Bulletin du Jardin botanique de l'État a Bruxelles, 29(3): 205-211 (Sep. 30, 1959)
Some Naming Problems in Rosa
Gordon Rowley
(John Innes Horticultural Institution, Bayfordbury.)

(1) REDHER, A., Bibliography of cultivated Trees and Shrubs hardy in the cooler temperate
regions of the northern hemisphere,
XXXVII — 825 p., Jamaica Plain (1949).
Manual of cultivated Trees and Shrubs hardy in North America exclusive of the subtropical
and warmer temperate regions,
ed. 2, XXX — 996 p., New York (1951).

Few venture deeply into rose nomenclature from choice. A glance at the « Index Kewensis » gives some idea of the extent of published « species » in the genus. There is no « Index Kewensis » for taxa below the rank of species, which have to be sought individually in a very extensive literature. Rehder courageously attempted to summarise the genus (1); he has earned our lasting thanks for sifting and standardising so great a body of information. His classification is based on that of Crépin, which in turn amalgamated the best of nineteenth century herbarium botany. The changed concepts of species today, and the many sister sciences now summoned to aid orthodox (morphological) taxonomy contribute few improvements to this system. A cytologist could have pointed out that diploid R. elymaitica and tetraploid R. hawrana are oddly out of place among the heterogametic Caninae, but by and large this remains the most acceptable name-list and classification we have today. Naturally, only a fraction of published names are dealt with, and the native American roses are covered more fully than, say, the European Caninae. In matters of nomenclature Rehder did much to bring rose names into line with the International Code. He seems to have committed only one grave misdemeanour: the invention of some quite uncalled-for latinised names ('R. X Bruantii', 'R. X Dawsoniana', 'R. X dilecta', etc.) for groups of garden hybrids like 'Teas' and 'Hybrid Teas'. The sooner these names are jettisoned and forgotten, the sooner will Rehder's reputation be cleared of its one blot.

(2) CHITTENDEN, F.-J., The Royal Horticultural Society Dictionary of Gardening, vol. IV, Pt-Zy,
Oxford, Clarendon Press, VIII p. et pp. 1713-2316, fig, (1951) et Supplement, XII p. et 334 pp. (1956).
(3) MCFARLAND, J. Horace, Modern Roses V, Pennsylvania, J. Horace McFarland Company,
Harrisburg, pp. 328-356 (1958).

In revising the Rosa sections of the « Royal Horticultural Society Dictionary of Gardening » (2) and of « Modern Roses V » (3), I have naturally retained Rehder's treatment as a starting point. A few cases arise that call for explanation, or for names or combinations which have not, apparently, been validated before. These are dealt with alphabetically in the list that follows. Many unresolved difficulties remain, but this will probably always be so in a genus so widely grown, so long hybridised and so voluminously written about.

  1. Rosa damascena MILL. var. semperflorens (LOISELEUR-DESLONGCHAMPS et E. MICHEL) ROWLEY comb. nov.
    R. semperflorens LOISELEUR-DESLONGCHAMPS et E. MICHEL, Nouveau Duhamel, VII, tab. 9 (1819), non tab. 18.
    R. bifera var. semperflorens LOISELEUR-DESLONGCHAMPS et E. MICHEL, loc. cit., p. 32 et 54 (1819).
    R. damascena var. bifera HORT. non REGEL, Tent. Ros. Monogr., p. 96 (1877).

This interesting 'Autumn Damask' rose was re-introduced to cultivation by Graham S. Thomas within the past few years. It appeared spontaneously as a sport on a plant of the 'Perpetual White Moss Rose' ('Rosier des Quatre Saisons Blanc Mousseux'), from which it differs in having rich pink instead of white flowers and a complete absence of « moss » on sepals, receptacle and pedicels. Both are pure 'Damask roses' and both flower sporadically in the autumn as well as in early summer, to a more marked extent than any of the other old Gallicanae. R. damascena var. semperflorens closely resembles the well-known R. damascena var. trigintipetala DIECK, the 'Attar Rose', but that never flowers again after July.

