- 381

The island people of the Massim area of Papua 1 have long been known for their inter-island voyaging tradition. However, although Malinowski 2 has analysed the kula trading cycle, and although he and other writers have described the construction of Massim canoes, 3 little detailed information has been published on the sailing and navigation techniques used by the Massim sailors—apparently because the ethnographers who worked in the area seldom ventured into the open sea in canoes. 4 This paper attempts to fill this gap in our knowledge by presenting some data I gathered on Massim canoes and inter-island voyaging while doing ethno-archaeological research in the area during 1967-1970. 5 The information presented is based primarily on personal sailing experiences; during my research in the Amphlett Group I had the opportunity and pleasure of accompanying Amphlett men on numerous inter-island voyages in their canoes. Amphlett men still make regular voyages to trade pottery 6 which is made by Amphlett women 7 for garden produce, and to make kula exchanges. 8

In the four inhabited islands of the Amphlett Group (Gumawana, Nabwageta, Kwatoita and Wamea) two types of canoes are built. the kewou is used for fishing and for transport to and from the gardens, some of which are located on the uninhabited islands of the group. It consists

- 382

FIGURE 1

South-eastern Papua

- 383

of a dugout log which is connected by five booms to a float. The kewou may vary in overall length between 1.8 and 3 metres (5 to 9 feet). The larger sailing canoe, called oga on Kwatoita Island and aidedeya in the three other Amphlett Islands varies in overall length between 4.5 and 8 metres (13.5 to 24 feet). Although canoes of this type measuring up to 17 metres (50 feet) in length have been recorded in the Massim, 9 the Amphlettans prefer their shorter versions because of their greater stability in rough seas and easier handling qualities.

The aidedeya is a single outrigger canoe. The round dugout log is heightened by two or three klinker-built strakes. The hull (kawakawa) is indirectly connected by twelve booms (giu) to the float (yamana). Slim wooden rods are lashed across the booms at right angles forming the platform (patapatala).

The slanting mast (taoyaina) which supports a triangular sail (naya) is stepped forward and rests with its forked end on the third boom from the bow. It is supported by a mast-shore (alivawina). If the mast-shore has a forked end it rests on the third boom from the bow near the middle of the platform; if it has a rounded end it rests in a hollowed piece of wood (velagina) which is lashed to the platform in its appropriate place. The mast is secured by two stays; both are fastened to the mast-head. One (vatowa) is tied to the bow end of the float while the other (koya) is usually tied to the weather-side end of the seventh boom from the bow. Further stabilisation of the mast is achieved by tightening a stay (veitawana) from the head of the mast-shore to the inside boom on which the mast rests. By altering the tensions of these stays the mast's position may be shifted.

The triangular sail (naya) is made of pandanus leaves and has an area of some 7.5 square metres. Its upper margin is tied to the yard (kanabalaina) at intervals of some 60 cm with short string-bark ropes (latuma). Its lower, shorter, margin is fastened to the boom (boliyaya) in the same manner. One end of the halyard (lisi) is tied to the yard near its front end. The loose end of the halyard is thrown through the forked end of the mast-head through which the sail may now be hoisted. Thereafter the yard's position is secured with a vang (aidau) which connects the head of the yard with the weather-side end of the first boom. By tightening or loosening this vang the yard may be raised or lowered. Once the sail has been adjusted to its required position the yard is held in position with two more vangs (tapotapo and lipi). They are fastened to the weather-side end of the seventh and ninth boom from the bow respectively. The front end of the sail's boom is tied to the mast with short string-bark ropes (alisala). With the help of a sheet from the aft end of the boom the sail is adjusted to take full advantage of the prevailing wind.

The canoe is steered with the help of a large steering oar (kunuo) which is thrust through the platform near the stern so that its blade is submerged between hull and float. Raising the blade allows the canoe to turn into the direction of the wind and by lowering it the canoe is kept - 384 on a steady course. A smaller steering oar is used from the lee side of the hull near the stern and is held in a fulcrum made of tree-bark.

Several paddles, a rowing oar (vaya), a wooden bailer and several 3 to 4 metre long wooden rods of mangrove wood which are used for poling complete the equipment of the aidedeya.

