American Rose Annual (1916)
Possibilities in the Production of American Garden Roses*
Department of Agriculture, Washington, D. C.
*Published by permission of the Secretary of Agriculture.

EDITOR'S NOTE.—As may be noted elsewhere, Dr. Van Fleet has produced several of the best climbers of the day. In this article he discusses hybridizing from the position of long experience, and attention to what he has here written will greatly aid aspiring workers with the rose.

The climbing rose, Dr. W. Van Fleet. Introduced in 1910; a cross between R. Wichuraiana and the tea rose Souv. du President Carnot.
Rosa setigera is one of the five American native or "wild" roses.
Ruby Queen—a Wichuraiana hybrid rose, introduced in 1899 by Dr. Van Fleet.
Rosa multiflora, var. Cathayensis—one of the important species introduced from West China by E. H. Wilson, of the Arnold Arboretum, by whose courtesy the photograph is used.
The Cherokee rose is not hardy in the North, but has "escaped" and run wild in the southern United States.
Harrison's Yellow is a fine old garden rose. the color plate is from "How to Grow Roses," by Robert Pyle.
Silver Moon, a lovely white climber, was hybridized by Dr. Van Fleet, between R. Wichuraiana and the Cherokee rose. Photograph used by courtesy of Bobbink & Atkins.
Prairie Queen is among the oldest of the climbing roses of American origin.
Baltimore Belle is an old climber of great beauty.

ROSE-BREEDING now, as in the past, progresses quite exclusively along commercial lines. The aim of most raisers of seedling roses, here and abroad, appears to be the prompt production of compact-growing and constant-blooming varieties, suitable for the production of cut blooms under glass or in the garden. Introductions of new roses of the Hybrid Tea, Tea, and dwarf Bourbon types suitable for glass-house culture or garden bedding in 1912 numbered ninety-nine, as against thirty-seven in all other classes combined. The following year the disproportion was even greater. The seasons of 1914 and 1915, from best information, would have shown further proportionate increase of the everblooming types if the deplorable war situation in Europe had not interfered.

This is all good work. Many splendid and highly useful varieties have been developed that will long beautify our greenhouses and gardens, providing large revenues for the florist and nurseryman as well. The great value of such roses as General MacArthur, Richmond, Radiance, My Maryland, Killarney, and the gorgeously tinted Pernetianas, appealing in the highest degree alike to growers and flower-lovers, is not to be depreciated. They are precious acquisitions, but it cannot well be denied that continuous-blooming roses, with their strong infusion of tender Oriental blood, are, with very few exceptions, children of exacting cultural conditions and cannot generally be relied on as home-yard plants.

More easily managed varieties than are now available, suited for common dooryard culture under the diverse climatic conditions of our broad country, are needed. For securing diversity of type as well as excellence of bloom, all available vigorous species of pleasing aspect, and their strong-growing garden forms, should be utilized, crossing and blending them together and with highly developed florists' varieties, in the hope that in some of the progeny will be combined the really desirable characters of the parent. This work should be widely carried out in all parts of our greatly diversified country, particular attention being paid to the raising of seedlings from species or varieties naturally adapted to the location.

The first American rose hybrids to gain recognition were probably the beautiful Noisette climbers raised in Charleston, S. C., as early as 1816. They are combinations of the everblooming Chinese rose and Rosa moschata, the wild musk rose of the Himalaya Mountains. Whether produced by intentional or natural hybridization, the introduction of this free-growing decorative type, which has since reached its highest development in the incomparable golden-flowered Marechal Niel, reflects high credit on the originators. In 1880, Harrison's Yellow, absolutely indispensable for dooryard adornment throughout practically our whole country, was sent out from a New York nursery. It bears evidence of admixture between the Asiatic Rosa lutea and the Scotch rose, R. spinosissima, and is the only form of the bright yellow Rosa lutea thoroughly at home in our climate.

