402 POLYNESIAN RESEARCHES. 
slender branches of the purau, or the leaves of a 
fine species of rush. The former was beautifully 
white and glossy, while the latter was of a yellow 
colour, and much more firm and durable, on which 
account it was preferred for hats. The only hats 
I wore in the islands during the subsequent years 
of my residence there, were made with this ma¬ 
terial ; and in that climate I should never desire 
any other. The use of hats increased so rapidly, 
that all the European thread in the islands was 
soon expended. There were no haberdashers’ 
shops at hand, whence a supply could be pro¬ 
cured ; recourse was therefore had to native pro¬ 
ductions. Some employed the long filaments of 
the dried plantain-stalk ; and others split the thin 
bark of the purau into fine threads or fibres, and, 
though not equal in strength to the twisted thread, 
both answered remarkably well. 
The bonnets were in many instances scarcely 
finished, when another difficulty met their possessors. 
They had observed that the wives and daughters 
of the Missionaries, however plain their dress, wore 
a riband and strings to their bonnets, and they 
had often observed a greater profusion of trim¬ 
mings attached to those worn by the wives of the 
captains, or the female passengers, in any of the 
vessels that touched at the islands; they therefore 
imagined that in point of improvement they might 
almost as well appear without a bonnet, as with 
one destitute of these appendages. These, how¬ 
ever, it was no easy matter to procure, and they 
would at that time, certainly, have been the last 
article a captain or trader would have thought of 
taking to the South Sea Islands for barter. A few 
of the chief women were furnished with an English 
riband, which was considered as valuable as an 
