Volume 52 1943 > Volume 52, No. 3 > Outrigger devices: distribution and origin, by James Hornell, p 91-100
                                                                                             Previous | Next   

- 91

THE OUTRIGGER-CANOE in one or other of its two forms, the double and the single, has a distribution coincident with the widespread wanderings of the Island People of the Far East; indeed, its range has never extended beyond the limits set by the voyaging and colonizing efforts of those islanders who have spread from Indonesia east and west throughout the whole of the equatorial belt of the Indo-Pacific region. Nowhere outside of this area is the outrigger-canoe to be found.

The focal point in its distribution, past as well as present, centres in the island world of Indonesia, stretching from Sumatra in the west to New Guinea in the east. Thence it was spread, partly by cultural diffusion but mainly by migration, throughout Oceania until to-day it covers the whole of Micronesia. Polynesia, and Melanesia (the Solomons excepted). In times long gone it was carried even as far as lonely Easter island by intrepid Polynesian voyagers sailing into unknown seas with splendid recklessness. To the west it spread to Ceylon, India, East Africa, and Madagascar. To-day its influence, still active, has reached out to the western coast of South America, where, on the coast of Colombia, a crudely-rigged outrigger-canoe has been introduced, probably by Spanish influence.

In its simplest form the hull of an outrigger-canoe consists of a hollowed-out tree-trunk—a dugout—so narrow and unstable that it is too dangerous to use in the open sea, particularly if lengthy voyages have to be made under sail. To overcome this lack of stability a balancing-device is fitted on one side in the case of those known as “single-outrigger canoes” and on both sides in the “double-outrigger”. In the single-outrigger type this stabilizing apparatus consists of a log of light wood, the float, boomed out from one side of the dugout-hull by means of two or more poles lashed athwart the gunwales. In double-outriggers the boom-poles, with their attached floats, project equally on each side, thereby ensuring a considerable measure of lateral equipoise.

With regard to the relative distribution of the two types, we find that the double-outrigger is limited rather strictly to Indonesia, the western end of New Guinea, and the East African coast. In the Comoro islands it is dominant in the islands facing Africa, whilst in those nearer Madagascar, the double gives way to the single (4, p. 8 and 7, p. 320). In Madagascar transition from the double to the single type has taken place comparatively recently, for the descriptions left by Houtman (8, p. 6) and by Richard Boothby, writing respectively in 1595 and 1644, show that the double form was general in their time, whereas to-day only a few rudely-built small canoes are faithful to the old-time double type—all others are single-outriggers, but generally the transition is not complete for although the second - 92 float has disappeared, the booms on that side have been retained, with their projecting ends connected by a thin pole that is a mere vestige of the original float.

The single-outrigger is characteristic of the whole of Polynesia and Micronesia and also of all Melanesia except the Solomons and the western end of New Guinea. It is the only type met with on the coasts of Ceylon and India and, as we have already seen, it is present, mingled with the double form, in Madagascar and the Comoro islands.

Local variations are infinite, and cannot be considered here in detail. 1 Most generally the number of booms is two in Polynesia, Indonesia, Ceylon, India, East Africa, and Madagascar, whereas in islands occupied by people in a low state of civilization a multiplication of the outrigger-booms is customary: the canoes of the Negrito tribes of the Andaman islands and those of the Oceanic Negroes of western New Guinea are typical examples. An Andamanese outrigger-canoe has never less than three booms, and in the largest the number rises to twelve; in New Guinea I have seen Papuan canoes with as many as eleven.

