30 A LIST OP 15KITISH ROSES 



{Waddell, Wats. Exch. Club, 1905) is named var. mutata Burn. & 

 Grem. by Sudre, while Dingier thinks it near var. ocnensis Kell., 

 though it has no subfoliar glands. I can find no satisfactory 

 name for a form with very small leaflets from Killin (Barclay). 

 V.-c. 36 '?, 58, 65, 81, 88 ?, 98, Co. Down ?. 



E. GLAUCA var. cenensis Kell. in Asch. & Graebn. Fl. Mit- 

 teleur. vi. p. 194, non Kern. I introduce this name tentatively, 

 as the representative of a subdivision of this subgroup made by 

 Keller to contain species with hispid peduncles and biserrate 

 leaflets with subfoliar glands on the primary nerves, homologous 

 with B. stephanocarpa in those with smooth peduncles. A speci- 

 men from Salop {Ley) belongs here, but the Co. Down form men- 

 tioned under B. fugax, having no subfoliar glands, can hardly 

 belong to it. V.-c. 40. 



E. Hailstoni Baker, in Eept. Bot. Exch. Club, 1867, p. 7. I 

 can add nothing to my remarks in E. p. 96, and though I strongly 

 suspect it to be merely an abnormal form, I retain the name until 

 further light can be thrown on it. V.-c. 62. 



SUBGROUP SUBCANIN/E. 



This subgroup, according to Keller's arrangement, contains 

 just one-third of the names in the whole group, and covers quite 

 a large range of varieties, which he classifies primarily, as he 

 does the first subgroup, into those with simple, slightly double, 

 and fully biserrate leaflets, with their subdivisions of those with 

 smooth and those with hispid peduncles. I think w^e have repre- 

 sentatives of all three primary subdivisions, though I propose, for 

 the present, to disregard that with slightly biserrate leaflets. I 

 have not seen British forms with hispid peduncles. Beyond a 

 suggestion by Dingier that one or two of our British roses belong 

 to the subgroup, I have no names of any of its members from 

 Sudre or Dingier. As already stated, I confine the subgroup to 

 forms in which the sepals do not rise above the disc, but Keller 

 and Dingier admit those in which those parts are almost sub- 

 erect, thus completely destroying any slight dividing line which 

 might be drawn between the two subgroups ; -indeed Keller's 

 definition would admit all Canina forms with glabrous leaflets and 

 somewhat spreading sepals. Mr. Barclay informs me that a con- 

 siderable number of Perthshire forms are probably referable to this 

 subgroup, but I have only seen about a dozen from Britain. 



Leaflets Uniserrate. 



E. Eeuteri var. subcanina Christ, Eos. Schw. p. 169. Christ's 

 definition confines his variety to plants like typical B. Beuteri but 

 with reflexed sepals, but for the present I include also those which 

 have them spreading. My specimens vary in the shape of their 

 fruit from globose to obovoid. The Co. Antrim specimen {Waddell) 

 belongs doubtfully to the group, its only distinguishing feature 

 being its woolly styles, which may occur in the Canina group ; its 

 long peduncles, reflexed sepals, and narrow stipules and bracts 



