DESKGLISK ON CLASSIFICATION OF SPECIES OF ROSA. 297 



very natural sections ought to have been followed by authors, correcting 

 what appeared defective in them, instead of trying to create new classifica- 

 tions, which have none of the value of those proposed by these two illus- 

 trious savans, who with good reason rejected the divisions of Linneus. If 

 it were desirable to make only two grand sections in the genus Rosa, they 

 might be established on a single principal character ; taken from the form 

 of the styles according to Desvaux ; from the leaves glandulose or gland- 

 less beneath, after Rau ; the entii'e or pinnatifid calyx divisions, deciduous 

 or persistent, after Walbroth, Loiseleur-Deslonchamps and Renter ; but 

 we must still say that this classification presents many great anomalies, 

 by interrupting in a grievous manner, the series of species which seem to 

 bind themselves together naturally. The prickles, after Reichenbach ; the 

 toothing of the leaves after Leman ; the form of the disk as proposed by 

 Dumortier; and the carpels used by Koch ; do not offer a better division. 

 We may succeed in establishing good sections, when all the species are 

 rigorously described, and known from an organographic and physiological 

 point of view. But who shall pretend to perfection when he attempts to 

 unravel the grand mysteries of nature ! 



The classification which I propose, is not, I know, more than any 

 other, beyond the reach of criticism ; and I must ask m}' readers to con- 

 sider it only as a fresh attempt to facilitate the knowledge of our French 

 species. In' selecting all that is good from my predecessors, I shall 

 endeavour to avoid errors as much as possible ; but, alas ! dare I pretend 

 to it ? I propose to divide the genus Rosa (for our French species) into 

 nine grand sections, which I think are sufficiently natural. After the 

 example of De Candolle, Lindley, Seringe, and a great number of authors, 

 I preserve the section Systi/Ice ; a very natural section, which may be easily 

 recognised at first sight, and which includes but a small number of species. 



i. Systyl^. — Styles united in a column. 



ii. Gallicanj*;. — Low shrubs ; prickles of two kinds, branches more or less 

 covered with slender prickles and glanduliferous setoe ; leaves orbicular 

 or oval, more or less coriaceous, pale or whitish beneath ; exterior 

 calyx divisions pinnatifid, canescent, not persistent on the fruit, 

 styles free or near together, but not united in a column. 



little knowledge wliicli I possess of English Roses I owe to Mr. Baker, whose 

 friendship, bora under the auspices of botcany, remains with me always as a pleasant 

 remembrance] 



