American Rose Annual 1921
Rose-Breeding in 1920 at Bell Experiment Plot PDF
Department of Agriculture, Washington, D. C.

EDITOR'S NOTE.—It will be observed in reading this most interesting continuance of the progress-records with which Dr. Van Fleet has favored us in the Annuals for 1916, 1917, 1918, 1919 add 1920, that he is not only continuing his painstaking hybridizations with the species previously reported upon, but extending them to other rare species. The ideals in mind are high, and the patient skill of the worker most notable. No such breeding work is proceeding anywhere else in the world, so far as can be ascertained.
   Following will be noted the advancing of the arrangement for the dissemination through the agency of the American Rose Society of the new roses which Dr. Van Fleet deems worthy of introduction.
    Attention is directed to Plate II, facing page 9, and to III, IV, V and VI.

THE season of 1920 at Bell Experiment Plot (Glendale, Md.) in most part was as favorable for outdoor rose pollination as the preceding year was adverse. The weather during the blooming months averaged more than ordinarily cool and clear, favoring the setting of fruits, and there was less interference from insect pests and disease than in any previous year. The harvest in matured seed was greater than heretofore and covered the widest range of species yet attempted, the special features borne in mind being hardiness, disease resistance, and good garden appearance of plants. Elegance, profusion, and continuity of bloom are of the highest importance, and every effort is made to develop these perfections in hybrid progeny. There are already too many weak varieties that produce a few exquisite blooms under exceptional conditions of culture, but new varieties that do not require incessant coddling are needed for American gardens.

Particular attention was given in 1920 to the utilization of the lesser-known northern species of both hemispheres, such as Rosa inodora, R. pulvurulenta, R. Murielae, R. hibernica, R. micrantha, R. involuta, and R. Jundzillii of the Old World, our native R. Macounii and R. nutkana, and the most northern forms of R. nitida, R. pratincola, and R. lucida or R. virginiana. There is considerable diversity in garden adaptability among these extremely hardy wild roses, and crossings with the several attractive new Chinese species and with highly developed florists' varieties may open up new features of value in the offspring. Seeds of most of the northern species are of slow germination, and probably several seasons will be needed to bring out their possibilities.

Taking the seedlings of past years in the order of bloom, we may begin with R. Willmottiae,* which opens its cheerful rosy purple flowers at Bell before the middle of May. This very distinct species has not generally succeeded in cultivation, but grows thriftily in the sandy loam soil at Bell, ripening a limited number of hips. The type has a most attractive drooping habit, the strong canes rising about five feet high before they curve, and it has attractive foliage throughout the growing season. In habit and coloration this wild rose gives, in early spring, essentially the decorative effect of the purple beggarweed, known to gardeners as Desmodium penduliflorum, a beautiful late autumn perennial, but on a much larger scale.

This striking new rose does not readily lend itself to hybridization, and the best progress is likely to be made with self- or chance-fertilized seedlings. Quite a number have already bloomed, some showing brighter colors and greater profusion of bloom than the parents. A hybrid with R. Hugonis, as might be expected, is so inferior in decorative effect to either parent as to be an almost pitiable object. The blend with R. altaica loses in grace of habit and foliage effect, but has large, blush-white blooms, very well placed. Hybrids with Teas and Hybrid Perpetuals show weakness in growth and are slow in reaching the bloom stage, but those with deep red forms of R. rugosa are very encouraging in color, form, and abundance of bloom, the general aspect being that of a graceful, bright-flowering, miniature R. rugosa adapted for edgings and rockeries.

Rosa Hugonis blooms about the same time and occasionally anticipates R. Willmottiae, its clear yellow flowers harmonizing well with the rose-purple of the latter species. About 250 hybrids and self-seedlings have bloomed to date, and a considerable number of new blendings are coming on. The hybrids with R. altaica and the white forms of R. rugosa run a close race in attractiveness. Both agree in boundless profusion of bloom, the large, widely opened flowers being so thickly placed as to almost hide the branches.

*See Plate IV. pages 28-29, for Hugonis-Altaica seedling.
†See Plate II, facing page 9.

All the seedlings with R. altaica* as the pollen parent have single flowers three inches and more in diameter, the shades varying from porcelain-white to sulphur-yellow. There is great diversity in character, poise, and finish of the blooms, suggesting need for critical selection to secure the best possible combination of characters. The plants are noticeably graceful in habit, the long shoots arching to the ground with the weight of buds and bloom, while the foliage is clean and persistent. The hybrids with Rugosa have larger flowers, several being quite double, and the shades vary from cream-white to blush-pink and salmon. The growth is heavier, with quite characteristic Rugosa foliage, and the plants bid fair to form large clumps, five or more feet high, at maturity. The combination of Hugonis with Rugosa has so gained in promise with repeated bloom that greater attention will be given it in future.† Several crosses of Hugonis with Harison's Yellow opened blooms in 1920, showing deeper coloring than the type and some duplication of petals. The foliage is dense, and the habit of the seedlings is upright.

