334 PROCEEDINGS OF THE AMERICAN ACADEMY 



Considering the limited amount of material at his command, his work 

 is excellent, though as yet in a measure tentative and provisional, and 

 witii a tendency to multiply species. He has examined critically all 

 the specimens of Willdenow's original herbarium, and has clearly de- 

 fined the characters by which the early species are to be distinguished, 

 and he was the first to perceive and indicate the limited range of 

 /?. lucida. His disposal of the northern and western species may be 

 tliought less satisfactory. He places under R. hlunda all the pubes- 

 cent and acicular and even the spiny western forms which correspond 

 to Lindley's R. Woodsii and to the R. Arkansana of Porter, while the 

 resinous-pubescent and acicular form he prefers to consider a variety 

 of R. acicularis, though at first thought distinct and named 7?. Bour- 

 geauiana. He revives Presl's R. Nutkana^ but would separate two of 

 its forms under the names R. Aleutensis and R. Durandii, the latter 

 based upon some peculiar specimens of Hall's collection that had been 

 referred by Dr. Gray to R. Kamtschatica. Finally, upon a specimen 

 collected by Fendler in New Mexico he proposes, with some doubt, 

 a new species as R. Fendleri. His general arrangement of our species 

 is under the following sections : — 



1. Synstyl^. — R. setigera. 



2. Alpine. — R. acicidaris and blanda. 



3. CiNXAMOME^E. — R. Nutkana, Durandii, Aleidensis, Californica, 

 and Fendleri'? 



4. Cauolin.e. — R. Carolina, lucida, nitida, parvijlora, and folio- 

 losa. 



5. Gymnocarp^. — R. gymnocarpa. 



A still later revision is that of Kegel of St. Petersburg, in his 

 Tentamen Rosarum Monograpldcn (1877). He retains R. hianda, 

 R. Carolina, R. lucida, R. nitida, and R. Woodsii. To R. Woodsii 

 he refers R. Nutkana, R. Lyonii of Pursh. and some Asiatic forms ; 

 and to ^. Carolina the R. Californica of Cham. «&; Schlecht., propos- 

 ing another R. Californica, Kegel, on specimens collected near San 

 Francisco by Tiling. He refers to R. acicidaris, var. Gmelini, Cre- 

 pin's var. Bonrgeauiana ; R. gymnocarpa to R. pimpincllifolia ; 

 R. fraxinifolia to R. cinnamomea ; R. setigera to R. moschata, but 

 R. rnhifolia to R. repens ; and R. MontezumcB to 7?. canina. 



In 1880 R. minutifolia, a very peculiar species from Lower Califor- 

 nia, was described by Engelmann in Coulter's Gazette, and R. spitha- 

 mcBa by myself in the Botany of California, the latter probably only 

 an extreme form of 7?. Californica. And last of all, in 1881, Pahiier 



