330 NOTE ON vav.5:a and ehytidandra. 



from all those previously examined in having only twice as many stamens as petals, — 

 conforming in this respect to the type of the anclrcccium in the order Meliacece, except 

 only that the filaments are not monadelphous to the top. The anthers being rather 

 smaller than usual, and containing little good pollen, while the pistil is well devel- 

 oped, I am led to suspect that the difference may be attributable to sex, and that the 

 flowers may be more or less polygamous, as in Aglaia, Sec. ; which is the more 

 probable, inasmuch as these occur on a specimen which bears, on a lower and earlier 

 inflorescence, some nearly mature fruit. It was apparently these decandrous blos- 

 soms that misled Mr. Rich, the Botanist on the Expedition, preventing him from 

 recognizing the plant which he had previously marked as a probable relative of 

 Canella* while these specimens were ticketed and even figured as a Styrax. The 

 drawing of the plant was accompanied by some erroneous analyses, in which I had 

 failed to identify the Vavtea, and therefore had left the specimens among other Styra- 

 cacecB without examination until now. 



The fruit of Vavaa proves to be a berry, as was anticipated from the fertilized and 

 half-grown ovary. It is rather dry, four or five lines in diameter, subtended by the 

 small persistent calyx, and three-celled or four-celled by thin dissepiments, which per- 

 haps are obliterated when only one seed matures. A single seed is sometimes matured 

 in each cell ; and in one instance both ovules were fertilized in the same cell. The 

 seeds are oval, about three lines long, smooth, destitute of any arillus, ascending 

 from near the base of the cell, closely sessile ; the linear hilum being attached directly 

 to the axis of the fruit without any funiculus : the testa chartaceous, or perhaps some- 

 what fleshy, its whole base occupied by a large orbicular chalaza, which is connected 

 with the hilum by an extremely short rhaphe. The hilum extends from near the 

 base to about the middle of the seed. There is a rather fleshy inner integument of 

 the seed, but no albumen. The embryo consists of a pair of orbicular-oval, plane, 

 flat or plano-convex, fleshy, peltate cotyledons, which are cordate by a narroAV and 

 deep sinus: the radicle is superior, remote from the hilum, slender, but wholly re- 

 tracted and concealed within the sinus. 



The carpological characters, therefore, manifestly confirm the suggested relationship 

 of this genus to the Meliacece, where the cxalbumincus embryo assigns it to the tribe 

 Trichilieie. 



Simple and undivided leaves occur, as is well known, in three genuine Meliaceous 

 genera. The cup-shaped disc is partially united with the andrcecium in Trichilia, Eke- 



* Botany of United States Exploring Expedition, 1. c. p. 246. 



