1914-15.] BOTANICAL SOCIETY OF EDINBURGH 309 



Hook, f., in both habit and form of leaves, calyx, and 

 corolla, but is very much smaller in all its parts, with 

 filiform petioles and scapes and more rounded leaves."' 

 And in the following year in Flor. Brit. Ind.. iii (l Vs i!.. 

 p. 485, under P. filipes, Watt, he says: 'Allied to the 

 Chinese P. obconica, Hance (P. poculiformis, Hook. £., 

 Bot. Mag - ., t. 6582), but much smaller and more delicate. 

 and the corolla tube is longer" : and then under P. Lister i. 

 King, we find : " I retain this species as distinct from 

 P. filipes. Watt, with great doubt, having very imperfect 

 specimens of this last : it differs chiefly in the more 

 orbicular lobulate leaves, broader calyx, and much shorter, 

 more funnel-shaped corolla-tube." 



I may say here that the leaf-shape, calyx, and corolla 

 characters used by Hooker as diagnostic are those that 

 are useful in the whole type. 



The near relationship of P. obconica, Hance. and 

 P. Listeri, King, was noted independently by Sir George 

 Watt. In one of the suggestive field notes attached to 

 his herbarium specimens, Sir George Watt says of 

 P. Listeri, King (No. 7105 from Manipur), to which he 

 at the time gave the name P. Robertiana, Watt: "This 

 is the plant collected by me at Tongloo in Sikkim. Bent to 

 Kew. Dr. King informs me it was gathered by Lister 

 there also. It is nearest of all to P. filipes, mihi. gathered 

 by Griffith in Bhotau. but I think it advisable to keep it 

 distinct. It is much more robust, and whereas the leaves 

 oi filipes are oblong, slightly cordate, those of Robertiana 

 are rotund-lobed or angled and deeply cordate. Filipes is 

 also very much more hairy, but being in an imperfect 

 condition and the figures made by Griffith exceedingly 

 rude, I think it advisable to keep my present plant distinct 

 so as to avoid adding characters to filipes gathered from 

 my present material that future research might prove 

 only confused both species. It is also nearly related to 

 P. poculiformis. Hook. f. (P. obconica, Hance). in Bot, 

 Mag., n. 6582, but differs in the shorter scape which in 

 Robertiana is shorter than the leaves, in the larger coarser 

 calyx, and deeper more rounded sinus in the petals. The 

 leaf of P. poculiformis is also more ovate oblong, much 

 larger and broader, and the calvx teeth much shorter. I 



