V 
70 CONTRIBUTIONS TO BOTANY. 
for purposes of generic distinction. An exception to this rule 
has been maintained in the extensive genus Lundia, which may 
always be recognized from all others by its densely pilose an- 
thers ; but this, notwithstanding, is an artiticial character ; for 
by its adoption we find a considerable deviation from one com- 
mon form in the calyx, corolla, and stamens in several species 
which, but for this character, would be referable to other 
genera. 
The anther-lobes are fleshy on the dorsal face, and formed in 
front of a very delicate membrane, where they open by a longi- 
tudinal suture, the two margins of wdiich are greatly thickened. 
In cabinet specimens these margins often separate from the de- 
caying membrane, and stand out like aristae, for which they have 
been mistaken by some botanists. 
There are several other points of structure in the Bignoniacete 
which merit attention j but enough has been said for the purpose 
here intended, of calling the attention of botanists to the study 
of this interesting family. I will now therefore proceed to par- 
ticularize some of the features hitherto unnoticed or insufficiently 
explained in Adenocalymna, Anemopcegma, Dolichandra, Mac- 
fadyena, and several other genera, notifying at the same time 
many new species collected by me. I have also brought together, 
in several new groups, a great number of species that have either 
fallen under my observation, or that (not having been seen by me) 
are recognizable from the ample descriptions of authors ; they are 
scattered amongst the genera Bignonia, Spathodea, Tabebuia, 
Tecoma, See. In the descriptions that follow, which are confined 
almost entirely to plants of the New World, I have endeavoured 
to detail the specific characters as laconically as possible com- 
patible with the object in view, and to expose more amply the 
features that distinguish each genus or each peculiar group ; 
for to the want of' such details we must attribute the confusion 
now existing throughout the family. 
In this early stage of the investigation, I have not attempted 
any arrangement of the genera; and though I recommend 
the system of distribution adopted by the illustrious DeCan- 
dolle, I have not thought it necessary to follow it here. The 
remarks now ofiered must be considered, as they are intended 
to be, rambling contributions of observed facts towards a better 
knowledge of the family; they are given as mere examples 
of the groups proposed, and are confined either to the plants 
of my own collection, wffiich have enabled me to study their 
characters more fully, or to a portion of those, more especially 
typical specimens, in the herbarium of the British Museum, and 
to a very few in the rich and extensive Hookerian collection at 
