134 LhiuuuuL Society. 



STEMOPTERA. 



I'criantlmim ovario adnatum, sujua libemni, subinfundibuliforme : favct 

 tuigidA sacculis ,\ interiopbui auota : limbo 6-partito, laciniis aeutis, 

 Bstivatione marginibua induplicatis, 3 alternii brevioribus. Stamina 

 3, fauci adnata : Jilamentis complanatis, e margine sacculorum orienti- 

 bus bifurcatis, ramulo aingulo anthcrifero a'ato. Ovarium turbinatuni, 

 1-loculare, placentis •'{ parietalibus. Stylus longitudine staminum. Stig- 

 mata 3, recurvata, apice glandulifera, Capsula 1-locularis, polyspemia, 

 ■ubtrivalvis, apice 3-fisao dehiscens. Placenta; .'>, parietales. Semina 

 Dttmeroaiasima, BCobiformia, testa, nuclco vix exccdente, reticulata, are- 

 olis elongatis oblique dispositis. 



Plants (brasilienses) rhizocarpeee, r a dice fibrosa. Cavils erectus, subdi- 

 chotome ramosus, ramis subjlc.vuosis, patlidis, subpurpurasceulibus. Fo- 

 lia pauca, scssiha, crcrta, bract eiformia, pallida. Inflovescentia termi- 

 nalis, uni flora. Flores reefer is majores, ebracteati, purpuraseentes, Ap- 

 teria? Nutt. haud absimiles. 



\ . Stemoptera lilacina. — In uliginosis ad Serra dos Orgaos Prov. Rio de 

 Janeiro. 



All the species are described at length in the paper, and their cha- 

 racters are further illustrated by drawings, with details of the parts 

 of fructification. The author remarks that upon the same principle 

 that Apostasiacece have been separated from Orchidecc, and Xyridete 

 from Restiacece, these plants ought to constitute an order distinct 

 from Burmanniacea? ; but the difference between the unilocular cap- 

 sule with parietal placentation and the trilocular capsule with axile 

 placentation, which at first sight seems to ofFer a wide and well- 

 founded distinction, appears of less value when we consider that 

 the extensive order Gentianea? presents similar differences, toge- 

 ther with every possible gradation of transition from one extreme 

 to the other. He therefore inclines to the view of preserving all 

 within the natural order Burmanniacece, dividing it into two sub- 

 families, viz. 1. Burmanniea?, which will contain only the single 

 genus Burmannia (and perhaps the Gonyanthes of Blume may be 

 found to belong also to this section) ; 2. Dictyostegece, com- 

 prising Dictyostega, Cymbocarpa, Stemoptera, Apteria, Gonyanthes, 

 and Gymnosiphon. He then proceeds to show the close affinity which 

 BurmanniacecB bear to Orchideae, which often also present nearly a 

 naked stem, with imperfectly developed leaves, and instances are 

 moreover known in which they exhibit three distinct stamens and 

 three stigmata : they have also an unilocular ovarium, with parietal 

 placentation ; there exists also a close resemblance in the structure 

 of the walls of the capsule, and there is hardly any difference in the 



