MR. J. MIERS ON THE HIPPOCRATEACEX OF SOUTH AMERICA. 415 
bursts along a vertical transverse suture, thus opening bivalvately, somewhat as in Hip- 
pocratea. The ovary is trigonoidly columnar, depressed at its summit, and with deeply 
ineurved or hollow sides, at the bottom of which are the three lip-shaped discal cups in 
which the stamens are inserted; it is trilocular, the cells being placed within its three 
rounded angles, each containing two ovules attached to the axis a little below its summit : 
its depressed apex is slightly umbilicated; and from the centre radiates a sessile adnate 
stigma, formed of three divaricating rays that become free as they advance horizontally 
beyond the ovary, which they overhang in the form of three obtusely terete salient points, 
opposite to and directed towards the stamens. We see here, contrary to the usual rule 
in Hippocrateacee, that the stigmata are opposite to the stamens, and alternate with the 
cells of the ovary—a structure I believe to be hitherto unexampled among Dicotyledones, 
at least where the ovary has complete cells. It was shown by Mr. Robert Brown, in 
1840*, that this structure is frequent in Monocotyledones, especially in those with a tri- 
locular fruit; he also pointed out some exceptional instances among Dicotyledones, as in 
Parnassia, Papaveracee, and Crucifere, where a similar deviation is observable. The 
ovary in Parnassia is formed of four carpels, the inflected sides of which are agglutinated 
together, so as to form four dissepiments, with four intermediate cells; but these dissepi- 
ments are so far incomplete that they do not unite in the axis, and each pair of placentæ 
thus rests on the edge of a dissepiment quite free; and under this arrangement the 
stigmata stand opposite to the cells. Doctor Asa Gray, in his analysis of Parnassiat , 
shows a contrary arrangement, but subsequently he appears to have reversed this opi- 
nioni, and to agree that the stigmata alternate with the cells, standing opposite to the 
placentee, as Mr. Brown demonstrated$, and as Gaertner's drawing well shows|. Mr. 
Brown, moreover, inferred that in all cases each carpel is normally furnished at its apex 
with two stigmatic points, which are to be regarded as lateral, and not terminal, and that 
in ordinary cases they become confluent in the same carpel, and therefore opposite to the 
placentæ in à compound ovary with complete cells: instances, however, occur, as in Iris 
and Morea, where the lateral stigmatic terminations of each carpel do not coalesce, but 
remain distinct, under the form of alternating bifid stigmata, resulting from the partial 
confluence of each with that of the adjoining carpel; in other cases, again, as in most 
of the fridacee, the same divarication takes place in the terminations of each carpel, 
accompanied by the complete agglutination of these lobes with those of the adjoining 
carpels, thus producing a trilocular ovary, with the placentations in the axis of its com- 
plete cells, surmounted by a style whose three stigmata alternate with the cells, contrary 
to the general rule in Dicotyledones. If this be the organization in Kippistia, it is the 
first decisive instance of such an occurrence among dicotyledonous plants; for Parnassia 
and Papaveracee cannot be cited as parallel cases, because their placentation is not in 
the axis of the cells, but partakes more of a parietal character. 
It is not necessary to state here any thing respecting the fruit of Kippistia, as it is 
subsequently deseribed. . 
* PL Jav. Rar. p. 110. T Gen. Un. St. Am. pl. 86. 
+ Benth. € Hook. Gen. Plant. i. p. 1004. § PL Jav. Rar. p. 110. 
|| De Fruct. i. 287, tab. 60. 
3K 2 
