322 TRANSACTIONS OF THE [Sess. lxxix. 



very different. The hirsute scape is always shorter than 

 the leaves, and the few flowers are on long pubescent 

 and hirsute pedicels. The flowers have remarkably large 

 corollas. 



P. obconica, Hance. Forms of slight divergence. 



The forms shown on Plates XXI, XXII, and XXIII are 

 taken to be the type form of P. obconica, Hance; and 

 for the reasons already given, they are the plant as first 

 described by Hance. They come from the area of original 

 find. They are like the cultivated form (as shown in 

 Plates XXIV and XXV), which we assume to be descended 

 from seedlings raised from Maries' plants. But within the 

 Hupeh area the plant shows variation. Plates XXXIV 

 and XXXV show specimens which illustrate this. 



Plate XXXIV is one of Dr. Henry's specimens (No. 1312) 

 collected in West Hupeh ; I presume somewhere in the 

 vicinity of Ichang, which was his centre. In it the glau- 

 cous hue is not seen save on the under side of the leaf ; the 

 form of leaf is more oblong, and there is some variety in 

 form in the several plants ; the lobing and denticulation 

 become more evident. The flowers, too, are smaller, and 

 the successive expansion within the umbel becomes very 

 evident. The vestiture of hairs is hardly different from 

 what I have described, save that the young scapes have the 

 brown villous hairs. 



Plate XXXV shows one of Delavay's specimens (No. 317) 

 in the Paris Herbarium gathered on rocks at the side of 

 the Blue River (Yangtze) at Ichang. In general aspect 

 it resembles Henry's specimen (No. 1312), but the whole 

 plant is densely clad with the long brown hairs — they coat 

 the scapes and bracts and even the base of the pedicels, and 

 form beard-like tufts at the base of the calyx. The upper- 

 most plant in the plate shows foliage in form more like the 

 cultivated plant (Plate XXIV) and the plant from Che- 

 pa-to (Plate XXIII), but the foliage generally is more oblong 

 and the leaf-margin more lobed. As in Henry's plant, the 

 under side of the leaf only has a greyish look. These two 

 plants are undoubtedly P. obconica, Hance. 



The physiognomy of these specimens recalls the plant 

 which Hooker figured (Bot. Mag. (1881), t. 6582) as P. 



