58 MR. BENTHAM ON LOGANIACE.E. 



1. Antonia, Pohl. 



The peculiar characters of this plant consist in the numerous 

 imbricated bracts surrounding the calyx, the short tube of the 

 corolla, and the broad peltate placentae, producing numerous ovules, 

 of which only one or two in each cell are ever found to enlarge 

 into seeds. It has been hitherto supposed that there are two 

 species, the one glabrous, the other more or less hairy, especially 

 on the under side of the leaves ; but the numerous specimens we 

 now possess from various parts of Brazil and Guiana show a gra- 

 dual passage from the one into the other, and it is seldom, even in 

 Pohl's original specimens, that the under side of the leaves is abso- 

 lutely without hairs. 



2. TTsteeia, Willd. 

 The great development of one lobe of the calyx, a circumstance 

 of which several examples exist in JRubiacece, and the constant 

 abortion of three out of the four stamens, afford good generic cha- 

 racters in listeria, although they do not appear of sufficient im- 

 portance to separate this single species into a distinct tribe. The 

 corolla, excepting in the number of its parts, and the placentas, 

 are as in Norrisia, with the capsule and seeds common to that 

 genus and Antonia. With regarcf to the reduction of the stamens, 

 it cannot be considered as any approach to the irregular flowers of 

 Scrophularinece, as it shows no tendency to didynamy, but it is 

 rather one of those exceptional anomalies such as that observable 

 in Carlemannia among Hedyotidece, where the stamens are reduced 

 to two, without any irregularity in the corolla. 



3. Noeeisia, Gardn. 



"Well described by Gardner, this plant differs from Antonia, 

 with which Wight proposed to unite it, in the want of the imbri- 

 cated bracts, in the slender tube of the corolla, and in the linear 

 placentae. Gardner describes and Wight figures the embryo as 

 reversed with the radicle uppermost, contrary to what we observe 

 in all allied Ginchonecd and Antoniece; but this may be a mistake. 

 The seeds of Griffith's specimens are almost all loose, and the two 

 ends are generally so exactly alike, that it is very difficult in 

 dissecting to be certain which end really lie uppermost in the 

 capsule. 



4. Gelsemitjm, Juss. 



This genus, most accurately described and properly placed by 

 Alph. DeCandolle in the ' Prodromus,' corresponds, as already 



