(1962)  Page(s) 216-219.  Includes photo(s).
 
(Contn'd)
Specimens of R. rapa from Lindley's herbarium are shown in Figure 1. The specimen at the right is evidently a root shoot, which Lindley described as, "very red, very densely covered with very unequal scattered crimson prickles: of these the largest are compressed and falcate, as they decreasi insize becoming gradually straighter until they change into setae." The branches were described by Lindley as "red, either unarmed, or furnished with a few, weak, pale, setiform prickles, now and then decreasing into setae." The specimen at the left has unarmed stems, oval leaflets, glandular-hispid pedicels, and large, glandular-hispid, hemispherical hypanthia as decribed by Bosc. Specimens of R. rapa from the garden of Dupont and a specimen from the garden of Lyell are shown in Figures 2 and 3. These specimens are in the de Candolle herbarium. They are similar to Lindley's specimens.
Loiseleur-Deslongchamps, Seringe (1825), and Torrey and Gray (1838) regarded R. rapa as a variety of R. virginiana. Koch (1869) considered t to be a cross R. virginiana and one of the varieties of R. chinensis. Dippel (1893) thought that it was a cross between R. carolina and R. gallica or R. chinensis. In 1872, the Belgian rhodologist, Crépin, published his revision of the roses in Willdenow's herbarium. Among three specimens therein, which had been identified by Willdenow as R. virginiana, there was one with double flowers. Crépin rejected Willdenow's identification of this double-flowrred rose, and suggested the possibility that it was a hybrid in whose production R. virginiana might have been involved. In his detailed description of the rose, Crépin noted...that the stems were unarmed, and that the hypanthia were short and campanulate. The leaflets resembled those of R. virginiana, but were thinner, and did not appear to have been lustrous. A large corolla, about two and one-half inches in diameter, extended beyond long sepals. In 1876, Crépin remarked that a horticultural variety of R. virginiana, sometimes called 'Multiplex', was being grown. "This variety, 'Multiplex'," said Crépin, " is similar to the double-flowered specimen of Willdenow's herbarium, and it is entriely possible that it is the same rose that has been described under the name of R. rapa. Rehder may have been referring to the opinion of Crépin, when he wrote in 1903 that Rosa rapa is supposed to be a double-flowered variety or perhaps a hybrid of Rosa lucida [virginiana]. In 1919 Rehder gave Rosa rapa Bosc as a synonym of "Rosa virginiana var. plena Hort." Mulligan (1951) also gave R. rapa as a synoynm of R. virginiana plena, and Rydberg (1918) gave it as a synonym of R. virginiana.
On the basis of the description of Bosc, and those of other authors, and the herbarium specimens of Lindley, and of De Candolle, it appears that R. rapa differs from R. virginiana in the following respects: 1. The stems do not have infrastipular prickles; 2. The leaflets are oval; 3. The stipules are markedly dilated; 4. The hypantia are large and hemispherical-turbinate; and 5. The flowers are double. Because of the variability of this plant, Rehder, Rydberg, and Mulligan erred, in my opinion, in making R. rapa a synonym of R. virginiana or R. virginiana plena. R. virginiana is a homomorphic species, not subject to this amount of variation. On the other hand, the related species, R. carolina, is exceedingly variable, it is possible for a variety of R. carolina to have the characters of R. rapa. It is proposed, therefore, that R. rapa be reclassified as a synonym of R. carolina var. plena (Marsh.) D. Lynes (1955) [R. carolina f. plena (Marsh) W.H. Lewis (1958)].

(1962)  Page(s) 213-216.  Includes photo(s).
 