When the canoe rests on an even keel, the float or floats usually rest lightly on the surface of the water, generally not less than a foot below the top of the gunwale and often at a considerably greater distance below it. As a consequence there is almost limitless diversity in the method of attaching the float to the boom. The simplest and most primitive is when the boom-ends are lashed directly to the float. In these the outer part of each boom should curve downward to allow attachment to the float at the correct height. Properly curved poles are often difficult to obtain, so an intermediate connection has to be provided. This ‘connective’ or ‘intermediary’ when single, may be straight, vertical and stanchion-like as in the Papuan canoes of Manokwari, or curved and arched in various ways or even recurved in lax S-form; sometimes it assumes a U or Y shape; it may be peg-shaped as in some Malagasy canoes. Frequently the connective is multiple, the units arranged in splayed pairs; the pointed lower ends are driven into the float while the upper ends are lashed against the sides of the boom. Two or more pairs are commonly present to constitute one multiple connective. This type is characteristic of Melanesia, but is by no means the only one found there.

The use of outrigger-canoes with hulls formed of simple dugouts is confined to river, lake, and inshore work; these seldom hoist a sail. For voyaging, greater freeboard and carrying-capacity are requisite; these are usually obtained by raising the height of the sides by means of one or several runs of planking (strakes), ‘sewn’ together, edge to edge, with coir twine. If a flare or outward inclination be given to the strakes, almost any width or depth of hull may be obtained. The result, when numerous strakes are built up, is to produce a wide and roomy hull comparable with that of a carvel-built boat of European design. Such large craft are the exception; when they do occur it is probable that their builders had had trading contact with people from a locality where the built-up design of hull was usual. The most noteworthy instance is that of the great two-masted sailing-ships of the Javanese, known to have existed in the eight or ninth century A.D. As seen in the sculptures of Boro Budur in Java, these vessels were equipped with massive outriggers of complex design (11, pp. 227–255). The possession of large sea-going vessels of this description enabled - 93 the later migrations of Indonesians to Madagascar to take the direct sea route across the Indian Ocean instead of the coastwise voyaging of the earlier migrants who went by way of Ceylon, Aden, and the east coast of Africa.

Vessels related in origin and construction to these fine old Javanese ships lingered on in Ceylon and Madagascar into the early years of the present century. The Sinhalese type is the more distinctive as being the larger and better equipped of the two; generally one of these ships carried about fifty tons of cargo. Like the Javanese craft they were two-masted, but in common with the sailing outriggers of Sinhalese fishermen the outrigger-device was single; it was fitted upon the port side as this was always the weather side when the vessels were making use of the alternation of land and sea breezes characteristic of the fair-weather season to which their voyaging up and down the west coast of Ceylon was restricted. The large and weatherly design of this Sinhalese hull is probably a legacy from the days when trade between Sumatra and Java on the one hand and Ceylon and South India on the other was active and frequent, for allusion is made in several old Tamil poems, quoted by Kanakasabhai Pillai in his Tamils of 1800 years ago (9), to commerce with Chavakam, meaning Sumatra and Java.

Small plank-built craft of the same general type, are occasionally to be seen in the western ports of Madagascar. These may justifiably be considered as degenerate survivals of the Indonesian fleets that conveyed the later streams of emigrants to the great African island. One of the very few remaining, which I had an opportunity to examine at Majunga in 1926, had the hull-planking sewn together as in the Sinhalese boats; the single outrigger was, however, on the starboard side, and each of the booms connected with the float in the usual Malagasy fashion by a stout peg which, it is significant to note, is of the same type is characteristic of a kind of small canoe still found on the north coast of Java (3, p. 103).

An exceedingly curious type of outrigger, the korakora of the Moluccas, must also be mentioned, for, as we shall see later, although it appears to be a freak design, actually it bears direct and intimate relationship to the river-craft from which all modern outrigger vessels have developed. This type is long extinct, but various old Dutch writers figure and describe it, and our own Dampier noted the extraordinary design of its outriggers.

It was a vessel owned only by native potentates. Those seen by Dampier (1729, vol. 1, pp. 335–6) were the state-barges of the Sultan of Ternate. They were roomy craft capable of accomodating “fifty or sixty people or more”. A lightly-built thatched cabin divided into three compartments occupied the central part of the vessel, the free ends of the deck being reserved for some of the crew whose duty it was to row or paddle the vessel.

