674 Mr. Bentham's Account of two new Genera allied to Olacineae. 



wanting. The anthers are always introrse, fixed by the back, bilocular, the 

 cells nearly parallel and opening longitudinally ; they aie usually versatile, 

 but adnate in Gomphandra. 



Thick matted hairs are present in the flowers of many species ; along the 

 middle nerve of the petals on the inside in Ximenia, Leretia and Pogopetalum ; 

 immediately above the part where the stamens cohere to the petals in most 

 species of Olax and in Scliœpjia fragrans ; at the top of the filaments in some 

 Gompliandra\ 



The pistillum is simple and sessile on the torus, which is sometimes thick 

 and of a glandular texture, but never encircling the ovary or projecting in the 

 form of distinct glands. The style is simple, terminating either in a thin, 

 truncate, apparently entire stigmatic surface, or in a two-, three- or four-lobed, 

 thick, fleshy stigma. 



The ovary is thick and fleshy, containing one, or (in Pogopetalum) three 

 small cavities, in which are one, two, three or four ovules suspended from the 

 apex of a placenta which arises from the base of the cavity, and is entirely 

 free, or more or less connate with spurious incomplete dissepiments, or with 

 the side of the ovary next to the style, which is in the latter cases excentrical. 

 In Ximenia, Ileisteria, Olax and Schœpjia the placenta is nearly central, bears 

 three or four ovules, and is more or less connected with as many spurious 

 dissepiments, which are exceedingly short in the Australian and some East 

 Indian species of Olax, but reach nearly to the insertion of the ovules in 

 other species of Olax, in Ximenia, and in Schajrfia. In Opilia and Cansjera 

 the placenta is almost entirely free, and has the appearance of an erect ovule, 

 but under a strong glass it is seen to bear a single minute ovule suspended from 

 the apex ; in the young bud, it has appeared to me that there are two ovules 

 in Opilia, three or four in Cansjera, a circumstance rendered probable by the 

 evidently compound nature of the stigma in both genera, but which, on ac- 

 count of the excessive minuteness of the parts, I am unable to ascertain with 

 certainty from dried specimens. After fecundation I never find traces of more 

 than one ovule. 



In Icacina, {Gomphandra ?,) Jpodytes and Leretia there are always two ovules 

 collaterally suspended, or nearly so, but (in Apodytes, at least), owing to the 

 lengthened thread by which one of them hangs, they are placed above one 



