FURTHER NOTES ON SCOTTISH ROSES 173 



" What is it in reality ? After having examined it 

 thoroughly, my opinion is that it is very near to ft. corii- 

 folia, Fr. If its sepals erected themselves on the fruits, I 

 think I should not hesitate to identify it with R. coriifolia. 

 Its general facies, its leaves, its wood are similar. Moreover, 

 and this is a very important point, it appears to ripen quite 

 as early as a R, coriifolia or R. glauca. If I were a 

 Deseglise, I should not hesitate to make a new species of it ; 

 but, according to the principles which guide me, it is not 

 possible for me to see in this form certainly a remarkable 

 one a veritable species. But, you will say, what do you 

 make of it ? Must one join it to R. coriifolia as a variety ? 

 That is perhaps the wisest thing to do until further light be 

 cast upon it. 



" The reflexion of the sepals certainly makes it difficult 

 for us to identify it with R. coriifolia, to which we must add 

 also the less important fact of the different colour of the 

 petals. 



" If you describe it as a new secondary species, do so 

 with reserves, and make reserves also if you describe it as a 

 variety of R. coriifolia. What appears to me certain is that 

 it is not a variety of R. canina, L., or of R. tomentella, Lem. 

 On the other hand, it is not any of the forms which 

 Mr. Baker has described under the name of R. arvatica. 



" As I have already said, I do not find on the Continent 

 anything similar to it. 



" It varies a little, whilst preserving its chief characters. 

 Sometimes the sub-foliar glands are abundant, with the 

 sepals profusely glandular on the edges, or scarcely at all 

 glandular. Sometimes the sub-foliar glands are wanting or 

 nearly so, or with only a few glands on the midrib. In two 

 cases, the sepals are not only glandular on the edges, but 

 also on the back, whilst one specimen has the pedicels also 

 thinly glandular. 



" In R. coriifolia, Fr., the fruits are usually rounded ; but 

 in R. sub-coriifolia they are long and pyriform. 



" In all your specimens the teeth are composite- 

 glandular, both edges of the teeth having glandular toothlets. 



" In fine, I repeat, this is an extremely interesting form, 

 and one which deserves to be made the subject of a special 



