A foliot is a horizontal bar with adjustable weights on each end that acts as the timing regulator in early mechanical clocks. Medieval horologists relied on it as the only practical way to slow down a falling-weight drive into countable beats. Paired with a verge escapement and crown wheel, the foliot's rotational inertia resists each escapement impulse, swings back, and releases the next tooth — splitting time into discrete ticks. Tower clocks like the 1386 Salisbury Cathedral clock used foliots to keep canonical hours within roughly 15 minutes per day.
Foliot Interactive Calculator
Vary sliding-weight mass and radius, swing angle, and impulse torque to see the foliot tick period, inertia, beat rate, and animated escapement motion.
Equation Used
The calculator estimates one foliot beat period from rotational inertia, angular travel, and net escapement impulse torque. Moving the two weights outward increases I = 2*m*r^2, which increases tick period and slows the clock; increasing impulse torque reduces the period.
- Two identical sliding weights are placed symmetrically about the verge.
- Weight inertia dominates; foliot bar inertia and bearing friction are neglected.
- Impulse torque is the net torque delivered to the verge after escapement losses.
- Swing angle is converted from degrees to radians in the calculation.
Inside the Foliot
The foliot is a dumb-bell shaped bar mounted at the top of a vertical shaft called the verge. Two pallets project from the verge at roughly 90° to each other and engage a horizontally-mounted toothed wheel called the crown wheel. A falling weight on a rope drives the crown wheel. As one tooth pushes against the upper pallet, the verge twists, the foliot accelerates against its own inertia, and the upper pallet finally clears the tooth — at which point the lower pallet swings into the path of a tooth on the opposite side and the whole cycle reverses. That swing-stop-reverse pattern is the tick.
The period of the swing depends almost entirely on the foliot's moment of inertia and the impulse force from the crown wheel. Slide the two end weights outward and the moment of inertia goes up — the bar swings slower and the clock runs slower. Slide them inward and the clock runs faster. That is the only rate adjustment available. There is no isochronism, no natural resonance, and no restoring spring — the foliot is a pure inertial regulator, which is why it drifts with any change in driving weight, friction, or pivot wear.
If the verge pivots wear out of round, the pallets engage the crown wheel teeth at the wrong depth and the clock either stalls or races. If the crown wheel teeth are unevenly cut — and on hand-forged 14th-century clocks they often were, with tooth-to-tooth pitch errors of 2-3° — each tick lasts a different length of time and the clock loses minutes per hour rather than per day. The dominant failure mode is pivot wear at the verge bearings, because every reversal slams the verge against its supports, and the typical wrought-iron-on-iron bearing wears measurably within a few months of continuous use.
Key Components
- Foliot Bar: Horizontal iron bar, typically 300-600 mm long on tower clocks, carrying two adjustable weights. Provides the rotational inertia that sets the tick period. Sliding the weights changes the moment of inertia and is the only rate-adjustment mechanism on the clock.
- Verge: Vertical shaft connecting the foliot to the two pallets. The verge twists back and forth through roughly 80-100° per tick. Pivots at top and bottom in plain bearings — pivot diameter is typically 6-8 mm on a tower-scale build, and surface finish at these pivots dominates rate stability.
- Pallets: Two flat flags fixed to the verge at approximately 90° to each other. Each pallet alternately catches and releases a tooth on the crown wheel. Pallet-to-tooth contact angle must be close to 52° for clean release — too steep and the tooth jams, too shallow and the foliot fails to reverse.
- Crown Wheel: Horizontal toothed wheel with saw-tooth teeth aimed sideways like a crown. Driven by the going train from the falling weight. Tooth count is usually odd (often 11 or 13) so that pallet engagement alternates between top and bottom on each beat.
- Driving Weight: Suspended on a rope wrapped around the going train barrel. Provides the constant torque that powers escapement impulses. Weight magnitude directly affects rate — heavier weight means stronger impulse, faster swing, faster clock. Typical tower clock weights run 20-50 kg.
Where the Foliot Is Used
The foliot ran European public timekeeping for roughly 300 years before the pendulum replaced it. You will not find one in a new build today, but you will find them in working condition in cathedral clocks, museum collections, and reproduction projects where horologists deliberately use period-correct mechanisms. Anywhere an interpreter, restorer, or educator needs the unmistakable irregular tick of a pre-pendulum clock, the foliot is the only honest answer.
- Heritage Horology: The Salisbury Cathedral clock (1386), one of the oldest working mechanical clocks in the world, uses an iron foliot and verge escapement and is exhibited in the cathedral's north transept.
- Museum Timekeeping: The Wells Cathedral clock movement, now displayed at the Science Museum in London, retains its original foliot regulator dating from around 1392.
