Every Fijian is a fisherman by instinct. At ten years old, with a little four-pronged spear, or with a bow and a four-pronged arrow, he is scouring the pools left on the reef by the receding tide, and by the age of eighteen his aim is unerring. He fishes for the pot, not for sport, and seldom does he come home empty-handed. The spectacle of a big fish swimming in the sea never fails to stir his emotion. A sanka darting across the bows of your boat will touch the most lethargic of your crew to tense excitement; no spear being at hand he will poise and cast your precious boat-hook at the monster, and fling himself into the sea to recover it. Even among the tribes of hereditary professional fishermen this emotion is never staled by use.

Wherever the sea runs up into sandy or muddy inlets there stands a fish-fence belonging to some village in the neighbourhood. The fence is from 100 to 200 yards long, built of reedwork supported by stout stakes driven deep into the mud, and shaped like the segment of a circle with its axis on the shore, and about the middle there is a bag-shaped annex with an intricate entrance so contrived that a fish making for the sea as the tide recedes will nose his way through it into the annex and not be able to make his way out again. There is a scene of wild excitement and confusion when the spearmen enter the annex at low-tide. Mad with terror, the great fish lash the water into foam as they dart hither and thither and leap clear of the water to escape the spear-thrusts.

These fences do not survive tempestuous weather. The waves soon make a breach in them, and the smallest hole renders them useless. When they are rebuilt it is generally at a different place, and ruined fish-fences may be seen at every inlet along the coast. But this is for another reason; after some months of use the fish appear to know their danger and to avoid the fence. Perhaps their range is very much restricted, and when the fence has caught all the fish in its immediate neighbourhood the sea at that point is depopulated for the time. At Nasova a superior fence was built of wire-netting. Its daily catch for the first few weeks was enormous—on some tides not less than 1500 fish of five pounds' weight and over—but a few weeks later the catch failed quite suddenly, and thereafter the trap was scarcely worth examining.

In the larger rivers the natives build stone fish weirs constructed to lead into a basket trap. A rope bristling with fibre streamers is dragged by men on both banks to frighten the fish down-stream, and the basket is filled.

But these are mere amateur expedients compared with the methods of the fisher tribes. These, as will be explained in another chapter, own no planting lands, but barter their fish for vegetables, or live upon the bounty of the great chiefs for whom they work. Their skill as seamen was unsurpassed, and in the great confederations they manned the big war-canoes.

In Fiji the royal fish is the turtle. Every considerable chief had turtle fishers attached to his establishment. He would allow them to take service with other chiefs for ten expeditions. The hiring chief paid them by results; for blank days they received nothing, but food and property were given to them for every catch, and a considerable present was made to them at the end of their engagement. The turtle men use nets of sinnet from 60 to 200 yards long and 10 feet wide, with meshes 8 inches square. The floats are of light wood 2 feet long and 5 feet apart; the weights pebbles or large shells. A canoe takes the net into deep water, and pays it out in a semicircle with both ends resting on the reef. This intercepts the turtle on his way back from his feeding-grounds in shallow water, and only a perfect knowledge of his habits guides the fishermen to choose the proper time and place. If the turtle takes fright at the net the men drive him forward by striking the water with poles, and stamping of the canoe deck, and the dipping of a float is the signal that he is entangled. The catch is announced by loud blasts on the conch, and the canoes are received with the same noise of triumph as when they brought back bodies for the cannibal ovens. The women meet them with songs and dances, and sometimes they pelt the crew with oranges and are chased from the beach with loud laughter.

