ROSA BAKERI 105 
mall green viscous glands, serrations open and many times 
decked with gland-tipped teeth; terminal ovate or obovate, either 
rounded at base or narrowed from below middle. qovole both 
rickly and rather pubescent, copiously setose. Stipules and 
bracts both hairy and somewhat glandular on back, ae at 
Seo Peduncles sometimes naked, som rei with a few 
erenten.” geese naked, oval or Leg aleagge oaks. Petals 
deeper seg and smaller than in type, and wavy towards 
borders. One or two sepals BRE the ia copiously pinnate 
and leaf pointed, all slightly glandular on back, hairy towards 
edges pe eek setoso-ciliate, erect-patent after pani fall. 
it or elliptic, ripe early September, most of the sepals 
adhering “tal after it changes colour. Styles vi at Near Borrert, 
but leaves very different in shape and texture, and very slightly 
gla pleat ou gaa Peduncles hardly at all si par os sepals 
subpersi 
In Fie. p. 237, Mr. Baker says 
approaches “f. pulverulenta M. Bieb.” (i.¢., R. pa Theil, 
his opinion being based no doubt on the glandular under surface of 
the leaflets, and the oblong oF atone fruit. Déséglise contrasts 
it with tomentella, though th ms to be little danger of con- 
founding it therewith. He sie eget that it is very near R. fetida 
Bast. Sn of the Tomentose), but eat no smell on rubbing, the leaves 
are less glandular, peduncles very short, glabrous, calyx-tube glab- 
rous, sepals much less glandular, flowers rose, fruit bright red. Ido 
not think Déséglise would have seen the similarity to fetida had 
he not credited R. Bakert with obscur sgt hispid or glabrous styles; 
) ) 
type, as they almost all present the eet, woolly head of styles so 
phaeictaristls of the corifolia grou r. Baker’s No. 30, how- 
ever, which Déséglise quotes as a Pay bey one or two others, have 
the styles only moderately hispid. This being an endemic species, 
and since the bulk of the authentic specimens agree with Mr. 
Baker’s description in having villous styles, it is, I think, justifiable 
to regard the eople | from which Déséglise drew up his deserip- 
tion as Ne s respect. 
udging from Mr Baker’s specimens collectively, 8. Bakeri 
should have slender curved prickles from longish bases, leaflets of 
ssa size, elliptical or oval, somewhat narrowed below, pube- 
nt both sides, but very thinly so above and not densely atieeth, 
pa ae all over the lower — No. 30, but usually only on 
midrib and secondary nerves, only; the primary 
teeth very coarse, much biserrate; icles very densely but shortly 
felted, and with glandular sete. Peduncles 2 or 3 together, very 
ants Dring ate oe ig oat slightly lesdidar hing in No. 30. 
oid, usually towards ellipsoid, rarely s oe Sepals 
a ami as diestchare fully bipinnate, strongly & ndular- 
ciliate, nh not much glandular on back. Style hay more 
or less villous head. 
