148 ME. G. BENTHAM ON INOCARPUS. 
under Thymelew. All these doubts may now be considered as 
removed by the reference of /nocarpus to Leguminose ; whilst 
Hernandia remains far away amongst t unisexual Monochlamydes, 
allied to those Euphorbiaces in which the albumen almost or 
completely disappears. 
The so-called nut, but rather the kernel or seed, of 7nocarpus 
edulis appears to be extensively eaten in its native country, and 
more especially in some of the eastern islands of the Indian 
Archipelago, where it is said to be very abundant. Rumphius 
says that, when boiled or roasted in ashes, it is sweet like the 
Spanish eatable acorns, much prized by the natives of several of 
the islands, and that in Machian they almost live upon it. Accord- 
ing to George Forster, it replaces chestnuts in the Society and 
Friendly Islands ; but is less agreeable, although sweetish, and is _ 
ill suited to weak stomachs. Roxburgh, who raised the tree in 
the Botanic Garden of Caleutta, says that the kernel is certainly 
eatable, but by no means palatable. Like the Htaballia, this tree 
appears to be hard-wooded and of considerable beauty. The 
flowers are described as of a pale yellow. 
According to Rumphius, the tree yields a resinous glutinous 
Juice, into which the Papuans steep the tips of their arrows, giving 
them a black colour; and this statement is copied by Forster and 
subsequent writers. There is some doubt, however, whether Rum- 
phius has not confounded two different trees in his article “ Ga- 
janum,” Herb. Amboin. i. 170, t. 65, universally referred to Inocar- 
pus onthe authority of Thunberg. The fruit is indeed so described 
by Rumphius as to leave little doubt as to its identity, but his 
representation of the flowers does not at all agree with those of 
Inocarpus. They are figured as borne on long pedicels in a short 
loose raceme ; and the petals are lanceolate, not linear. In all our 
specimens, wild or cultivated, they are either closely sessile, or the 
pedicel is so short as to be scarcely perceptible even after the 
fruit is considerably enlarged, although the stipes of the fruit may 
then, after the calyx has fallen off, answer the appearance of a 
pedicel. Forster says indeed that the flowers are “ brevissime pe- 
dicellati," and that the inflorescence is a raceme, and not a spike ; 
but, in the loose sense in which these words were formerly taken, 
he may mean that the flowers are distant from each other, and 
not close together. Forster also describes the flowers as occa- 
sioually 6-merous, with twelve stamens ; but this must have been 
accidental. We find them always 5-merous, and Roxburgh 80 
describes them. 
