20 MR. G. BENTHAM’S SYNOPSIS OF DALBERGIES, 
gested by а general consideration of the Asiatic species, and these 
groups have been adopted by Miquel in his ‘ Flora van Neder- 
landsch Indie,’ with the addition of two new small genera, but 
none of us had attempted any comparative examination of the 
American and Asiatie species. This investigation I have now been 
compelled to take up, in order to give some precision to the generic 
characters for Martius’ Brazilian Flora, and the result has been 
а redistribution of some species, and a reduction of seven or eight 
small genera whose characters proved to be of no more than spe- 
cific value. 
After a general examination of all the species of which I had 
specimens, the first difficulty occurred in establishing a line of de- 
marcation between such of these genera as should be retained 
among Dalbergiee, and the arborescent Galegee with which they 
are so closely connected, and I confess that I have been unable to 
find any character more precise than that of the pod, indehiscent 
in the former, opening in two valves in the latter. This distinc- 
tion may indeed be often inferred from the inspection of the ovary, 
of which the cavity is usually larger in proportion to the ovules, 
where the valves are afterwards to become free, than when they 
remain firmly adherent all round the seeds, and it is frequently 
accompanied by differences in habit, and other unimportant cha- 
racters ; still the passage from the one to the other is, in some 
instances already alluded to, so gradual, that the limits between 
the two tribes remain far from satisfactory. 
This definition has obliged me in the first place to bring into 
Dalbergiee all the true Lonchocarpi as well as Miillera and 
Piscidia, in which the valves cohere as closely round the seed as 
in any admitted Dalbergiee, and, on the other hand, I have retained 
in or referred back to Galegee the Gliricidias included by DeCan- 
dolle under Lonchocarpus, Arnott’s genus Millettia, and all those 
closely-allied species for which I had proposed the genera Mundulia 
and Otosema, in all of which the pod, though often thick and hard, 
or almost woody or fleshy, still appears always to open in two valves 
when quite ripe. To the same group belong Marguartia, Vog., 
described as having the stamens free,—an error arising from the 
examination of very young buds before the staminal sheath has 
grown up; Padbruggea, Miq., in which, however, the ripe pod has 
not been seen; Berrebera, Hochst, and Fornarinia, Bertol., all 
evidently congeners of Millettia, or of its two above-named allies. 
How far these three genera should be reunited or remain distinct, 
and how far they should be distinguished from the American ӨЙ- 
