OP ARTS AND SCIENCES. 335 



found in the mountains of Coaluiila, Mexico, a rose that was de- 

 scribed the next year in my list of his collectiou under the name of 

 R. Mexicana. 



The material at hand in the preparation of the following revision, 

 while none too ample for correct results, has been enough at least to 

 furnish all that was desired of dubious and difficult forms. The col- 

 lections that have been examined, besides those of the Gray Herbarium, 

 have included the large accumulations of Dr. Engelmann, those of the 

 Torrey Herbarium, of the Philadelphia Academy of Sciences, of the 

 Department of Agriculture at Washington, and of the Arnold Arbo- 

 retum, besides the private collections of J. H. Redfield, of Philadelphia, 

 H. G. Jesup, of Hanover, N. H., and J. Donnel Smith, of Baltimore, 

 and contributions from Dr. A. Gattinger of Nashville, Dr. C. Mohr 

 of Mobile, Howard Shriver, of Wytheville, Virginia, and Warren 

 Upham of IMinneapolis. 



Class fjlcatton, Synopsis, and Descriptions of Species. — Th^e differ- 

 ences which have to be taken into consideration in dotermininw the 

 species of the genus Rosa are so variable, and present such a multitude 

 of combinations, that there are few genera whicli illustrate more fully 

 the different views that can be taken by different botanists respecting 

 the specific value of the same characters. The number of species ad- 

 mitted by Linna?us in 17G2 was fourteen ; by Sir J. P^ Smith in 1816, 

 fifty-seven ; by Lindley in 1820, seventy-eight ; by Seringe in 1825, 

 ninety-one, besides fifty-one which he classed as imperfectly known. 

 Bentham & Hooker, in the Genera Plantarum (1805), limit the num- 

 ber of species then known to thirty. The number of species credited 

 to Great Britain by Lindley was ten; by Baker (1871), eleven; by 

 Hooker (1871), seven, with six subspecies; and by Bentham, in the 

 same year, five " probably real species.'' Deseglise in 1876 makes 

 the whole number of Old World .''pecies 410, of which 323 are 

 European (66 in Great Britain), To Asiatic, and four African, — eight 

 being of uncertain habitat. Crepin's revision of the European species 

 (1869) accords in the main, at least provisionally, as respects the 

 weight and number of species, with Deseglise. Nyman in the Con- 

 spectus Florce European (1878) enumerates forty European species, 

 with fifty-two subspecies. Kegel in his Tentnmen (1878) recognizes 

 a total of fifty-six species, of which seventeen occur in Europe, thirty- 

 four are Asiatic, and five American, several of our species being re- 

 ferred to foreign types, as we have seen. On the other hand, Gandoger 

 {Tahulce Rhodologice Europceo-OrientaJes, 1881) distributes the Old 

 World forms into twelve genera and 4,266 species. 



