46 ANNALS OF SCOTTISH NATURAL HISTORY 



British Islands. It becomes more rare southwards, where it 

 is ultimately replaced by R. point/era. It is a species of the 

 mountains, at least in the centre and the south of Europe, 

 and not of the plain, like R. tomentosa. If, in nature, R. mollis 

 is always easy to distinguish from R. tomentosa, this is not 

 the case with herbarium specimens, and numerous are the 

 confusions that exist in collections. As we have stated 

 above, these two types are essentially distinct, not only in 

 their morphological characters, but also by differences in 

 their biology. Once again I draw the attention of phyto- 

 graphers to the distinctive characters of these two species. 

 R. pomifera, Herm. (which is closely related to R. mollis, to 

 which one might probably unite it specifically), does not yet 

 seem to have been met with in a wild state in the British 

 Islands, in which it perhaps does not exist, at least in its 

 typical condition. 



I come now to hybrids known under the names of 

 R. hibernica.) Sm., R. involuta, Sm., R. Sabini, Woods, 

 etc., which for a long time have been admitted as species 

 peculiar to the British Islands. Regarded as autonomous 

 types, these Roses made the rhodological wealth of this 

 country. But at the present day these same Roses have been 

 met with on the Continent, and under conditions such as to 

 give rise to the belief that they are not true species, but 

 rather hybrids. My own study of them has led me to 

 accept this conclusion. While one of the parents of these 

 alleged species is beyond doubt R. pimpinellifolia, and for 

 some R. tomentosa and R. canina is certainly the second, 

 one may conclude with good cause that in the production 

 of others R. mollis has intervened, or perhaps even R. rubi- 

 ginosa or R, micrantha. 



At present R. involuta, Sm., is made up, according to the 

 " London Catalogue," of twelve varieties. These are very 

 hard to characterise, or even to distinguish from one another, 

 a thing very naturally explained if they are regarded as 

 hybrids. Their study should be undertaken not on materials 

 in herbaria, but on living plants in their natural habitats. 

 They must be compared carefully with the other Roses that 

 grow near them, and search should be made for the types 



