102 ROSA CANINA. 
to become persistent in Mr. Winch’s glaxcophylla; 
nearly simple in R. canina 3 of Woods; and glandular 
on the outside in glaucophylla. The disk is flat in R. 
surculosa 3 of Woods. The fruit is nearly round in R. 
canina < and sarmentacea 5 of Woods; at the same time 
very small in R. teneriffensis of Donn; and rough in 
R. canina y of Woods and andegavensis of Batard. In 
the latter the peduncles are hispid. Such is the most 
common R. canina and that with which Linnzeus was 
best acquainted. The Teneriffe plant is very impatient 
of cold, flowers sparingly, and produces little mis- 
shapen fruit. Mr. Winch’s glaucophylla is a remark- 
able variety with obovate fruit and nearly persistent 
sepals. 
The difference of 6 seems to be nothing more than 
the smaller size of every part; an appearance which is 
by no means uncommon in this country, although I 
have never observed it in so remarkable a degree as 
Rau's figure indicates. The specimen figured by Re- 
douté is evidently an approach to a more robust mode 
of growth. Perhaps it is the same sort of variety of 
canina, as the dwarf Chinese Rose of the gardens is of 
R. Indica. 
Iam acquainted with y only from a specimen in 
Sir Joseph Banks's herbarium, from Forskahl, marked 
R. ewgyptiaca. It is distinguished by the unusually 
deep serratures of the leaves and its very long re- 
ceptacle. Forskahl mentions no Rose as having been 
found by him in Egypt; can this, then, be what he 
calls R. indica, found on the mountains of Arabia felix ? 
A has the lower sides of the leaves hairy, the upper 
surface shining, and the sepals and flowerstalks usually 
hispid. In other respects it is not to be distinguished 
from R. canina a. The R. collina of English Botany 
belongs to a very different plant, R. systyla. R. platy- 
phylla of Rau has smooth peduncles; and R. ake 
-phylla of the same author has downy petioles with a 
naked under-surface to the leaves. R: fastigiata of 
Batard has no character whatever to distinguish it. 
