A LIST OF BRITISH ROSES 27 



SUBGROUP ACICULAT^. 



This subgroup contains species with biserrate leaflets and 

 glandular peduncles, that is, it bears the same relation to the sub- 

 group Deseglisei as the Canescentes do to that of Eu-dumetorum. 

 It contains only one or two obscure forms. 



E. CANiNA var. aciculata Eouy, Fl. Fr. vi. p. 295. I use this 

 name to cover a very peculiar form collected in Surrey. It has 

 rather small, rather nari'owly elliptical, fully biserrate leaflets, 

 thinly hairy beneath, long, hispid, solitary peduncles, ovoid fruit, 

 very strongly gland-fringed sepals, and thinly hispid or subglabrous 

 styles. Its peculiarity lies in its fine glandular and eglandular 

 acicles on the flowering branches. Its main prickles are few, 

 small, and hooked. Sudre thinks it near var. aciculata Eouy, but 

 different. I do not know the variety except from Eouy's key, 

 which does not mention the aciculate branches. The name is 

 provisional only. V.-c. 17. 



A South Devon plant [Savery), quite distinct from the last, 

 comes into this subgroup, but I am totally at a loss for a name. 

 It has suborbicular, cuspidate, biserrate leaflets, slightly hairy 

 on midribs only beneath, glandular peduncles in clusters of three, 

 somewhat gland-ciliate sepals, and glabrous styles. The fruit is 

 not formed ; the prickles are few, rather slender and arcuate, and 

 flowers white. 



GEOUP GLAUCA. 



■ In most cases the members of this group are well marked off 

 from those of other groups by their glabrous leaflets, dilated 

 stipules, especially the upper ones and bracts, short peduncles, 

 erect long-persistent sepals, and especially by their broad woolly 

 head of styles, but a considerable number of individuals may 

 lack one or more of these characteristics, notably those of the 

 subgroup SuhcanincB, in which the sepals are reflexed or not more 

 than spreading, instead of rising or suberect. Since individuals 

 of the Ganina group may often have spreading or sometimes more 

 or less rising sepals, as well as one or more of the other features 

 of this group, it is often extremely difficult to draw the line 

 between them. The style characters are the most reliable. It 

 is rare in the South of England, becoming more frequent in the 

 North, and common in Scotland, where a large numljer of forms 

 occur which have not been satisfactorily identified. I have treated 

 the group at somewhat greater length than the foregoing ones, 

 but fear it is not at all well elucidated. It will be seen that it is 

 in as much need of revision as the subsection Villosce. 



SUBGROUP REUTERI. 



This contains those species and varieties of the group which 

 have their sepals suberect or at least rising above the disc. They 

 may be subdivided either into those with smooth or hispid 

 peduncles, or those with uniserrate or biserrate leaflets. I have 

 followed tlie former method, but if tlie subgroup prove a large 

 one, as I expect will be tlie case when the Scottish forms have 



