39. R. BALFOURIANUS. 2o9 



leaflet. On the wliole I tliink it best to place these plants 

 with 7?. Balfonrianus, and not to attempt to se[)arate them 

 even as varieties until we are better acquainted with them. 



I possess a specimen of the R. cccsius 8 nudatus of Lees 

 (in StpeWs llandh. 54) obtained from Leighton's Herbarium, 

 to which it was given by Mr Lees. It grew at Henwick 

 near Worcester, and very much Resembles some states of li. 

 coryUfoUus y purpuj'eus, but seems to possess the characters 

 of Ji. Balfunrianus. 



M. Questier sent the R. tenuiarmatus (Lees) with the 

 name of R. Bdlfourianus. His specimens agree well with the 

 authentic plant of Lees. I also place here some specimens 

 called R. nemorosus by M. Questier and myself. A speci- 

 men received as R. dumetoruni from Mr Lange, gathered 

 at Apenrade in Sleswig, is exactly the R. Balfom-ianus. 



The above remarks will show that this is a very variable 

 species which may ultimately require such division : but the 

 series seems complete from typical R. Balfourianus to 

 typical R. tenuiarmatus. 



I am quite unable to conjecture the reasons which 

 caused Dr Bell Salter to join this plant to R. Salteri^ with 

 which it seems to have very little in common. 



Mr Borrer obtained a specimen of this species from the 

 Horticultural Society's Garden as the R. vulgaris of Lindley 

 who quotes it. corijUfoUus (Sm.) as a synonym in his second 

 edition of the Sijnopsis. The R. vulgaris of his first edition 

 is R. villicaulis. 



I refer the R. coryUfollus of Johnston (E. Bord. G2 and 

 fig.) to R. Balfourinnus with some slight doubt. He held 

 the opinion that his R. corylifolius " is apparently diflferent " 

 from my plant so-named. His specimens and description 

 agree in most respects with R. BaJfourianiis. As the 

 specimens are in flower it is not possible to determine from 

 them the condition of the fruit-calyx, and it is only on the 