The Queen of the Prairies type of climber, formerly planted in great numbers but now superseded by the new Wichuraiana and multiflora ramblers, was produced, it is well known, about the year 1848 by Samuel Feast, a Baltimore nurseryman. The wild Michigan or Prairie rose, R. setigera, is plainly the dominant parent, and the perfume rose of southern Europe, R. gallica, is assigned as the other factor, though the progeny, like the native setigera, is scentless. Baltimore Belle is thought to have had a Noisette rambler as the pollen parent.

Hitherto there is little evidence of intentional breeding work, in the modern sense, among rose-growers. Superior seedlings or mutations of chance origin were propagated when observed and disseminated by nurserymen. Rose-breeding for the avowed purpose of developing new varieties adapted to the needs of the country may be said to have been initiated by the late H. B. Ellwanger, of Rochester, N. Y., writer of one of the best rose books ever published. He worked with the best varieties and species that came to his hand, and made an especial effort to inject diversified blood into Queen of the Prairie and other of the Feast climbers, but they proved then, as since, sterile, refusing to produce seeds or fertile pollen under all available conditions. Marshall P. Wilder, a seedling of General Jacqueminot, introduced in 1884, is the best known of Mr. Ellwanger's productions and is a most excellent garden rose.

Few notable results appear to have been accomplished by American rose-breeders after Mr. Ellwanger's efforts until the closing years of the past century, when the general dissemination of the hardy and vigorous Asiatic species, Rosa rugosa, R. multiflora, and R. Wichuraiana, suggested new possibilities of combination with existing successful types. The rugosa rose appears to have been first used in this country by E. S. Carman in New Jersey and J. L. Budd in Minnesota, though Jackson Dawson, of the Arnold Arboretum, produced hybrid seedlings as early as 1892. Few of these early crosses have been widely cultivated, Agnes Emily Carman, progeny of rugosa pollinated with Harrison's Yellow, and the Arnold and Ames roses, both rugosa and "Jack" crosses, being best known.

Rosa multiflora and R. Wichuraiana next received attention, the first at the hands of Jackson Dawson, resulting in the well-known Dawson climber (multiflora x Jacqueminot), and others. The first hybrids of R. Wichuraiana given to the public were sent out by W. A. Manda, of New Jersey. The variety Gardenia (Wichuraiana x Perle des Jardins), often termed the "hardy Marechal Niel," never has been superseded and is still widely grown. The work was taken up by M. H. Walsh, of Massachusetts; Jackson Dawson; James A. Farrell, of Pennsylvania, and others, the appearance at this time of the superior climbing form of multiflora, known as Crimson Rambler, lending new interest to the production of tall-growing roses.

The writer's experience in raising hybrid seedlings from various rose species has convinced him of the great desirability of disseminating information-none too plentiful—of the breeding characteristics of native and Old-World rose species, in order that efforts for their utilization may not be too greatly duplicated. Much has been done with a few of the rugged and free-growing natural types, but the surface of rose-breeding for American home-yard adornment has barely been scratched. A brief review of the most promising species for immediate effort may be in order, taking first those native to or already well disseminated in North America.

Rosa rugosa must be regarded as of first importance. While it appears to be found naturally growing only in Japan and eastern Siberia, it is at home wherever planted in the United States and Canada. The plant endures great heat and is scarcely destructible by cold, but the blooms are more lasting and better appreciated in northern latitudes. Rugosa varieties and hybrids bid fair to become the most reliable and highly prized bush roses for the northern and Prairie States, and, when the choicer forms are known, to be valued far down toward the frostless regions. Rugosa hybrids, as a rule, carry their vigor, beauty of foliage, frost- and disease-resistance well into the third and fourth dilution with Tea and Remontant blood, while gaining greatly in beauty of bloom and coloring; but the faults of excessive spininess and weak flower-stems also persist; and the rugosa type may be regarded as especially adapted for the garden and not likely to produce varieties having value for cutting and exhibition.

