8 A LIST OP BRITISH ROSES 



or more others, and present an almost interminable array of 

 species, varieties, and forms, it will readily be seen how very much 

 larger a series must i-esult from their hybridization, not only on 

 account of the large possible number of combinations, but also 

 from the inevitable variation resulting from their interbreeding. 

 The following account, therefore, is a very imperfect one. 



GROUP PIMPINELLIFOLI.E x VILLOS^. 



There is no doubt that B. involuta Sm. has been treated as a 

 very large aggregate, and has been made to include hybrids with 

 the subsections Eu-canina {B. hibernica Tempi.) and BuhiginoscB, 

 besides those with the Villosce, of which latter hybrids it has of 

 recent years been held to be the type (agg.). 



E. sPiNOSissiMA (agg.) X MOLLIS. I havo seen specimens 

 which should probably be referred to this parentage from v.-c. 79, 

 89, 92, 94, and 108. 



E. SPINOSISSIMA (agg.) X TOMENTOSA (agg.). It is probable 

 that most of the succeeding forms belong to this hybrid or to 

 B. sinnosissima (sbgg.) x oniissa (agg.), of one of which they should 

 be regarded as varieties only, certainly not as species ; but, as 

 already stated, it contains an admixture of B. sjnnosissima (agg.) x 

 mollis, at least under the first three names. 



E. INVOLUTA Sm. Fl. Brit. p. 1398 (var. Smithii Baker, Monogr. 

 p. 207). Specimens in Mr. Druce's herbarium from Arran and 

 Derby, and Mr. Marshall's from E. Kent, may, I think, be referred 

 here, but it is difficult to separate it from one or two of the other 

 segregates. V.-c. 15, 57, 100. 



E. Sabini Woods in Trans. Linn. Soc. xii. p. 188. This runs 

 so completely into B. Doniana that I think it is best to regard 

 them as synonymous, though extreme forms might be segregated. 

 V.-c. for seg. 62, 83 ; for agg. 38, 58, 64, 65, 82, 90, 96, 106, 108, 

 E. Mayo, Antrim. 



[B. Doniana Woods in Trans. Linn. Soc. xii. p. 185. Included 

 in the above ; also, as a segregate, from v.-c. 13, 83.] 



E. gracilis Woods in Trans. Linn. Soc. xii. p. 186. I have 

 seen no further specimens referable to this or to the remaining 

 segregates of the group, except B. Wilsoni,_QXidi can, therefore, 

 only quote the vice-counties given in B. E. pp. 33-40. V.-c. 17 ? 

 (distributed as var. Bohertsoni), 21, 62, 70. 



E. INVOLUTA var. Eobertsoni Baker (Monogr. p. 206). V.-c. 

 62, 66, or 67, Derry. 



E. INVOLUTA var. gracilescens Baker [I.e.). V.-c. Antrim. 



E. INVOLUTA var. occidentalis Baker (/. c. p. 207). Ireland. 



E. Wilsoni Borr. in Hook. Brit. El. ed. 3, p. 231. As stated in 

 B. E. p. 47, I include in this B. hibernica var. cordifoUa Baker. 

 Some of S. A. Stewart's specimens from Knockagh, labelled B. in- 

 voluta, are very near this. V.-c. 44, 68, Derry. 



