208 THE JOURNAL OF BOTANY 



station seems an unlikely one for B. media or B. marina, and they 

 would not be surprised if it proves to be B. rupestris. As a matter 

 of fact, the specimen belongs to B. media, but it differs from the 

 type not only in its place of growth, but in having a woody root- 

 stock; and the pedicels, instead of having glabrous calyces, are 

 distinctly glandular. 



Geranium purpureum Vill. Dry sunny banks, St. Ouen's, Jersey, 

 21st May, 1899. This plant seems to correspond well with the very 

 full description of G. purpureum Vill. in Lowe's Flora of Madeira, 

 except that the carpels are downy, whereas Lowe describes them as 

 smooth. The Rev. R. P. Murray tells me that it is exactly the 

 form which is abundant in Portugal. It is distinguished from 

 (t. Eobertianum L. by (1) the erect habit ; (2) the absence of the 

 villous hairs, and consequently of the very characteristic odour of 

 (r. Robert ianum : (3) the smaller size of the parts of the flower. 

 What relation does this bear to the G. purpureum auct. angl. of the 

 London Catalogue.^ In Brebisson's Fl. de Normandie two species 

 (besides G. Robertianum) are given: (1) G. minutijiorum Jord. = 

 G. purpureum Vill. pro parte, of which G . modestum Jord. is put 

 down as a variety ; (2) G. Lehelil Bor., the description of which 

 will not tit the Jersey plant. In the Student's Flora, G. purpureum 

 Vill. = G. modeatum. Jord. — G. Lebelii Bor., and according to 

 Reichenbach G. Rail Lindl. also = G. purpureum Vill., though 

 English authorities seem to regard it as a ** shaggy" maritime 

 form of G. lucidum L. In Smith's Fnt/lish Flora, Geraniwn lucidum 

 saxatile, foliis Geranii Robertiani (an excellent description of the 

 Jersey plant) is given as a var. of G. Robertianum L. Can anyone 

 disentangle these synonyms? In Lloyd's Flore de V Guest de la 

 France, G. purpureum Vill. is the only species given besides 

 G. Robertianum, and it is made to include G. modestum and 

 G. ininutijiorutn Jord. — L. V. Lester. "This is the plant we 

 call purpurewa in Britain. It differs from G, purpureum Vill. in 

 having downy carpels." — E. F. Linton. " G. purpureum Vill. is 

 distinguished from G. Robertianum by its shorter and narrower 

 petals, and its carpels being more closely or thickly rugose. G. 

 modestum Jord. is a form of G. pur pur emu with a less hairy calyx. 

 G. minutiflorum Jord. is a southern maritime var. of G. purpureum.'' 

 W. R. L. 



Rosa pimpinellifolia x canina = R. hibernica Sm. var. glabra 

 Baker. Hedges near Hoylake, Cheshire, 5th August, 1899. I am 

 not at all sure that I ought not to have labelled these R. pimpinelli- 

 folia X (jlauca. In either case it is a good example of how a hybrid 

 may exceed either or both of its parents in frequency. I saw only 

 three or four plants of R. pimpinellifolia L., and those not within a 

 quarter of a mile of the hybrid, but I only searched a portion of the 

 coast sandhills, where it probably grows. Canina forms were also 

 few, and I saw no glauca or subcristata at all, though the latter is 

 stated to be frequent in the district in the Flora of Cheshire. The 

 hybrid is so abundant as to fill many of the hedges, and, except one 

 bush of R. Doniana, or possibly R. Robertsoni, it belongs exclusively 