From an aesthetic point of view the most striking features of the canoe are the elaborately carved and painted cutwater (bogalugalu)—made up of a vertical prow (yawawa)—and a breakwater (yagimo). 10

A crew of four can handle the aidedeya adequately. One man near the bow works the vang of the mast-shore, the stay of the yard and the halyard. Another man midships bails water and adjusts the two other vangs which run from the yard to the seventh and ninth boom from the bow. A third man steers the canoe from the lee side, sitting in the hull near the stern. The captain's position is, in general, near the extreme stern-end of the platform from whence he works the large steering oar and the sheet. Although he is in charge of the canoe at all times during a voyage, the other crewmen may frequently change positions with one another. 11

Near one end of the platform some smouldering logs are placed on a sand-covered fireplace over which food may be cooked in a clay pot during long voyages.

All goods are loaded in the central part of the hull (gebobo). To increase the loading capacity of the gebobo when clay pots are to be transported, an additional lee platform is added which extends some 50 cm beyond the lee side strakes and runs over the full length of the gebobo (Plate 1).

Figure 2 shows the increase of the carrying capacity of an aidedeya as it submerges deeper into the water. 12 The calculations are based on a canoe measuring 6.8 metres (20 feet) in length and having three strakes on its lee side and two on the weather side. The height from the top of the gunwale to the exterior bottom of the hull measures 0.9 metres. The float produces a constant uplift of 206 kg but serves only as a balancing device.

Malinowski 13 recorded the payload of one canoe of the aidedeya type. It carried 18 men, which may be equivalent to a payload of 1,080 kg. On another occasion he estimated 14 the payload of an aidedeya which carried clay from Fergusson Island to the Amphletts to be 2,000 kg. Clearly, the latter payload would have required a canoe much larger than the one I have been discussing. My informants were not able to tell me if canoes capable of carrying such a payload were once used in the Amphletts.

- i

PLATE 1

Gumawana Island: A lee platform is constructed to provide additional loading space for the pottery.

- ii

PLATE 2

Hughes Bay: Aidedeyas moored near the beach. Islands of the Amphlett Group are in the background.

- 385

FIGURE 2

Carrying capacity of the aidedeya in relation to submergence of hull

I turn now to the sailing techniques used by the Amphlett men when they travel from island to island. This section is devoted primarily to an account of voyages, summarised in Table 1, made by aidedeya in 1968-1969. I was a passenger on most of the voyages discussed; for the remainder first-hand information was obtained from crew-members.

At present the Amphlettans carry out voyages to several places on Goodenough Island, to the north shore of Fergusson Island, to Dobu Island to the south, to the Bwaio Peninsula of south-east Fergusson Island, and to the Trobriand Islands in the north. None of these voyages is much longer than 40 nautical miles (75 km), a distance which can take up to four days to cover if the winds fail and the men have to pole the canoe. The average sailing speed, as I recorded it on various journeys, is around 2.5 knots.

Prospects for sailing are determined largely by the monsoonal seasons. Wyrtki 15 says that, in this region, a prevailing easterly to south-easterly

- 386

- 387

wind blows from May to November and a prevailing north-west to north-easterly wind from December to April. During the south-east season, strong steady winds often make sailing impossible so that most voyages are made during the northerly wind season. But even then voyages are often delayed because of adverse wind and weather conditions.

The Amphlett sailors distinguish by name five prevailing winds, the directions of which are related to the positions of their islands. These are:

• (1) a south-easterly wind, bolimana, coming from Sanaroa Island
• (2) a southerly wind, youya, from near Mount Koyotabu on Fergusson Island
• (3) a westerly wind, koibwaga, from Goodenough Island
• (4) from the north-west, kotala, from the direction of the Lusancay Group, and
• (5) a more northerly wind, yaveana, from the Trobriand Island.

When we sailed from Gumawana Island 18 to Hughes Bay with a bolimana we set the sail immediately after the canoe had been poled clear of the reef which fronts the hamlet shore line. Hughes Bay was clearly visible at all times during the voyage. When we reached the sheltered bay near Fergusson Island it became necessary to pole the canoe for the remainder of the distance. The return to Gumawana Island required, however, more effort. If we had tried to set sail in Hughes Bay we would not have made our destination in a bolimana; therefore the canoe was poled to a known point east of Hughes Bay, called Maikau, from which Gumawana or Wamea Islands can easily be reached. An aidedeya from Nabwageta or Kwatoita Islands will use the Moita point for its departure. The latter point is not as far east as Maikau but suffices as a safe departure point to reach the Amphlett Islands lying further to the west. From these two points the Amphlett Group is clearly visible again.