The hen turtle is taken when she crawls on shore to lay her eggs, and the nest itself is robbed when eyes are sharp enough to detect the place where she has so cunningly smoothed the sand over it. But in Kandavu the turtle is actually taken in the sea without nets, and this is sport indeed. Two men go out in a light canoe; the one paddles in the stern while the other lies upon his stomach with his head projecting over the bow, and with a heap of pebbles under him. With scarce a ripple from the paddle the canoe is gently propelled to and fro over the bottom where grows the green sea-grass which is the turtle's favourite pasture. The watcher in the bow lifts his hand; the motion is checked; he takes a pebble from the heap beneath him, and drops it gently into the water. Down it goes pat upon the shell of the feeding turtle. Unsuspecting danger, the beast crawls lazily out of range of such accidents and begins to feed again. Steered by hand-signals from the watcher the canoe swings her head over him again, and another stone taps rudely at his shell. It may need a third or even a fourth to convince him that this rain of solid bodies from the upper world is more than accidental, but this unwonted exercise at meal times has bereft him of breath. Air he must have, and he makes slantwise for the surface. Then the sport begins; the watcher snatches off his sulu and plunges down into the depths to meet him. The art lies in seizing him by the edge of the fore-flipper, and in turning him over before he reaches the surface. It is a slippery handhold, but the hand that grasps the limb higher up will be nipped between the flipper and the sharp edge of the shell, and to seize a turtle by the hind-flipper is to be the tin can tied to the puppy's tail. Having seized your flipper by its edge, you must turn the beast over on his back (if he will let you) and propel him to the surface, where your companion will help you to hoist him on board. The turtle spends his few remaining days lying on his back, and throughout Western Fiji he dies the horrible death which is prescribed by custom: an incision is made at the junction of the hind limb with the under shell, and through this the entrails are drawn out. After their removal, and even during the process of dismemberment, he continues to live. I have often reasoned with the natives against this cruelty, and they have listened to me with amused surprise; "It was the way of our fathers," they said; "if we cut off his head he would not die any sooner, and the meat would be spoiled." When a great feast is in preparation turtle-fishing begins several weeks in advance, and the beasts are kept alive in a stone or wickerwork enclosure in shallow water, which is called a mbi. They can thus be kept alive for several months. There was a tragic note in the fate of one little turtle captured when he was no bigger than a soup plate, and presented to an European as a pet. The owner had moored him to a stake by a string fastened to his hind-flipper, and for several days and nights he swam bravely but fruitlessly towards the open sea. But when, in pity for this wasted expenditure of energy, his owner built a wickerwork mbi for him, and cut him loose, and he had explored every inch of his cage for an opening, he abandoned the hope that had buoyed his spirits, and died in twenty-four hours—a victim, one may suppose, of a broken heart.

The Fijian nets are so like our own that a newcomer may believe that they have been imported. They are made of hybiscus fibre, and the mesh and knot are identical with those of the European net-maker. Long seines are used occasionally, but a commoner practice is to drag the rau—a rope of twisted vines, bristling with cocoanut fronds, several hundred yards long. The ends are brought together, and the fish are speared and netted in the narrow space enclosed by the rau.

The women do most of their fishing with two-handed nets mounted on sticks six feet long. A line is formed with two women to each net, standing to their waists in the sea. As the fish make for the sea in the ebbing tide they are scooped up and held aloft; the ends are brought together, and a bite in the head from one of the women kills the fish before it is slipped into the basket hanging from her shoulder. The kanathe, a kind of mackerel, and the garfish spring high out of the water in their efforts to escape, and it needs very dexterous manipulation of the net to intercept them; sometimes women receive ugly wounds in the face from these fish.

Eels grow to a great size in the rivers, and in the inland districts the women mark their lairs in holes in the bank, and stupefy them with a vegetable poison extracted from the stalk of a climbing plant, or with tobacco. A sort of sponge made of bark-cloth is saturated with the poison, and is quickly immersed and pushed into the mouth of the hole; the poison distils into the surrounding water, and after a few minutes it is safe to explore the recesses with the naked hand. The narcotic effect of the poison is only temporary; left to itself in clear water the fish would recover in about five minutes.