Several curved outrigger-booms projected from each side; upon these at some distance apart two tiers of light beams (bamboos, the prototypes of true outrigger floats) were attached, parallel with the hull. They did not reach low enough to touch the water being, one, three, or four feet above it; “they serve”, says Dampier, “for the bargemen to sit and row and paddle on. . . The lower tier of these beams is not above a foot from the water; so that, upon any the least reeling of the vessel, the beams [floats] are dipt in the water and the men that sit are wet up to their waste, their feet seldom escaping the water. And thus, as all our vessels are rowed from within, these are paddled from without”.

- 94

In Polynesia, the islanders instead of developing the dimensions of the hulls of their outrigger-canoes to cope with the transport of considerable cargo or numerous passengers—migrants or more usually a body of raiding warriors—as was done in Indonesia, devised and developed as an alternative, the double-canoe.

This cumbersome craft is formed by lashing two canoe hulls together at a varying distance apart by means of two or more horizontal beams which represent morphologically the poles used to boom out the outrigger in single-outrigger canoes. The finest examples were built by the Tongans, the Samoans, the Maoris of New Zealand, and the people of the Low Archipelago or Tuamotus. The Fijians, although mainly of Melanesian origin, possessed magnificent double-canoes, but they usually employed Tongan shipwrights to build them. The Tahitians and Hawaiians were also noted for their skill in building double-canoes.

The hulls of large double-canoes consisted generally of a dugout base, often little more than the equivalent of a keel timber; on this, sides were raised of planking sewn together. In Hawaii giant logs of Oregon pine occasionally drifted ashore; these were greatly prized, for they were often so large as to serve as entire hulls without the need of raising the sides by means of planks sewn on; the difficulty was to obtain a pair of approximately equal size; sometimes a log was kept for years before this aim was achieved.

On the cross-beams connecting the twin hulls of double-canoes a platform was laid; upon this a shelter-hut or cabin was built to accomodate the chief and his henchmen. The store of food, usually the only cargo carried, was stowed in the holds of the hulls; sometimes, if migration was the object, the women were allotted the hold of one hull.


In searching for the place of origin of the outrigger-canoe, the most probable location is the centre whence the type radiated to other localities. This we know was Indonesia, for from there it passed eastward at a very early period with the swarms of migrant Indonesians who, voyaging through Micronesian and Melanesia, found new homes far out in the Pacific among the rocky islands and coral-atolls which we now term collectively ‘Polynesia’. At a rather later date outrigger people from Sumatra and Java voyaged westward, leaving a record of their passage in Ceylon and the southern and western shores of India; thence, coasting onward, they reached the shores of East Africa, where, after a temporary halt, they finally established a permanent home in Madagascar, which seems at that period to have been uninhabited, although there are Malagasy legends that suggest the presence of a pygmy race long since exterminated or absorbed.

At a later date, so far as we are able to judge sometime between the eighth and tenth centuries A.D., a second Indonesian immigration into Madagascar took place; this time the craft employed were of the type seen in the Boro Budur sculptures, stout two-masted vessels well fitted to sail direct to their oversea destination. Finding friends of their own race and speech already settled in the island, the newcomers appear to have abandoned seafaring adventures apart from raiding expeditions across the Mozambique channel, made generally with the object of securing slave labour. Occasional trading trips appear to have been made to Aden (7, p. 316), while communication - 95 with the Comoro islands was frequent, and led to settlement of some if not all the islands by Indonesians who at a later date formed a hybrid population through intermixture with Arabic-speaking people.

Although the dispersal centre and the area where the outrigger-canoe developed into a serviceable craft whereby fishing and travel at sea became possible without running undue risk, is almost certainly to be located in Indonesia, this does not imply that it originated there. All other types of primitive sea-craft had their genesis on inland waters, in marshes or on rivers, and it would be strange if the outrigger-canoe were an exception.

