o/* Aracliis fl»rf Voandzeia. T57 



except that the calyx is rather less deeply divided ; they are also usually sterile ; 

 at least, after examining- several flowers of seventeen or eighteen species, I have 

 only once observed a tendency on the part of the style to separate itself from 

 the ovarium, whilst I have continually seen this organ wither and fall off with 

 the calyx. The fertile flowers of Sti/lnsanthes also resemble Aracliis in the 

 most important peculiarity that they are entirely without floral envelopes. The 

 ovarium is, however, nearly sessile between the two bracteolœ, which always 

 exist in both genera, and it is terminated by a short hooked style with a 

 somewhat thickened stigma. During maturation it swells and separates into 

 two articulations, without any elongation of the torus. 



It is not impossible that the hermaphrodite flowers of Sii/losanfhes may also 

 be occasionally fertile, but the hook which constantly terminates the legume 

 seems to show that it proceeds from the hooked style of the naked flowers, 

 and not from the straight end of the ovary of the perfect ones. 



The chief point in which Gracilis difters from Hedi/sarece is in the legume, 

 which does not separate into distinct articidations, a character regarded as ab- 

 solute in aW Hediisarecv \\\{\\ more seeds than one ; and this is the reason that, 

 notwithstanding the remarkable resemblance of the flowers to those oï Sti/lusan- 

 thes, no one has as yet, to my knowledge, proposed bringing them together; 

 but even in this point it will, perhaps, be found that Arachls is nearer to He- 

 dt/sared' than to any other tribe. As in the Hedysareœ, it has no valvular 

 dehiscence, and its surface is marked with those remarkable reticulations, 

 which, as far as I am aware, are peculiar to Hedysareœ ; the young legume is 

 often very much constricted between the seeds ; and even the non-articulation 

 may be accounted for by the underground maturation, a circumstance which 

 has usually the effect of rendering indéhiscent the iniderground legumes in 

 amphicarpous species of which the upper legumes are dehiscent. 



f^oandzeia resembles Arachis in having sterile perfect flowers and apetalous 

 fertile ones, which enter the ground to ripen the legumes, it has therefore 

 been usually associated with Arachis; and I myself, when siuày'mg Phaseoleœ 

 at Vienna, having no specimens of J^oandzeia at the time, referred it Mnth 

 Stt/Iosanthes to Hedysareœ : but in this I was mistaken ; the underground 

 fruits of most amphicarpous Legumlnosœ proceed from apetalous flowers, so 

 that that character alone would not connect Foandzeia with Arachis any more 



VOL. XVIII. Y 



