36. R. HUMIFUSUS. 241 



former E. humi/usus, from Inverarnan, to this species. It 

 is not the H. hirtus (Weilie), as figured and described in the 

 above-mentioned work, whicli seems to be very nearly allied 

 to R. Bellardi, and will be found, together with that plant, 

 placed under R. glandulosus in this essay. 



Mr Lees' R. pallidiis (3 foliosus is apparently a form of 

 this species. My specimens of it were gathered at the foot 

 of the Great Doward Hill, Herefordshire. Its stem is much 

 more prickly ; its panicle very long, leafy near the top, with 

 large and mostly simple leaves. I suspect that the plant 

 was in an unnatural condition, caused probably by pecu- 

 liarity of situation. The panicle of my specimen was 

 gathered when too young. 



Mr Baker finds an interesting plant on the hill side 

 above Byland Abbey in Yorkshire, which I refer to this 

 species ; but M. Genevier says it is very closely allied to 

 R. offensus (Miill.) of which I cannot find any account ; 

 it has ternate leaves and thus confirms my opinion that 

 R. humi/usus is one of the Bellardiani. Mr Baker considers 

 it to diflfer from R. humi/usus by its " more hairy leaves and 

 adpressed sepals." I do not consider the former difi'erence 

 of much consequence, indeed some of my specimens of 

 R. humi/usus have quite as hairy leaves ; and the sepals 

 are loosely adpressed. He finds another plant near the 

 same phice, which M. Genevier names R. saxicolus Miill., 

 and which I consider as liardly differing from our R. humi- 

 /usus in any respect. 



I now refer the plant mentioned in my Synopsis and in 

 Flora Hert/ordiensis as R. liorridissimus to the present 

 species. 



Garke and Sonder combine R. humi/usus with R. pyg- 

 mceus (Glinth.). I have no practical acquaintance ^^'ith that 

 plant, but Metsch states that he has examined the original 

 specimens of Weihe, and finds that R. humi/usus is certainly 



21 



