A cartwheel is shaped like a shallow cone

A cartwheel drawn in fine line: felloe, spoke, hub, tyre, each named. Beneath it a side elevation, and one figure picked out in red.

Dish, 25mm. That's not a decorative flourish. It's the reason the wheel doesn't fold.

Why isn't the wheel built flat?

Because flat is the weak shape for this job.

A cart doesn't just roll forward — it lurches sideways, over ruts, into kerbs, on a camber. A flat wheel meets a sideways knock with nothing but the stiffness of its own rim, and a rim is the easiest part of the wheel to bend. So the wheelwright sets the spokes leaning outward from the hub, not upright, which turns the wheel into a shallow cone rather than a disc.

Looked at side-on, that lean is the whole drawing. It looks like a fault in a flat sketch and turns out to be the argument.

What does the dish actually do?

It turns a sideways push into a straight push along the spoke.

A spoke set at an angle takes a lateral load along its own length rather than across it, and wood resists a load along its length far better than a load trying to bend it sideways. Set every spoke the same way and the whole wheel gains that resistance at once. The dish is the geometry doing the work that a thicker, heavier rim would otherwise have to do.

Which parts actually belong in the diagram?

Felloe, spoke, hub and tyre — the four that a wheelwright would point to first.

The felloe is the curved rim itself, built up from several shaped sections rather than bent from one piece. The tyre is the iron band shrunk on hot to clamp the felloe segments together as it cools. Between hub and felloe, the spokes carry the dish. A diagram naming all four, with a side elevation showing the lean, reads correctly to somebody who's built one — which is the same test a lettering guide has to pass.

Can I put a real wheelwright's stamp on it?

Not somebody else's — but that's a narrow restriction, not a wide one.

The geometry, the terms and the dish figure are trade knowledge, not anybody's property, in the same way a torque sequence is a fact rather than a brand. A specific firm's stamp or a museum collection's own drawing is a different matter — that's a document somebody made, and if a real one is the point, asking is the job. Left generic, the diagram is free to be exactly this precise.

Does the idea work for a wheel that isn't a cart's?

It works for the wheel most people actually own.

A bicycle wheel is dished too, for a different reason — the rear hub is asymmetric, so the spokes on one side run steeper than the other to keep the rim centred. Anyone who's trued a wheel has already met this problem from the other direction. A cutaway of a hub and its spoke angles reads just as well on somebody who's never seen a cart.

Isn't this just nostalgia for a dead trade?

The trade is genuinely small now, and it still has orders on the books.

Heritage carriages, museum wheels, showground vehicles and working farm carts all still get built and repaired by hand, by people who cut this exact profile this week. That's a real, current practice, not a re-enactment of one.

Does it print well?

Yes — it's line work with a single figure to protect.

Flat, hard-edged linework on a sand ground survives fabric cleanly at any size. The one thing to watch is the labelling: part names set at drafting scale drop below the weave's floor, so they want to be set larger and fewer than a real technical sheet would use. Charcoal for the drawing, one red for the dish figure, keeps the eye going straight to the number that matters.

How do I get one made?

Describe the wheel at JustOG — which parts to label, the elevation, and whether to include the dish figure. Pick a direction, drag the crop frame, see it composited on the real garment, and it's made to order and shipped.

Designs other people have published are in the shop.

Twenty-five millimetres, and the cart doesn't fold.