The hamlets in Basima are situated at S 9° W from Gumawana Island. When we set out for Basima a bolimana prevailed and the men informed me that we could not reach our destination in a direct approach. The canoe would make the shore of Fergusson Island too far to the west, thus leaving a considerable distance for poling. In order to avoid this we paddled the canoe across to Wamea Island and poled along its shore until we reached the most southerly point. From here the new course to Basima read S 18° W. After a short rest we prepared the canoe for sailing and proceeded into the open sea. When we had secured the yard and vangs everyone settled down to enjoy the trip. We baited our fishing lines and prepared them for trawling. Occasionally, minor adjustments to the sail were necessary. We passed Gudulai Island close to its eastern shore and saw some Wamea people working their gardens on this island. As we approached Fergusson Island, the bolimana changed almost to a youya. Our progress subsided accordingly since the sail had to be kept close to the wind. Consequently we made our landfall some 1.3 km. short of our destination and had to pole the remainder of the distance.

- 388

The return voyage is effortless in a bolimana as the canoe can be sailed before the wind. Our slow progress of 1.9 knots is explained by the fact that we were becalmed for some 2.5 hours halfway between Basima and Gumawana Islands. A light swell made it necessary to furl the sail in order to avoid any damage to it as it was continuously tossed from side to side. In a becalmed sea the men take only very reluctantly to the paddles. A considerable amount of time passed before they decided who was to use the rowing oar, who was to paddle and who was to steer. Since there was no urgency about our progress the captain did not interfere in the discussion. Once everyone had found his place one man decided to have a smoke before starting to paddle while someone else felt like chewing betel-nut. Then, those of us who were paddling became tired and were eager to have a rest. During two hours, only once was a co-ordinated effort made to move the canoe. The men glanced hopefully in the direction from which they expected a wind to rise, and sometimes broke out in loud cries calling the bolimana. 19 During the afternoon a bolimana did rise and we reached Gumawana Island without any further delays.

The voyages to the Bwaio Peninsula and to Dobu Island are the longest ones made by the Amphlett men; I did not have the opportunity to join such voyages, but met, at a later date, two aidedeya from Gumawana in Bwaio on my return from Normanby Island.

Near Cape Labillardiere to the west of the Amphlett Group on Fergusson Island, is the clay quarry, Yayavana, from whence the Amphlett men import the clay for the manufacture of pottery. With a crew of four men we set sail as soon as we cleared the reef and followed a course of W 24° S. A bolimana helped us to make good progress. Our first landfall was on Tuyatana Island where the Nabwageta people cultivated gardens. Two men climbed several coconut trees to cut off some green nuts for drinking and coconut leaves to make baskets in which the clay was to be transported. Then, as the wind had increased, we eagerly hoisted the sail. After an hour of sailing, during which Cape Labillardiere was visible at all times, the wind suddenly died. Nobody was eager to paddle. The sail was furled and some men started to weave the baskets. Others smoked, talked, chewed betel-nuts and occasionally uttered cries to make the wind come. Eventually a breeze rose which developed into a strong yaveana. As will be explained later, an aidedeya must be sailed with its outrigger side facing windwards. Therefore, since we did not want to miss our landfall, we had to take down the mast, turn the canoe around 180° and erect the mast again in the “new bow” to take full advantage of the yaveana. Gradually the sea became more and more choppy and everyone was now fully occupied with attending to his duties. We beached the canoe in the late afternoon at Namanamaletta, a trade store near the Cape.

The following day was spent quarrying clay. The day after during the early morning we loaded some 400 kg of clay in the gebobo of the - 389 aidedeya. By noon the sea was still calm and it was decided to delay our departure for one day. The next day again brought no wind, and the captain gave orders to start the return voyage by poling east along the shore hoping that a koibwaga would soon come.

Wherever possible the men prefer poling to paddling. In order to avoid paddling they willingly follow the perimeter of large bays which could be crossed more quickly by paddling across at narrow points. Two men poled the canoe. One stood in the bow of the hull and the other near the stern on the inside stringers. The man poling near the stern kept the aidedeya on course by pushing with the pole either on the lee side or on the weather side. The man near the bow was constantly on the alert, looking for shallow water and pushing the bow around coral patches. At times he shouted his manoeuvres to the man in the stern for supporting action. Occasionally the canoe came off course altogether either because the sea-bed was suddenly beyond the reach of the pole or because one of the men temporarily paid little or no attention to the other's actions. On such occasions the hull often scraped submerged coral and twice we ran aground on the sandy bottom. The men took turns poling and in five hours we covered a distance of some 16 km.