The rugosa type has been hybridized with almost all the cultivated species and with many of the garden forms, resulting in a considerable number of varieties of great botanical interest and not a few of considerable garden beauty. Apparently not much more is to be expected from primary crosses except as a starting-point with newly found species, but the hybrids should be extensively used whenever they prove fertile, which is very rarely the case among those bearing double flowers. The most reliable of the latter are Delicata, pink; Germanica, deep red; and Souv. de Pierre Leperdreiux, crimson. Double white-flowered rugosas, in my experience, rarely fruit, and if seeds are occasionally produced they germinate feebly, if at all. The single white variety, however, seeds freely when properly fertilized with good pollen from double roses, and the resulting progeny usually shows a fair percentage of plants with well-formed double blooms, fragrant and freely produced. The work of building up acceptable rugosa varieties, it must be admitted, is excessively slow—one must be content with the prospect of one or two real successes in a lifetime of effort.

Rosa Wichuraiana, on the other hand, gives quick and generous response to all well-considered breeding efforts. Most of the charming new rambling varieties, now so plentifully grown, are the direct result of pollinating the type or natural wild form of Wichuraiana with the best garden roses, including Teas, Hybrid Teas and Remontants. So susceptible is this oriental species to foreign pollen that if grown near other varieties blooming at the same time it rarely reproduces itself from seed. No other species so richly rewards the breeder impatient for results, but its ease of manipulation has rather flooded the market with varieties lacking distinctiveness.

What the rose-loving public now demands is large-flowered Wichuraiana climbers of varied coloring, having blooms approaching the size and finish of the indoor and exhibition varieties, but with the vigor and foliage advantages of the type and earlier hybrids. To secure these qualities it appears necessary to use as seed-producers Wichuraiana hybrids resembling. the type, but one or more removes from the wild form.

Thus the climber, Dr. W. Van Fleet, was raised from seeds of a Wichuraiana x Safrano seedling (showing little of the tea-scented pollen-parent except in reddish shoots and longer stamens), pollinated with Souv. du Pres. Carnot. The result practically places the highly finished Carnot blooms on a rampant hardy climber, from which buds with 18-inch stems may be cut by the armful in its season of bloom.

Silver Moon is the offspring of Cherokee rose pollen on the stigmas of a cross between Wichuraiana and Devoniensis, the latter a strong-growing Tea rose, possibly having traces of the Indian Rosa gigantea in its composition. This hybrid differed from R. Wichuraiana only in its fewer large blooms produced late in the season, and the very sparing way in which it fruited. Wichuraiana hybrids of the better class are rarely fertile, and often tend to dwarfness in their seedlings, when they can be bred at all. A race of everblooming Wichuraiana of bushy growth will no doubt be eventually developed.

Rosa multiflora in its typical form does not promise much, but the progeny of Crimson Rambler and the charming dwarf Polyanthas offer fine grounds for future breeding work. The weakness of this type is the susceptibility of the foliage of many varieties to mildew.

Rosa lutea (Harrison's Yellow).—Reference has already been made to this charming variety. The pollen has been quite extensively used on R. rugosa and other species, but thus far has given little result except in the production of the dark crimson Rugosa hybrid, Agnes Emily Carman. I have raised some very attractive yellow and coppery flowered crosses of Harrison with rugosa alba, but only disappointment has followed its use with other varieties. Plants of Harrison's Yellow in dry situations occasionally seed with some freedom; but, although many hundreds of chance or self-fertilized seeds have been sown, I have never known one to germinate, and have never been able to secure seeds by pollinating its blooms from other roses, though as many as 600 trials have been made in a season. All seeds produced by this fine old variety should be planted in the hope that some will grow and in time help to solve the riddle of its origin.

Seeds of other forms of R. lutea, such as Persian Yellow, Austrian Copper, etc., are quite as refractory, none germinating under my observation. Persian Yellow is, however, the pollen parent of Lord Penzance, one of the best of the Sweetbrier hybrids, and also through Soleil d'Or, is the dominant parent of the new Pernetiana race. It may well be used in this country.

Rosa bracteata (Macartney).—This beautiful white-flowered climber, native of eastern Asia, but fairly well established in the South, has been used very little for breeding purposes. But few varieties are known, the chief one, Maria Leonida, not always expanding its blooms. It is usually termed a tender species, but is quite hardy at Washington, D. C. I have found it to seed freely when pollinated with many varieties and other species. The few crosses that have bloomed include a lovely pink-flowered hybrid with Rosa carolina, with extremely long-pointed buds; and a fragrant, double-flowered, pure white variety of bushy form, the result of using pollen of the scentless Frau Karl Druschki. R. bracteata promises well for the production of varieties suited to at least the South.

