120 ANNALS OF SCOTTISH NATURAL HISTORY 



" Your rich series of R. tomentosa is extremely interest- 

 ing. If I had time I could make abundant remarks upon 

 it ; but that will be for a later date. I shall limit myself to 

 a few short observations. 



" No. 145 is remarkable for the numerous glands on the 

 upper surface of the leaves. 



" Nos. 143 and 160 are similar to it in this respect. 



" No. 147 is a variety with leaflets simply serrate belong- 

 ing to the group R. cinerascens, Dumt. 



" No. 157 belongs to the same group. 



"No. 149 has the teeth almost all simple, a few with 

 one or two glands. 



" No. 158 has the teeth irregular, simple, double, or with 

 two ' denticules.' 



"No. 140. In this form the flowering branchlets and 

 certain parts of the branches are ' heteracanth.' This would 

 seem to indicate R. involuta, Sm., but I think it is only a 

 simple variety of R. tomentosa. 



" No. 153 is probably what you think it, a var. of R. 

 tomentosa ; but, singular enough, the serration is irregular. 



" Nos. 178 and 179 have a resemblance to certain forms 

 of R. coriifolia, with leaflets glandular beloW r ." 



The parcel of 1895 consisted of 43 Nos. from localities 

 in Mid Perth, Fife, Stirling, Roxburgh, Selkirk, and Cheviot- 

 land. Professor Crepin reports on it as follows : " All these 

 Nos. have given 82 inflorescences one -flowered, and 79 

 many-flowered, which almost gives the proportion I to I, as 

 in my statistical work on this subject. 



" All these numbers constitute a very fine series of 

 variations. I have not sought to classify these variations. 

 I shall wait, before doing so, till I find a satisfactory basis of 

 classification, which I hope to be able to establish at a 

 later date. 



" In comparing your specimens of R. mollis with those 

 of R. tomentosa, one can easily recognise the characters 

 proper to each species characters based on the form of 

 the prickles, the shape of the stipules, the mode of erection 

 of the sepals and their shape, the date of ripening, etc. 



" I have already put forth the idea that there may exist a 

 hybrid of R. mollis and tomentosa. Perhaps this hybrid has 