Blends of Hugonis with garden forms of R. lutea, such as Persian Yellow, Austrian Brier, and Austrian Copper, have yearly been attempted and only two weak seedlings secured, but the past season better success in securing seed was had by the use of the type known to botanists by the libelous designation of R. foetida. In plant and flower this type is scarcely to be distinguished from the familiar single-flowered form of Austrian Brier, but appears to have more active pollen. There is no unpleasant odor in leaf or flower, but rather a faint sweetbrier fragrance in the young growth on damp mornings.

A hybrid of R. omeiensis, with R. Hugonis as the pollen contributor, is attractive from its dense, arching habit, forming a most symmetrical mound of growth, with its fine glossy foliage enduring until December. The only bloom developed on the plant last spring resembled that of Hugonis in its yellow coloring and in having five petals, instead of four, as in the seed parent. The attractive flowers of R. omeiensis, as well as those of the closely related R. sericea, are pure white and composed of but four broad, rounded petals. All other well-known rose species have five-petaled blooms.

The self or chance seedlings of Hugonis are remarkably uniform in flower, foliage, and habit, rarely being distinguishable from the wild parent, but vary to a considerable degree in fruitfulness. We now have seedlings far superior to the type for breeding, producing good seeds in quantity and accepting a diversity of pollen.

The closely related R. xanthina, sent over from China in single- and double-flowered forms by the late agricultural explorer, Frank N. Meyer, shows considerably more variation in its cross and self-pollinated seedlings, the flower colors ranging from paper-white to a somewhat deeper yellow than has yet been developed in Hugonis. Many seedlings are free from the penetrating formic acid odor so objectionable about the young growth of the type in moist weather, and will consequently be preferred for planting near dwellings.

Hybrids of R. xanthina with R. Hugonis are very erect in habit, with dense foliage and attractive brown-red stems in winter. They tend to grow taller than either parent species, and should form excellent material for dense boundary hedges. The widely opened white or yellow blooms are very thickly placed in early spring, and are followed by glossy red or purple-black fruits that last until midsummer. The blends with R. altaica are less striking than the Hugonis-Altaica hybrids, as the flower colors run in lighter shades, but have merit in their good foliage and their interesting armature of broad-based prickles.

Little benefit has been secured by working with R. altaica except the fine hybrids with Hugonis. While so attractive in itself that it is grown as the "Northern Cherokee Rose," though it belongs to the R. spinosissima group, only resembling the Southern Cherokee rose in its beautifully finished white blooms, it appears to fall off at once when crossed with other species and varieties.

Rosa hispida, a nearly related form, very erect in habit, with straw-colored blooms, hybridises readily with many varieties, and produces charming seedlings with single to densely double blooms of a great range of color, soon showing as many as three shades of pink in the same petal. Rosa hispida x Sunburst produces large, full buds and blooms, bright canary-yellow until fully expanded, and should make a fine hardy border rose.

Rosa bella, among the new-comers, is likely to command much future attention. It is dwarfer and more compact in growth and much hardier than other species recently introduced from North China. The cherry-red blooms cover the branches soon after R. Willmottiae and R. Hugonis pass out of bloom, and continue to appear for a longer period than most wild roses of the season. They are slightly fragrant in themselves, but the very agreeable fruity odor noticed about the plant in the morning hours, and which may be likened to the fragrance of Azalea nudiflora, or Pinxter Flower, appears to be produced by a glandular surface on the young growth and pedicels. The type plants at Bell mature few hips, but seedlings are more fruitful, and a series of promising hybrids have been raised, the best one being a blend with R. Malyi, having much brighter red flowers than either parent, with intensified fragrance, good compact habit, and ornamental fruit. The outcome of crossing this fine species with highly developed florists' roses is awaited with interest.

Rosa Moyesii will always challenge attention in its typical form, the deep red blooms standing out distinct from all other wild roses, but is far from easy to propagate and grow. Plants show great vigor for a few seasons, throwing up strong, tall canes, but there is a constant tendency to die-back in the mature wood, irrespective of disease or winter injury. This species, in common with most introductions from North China, appears to thrive best in light soil and on its own root system. It is not thoroughly at home on any stock we have yet tested, but it is difficult to secure plants except by budding or grafting, as cuttings and layers rarely form roots. Sell-pollinated seedlings grow well but do not retain the intense red coloring of the type, the flowers opening into crimson shades and even light rose and pink. Hybrids with American rose species, such as R. Engelmannii, R. arkansana, and R. nitida have all the depth of color of Moyesii, and added brilliancy from the yellow stamens conspicuously displayed in the center of the blooms. Moyesii itself has reddish brown anthers and filaments, giving the interior of the flower a rather somber aspect when not in bright sunshine. The hybrids with Old World species that have bloomed are R. Malyi x R. Moyesii, R. canina x R. Moyesii, and R. Moyesii x R. cinnamomea.

