tis Savi may perhaps be referred to R inodora 
Species that has 6 
136 THE BRITISH ROSES 
1 
continental ones, so that this subgroup is likely to become an 
unwieldy one. ‘There are, for example, both glandular and 
The subgr 
glabrous-leaved and glabrous-styled species, viz. R. fetida Bast., 
- scabriuscula Sm., “RB. Jundzilliana B non +» an 
- Lomentosa var. sylvestris Woods. To these must be added, I 
think, F. confusa Pug., which is the best name that can be given 
& 
w 
c4>) 
Rn 
n 
= 
ubsection Rubiginose.—This is a fairly natural and distinct 
subsection, though its less glandular members sometimes make a 
e 
will mark that off. The rest 
aggregate L. Hglanteria Linn., consisting of a series of forms not 
readily separable. 
The group of R. elliptica in Britain is reduced to R. Billietit 
Pug. which I do not think has been found except at Annesley, 
and f. cryptopoda Baker, which is suspiciously near the R. cori- 
folia group, 
The group of R. micrantha is formed mainly of the inseparable 
Species L. micrantha Sm. - permixta Déségl., which are 
generally distributed and common on caleareous soils. R. hystrix 
is @ rarer species, which has well-marked features, while 
m. is a 
a, = aro Déségl. and R. micrantha var. Briggsii Baker are 
ver 
ery local. 
The group of R. “grestis consists of R. agrestis Savi and 
f. sepium Thuill., which are ractically synonymous, and in the 
aggregate form a somewhat rare Briti species, to which R. bel- 
nensis Ozan. approaches as a more hairy-leayed form. Much of 
our material which has been labelled R. sepiwm Thuill. or R. agres- 
3 Fries, which is a 
as been much misunderstood in Britain. 
