70 FLORA VITIENSIS. 
* 
stipulis parvis; spicis axillaribus laxis; floribus albis, ad axillas bractearum parvarum sessilibus 
v. subsessilibus, odoratis, bracteolis minutis inconspicuis.—Renia, Noronha, in Vorhandl. Batav. Ge- 
nootsch. vol. v. p. 64. 
Jnocarpus is a genus the position of which in the natural system has been involved in great doubt. 
Jussieu referred it to the neighbourhood of Sapotacee, Sprengel to Laurinee, and Guillemin to Apocynee ; 
whilst Endlicher, combining it with Hernandia, formed it into a separate group, which he placed at the end 
of Thymelee. More recently, Bentham (Journ. of Linn. Society, vol. vi. p. 146) has expressed an opinion, 
which seems also to have been that of Solander (Prim. Fl. Ins. Pacif. p. 255, ined.), who regarded it as 
allied to Cynometra (* Fructificatio in multis convenit Cynometris, preeipue si suspicari liceat. Garciniwm 
lacinias corolle pro petalis distinctis descripsisse. Fructus ne quidem Cynometre nomen Leguminis mere- 
tur. Habitus autem diversa indicant genera, in Cynometris peculiaris foliis 2-natis.”). In the artificial 
system it was always placed in Decandria, aud near Sophora. What has puzzled botanists most is the co- 
hering of the petals and the drupaceous fruit. Bentham justly remarks that the former character is found 
in Trifolium, and the latter, amongst others, in Dipterye. But I must frankly own that I am not quite 
converted to the view of regarding Inocarpus as a Leguminosa, when it has neither a papilionaceous corolla 
nor a genuine legume. _My own investigations led me to look upon it as an anomalous Chrysobalanea. It 
has the habit, the simple coriaceous leaves, the inflorescence, the drupe, ete., peculiar to that Order. But 
the style is not basal, and the ovules not quite erect. 
l. I. edulis, Forst. Char. Gen. p. 65. t. 33; Prodr. p. 34; Plant. Escul. p. 50; Icon. (ined.) 
t. 143; Geertn. Fruct. vol. iii. t. 199 et 200. f. 1; Roxb. Pl. Coromand. vol. iii. t. 263.— Amotum 
fagiferum, Sol. Prim. Fl. Ins. Pacif. p. 255, et in Parkins. Drawings of Tahit. Plants, t. 47 (ined.). 
Nomen vernac. Vitiense, * Ivi."—Common throughout Viti (Seemann! n. 371; Sir E. Home! 
Barclay! n. 3453). Also collected in the Marquesas (Barclay !) and Society Islands (Banks and 
Solander! Forster!), Uvea or Wallis Island (Greffe! n. 34), Eromanga (M‘Gillivray !), and Java 
(Horsfield !). 
The Ivi, or Tahitian Chestnut, as it has been called by voyagers (Inocarpus edulis, Forst.), is one of the 
commonest trees in Viti, and when fully grown has a most venerable aspect. I still see in my mind's eye a 
fine group on the banks of a rivulet between Wairiki and Somosomo, producing a dense shade. Sixty, often 
eighty feet high, the Ivi bears a thick erown of oblong leathery leaves, small white flowers emitting a deli- 
cious perfume and kidney-shaped fruits, whieh contain a kernel resembling chestnuts in taste. The kernel 
is either baked or boiled, and eaten without further preparation, or grated on the mushroom coral ( Fungia), 
and made into puddings or bread (madrai). The stem is most singular. When young, it is fluted like a 
Grecian column; when old, it has regular buttresses of projecting wood. Ferns, orchids, and Hoyas 
frequently take up their abode on the soft spongy bark. The roots of old trees appear above the ground 
somewhat like those of the Bald Cypress of North America (Taxodium distichum, Rich.). Thousands of 
seedlings are continually springing up around the old plants, and nothing, save the dense shade of their 
parents, and the close proximity in which they grow to each other, exercises a check upon their engrossing 
all the adjacent ground. 
The tree is termed “If”? in New Guinea, and * Hi" in the Society Islands; “Nias” in Mallicollo, and 
“Emmer” in Tana. The Tahitians apply the names * Rata” and * Mape" to the fruit, which signifies also 
the kidney of an animal. Ellis (Polynesian Researches, vol. i. p. 374) furnishes the following popular 
account of it :-— : 
^In certain seasons of the year, if the breadfruit be scarce, the natives supply the deficiency thus 
occasioned with the mape or rata, a native chestnut. Like other chestnut-trees, this is of stately growth 
and splendid foliage. It is occasionally seen in the high grounds, but flourishes only in the rich bottoms 
of the valleys, and seldom appears in greater perfection than on the margin of a stream, From the top of 
a mountain I have often been able to mark the course of a river by the winding and almost unbroken line 
of chestnuts, that have towered in majesty above the trees of humbler growth. ... The trunk, which is the 
most singular, usually rises ten or twelve feet without a branch, after which the arms are large and spread- 
ing. During the first seven or eight years of its growth the stem is tolerably round, but after that period, 
as it enlarges, instead of continuing cylindrical, it assumes a different shape altogether. In four or five 
poe round the trunk small projections appear, extending in nearly straight lines from the root to the 
ranches. The centre of the tree seems to remain stationary ; while these projections inereasing, at length 
seem like so many planks, covered with bark and fixed round the tree, or like a number of natural but- 
tresses for its support. The centre of the tree often continues many years with perhaps not more than 
two or three inches of wood round the pith, while the buttresses, though only about two inches thick, 
extend two, three, and four feet, being widest at the bottom. I have observed buttresses not more than 
two inches in thickness projecting four feet from the tree, and forming between each buttress natural 
