Jour. Trenton Nat. Hist. Soc. 2(1): 1-7 (Jan 1889)
North American Roses; Remarks on characters with classification
By G. N. Best, M.D.

To the botanist who yearns to enrich synonymy the Roses offer at once an inviting and productive field. It is not even necessary for him to go from bush to bush; by taking advantage of the successive stages of growth, from one alone he may cull species and varieties ad libitum. This may seem an exaggeration, but such is not the case, nor is the following instance unique. Amos Eaton is authority for having said that Dr. Bigelow, author of Florula Bostoniensis, once on a time collected three specimens from the different parts of a single bush, sent them to Sir Joseph E. Smith, then one of England's most distinguished botanists, who in reply made two of them different described species and the other an undescribed species.

Rafinesque in his Prodrome d'une Monographie des Rosiers de l'Amerique Septentrionale, 1820, makes of the Roses of this country twenty-nine species. Gray in Flora of the Northern Portion of the United States, 1848, reduces the number to four. Rafinesque collected in New York, New Jersey, Pennsylvania, Maryland, Virginia, North Carolina, Kentucky, Ohio, Indiana and Illinois. It is therefore evident that in Gray's reduction are embraced all, or nearly all, the forms with which Rafinesque was acquainted. This comparison shows the extremes, the mean being filled by many conflicting authorities and unsettled descriptions.

Nor has this confusion of the Roses been confined to this country; in the Old World the like has been going on, only on a grander scale. Volume after volume has been written, classification after classification presented, each author stoutly protesting against the accuracy of those differing with him, and as warmly insisting on the correctness of his own views. Déséglise in 1876 makes of the Old World Roses four hundred and ten species. Gaudoger in Rhodologiae Europae-Oriertales, 1881, is content with no less than four thousand two hundred and sixty-six. In fact a history of the perturbations in nomenclature of the genus Rosa would inspire serious doubts as to the saneness of mind and the stability of matter.

The question naturally arises, what are the peculiar difficulties met with in the differentiation of the forms of Roses? A correct answer would be the finding the value of the unknown quantity in the rhodological equation. An attempt, however, is to be made, and in doing so certain fundamental principles are to be considered, applicable not only to the special subject under treatment but to descriptive botany in general.

Extreme variability is one of the distinguishing characteristics of the Roses; environment and hybridization seem among the most potent factors in its production. Changed conditions modify more or less all plants; in cultivation are seen the extremes of these modifications. It is quite reasonable to infer that like changes, a question of degree rather than of kind, are going on in the wild state. In the germs of every plant there are apparently certain potentialities which, when brought into action by.the stimulus of environment, either advance the organism in the scale of development or return it to some stage of its evolution when surrounding conditions were somewhat analagous. More commonly, however, these tendencies are so evenly balanced as to retain it in a comparatively fixed condition.

*Origin of species, Chap. IX., Chas. Darwin.

Of hybridization among uncultivated plants, although often invoked as a cause of variability, comparatively little is known. The experiments, however, of Kölreuter, Gärtner, Herbert, Darwin and others* seem to show that hybrids are not so universally infertile as was formerly supposed. Even when the first cross is barely fertile, the hybrid thus engendered not rarely becomes in succeeding generations quite prolific, often more so than either of the antecedent species. It would seem, furthermore, that these hybrids are disposed to exhibit a closer resemblance to either the one or the other parent species than to a mean between them. Curiously enough they occasionally bear little or none to either, showing a freedom quite unlocked for.

From the foregoing considerations, both environment and hybridization, acting either alone or conjointly, it would seem probable that there are forms of plants, fertile hybrids in reality, capable, when placed under special conditions of growth, of producing other forms more or less closely resembling one or the other species from which they are derived, a mean between them or a form of a different type from either. The modifications thus produced would constitute the basis of three or four tolerably well defined species, or a polymorphous species with as many varieties. Facts gleaned from a study of genus Rosa, so far at least as certain forms are concerned, seem to give countenance to the hypothesis of fertile hybrids with the revertible characters just alluded to.

To establish this hypothesis, direct experiments are far from conclusive; inferentially they have their value. It is impossible to duplicate the doings of Nature in sequestered nooks where she loves to play her pranks. Special conditions make exceptional results possible. The experiments of Herbert show that much care is needed to ensure fertility in first crosses; when once effected, each succeeding generation becomes more and more hardy and prolific. In some genera fertile hybridity is common, in others rare or impossible; in none is it produced except under special conditions, a lack of which is one reason why species preserve their individuality.

Growing out of and depending upon, in a measure, the variability met with among Roses, another difficulty is the want of accord as to the values of certain characters, some authors holding that one and some another should be considered of specific rank, basing their claims on the distinction recognized in the species of other genera. Among the more modern rhodologists, however, there are indications of a better understanding on these points, and characters that were considered half a century ago as those of the first order, are to-day relegated to an inferior rank.

