Bark Canoes in Arnhem Land, Australia

This short video clip, filmed in 1948 in Arnhem Land, Northern Territory, Australia, shows a few of the steps of building a bark canoe and a bit of paddling what appears to be a different canoe at the conclusion of a successful sea turtle hunt.

Source: (Australia) National Film and Sound Archive

What we observe of the canoe under construction:

• The bark was stripped in a rectangular section, not the ovoid or lozenge-shaped sections as were used for the canoes of the Murray Valley and as seen in the image at the bottom of this post.
• The bark appears to be charred on the inside of the hull. This might have been done to make it more flexible, to harden it, or to drive out insects and/or moisture and prevent rot.
• The bark is folded roughly in half and held upright by stakes driven into the ground, which define the width of the hull.
• The ends are stitched closed with what we presume is vegetable fiber. The stitching is on an angle for a bit of bow or stern overhang. The bark is cut back to a line several inches outboard of the stitches after they are in place.
• A "thwart" is tapped into place to open the bark envelope to the desired width and keep it from closing up.
• There are gunwale timbers of round poles (seen briefly at 0:24) that stiffen the sides but do not extend to the ends of the boat. We assume they are stitched/tied to the top edges of the bark, but can't make out that detail.
• Abaft the thwart (assuming that we're looking toward the bow), a withy is tied between the opposite gunwales, apparently to keep the bark from opening too wide (which we assume could occur as it softens when wet).
• The ends are caulked with a liberal application of tree sap, probably mixed with some hardening agent like ash or charcoal.

The canoe shown in use appears to have a different bow shape than one under construction. The stem is vertical; there are no gunwale timbers; and the visible thwart is shaped flat, not like the unshaped round stick as appears in the construction segment. In spite of the lack of gunwale timbers, the boat is well-shaped and apparently stiff, handily carrying its crew of four. Paddlers #1 and #4 have full-length paddles; crewman #2 paddles mainly with his hands but briefly uses a very short paddle or perhaps some other available implement; crewman #3 paddles with both hands only.

We've written previously about canoes in Arnhem Land as they appear in the movie Ten Canoes. The construction method shown there appears to be the same, but the bow shape is different. In Ten Canoes, the craft have a ram bow or reverse overhang that may have been useful in parting the tall grass in the swamps where they were used. In contrast, the canoes in the present video were used at sea, where an overhanging bow would serve to ride up on waves and keep the interior dry, and a straight stem would at least not tend to shovel water into the boat.

Gunwale timbers similar to those shown in this video appear in other Australian bark canoes, like the one shown at the bottom of the image below. The shape of the bow in this example, however, is quite different from either of the boats shown in the present video.

Three types of Australian bark canoes. The one at top is the Murray Valley style, with the ends plugged with mud or clay. The ends of the canoe in the middle are bunched and tied. The canoe at the bottom has timber gunwales and a stitched end of complex shape. (Source: Edwin Doran, Jr., Wangka: Austronesian Canoe Origins)

The Australian Overlapping-Triangle Raft

Reader Geoff Cater steered me toward his fine summary page of indigenous Australian watercraft where, along with excellent images of various canoe types, there are a number of rafts.

The type that I find most intriguing consists of two triangular platforms, one stacked atop the other and facing in opposite directions. In two of the examples shown, fastening on each layer is by means by wood rods inserted through holes bored through the logs. In the model, the logs are fastened to their neighbors with thread, which may indicate lashed construction on the example from which the model was based, or might have been simply the modelmaker's expedient. Also in the model, the layers are attached to one another with a nail; it's unclear how the layers are attached in the other examples.

The photos and captions below are borrowed directly from Mr. Cater's webpage, for which thanks are given.

1931: A model raft made from six cylindrical wooden rods, collected in 1931 by Gerhard Laves from Bardi people at Cape Leveque, WA. Description: A model raft made from six cylindrical wooden rods with pointed ends and joined together with thread. A second "deck" of five pointed cylindrical wooden rods are attached to the first with a metal nail. Source: National Museum of Australia.

