24 THE SUBSECTION EU-CANINE OF THE GENUS ROSA 
or very slightly hairy. Resembles sepiwm in shape of leaves and 
in styles.” 
an exceptionally thinly glandular form, but I have been unable to 
or less glandular on the whole lower surface. Equally, Mr. Baker’s 
description does not fit his own plants. Before he wrote the 
Monograph Mr. Baker evidently realized that his specimens did 
m 
lutetiana. Leaflets obovate-oblong, naked above, hairy on the ribs 
Ww ; ; copiously compound; the accessory teeth 
gland-tipped; the petioles densely pubescent and glanduloso- 
setose, and the glands often extending to the midrib beneath ; t 
bracts, stipules, and sepals copiously gland-ciliated. Fruit ovate.” 
The inference from his description is that the glands beneath do not 
extend to the side nerves, and sometimes not even to the midrib. 
Specimens which correspond most nearly with Mr. Baker’s 
description, and labelled by him R. arvatica, have elliptical (hardly 
i a it s 
spreading, fully pinnate, glandular-ciliate but not glandular on 
back. ¢ Styles glabrous. 
t 
Baker of the Monograph, but there is another plant represented 
t . : a : ; 
named by Mr. Baker, but the third is his No. 26 quoted in the 
broadly 
of that species. They are very thinly hairy on the midrib only, and 
d 
not much glandular. The prickles are small, slender, and straight, 
