356 FLORA NIGRITIANA. 
apices ramulorum pedicello 3-4 lin. longo fultis. Bractee 
minute, subulate. Calyx fere 3 lin. longus, glaberrimus, 
levis, basi acutus, supra ovarium paullulum constrictus ; 
dentes lati obtusissimi brevissimi v. fere obsoleti. Petala 3 
lin. longa. | Ovarium glaberrimum, per anthesin mediantibus 
costis fere ad apicem calyci adnatum, vertice libero nudo 
calyce tamen breviore. Stylus basi attenuatus, apice trun- 
catus. Capsula membranacea, vix valvatim dehiscere videtur 
sed pericarpium membranaceum cum calyce tenuissimo saltem 
in speciminibus siccis disrumpitur. 
This plant is so very near in most respects to the American 
genus Spennera, that I should have considered it as a species 
of it, but that in the. present state of our knowledge of 
Melastomacee it seems necessary to keep separate those genera 
where the cells of the ovary are reduced below the number of 
parts of the floral envelope, from those which are strictly iso- 
merous. 
1. Spathandra cwrulea, Guill. et Perr. Fl. Seneg. p. 113. t. 71. 
—Sierra Leone, Vogel, Don; Senegambia. 
It is to the tribe of Memecylee, not to that of Charianthee, 
that this genus should be referred; indeed it is only to be 
distinguished from Memecylon itself by the three-ribbed leaves 
and the cotyledons, which are more fleshy and less plaited, 
although in Spathandra cerulea they are to a considerable 
- degree irregularly wrinkled and folded. The unilocular ovary 
is not alluded to in the ordinal character either in Endlicher’s 
Genera or in Lindley’s Vegetable Kingdom, although in both 
works Spathandra and Memecylon are included among Melas- 
tomacee : the little importance, however, of this anomaly ™ 
certain calyciflorous Orders is shown, as well by its occasion 
occurrence among Myrtacee and Rubiacee, as by the very 
natural genus Mouriria, (the American representative of Meme- 
cylon), which includes both unilocular and plurilocular species 
This circumstance is much in favour of the supposition, that 
in these Orders the placentz often proceed on Schleiden's 
theory from a prolongation of the axis, and not from the 
margins of the carpellary leaves. 
