512 
POLYNESIAN RESEARCHES. 
taining several springs, and having within its precincts 
the principal temples of their tutelar deity. The walls 
are of solid stone-work, varying in height from six to 
twelve feet. They are even at the top; in some places 
ten or twelve feet thick. Openings in the wall appear at 
intervals for ingress and egress, but during a siege, these 
were built up with loose stones, when it was considered 
a pari haahuea, an impregnable fortress, or, as the term 
indicates, place of refuge and life. Such as fled to 
the rocks or mountains were called meho. 
If those who had fled were numerous, and the conquer¬ 
ing army wished to reduce them, the war often assumed 
a protracted form. When the assailants had determined 
on reducing them, they endeavoured to decoy them out; 
if they failed, they seldom succeeded in scaling or forcing 
their ramparts. Famine often reduced the besieged to 
the greatest distress, so that they ate the pohue, or wild 
convolvolus stalks, and other rude kinds of food. They 
frequently made desperate sallies upon the besiegers, but 
were often driven back with great slaughter. 
In a sally made during one of the wars which occurred 
in the year 1802, called in the annals of Tahiti, “ the war 
of Rua,^^ this chief, and a number of his fighting-men, 
were taken, and killed on the spot by the king’s order. 
The next day the king marched to the fortress, but found 
it well manned, and the greatest determination to resist 
manifested by the warriors. 
An ambassadress, with a flag of truce, passed between 
the parties, and the besieged manifested an uncommon 
degree of dauntless obstinacy. When told of the num¬ 
bers and the persons slain, they appeared as if but little 
affected by it, pretended not to know them, excepting 
the chief, who, they said, it was far more likely had been 
