296 
POLYNESIAN RESEARCHES. 
A pair of ordinarily sized canoes is usually selected, with 
a kind of basket-work fixed between them, to contain the 
fish. To the forepart of the canoes a long curved pole 
is fastened, branching in opposite directions at the outer 
end; the foot of this is fixed in a kind of socket, between 
the two canoes. From each of the projecting branches, 
lines with pearl-shell hooks are attached, so adjusted as 
to be kept near the surface of the water. To that part 
of the pole which is divided into two branches, strong 
ropes are attached; these extend to the stern of the 
canoe, where they are held by persons watching the 
seizure of the hook. The tira, or mast, projects a con¬ 
siderable distance beyond the stem of the canoe, and 
bunches of feathers are fastened to its extremities. This 
is done to resemble the aquatic birds which follow the 
course of the small fish, and often pounce down and 
divide the prey which the large ones pursue. As it is 
supposed that the bonitos follow the course of the birds, 
as much as that of the fishes, when the fishermen per¬ 
ceive the birds, they proceed to the place, and usually 
find the fish. The undulation of the waves occasions the 
canoe to rise and sink as they proceed, and this produces 
a corresponding motion in the hook suspended from the 
mast; and so complete is the deception, that if the fish 
once perceives the pearl-shell hook, it seldom fails to dart 
after it; and if it misses the first time, is almost sure to 
be caught the second. As soon as the fish is fast, the 
men in the canoe, by drawing the cord, hoist up the 
tira, and drag in the fish, suspended as it were from a 
kind of crane. When the fish is removed, the crane is 
lowered; and as it projects over the stem of the canoe, 
the rowers hasten after the shoal with all possible 
celerity. 
