- 144

After decades of discussion, no consensus has been reached on the question of purposeful versus accidental voyaging among the Pacific Islands. 1 An equally lengthy and unsettled debate continues on the number, time, and significance of contacts across the oceans between Old World and New. 2 Although these topics lie behind the scenes, this paper is simply a description of two oceanic canoes. As factual data accumulate 3 they increase our understanding and may contribute to an eventual meeting of minds on the larger problems at issue.

The purpose of this paper is to present a reasonably precise and careful description of two types of Pacific outrigger canoes, the single-outrigger wa of the Caroline Islands and the double outrigger vinta of the Sulu Archipelago, and to compare these craft with a modern analogue, the trimaran. Detailed drawings allow careful inspection of the forms of these craft, and a polar diagram permits accurate comparison of their sailing ability. The primitive craft compare quite favourably with the modern boat and in some ways are superior. On some points of sailing they are faster, and their seaworthiness for long voyages is not to be questioned.

Standard field techniques of note-taking, interviewing, and photography (both 35 mm still and 16 mm motion pictures) used during the field work in Spring, 1970 hardly need mention except to note that motion pictures of certain aspects of setting sails and changing tacks add detail to our knowledge of these procedures. In addition, however, lines drawings (hull contours as viewed from above, the side, and end-on) were made which provide an accurate description of the exact shape of hulls. 4

The sailing performance of the canoes was measured with simple instruments which give a reasonably accurate picture. 5 Wind speed was

- i

PLATE 1

The single-outrigger wa of Puluwat Atoll

- ii Page is blank

- 145

measured with a small Dwyer anemometer, later calibrated against an accurate meteorological instrument; boat speed was measured with a Sims speed meter. Both of these instruments are simple pitot tubes. Wind angle was measured with a vane constructed in the field, mounted on deck, and read against a large protractor. Leeway angle was measured visually with a protractor and later estimated from the wake angle; this critical angle is the least accurate of all measurements. Dozens of observations were made on each trial, permitting the scatter of data points to be averaged out with some confidence.

Since wind speed and direction vary constantly and rapidly, it is necessary to measure these parameters as simultaneously as possible with boat speed and leeway angle. Figure 1 indicates how this is accomplished. The desired data are true wind speed (VT), boat speed (VS), and angle of course made good to the true wind (Σ). The combined directions and speeds of true wind and boat, however, result in an apparent wind of different velocity (VA); this can be measured as the boat moves along, almost simultaneously with Vs. If at the same time one measures the angle between apparent wind and boat heading and adds the angle of leeway, then angle beta is obtained. By means of trigonometry, or graphically, one can calculate VT and gamma, which, with VS, are the three required pieces of information. 6 A correction factor for height of measurement of wind relative to centre of effort of sail must be included in the calculations. In order to make different wind speeds comparable, the ratio VS/VT is used instead of boat speed alone. This may be done because the boat speed of these types varies directly with wind speed until the latter exceeds about 8 to 10 knots.

The principal reason for choosing the wa of Puluwat Atoll, west of Truk (Plate 1), as the type example for single-outrigger canoes, was that Gladwin had recently worked on the island and kindly let me use a typescript of his book. 7 It was evident that the measurements I was interested in taking filled one of the few empty niches in his field observations. The construction, method of tacking, seaworthiness, and voyaging capacity of these particular Micronesian canoes are well documented, by Gladwin and others, hence I was able to concentrate specifically on obtaining one limited kind of data.

Figure 2 is a schematic drawing which illustrates general shapes and dimensions of hull and sail of a wa named Mikael which had the reputation of being a fast sailer. The plan view and float side-elevation of Figure 3 are drawn carefully to scale and provide certain construction details. A midship cross-section and the three views which together constitute a lines drawing and an accurate description of the complex hull shape make up Figure 4.

Choice of a Sulu vinta 8 as the double-outrigger canoe to be studied was conditioned by scientific, political, and logistical considerations.

- 146

FIGURE 1

The sailing triangle

Relatively little is known about Sulu, in contrast to the main area of the Philippines, but proximity to Indonesia promised a wealth of boat types, as indeed turned out to be the case. Some seven types were identified and ultimately will be described and discussed. A knowledge of English by many people and a considerable residual of warm feeling toward Americans suggested greater facility in field work than did Indonesia; contacts with several priests of the Oblate order, who were particularly kind and helpful during the field work, also boded well. A considerable shortening of travel distance, as compared to that through Djakarta, also was persuasive. The actual site of study and boat measurements was the village of Bongao, off the south-west end of Tawi Tawi Island, itself near the south-west end of the Sulu chain. A Samal fisherman, Mr Arenaki Nardi, was engaged together with his vinta for the study; his name is given to the boat for identification purposes.

