2 MR. G. BENTHAM’S SYNOPSIS OF DALBERGIEA, 
bittorent i in appearance as are, for instance, the pods of Drepano- 
carpus lunatus and Macherium villosum, they are both formed, as 
it were, upon one plan, easily traced through the intermediate 
‘species, and the two genera are very unnaturally separated. On 
the other hand, a cursory glance at the pods of Macheriwm and 
Tipuana would suggest their generic identity, whilst a closer ex- 
amination shows a considerable difference in the origin and forma- 
tion of the winged appendage, and the distinctness of the two 
genera is confirmed by differences in the flower without any ap- 
pearance of intermediate links; Tipuana being, in fact, much 
nearer to Pterocarpus than to Macherium. So also the degree to 
which the corky thickening of the pod is developed in certain 
cases makes a very great difference in the external appearance of 
the fruits of species which may be otherwise so closely allied that 
they cannot be generically separated ; whilst the tendency to this 
corky development is observable only in certain genera, although 
it does not take place in all the species of those genera. 
These considerations have obliged me to modify in many in- 
stances the circumscription and characters of the genera/I pro- 
posed in my Vienna paper, and to reduce several of those which I 
considered as new, more especially those which I had founded on 
the fruit alone, without having seen the flower—always a very 
unsafe course, and particularly so among Leguminose. In one 
instance, indeed (that of the genus Corytholobiwm, which I had 
taken up from Martius's fruiting specimens and MS. notes, with- 
out either he or myself having seen the flowers), the plant turns 
out to be Polygalaceous. In a few other cases the flowering and 
fruiting specimens had been mismatched ; a source of error very 
frequent in describing the tree-vegetation of the tropics. Travel- 
ling collectors have very seldom the opportunity of gathering 
specimens in both states from the same tree, and where these 
collectors are not very methodical and precise in the notes they 
take at the time of collecting, the memoranda attached to the loose 
fruits or flowers of their specimens in herbaria are not always to 
be depended on. Owing to the intricacy with which the branches 
of different trees and climbers are intermixed in tropical forests, 
fruits, flowers, and foliage, which the careless traveller fancies he 
has gathered from different branches of one tree, have belonged, in 
fact, to different trees ; and mistakes are still more frequent when 
he picks up from the ground loose flowers or fruit, apparently fallen 
from the tree overhead. Moreover, I have occasionally met with 
convincing proofs that the memoranda attached to specimens had 