1931-1994: Double triangular raft, pole and paddle, Western Australia. Illustration by Xiangyi Mo. Barlow: Aboriginal Technology: Watercraft (1994) page 27. The illustration by Xiangyi Mo is possibly based on the model raft held by the NMA, above. The grass matting is possibly represents a seat or, if mixed with mud, a hearth for the fire often carried on board. Below are a punting-pole and a short-staffed spoon or pudding-stirrer paddle, as used by George Water in the photograph [below]. Their use throughout the continent and Tasmania indicates the pudding-stirrer's great antiquity, perhaps back to the last of the migrations from SE Asia.

1916: Worora youth on a mangrove tree raft (or 'kaloa'), George Water, Western Australia, 1916. Photograph by Herbert Basedow (1 of 2). Basedow, Herbert: The Australian Aboriginal. F.W. Preece & Sons, Adelaide, 1925, plate 22.
National Museum of Australia

In the model and, apparently, in the 1916 photo, the ends of the two layers line up pretty well. In the drawing, the two layers are offset, so that the logs' wider ends extend beyond the other level.

Why did the raftmakers face the two triangles in opposite directions? If two layers of logs were needed for buoyancy, it might make more sense to place the narrow ends one atop the other to form a roughly conventional "bow," since a forward-pointing triangle is more hydrodynamic. 

But as the photo shows, the bottom layer seems to provide enough buoyancy to hold the top layer and its passenger nearly or entirely clear of the water. By reversing the directions of the triangles, the logs of the top one do not nestle between the logs of the bottom one, so the top one actually rides higher and drier, while the forward-pointing triangle on the bottom still provides the desirable hydrodynamic efficiency.

Both the drawing and the photo show short, one-handed "pudding stirrer" paddles, and the drawing shows a pole as well: no doubt poles were commonly used for propulsion in shallow water.

Bark Canoes of Australia's Murray River

The bark canoes of the Australian Aborigines of the Murray River were among the simplest true boats ever built. Even so, they played an important role in the peoples' subsistence economy, and they had a kind of natural grace.

The Murray River forms much of the border between New South Wales and Victoria before entering South Australia, entering the Indian Ocean through Lake Alexandrina, south of Adelaide. (Map from Wikipedia. Click any image to enlarge.)

The Murray, with its major tributary the Darling, is the largest river system in Australia, running over 2,300 miles through a variety of terrain and ecosystems. It's not surprising that such a dominant geographical feature should have played a role in the material culture of a people. The Aborigines built bark canoes for hunting and fishing, and for transportation across the rivers in the system.

A tree from which a section of bark was cut for a canoe. Even though no bark canoes of the type described here have been built for over 100 years, several of these slow-growing "canoe trees" still remain, showing the boat-shaped scar.

The canoes were formed from a single piece of bark of the river red gum tree (Eucalyptus camaldulensis). A man would climb a tree and cut a lozenge-shaped section of bark, using either a stone axe or sharpened sticks. On occasion -- especially where the bark being taken was high on the tree -- the builder would cut hand- and toe-holds into the bark as an aid to climbing. Flat, flexible sticks were pressed between the bark and the wood all around the cut, being forced in bit by bit. Before the bark came away from the trunk, a line was passed around it to prevent it from falling to the ground. (The wood of the river red gum, by the way, is too dense to serve as a good dugout canoe, and even the bark is somewhat difficult to work.)

Murray River bark canoe under construction. Note the prop sticks around the edges.

Once flat on the ground, small sticks were placed under the bark to support its perimeter edges and small fires were built on top to evaporate much of the sap. This also caused the center to sag down, increasing the curvature of the hull and the height of the sides. A few sticks were placed across the hull from sheer to sheer to prevent the bark from curling inward too far, and sometimes these were left in place, in the nature of thwarts, when construction was complete. 


The end of the bark that was lower on the tree was always at the stern. This was much thicker than the bark at the other end, and often it could not be bent up sufficiently. In such cases, the stern would be closed in with a transom made from clay mixed with grass or other vegetable matter. 


The canoes were propelled with a pole about 14' long. In deeper water, this was used in the manner of a double-bladed paddle. In shallow water, the canoe was poled, always using the thicker end of the pole against the bottom. The narrower end was sharpened and served as a fish spear. The paddler always stood, except when hunting, when he crouched. Passengers would crouch or sit.

A funeral procession on the lower Murray, 1864. The corpse is in the foreground boat, covered with leaves. Note the small size and minimal freeboard of some of the canoes.

