A LIST OF BRITISH ROSES 41 



K. Sherardi Davies, Welsh Bot. p. 49. Though typically the 

 fruit of this species should be globose or almost so, several of 

 Smith's own specimens have it quite ovoid. Thick, densely, softly 

 tomentose leaflets, eglandular beneath except sometimes on primary 

 nerves, often subsimply toothed, and rather stout, falcate prickles, 

 are its leading features. Continental authors place it, under the 

 name of B. subglobosa, in the Tomentosa group, but if there is 

 anything in the persistence of the sepals, it belongs to the present 

 one, though just similar forms occur with the sepals deciduous, 

 which belong to the next. The species is probably much com- 

 moner than the following records show. The S. Devon example 

 was referred to B. tomentosa var. leiocarim Boullu by Sudre. 

 V.-c. 3, 8, 15, 16, 17?, 31, 34, 40, 43, 58?. 



E. OMissA var. submollis Ley, in Journ. Bot. 1907, p. 205. 

 This is very near its neighbours on either side. I do not place very 

 great reliance on the absence of subfoliar glands — indeed, the 

 hairs so often conceal them that it is difficult to say whether they 

 exist or not ; thus the chief distinction from B. resinosoides is 

 lost. The prickles in this and the neighbouring species vary by 

 insensible gradations from straight to hooked, so that it is im- 

 possible to draw the line between var. siibmollis and B. Sherardi 

 or var. pseudomollis . If it can be segregated at all, specimens 

 from the following vice-counties may be referred to it. V.-c. 23 ?, 

 35, 36, 40, 43, 70, 88, 98, Antrim, Armagh?. 



E. TOMENTOSA var. PSEUDOMOLLis Baker fil. in Journ. Bot. 

 1892, p. 341. I can see no difference in specimens referred by 

 Ley to this variety and to var. submollis. By his key, var. 

 liseiulomollis differs in little but its falcate prickles, but specimens 

 with quite falcate and stoutish prickles have been referred by him 

 to var. submollis. The type specimen shows that var. liseudomollis 

 may have some subfoliar glands on some of its leaflets, those of 

 var. submollis being always eglandular, but this is a weak dis- 

 tinction. The Boar's Hill plant (Bot. Exch. Club. Eept. 1893) is 

 just as near var. uncinata, to which Ley referred it. V.-c. 20, 

 22, 36, 42, 43, 80. 



E. TOMENTOSA var. UNCINATA Lees, in Eept. Bot. Eecord Club, 

 1884-1886, p. 123. This variety stands at the end of a practi- 

 cally unbroken series from B. resinosoides, or at least from B. 

 Sherardi, and though no doubt fairly distinct forms may be found 

 among them, they seldom agree with any of the described ones. 

 It would seem best to reduce var. submollis, var. liseudomollis, and 

 var. uncinata all to varieties of B. Sherardi Dav. Var. uncinata is 

 too close to \iix. pseudomoUis, differing only in some of its prickles 

 being quite hooked and rather stout, though not large, which give 

 it a Coriifolia look. I have, however, specimens from Llys-y- 

 Gwynt, as well as of var. pseudomollis from Cowlcigh Park, 

 their respective original stations, and it is difficult to see how 

 they ditt'er, except in unimportant details. My specimens were 

 labelled B. tomentosa var. serotina Coste by Sudre. V.-c. 36?, 

 40, 42, 49. 



