276 FLORA INDICA. [Fumariacea. 



Variable in foliage and size ot flower, as are all the species of the genus ; also in 

 the form of the sepals and petals ; the inner petals have the middle lobe larger or 

 smaller than the lateral, and more or less fimbriated. 



3. H. leptocarpum (H.f. et T.) ; floribus pallide purpureis, se- 

 palis ovato-lanceolatis, petalis exterioribus late obovatis interioribus 

 trifidis lobo intermedio spathulato, siliquis gracillimis. 



Hab. In Tibetia occidentali, frequens arvis alt. 9-13,000 ped. ! in 

 Himalayse Tibeticse arenosis: Sikkim, alt. 13-14,000 ped. ! — (PI. Jul.) 

 («. c.) 



Serla gracilis, annua, diffusa, statura valde variabilis, 3-poll. v. bipedalis. Wolia 

 radiealia plurima, patnla, lineari-oblonga, pinnatisecta, 2-4 poll, longa, pinnis \ unc. 

 longis, remotis, late ovatis, pinnatisectis, lobis dentatis acuminatis. Caules plurimi, 

 basi decnmbentes, ssepius elongati, simpUces v. pluries dichotome ramosi, bracteis 

 sectis. Pedicelli filiformes, bracteolis setaceis involucrati. Flores pallide purpurei 

 V. lilacini, \-^ poll. lati. Sepala petalis \ breviora. Petala exteriora apice sub- 

 coriacea, viridia, interiora minora, vix ad medium flssa, lobis lateralibus obtusis, in- 

 termedio late oblongo, sessUi, cucullato, marginibus incurvis integerrimis. Stigmata 

 2, recnrva. Siliquts poEicares, vix J hn. latse, gracillimae, compressaj, 8-10-sperm£e, 

 articulis indehiscentibus, facile solutis. Semina oblique oblonga, loculos implentia ; 

 testa subcoriacea, brunnea. Albumen carnosum. 



Very closely allied to the Siberian Chiaeospermum erectum, Bernh., in habit, co- 

 lour of flowers, form of sepals, and slender siliqua, but the latter are very different in 

 structure, showing no trace of valvular dehiscence, but breaking across at the joints 

 even before the ripening of the seeds, which adhere firmly to the cavity of the peri- 

 cai*p. The C. lactiflorum, Karel. et Kiril., of Soongaria, seems the same as C. ereC' 

 tum, the inner petals in our (authentically named) specimens of it differing in noway 

 from those of other species of Rypecoum ; and the character of their middle lobe 

 being antheriferous, is either inconstant or founded in error, and possibly arises 

 from the pollen being sometimes adherent to its cucullate face. 



Our Sikkim specimens are very much smaller than most of our western Tibetan 

 ones, but agree in all essential characters. 