(4) CARRIÈRE, E.-A., Production et Fixation des
variétés dans les végétaux, p. 55 (1865)
. [54]

Because of its marked autumn flowering, there is good reason to identify this rose as the 'Quatre Saisons' rose of older writers and the 'Rose of Paestum' mentioned by Pliny for this very feature. Its relationship with the 'Perpetual White Moss' is not yet clear. Is it a single mutation? If so, it is the most striking on record in the genus. Perhaps we are dealing with a chimaera, originating from the use of the 'Pink Damask' as stock for budding the (White Moss). The problem could be solved by raising the 'Perpetual White Moss ) from root cuttings — if one can persuade it to grow on its own roots. The sporting takes place not infrequently and is illustrated on a colour plate showing both types of flower on a single shoot in Carriere (4). Here we are told that « This variety (i.e. the 'Perpetual White Moss') is the result of an accident which apparently occurred for the first time at Thionville around 1835 ».

  1. Rosa gallica L. vel R. rubra BLACKW.
(5) BLACKWELL, El., Herbarium Blackwellianum emendatum et auctum, ed. ch. J. Trew, I,
tab. 78 (1757 — some authorities cite « 1750 » but I have not seen any copy so dated).

In the Royal Horticultural Society Journal for 1941, Dr. C. C. Hurst took up the name Rosa rubra BLACKW. in place of R. gallica L. 1759 non 1753, on the grounds that Linnaeus' 1753 « R. gallica » was in fact a variety of R. canina L., and the mistake was not corrected until 1759, by which time Trew had published the new edition of Mrs. Blackwell's Herbarium in which the name R. rubra appeared (5). This work, however, has no standing in matters of priority since it does not consistently use Linnaean binomials, and is hence outlawed under Art. 70/5 of the International Code.

(6) ROWLEY, G., Nomenclature proposals for the Montreal Congress
V, Comments and proposals, Taxon, VII, pp. 153-155 (1958).

It is well known to students of the roses that Linnaeus treated the genus exceptionally badly, and interpreted the species differently in different publications. Many of his names, such as R. eglanteria and R. rubiginosa, are still subjects of controversy today, to which our current Code offers no solution. Indeed, by citing Rosa villosa L. under Art. 65 as an example of a nomen confusum which must be rejected it invites numerous other name changes if similar treatment is accorded to all Linnaean rose names. This example should be dropped from future editions of the Code (6).

  1. Rosa hemisphaerica HERRM. var. Rapini (BOISSIER et BALANSA) ROWLEY comb. nov.
    R. Rapini BOISSIER et BALANSA ex BOISSIER, Diagn. Plant. Nov., ser. 2, VI, p. 72 (1859) et Flora Orient., II, p. 672 (1872).

Examination of living material and numerous herbarium specimens at Kew convinces me that R. Rapini BOISSIER et BALANSA is the single-flowered wild species from which the full-flowered 'Sulphur Rose' sported some time during or before the seventeenth century. By all that is logical this name should stand with hemisphaerica subordinated as a garden variety of it. However, as the epithet hemisphaerica was published earlier (1762) we have no option but to make the species a variety of the garden rose.

This topsy-turvy situation, which recurs for Rosa chinensis JACQ., R. Roxburghii SWEET and R. xanthina LINDL., makes a laughing stock of our botanical nomenclature. At present it can be overcome only by rigid application of Art. 67, whereby all double flowers are treated as monstrosities and names based on them discarded in favour of new ones. Such a drastic procedure would not be in the interests of a stable nomenclature. Again I suggest a minor alteration to the Code (6) so that flower-doubling is not considered a monstrosity in the sense of Art. 67, and a species epithet can be considered as applying to the wild species even if described originally on a double-flowered cultivar.

