220 Mr. J. Miers on Bursinopetalum. 
making a cross-section, I observed that, at the junction of the 
petals, the margins were not square and yalvate, but were all 
considerably inclined, their chamfered edges overlapping each 
other in a direction corresponding to the indications seen at the 
apex. The extent of this overfolding in the latter species (which 
is the one mentioned by Mr. Thwaitcs) is small, so that under a 
hasty examination it might be mistaken for a valvate junction 
of the margins; in the former species, the imbricate overlapping 
of the edges is so considerable as to admit of no doubt. 
In the general habit of Bursinopetalum, in the form of its 
simple leaves and simple joints, there is nothing to show any 
approach to the Araliacee : its inflorescence is always in a tri- 
chotomous panicle, not umbellate as in that family ; the ovary 
is only half inferior, and always unilocular, with a single sus- 
pended ovule: in these respects, as well as in the estivation of 
the corolla, Bursinopetalum is irreconcilable with Araliacee. On 
the other hand, the form of its calyx, its five petals imbricated 
in estivation, deeply inflexed at the apex, and with an internal 
carinated nervure—its five alternate stamens rising from the 
margin of the ovary, with subulate filaments spreading towards 
the base—a one-celled ovary, with an ovule suspended from a. 
longitudinal parietal expansion of the placenta—a simple style 
—a dry 1-locular drupe, with a coriaceous indehiscent putamen, 
enclosing a single seed that fills its: cavity and is moulded in 
a horse-shoe form round the longitudinal parietal expansion, 
which extends nearly to the centre of the cell—a fleshy albumen, 
enclosing in its summit an embryo with a superior terete radicle 
—are all characters surprisingly in accordance with Villaresia, 
the principal point of its disagreement being its inferior fruit. 
With this single exception, Bursinopetalum is quite conformable 
with the Aqguifoliacee; and even this exception, making due 
allowance for its aberrance, in great measure may be reconciled. 
The ovary, with its fleshy summit assuming the appearance of 
an epigynous disk, as in Styracee, is half superior, that is to 
say, its cell rises above the level of the insertion of the stamens ; 
and we find a parallel of this instance in Halesia. I have shown 
how it happens in that genus*, contrary to what occurs in Styrax 
and its congeners, that the superior moiety of the ovary remains 
almost unchanged, while the principal growth takes place in the 
lower moiety, from which it results that a half-superior oyary 
becomes converted into an inferior fruit. Now, precisely the 
same occurrence takes place in Bursinopetalum; and if on this 
account we were to deny its right to rank in Aquifoliacee, then, 
for the same reason, we ought to exclude Halesia from the Sty- 
racee, which few would venture to propose. For the same reason 
* Ann, Nat. Hist, 3 ser, iii, 137; Contributions to Botany, i, 168, 
