There is a moment in the building of a skin on frame craft when the thing declares its own logic. The ribs go in. The lashings tighten. What was a collection of bent wood and cordage becomes, suddenly and without announcement, a vessel. Not finished. Not even close. But unmistakably itself.
This is an account of getting there.
The Frame
I chose not to build on a strongback. The decision was partly practical — I wanted to understand what the simplest possible system could achieve — and partly the kind of stubbornness that produces either useful knowledge or wasted timber. Here, I think, it produced useful knowledge.
The method: a single temporary thwart, bent outwales sprung around it to find their curve, four formers dropped into the gap. The hull shape declared itself longitudinally. Stringers laid over. Ribs bent in over the outside of the stringers, then transferred beneath the frame once dried.
I have no photographs of this process. The day job was in one of its periodic states of mania and I was building at night, which is its own kind of knowledge — how noises travel so much farther and louder in the quiet air, and proximity to neighbours mean hand tools and strict timekeeping. The next build will be documented properly. This one had to be lived first.
On Rigidity and Give
Every connection in a skin on frame hull is a negotiation between rigidity and movement. Get it wrong in either direction and the craft fails — not slowly, with warning, but brittlely, at the worst possible moment. The lashings are not fastenings in the conventional sense. They are a system for distributing load across the whole frame, so that what the water gives the hull, the hull passes on without concentrating it anywhere fatal.
This is, it turns out, a reasonable philosophy for other things besides boats.
The Ribs: A Taxonomy of Failure
Instead of floorboards I chose half ribs and three-quarter ribs between the full ribs. The reasoning was sound: feet and knees stay clear of the skin, weight stays low, and the rib lengths could be placed strategically — a full rib under the kneeling thwart, three-quarter ribs at the seats, half ribs elsewhere. The logic was correct. The lashing time increased considerably.
The ribs are of varying widths and species. This is not aesthetic. It is the record of an investigation.
Sweet chestnut — air dried and green — was the first candidate. Lightweight. Famously rot-resistant. It begins bending beautifully and then, without warning, opens at a growth line. The inter-growth shear strength is too low for steam bending at this radius. The rib fails not dramatically but disappointingly, like a sentence that loses its nerve halfway through. Worth attempting. Not worth repeating.
Larch — largely knot-free stock salvaged from elsewhere — steam bent excellently. But tap an oak dowel home through larch at the outwale junction and it cracks. The wood has no patience for that particular insult. Filed under: knows its limits.
Green oak — the old answer, the one that has always worked, the one you return to when the experiments have run their course. The rest of the ribs are green oak. They bent without complaint.
Thwart Construction and the Problem of Sweet Chestnut
The thwarts are sweet chestnut — beautiful, lightweight, prone to splitting along the grain if you look at it unkindly. A small larch strip is glued to the underside of each thwart to resist this tendency. The thwart ends bear against the outwale through a small block rabbeted in and glued with casein glue, the bolt passing through the scupper, the whole assembly pulling on the gunwale when lifted rather than concentrating load at a single point.
Casein glue dries very dark on high-tannin woods. The staining around each glued block is not a mistake. It is the glue announcing its presence, which seems honest.
The final rib at each stem presented a problem of radius — too tight for single-thickness bending without catastrophic failure. The solution was to break the rib deliberately at the apex of the curve where it meets the hog, then lash the surrounding stringers to it on both sides. A broken rib that shares its load is more useful than a whole rib acting alone. There is something in this worth thinking about beyond boatbuilding.
On Japanese Tools
The Japanese woodworking tradition developed without the heavy static workbench that organises Western carpentry. The carpenter's body — feet, knees, weight — became the clamping system. The tools that evolved from this way of working are lighter, smaller, and designed for the pull stroke rather than the push: less force required, more control available, one hand sufficient where two would otherwise be needed.
For building a skin on frame hull — where the work is always in situ, always awkward, always demanding precision in positions a workbench cannot help you reach — this turns out to matter enormously. The kanna works with one hand on the pull stroke. The saws cut on the draw with a kerf so fine and a finish so clean that the wood seems to have agreed to the separation. The spokeshave here is small enough to get into places where a Western equivalent would simply refuse to go.
A modest budget, in Japanese tools, gets you remarkably good steel.
The adjustable countersink bit deserves separate mention. It can be set for any metric screw to give an exact depth and diameter, and does not drill the pilot hole simultaneously — which, when working with the small scantlings of a skin on frame canoe, is the difference between a clean joint and a split rib.
Mistakes and What They Cost
Western Red Cedar from Canada is expensive in the UK. British-grown cedar runs too knotty for clear lengths. I bought the Canadian stock — lightweight, rot-resistant, exactly what this build needed — and then learned what it costs in fragility. Any ding, any tool slip, any moment of inattention produces a stress riser in the surface. These need to be shaved smooth immediately. The wood is unforgiving of carelessness in a way that green oak is not.
I also managed to crack the top of one rib with sloppy drilling, and split slightly into the inwale beside it. Minor damage, easily repaired when the glue is next mixed. Chalked up. Moved on. The hull does not care about the builder's embarrassment.
Oiling the Frame
For this build I chose Roslag Mahogany — pine tar, linseed oil, gum turpentine — rather than mixing my own boat soup. I wanted to understand what a centuries-old Nordic formulation does before I presume to improve on it.
The smell alone is an argument for natural materials. Sweet, vinegary, woodsmoke, something close to an old forest on a warm day. It brings out the grain of the sweet chestnut and oak in a way that no synthetic finish approaches — not because it is more beautiful, but because it is more true to what the wood actually is.
The practical function: linseed oil and turpentine together reduce the internal friction of the tarred marline used for lashing. Saturated in the oil, the lashings can be pulled genuinely tight through the frapping turns without pre-stretching, even in natural fibre. The oil does not merely preserve the frame. It makes the lashing system work better. These things are connected. They usually are.
Copper Rivets
Every boat should have at least one copper rivet. The seat supports could have been dowelled and lashed. I used rivets. There are two rove types — dished and flat washer. For clinker planking, use the dished roves: they keep the planks tighter under load. Where water ingress is not the concern, flat washers serve. I used both, because it was what I had, and this turned out to be fine.
Next post - Skinning the frame
Join the conversation