Mr. Woows on the British Species of Rosi. 220 
are R. canina, R. systyla, and R. arvensis; and from each of these 
it may perhaps be difficult to give a description which shall accu- 
rately distinguish it, while in habit it is considerably different 
from either. From the first it may however, T think, always be 
known by the porrect styles, the entire pinne of the calyx-leafits, 
the peduncle almost always furnished with hairs or sete, the 
shape and flatness of the leaflets, and the strong and hooked aculei 
of the footstalk. ‘These marks seem indeed amply sufficient, but 
I am afraid they are all more or less uncertain. I have never 
seen the glands of the peduncle extending themselves on the re- 
ceptacle or calyx ; in R. canina, when glands are found on the pe- 
duncle, they are also generally to be observed on the fruit, and still 
more on the calyx; but this character likewise sometimes fails, 
A better distinction in the living plant is found in the enormous 
surculi covered with beautiful blue wax, and bearing great cymes 
of flowers. In the most favourable circumstances it is only by 
accident that R. canina has more than four flowers. In this plant 
if any surculi are produced, and it is rarely without them, the ob- 
server will not often be disappointed in searching for eight or ten, 
. and he will sometimes find double that number; but even this 
mark is not very decidedly exhibited in the variety 8, which 
seems however to unite better with this species than with any 
other. From R. arvensis it may be known by the styles, which 
are here hairy and but just protruded, not smooth and collected- 
into a long cylinder, as in that plant. It is also a much more up- 
right plant, the surculi being rather erect than decumbent. From 
R. systyla also a due attention to the styles will distinguish it; 
and the shape and flatness of the leaf give a decidedly different 
appearance to the present plant. 
25. Rosa 