Despite the considerable number of photographs of vintas which are seen in scientific and popular accounts, there was relatively little good information about them in the literature until recently (Plate 2). A few comments on Sulu canoes are given by Folkard, 9 Hornell contributes some verbal description, names for parts and photographs, but no draw-

- 147

FIGURE 2

Sail plan of the wa Mikael

- 148

FIGURE 3

Deck plan and float elevations of the wa Mikael

- 149

FIGURE 4

Midship cross-section and lines of the wa Mikael

- 150

FIGURE 5

Sail plan of the vinta Nardi

- 151

FIGURE 6

Plan, midships section, and lines of the vinta Nardi, (Photograph by Henry A. Morss Jr.)

- 152

FIGURE 7

Sail and deck plans of the trimaran Coqui. (Taken from brochure of the manufacturer, Warren Products, Warren, Rhode Island.)

- 153

FIGURE 8

Lines and sections of the trimaran Coqui

- 154

FIGURE 9

Sailing speeds of the wa, vinta, and trimaran

ings 10 and even less information is provided by Haddon. 11 One photograph appears in Hornell's major summation on boats 12 and is now easily available in the recent reprint 13 while some material mostly concerned with artistic endeavours but accompanied by three excellent photographs of details is to be found in Szanton. 14 The recent paper by Spoehr 15 provides the most detailed and useful description yet to appear.

The above descriptions are not expanded here except in the limited but critical fashion which is the subject of this paper. The schematic drawing of Figure 5 gives a general outline of hull and sail plus certain dimensions. Figure 6 provides a plan view, a midships cross-section, and the three line drawings which define the hull shape.

- iii

PLATE 2

Vinta Nardi off Bongao, Sulu

- iv

PLATE 3

The trimaran Coqui

- 155

The trimaran is the modern analogue of the double-outrigger canoe. 16 Although experiments with multiple-hulled craft in the West go back at least to the seventeenth century and several fast double-hulled yachts were built by Nathaniel Herreshoff in the nineteenth century, 17 the present wide popularity of the type has developed in the last 25 years. Trimarans, which have been publicised in largest measure by Piver, 18 solve the stability problem ubiquitous in sailing craft by means of floats extended on both sides of the main hull. In contrast to double-outrigger canoes, however, the floats are often quite large and buoyant and contribute considerably to drag which slows the boat.

It happens that performance data are available for a particular type of modern boat known as the Arrowhead trimaran. The individual boat is the Coqui (Plate 3), used by its owner as an experimental craft, and the configuration for which data are available is that of 1968. 19 It may be taken as a representative modern trimaran, certainly well constructed and seaworthy, of reasonably good performance. Some trimarans (including the Coqui in a later configuration) are much faster; others are appreciably slower.

Figure 7 is the manufacturer's illustration of the sail plan, a typical modern Bermudian mainsail and jib, and a deck plan. The forward dagger-board unique in the Coqui is omitted from this drawing. A longitudinal section and cross-section, as well as the usual three lines drawings, for both main hull and float, are given in Figure 8.

A number of critical characteristics have been taken from the drawings above and from other data on the three boats and arranged in tabular form for ease of comparison (Table 1). By coincidence, the three boats are quite similar in size. Overall length, waterline length, and sail area are very nearly the same and give increased confidence that comparisons among the boats are justified. The vinta and trimaran are estimated to be the same weight, but the wa, in part because of larger crew, is much heavier. The sail area/weight ratio (which is preferably used as shown to give a measure of power) is exactly the same for vinta and trimaran, but less efficient for the wa. 20 Both the wa and the trimaran have adequate freeboard for the open sea, but the vinta is somewhat inadequate in this respect. Probably the most crucial difference lies in length to beam ratio, a fundamental factor in reducing drag and increasing speed. The Pacific canoes are notably more narrow (ratios of about 10 and 13 respectively) than the trimaran (ratio of about 7).