If this be correct, Indonesia itself must be ruled out; its islands have few rivers of great volume or extent, and marshland, apart from being scarce, is not the home of any considerable or important fishing tribe. It is to the mainland therefore that we have to look. There, in Farther India—in Burma, Siam, French Indo-China, and the extreme south of China—three mighty rivers, the Irrawaddi, the Salween, and the Mekong, have been from time immemorial the great highways of migration available to the peoples pressing southward from the cold and dreary mountain lands of the north to the fertile lower valleys and rich deltaic plains. Without these, and various minor rivers, traffic in these countries would stagnate; the densely afforested land along their banks made land-transport too difficult, slow, and expensive to be utilized to any appreciable extent before the advent of the railway. Everything went by water, a necessity that led to great activity and enterprise among the boat-building fraternity. All sizes of native-built craft were, and still are, to be seen on these rivers and especially on the Irrawaddi, in spite of the ever-growing competition of railways, river-steamers, and motor-barges. Nowhere else can the development of the dugout-canoe into sailing ships of large size be so clearly traceable, step by step, as on this great Burmese river. The dugout in general use by the river-people, scattered in innumerable villages along its banks, is the most beautiful of its class. Hewn out by men possessed of inherited skill and inborn artistic feeling, the huge teak baulk is fashioned into a long and relatively narrow craft sheering upward in graceful curve toward the two ends which rise well above the water-line like the horns of a gigantic crescent.

When we examine larger craft, even those of the greatest size, we find a dugout-canoe of the form described persisting as the basal under-body of the vessel. Upon its sides and ends, planking is raised to whatever height may be required. Frames are added to which the planks are nailed and a superstructure is raised at the after end to form a lofty poop. Every gradation in size is found until that miracle of Burmese ship-craftsmanship is reached, the square-rigged “rice-boat”. Even here the basal dugout can still be traced, overshadowed though it be by the high planked sides and the lofty poop where the steersman, seated at a commanding height, steers with the aid of an enormously powerful quarter-rudder. The sides of the high poop-platform where the steersman sits as on a throne are decorated with elaborate carving, whereon clever artists in wood have lavished all their skill, decorating them with elegant and intricate floral scrolls that owe their prime motive to some old studies of the lovely creepers and flowering shrubs so frequently seen draping the trees and growing in forest-clearings adjacent to the village houses; these have become conventionalized, like the Acanthus leaf by the Greeks of the classic age, into designs superbly decorative.

- 96

These Burmese vessels possess many other features of outstanding interest, but the only one which concerns us here is the outboard punting or poling gangway that runs along each side of the old type of large Burmese rice-boat as it existed down to the early years of the present century. This gangway or gallery, the counterpart of the outboard rowing platform or “parodos” of ancient and medieval galleys, is, or was, a light passageway between the .two ends of the boat, fitted laterally at gunwale level, there being no continuous deck.

A number of very long and stout bamboos run lengthwise to form a lattice-work floor framework and upon these is laid in turn a light flooring of square-framed gratings, formed of short lengths of split bamboo, arranged in two rows of chequer-board areas. Along the outer margin a light handrail is erected, supported on numerous short stanchions, to provide fulcra for the sweeps which work in grommets fitted around it at certain of the places where stanchions occur.

One projecting end of a strong crossbeam fitted across the hull near the fore end, of another some distance forward of the stern, and of others disposed at intervals between the two ends of the vessel afford support to the bamboos forming the longitudinal members of the floring framing. To secure the ends of these bamboos from slipping a thick baulk of timber, having three large ornamental crenellations cut out of the upper edge, fits down upon each of the crossbeams at each end of the “parodos” in yoke-fashion, locking them in position as in a vice. The two ends of the bamboo forming the handrail of each gangway rest upon the upper edge of these two yoke-baulks, to which they are secured by lashings.

To prevent accidental damage to these comparatively frail gangway-structures, a fender is suspended by stout lashings under the outer edge; sometimes the fender consists of a stout pole, sometimes a bundle of bamboos is employed for the purpose.

Outboard poling-gangways showing structual affinity to those of the Burmese type of rice-boat are also in use by some of the larger craft plying on the greater rivers in the province of Kiangsi in southern China. In these vessels a light and narrow outboard-platform along each side is supported upon the projecting ends of booms laid athwart the hull. Unlike the Burmese gangways these have no handrail.

