*- 68 THE BRITISH ROSES 
confounded it with both, while —— thinks that R. omissa is 
merely glandular mollzs. This r view is clearly untenable if 
separate Li. mollis from its allies, but Mr. tee I think, is of 
ago that well-defined R. mollis is not known, or is rare, on 
the ent. 
aca. are three emer on the sheet in Smith’s herbarium, 
of which the types quoted by him in the remarks following his 
one age are Nos. 1 and 3. The first is cultivated from a 
ottish specimen brought by Jackson, and the other a wild one 
fon between Edinburgh and Ravelstone Wood. They both have 
finely and esta Madi aieo, more Dataaly so ovis eglan am 
lar except on of the midribs. imary teeth open and 
shallow, some of ahead with one or two conspicuous secondary 
teeth, but all with small fine glandular denticulations. Petioles 
very densel y, but very finely pubescent, in No. 1 a good deal 
glandular and aciculate, in No. 3 much less so. Stipules broad, 
with broad shortly acute or cuspidate auricles, which do not 
in, 
Ss : 
densely woolly. In No. 2 the peduncles are much shorter, fruit 
ger, sian erect, subconnivent, Pm pinnate, densely glan- 
dular-hispid 
nature of the fruit Theale the best character of distinction. 
erate they flower at the same time, this species ripens its 
t least a month earlier. By the end of August its bright 
nivent and eeakingly compound sepals, and deliquesce with the 
early frost. ping Aepees. tee ‘eas alice fruited 
forms arecommon. The ciliation of the petals, which is + maekaineed 
_ by Déséglise and Fries as a character of R. mollissima, is 
occurs tosa 
> Sem ween the two in other points that t often dried spec 
_ mensin flower are hardly distinguishable 
_ The following no aie tear ngs Siten an examination of all 