- 156

The hundreds of data bits measuring performance can be reduced to the polar graphs of Figure 9. Curves are drawn which represent the speeds of the three boat types when sailing at various angles (gamma) to the true wind. Speed is represented radially by circles at increasing intervals of boat speed/wind speed ratio (VS/VT). No sailboat, of course, will sail directly into the wind, but as the bow of the boat is inclined farther and farther away from the wind the speed increases, then decreases slightly. The trimaran reaches its maximum speed to windward (represented by the tangent to its curve) at about 54 degrees and is superior in performance to the other boats at increasing angles to the wind to about 70 degrees. At that point, the speed of the wa surpasses the trimaran and is the greatest of the three boats to a gamma angle of about 95 degrees. Because of its poorly setting and relatively inefficient sail type, the vinta does not sail very well closehauled, but as the wind comes abeam its speed rapidly increases. From gamma 95 on around to a downwind run it is notably faster than the other boats.

- 157

Critical performance data are extracted from the polar graph and presented in Table 2 for easy comparison. If the rank orders of speed for three critical points of sailing—closehauled for maximum progress against the wind, maximum speed, and speed downwind—are listed, then added, it is apparent that the order of merit is vinta, wa, and trimaran. This by no means includes all factors to be considered in judging the three boats, but it indicates conclusively that these primitive craft are superior to a modern boat on significant points of sailing.

Although the performance data for this paper are limited and should be amplified by studies of the same and other primitive craft using more sophisticated and accurate measuring devices, the fundamental point appears to be adequately documented. No one has doubted that Pacific canoes sailed fast, but we know now with some accuracy just how fast they sail, and, in addition, that they compare favourably with modern fast sailing yachts. This level of performance comes as something of a surprise and requires an explanation.

In part, present-day “primitive” craft are less primitive than their predecessors because of the use of canvas for sails. As Gladwin has noted, 23 voyages in Micronesia are quicker and more efficient with the substitution of modern materials for pandanus matting. It might be added that modernity proceeds apace; two canoes on Satawal Island now have dacron sails and outperform the canoes with canvas sails. 24

The most fundamental explanation lies in hull shape. Pacific canoes are well shaped, have fine entrances and runs, and have nearly ideal length to beam ratios for minimum drag. Mr Edmond Bruce, an expert in these matters, concurs with this explanation and in addition suggests that wave interference between hull and floats may have an adverse effect on trimaran performance. 25

Certain obvious deficiencies in the Pacific canoes should be mentioned. It is clear that neither of the boats measured has enough lateral plane to inhibit leeway sufficiently and thus permit good closehauled sailing. The hard chines (sharp angles at the bottom) in the vinta underbody are probably deleterious also when closehauled; the rectangular lug sail is most inefficient when sailing on the wind because of its unsupported luff (front edge). The chines which form part of the entrance angle on the wa also affect performance. It was possible to identify the faster canoes on Puluwat by their finer entrance angles on main hull and float and by the smoothness of fairing into the main hull body.

A few comments on voyaging capacity of Pacific canoes may be added to many in the literature. 26 Voyages of 150 miles across open ocean by Puluwat and other west-central Carolines canoes are commonplace. 27

- 158

A canoe from Satawal made the 500-mile voyage to Saipan in 1970, and a second such voyage is planned for 1971. 28 Star courses between islands are known on Puluwat for all major islands from Tobi, south-west of Paalau, to Makin in the Gilbert Islands—clear evidence of repeated trips over various parts of this 3,000-mile-long region. 29

Although vintas operate in somewhat more protected waters, the specific boat illustrated in this paper has been used for fishing from Bongao, Sulu to Leyte in the Central Philippines, a distance of about 600 miles. The large 100-foot double-outrigger canoes called basnig, now motorised for the most part, but with a few sailing examples still in operation, have been observed far at sea in 12- to 15-foot waves. 30 A Sulu vessel (but not an outrigger type) was used in 1779 for a trip extending from the north tip of Borneo to east of Geelvink Bay in New Guinea, a distance of over 1,500 miles, much of it in the open Pacific. 31

The examples cited above should help allay any doubts that exist as to the voyaging capacity of Pacific canoes of the types described here. Seaworthiness is implicit in the ability to make such voyages. A specific point illustrating the latter is the technique known to Caroline Islanders for righting their canoes after they have capsized at sea. In brief, the mast is rigged from under side of float to a sheer legs erected above the bottom of the capsized boat. Four men climb quickly up the inclined mast, their weight forcing the float to submerge to a point directly underneath the main hull. Past this point the float's own buoyancy takes it back to the surface in righted position whereupon the canoe is bailed, rerigged and continues on its voyage.

