o48 Mr. MiERs on some new Brazilian Plants 



masses, according with the description given of that of Dicti/ostega. That 

 portion of the ovarium invested by part of the perianthiuin is of an ol)long 

 sliape, tapering at its base, and of a deep reddish-purple colour, but the 

 upper portion is free, of a pale colour, and tapers upwards in the forui 

 of a sharp cone, froin the sunmiit of which rises the erect filiform style, 

 which attains the height of the stamens, and then divides into three very 

 divaricate, sigmoid stigmata, each of which forms at its extremity a some- 

 what upright funnel-shaped cup, with an oval orifice, drawn together on 

 the two sides, and filled with a yellowish viscid fluid. The capsule very 

 much resembles that of Dicfi/ostega, but tlie seeds are different: they are 

 oblong, quite opake, of a j^eHowish-brown colour, and suspended by a slen- 

 der umbilical cord. The testa is marked with very prominent reticulations, 

 appearing by a common lens as if covered with twisted longitudinal la- 

 niellie ; but under a higher power it is seen to consist of elongated, hexa- 

 gonal cells. 



The plant was found by me in the Organ Mountains, in March 1838, in a 

 swampy situation under the shade of a large block of granite. 



I cannot close the enumeration of these plants without alluding to the two 

 species regarded by some as distinct from Barmannia under the name of Tri~ 

 pterella ; the one is Tripterella eapitata of Michaux, which Von Martins con- 

 siders to be the same as his Brazilian Burmannia eapitata ; but judging from 

 the specimen in the Herbarium of the British Museum, I am inclined to 

 believe they are two distinct species : the other is Tripterella cœrulea of 

 Elliott and Nuttall, which is the same as Burmannia hijida of Linneeus. It 

 is chiefly on the authority of Mr. Nuttall that Tripterella has been retained 

 distinct from Barmannia ; but as I cannot find that he has offered any evi- 

 dence of an existing difîerence, there seems no good ground for this sepa- 

 ration. 



From the facts now adduced, we find that the Burmannîaceœ comprise two 

 very distinct groups of plants, namely, those having a trilocular ovarium with 

 central placentation, and those having a unilocular ovarium with parietal pla- 

 centation. The first consists only of the two genera, Burmannia and Gony- 

 anthes ; the second contains four genera, viz. Dictyostega, Cymhocarpa, Apteria 

 and Gymnosiphon : that they are all very closely related there cannot be any 