Canoes varied in size from one-man boats as short as 8' to boats over 20' capable of carrying as many as eight people. A big boat might have freeboard of 8", and beam was usually around 30", rarely more than 36". Stability apparently varied, with many European descriptions emphasizing the canoes' tenderness, while at least one claimed that one could climb into a large canoe from the water, over the side, without upsetting it. There are numerous recorded instances of both Aborigines and Europeans upsetting or swamping the canoes. Most Aborigines were apparently good swimmers, and this was rarely a disaster for them. 


Aborigines would fish and hunt from the canoes. One fishing technique was to stir up the bottom of the muddy river with the pole in an area suspected of harboring fish. When fish then fled into areas of vegetation, the paddler would follow their progress and deduce their location by the movement of the plants, and then thrust his spear "blind," so to speak, but often successfully.


For night-fishing, a small fire would be built near the bow on a platform of mud, bark and sticks. The fire would attract fish, which could then be speared by sight. One man would tend the fire, while another would handle the boat and wield the spear.


Early European explorers and settlers used the canoes, sometimes stealing them from Aborigines; sometimes acquiring them through barter or commissioning their construction, and occasionally building them for themselves after having observed the native techniques. One European recorded using a bark canoe to ferry as many as six sheep at a time across a river. Another told of carrying 400-pound bales of wool. 

On the Darling River, 1904. The canoe is in the South Australia Museum.

Of course, as European settlement progressed, European-style boats soon came into use, and Europeans generally ceased using bark canoes. Then, as Europeans came to dominate the riverfront land, they prohibited Aborigines from taking bark from their trees, and the use of the bark canoe died out entirely.


(Most content and all photos from: Aboriginal Bark Canoes of the Murray Valley by Robert Edwards.)

Australian Bark Craft, Canoes and Otherwise

A few weeks ago I presented material from Edwin Doran on the bark canoes of Australia. I can now elaborate on that a bit, based on images from Paul Johnstone's The Sea-Craft of Prehistory.

Doran's line drawing of an Australian canoe that was built by bunching and tying the ends of a half-tube of bark was obviously based on this fine early 19th-century engraving, from Lesueur and Petit's Atlas to Peron's "Voyage of Discovery." (I've transcribed the name of the source verbatim from Johnstone's caption, but unfortunately, there is no bibliography or footnote citation for this source.) In the engraving, one can see thwarts, and ties (probably bark) running across the boat, from shear to shear just above the thwarts, , but no gunwale members, ribs, or other structure. The Aboriginal people paddling in the background provide scale, showing that this was a small two-man canoe indeed. They are paddling with single-handed paddles, and they have a small fire going amidships. In the background, a man is carrying a similar boat above his head. From the abundance of ducks and other waterfowl in the image, one can safely speculate that the canoes were used for waterfowling, egg collecting, or both.

My earlier blog included Doran's description of the bark rafts of Tasmania, which stated that they were extremely temporary craft, becoming waterlogged in about six hours. Here's an image, from the same source as the above image, by way of Johnstone.

As Johnstone notes, these are similar in design to bundled-reed floats that were used widely in many cultures around the world, the Peruvian caballito pictured just a couple posts ago being an example. But given their short working life, Johnstone seems justified in calling them among the most primitive of craft. I have no information on what material was used to bind them together, nor on the type of bark used in the bundles.

Again, the two paddlers in the background show that this is a very small tandem craft, and it doesn't look particularly stable either. They carry poles, but apparently no paddles, and a pair of poles appear in the foreground as well. As any modern canoe poler can attest, however, poles can serve entirely adequately to propel a boat in deep water, as their projected area under water is often not much less than that of a paddle blade.

Bark Canoes of Australia

A few weeks back I wrote about the bark canoes of Australia's Arafura Swamp, noting that the design was unique to that area. Bark canoes of other styles were used elsewhere in Australia, as shown below in this image from Edwin Doran, Jr.'s Wangka: Austronesian Canoe Origins, which I cited a recent post. As always, the image may be clicked to enlarge.

To quote Doran's complete description (items in [square brackets] are mine):

Among the simplest, and presumably most ancient, types of watercraft in Austronesia are boats made of bark. The simplest of all are formed of bark sheets plugged at the ends with clay [top image], whereas others are tied at the ends [center] and more complex ones have sewed ends [bottom]. The principle location [i.e., within Austronesia] of bark canoes is Australia, but a few relicts are found in Indonesia, probably indicating a much wider distribution.