We spent the night on the beach and made Hughes Bay (Plate 2) during the following day, poling all the way again for lack of wind. After camping on the beach once more we awoke in the morning to find a fresh bolimana had come. We poled to Maikau point and set sail for Gumawana Island.

Two days are usually required to reach Goodenough Island. During the first day we followed the course to Yayavana where we spent the night. A fresh bolimana carried us across Moresby Straits on the second day. We passed the Barrier Reef Islands in Moresby Straits to the north and beached the canoe near the patrol post of Bolu Bolu. I did not make the return voyage but was informed that at times it is rather lengthy in adverse wind conditions and follows from Fergusson Island along the course already described.

The voyage to the Trobriand Islands is only slightly shorter than the long voyages to the Bwaio Peninsula and Dobu Island. Nowadays the Trobriand Islands are only visited from the Amphletts for the purpose of the kula. By the time I expressed my wish to join such an expedition the main kula activities were completed. A month earlier several aidedeya from Kuyava Island had concluded the kula transactions with a visit to Nabwageta Island. Nevertheless three men were keen to take me as it gave them a chance to see friends again and to talk over future kula exchanges. In this respect the voyage turned out to be a success: they obtained seven mwali (armshells) for which they promised return gifts in the form of bagi (necklaces). 20

Katubai, the Amphlett Police Constable and most important kula-man at present in the islands, insisted on having his aidedeya overhauled before the voyage. His 6 metre (18 foot) canoe had all brittle lashings (oiga) replaced, fresh caulking (bwoada) added to leaking places between - 390 the strakes, and the sail mended in several places. By 5 April the work was completed and we started to talk about the voyage and the dangers of the sea 21 while waiting for a good wind to rise. On 7 April a strong kotala was blowing. Katubai, as owner and captain, decided to leave on this day. Lupweni and Dibilai brought their belongings and we loaded the aidedeya. Our food supply consisted of several baskets of yams, two bunches of cooking bananas, green and ripe coconuts and several sticks of sugarcane. We took along two clay pots for our own use and a large quantity of betel-nuts.

When we got under way it was mid-morning. We poled the canoe along the western shore line of Gumawana to Yagianeta, a large rock in the sea which marks the most north-westerly point of the island. By now the kotala was blowing only moderately and when we had erected the mast and hoisted the sail Katubai followed a course of N 2° W into the open sea. There was no land in sight. The sail had to be kept close to the wind, which gave us only slow progress. To the south the rugged outlines of the Amphlett Group faded away. Mount Koyotabu on Fergusson Island was also visible. After some 30 minutes Lupweni pointed to a faint shadow on the horizon; the island of Yakum had been sighted. In the distance we had covered so far without sight of land ahead we had drifted a little to the west. Katubai held the sail even closer to the wind giving Dibilai instructions to follow directly towards Yakum. The required course was apparently beyond the windward sailing capacity of the aidedeya. I calculated that we were pointing 60° into the wind and our progress quickly became almost negligible. Katubai then instructed Dibilai to fall off slightly to windward and steer for a point east of Yakum Island. Repeated efforts were made to steer directly for the island but eventually we ended up some 600 metres east. We let the sail down and without any arguments started to paddle west towards Yakum. We made only slow progress despite great efforts paddling almost directly into the oncoming wind and waves. With one man bailing water constantly, only three could paddle and one of those had to steer the canoe at the same time. As we approached the reef around Yakum we had difficulty finding a good passage through the coral. The aidedeya was then poled to the lee side of the island and moored in the water near the beach.

Yakum Island is of an oval shape measuring only some 60 metres across and consists solely of raised coral. Its highest point is only some 1.5 to 2 metres above the high water mark. Coconut trees and some scrub make up the total vegetation. An enormous population of mice are the only land animals. These apparently reached the island accidentally in the sleeping mats of the sailors. The western and northern shores are covered with bright coral sand, whereas large slabs of dead coral extend on to the island out of the sea at its eastern and northern shore. The surrounding reefs abound in fish. During the late afternoon thousands of seagulls come ashore to spend the night on land. They are without - 391 fear of men and I was able to approach within half a metre of them before they flew off. In the soft coral sand of the beach above the high water level Dibilai discovered two places where sea turtles had buried their eggs. These were collected and formed a substantial part of our evening meal.

