POLYNESIAN RESEARCHES, 
287 
space of water within in the shape of a wedge^ the point 
of which terminates in the circular enclosure. These 
walls diverge in a direction from the sea^ so that the fish 
which enter the lake are intercepted only in their re¬ 
turn. They are so numerous through the whole extent 
of the shallow parts of the lake, that it seems hardly pos¬ 
sible for a fish to escape. The enclosures are exceedingly 
valuable; fish are usually found in them every morning, 
and furnish a means of subsistence to the proprietors, 
who have no other trouble than simply to take them out 
with a hand-net. They are also excellent preserves, in 
which fish may be kept securely till wanted for use. 
Each enclosure has its distinct owner, whose right to the 
fish enclosed is ahvays respected. Most of the fish from 
the lake are taken this way. The net and the spear are 
occasionally employed, but they rarely use the line here. 
They have a singular mode of taking a remarkably 
timorous fish, which is called au or needle, on account of 
its long sharp head. The fishermen build a number of 
rafts, which they call motoi; each raft is about fifteen or 
twenty feet long, by six or eight wide, and it is mtade wdth 
the light branches of the hibiscus or purau. At one edge 
a kind of fence or skreen is raised four or five feet, by 
fixing the hibiscus poles horizontally, one above the 
other, and fastening them to upright sticks, placed at 
short distances along the raft. Twenty or thirty of these 
rafts are often employed at the same time, the men on 
the raft go out at a distance from each other, enclosing 
a large space of water, having the raised part or frame 
on the outside. They gradually approach each other 
till the rafts join, and form a connected circle in some 
shallow part of the lake. One or two persons then go in 
a small canoe towards the centre of the enclosed space, 