The very simple bark rafts of the Tasmanians allowed some mobility on the water but became waterlogged in about six hours and were never used for distances offshore greater than about eight miles.

The sewn canoe shown differs from those of Arafura not only in the shape of the end but, significantly, in its having gunwale members lashed to the sheerline. The Arafura boats were entirely without added structural members.

Although Doran says that bark canoes were "Among the simplest, and presumably most ancient types of watercraft" (emphasis added), he is not implying that they are the simplest or most ancient. That distinction he applies to log rafts, noting that true rafts rely on the inherent buoyancy of their material of construction, rather than on a designed, built structure that displaces water. Also in support of his argument: it is a simpler matter to tie one floating log to another, thus creating a raft without any tools at all, whereas even the simplest bark canoe requires tools of some sort to strip the bark from the tree.

Doran shows the distribution of the various types of bark canoes in the following map, along with the distribution of dugout-based outrigger canoes with constituted the focus of his study.

Ten Canoes

The 2006 movie Ten Canoes by Rolf de Heer is a mesmerizing look at the pre-European-contact culture of Australia and well worth viewing even if it didn't feature indigenous boats. A story-within-a-story within yet another one, they're all set in the Arafura Swamp in Arnham Land, in what is now the Northern Territory, on the northwestern shore of the Gulf of Carpentaria. Director de Heer's inspiration for the movie came from a 1936 photograph by anthropologist Donald Thomson, that showed 10 solo bark canoes being poled through long grass in the swamp.

The story was built around that image, but when he showed up there to film, seeking local craftspeople to make the props, and local residents -- all of them Australian Aborigines -- to fill all the movie's roles (none of whom had any acting experience), de Heer found that knowledge of the swamp's indigenous canoe had disappeared from the region. de Heer was able to recruit one individual with some bark canoe building experience from a neighboring district, but as his first canoe began to take shape, it was clear that its design was not that of the Arafura Swamp as photographed by Thomson. So combining the technical skills of the experienced builder (whose name was Pussycat, and who is pictured below) with Thomson's detailed drawings and photographs, de Heer was able to guide the process and build the ten canoes needed for the movie with probably a fairly high degree of accuracy. Certainly, if Thomson's book constitutes the only careful documentation, it might be impossible to come up with a more accurate reproduction, since native knowledge of the type vanished two or three generations ago.

I haven't seen Thomson's book (nor do I know its title), but I was able to make out the following details from the movie, which shows certain steps of the building process, and the DVD's "special feature" (far more interesting than most) that shows more of it.

Bark is removed from the standing tree (type unknown) by first stripping vertical fibers, then making a long vertical cut, and prying the bark from the trunk with long poles. The bark is far rougher and stiffer than birch: when it comes off the tree, there is no tendency for it to fold over. Instead, it retains its longitudinal shape, and it is carried as a cylinder to an area of shallow water, where it is submerged and allowed to soak.

After a time (duration unknown), it is removed from the water. It can now be unrolled to lie flat. It is flopped onto a low, live fire to soften further. It chars somewhat in the process.

Removed from the fire, it is folded in half lengthwise, but only at the two ends, the section amidships being allowed to sag open. The ends are held between pairs of vertical stakes which are lashed tightly together. The stern is left square, and it is sewn several inches inboard of the end with tree fibers -- perhaps the same ones that were stripped off the bark before it was removed from the tree?

The bow is cut down to a "ram bow" shape. Starting perhaps 2 to 3 feet inboard of the front end, the top "gunwale" edge of the bark is cut in a downward slope toward the front. This is also sewn with bark fibers. A simple stopper knot is tied at the end a bundle of fibers, then sewing proceeds in a simple in-and-out pattern. More fibrous material is shoved inside the seam, and this may serve as a kind of caulking.  Just aft of the stitching, the top edges of the bark are pushed outward and down.

One or two simple wooden poles are lashed across the boat amidships to hold the hull apart, and shorter ones ("stretchers") are set well below the sheer well into the ends to provide some width and buoyancy there. The ends of these "thwarts" bear directly against the inside of the bark. The midships thwarts are quite long, the result being that the boat takes on a substantial sheer, while midships it is very shallow -- just a couple inches of freeboard when the paddler sits amidships. The run is rather flat aft, and it rises well above the waterline forward, but I could not tell how the bottom was shaped thus. The leading point of the "ram bow" is thus well above the waterline, and this is supposed to allow the boat to pass more easily through the long grass that grows in the swamp's shallow waters. The design is unique to the Arafura Swamp.

