Mr. Bentham's Account of two new Genera allied to Olacinese. 681 



structure of the ovary ; Endlicher * associates it with Symplocos, from which it 

 differs in the aestivation of the corolla and the ovary incompletely divided, two 

 points in which it agrees remarkably with Olax and Ximenia. The objection 

 as to the adherence of the ovary entirely loses its force, when it is considered 

 that the greater part of that organ as well as the margin of the calyx are free, 

 and that an adherence almost as complete exists in some species of Olax. 

 The gamopetalous corolla is also a character of little consequence in orders 

 where the {estivation is valvate, and it exists to a considerable degree in Olax 

 itself. The stamens of Schoepjia are more closely adherent to the corolla than 

 in otlier Olacinece, but the filaments are filiform and prominent from the base 

 of the corolla, and not confounded with its substance, as in those orders where 

 they are truly epipetalousf. 



Cansjera has been usually referred to Thymelece, from which it differs in the 

 nature of the floral envelopes, in the position of the stamens, in the structure 

 of the ovary, and in that of the fruit ; and in all these points it agrees with 

 Opilia, from which it differs only in the coherence of the petals. 



The close affinity of Apodytes to Icacina will be evident from a comparison 

 of my figure and description with those of Adr, de Jussieu. Pogopetalum dif- 

 fers, however, in the compound nature of the ovarium ; but each cell with its 

 ovules is so exactly the single cell of Apodytes, and the flowers agree so well 

 with the general character of Olacinece, that the approximation will not be 

 found, it is believed, to violate materially any important natural affinity. It 

 is true that the ripe fruit and seed are as yet unknown, which leaves some doubt 

 on the matter, though it does not occur to me that it comes near to any other 

 order. Leretia is in almost all respects intermediate between Icacina and 

 Pogopetalum. 



It will be observed, however, that the three gi-oups, which I have designated 

 as tribes, present some material differences from each other, especially in the 

 structure of the ovary, and may perhaps, when better known, be considered as 

 so many distinct orders. But with our present imperfect knowledge of seve- 

 ral genera, and considering their evident connexion in many material points, I 



* Genera Plantarum, p, 744. 



t Mr. Brown informs me that he considers Schoepjia as a true Santalaceous and not an Olacineous 

 genus ; at all events, it is an additional proof of the close affinity of the two orders. 



