POLYNESIAN RESEARCHES. 
283 
head was fastened to the open work at the bottom by- 
strings of finely-braided cinet^ made with the fibres of 
the cocoa-nut husk. The drums beaten as accompani¬ 
ments to the recital of their songs^ were the same in 
shape^ but smaller. They were all neatly made^ and 
finely polished. The large drums were beaten with two 
heavy sticks, the smaller ones with the naked hand. 
When used, they were not suspended from the shoulders 
of the performers, but fixed upon the ground, and con¬ 
sequently produced no very musical effect. The sound 
of the large drum at the temple, which was sometimes 
beaten at midnight, was most terrific. The inhabitants 
of Maeva, where my house stood within a few yards of the 
ruins of the temple, have frequently told me, that at the 
midnight hour, when the victim was probably to be 
offered on the following day, they have often been 
startled from their slumbers by the dull, deep, thrilling 
sound of the sacred drum ; and as its portentous sounds 
have reverberated among the rocks of the valley, 
every individual through the whole district has trem¬ 
bled with fear of the gods, or apprehension of being 
seized as the victim for sacrifice. 
The sound of the trumpet, or shell, a species of murex 
used by the priests in the temple, and also by the herald, 
and others on board their fleets, was more horrific 
than that of the drum. The largest shells were 
usually selected for this purpose, and were sometimes 
above a foot in length, and seven or eight inches in 
diameter at the mouth. In order to facilitate the blow¬ 
ing of this trumpet, they made a perforation, about an 
inch in diameter, near the apex of the shell. Into this 
they inserted a bamboo cane, about three feet in length, 
which was secured by binding it to the shell with 