The canoes are decidedly "primitive" -- i.e., of extremely simple construction. The stiff bark is itself the canoe's only structure. They have no gunwales, no ribs, no internal strengthening of any kind, other than the simple cross-pieces whose purpose is only to hold the sides apart.

In the film, actors propel the boats in swampy areas by poling from a standing position. In open water, they paddled, seated amidships, with a small one-handed paddle in each hand. Some of the actors paddled with both hands moving together, typically leaning their torsos forward for the "catch" and leaning back during the stroke. This looked strenuous. Other actors paddled the two hands alternately, maintaining a steady motion like swimming, and this looked much easier to maintain over long distances.

On the set-up screen of the DVD I rented from Netflix, the scene shown above was shown, in which the ten canoeists paddle for a couple minutes across a stretch of open water. I repeated this scene over and over, perhaps 15 or 20 times. Although it's just a movie, and its accuracy is dubious, it had an amazing, mezmerizing sense of authenticity -- it really felt like one was watching pre-contact indigenous canoeists on a hunting/gathering trip.

(All photos are from the movie's website.)

First Boats, First Voyagers?

A good friend gave me a copy of Jared Diamond's Pulitzer Prize-winning Guns, Germs, and Steel: The Fates of Human Societies for Christmas. I'm about halfway through it -- good stuff, indeed. Early on, there's an interesting passage about the probable use of watercraft a long, long time ago. I'll quote at some length:

The Great Leap Forward (what Diamond calls the evolution from earlier versions of man to Cro-Magnons, the first fully modern humans) coincides with the first proven major extenion of human geographic range since our ancestors' colonization of Eurasia. That extension consisted of the occupation of Australia and New Guinea, joined at that time into a single continent. Many radiocarbon-dated sites attest to human presence in Autralia/New Guinea between 40,000 and 30,000 years ago (plus the inevitable somewhat older claims of contested validity). Within a short time of that initial peopling, humans had expanded over the whole continent and adapted to its diverse habitats, from the tropical rain forests and high mountains of New Guinea to the dry interior and wet southeastern corner of Australia.
During the Ice Ages, so much of the oceans' water was locked up in glaciers that worldwide sea levels dropped hundreds of feet below their present stand. As a result, what are now the shallow seas between Asia and the Indonesian islands of Sumatra, Borneo, Java, and Bali became dry land. (So did other shallow straits, such as the Bering Strait and the English Channel.) The edge of the Southeast Asian mainland then lay 700 miles east of its present location. Nevertheless, central Indonesian islands between Bali and Australia remained surrounded and separated by deep-water channels. To reach Australia/New Guinea from the Asian mainland at that time still required crossing a minimum of eight channels, the broadest of which was at least 50 miles wide. Most of those channels divided islands visible from each other, but Australia itself was always invisible from even the nearest Indonesian islands, Timor and Tanimbar. Thus, the occupation of Australia/New Guinea is momentous in that it demanded watercraft and provides by far the earliest evidence of their use in history. Not until about 30,000 years later (13,000 years ago) is there strong evidence of watercraft anywhere else in the world, from the Mediterranean.
Initially, archaeologists considered the possibility that the colonization of Australia/New Guinea was achieved accidentally by just a few people swept to sea while fishing on a raft near an Indonesian island. In an extreme scenario the first settlers are pictured as having consisted of a single pregnant young woman carrying a male fetus. But believers in the fluke-colonization theory have been surprised by recent discoveries that still other islands, lying to the east of New Guinea, were colonized soon after New Guinea itself, by around 35,000 years ago. Those islands were New Britain and New Ireland, in the Bismarck Archipelago, and Buka, in the Solomon Archipelago. Buka lies out of sight of the closest island to the west and could have been reached only by crossing a water gap of about 100 miles. Thus, early Australians and new Guineans were probably capable of intentionally traveling over water to visible islands, and were using watercraft sufficiently often that the colonization of even invisible distant island was repeatedly achieved unintentionally.

Wow. 40,000 years ago! You go, boat people.