24 МВ. Ө. BENTHAM’S SYNOPSIS OF DALBERGIE Р, 
is nearly circular in its transverse section. It would thus include 
Cyanobotrys, Zuccar., which differs in the pod being much less 
constricted between the seeds. Its flowers and foliage, if rightly 
matched by Zuccarini, are precisely those of another Lonchocarpus, 
the West Indian L. violaceus. 
EvcnnarzsTA. 
А single species, the history and characters of which have been 
well elucidated by Bennett in the ‘Plante Javanice Rariores.’ 
In many respects it comes very near to some Sophoras, but the 
united stamens refer it without doubt to Dalbergiee. In the 
present arrangement it is intermediate, as it were, between the 
Lonchocarpuses and the Geoffroya groups. The inflorescence, ovary, 
and seeds are those of Geoffroya, whilst the foliage, united keel- 
petals, and thin pericarp bring it nearer to the Lonchocarpoid 
genera. 
ANDIRA and Gxorrhoxa. 
These two genera, originally united under the latter name, are 
certainly very closely allied to each other. Considered together, 
their flowers and foliage have nothing remarkable, resembling so 
much those of some other Dalbergiee that some of their species 
have in that state been occasionally confounded with Macheriwms 
or Lonchocarpuses. They may, however, be generally distinguished 
from both those genera by their free keel-petals and more pendu- 
lous ovules; and their fruit is very different, having a more or less 
succulent external sarcocarp, with the endocarp more or less 
hardened, and occasionally thick, and almost woody. In the half- 
grown or unripe state in which the fruits are sent by collectors 
for the herbarium they show no generic difference between Andira 
and Geoffroya, and but very little in the several species, varying 
chiefly in size and somewhat in shape,—more or less compressed, 
ovoid, or nearly globular, straight or oblique, obtuse or pointed. 
The fully ripe pods are but little known; they have been but very 
seldom and very vaguely described when fresh, and in our collec- 
tions they are very scarce, and generally more or less contracted 
or otherwise altered by desiccation ; so that, whatever diversity 
they may really show, they afford very little assistance in the 
characterizing species, still less in the delimitation of the two 
genera. l should, indeed, have proposed their reunion but for 
rather prominent differences in habit. Andira has branching 
panicles of violet, pink, or purplish sweet-scented flowers, whilst 