I have described Yakum Island in some detail since, because of its position, it is used as a landing place by every canoe from the north and south. Occasionally crews shelter there from violent storms lasting up to two weeks. Both Malinowski 22 and Fortune 23 report that the canoes travelling on this route supposedly hug the lee of a reef which extends from Vakuta Island to a point some 12 nautical miles east of Wamea Island. The route from Gumawana to Vakuta, however, lies always to the west of the reef. Therefore the canoes are exposed to the full winds in the north-west season and only travel on the lee side in south-easterly winds. Yakum Island, a dominant landmark of the route, actually lies some 15 nautical miles to the west of the large reef. I found that the latter affords no shelter whatever from the easterly winds, and the men sailing between the Trobriands and the Amphletts stay well away from it. Katubai and Lupweni knew several legends about the dangers of this reef to canoes and their crew.

We spent the evening sitting on the beach. Later that night two men went to sleep on the canoe to guard against its breaking loose from the mooring.

The following day, 8 April, after a short but heavy rainstorm, we poled the canoe over the reef into deep water before the mast was erected and the sail hoisted. Katubai followed a course of N 30° E. A moderate bolimana helped us to make good progress. There was no land in sight ahead, and Yakum Island and the Amphlett Group were used for orientation. By 1 p.m. wind and waves had increased to such an extent that Katubai found it necessary to lower the yard so that the end of it came to rest on the platform of the canoe. Our progress was greatly reduced but the area of sail exposed to the wind was sufficient to keep the canoe on course. A strong rain followed, and, because a sail of pandanus leaf becomes soft when wet and will then tear easily, we were forced to lower the sail briefly at 1.45 p.m. We then sighted Muo Island at N 30° E. Some minutes later the coconut trees on Vakuta Island appeared at N 70° E. Katubai altered the course to the south-eastern end of Vakuta as soon as it became visible. Just then we hooked a tuna on one of our lines. Katubai prepared the fireplace and after 30 minutes the fish was boiled and eaten. After that, we reached the hamlet of Wokinai without any further delays.

By 14 April we were ready to leave Vakuta Island. Our next destination was Sinaketa village on the west coast of Kiriwina Island. A bolimana enabled us to set sail immediately and we made our way in a northwesterly direction following the coastline some 200 metres out at sea.

Near the hamlet of Gilibo, at the sea passage between Vakuta and - 392 Kiriwina Islands, the wind suddenly shifted and we now faced the youya. It soon became evident that we could not hold the canoe in the youya for long enough or far enough from the shore to make Kabinakau Point. After less than a mile of sailing we were right along the shore and had to take the sail down. From here the men poled to the point opposite Muo Island from whence we could sail the rest of the way to Sinaketa which now lay in a north-easterly direction.

From now on the weather deteriorated more and more. The sea became very choppy under a strong bolimana which prevailed for the next four days. Although the wind was still reasonably strong, by 19 April we departed to cross the shallow lagoon from Sinaketa to the village of Kavataria. The course lay almost due north and the distant coast was visible at all times. We covered some 8.5 nautical miles in two hours, equivalent to an average speed of 4.3 knots. About half way to Kavataria a slight mishap occurred. None of the men was very familiar with the seas we were now sailing and the canoe was grounded on a sandbank in the lagoon. The sail was let down quickly so that it would not be torn in the strong wind, after which we all jumped into the water to push the aidedeya free again.

We left Kavataria on 26 April. The weather was reasonably fair and the bolimana was still blowing strongly. After a short halt at the hamlets of Bulakwa and Kaisiga on Kaileuna Island we proceeded to Kuyava Island in a south-westerly direction. By now we had left the reasonably quiet waters of the lagoon, and waves of 1 to 1.5 metres in height kept everyone on the alert. Water from the sea splashed freely into the hull and Lupweni bailed constantly. Katubai did not lower the sail and as conditions did not get worse the men enjoyed the wind, the waves and the speed of the canoe.

In Kuyava we hauled the aidedeya on to the beach. A number of logs had to be laid out in front of the canoe, and a dozen or so men helped us drag the canoe over the logs to a point above the high water mark.

In the following week we experienced extremely strong south-easterly winds which were accompanied by heavy rainstorms.

Two days before our arrival a large double-hulled canoe from Mailu Island, near the southern coast of Eastern Papua, had reached Kuyava. The captain of this canoe was also frustrated by the unfavourable wind as it delayed his departure to the south. From Kuyava to Mailu Island they would need between 8 and 11 days in good winds. By the beginning of May the kotala was blowing again. The Mailu captain took this opportunity and left the same morning. Katubai, on the other hand, was not certain whether the kotala would last through the day and suggested that we wait until the morning. After a rainy and stormy night we found the kotala still blowing strongly.

