ROSA CANINA. 101 
yet on such differences, which in less variable genera 
would scarcely have been trusted, have writers on 
Roses attempted to establish their species. Pubescence 
has received much attention ;‘on its absence, presence 
and quantity R. collina, dumetorum and canina of 
authors and bractescens of Woods are divided from 
each other. Yet examine for a moment R. canina, as 
it grows in every hedge. A careful observer will pre- 
sently discover on the same plant some leaves entirely 
naked, and others in which the midrib and primary 
veins of the under surface are clothed with hairs. Here, 
then, is the first approach to pubescence, which, be- 
coming inereased in quantity, distinguishes. R. collina ; 
this usually has hairs on the upper surface of its young 
leaves, but none on the old ones. In R. dumetorum 
there is a further increase of pubescence, which then 
covers both sides of the leaves, and, becoming very 
dense and permanent, forms R. bracteseens of Woods. 
The distinction between simple and double serratures 
in this species I confess myself unable to understand. 
I have attempted to draw a line of separation between 
them, but without success. They have no limits; for 
no one can always say whether the serratures of a par- 
ticular leaf are simple or double. But the value of 
these and similar characters has been already dis- 
cussed. It is therefore unnecessary now to extend their 
examination. 
The foregoing description applies strictly to R. ca- 
nina « When this is weak and grows in woods or 
shady places among grass, it has straight prickles and 
becomes R. nuda of Woods; with very distant aculei 
it is R. andegavensis of Batard; with very dense ‘ones 
it is R. canina § of Rau. The stem is slightly setige- 
rous in a plant in Mr. Sabine’s garden. Another of the 
same collection has the leaves bipinnate. The leaflets 
are dull in R. sarmentacea and canina B of Woods; 
much rounded, with blunt serratures, ma plant from 
Mr. Lyell; irregularly serrated in R. sarmentacea of 
Woods and affinis of Rau; pubescent on the upper sur- 
face in affinis. The sepals are assurgent, and disposed 