Rosa laevigata (Cherokee).—Though long considered a native, the Cherokee rose is now believed to have originally been imported from China or Formosa. It is widely naturalized in the South, extending on the banks of irrigating canals far into Texas. Where it is sufficiently hardy to bloom well it is highly prized for its large and beautiful white blooms and shining deep green foliage. Countless attempts have been made to blend it with the choicer garden roses, but failure has been so constant that Cherokee rose-breeding has been pronounced impracticable. The writer has squandered whole seasons of work on the Cherokee, and has little to show for it except Silver Moon and a bushy seedling producing apple-blossom-pink, semi-double blooms, of exquisite fragrance but of little garden value. Scores of hybrid offspring of the choicest parentage have been grown from this species only to perish before flowering, often without divesting themselves of immature foliage. A hybrid, Cherokee x Marechal Niel, promised much at the outset, repeatedly sending up shoots 8 to 10 feet high, only to have the juvenile-looking foliage fall before full development, and the shoots wither away. This variety was grown in the greenhouse and outside, on its own roots, budded on both parents and on other stocks, in the East, and was also well established in a favorable location in California; but it perished after four years of trial without developing a bloom. The Cherokee rose, like Harrison's Yellow, is indeed a hard nut for the rose-breeder to crack; yet it has developed varieties of value like Anemone, a lovely pink-flowered form, thought to have an infusion of Tea-rose blood; and efforts to blend it with other types should not be abandoned.

Rosa rubiginosa (Sweetbrier) and R. canina (Dog rose).The pleasing results obtained by the late Lord Penzance by crossing in England the Eglantine or wild Sweetbrier rose with various garden roses is known to all in the form of the Penzance hybrids, charming as hedge or dooryard roses, with their bright-colored blooms and scented foliage. Trials in this country, however, have not added new varieties comparable to those produced abroad, and the same may be said of the Dog rose, now used in hybridization to some extent abroad. Both species have foliage very susceptible to fungous troubles in our climate, and have hitherto failed to give good results, when used for breeding purposes. Many indifferent forms of the Sweetbrier of the Rosa agrestis type, with foliage of faint odor or altogether scentless, abound here, and care should be taken to secure the true R. rubiginosa if it is desired to use the species.

Rosa setigera, the Prairie rose.—This valuable native should be used freely where hardiness and vigor are especially desired. Although the Samuel Feast seedlings have never been changed by the efforts of breeders, the type remains to be worked with, and promises well when combined with free-growing Asiatic species. American Pillar, which has won wide popularity here and abroad, came from a Wichuraiana x setigera hybrid, pollinated with a bright red Remontant rose. Not all crosses with setigera, however, are good. When the species was bred with Hybrid Teas, the result was a number of exceedingly bright-colored varieties with thin unattractive foliage.

Rosa carolina.—This tall and vigorous native of our eastern lowlands without doubt has possibilities of a high order. The long-budded hybrid with the Macartney rose, referred to, appears desirable. Other cross-bred seedlings are on the way.

Rosa californica.—This promising species may be regarded as the Pacific Coast representative of R. carolina, though there are some rather important botanical and horticultural differences. The large panicles of pale pink flowers are borne through a long season, and there appears to be a decided tendency toward doubling of the blooms and the production of autumn inflorescence when grown in the East under good cultural conditions. R. californica has been hybridized with R. rugosa in Europe, and readily accepts the pollen of other species and varieties. I have cross-bred seedlings of this species under way.

