Rosa lucida or virginiana (Species)

Bulletin of the Torrey Botanical Club, 47:52-53 (1920)
This has usually been regarded as a synonym of R. blanda. This may have been due partly to the fact that Miller described R. virginiana as unarmed, partly perhaps to the fact that at least one of the specimens on which R. blanda was originally based belonged to the species here treated. Another character assigned to R. virginiana by Miller, viz., "the shining leaves," does not very well apply to R. blanda as usually understood. The name R. virginiana Mill, was substituted for R. lucida Ehrh. in the New Gray's Manual by Robinson and Fernald. I therefore wrote to Professor Fernald, asking him kindly to let me know the reasons for the change made. In answer I received the following letter, which I take the liberty of publishing:

Rosa virginiana Miller, Gard. Dict. ed. 8, no. 10 (1769), is represented by a fine sheet in the herbarium of the British Museum, marked "Rosa virginiana Mill Dict. No. 10!" James Britten and J. G. Baker who called my attention to it say there is absolutely no question about its authentity. There are three fruiting branches and they are perfectly good R. lucida Ehrh. Crepin recognized it and has written on the sheet "R. lucida Ehrh. Cr." and J. G. Baker (Jour. Linn. Soc. XXXVII. 74) in his Revised Classification of Roses so treats it. I took a photograph—an excellent one nearly life-size—and it shows the characteristic broad-base and curved infra-stipular prickles at two points.

It is therefore plain that R. virginiana Mill, is the oldest name for the rose usually known as R. lucida Ehrh. To me it seems that R. carolinensis Marsh, applies better to this species than does either of the two species described by Linnaeus under the name of R. Carolina. R. rapa Bosc is apparently a double form of this species.

Mr. Best reduced this species to a variety of R. humilis. He had collected a great number of rose-specimens in New Jersey. Some of these were presented to Columbia University. These show many gradations between R. lucida Ehrh. and R. humilis Marsh, (i.e., the original R. Carolina L.), and also between these and another form, R. humilis villosa Best (R. Lyoni Pursh). Best concluded that all should be regarded as a single variable species. He has been followed by N. L. Britton and C. K. Schneider, the latter using the name R. virginiana lucida Best. I doubt if Best ever used said combination, at least in print. In my opinion several of Mr. Best's specimens are of hybrid origin, and this circumstance would give a satisfactory explanation for the intergradation, which is rarely met with elsewhere.

Rosa blanda Willmottiana Baker, according to the figure, has nothing to do with R. blanda, but belongs without doubt to R. virginiana.