GROUP OF ROSA SHERARDI 67 
ich he has been naming R. Andrzeiovii are a dis- 
tinct from other members of the Sherardi group to merit a new 
ame ave not seen enough of them to gore an opinion on 
this point, nor do I_ know more of Rouy’s var. losa than 
appears in his key. Beyond all this, Steven's Andraciovis is an 
Eastern European species, unlikely to occur in Britain 
GROUP OF ROSA SHERARDI. 
nee the recognition of R. omissa Déségl., or at least forms of 
it, as a British ga we the desirability of segregating species and 
varieties, which have affinity therewith, from 
- pomifera R. 
ec 
fi. Sherardi Davies, one of its members, is a much older name, it 
is proposed to use it at the head of the group. . farinosa 
Bechst., one of our supposed British species, actually pier 
ie. Sherardi by three years, but that is a somewhat o 
aberrant ces of which I have seen no aut eit “material, 
nastag: ver it is usually regarded on the CGantinsiit as belongin 
o the ¢ pd sis group, though its affinities seem much stronger 
with the present o 
The leading feat of the group of BR. Sherardi are a low, 
compact, rather stiff habit, just intermediate between that of the 
- pomifera and RK. tomentosa groups. The internodes are seldom 
el straight but they are not as a rule obviously flexuose as in 
mentosa e prickles are slender or moderately stout, straight 
or cnn falcate, very rarely hooked, and even then less stout 
than they freque ently are in the next group. Its leaflets are 
ated densely and softly tomentose, but subglabrous varieties 
ey vary much in shape, but in general are rather close 
ay a somewha se, and are bluish green in colour 
stipules have more acuminate auricles than in the pomifera group, 
and the upper ones are not SF ie sec Nor the 
ur 
long. The sepals are somewhat pin rade ase, and rounded on 
the back, spreading-ascending or even erect and connivent after 
flowering, persielvn till the fruit is ripe, then deciduous. Occa- 
sionally some of the sepals may fall before the ripening of the 
fruit, but never leo it is fully fon The styles are villous, rarely 
only hispid. 
ISioueh! the above definitions are drawn up to include the 
‘eenrasiaaly classified 
as tomento, eras on ere Continent. Apart from the eres 
itch 1 is not always determinable in a atta specimen, 
best and most easily observable feature is oS of the : 
