A LIST OF BRITISH ROSES 39 



R. POMiFERA Herrm. Dissert. Eos. p. 16. I have no speci- 

 men of the typical plant in my herbarium, but I do not think 

 that B. recondita should be separated, except perhaps varietally. 

 V.-c. 80, 106. 



[B. recondita Pug. in Desegi. Rev. Tom. p. 46. This seems to 

 be nothing whatever but a form of B. i)omifera with more sub- 

 foliar glands than usual. By description its leaflets are smaller, 

 but Deseglise does not lay stress on this to distinguish the two. 

 I doubt if it is even varietally distinct. My own specimen of the 

 Deal plant (see B. R. p. 57) has quite as many subfoliar glands as 

 that from Gloster. V.-c. 15, 34, 77 ?, 91, 98.] 



R. Grenierii Desegi. Ess. Monogr. p. 128. This species 

 may be said to cover examples that are intermediate between 

 B. pomifera and B. mollis, and though I follow the continental 

 practice of classifying it near B. poviifera, authentic specimens 

 look to me nearer B. mollis. Probably this is a case where the 

 habit would decide. Ley's ''B. poviifera" from Brecon is placed 

 here by Sudre and Dingier, and I suspect much of the Scottish 

 B. mollis might be refex'red to it. V.-c. 42, 62. 



SUBGEOUP MOLLES. 



R. MOLLIS Sm. E. B. tab. 2459. As with B. pomifera, this 

 species can be segregated into forms without subfoliar glands, 

 which constitute the type, and forms with the glands numerous 

 and conspicuous, which seem to me much commoner than I 

 supposed when I wrote my British Boses. I have not yet got 

 any continental name for the latter form which Ley used to call 

 " glandular mollis." V.-c. for type, 62, 64, 70, 75, 79, 80, 83, 95 ; 

 and for "glandular mollis," 65, 69, 70, 78, 90, Perth. 



R. MOLLIS var. ccerulea Woods, in Trans. Linn. Soc. p. 192 

 (sub B. villosa). This is not separable either by description or by 

 author's specimens by any hard and fast line from the type, both 

 fruit and peduncles having permissibly some glandular seta3. 

 Moreover, I have seen bushes, and Mr. Barclay says they are 

 frequent, on which both smootli and glandular fruit and peduncles 

 occur. My records include specimens with a few glands on 

 peduncles or on the upper part of fruit, only one or two having 

 both quite smooth. Ley considered the variety to run very close 

 to B. oviissa, and Sudre has placed one or two examples of it to 

 that species. The W. Sutherland examples {Marshall) are sug- 

 gestive of B. villosa var. coeridea x B. suberecta. V.-c. 36, 42, 

 Yorks, 69, 74, 79, 88, 92, 108, Inverness, Antrim. 



R. rsEUDO-RUBiGiNOSA Lej. Fl. Spa, p. 229 {B. arduenncnsis 

 Crep.). This is hardly separable from "glandular mollis," unless 

 it bo by its long and straight prickles, and even more glandular 

 leaflets, very often glandular on both surfaces. The leailets also 

 are smaller, narrower, and darker in colour, and by description 

 should be subglabrous, but Crepin says he places little reliance on 

 the amount of hairiness, w'hich makes its segregation from "glan- 

 dular mollis " even more diflicult. V.-c. 62, 69, 79, 89. 



