NOTES ON FISHING METHODS OF THE SOUTH SEAS. 157 
for a small number of fish. Sometimes four or five hauls were made 
without taking a single specimen. In many places the bottom was 
quite smooth; but owing to the clear water the seine was visible a long 
distance, and this, combined with the natural shyness of the fish, 
caused seining to be abandoned. Satisfactory results were obtained 
only by stretching the seine across the mouth of narrow passages and 
driving the fish into it, a man being stationed at each end to lift it up 
when they crossed the foot line. Only a few at a time could be caught 
in this way, for they would invariably dart under rocks or hide under 
coral. Turning over loose rocks and bowlders was found to be a fairly 
successful way of capture. As this caused the water to become very 
muddy, the fish then would be as likely to dart in one direction as 
another in their efforts to escape, and a considerable number were 
caught. On another occasion when this method was employed 80 
fish were taken, including 6 species not previously taken by us. The 
following day 100 fish were captured, and the next day fully as many 
more. 
Several hauls were made with the seine on the beach near the outlet 
of the passage on the lagoon side, it being quite free from coral and 
other material injurious to a net; nothing was obtained. Previous to 
setting the seine fish were schooling near the shore, but they soon dis- 
appeared and did not return. Running parallel to the beach some 
little distance back, and almost hidden from view by a thick growth of 
vegetation, we found a slough about 30 feet wide and on an average 
3 feet deep, in which fish were plentiful. The bottom and banks were 
composed of fine coral sand mixed with decayed vegetation. In many 
places were branches of cocoanut trees, which made it difficult to haul 
the seine. If an attempt was made to clear away the obstacles the fish 
scattered in every direction. By forming the seine into a kind of trap 
and driving the fish into it we collected a variety of species. Ata 
deep pool near the mouth of the slough fully an hour was spent in 
capturing two beautifully colored fish. 
Our experience showed that with our apparatus only a few fish at a 
time could be taken, the color of the nets, seines, and hand lines having 
considerable to do with the many unsuccessful trials. 
A trammel net was set in 7 fathoms of water close to a coral reef. 
Fish were numerous around it, but nothing was caught. 
In a narrow cut leading from the lagoon to the outside reef, through 
which the tide ebbed and flowed, a wire trap was set in 3 feet of water, 
baited with pieces of fish and a live fish for a decoy. In the cut were 
many holes, varying in depth from 5 to 6 feet and 10 to 20 feet across; 
some were much wider at the bottom than at the top, thus form- 
ing overhanging shelves under which fish lurked. Standing a short 
distance away, in a position commanding a full view of the openings, 
a sight was presented very much like a row of aquariums. A step or 