  1. Rosa X proteiformis ROWLEY nom. nov.
    R. rugosa THUNB. var. alba ROBINS. X unknown diploid.
    R. X heterophylla COCHET, Journ. Soc. Nat. Hort., 1897, p. 779, non WOODS, Trans. Linn. Soc., XII, p. 195 (1818).

This curious little rose is still in cultivation although of botanical rather than horticultural interest. The first leaflets of the season expand normally, but subsequent leaflets become narrower, paler and crumpled after the fashion of R. multiflora THUNB. var. Watsoniana (CRÉPIN) MATSUM. Perhaps the condition is caused by virus. Clearly the plant is a hybrid of R. rugosa, but the other parent is unrecorded. It is not R. foetida HERRM. as Cochet reported since that is a tetraploid and gives with R. rugosa to a triploid hybrid, the rose 'Agnes', whereas R. X proteiformis is diploid with 2n = 14. The rose 'Adiantifolia' is another garden hybrid referable to, if not identitical with, the above.

  1. Rosa X pteragonis KRAUSE
    R. hugonis HEMSL. X sericea LINDL.
(7) R. X pteragonis KRAUSE, Rosenjahrbuch, 1938, pp. 19 et 21.

In consequence of the union of R. omeiensis ROLFE with R. sericea LINDL. (see below), the name R. X pteragonis KRAUSE (7), now covers all hybrids of R. hugonis with sericea and omeiensis. Their nomenclature is thus:

  1. Rosa Rubus LEV. et VAN. f. nudescens (STAPF) ROWLEY comb. nov.
    R. Ernestii STAPF f. nudescens STAPF, Curtis Bot. Mag., CXLVII, tab. 8894 (1938).
(8) in SARGENT. Plantae Wilsonianae, II, p. 308 (1915).

The original publication of Rosa Rubus LEV. et VAN. is rather confusing, and Dr. Stapf (loc. cit.) regarded garden plants as so different as to warrant a new name, R. Ernestii. Rehder and others, however, have retained these under R. Rubus and are followed here. Two strikingly dissimilar forms are grown today: one quite glabrous and the other softly downy on stems and leaves. For these extremes Stapf published the name R. Ernestii f. velutescens and R. Ernestii f. nudescens. Under present rules one of these, the type forma, must bear the specific epithet unchanged as R. Ernestii f. Ernestii (or R. Rubus f. Rubus on transfer). From Redher and Wilson (8) this is evidently the hairy form, so that the other becomes R. Rubus f. nudescens, a new transfer as indicated above.

  1. Rosa sericea LINDL. and R. omeiensis ROLFE

Hurst's hybridisations and living material now in the National Rose Species Collection at Bayfordbury suggest that these two are best united as a single species under the older name R. sericea LINDL. Parallel variations, like the pteracantha prickle form, occur in both, and all combinations of leaflet number, dentition, pedicel form and so forth. A key follows to the better subdivisions:

  Fruit red:
    Plant unarmed var. denudata.
    Plant with prickles and bristles:  
      Leaflets numerous, up to 17 var. polyphylla.
      Leaflets rarely more than 11:  
        Pedicels slender, green var. sericea.
        Pedicels fleshy, red var. omeiensis.
    Plant with prickles, bristles and dense glands var. Hookeri.
    Plant with long-decurrent winglike prickles var. pteracantha.
  Fruit yellow var. chrysocarpa.

Observation. The roses 'Heather Muir' and 'Praecox' are luxuriant garden selections.

  1. Rosa spinosissima L. var. rosea (KOCH) ROWLEY comb. nov.
    R. pimpinellifolia L. γ rosea KOCH, Syn. Fl. Germ. et Helv., p. 194 (1837).

This is the single pale pink-flowered (Scotch Rose), a rare wildling, for which no valid name has so far been traced under R. spinosissima.


I am indebted to Mr. G. H. M. Lawrence of the Bailey Hortorium for helpful criticism and service in preparing the above.