A TRIBE OF LEGUMINOSJR. 11 
In other respects, also, the two genera form a natural group ; the 
leaflets are either all alternate, or rarely a few pair in each leaf 
may be found opposite or nearly so, and always without stipelle ; 
in one Dalbergia and one Ecastaphyllum the leaf is reduced to a 
single large terminal leaflet. The inflorescence is more decidedly 
сушове than in any other genus of the tribe, and perhaps of the ~~ 
whole order. The primary branches of the panicle are usually 
alternate, the next ramifications are dichotomous, and the ulti- 
mate branchlets bear several flowers, sessile or pedicellate, ar- 
ranged singly along their upper or inner side only, with a small 
bract under each pedicel, and two bracteoles under the calyx. 
Sometimes the inflorescence is dichotomous from the base, with 
innumerable small flowers forming a perfect flat cyme. Some- 
times it is only in the last ramifications that the cymose arrange- 
ment can be traced, or the whole is reduced to one or to very few 
racemes, but these are then always unilateral. A similar arrange- 
ment, although much less decided, may be observed in several 
Macheria, but is quite lost in Pterocarpus, Lonchocarpus, and 
their allies. The flowers in Dalbergia and Ecastaphyllum are 
white, straw-coloured, or purple-red, never, as far as I can learn, 
really yellow, and they are generally the smallest in the tribe ; in 
Some species not above one line long, generally from that to three 
lines, and only very rarely five, or even near six lines long. 
The pod was the character formerly relied on chiefly, if not 
solely, for the separation of Dalbergia from other Leguminose— 
an oblong or linear thin flat indehiscent pod with one flat seed in 
the centre, in a cavity not much bigger than itself, or, if more than 
one seed, then always at some distance from each other, and each 
in its own cavity. The consequence of such a limitation was the 
Including in Dalbergia species very different in foliage, inflores- 
сепсе, and flowers, now referred to Derris and Lonchocarpus, and 
Would now necessitate the adding to it Platymiscium, Leptolobium, 
and many other widely distant genera which have the same pod, 
and the exclusion from it not only of Ecastaphyllum, but also of 
those species universally retained in it which constitute my sec- 
tion Selenolobium. I have now, however, as already observed, 
Very much restricted the generie importance I would attach to 
modifications of the pod, and were I now first grouping the species 
before me, I should certainly have proposed the uniting them all 
under Dalbergia, divided into three artificial sections :— Dalbergia, 
With a straight, long, thin рой; Selenolobium, in which it is thicker 
and lunate or reniform ; and Ecastaphyllum, where it is orbicular »* 
