Back to the Proa page or to The Cheap Page 

A Gilbert Islands Canoe - 1945


Here is an excellent brief description of how the Gilbertese canoes are constructed. It appeared as a letter to the American Anthropologist from Drews, who apparently was on Tarawa after it had been recaptured. The careful depiction of how the hama is braced really solves a lot of conceptual problems if you'd like to build a platform with a more traditional bracing system. The Y-shaped thing he can't quite figure out is for the steering paddle.

from: American Anthropologist, 1945.


The Gilbertese canoe illustrated was sketched in a village on an islet at the southern end of Tarawa Atoll. Little concerning its construction or its ownership could be learned because the natives at this village spoke no English. The author was on duty in the area too short a time to pursue the study more thoroughly.

Canoes built by the Gilbertese are of plank construction in marked contrast to the dugouts built by the natives of the Ellice Islands, the neighboring group of atolls to the south. In an earlier day, Gilbert Islanders hewed the planks for their canoes from trunks of trees growing on the islands or from logs cast up on the beaches. At present, however, according to a native informant contacted briefly at a later date, planks are obtained either new from traders, as driftwood lumber, or from previously constructed canoes.

The canoe described here was twenty-seven feet long, eighteen inches in beam, and thirty inches from keel to the upper edge of the gunwale. Throughout its construction no other fastening medium than sennit cord, from coconut-husk fiber, appeared to have been used. The planks along the side (Fig. 1), five in number, were about three-fourths of an inch thick. Since no one plank was long enough to extend from stem to stern, numerous joints were necessarily made to obtain the length desired. Two of these are shown (Fig. 1). Both were good examples of the carpenter's art. The final junction was achieved, as everywhere else in the boat, by tying with sennit cord.

Fig. 1


At their edges the planks were tied together at intervals of four inches through matching awl-holes (Fig. 1, d). At places of stress, particularly where the bottom plank met the keel and also where the ends of the planks met the bow and stern pieces (Fig. 2 a) the intervals between the ties were reduced to two inches (Fig. 1 b).

The one-piece rib units require little description. However they were beautifully worked to an even thickness of about two inches with the same care of construction shown in all the workmanship in the boat. At the center of the bottom of each rib unit was a notch clearly shown in the sketch (Fig 4 a), which permitted water to gather at one place in the boat thus facilitating bailing. Lashed securely to each pair of ribs was a thwart (Fig. 4 b) about one and a half inches square in cross-section.

Fig. 2

To these cross pieces were secured the slats which formed the half deck (Figs. 2 d, 4 c) built on the outrigger side of the boat. To their opposite ends were fastened the flaring weather boards (Fig. 4 e) and the wider deck boards (Figs. 4 f, 3 e, 2 e) on both sides of the canoe.

The outrigger float (Fig. 3 a) had been employed formerly on some other craft where the natural forks (Fig. 3 b) used to elevate the boom had been placed considerably closer together than in this boat. The great skill of the canoe maker who had originally used the float was amply shown by the care with which he had faired the sennit fastenings into the body of the float (Fig. 3 c) to present the least possible resistance to the water. The three parallel outrigger poles and the crossed strengthening members were of material two inches wide and an inch and a half thick (Figs. 3 d, 4 d). The parallel poles extended twelve feet beyond the side of the boat and were spaced two feet apart. Each pole was lashed rigidly to a corresponding thwart. The detail of the outrigger pole assembly may be seen in Fig. 3 and requires no comment.


Fig. 3

No mast was in evidence, although what appeared to be a stepping block was tied securely to a thwart and to the strips that formed the half deck. The block was secured at the outrigger side of the boat amidship.

Two blocks (Fig. 2 b) apparently designed to hold the butts of fish spears or paddles were tied to the half decking near both stem and stern of the boat in the same relative position as the one shown. The other portion of the rest consisted of a pair of poles with a forking branch left intact (Figs. 2 c, 1 c) fastened so that the fork end extended approximately eighteen inches beyond the side of the boat opposite the outrigger. These were located, one where the bow piece met the center section of the canoe, and the other where the stern piece did the same, as shown in Figure 2.

 Fig. 4
Only one paddle was seen, and it may or may not have been a representative example. The blade was about six inches in width by a foot in length, and tapered to a rounded point. The two-and-a-half-foot handle of the paddle had been extended by the addition of a shaft an inch and a quarter in diameter and somewhat more than five feet long. The joint uniting the two pieces was a simple tapered joint securely tied with sennit cord.

The bow and stern were identical. A flaring pair of boards (Fig. 2 f) at the bow were apparently designed to make the knife-like boat drier and more comfortable in a sea. Even for lagoon use, the bow boards would be helpful because at the down-wind end of this atoll's big lagoon the strong afternoon breezes are capable of kicking up a sizeable chop. Naval whaleboats have been swamped within the lagoon on a brisk day.

However, the same winds that threaten the canoes with swamping assist them at flying nimbly around the enclosed waters. The natives are not unaware of the graceful appearance of their craft and expertly take great pleasure at displaying them at their best, i.e. heeling with the wind, with the float raised above the surface to reduce the drag.

to the Top or to The Cheap Page 
1.0 COD 3/21/99