42 ANNALS OF SCOTTISH NATURAL HISTORY 



while in R. glauca they raise themselves more or less rapidly 

 after flowering, and remain habitually erect on the receptacle 

 until complete maturity. It is not absolutely necessary to 

 assure oneself of the condition of the sepals to distinguish 

 R. glauca from R. canina. In the former the large head of 

 stigmas, tomentoso - whitish in colour, is very different in 

 aspect from that of R. canina, which is less thick, with the 

 hairs much less abundant, and sometimes is glabrous. In 

 R. glauca the pedicels are usually shorter, more or less 

 hidden by the bracts and dilated upper stipules ; the corolla 

 is usually a rather bright red, in place of being pale rose or 

 whitish. The bush of this species is usually more squat 

 than that of R. canina. R. canina, in the varieties of the 

 group R. andevagensis, Bast., and R, Deseglisei, Bor., may at 

 times be mistaken for varieties of R. stylosa, Desv. ; but it is 

 easy to distinguish the latter by its glabrous styles, united 

 into the stylar column, by the narrow upper stipules, and by 

 the constantly smooth receptacles. 



In its varieties with the leaflets glandular below (groups 

 R. scabrata, Crepin, and R. Blondeana, Rip.) R. canina- may 

 be confounded with R. Jundzilli, Besser ( = R. trachypliylla, 

 Rau), or with R. septum, Thuill. Since R. Jundzilli is not 

 found in the British Islands, British botanists need not attend 

 to it in this connection. R. sepium is distinguished from 

 the varieties of R. canina having glandular leaflets by its 

 styles being almost always glabrous and not shaggy, and by 

 its leaflets being elliptical and more or less strongly narrowed 

 at the base, with glands possessed of a decided scent of a 

 rennet apple, and not odourless. 



Rosa obtusifolia, Desv. (including R. tomentella, Lem.) 



The true R. obtusifolia, Desv., which cannot be con- 

 founded among the forms of R. dumetorum, Thuill, does not 

 differ from R. tomentella, Lem., except in the teeth of the 

 leaves being simple, and not compound-glandular. This 

 single difference will not permit us to separate these two 

 Roses specifically. The subordinate species constituted by 

 them ought to receive the older name that of R. obtusifolia. 



The variety with simple teeth appears to be localised 