Other native species.—With the exceptions above noted, I have found most native species intractable to hybridization, seldom producing sound seeds under controlled pollination, and showing little change of type even when hybrids have been produced. R. virginiana, R. blanda, R. Fendleri, R. Woodsii, R. humilis, R. Sayi, R. nitida, and others have repeatedly been tried without encouraging results, the efforts extending over twenty years. Most of these species will cross with Rosa rugosa, but except for botanical collections the immediate progeny is seldom worth growing, and the succeeding generations of self- or cross-pollinated seedlings show little change. Rosa virginiana and others, however, have given rise to charming double and white-flowered mutations under varied cultural conditions, and should not be neglected. The delightful fragrance of many species of this group is an encouragement, as also their entire adaptability to climatic conditions. There are some other native species of promise that do not appear to have been used for the production of new varieties.

New exotic species.—Some very beautiful rose species of great breeding interest have been discovered of late years, mainly in central and western China, and are slowly being introduced to American and European cultivation. At present, the most promising are:

Rosa Soulieana.—An extremely vigorous species of multiflora-type, less hardy in wood but with far more resistant foliage and larger blooms, very plentifully produced. Some botanical specimens of R. Soulieana show light yellow flowers, but those available in this country are of the white-flowered type. R. Soulieana appears to be quite as readily hybridized as R. Wichuraiana and may confidently be expected to produce varieties of interest. The only seedlings yet bloomed, with Cabbage or Rosa centifolia varieties as pollen parents, have delightful semi-double blooms, shell-pink and light crimson in color, fragrant and beautifully formed. The hybrid plants are rampant in growth, with very spiny stems, and preserve the resistant foliage of the species. The cross with Rosa setigera, as yet unbloomed, is especially vigorous.

PLATE V. Flowering Branch of ROSA HUGONIS. Flowers clear yellow. This photograph supplied by Dr. W. Van Fleet, shows them about half natural size. (See page 35.)

Rosa Hugonis and R. xanthina are early-blooming, yellow-flowered species of great promise as ornamental shrubs for dooryard embellishment in this country in their natural state, and should be capable of developing varieties of even greater attractiveness. There is some confusion between the two forms as introduced to cultivation here, but the one known as R. Hugonis appears to afford the better opportunities for breeding, as it seeds with great freedom and appears to hybridize readily with other wild and cultivated roses. Seedlings of Hugonis X rugosa, white variety, and Rosa altaica, are the first to bloom, to both of which it has imparted much of its yellow coloring. R. xanthina has given plants with double yellow blooms from collected seeds, but as yet does not fruit well here; more fertile forms may in time be imported. The seeds of R. Hugonis are very slow in germination, only a few coming up within two years. As the ungerminated ones appear sound, some may eventually grow.

Rosa Moyesi and R. Fargesii are two beautiful Chinese species of the R. macrophylla type, little known in this country. They are strong growers when established on their own roots, quite hardy, and produce showy blooms of great substance, deep brownish red in color. R. Fargesii has the darker flowers, but both are exceedingly attractive, and if they will "nick" with improved garden roses, very distinct varieties may result. I have not heard of either species fruiting here, but the pollen has been used here and abroad, apparently with good results.

Rosa setipoda is also an unusually attractive species, with nearly spineless branches, fine large resistant foliage and large clusters of lively pink blooms. Like the preceding species, it has not yet fruited under my observation, but the pollen appears effective on other species and garden varieties. It should impart good characters to its hybrids if they are successfully produced.

Rosa Willmottiae has a very distinct and pleasing appearance when in bloom. The bright pink blossoms are freely produced in graceful sprays very early in the season, but the pretty foliage does not well endure our summer conditions, falling an early prey to "black spot" and other fungous troubles. It fruits sparingly, and the pollen is effective at least on R. rugosa and R. setigera, but the foliage weakness is so pronounced that little can be expected from hybrids in our climate.

Rosa sertata and R. floribunda are Chinese novelties, highly praised from the European standpoint, but their utility for the Western hemisphere has not yet been tested. R. sertata is an excellent grower in Washington, D. C., and is said in effect to be a very superior wild rose of the R. Willmottiae type.

The list of hopeful species is far from being exhausted. Only the most prominent have been mentioned. New species and highly important local types are constantly brought forward. The era of intelligent rose-breeding for outdoor effects has scarcely dawned, and the wealth of material at hand suggests the widest use by patient and hopeful workers.