58 
CONTRIBUTIONS TO BOTANY. 
lately Dr. Seemann has supported this opinion. DeCandolle 
divided it into two sections — the Tanaeciece, possessing a bilocular 
ovary, and the Crescentiece, a 1-celled ovary — all being distin- 
guished from the Bignoniea by their indehiscent fruit and 
apterous seeds. Dr. Seemann, in maintaining its claims to ra nk 
as a distinct family, also separates it into two sections under the 
same names*; but he simply distinguishes the Tanaedea by a 
persistent, and the Crescentiece by a deciduous calyx; and he 
affirms, contrary to the statements of preceding botanists, that 
all alike possess at an early stage a unilocular ovary with pa- 
rietal placentations, the fruit becoming bilocular by the subse- 
quent enlargement and confluence of the placentse : this view 
is not confirmed by the analyses I have been able to make, and, 
as regards the Tanaeciece, is not supported by the evidence on 
record, which I here reproduce. 
First, as respects Colea, the several details of C. Mauritanica 
(Bot. Mag. tab. 2817), of C. Telfairii {ib. tab. 2976) t, and of 
C. floribunda (Bot. Reg. v. 27, tab. 19), all prove most distinctly 
the presence of a broad membranaceous wing around the seeds, 
as in Bignonia ; and the capsules, though covered by a some- 
what fleshy epicarp, indicate, by well-marked grooves, the sutural 
lines of their dehiscence into two valves. Prof. Bindley has 
remarked that no instance is known of the existence of winged 
seeds in indehiscent pericarps; for as the function of the wing 
of the seed is to carry it from a height to a long distance by the 
force of the wind, this object could not be effected were the fruit 
indehiscent. Colea, with its w'inged seeds in a 2-valved capsule, 
and its ecirrhose pinnated leaves, may probably find its place 
near Tecoma, among the Catalpece ; but if, as Sir Wm. Hooker 
states, the valves of its capsule be parallel to the dissepiment, it 
must belong to the Bignoniece. In regard to the structure of 
the ovary in Colea, Prof. A. DeCandolle found it to be distinctly 
bilocular. We have not as yet sufficient knowledge of the struc- 
ture of Phyllarthron and Periblema to enable us to judge of their 
true position : in the latter the ovary is bilocular, with only two 
ascending ovules in each cell, attached to the dissepiment, and 
the calyx is enclosed in a tubular ventricose 4-fid involucre, 
which characters, as Prof. DeCandolle remarks, are quite foreign 
to the order. Of Phyllarthron very little is known. Even in re- 
gard to Tanaecium, our information concerning the structure of 
the ovary, fruit, and seed, as far as has been heretofore known, 
has not been sufficiently positive. The genus was established 
* Bot. Herald, 181 ; Proc. Linn. Soc. ii. 269. 
t If the presence of a wnng on the seed of C. Telfairii be questioned, 
there can be no doubt of its existence in C. floribunda. 