The function of these gangways varies. In the Burmese rice-boats, where they are relatively broad, they serve both as rowing-galleries like the “parodos” of ancient galleys and, alternatively, as punting-gangways along which the crew tramp to and fro, propelling the vessel laboriously forward with long bamboo poles whenever the depth of water and the rapidity of the current be sufficiently reduced to render this practicable in the season when the wind, being adverse, prevents progress under sail.

In Kwangsi, where sailing conditions are usually more favourable, the function of the platforms is restricted to that of serving as poling-gangways. When sweeps are used, they are worked from the decking at the fore end of the boat.

Working backward through the stages through which these side-platforms have developed, we come finally to the condition where they consisted of nothing more than one or two bamboos lashed upon the projecting ends of a few short poles laid athwart the hull. This stage is to be found upon the upper reaches of the Irrawaddi and Mekong and their major tributaries, where smaller and lighter craft are employed for cargo-transport than upon the main sections of these - 97 rivers. Heine-Geldern (2, pp. 596 and 602) figures one of these on the river Shweli in the northern Shan States, showing two bamboos lashed outboard on each side in this manner, and states that such floats are intended not only to make the canoe more stable, but also to serve as breakwaters in the rapidly rising storms of the Mekong; at the same time they act as air-chambers which keep the craft afloat even when it becomes waterlogged.

The subsidiary function of outboard platforms along each side for stabilizing a vessel and enabling it to retain or to regain its equiblibrium is, in emergencies, of vital importance; should an unexpected squall strike the vessel they act as powerful stabilizing fins. As soon as the craft heels over to the blast to any considerable extent the lee platform strikes the water; in so doing it offers notable resistance to further immersion, acting precisely like a broad and powerful rolling chock; by this action the danger of further heeling is reduced or obviated and a capsize thereby prevented.

An accident of this nature may not happen except rarely during a river trip, but it is an ever present danger at sea, once the friendly shelter of a weather-shore be left behind. Hence the side-platforms of those craft which came to be used by the early Indonesians after they had left the motherland for homes on the islands lying between Asia and Australia, invaluable though they were on inland waters, were found to be too resistant and clumsy for work at sea. Little by little they were reduced in size as experience was gained, till eventually they came to be reduced to a single log, the float (rarely two), boomed out some distance from the hull on opposite sides.

The only exception is the korakora of the Moluccas described on page 93, where the outboard paddling-platforms may be considered as direct modifications of the rowing- and poling-gangways of the old Burmese rice-boat. Dampier, as I have noted, comments upon the great defect of this arrangement whereby “on the least reeling of the vessel” the paddlers on the outer platform were immersed up to the waist in water—evidence that this arrangement would entail disaster if the vessel were caught in bad weather.

Thus it was that the double-outrigger canoe came into existence if the argument just put forward be correct. Before I visited Indo-China and had an opportunity to examine the river craft in use there, I was inclined to consider that the single-outrigger was the earlier type; this view I have abandoned for the reason given above. There is now little doubt that the double-outrigger antedated the single type, and that it originated from the balance-device and the poling-gangways of the river-boats of the Indo-Chinese region. It also explains why the single-outrigger is all but absent from Indonesian waters where, indeed, it occurs solely as a degeneration from the double-outrigger in two exceptional instances (3, pp. 89 and 103).

That the double-outrigger originated from a balancing-device as described, finds confirmation in the story of an incipient form of double-outrigger which I met with in Colombia (S.A) in 1924. At the town of Charcas, a few miles from the mouth of the river Iscuande, I found many small dugout canoes with a balance-pole of light balsa-wood lashed to the outside of the gunwale on each side (see text-fig.).

