1896.] D. Prain — Some additional Fumaiiaceas. 11 



ments and the fact that the epidermis remaing entire after the segments havo 

 fallen away. This is characteristic of the Indian specimens also, whether the outer 

 petals be lobedor entire, and it is on this account that the writer makes the identi- 

 fication noted above. Thus considered the plant is seen to be a very distinct 

 geographical form occupying the eastern portion of the Mediterranean and Central 

 Asian region. The differences implied by their fruit-characters are however so 

 decidedly only differences of degree, that in a monographic review of the genua 

 ifc would probably be preferable to unite jff. parviflorum with H. pendulum as M. 

 Boissier has proposed. For the purposes of a local Flora it is obviously better to 

 follow Sir J. D. Hooker and Dr. Thomson in separating them. 



2. Hypecoum leptocarpdm H. f. ^ T. Flor. Ind. 276 (1855) ; Flor. 

 Brit. Ind. i. J 20 (1872); Franchet, Bull. Soc. Bat. Fr. xxxiii. 39 L 

 ( 1886) ; Maxim., Flor. Tangut. 37 ; Enum. Mongol. 36. 



Add to localities of F. B. I. : — Badakshan, Giles ! Pangi, Heyde ! 

 Kamaon, Duthie! Bootan, Chumbi and Phari, Dr. King's Collectors! 

 DiSTRiB. E. Tibet {TJiorold !) S. E. Tibet {King's Collectors!) N. Tibet 

 (Przewalski !) China; Kansu (Potanin!) Szechuen (Pratt!) Yunnan 

 ( Belavay ! ) 



This very distinct species comes just within the northern border of the Indian 

 region. Its area lies to the east of that occupied by the preceding but without 

 overlapping it. Very nearly related to this and perhaps only varietally distinct 

 is JET. chinense Franchet, [Fi. David, i. 27 (1884)]. This differs somewhat from 

 H. leptocarpum in foliage and differs moreover in having yellow petals. The colour 

 noted for the petals of H. leptocarpmn are " pale purple " (Hooker) and " pink," 

 "rose," "slate-coloured," "bluish- white," "white with purple-streaks" (various 

 collectors sent by Dr. King) ; in one gathering from Chumbi, the petals have been 

 noted as " yellow." This gathering therefore, agrees with M. Franchet's plant, 

 which comes from the neighbourhood' of Pekin, as to flowers ; at the same time 

 it has the foliage of the other specimens and could not be separated, even as a 

 variety, from H. leptocarpum. The existence of this form strongly supports M, 

 Franchet's suspicion (loc. cit.) that H. chinense is merely a variety of H. leptocar- 

 pum. In Northern Tibet and Mongolia the flowers, Mr. Maximowicz says, are 

 always pale-yellow, never blue. 



DICENTRA BoRKH. 

 Key to the hidian species. 



* Bracts elongate, capsules narrow linear, coriaceous : — 



t Bracts as long as pedicels ; capsules tornlose, seeds 



opaque ... ... ... ... •.. 1- D. toridosa. 



ft Bracts shorter than pedicels ; capsules not tornlose, 



seeds shining ... ... ... 2. D. Roylei. 



** Bracts very small, capsules broad (seeds shining) : — 



t Capsule membranous, acute at both ends, early dehis- 

 cent ... ... ... ... ... 3. D. Macrocapnos. 



ft Capsule fleshy, ovate-cordate, tardily or not dehiscent 4. D. scandens. 



