172 Mr. Woods on the British Species of Rosd. 



yellow-flowered Roses are perhaps constant in their colour; but 

 this is by no means the case with the other species. The globu- 

 lar fruit in some divisions of the genus appears to be important ; 

 in others it is extremely uncertain. If the bristly fruitstalks are 

 ever of any value, it can only be when they are used very cau- 

 tiously to separate one or two allied species in particular subdi- 

 visions. 



The characters which appear to me most constant in this genus 

 are the presence or absence of setse on the stems ; the prickles 

 straight or hooked, equal or unequal ; the tendency towards the 

 formation of the upper stipulae without leaves, or at least with 

 leaves of fewer folioles, and expanding into bractcfe. Next to 

 these are the simple or compound form of the leafits of the ca- 

 lyx, and the simple or compound serratures of the leaves. In the 

 latter subdivisions I have made use of the shape and flatness 

 or hoUowness of the leaflets ; and sometimes, though unwillingly, 

 I have been obliged to depend on the pubescence, not finding 

 any other describable character to discriminate plants whose 

 difference of habit seemed to announce the necessity of sepa- 

 ration. 



This arrangement is not without its disadvantages, principally 

 on account of the deciduous nature of the setae in two, or perhaps 

 in three, families of the genus. Of these, however, JR. cinnamomea 

 is the only British plant; and a moderate attention to the descrip- 

 tion M'ill easily teach the difference between this plant and Rosa 

 riUosa, the only species with which a specimen devoid of setce is 

 in danger of being confounded. 



ROSA. 