On 2 May we left Kuyava Island. Low banks of grey rain-clouds often obscured the Amphlett Islands to the south. We could also see rain in the distance over the open sea. We had entered unsheltered waters and the waves were running about 1.5 metres high, making bailing essential right from the start. Katubai and the men were not as cheerful as on - 393 previous occasions. In fact, Katubai soon started to recite magic spells designed to hold the rain off and to soothe the kotala. In these severe circumstances we steered for Yakum Island. Dibilai and Lupweni had to check the vangs, the stays and the halyard more often than usual. Katubai interrupted his spells only occasionally to give orders. The spray and the occasional brief shower kept everyone and everything wet. Most of the incoming water splashed into the hull over the breakwater. For a while the wind subsided a little. Because progress was good, Katubai decided to steer for Nabwageta Island in the Amphletts thus passing Yakum to the west. As everybody was eager to get home no objections were raised. Yakum Island lay about 1 mile to the east when the kotala increased again quite suddenly. During the short time which was required to lower the sail the aidedeya was temporarily blown off course and a succession of breaking waves almost filled the hull. It was impossible to move about on the canoe freely and we all held on desperately to secure parts of the aidedeya. During this time the wooden bailer was lost, a plastic bucket was blown overboard and several paddles disappeared in the sea. Dibilai quickly grabbed the metal dish in which our cooked food was stored and emptied it overboard. With it he started bailing at an enormous speed and for some time he fought an even battle against the incoming water. Katubai and Lupweni were fully occupied in steering the canoe, and I was not the only one on the canoe with a worried expression on his face. Without the skill and experience of these men we could easily have met with a more serious accident, particularly as Yakum Island was out of reach when the storm broke loose.

When the kotala subsided a little and everything was under control again, great relief was expressed by everybody. We reached the Amphlett Group that same evening at 7 p.m.

The mast, the sail, the yard and the boom of the aidedeya are all secured with movable running rigging. This greatly improves the performance of the aidedeya as many adjustments are possible to take advantage of strong winds without having to lower the sail. The standing rigging on a European launch, for example, where the stays cannot be adjusted beyond certain limits may quickly reach their maximum stress; and in order to reduce this stress it is necessary to reduce the area of sail exposed to the wind.

The outrigger with its float also adds considerably to the sailing performance, as it is always turned windward. As Gladwin 24 points out in reference to Caroline Island canoes, this “is done so that the force of the wind on the sail will tend to lift the outrigger float out of the water, not push it down”. Thus the float acts as a stabilising agent by its weight rather than its buoyancy. When the float becomes submerged in a wave its increased drag swings the canoe slightly around into the wind, thereby relieving some of the wind pressure on the sail. The canoe slows down temporarily and allows the float to rise again.

- 394

The aidedeya gives its greatest speed when run fully before the wind. However, when close-hauled the speed is reduced greatly. According to my rough calculations the aidedeya was able to sail to within 60° to 70° of the wind. However, since the aidedeya drifts considerably to leeward when sailing close to the wind, another 10 to 15 degrees should be added to reach a more realistic estimate of the aidedeya's windward sailing ability. 25

Having the outrigger side facing windward, does, on the other hand, severely limit the manoeuvre of tacking upwind. When sailing upwind, the canoe would first have to sail at an angle to the wind in one direction, then be turned around to point in the opposite direction with a similar angle but with the wind now on the other side. Tacking in the manner of a European sailing vessel would change the outrigger side from the weather side to the lee side thereby endangering the safety of the canoe. Therefore, as described in the voyages above, the only solution is to physically change the canoe's direction end for end. This, in turn, means that all sailing gear has to be dismantled and carried to the other end of the canoe, where it has to be erected again. In the Amphletts tacking is disliked by everyone and is only carried out in cases of necessity. I have never heard of anyone making voyages to destinations which could only be reached by tacking upwind.

The Amphlett Islanders do not appear to have developed sophisticated techniques for orientation and navigation. For example, although Amphlett men commonly know many stars by name they do not attempt to use their knowledge of the stars to guide them when sailing at night. 26 The relative lack of sophistication in the navigation techniques of the Amphlett Islands, as well as those of their neighbours in the northern Massim, can probably be explained by the character of the voyages made in the area. The voyages are all short, like those described above. Land, except during bad weather, is always visible, either ahead or astern—the D'Entrecasteaux Islands stand out from as far away as Vakuta and Kuyava Islands in the Trobriands. 27 And the island targets are all large so that a landfall can hardly be missed.

