682 Mr. Bentuam’s Account of two new Genera allied to Olacineæ. 
have thought it more convenient to keep them together under the general 
character above given. Those species of Olax in which the dissepiments of 
the ovary are almost obliterated form the connexion with Opilieæ, and, as far 
as known to me, Gomphandra appears to join Opilieæ with Icacinew, whilst 
Pogopetalum is in many respects equally allied to O/acec and Icacinee. 
The first tribe, where the double nature of the floral envelopes is most evi- 
dent and complete, is that which comes nearest to the various polypetalous 
orders, with which Olacine have been compared. But among these, I can- 
not admit the affinity to durantiacee. The structure of the seed, where 
it is so constantly and materially different as in these two orders, becomes 
one of the most important characters in the classification of Dicotyledons; 
and it is here accompanied by considerable differences in the fruit, in the 
ovary, in the estivation of the corolla, in the texture of the vegetable organs, 
and other points of habit. And those peculiarities, in which Aurantiacee 
appear to resemble O/acinec, such as the tendency of the petals to cohere by 
means of the stamens, the persistent calyx, &c., occur in other orders which 
are less removed by essential characters. 
Meliaceæ, nearly allied to Aurantiaceæ, are about equally distant from 
Olacineæ in essential characters, and more so in habit. 
Humiriaceæ, a small order usually associated with Meliaceæ and Aurantia- 
cee, differ from them in habit, in the ovary, the fruit and the seed; in all of 
which points they approach much to Olacineæ. The dissepiments of the ovary 
are, when young, not quite complete, and the embryo lies in the axis of a 
copious albumen. They differ chiefly from Ofacinec in the æstivation of the 
corolla, which is not so decidedly valvate, in the stamens often indefinite in 
number, in the dilatations of the connectivum of the anther, the presence of a 
disc round the ovary, the fruit more generally plurilocular, and the larger 
embryo with a long radicle; but, upon the whole, they are probably, among 
dichlamydeous plants, those which come nearest to Olacinec. 
Styraceæ*, which are perhaps more allied to polypetalous orders than to 
those which are essentially gamopetalous, are very near also to Humiriacee as 
well as to Olacinee. The structure of the fruit and seeds may indeed be 
compared to those of Schepfia, that of the ovary in some respects to Pogope- 
* Including Symplocee and Halesiaceæ of Don. Vid. Trans. Linn. Soc. Lond., vol. xviii. p. 230. 