Strangest of all fishing is that of the mbalolo, which is still an annual festival in the districts where it is taken. The mbalolo is a marine annelid about six inches long and of the thickness of vermicelli. It is found on certain sea reefs in various parts of the Samoan, Tongan and Fijian groups, and probably elsewhere in the Pacific. For ten months in the year it is never seen at all. Somewhere deep in a reef cavern it is growing to maturity, but on the night of the third quarter of the October and the November moons it swarms in myriads to the surface and dies, phoenix-like, in the propagation of its kind. So exact a time-keeper is it that it gave names to two months in the native almanac. October was called the Little Mbalolo, because the swarm in that month was comparatively insignificant; the Great Mbalolo was November, and preparations for the fishing in that month were made several weeks in advance. The fact—and it is a fact—that an annelid should observe lunar time would not be very remarkable in itself, but it seems that the Mbalolo observes solar time as well. As Mr. Whitmee has pointed out, the moon directs its choice of a day, and it follows that the creature cannot maintain regular intervals of either twelve or thirteen lunations without changing the calendar month of its reappearance. For two years it rises after a lapse of twelve lunations, and then it allows thirteen to pass, but since even this arrangement will gradually sunder solar and lunar time it must intercalate one lunation every twenty-eight years in order to keep to its dates. It has now been under the observation of Europeans for more than sixty years, and it has not once disappointed the natives who are on the watch for it. What are the immediate impulses of tide or of season that impel it to rise on its appointed day no one has attempted yet to show.

Consider for a moment how many centuries must have passed before the desultory native mind became impressed with its regularity. Even on the night of the Great Mbalolo it is not a conspicuous object on the sea. Mere chance must have brought the fisherman into a mbalolo shoal; years must have passed before a second chance again revealed its habits; decades before the unmethodical mind of natural man had realized its annual recurrence and had noted the day and the hour.

It is only at certain points in the sea reef fringing outlying islands that there are mbalolo holes. The canoes congregate there before midnight. The behaviour of the fish is the first signal; they are there in hundreds, dashing hither and thither in a criss-cross of phosphorescence. Towards morning they lie, stupid from surfeit, flapping their fins helplessly on the surface, and are speared in great numbers. It is an orgie of rapacity and greed. Salala gorge themselves on mbalolo; sanka devour the salala; rock-cod swallow the sanka; a few sharks fill their bellies with rock-cod; and man, as usual, preys upon all alike.

As the night advances the surface of the sea is oily and viscid with the interlaced bodies of millions of mbalolo that feel slimy to the touch as one stirs the water. There are breaks in the mass, and natives have assured me that through these they have seen an oscillating stalk, about the thickness of a man's thigh, coiling up from the depths—a fountain of worms spouting from some chasm in the reef. The fishermen scoop up the worms with cocoanut baskets and empty them into the canoe until the hold is full. The masses of worms are boiled, cut into slabs, and sent, like wedding-cake, all over the country, packed in banana leaves. To the European taste these dark-green masses, though unappetizing to look upon, are not unpalatable. They taste like caviare.

Mr. Whitmee, who made a scientific examination of the mbalolo in Samoa, took a glass jar with him to the fishing, and watched the behaviour of the worm in captivity. His catch included both brown and green worms, the brown being the males and the green the females. They varied in length, and as they swam incessantly round the jar with a spiral motion he noticed that the shorter ones of six inches long had two screw turns and the longer at the most three. Fished up by the finger and thumb they broke spontaneously into short lengths at their jointings.

At eight o'clock the mbalolo have disappeared. If they break up earlier the natives believe that there will be a hurricane between January and March. As the sun gains power the mbalolo may be clearly seen in dense patches with individual worms bridging the clear water between. They are now more active than in the night, the closer masses even churning the surface of the water. A little before eight they begin to disintegrate and break up; the sea becomes turbid and milky, and when it clears they are gone. Mr. Whitmee's captives in the glass jar behaved like their fellows in the sea. After swimming more rapidly for a few moments they gave a convulsive wriggle and broke into half-a-dozen pieces each, which wriggled about near the surface, squirting out their contents. The vase looked as if a teaspoonful of milk had been emptied into it, and the little transparent envelopes of the fluid sank empty to the bottom, just as the green worms discharging their cargo of eggs began also to settle down. After a few minutes' immersion in the fertilizing fluid the eggs themselves sank gently to the bottom, where they lay among the husks that had given them birth and being. Under the magnifying glass a faint whitish spot was detected on each of the tiny green eggs. Thus by a voluntary act of self-immolation the worms had handed on their lives to a new generation.