BOTANICAL INFORMATION. 239 
Scrophulariaceous or axile group, the four placente are united 
together in the centre of the ovarium, whilst in the Bignonia- 
ceous or parietal group, they are more or less removed from 
the centre in pairs, the interval being either empty or occu- 
pied by a spurious cellular dissepiment. This distinction is 
of little or no importance in some calyciflorous orders, while 
among Corolliflore it appears to be usually connected with 
great general differences, 
The main difficulty experienced by De Candolle in the 
Bignoniacee consisted in the very imperfect state in which 
specimens are usually sent home by collectors. Most of the 
species are large climbers, which attract the notice of travel- 
lers by their handsome flowers, but are difficult to gather or 
to dry well when gathered, and the fruits are generally entirely 
neglected, or what is worse, mismatched. - Yet it is the fruit 
and seed which appear to afford the best characters, and 
are accordingly endeavoured to be made use of for the de- 
mareation of the tribes and genera. Unfortunately, how- 
ever, after establishing genera from the structure of the fruit 
of one or two species, the author has been obliged to asso- 
ciate with them many other species from similarity of aspect, 
and notwithstanding the excellent eye for a natural group 
Possessed by the late Professor De Candolle, it is probable 
that considerable alterations will hereafter require to be made 
in the circumscription of several of the genera: we have not 
indeed as yet sufficient data to determine satisfactorily the 
validity of the primary divisions derived from the dehiscence - 
these plants the placentæ proceed from an elongation of the floral axis, or 
from the inflexed margins of the carpellary leaves, for however we might in 
the case of some of these orders (or perhaps only in some genera) incline to 
the former opinion, much sound argument may also be adduced on the other 
side ; and practically speaking, for systematic purposes, the decision is of 
little Or no importance, Under both theories, the placentæ must be 
admitted to be always double in number to the carpellary leaves, often 
intimately connected with each other (with or without the intervention of 
à central column), and with the margins of the carpellary leaves or with 
mr only—or, as the fruit ripens, more or less detached or de- 
