I > . »■ H.I.N- | 



K0XB1 10.111 \ 





I. XXII. ROX URGHIAC1 



Twining ahrabe with tuberoua ro its! Leaves ret 

 ondan reins connecting Beveral primary ribs Flowi 

 foetid. Perianth ■ •! i large petaloid 

 divisions. Stamens I, hypogynou 

 anthers adnate, opening inwards, 

 pointed, with connectives projecting 

 far 1. ivond tin- cells, which separate 

 from tin- latter a- tar as their I 

 '•i:in Bnperior, I -celled, with ■_' po- 

 lyspermous placentae arising from the 

 \ . iry I:,-'- of tin' pericarp : Btyle none : 

 Btigma somewhat pencil-shaped ; 

 ovules mi, aniitropal. Pericarp 1- 

 oelled, 2-valved, with 2 clusters of 

 - at tii>- base. & eds attached to 

 long cords covered with loose hairs 

 just below tlic Beads ; embryo taper, 

 in tho axis hi Beshy albumen, with the 

 pluxuul lying within a -lit. 



. 



i i i 

 snus is ma sufficiently ■ 



The affinity of this singular _ 

 refer it t" any known Natural ( h*der : by Endlicher it is i 

 ami there ran in fad 1»- ii<- doubt about 

 tanist includes in the Smilaceous Order. I, however, formerly i 

 allied to Arads than t,, anything else, and Mr. Grifl : : 

 opinion as to consider it certainly one of the class of which \ 

 which he has apparently been influeno d by the diacovi rj 

 embryo. But this character has lost it- valu< 

 Jussieu that a slit embryo is found v. • 

 inflorescence is indispensable to Arads ; bo that this ri< 

 ran hardly be maintained. It would rather ■ 

 recruits to which we have still to look. In the meanwhi 

 ■ tendency towards Avals on the pan of Did 

 100 bthoms long. Mr. Griffith regards the pistil 

 carpel only, as • is indicated by the obliquity 



Pig. CL.-Roxbutghia glon, 



