ON THE GENUS VILLARESIA. 263 
B. Stamens inserted outside the disk. 
Ovules erect, with a dorsal raphe. Petals imbricated . Celastracea. 
Ovules suspended, with a dorsal raphe. Petals iuibriciid. Aria 
Ovules suspended, with a dorsal raphe. Petals valvate. . 
This group (Celastrales) marked with the above prominent character 
might constitute the first of three sections into which the class Frangu- 
lacee of Endlicher may be divided; the second section would consist 
of the Rhamnales, distinguished by another and still more peculiar deve- 
lopment of the raphe; while the third would comprise such Orders 
as possess the ordinary ventral raphe, the stamens inserted inside or 
outside the disk, and erect or suspended ovules, such as Staphyleacee, 
Chailletiacee, Goupiacee, and others. It was under this view that I 
suggested, some time ago, to make Goupia the type of a distinct 
family, as it differs from all the Celastrales in having a ventral raphe, 
but otherwise agreeing with Hippocratacee in the insertion of its 
stamens inside of a bell-shaped disk. Goupia is certainly misplaced 
among the Celastracee in the ‘ Genera Plantarum,’ essentially differing 
in the position of its stamens. . 
It is manifest that Vil/aresia, from its organization, cannot be re- 
ferred to Olacacez, nor yet to Icacinacee, on account of the zstivation 
of its petals; the genus consequently falls into Aguifoliacee, as I 
showed long ago, differing from other genera of the family in the 
number of cells in its ovary. By referring to my diagnosis of the 
genus (Ann. Nat. Hist. 3rd ser. ix. 110) it will be seen to tally with 
every character which the authors of the ‘Genera Plantarum’ assign to 
Aquifoliacee. 
hile on this subject, I will offer a remark concerning Bursinopeta- 
lum, which Dr. Seemann in this ‘ Journal’ for July last (p. 205), refers 
to Cornacee, in which opinion he is in accord with the authors of the 
‘Genera Plantarum.’ I had previously endeavoured to show that the 
genus does not belong to Araliacee, where it had been assigned, but 
impressed by the singular resemblance in the structure of its seed to 
that of Villaresia, especially as it agreed in the dorsal position of its 
prominent raphe, according also in the number and position of its sta- 
mens and petals, the apical inflection of the latter, and their subimbri- 
cate estivation, I was induced to suggest its affinity to it; indeed, ex- 
cepting its inferior fruit, there is nothing to prevent the genus ranging 
by the side of Villaresia. In regard to this exception, I observed that 
