POLYNESIAN RESEARCHES. 
387 
cocoa-nut^ or fern leaves^ is first laid on the upper part 
of the thatch, and afterwards a species of long grass, 
called aretUy is curiously fixed or woven from one end 
to the other, so as to remain attached to the thatch, and 
yet cover the ridge of the house. 
The roof being finished, they generally level the 
ground within, and enclose the sides. In the erection of 
my house, this part was allotted to the king’s servants. 
About thirty of them came one morning with a number 
of bundles of large white purau poles, from two to three 
inches in diameter. After levelling the floor, they dug 
a trench a foot deep round the outside, and then, cut¬ 
ting the poles to a proper length, planted them an inch 
and a half or two inches apart, until the building was 
completely enclosed, excepting the space left for a door 
in the front and opposite sides. In order to keep the 
poles in their proper place, two or three light sticks, 
called tea, were tied horizontally along the outside. 
Partitions were then erected in the same manner, as 
we were desirous, contrary to the native practice, to have 
more than one room. The house was now finished, and 
in structure resembled a large birdcage. In two of 
the rooms we laid down boards which we had brought 
from Port Jackson, and either paved the remainder of 
the floor with stones, or plastered it with lime. The 
outside was skreened with platted cocoa-nut leaves, lined 
with native cloth. This also constituted our curtains, 
and, hung up before the entrance to some of the apart¬ 
ments, answered the purpose of a door. Thus fitted up, 
our native house proved a comfortable dwelling during 
the months we remained at Afareaitu. 
The houses of the natives, although varying in size 
and shape, were all built with the same kind of materials, 