In the larger craft each balsa-log is lashed in true outrigger fashion to the outboard ends of two stout bamboo outrigger-booms in the usual manner of the double-outrigger having direct attachment, except that here the balsa-logs, which may be considered as incipient floats, are

- 98
A small Colombian canoe with a balance-log attached to each side. Between each end of the median crossbar or ‘boom’ and the balance-log, a lenticular piece of balsa-wood (tojin) is intercalated to depress the balance-log somewhat

each depressed a few inches below the gunwale by means of a small pad or chock of wood placed between the float and each of the boom-ends, the connection securely lashed up by stout bast-fibre straps. The distance between each balsa-log and the side of the canoe is only about eight inches.

In the largest of these Colombian canoes, the sea-going sailing imbabura, engaged in local coasting trade, we have a built-up vessel furnished with three athwartship booms projecting outboard from eight to twelve inches on each side. As in the smaller craft already mentioned, the booms are horizontal, with the balsa-logs lashed to their ends; as a consequence these balancing-logs are elevated above the water by a distance little less than the amount of free-board existing— considerable when the vessel is light, much reduced when it is fully laden (5, article 102).

Admiral Paris (10, pl. 130 and p. 150) figures and describes a Chilian canoe almost identical with the simplest of these three Colombian examples, but none is known there to-day. This old record suggests that the type may have been prevalent along the greater part of the west coast of South America in the early nineteenth century.

In all these examples the balsa-logs do not reach the water unless the canoe heels over greatly; they serve solely as a balancing-device; they cannot be considered as anything but an imperfect and undeveloped form of outrigger. As to the origin of the device, this must be due either to independent invention in South America or it must owe its introduction to some vague recollection of the outrigger-device used in the Philippine islands that had been noted imperfectly by a Spaniard who had seen service in that region. This possibility is strengthened by the fact that in the Philippines the canoes are all fitted with double-outriggers having the float attached directly to the boom-ends; the booms generally have little downward curve outboard so that in many cases the floats do not reach to the water when the vessels are on an even keel; in the large two-masted coasters of Misamis Province they are at gunwale-level, and when beached the vessels have to be supported by a crutch placed under each of the floats to prevent them falling over and damaging the outrigger on that side (3, pi. 5, p. 76 and 5, art. 102).

The Origin of the Single-Outrigger Canoe.

The question now arises—Why has the single-outrigger superseded the double-outrigger in the oceanic island groups of the Pacific, in Ceylon and on the coast of India? From long years of sailing and of weather conditions in these localities I find a feasible explanation to lie in the dangerously-exposed conditions that beset navigation there when compared with the relatively sheltered seas of the Indonesian Archipelago, where islands are so thickly clustered and so near to one

- i
Fig. 1.
— Photo by J. Hornell., Transitional form of outrigger-canoe, Nossi-bé, Madagascar, 1926. On one side the float has degenerated into a slight pole connecting the ends of the two booms.
- ii
Fig. 2.
—Photo by J. Hornell., A further stage in the transition from the double to the single form of outrigger; the booms are still retained projecting outboard on the off side, but all trace of the float on that side has disappeared. Diego Suarez, 1926.
- iii
Fig. 3.
—Photo by J. Hornell., A single-outrigger canoe belonging to Ratnagiri, west coast of India. The booms are connected directly to the float by means of the Spanish windlass device.
- iv
Fig. 4.
Two old-type Burmese rice-boats with broad outboard rowing-platforms.
- 99

another over the whole area that we may consider the Indonesian Sea as possessed of most of the amenities typical of a land-locked bay. Voyaging among its islands, the seaman is seldom out of sight of land for long; friendly havens are generally close at hand, and light canoes are able to make a landing almost anywhere on the nearest island. Under these favourable conditions there is no urge to make any change from the double-outrigger type which, as we have seen, appears to have developed from the balancing equipment of canoes and other craft employed upon the rivers of Indo-China.