In 1809, the French naturalist, Louis Augustin Guillaume Bosc included in a monograph on the genus Rosa seven American roses, namely: 1. Le Rosier à Épines rouges, Rosa rubrispina Bosc; 2. Le Rosier luisant, Rosa lucida Ehrh; 3. Le Rosier à petites Fleurs; 4. Le Rosier de la Caroline; 5. Le Rosier en Corymbe, Rosa corymbosa Ehrh.; 6. Le Rosier de Pennsylvanie; and 7. Le Rosier Turneps, Rosa rapa Bosc.
Le Rosier de la Caroline and Le Rosier en Corymbe had been observed by Bosc in the marshes of Carolina, and plants of the former were grown by Dupont from seed collected by him. The others he described from French cultures.
The first species, Rosa rubrispina Bosc, may be confidently referred to as Rosa nitida Willd., published in the same year; Rosa lucida Ehrh.is a synonym of Rosa virginiana Mill.; Le Rosier à petites Fleurs is Rosa parviflora Ehrh., now treated as a synonym of Rosa carolina L. 1753; Le Rosier de la Caroline is Rosa palustris Marsh. (Rosa carolina L. 1762);  Rosa corymbosa Ehr. is a synonym of R. palustris; Le Rosier de Pennsylvanie appears to be a variety of R. palustris with glabrous hypanthia and pedicels.
The identity of R. rapa Bosc is uncertain. It has been variously treated as a species, as a hybrid, as a variety of R. virginiana, and more recently as a synonym of R. virginana or R. virginiana plena...In naming R. rapa, Bosc used the latin equivalent of its common English name, 'Turnip Rose'. Apparently, he was unaware that the same rose had been described in 1807 by Persoon under the name of R. turgida. Although R. turgida is the legitimate name, Desfontaines (1815), in his catalogue of Jardin du Roi, used Bosc's name, and made R. turgida Pers. a synonym. Subsequent authors, without exception, used the name R. rapa.  In 1802 this rose was recorded as R. fraxinifolia by Dumont de Courset, with the comment "très belle suivant Dupont", and it was described by him in 1811 under the same name. 
Persoon gabe the habitat of R. rapa as *Scotia?". Dumont de Courset (1811), Desfontaines, Poiret (1816) and Trattinnik (1823) followed Persoon in giving Scotland as the country of origin, but did not express doubt. Bosc said that it is probably of American origin, and Spach (1834) said that it grows in the southern Unites States. According to Lindley (1820) it was collected in the southern Unites States by Fraser. Between 1785 and 1807, John Fraser, a native of Scotland and resident of London, made seven trips from England to the United States for the purpose of collecting plants and seeds (Hogg 1852).
Bosc's description of Rosa rapa states that "it has very large hemispherical hypanthia, covered with scattered pedicillate glands, as are the peduncles; the stems are armed with few prickles, ans sometimes naked; the leaves composed usually of seven oval, dark green, glabrous, shining leaflets; the flowers red, slightly fragrant, one and one-half inches or more in diameter. It is probably originally from America and flowers at the end of spring. Its hypanthia approach the size those of the turbinate rose [R. francofurtana] and its leaves those of the preceding [R. virginiana]; however, they are neither as shining or as coriaceaous. It is therefore inappropriate for it to have been given the epithet lucida [virginiana] in some collections; the English name [Turnip Rose] that I have adopted fits it better; it is cultivated with single, semi-double, and double flowers.It is a very beautiful species which is scarcely more than two or three feet high...."
The immature hypanthia were described by Lindley as cyathiform, by Bosc and Trattinnik (1823) as large and hemispherical, by Persoon and Prévost (1829) as turbinate-hemispherical, and by Dumont de Courset (1811), de Pronville (1818), and Thory (1821) as turbinate. The hips of R. rapa were described by Thory and Trattinnik as globose, by de Pronville as turbinate, by Loudon (1854) and Spach as hemispherical, and by Lindley as nearly hemispherical, with a very wide mouth filled up by densely villous styles. The resemblance of the hip to the shape of a turnip is implied in the horticultural name, 'Turnip Rose'. Loudon called it the 'Turnip-fruited Rose'.
The stems of R. rapa were described by Dumont de Courset (1811) and Prévost as unarmed, by Thory as unarmed except for a few prickles on the mature wood, and by Bosc and Lindley as either unarmed or furnished with afew prickles. 
According to Bosc R. rapa was grown ingardens with single, semidouble anddouble flowers. Loiseleur-Deslongchamps (1819) knew it only as semidouble, and de Pronville and Lindley only as double. Loudon said that only the double-flowered form is grown in England.
Poiret (1816) thought that R. rapa resembled R. francofurtana in some respects. Persoon remarked that R. rapa is not unlike R. francofurtana, but its hypanthia are not constricted in the middle, and its leaflets are not pubescent, as in the latter species. De Pronville said that R. rapa bore some resembalnce to R. virginiana, and Dumont de Courset (1811) said that it resembled R. virginiana at first glance.