SUBSECTION VILLOSE 53 
gation of very heterogeneous material under one name, or the 
creation of new species or varieties, or at least the adoption of 
Continental names hitherto unknown, or little known, to British 
bot sree 
tions. It is more than likely that several of our forms, named or 
unnamed, are covered by Continental names, which would sooner 
or later be found to be preferable, and would displace any hastily 
given. Also any new combination would, in all probability, have 
to be changed again when the subsection is thoroughly revised, as 
it must be, with the whole genus, in the near 
publi n by 
M. Sudre and Prof. Déséglise to specimens of Mr. Ley till I have 
em as their authors named them, but I have ranged them 
: ae ee, groups to which they must, in my opinion, eventually 
e refe 
I Lae for the purposes of this paper and not as a final 
» arrangement, to subdivide the subsection into three groups, vz. 
ER. pomifera (including R. mollis), R. Sherardi “ Fi. tomentosa. 
Most authors make two subsections, one containing I. pomifera 
and R. sedi the other R. omissa (which excludes R. Sherardt) 
RR. tomentosa. The main subsectional difference relied upon 
ous as ei canina and Ff. coritfolia. It seems hardly con- 
eieals the groups may be defined as follows :— 
Group of R. pomifera. Of low growth. Stems rigid, with 
straight internodes. Prickles* slender, straight. Leaflets usually 
dense sely pubescent, fully biserrate. U tipules with faleately 
incurved auricles (according to Continental authors). Peduncles 
atte short. Flowers deep r Fruit becoming soft an 
arly, crowned by the subaidios persistent sepals, which 
roup 
habit intermediate between that of the pomifera and tomentosa 
groups. Prickles straight or somewhat falcate, rarely uncinate, 
slender, seldom stout. Leaves, as a rule, densely pubescent, but 
- oh n the middle part of the old stems (which ee 
aowering branches) should be examined. Those on the flowering. _them- 
selves, or on p 
