely 'Cinnabar' (syn. 'Tantau's Triumph') (1945), 'Floradora' (1944) and 'Kathe Duvigneau' (1942), all with red flowers and all raised by Tantau in Germany, but these I have not seen. If indeed the parentage of 'Floradora' is as stated, this may account for the great vigour of 'Queen Elizabeth' which is a reputed cross between 'Charlotte Armstrong' and 'Floradora'.

At some time before 1905, R. roxburghii was crossed with R. rugosa Thunb., ever a fertile parent, and the result partakes equally of both, being a good dense bush of some 2 m, well clad in its crisp foliage, amongst which the wide pale flowers tend to hide themselves. The stems are prickly like those of R. rugosa and the bark does not peel. It is interesting that the rounded, large heps are bristly and have some orange colouring. It was named R. x micrugosa Henkel.

*Platyrhodon microphylla, from China and Japan, the so-called Chestnut Rose with cup-shaped fruits covered with fleshy spines, seems more promising in its fertility since I have succeeded in raising the second generation of a cross with Rosa rugosa. The results, however, although extremely interesting from the scientific point of view, are not very promising horticulturally, since the grandparent R. rugosa has been reproduced in facsimile several times, while the others resemble the parent hybrid with strange mutational variations.

CybeRose note: Later generation seedlings duplicating one parent more than the other suggests that cytoplasmic inheritance might be involved.

During his experimental work with the parentage of roses at Cambridge in the second quarter of this century, Dr C.C. Hurst* raised seedlings of this cross, one of which was named R. x micrugosa 'Alba'. Apart from being of rather more upright habit, it is in other respects a replica of the original but of important garden value because the white flowers are produced not only at midsummer, but onwards throughout the growing season. They are, moreover, very fragrant. This might prove to to be a fertile parent and thus bring both species into today's hybrids. They would be very hardy.

In 1926 Dr Hurst raised from open-pollinated seed from Kew a seedling of R. roxburghii which was subsequently named 'Coryana' in 1939. It is thought that the other parent might well have been R. macrophylla Lindl. Its name commemorates a great benefactor of the Cambridge University Botanic Garden, Reginald Cory. Like R. micrugosa it inherited the good bushy habit and small leaves of the former species and is also apt to hide its flowers under the rich greenery, and has peeling bark. The flowers are borne with a horizontal poise as in R. macrophylla and are of rich dark pink with yellow stamens: they open wide and flat. It is a handsome shrub and, in common with R. x micrugosa the chromosome number is 2n=14.

Sir Frederick Stern hybridized R. roxburghii with R. sinowilsonii Hemsl. which is a very vigorous, but tender, species of section Synstylae DC. of the genus. It has magnificent foliage and trusses of small white flowers. It is difficult to envisage the reasoning behind this cross, but the result, 'Roxane', which still grows in Stern's garden at Highdown in Sussex, has made a good bush to about 1.5 m in height and width. It produces plentiful, single, flat, deep pink flowers at midsummer, well displayed over glossy, light green foliage. The heps are green and prickly.

Thus, R. roxburghii has proved to be a 'willing' parent, and might be considered by breeders as worthy of further attention in producing very hardy shrub and other roses.

Graham Stuart Thomas,
Briar Cottage, 21 Kettlewell Close,
Horsell, Woking, Surrey