The Amphlett sailors could not recall that anyone from their islands had ever met with a serious accident while voyaging in the aidedeya. 28 However, they did tell me that some fifteen years ago three men from Fergusson Island set out in a sailing canoe to cross Moresby Straits to Goodenough Island. A storm wrecked their canoe and two of the men drowned while the third was able to reach Goodenough Island by - 395 swimming. Benjamin, a Methodist mission sister, witnessed two cyclones on Dobu Island. The first one broke loose in 1898 and was followed by a second in 1900. 29 Although no lives were lost on Dobu Island, 15 people died on Normanby Island. During the storm the three mission launches were broken from their moorings and swept out to sea with local crews on board. During the following day a search party found the crews of the launches in the water clinging to the oars of the sunken boats.

In mid-November 1967 a cyclone, code-named “Annie”, swept through the Lousiade Archipelago. 30 A trading vessel, “Bev”, owned by one of the European trading companies in New Guinea, was at the time on its way from south-eastern Papua to Misima Island. The gale-force winds forced the captain to drop anchor on the lee side of a small uninhabited island. Shortly thereafter, the vessel broke the anchor cable and started to drift to the north. Eventually it capsized and one Papuan and one European lost their lives. At the same time 12 men from Misima Island were out at sea in two canoes when the storm hit them. One of their canoes was wrecked but they all returned safely to land in the other one. A launch rescued 14 Wari Islanders who had found refuge on a reef after their two launches sank during this storm. They had lost one man in the raging seas. In general, I found that the Amphlett men were careful in their manoeuvres and always showed good judgment as far as the wind conditions were concerned. Although they enjoy the speed of their craft, they never overestimate its limits. They say, however, that the younger men are more daring sailors, and that from time to time they have to pay the penalty of tipping the canoe over. This is, incidentally, a matter of great amusement to the older, more experienced men.

Accidental drift voyages, brought about by adverse wind conditions or malfunctioning of equipment, seem to be more frequent. Although I could not obtain any such information in the Amphletts, Malinowski 31 records several instances: a crew from Woodlark Island drifted all the way to Kiriwina Island. From Vakuta Island several canoes were stranded either in the Amphlett Group or further south on Fergusson Island, where before pacification the whole crew of one canoe except for one man fell easy prey to the cannibals of these shores. Captain Hilder 32 reports two drift voyages, lasting up to three weeks, by a group of canoes from Woodlark Island to Simbo in the Solomon Group, some 200 miles to the ENE.

The currents of the Southern Solomon Sea are strong and correspond closely to the prevailing seasonal winds (Table 2). However, the Amphlett sailors never mentioned to me the surface currents of the sea and seemed to be ignorant of their existence. This apparent ignorance may be explained by the long reef which extends from Vakuta Island south to a - 396 point east of the Amphlett Group. This reef seems to direct the westerly currents towards the north-west, thus leaving the waters immediately east of the reef free of strong currents.

The Amphlett men are by no means the only sailors in the north-western Massim at the present. During my six months in the Amphlett Group and seven months on Goodenough Island, visitors from various parts stopped at one or the other of the islands.

From Wagifa Island, off the south-eastern tip of Goodenough Island, five canoes arrived in Nabwageta Island in April 1968. They had travelled one way some 35 nautical miles (63 km) in 4 metre (12 feet) single outrigger canoes using canvas sails with European type rigging. Occasionally some men from Hughes Bay sail across to Gumawana Island bringing garden produce which they exchange for clay pots. During the middle of April 1968 three double-hulled canoes from Mailu Island anchored overnight near the village on Gumawana Island on their yearly voyage to the Trobriands, some 225 nautical miles (405 km) distant from their home. One of these canoes was the one I met again on Kuyava Island in 1969.

The Sim Sim Islanders from the Lusancay Group, west of the Trobriands, annually visit the western coast of Goodenough Island. Their canoes are identical to the aidedeya. In 1968 a fleet of eight canoes made the voyage. They covered a distance (one way) of 46 nautical miles (83 km.).

The people from Baniara near Cape Vogel on the mainland of New Guinea, sail at times to south-western Goodenough, where they trade pottery for garden produce. One sea-going canoe was built and used some years ago by the Yauyaula people on west Goodenough for the transportation of copra.

