170 
CONTRIBUTIONS TO BOTANY. 
carp, and are filled with a soft pith, which extends in a thin 
layer over the whole of the nut ; the latter is indehiscent, fusi- 
form, sharply pointed at both ends, thick, and osseous, with 
about eight deep grooves. It is justly remarked by DeCandolle, 
that in the ovary, as well as in the fruit, the cells are placed 
opposite to the wings, not alternate with them, as figured by 
Gaertner. 
From the above facts it is evident that, in order to eompre- 
hend the nature of the structure of the fruit in Halesia, it is 
necessary to trace the progress of the growth of its ovary from 
its earliest stages to the period of its ultimate development : we 
thus see how by degrees eaeh ovule is removed from its normal 
basilar position, and apparently carried up towards the centre, 
and is afterwards found in a parietal sac attached to the middle 
of the wall that surrounds a large central vacant space in the 
growing ovary ; but how this open sac extends itself into a com- 
pletely closed cell, at first membranaceous, and then hardened 
by thick osseous deposits, I am unable to explain. I can only 
vouch for the truth of the facts as they are here recorded*. 
In regard to the afllnities of the Styracecs, I have, in a pre- 
ceding page, stated my conviction, following the mode of infer- 
ence adopted by Jussieu, that, from the structure of its ovary, it 
should hold a position near the OlacacecB, Humiriacea, and Me- 
liacea-\. Later botanists have lost sight of the true affinities of 
this order, from having been led away by its supposed connexion 
with the Symplocacece, — a misconception which I have endea- 
voured to rectify. It has from this cause been placed near 
EbenacecB, on account of the partial agglutination of its petals 
and stamens into a tube ; but this character ought never to have 
been thus considered as one of such primary importance, because 
that union is never perfect ; for in Styracete, even more than in 
SymplocacecB, those parts may always be separated readily and 
without laceration. Prof. A. DeCandolle (Prodr. viii. 245), 
though he admits in some degree its relationship towards 
Meliacea and Humiriacece, yet considers such an affinity to be 
distant, on account of its ovaiy being sometimes inferior, also 
* A verj' analogous circumstance is recorded by Mr. R. Brown in the 
case of Persoonia (Linn. Trans, x. 35), where the ovarium is unilocular 
and contains two ovules ; after fecundation, a cellular substance is inter- 
posed between them, and this gi-adually indurating, acquires in the ripe 
fmit the same consistence as the putameu itself, from whose substance it 
cannot be distinguished ; and thus a cell, originally unilocular, becomes 
bilocular. The same occurs in Tribulus, where each cell of the ovary 
becomes thus divided into four osseous superposed cellules ; and a similar 
growth takes place in Bontia. 
t See also my remarks on the affinity of Styracece with Olacacece, Ann. 
Nat. Hist. 2 ser. viii. 163; huj. op. p. 23. 
