so. E. KOEHLERI. 20? 



those of R. palUdus are usually small racemes. The j^etioles 

 of the latter are not furrowed, the filaments are pink, and 

 the anthers fuscous. 



Mr Borrer referred the B. affinis of Smith {Enr/. Fl. ii. 

 405) to R. Koehleri y pallidus; but the specimen so named 

 and identified with Smith's plant in his Herbarium is R. 

 rosaceus. It was gathered at Woodmancote near Henfield. 

 It appears probable therefore that Borrer's var. pallidus in- 

 cludes my R. rosaceus. Mr Edw. Eorster also considered 

 the typical form of R. Koehleri to be the R. affinis of Smith, 

 but it seems nearly impossible that Smith can have had one 

 of the Glandalosi before him when drawing up his descrip- 

 tion of R. affinis. 



The R. fusco-ater of Lindley's Synopsis, ed. 2, is shown 

 to be R. Koehleri y pallidus by the sj)ecimen from the Hort. 

 Society's Garden in Herb. Borrer. 



My R. fusco-ater y aculeatus seems to be properly re- 

 ferred to R. pallidus. Its stem and flowering-shoot have 

 very few hairs or hair-like setas, but an abundance of aciculi- 

 form ones and aciculi and prickles. All nevertheless stand- 

 ing quite separate from each other, and having much less 

 compressed bases. Its leaves are whiter beneath and more 

 felted. Its panicle is more open and more pyramidal, and 

 the terminal flowers are on shorter stalks. 



A plant gathered by Mr H. C. Watson, at Chessington 

 in Surrey, has precisely the same kind of prickles as R. 

 Koehleri, and perhaps about as many of them, but they are 

 very small and short, and therefore leave much of the cuticle 

 naked. The only leaf which I have seen has four leaflets; 

 the two on one side being those of a palmate leaf, both 

 stalked, and the basal one directed backwards so as to be 

 quite clear of the other ; on the opposite side the single 

 leaflet is dilated externally but not lobed; they are very 

 slightly felted beneath, the veins are scarcely at all hairy, 



