CONCLUDING REMARKS 137 
Conctupinc Remarks. 
t will be felt, I fear, that not only have I done nothing 
Svigiais but that I have left the Sah in confusion, owing to my 
grouplag EMO ss he S ina the plan of 
that I soon became cena at nde with the leading species 
of my own neighbourhood, but it is not so with Rosa. This ma 
n to 
belong to different species or even groups, while stat which 
characters. It is the latter which I think are at fault. Too 
much stress may be laid upon variations in ee clothing, 
size of leaflets, form of prickles, &c., while t ural affinities 
deduced from habit and geographical distribution shay be lost sight 
of. These latter are, of course, more cult of observation, and 
ett moreover, nothing more convenient could be made of 
artificial arrangement suitable for a herbarium, but it eevulbs 
in ineongruit es. 
not think it of great importance whether we have few or 
any ries in our list, but there is no doubt that we have a 
considerably larger n umber of forms than have so far receiv 
mes in Britain; these are quite aig and worthy of 
distinction, probably as varieties only, to which rank several of 
een ies should It is ee duty to bring our- 
into line with continental botanists, and to determine what 
hee oa ee the limits of the 
forms for which we have nam 