In recent years sea-going canoes have come to be replaced in some areas of the Massim by European-designed cutters and motor launches. Technical education programmes have been inaugurated and the Massim people have begun to earn money which enables them to purchase modern boat-building materials and marine engines. As early as the late 1940s the southern Massim people, aided by technical training provided by the Kwato Extension Association, 34 had begun to build - 397 their own cutters. Belshaw, 35 for example, records that the Wari Islanders had seven cutters in 1955. Several local government councils and a few enterprising individuals or families in the D'Entrecasteaux Group have recently purchased motor launches which are used primarily to pick up copra from producers and transport it to buying agents in the main ports. As the Amphlett men are considering buying a motor launch, to be used in pottery trading, it seems likely that the spread of modern means of transport may soon begin to alter the inter-island voyaging techniques described in this paper.

This research was financed by the Australian National University. I wish to thank Professor Jack Golson and Dr Ben Finney for their valuable help and suggestions. Fathers G. Doody and Toomey and their staff from the Catholic mission on Kiriwina Island assisted me greatly with their generosity and hospitality after the strenuous days on the canoe.

In particular I wish to extend my appreciation to Katubai, Lupweni, and Dibilai, and to all people in the Trobriand Islands who became involved in my work. Without their patience and assistance this research could not have been accomplished.

• ARMSTRONG, Wallace E., 1928. Rossel Island. Cambridge, University Press.

• BECHTOL, Charles, 1962. “Sailing Characteristics of Oceanic Canoes”, in Polynesian Navigation. A Symposium on Andrew Sharp's Theory of Accidental Voyages (ed. By Jack Golson). Supplement to the Journal of the Polynesian Society, Memoir No. 34, pp. 88-101. Wellington, Polynesian Society.
• BELSHAW, Cyril S., 1955. In Search of Wealth. American Anthropologist, 57, No. 1, Part 2, Memoir 80.

• DAVIES, H. L., and D. J. IVES, 1965. The Geology of Fergusson and Goodenough Islands. Department of National Development Bureau of Mineral Resources, Geology and Geophysics. Report No. 82. Canberra. The Bureau of Mineral Resources, Geology and Geophysics.

• FORTUNE, Reo, 1932. Sorcerers of Dobu. London, Dutton.

• GILMOUR, Reverand M. K., 1904. “A Few Notes on the Kiriwina (Trobriand Group) Trading Expeditions”. Annual Report on British New Guinea.
• GLADWIN, Thomas, 1970. East is a Big Bird. Cambridge, Massachusetts. Honolulu University Press.

• HADDON, A. C., and J. HORNELL, 1937. Canoes of Oceania. The Canoes of Melanesia, Queensland, and New Guinea, Volume 2. B. P. Bishop Museum Special Publication 28.
• HILDER, B., 1959. “Polynesian Navigational Stones”. Journal of the Institute of Navigation, 12 (1):90-97.

• JULIA, Sister, 1912. Victoriana: Missionary Sister in Papua New Guinea. Geelong, Thacker.

- 398

• Koninklijk Nederlands Meteorologisch Instituut, No. 124, n.d. Zeegebieden rond Australie. Oceanografische en Meteorologische Gegevens. 'S-Gravenhage. Staatsdrukkerij - en Uitgerijbedrijf.

• LAUER, Peter K., n.d. (a). “Preliminary Results of Ethnoarchaeological Investigation in the Northern Massim”. Asian Perspectives, 13 (in press).
• —— n.d. (b). Pottery Traditions in the D'Entrecasteaux Group, Southeastern Papua. (Ph.D. Thesis currently in preparation.)
• —— 1970. “Amphlett Islands' Pottery Trade and the Kula”. Mankind, 7 (3):165-76.

• MALINOWSKI, Bronislaw, 1922. The Argonauts of the Western Pacific. London, Routledge and Kegan Paul.

• SELIGMAN, C. G., 1910. The Melanesians of British New Guinea. Cambridge, University Press.
• SILAS, E., 1926. A Primitive Arcadia: Being the Impressions of an Artist in Papua. Boston, Little, Brown.

• WHITEHOUSE, E., 1922. “Canoes, Trobriand Islands, Territory of Papua”. Anthropological Report, 2:113-118. Port Moresby.
• WYRTKI, K., 1960. “The Surface Circulation in the Coral and Tasman Sea”. Technical Papers, Division of Fisheries and Oceanography. C.S.I.R.O. Australian Papers No. 8, Canberra.