The excellent speed and performance, seaworthiness, and general voyaging capacity of the wa and the vinta cannot be seriously questioned. Their performance in comparison with modern sailing yachts is remarkably good. Although these characteristics cannot be used to prove that long, purposeful inter-island and trans-oceanic voyages occurred, the opposing argument certainly is not sustained. No one can argue seriously that such long voyages did not take place because the available watercraft were inadequate.

• BLAKE, W. M., 1935. “Taking off the Lines of a Boat.” Mariner's Mirror, 21:5-11.
• BRUCE, Edmond, 1966. “Performance Measuring Instruments for Sailing Craft.” Amateur Yacht Research Society Publication Number 56, pp. 36-48.
• —— 1970. Personal communication, 29 December.
• —— 1971. Personal communication, 19 January.

• CHAPELLE, Howard I., 1935. The History of American Sailing Ships. New York, Norton.

• FINNEY, Ben R., 1967. “New Perspectives on Polynesian Voyaging”, in B. Highland et al. (eds.), Polynesian Cultural History. Bishop Museum Special Publication 39. Honolulu, Bishop Museum Press.

- 159

• FOLKARD, Henry C., 1901. The Sailing Boat. London, Stanford, 5th ed.
• FORREST, Thomas, Captain, 1779. A Voyage to New Guinea, and the Moluccas, from Balambangan: . . . London, Scott.

• GLADWIN, Thomas, 1970. East Is a Big Bird. Cambridge, Harvard University Press.
• GOLSON, Jack (ed.), 1962. Polynesian Navigation: a Symposium on Andrew Sharp's Theory of Accidental Voyages. Memoir No. 34. Wellington, The Polynesian Society.

• HADDON, A. C., 1920. “The Outriggers of Indonesian Canoes.” Royal Anthropological Institute Journal, 50:69-134.
• HOGG, John, 1966. “Yacht Performance Measurement.” Amateur Yacht Research Society Publication No. 56, pp. 49-69.
• —— 1967. “The Calculation and Presentation of Sailing Measurements.” Amateur Yacht Research Society Publication No. 61, pp. 14-19.
• HORNELL, James, 1920. “The Outrigger Canoes of Indonesia.” Madras Fisheries Bulletin, 12:43-114.
• —— 1946. Water Transport: Origins and Early Evolution. Cambridge, University Press.
• —— 1970. Water Transport: Origins and Early Evolution. Newton Abbot: David and Charles (reprint of 1946 edition).

• LAUER, Peter K., 1970. “Sailing with the Amphlett Islanders.” Journal of the Polynesian Society, 79:381-98.
• LEVISON, Michael; FENNER, SENTANCE, WARD, and WEBB, 1969. “A Model of Accidental Drift Voyaging in the Pacific Ocean with Applications to the Polynesia Colonization Problem.” Information Processing. (Amsterdam, North-Holland Publ. Co.), 68:1521-6.
• LEWIS, David, 1970. “Polynesian and Micronesian Navigation Techniques.” Journal of the Institute of Navigation, 23:432-47.

• McCOY, Michael, 1971a. Personal communication, 12 January.
• —— 1971b. Personal communication, April.
• MARCHAJ, C. A., 1964. Sailing Theory and Practice. New York, Dodd, Mead & Co.
• MORSS, Henry A. Jr., 1969. “Polar Curve of Coqui Arrowhead Trimaran.” Amateur Yacht Research Society Publication No. 70, pp. 63-6.

• PHILLIPS-BIRT, Douglas, 1962. Fore and Aft Sailing Craft. London, Seeley Service.
• —— 1966. Ships and Boats. The Nature of Their Design. London, Studio Vista.
• PIVER, Arthur, 1961. Trans-Atlantic Trimaran. Mill Valley, Calif., Pi-Craft.
• —— 1965. Trimaran Third Book. Mill Valley, Calif., Pi-Craft.

• RILEY, Carroll L.; KELLEY, PENNINGTON, and RANDS (eds.), 1971. Man Across the Sea. Austin, University of Texas Press.
• ROMMY, Captain C., 1970. Personal communication, 4 May.

• SHARP, Andrew, 1964. Ancient Voyagers in Polynesia. Berkeley and Los Angeles, University of California Press.
• SPOEHR, Alexander, 1971. “The Double Outrigger Sailing Canoe of Zamboanga and the Sulu Archipelago, Southern Philippines.” Occasional Papers of B. P. Bishop Museum, 24:115-26.
• SZANTON, David, 1963. “Art in Sulu: A Survey,” in Frank Lynch (ed.), Sulu's People and Their Art. Inst. of Philippine Culture Papers, No. 3, pp. 27-66.