Out in the Pacific, conditions are radically different, and the risks incidental to long voyaging immensely increased. Islands are often widely scattered, small generally, easily missed and with coasts either encumbered with surf-beaten reefs or fronted by forbidding cliffs. Weatherly craft are essential to navigation amid the islands. This is not afforded by the double-outrigger. In bad weather it becomes a grave handicap; its lee float plunging deep in the waves when the craft heels over under the impact of a tropic squall proves a dangerous drag. To steer a course is, in consequence, difficult, and the outrigger-frame, exposed to exceptional strain, is liable to break up at any moment, an accident that entails immediate capsize. That experience of these disabilities and consequential dangers led eventually to the discarding of the outrigger-frame and its float on one side, usually the port side, is certain, for we find that wherever long voyages have to be undertaken, or if fishing has to be carried on out of sight of land — apart from the Indonesian Sea — the single-outrigger is employed invariably in preference to the double. Polynesia has already been instanced; we find that the same conditions have led to the same result throughout Micronesia, and the whole of Melanesia except in the extreme western fringe where it marches with Indonesia. Equally, do we find that the single-outrigger has ousted the double one among the fine sea-going fishing-craft of Ceylon and of India that employ the outrigger. In Madagascar and the Comoro islands this change over is even now to be seen in progress. When Houtman (loc. cit) wrote circa 1595 of the conditions then prevailing, it is manifest from his drawing and description of the canoes then in use there, that the double canoe was the type generally employed; the same conditions appear to have prevailed in Richard Boothby's time (c. 1644). To-day the single-outrigger is dominant in Madagascar while in the Comoros it is general in the islands facing Madagascar—the double being more frequent in those turned toward Africa; the latter in common with the double-outriggers of the mainland, are of small size, used only in inshore fishing. That the change over in Madagascar is of comparatively recent date is marked by the presence in very many Malagasy outrigger-canoes of a vestigial float, a rod or slight pole, connecting the boom-ends on the side opposite to that which is fitted with the functional one. In numerous instances even this vestige of a float disappears, although the booms on the same side are left projecting free, without even a trace of any pole, however light, connecting their outboard extremities (4, pp. 6 and 7).

- 100
  • 1. Haddon, A. C. and Hornell, J., The Canoes of Oceania: B. P. Bishop Museum, Special publication No. 27, 3 parts, Honolulu, 1936-1938.
  • 2. Heine-Geldern, R., Urheimat und früheste Wanderungen der Austronesier: Anthropos, vol. 27, pp. 543-619, 1932.
  • 3. Hornell, J., The Outrigger Canoes of Indonesia: Madras Fisheries Bulls vol. 12, pp. 43-114, Madras, 1920.
  • 4. Hornell, J., Les Pirogues à balancier de Madagascar et de l'Afrique orientale; La Géographie, vol. 24, pp. 1-23, Paris, 1920.
  • 5. Hornell, J., South American Balanced Canoes: Man, 1928, article 102.
  • 6. Hornell, J., South American Balsas—The Problem of their Origin: Mariner's Mirror, vol. 17, pp. 347-355, Cambridge, 1931.
  • 7. Hornell, J., Indonesian Influence on East African Culture: Jour. of the Royal Anthropological Inst., vol. 64, pp. 305-332.
  • 8. Houtman, C, Histoire de la Navigation hollandaise aux Indes orientates en 1595, Amsterdam.
  • 9. Kanakasabhai Pillai, The Tamils 1800 years ago, Madras, 1904.
  • 10. Paris, F. E., Essai sur la construction navale des peuples extra-Européens, n.d. [1841].
  • 11. Van Erp, Th., Voorstellingen van Vaartuigen op de Reliefs van den Boroboedoer: Nederlandsch-Indië, Oud en Nieuw, 8th Year, pp. 227-255, December, 1923, Den Haag.

In the article “The Feather Cloak of Tahiti” in the March Journal, 1943, would readers kindly make the following corrections on page 15:

in par. 2, line 1, for ‘mourning-custom’ read ‘mourning custom’.

in the same par., fourth line from end, for ‘less friable’ read ‘more friable’.

These errata were not the fault of the author, whose text was correct: unfortunately there was not time to send the proofs to Hawaii for reading, or the errors would not have appeared.

1   For particulars of the most important of these variations see the works listed in the bibliography under the names of A. C. Haddon and J. Hornell.