18 A LIST OF BRITISH HOSES 



[B. firmtda God. Fl. Jur. Suppl. p. 71. A very ambiguous 

 species, said by Godet to be synonymous with li. mucronulata 

 D6segl., though he describes it. as biserrate. It should also have 

 straight prickles and roundish ovoid fruit. Both Dingier and 

 Sudre agree to place a Surrey and a Cheshire specimen to this, 

 though Dingier thinks it a doubtful species. Botli my specimens 

 have very hispid, almost woolly styles, but, in spite of this, I think 

 they might be included in B. sylvulanim.] 



K. CANiNA var. parisiensis Eouy, Fl. Fr. vi. p. 288. This is also 

 very near B. sylvularum, but is described by Eouy as having the 

 habit of B. senticosa Ach., wuth biserrate leaflets. I have a 

 specimen from Cheshire agreed to by Sudre. It has very small, 

 very strongly biserrate leaflets, narrowed at each end, rather 

 short, curved, subconical prickles, ovoid fruit, and glabrous styles. 

 A Herefordshire specimen {Leu) n^atches it very closely, and one 

 from Bucks [Driice) probably belongs here. V.-c. 24, 36, 58. 



SUBGROUP ANDEGAVENSES. 



This is a small subgroup, and though my specimens have been 

 referred to at least fifteen different names, they can probably be 

 reduced to less than ten, none of which are common, and several 

 appear to be quite rare. In the aggregate its members differ 

 from those of the last three subgroups by the presence of stalked 

 glands on the peduncles. They may best be divided into those 

 with simply and those with doubly serrate leaflets, though an 

 intermediate subdivision might be made corresponding to the 

 TransitoricB. 



Leaflets Uniserrate or nearly so. 



E. andegavensis Bast. Fl. Maine & Loire, p. 180. As a 

 segregate, i. e. the form with hispid styles, this appears to be 

 much less frequent than the next ; in fact, I only have two speci- 

 mens of it, named by Sudre. They are from adjacent stations in 

 Surrey, but I have seen it certainly from Northants and Wilts, 

 and though not common, I have little doubt that it is widely dis- 

 tributed. V.-c. Wilts, 13, 17, 32, 40 ? 



E. AGRARiA Eip. ex D6segl. Cat. Eais. p. 181. This may be 

 regarded as a glabrous-styled form of B. andegavensis, and as such 

 is only varietally distinct. Its leaflets are said to be broader, but 

 both broad and narrow-leafleted forms occur in both. V.-c. 3, 15, 

 16, 17, 36?, 57, 58. 



\B. edita D6s6gl. Cat. Eais. p. 178. I have a specimen from 

 Cheshire for which this name is suggested by Dingier, and to 

 which it is referred by Eouy's key, but it does not in the least 

 resemble B. surculosa, as a form of which the species was founded. 

 It has medium-sized suborbicular leaflets, very pubescent petioles, 

 an occasional glandular seta on one or two of the peduncles, sub- 

 globose fruit, and thinly hispid styles, colour of flowers unknown. 

 Sudre refers it to B. pubens D6s6gl. & Ozan., of which I know 

 nothing but the description, with which my specimen does not 



