OF ARTS AND SCIENCES. 329 



Donn also, at about the same time, in his Catalogue oC the Cambridge 

 Garden iu England, gave some additional names to already knowu 

 species ; his H. Jlorida being R. Carolina^ the R. Cherokeensis the 

 same as R. Icsvigata, aud his R. fenestrata the smooth form of R. 

 setigera. 



Sir James E. Smith in his revision of the genus in Rees' Cyclo- 

 pedia (1816) describes fifty-seven species, of which eleven are cred- 

 ited to America ; but these include R. gemella, R. Lyonii^ R. pendulina, 

 and J?. Icevigata, and R. rublfoUa as well as R. setigera, — thus leav- 

 ing but six actual native species. Under R. Carolina (" common in 

 our shrubberies ") he mentions " R. Pennsijlvanica of the gardens." 



In 1817 R. MontezumcB, HBK., discovered by Humboldt and Bon- 

 pland in the high mountains of Mexico (but in all pi'obability only a 

 naturalized form of the European R. canina) , was described by them, 

 and also figured by Redoute in Les Roses. This latter work, pub- 

 lished between 1817 and 1824, the text by Tliory, included figures 

 of R. luc'da, R. parvijiora (double), R. rubifolia, R. Carolina, and 

 of several forms of the latter under the name of R. Hudsoniana. 

 R. hlanda is referred to R. cinnamomea as a variety. Two other 

 publications illustrating the genus appeared early in tlie century, which 

 have not been accessible to me; viz., Roessig's Die Rosen (1802-1820), 

 containing figures of R. Carolina and '' R. Virginiana simplex" and 

 Andrew's Roses (1805-1828). 



Of far more value is the Rosarum Monograpliia of Lindley (1820), 

 in which both the grouping of the species and the descriptions are 

 more satisfactory than anything that had preceded, and are accom- 

 panied by several figures that are in most cases excellento He ad- 

 mits fourteen American species, viz. 7?. nitida (or " Dwarf Labrador 

 Rose"), R. rapa ("as distinct from R. lucid/t as a species can be;" 

 he speaks of specimens from the Southern States, probably R. humilis), 

 R. lucida and R. laxa, Lindl., R. ■parvif.ora ("a universal favorite"), 

 R. Woodsii, Lindl., R. Carolina, R. hlanda and R. fraxinifolia (the 

 smooth form of 7?. hlanda), R. sfricta (the same as the R. pendidina 

 of Linnseus), R. ruhifolia and 7?. setigera, without perceiving them 

 to be the same, R. Icevigata (believing it a native of Georgia, he tries 

 to distinguish this from R. Sinica), and R. Montezumce. His R. laxa 

 appears, from his figure and description, to be a glaucous garden form 

 of R. lucida. The R. Woodsii, said to be a native of the country 

 near the IMissouri, was described from a cultivated specimen. It had 

 previously been sent to France by an English nurseryman as an Amer- 

 ican rose with black and yellow flowers, and had io been advertised 



