160 Mr. Woops on the British Species of Rosa. 
tion, and urged my endeavours to find fixed and distinct charac- 
ters, to distinguish plants marked by such differences in the ge- 
neral appearance. ‘Though I feel that I have little reason to con- 
gratulate myself on the success of these attempts, and have indeed 
been successively obliged to relinquish many of the characters on 
which that general difference of appearance depends, yet I ven- 
ture to offer their imperfect results to the notice of the Linnæan 
Society. In an obscure or intricate subject, the faithful record of 
observations is always valuable. 
With views no more exalted, it may be considered as incon- 
sistent to attempt a Synopsis of the British Roses; but in fact I 
did not perceive any other mode in which the remarks I had col- 
lected could be so well arranged ; and the attempt once made, I 
exerted myself to give some consistency and value to the essay, by 
putting in systematic order the materials within my reach. That 
it is still imperfect I am aware; but I flatter myself it will not 
be found useless by the future investigator of this most interesting 
genus. : | 
. It appears to me that the principle to be attended to in the 
subdivision of genera, is to keep together those species which are 
most nearly allied in nature. In the formation of the genera 
themselves, it may be necessary to attend exclusively to the or- 
.gans of fructification, as the most important parts of the plant ; 
but in their sections we must find a character in any part which: 
will keep similar plants together. With all this latitude of choice, 
the accomplishment of the object will be found often of very 
difficult attainment; and after all our labours, the best arrange- 
ment which can be made may still present some important aber- 
rations. à | 
On examining by this general rule the usual division of the ge- 
nus Rosa into those ** fructibus ovatis" and “ fructibus subglobo- 
sis,” 
