2 
or more-celled with axile placentas, whilst in the latter the eee 
tion is parietal. In many species of Gardenia, however, we have found 
that the placentas towards the base of the ovary are frequently coherent, 
and a transverse section at this point would be the same as one across 
the ovary of a typical species of Randia. It has long been our opinion 
that these two genera should be entirely recast, the distinguishing 
characters being taken from the shape of the corolla ‘tube and the nature 
of the stipules. These characters, associated with the number of the 
ovarian cells and the position of the placentas, would probably result 
in the segregation of a few (4-6) smaller and fairly distinct genera. 
The proper position of the Mrs here figured, in spite of the axile 
placentation, seems to be in the section Rothmannia, which was described 
as a genus by Thunberg.—J. Hurcuinson. 
Fic. 1, calyx and pistil; 2, corolla laid open; 3, transverse section of ovary ; 
4, fruit; 5 and 6, seed. Aili ‘enlarged except 5. 
— 
