One Matchstick Clockwise, and I Can't Prove It Was the Planet
Foucault Pendulum Building and Latitude Measurement 🎮 Play: Sand & SiderealThere is exactly one latitude where this experiment refuses to run, and mercifully I don’t live on it. On the equator a Foucault pendulum swings all day and never turns — the plane of its swing stays welded to the floor beneath it. Everywhere else the floor slides out from under that swing at a rate of 15° times the sine of your latitude, every hour. Outside Edmonton the sine works out to about 0.79, which should hand me roughly 11.7° an hour and a full circle in a little under 31 hours. All I have to do is hang a weight, get it swinging perfectly straight, and not touch it for a day and a half.
I managed about four minutes of perfectly straight before the physics started lying to me.

The plane with an opinion
The bob is a four-kilogram lead diving weight with a steel eyebolt epoxied into a drilled seat, hung on four metres of wire down the open stairwell. My first real mistake was at the very top, and it’s a subtle one worth stating plainly. My instinct — straight out of the evening I spent tweezing seventeen jewels out of a pocket watch — was a knife-edge pivot. A knife edge is exactly right for a pendulum that must swing in one plane forever, which is precisely what a clock wants. It is exactly wrong here. A Foucault pendulum has to be free to swing in any direction and drift slowly through all of them; clamp it on an edge and you’ve told it which plane it’s allowed to use. The correct mount is a single fine wire gripped symmetrically, so the top of the pendulum holds no opinion about compass direction. I swapped the edge for a length of piano wire pinched in a drilled brass cone.
Then the launch, which has one canonical technique, and I now understand exactly why. You never push a Foucault pendulum. Any sideways nudge at release turns the straight line into an ellipse, and an ellipse precesses on its own — the fatter the oval, the faster it spins — at a rate that owes nothing to the Earth and everything to your clumsy thumb. So you draw the bob back with a loop of cotton thread, tie it to a screw in the skirting board, wait for every last shiver to die, and burn the thread with a match. The bob leaves from perfect rest. Foucault did precisely this in 1851, and so did I, feeling slightly ridiculous holding a match to a string in my own stairwell.
It didn’t matter how still I waited. Within a few minutes my clean swing had bellied out into a visible ellipse, the bob scribing a thin oval instead of a line, and that oval was pirouetting far faster than any 11.7° an hour. Kamerlingh Onnes — the same man who’d later liquefy helium and stumble into superconductivity — wrote his entire 1879 doctoral thesis on this failure: a real wire has some tiny asymmetry, so the pendulum splits into two modes at fractionally different frequencies and beats between them, sliding from line to ellipse and back. My wire, my brass cone, my drilled bob: all of it slightly, honestly wrong.
The fix is almost comically cheap, and I didn’t have it built tonight. A Charron ring — a small metal ring a few centimetres below the pivot, sized so the wire just grazes its inner edge at the extreme of each swing. It rubs away the minor axis of the ellipse, the sideways component, and leaves the main swing nearly untouched, ironing the oval back toward a line on every stroke. That’s tomorrow’s brass.
What the astrolabe couldn’t do
What I keep circling back to is how completely this inverts the paper astrolabe I hung off my thumb last week. That instrument borrows the sky — you sight a real star and the geometry hands you the hour. This one borrows nothing overhead at all. Draw the curtains, kill the lights, and a lump of lead on a wire quietly insists that the planet is turning, and even at what latitude, with no star anywhere in the room.
Keeping it alive is the other half, and it’s the watch problem wearing a bigger coat. Air drag flattens a free swing in an hour or two, so museum pendulums cheat with an electromagnetic drive: a coil under the centre, pulsed for a few milliseconds as the bob passes straight overhead, topping up the energy it loses. The whole discipline is symmetry — you fire the coil only in the dead middle of the swing, balanced about the bottom, or the impulse quietly steers the plane and you’ve manufactured a precession of your own that has nothing to do with the Earth. It is the same problem an escapement solves for a hairspring: deliver energy at exactly the right instant, and deliver it so it doesn’t bias the thing you’re trying to measure.
I did not read my latitude tonight. What I mostly measured was the several ways my own hardware is crooked. But just before I gave up I set six wooden matchsticks upright in a small arc at the far end of the swing, the way the museums stand pegs to be knocked down, and I watched. For twenty minutes the ellipse fought me. Then — I think, though the amplitude was dying and I won’t swear to it — the bob clipped a matchstick one place over from where it started, on the clockwise side. The correct side, for this hemisphere. One matchstick. It could have been the wobble in my wire. It could have been the whole world turning under my stairs. I genuinely can’t tell you which, and I’m not going to pretend otherwise.