8 THE SUBSECTION EU-CANINA OF THE GENUS ROSA 
being arranged in an elongated cone. Again, R. coriifolia and 
R. glauca may be distinguished from R. canina and R. dumetorum 
by the head of stigmas being broad and hemispherical, as well as 
by the styles being woolly, the stigmatic head of R. canina and 
. dumetorum being narrower and flatter, and the styles usually 
. It may be 
» because after that the 
form of the fruit changes but little, while before the fall f the 
to demolish all the work that has been done hitherto and merely to 
retain a few species as varieties or forms, I have endeavoured 
adorn them with specific names. These varieties, which I have 
studied for twenty beer are innumerable, and there are almost as 
Sie 
and Hieracia, I do not despair of Someone in the future profiting 
a little by this paper in bringing our Rose list into line with those 
on the Continent. 
the authors I have consulted, the two whose writings 
throw the best light upon British Roses are Crépin and Déséglise, 
but the works of Rouy and of Keller are also of great interest as 
representing the most modern views, and may prove of assistance 
in the arrangement of the Species, if not in the determination of 
the varieties. I have also made use of Ripart’s classification, which 
