Mr. Woods on the British Species of Rosa. i.69 



length; and bearing in the following summer bunches of six\)r 

 eight flowers ; or in Rosa stirculosa, which affords an excellent 

 example of these modes of growth, perhaps even of twenty-four 

 flowers. In R. arvensis, and still more in some foreign species of 

 that tribe, these shoots frequenll}' bear cymes in the same year in 

 which they are produced ; consisting in R. arvensis of fifteen or 

 sixteen flowers; in R. indica of twenty or thirty ; in R. moschata, 

 as I am informed by my friend Mr. Borrer, who has taken the 

 trouble to count them, sometimes as many as two hundred and 

 sixty-five. As branches are yearly produced from these surculi, 

 their strength diminishes, and the original character of the plant 

 returns till new root-shoots make their appearance. These are 

 produced when the plant is partially destroyed ; nor do I know 

 that they ever occur except in consequence of some injury to the 

 original growth. They do not indeed always vary to the extent I 

 have described ; but they constantly differ in this manner from 

 the other parts of the plant, though not in equal degree. 



In the Latin descriptions no ambiguity can possibly occur from 

 the use of the term "foliolum," as applied to the parts of the 

 calyx and those of the leaf. In the English observations I have 

 endeavoured to avoid confusion, by calling the first leajit and the 

 latter leaflet, a distinction I did not adopt till I felt the want of 

 it. The shape of the leaflet is taken principally from the ter- 

 minal one, which I consider as the most perfect; all those of the 

 earlier leaves are uncertain in their shape, always rounder than 

 the others, sometimes retuse : these are to be rejected, and the 

 shape of the leaflet deduced from those expanded later in the 

 season. 



The stipulae of all British Roses are linear-decurrent on the 

 petiole of the leaf, and generally edged M'ith glands ; in some 



VOL. XII. z species 