Pubescence and glandulosity, once thought to be of the highest importance in a diagnostic point of view, are now regarded as little more than mere accidents of growth and may be present or absent in the same species. Rosa setigera, Michx., the only Synstylae in North America, exhibits forms that illustrate every, gradation from glabrity to tomentosity. Rosa rubiginosa, L.,. that naturalized foreigner, the under surfaces of whose leaflets are usually studded with sessile glands, may offer forms nearly or quite glandless. Through the kindness of Mr. J. H. Redfield, the genial conservator of the herbarium of the Academy of Natural Sciences, Philadelphia, I was permitted to examine its extensive collection of Roses, curious to find specimens of Rosa micrantha, Hooker, which "stands midway between Rosa canina, L., and Rosa rubiginosa" and described as being nearly or quite destitute of glands. Two specimens rewarded my search, one Rosa canina, the other Rosa rubiginosa, with glands obsolete or rudimentary!

Prickles, those acicular growths to be distinguished from the stouter ones erroneously called spines, are of little more significance than pubescence or glandulosity. These characters in fact seem to be common property among Roses, epidermal appendages whose presence or absence depends upon the location or soil in which the bush grows. The only exception is Rosa nitida, Willd, whose prickles appear diagnostic; future investigations may show, however, that in certain forms these disappear and are replaced by infra-stipular spines.

*Torr. Bull. Vol. XIV, No. 12, p. 256.

Spines are characters of the first order, not so much by their shape as by their arrangement on the stem and branches; they may be either present or absent in forms of the same species, as in Rosa humilis* Marshall; when present their position, infra-stipular or scattered, is often diagnostic.

Sepals are likewise characters of the first order. A slight lobing of the exterior is to be regarded as a passing variation from the entire form and not of specific import; quite different however when they are pinnately lobed or laciniate. The sepals by their behavior after anthesis and by their adnation to the fruit are quite significant. They may te deciduous from the apex by a clean circumcision, as in Rosa Carolina L.; through the apex as in Rosa gymnocarpa, Nutt; they may be persistent as in Rosa blanda, Ait.

The styles, leaflets, stipules, petals, fruit, time of flowering, subterranean growth, durability of bush, all furnish more or less important indications, much modified however by the vigor of growth.

†Hist. and Revis. of the Roses of N. A., Vol. XX. Proc. Amer. Acad. Arts and Sciences

The study of the North American Roses is yet in its infancy. The Alpinae. and Cinnamnmeae especially need revision. With the exception of the monograph by Dr. Watson† and M. Crépin's work preliminary to his Monographie Generale, very little of a serious character has been attempted during the last half century. M. Crépin has devoted many years almost exclusively to his favorite study, the Roses; has a herbarium unequaled, and is today perhaps the highest authority in rhodology. In his Primatiae Monographiae Rosarum,—quatriéme fascicule,—Nouvelles Remarques sur les Roses Americanies, et seq., he has given evidence of his knowledge of our Roses and his ability to deal with the difficult problems connected therewith. To his appeal for materials from this country, especially the West, American botanists should give heed. The subjoined classification is a modification of his, in which are enumerated all the native species of North America. The naturalized are Rosa canina, L., Rosa rubiginosa, L., Rosa laevigata, Michx. and Rosa bracteata, Wendl.

GENUS ROSA

SUBGENUS A.—Styles connate in one column projecting beyond the disk.
Sect. 1.—Symtylae. Sepals deciduous; spines scattered, not infra-stipular; usually not prickly.

Rosa setigera, Michx.

SUBGENUS B.—Styles separate, usually included.
Sect. L Alpinae.—Sepals persistent, usually connivent; spines often absent, when present not infra-stipular; rarely unarmed, commonly more or less prickly.

ROSA ACICULARIS, Lindl.
ROSA BLANDA, Ait.
ROSA SAYI, Schwein.
ROSA ARKANSANA, Porter.

Subsect. A. Cinnamomeae. Sepals persistent; spines when present infra-stipular; usually more or less prickly; fruit smooth or glandular-hispid.

ROSA NUTKANA, Presl.
ROSA PISOCARPA, Gray.
ROSA CALIFORNICA, C. & S.
ROSA FENDLERI, Orepin.
ROSA WOODSII, Lindl.

Subsect. B. Minutifoliae. Sepals persistent; spines infrastipular; densely armed with decidous pickles; fruit prickly; leaflets small.

ROSA MINUTIFOLIA, Engelm.

Sect. II. Carolinae. Sepals deciduous; spines usually present and infra-stipular; more or less prickly, rarely unarmed; base of calyx persistent on fruit.

ROSA CAROLINA, L.
ROSA HUMILIS, Marah.
                      "        "        var PLENA, Best.
                      "        "        var LUCIDA, Best.
ROSA NITIDA, Wittd.
ROSA FOLIOLOSA Nutt.
ROSA MEXICANA, Wateoft. ,

Sect. III. Gymnocarpae. Entire calyx deciduous; spines irregularly infra-stipular or scattered; more or less prickly; fruit small.

ROSA GYMNOCARPA, Nutt.

G. N. Best, M.D. Bibliography