64 
CONTRIBUTIONS TO BOTANY. 
the SesamecB) but, from the great difference in the normal consti- 
tution of their carpels, this union cannot be maintained. End- 
licher, on the other hand (Gen. Plant. 709 and 723), places the 
Sesameee at a considerable distance in the system from Pedalinece, 
ranking the latter before Orobanchacece, and considering the 
former as a suborder of the Bignoniacece, from which they cer- 
tainly differ very widely. In many respects the Sesamacece will 
be found to approach VerbenacecB : there is a great similarity in 
the form of their calyx and corolla ; but in the one their divisions 
are five, and in the other four ; Priva, however, is 5-merous, as 
in Sesamum. Ceratotheca has four didynamous stamens, as in 
Verbena ; while Priva has the rudiment of a fifth stamen, as in 
Sesamum. Ischnia verbenacea has flowers like those of Verbena, 
and has the habit of that genus. In Tamonea and Verbena 
{Shuttleworthia) the anthers are furnished with apicular append- 
ages, as in Sesamum and its congeners; and the capsule of 
Priva splits into two halves, each 2-celled, with erect seeds, 
thus offering much analogy with the fruit of Sesamacece : the 
seeds in both cases are exalbuminous, with an embryo of similar 
form. The capsule, also, is often echinate or cornute in Priva 
and Tamonea, as in Sesamum and Ceratotheca. In habit there 
is also much accordance between Priva and Sesamum. 
Tourretia has always been considered as a doubtful genus of 
the Bignoniacece ; but if the structure of the fruit be correctly 
described, it evidently belongs to Sesamacece, as Fenzl long since 
indicated * : it has a 4-celled capsule, which opens only at the 
apex by a gaping transverse fissure, greatly after the manner of 
dehiscence in Sesamum-, and the seeds are affixed to a central 
axis, as in that genus, and are erect, not transverse as in Bigno- 
niacece, showing that this structure results from the combination 
of four carpels, whose placentiferous margins meet in the axis — 
a structure quite incompatible with Bignoniacece. In regard to 
its echinate fruit, the retrorsely uncinate spines that cover its 
capsule are precisely analogous to those of Harpagophytum, 
which, from its axile placentation, certainly belongs to Sesdma- 
cece, and not to Pedaliacece (we find a corresponding tendency 
to the production of spines in the capsules of Ceratotheca and 
Sporledera) ; and its seeds have a similar rugosely expanded 
border. The principal difference is in the habit of the plant, 
which, though herbaceous, has conjugate leaves, its leaflets 
being again palmately divided ; and they have an intermediate 
cirrhus, which is also pinnately branched. Harpagophytum and 
Sesamum have tripartite or palmatifid leaves, laciniately divided, 
but they have no cirrhus. But as Bccremocarpus and Calampelis 
* Fenzl, Denksch. Regensb. iii. 211 ; A. DeCandolle, Prodr. ix. 236 in 
adnotat. 
