20G MR. O. BENT HAM OK EUPHORBIAC'EiE. 



and in the division of the ovary-cells into two compartments by 

 the protrusion of a spurious dissepiment between the two ovules, 

 a character unique in Euphorbiaceao* Two species have the dis- 



tant styles common in Buxus ; in a third (S. columnaris, Muell. 

 Arg.) they are close together and united at the base. Mueller, 

 in the 'Prodronius' (xvi. 1. 7), adopts Baillon's Buxaceae as a 

 distinct order, which however he leaves in close proximity to 

 Euphorbiacese ; and he adds to it the Californian Simmondsia, 

 notwithstanding the uniovulate cells of the ovary. Baillon, again 

 (* Histoire des Plantes,' vi. 47), agrees to the admission of Sim- 

 mondsia, but places the whole group in Celastrineae, according 

 to peculiar views which I feel quite incompetent to understand. 

 To me it appears that Buxeae, thus extended, constitute a marked 

 and well-defined though very heteromorphous group, but not of 

 a higher grade than that of a tribe of Euphorbiacese, allowing, as 

 in Stenolobeae, a specially exceptional essential character to over- 

 ride that derived from the uni- or biovulate cells of the ovary. 



4. Phyllanthe^. 



The division of the great mass of tropical Euphorbiaceae into 

 great groups according as the ovary-cells have one or two ovules 

 in each, originally sketched out by Adrien de Jussieu, and almost 

 universally adopted, is very near being a natural one. The in- 

 florescence of the biovulate Phyllanthece, always axillary, is, in 

 the typical genera, usually in sessile clusters or cymules, at least 

 in the males ; the calyx is rarely valvate, and occasionally dis- 

 tinctly 2-seriate, the sepals of the two series sometimes dissi- 

 milar ; the petals, in the few genera where they are present, are 

 usually small and scale-like, not always readily distinguishable 

 from lobes of the disk ; and the stamens, rarely more than twice 

 the number of sepals and mostly isomerous with them, have the 

 single or outer series opposite the sepals. In the uniovulate 

 Crotonece the inflorescence is usually spicate, racemose, or pani- 

 culate, and in several genera terminal ; the male calyx is uni- 

 8eriate and often valvate ; the petals, when present, are quite 

 corolline in their aspect and insertion ; and the stamens are often 

 indefinitely crowded in the centre of the flower — but when definite, 

 those of the single or the outer series are alternate with the 

 sepals. In the fruit, however, both these great tribes vary in the 

 same manner ; and the concordance of the above characters is by 

 no means constant. The latter groups of our series of Phyllan- 