The Malyi blend was noted in the American Rose Annual for 1919 (page 32). The blooms almost duplicate those of Moyesii in color and form, but the habit is dwarf and compact and the plant apparently very hardy. The Canina-Moyesii crosses gain in beauty with respect to bloom, and should certainly be widely planted. While habit and foliage are much like a moderate-growing dog-rose, the blooms are vivid ruby-red, lighted with prominent golden anthers and very thickly placed on the two-year canes. Like the other Moyesii hybrids, they are vigorous growers, hardy and free from disease. The hybrid with R. cinnamomea—the only good one yet secured with Moyesii as a seed parent—first bloomed the past season, and promises to be the best of all in size and brilliancy of bloom, fairly beating Moyesii in intensity of coloring. The plant appears to have no defects except that of profuse suckering, common to the Cinnamon rose, which, however, will insure rapid increase of the hybrid plants on their own roots.

We have undertaken to hybridize Moyesii with every ornamental hardy rose species of both hemispheres that can be grown about Washington, and also with promising garden varieties that may become available, as far as these crosses can be effected by painstaking work. More than thirty of these blendings have been carried to the point of seed production, and new varieties combining the unique characteristics of this beautiful importation from China may yearly be expected.

A few other of the new oriental roses are showing possibilities for breeding. Rosa setipoda produces striking hybrids with R. nitida and R. lucida, while R. sertata has at last produced seeds when fertilized with pollen from diverse species and varieties. Rosa multibracteata and R. filipes, the latter quite near R. Helenae, are now giving encouraging results, and may add unique features to our rose-gardens. A considerable number of hybrid seedlings of R. multiflora cathayensis, R. Jackii, and R. Gentiliana have reached their second year, and new crosses are under way. Cathayensis and Jackii are valued for extreme hardiness and the comparatively large blooms the typical forms produce. Rosa Gentiliana is quite tender at Bell, but develops a charming profusion of milk-white flower-clusters on the branches that escape winter killing, so that it appears well worth developing for southern localities.

*See Plate V. page. 28-29.

Rosa Wichuraiana and R. Soulieana continue to produce beautiful novelties that should be distributed for trial throughout the country. A hybrid of Wichuraiana and Pernetiana shows very large and full flowers.* Bell Experiment Plot does not possess equipment for extensive propagation of new hybrids and selections, and it is hoped that a working arrangement will soon be perfected with the American Rose Society by which the most promising new roses produced here may be increased and widely disseminated at very moderate cost to the purchasers.


The rose varieties of old gardens are not neglected at Bell. The typical Scotch rose of our forefathers, R. spinosissima, is yielding beautiful new seedlings with intense red as well as paler shades in both single- and double-flowered forms. The Boursault rose, R. Lheritierana, is capable of developing really fine new varieties, and R. gallica officinalis, source of the red rose petals and much of the rose attar of commerce, is so bright and attractive that it appears well worth blending with R. Moyesii and other deep-colored wild roses.

The possibilities of the sweet brier in its less known forms, such as R. agrestis, R. Seraphinii and R. ferox, are far from exhausted as regards the development of varieties with fragrant foliage and handsome flowers. The much-admired Harison's Yellow gives an occasional seedling of the Scotch rose type that nearly equals its parent, and one or more may eventually be raised that will supersede it in garden value.

Time is an insistent element in breeding work of this character as several seasons are often needed to carry the seedlings through all phases from pollination to characteristic bloom, but each succeeding year brings its encouragement as well as attending disappointments.

PLATE II. Some Chinese Roses and Their Van Fleet Hybrids as Grown at Bell Experiment Plot. Photographed May 20, 1920.

A. Rosa Willmottiae x R. altaica
B. R. Willmottiae: violet-purple
C. R. Hugonis x Harison's Yellow; semi-double, primrose-yellow
D. R. Hugonis seedling; deep cream
E. R. altaica; white
F. R. Hugonis; canary yellow
(See page 26)
PLATE III. ROSA WILLMOTTIAE in bloom at Bell Experiment Plot (See page 26)
PLATE IV. Hybrid of Rosa Hugonis X R. altaica, two years old, blooming May 20, 1920, at Bell Experiment Plot (See page 26)
PLATE V. Van Fleet Hybrid of Rosa Wichuraiana X R. Pernetiana at Bell Experiment Plot (See page 31)
PLATE VI. New Hybrid Rose, "W. C. 124" (Rosa Wichuraiana X Hybrid Tea) in bloom in 1920 at Bell Experiment Plot.
Originated by Dr. W. Van Fleet; received three first prizes at Portland (Ore.) Test-Garden in 1920. (See pp. 33, 117)