CONTRIBUTIONS TO BOTANY. 
55 
anomalous genus Oxycladus will naturally find a place, as also 
the Monttea and Reyesia of Gay* — genera • evidently allied 
very closely to it, but of which little is yet known. We have 
thus a third natural tribe, the Platycarpea. 
I have already described (in p. 50) the development that takes 
place in the fruit of Jacaranda, which is quite analogous to that 
of Fridericia, Calampelis, and Eccremocarpus. Here the ovary, 
as in the Platycarpea, is normally constituted of two carpels 
only, which are in like manner placentiferous on their midribs, 
but they are differently arranged, being placed with their pla- 
centiferous lines opposed to each other, and conjoined picr. ] 4 . 
by their sterile margins, as in fig. 14. The ovary is / > 
therefore unilocular, with two opposite longitudinal I ' 
parietal placentae, and the result is a compressed, 1- )£ 
celled, 2-valved capsule, with the seeds attached to the /( 
middle of the valves, which open along the sutural \ 
line of union of the original carpels. These genera thus con- 
stitute a fourth very natural tribe, the Eccremocarpea, a group 
of greater extension than the subtribe of the same name of 
DeCandolle. 
It appears to me that the fourth subtribe of DeCandolle, the 
Incar villecR, cannot be maintained. That group consists only of 
Incarvillea and Amphicoma. In regard to the former genus, I 
am able to confirm the accuracy of the details given by Correa da 
Serra (Ann. Mus. viii. 391, tab. 63. fig. 2), which prove that the 
structure of its fruit and the position of the seeds, in the only 
known species, are precisely the same as in Argijlia, which has 
been noticed in page 160. The resilient process often observed 
in the anthers, which seems like an arista, has been urged as a 
distinctive character; but this arises (as was long ago indicated 
by Mr. Brown, PL Jav. Rar. Ill), not from any emanation of the 
connective, but from the rending of the thickened nerve-like 
sutural margins of the anther-cells, which separate at the base 
and remain attached at the apex, as I have shown to occur in 
Argylia. It was upon this circumstance that Presl was induced 
to found his Oxymitra, a genus which cannot stand. A similar 
resilience in the anther-cells is occasionally met with in other 
Bignoniaceous genera, for instance in the Pyrostegia o Presl 
[Bignonia venusta). The general habit of Incarvillea is quite 
that of Argylia, as is acknowledged by DeCandolle ; and there is 
a remarkable similarity in its leaves, which are in like manner 
bipinnatisected, with linear segments; it has also a terminal 
raceme, with large handsome crimson flowers. Incarvillea, 
therefore, may safely be placed among the Caialpeat, and near 
Argylia. 
* Gay, Chile, iv. 416, tab. 51 j ibid. 418, tab. 52 ; Wain. Ann. iii. 92, 95. 
