141 



Inflorescence^ flower, etc. — The inflorescence of Rosa 

 berberifolia is, so far as I hâve seen, in herbaria or 

 gardens, uniformly 1-flovvered and wilhout trace of brac- 

 teoles, a circumstance ihat leads to the inference lliat 

 a cymose inflorescence such as other Roses hâve, though 

 far from impossible is not likely to occur in ihis species. 

 The color of ihe spot at the base of the petals varies in 

 intensity and even in tint from purplish brown to rich 

 crimson. The stamens and anihers are also purplish and 

 there is a band of a similar color, herelofore unnoticed, 

 around the top of each style just benealh the large flat 

 reniform stigma, so that on iooking down into the flower, 

 three rings or bands of color are visible, one constituted 

 by the petals, one by the stamens and the third by the 

 styles. Doubtless this arrangement of the color in three 

 separate, but concentric, rings lias référence to insect 

 visitaiions and botanical travellers in Persia and Affgha- 

 nistan should bave their attention directed to this matter. 



The carpels are raised upon a short stipes covered 

 with coarse white setae, the obovoid ovary is itself 

 entirely glabrous with a single pendulous ovule in the 

 interior. The styles project beyond the moulh of the 

 réceptacle and are densely covered with long white, 

 cottony hairs which bind the styles into one mass though 

 they can be readily separated by the needle. The stigmas 

 bave already been alluded to. 



Acting on the suggestion of our great Rhodologist 

 M. Crépinlli, 1 bave examined the « insertion of the car- 

 pels » and find it basilar. The low^r part of the recepta- 



(1) See Crépin in Bull. Soc. Bol. Belg., 1889, part. II, pp. 87 & 88. 



