G2 LETTER XXXr. 



The stamens are numerous, and much shorter than the 

 petals ; they originate on the outside of a roundish, 

 flat, green cushion (^jig. 4. a.), that surrounds the 

 stigTiia, and caps the ovary. Tlic latter is inferior, 

 containing about eight cells, divided off* from each 

 other by strong dissepiments, but, what is very re- 

 markable, not bearing the ovules at the point where the 

 dissepiments come in contact, but producing them from 

 the centre of the back of each cell {fig. 4. h.). Hence 

 in this species we have the singular instance of a many- 

 celled ovary, with true complete dissepiments, and com- 

 mon parietal placentation. The stigma is sessile, and 

 divided into as many rays as there are cells in the 

 ovary. I must now warn you, that, although the spe- 

 cies before us has this curious arrangement of the in- 

 terior, yet you will not find the same structure in all 

 species ; on the contrary, in some, the back of the cell 

 simply presents a fleshy hump, from the lower edge of 

 which, and the base of the cell, the ovules originate ; 

 or, as in most cases, they simply grow from the lower 

 part of the inner edge of the cell. 



The latter structure is that of the ripe fruit I send 

 you for examination {fig. B. 5.). You will find that 

 it divides at the top into five valves, which close up 

 when the fruit is wet, and open when dry. Each of 

 its cells contains a considerable number of seeds {fig. 

 B. 6.), hanging from long stalks, that grow from the 

 lower part of the centre of the fruit. The seeds are 

 angular, and tuberculated {fig. 7-)» ^^^ contain a 

 curved embryo, lying on one side of the albumen. 



I have omitted to state, that in this and all the spe- 