- Clock Reconstruction: The De Vick clock built for Charles V at the Palais de la Cité in Paris (1364) is reconstructed in working form at the Conservatoire National des Arts et Métiers using a period-accurate foliot.
- Living-History Education: Reproduction tower clocks built for sites like Castell Coch and various open-air museums use foliot regulators to demonstrate medieval timekeeping for school visits.
- Horological Research: Antiquarian Horological Society members reconstruct foliot-regulated chamber clocks to study the rate-stability behaviour of early escapements before pendulum control.
- Film and Theatre Props: Working foliot clock props built for productions set in the medieval period — the visible swinging bar reads correctly to camera and audibly ticks at roughly 30-60 BPM, much slower than a pendulum clock.
The Formula Behind the Foliot
The tick period of a foliot is set by the balance between the impulse torque from the crown wheel and the rotational inertia of the foliot bar with its end weights. There is no resonant frequency to lock onto — the foliot just swings until friction and the next tooth engagement stop it. At the low end of the practical adjustment range, with the weights pulled all the way in, the moment of inertia drops and the clock might tick once per second. Slide the weights all the way out and the period stretches to 4 seconds or more. The horological sweet spot for a tower-scale foliot sits around a 2-second tick period — long enough for a clean swing-and-reverse, short enough that small impulse variations average out across an hour.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| T | Tick period (one beat, half a full swing cycle) | s | s |
| I | Moment of inertia of foliot bar plus end weights about the verge axis | kg·m<sup>2</sup> | lb·ft<sup>2</sup> |
| θ | Angular swing per beat (typically 0.7-1.7 rad) | rad | rad |
| Mimp | Net impulse torque from crown wheel tooth on pallet | N·m | ft·lbf |
Worked Example: Foliot in a reconstructed medieval tower clock
A horological restorer in Ghent is rebuilding the foliot regulator on a 15th-century tower clock movement that originally hung in a Flemish belfry. The foliot bar measures 500 mm tip-to-tip and carries two 0.6 kg adjustable weights. The driving weight delivers an effective impulse torque of 0.25 N·m at the verge, and the design swing per beat is 1.2 rad (about 69°). The restorer needs to predict the tick period and check whether sliding the end weights between 100 mm and 240 mm from the verge axis covers a usable range.
Given
- Lweight = 0.170 m (nominal radial position of end weights)
- mweight = 0.6 kg each (×2)
- θ = 1.2 rad
- Mimp = 0.25 N·m
Solution
Step 1 — compute the foliot's moment of inertia at the nominal weight position of 170 mm. Treating the end masses as point masses dominates over the bar itself:
Step 2 — apply the tick-period formula at nominal:
That is a tick a bit faster than once per second — a brisk, steady beat that an observer hears as a clear swing-and-reverse, very different from the rapid ticking of a pendulum mantle clock.
Step 3 — at the low end of the adjustment range, the weights slide inward to 100 mm. Moment of inertia drops sharply:
At this setting the clock runs roughly 70% faster than nominal — a hummingbird tick by foliot standards, and the impulse-to-inertia ratio gets so high that the foliot tends to overshoot and slap its bearings. Step 4 — at the high end, weights slid out to 240 mm:
This is the sweet spot for a tower clock — a slow, deliberate 1.6 s beat that matches the visual cadence of the Salisbury and Wells movements. Push the weights any further out and the impulse can no longer reliably reverse the bar before the next tooth catches, and the clock stalls.
Result
The nominal tick period at 170 mm weight position is approximately 1. 15 s — fast for a tower foliot, slow for a domestic clock, and audibly distinct as a separate swing each way. Across the practical adjustment range the period varies from roughly 0.68 s at the inner stop to 1.63 s at the outer stop, with the genuine horological sweet spot near the outer end where the bar swings deliberately and impulse variation averages out. If the restorer measures a period that drifts by more than 5% over an hour, the most likely causes are: (1) verge pivot wear opening up the bearing clearance and changing pallet engagement depth, (2) crown wheel tooth pitch error from hand-filed teeth varying by 2-3°, or (3) a frayed or stretching driving rope reducing impulse torque as the weight descends.
Choosing the Foliot: Pros and Cons
The foliot competed with nothing during its first three centuries because nothing else existed. Once Huygens applied the pendulum to a clock in 1656, the foliot became obsolete for accuracy work overnight. Today the comparison matters when you are deciding whether to build a period-accurate reproduction or to fit a more modern regulator into an old movement.
| Property | Foliot + Verge | Pendulum + Anchor Escapement | Balance Wheel + Lever Escapement |
|---|---|---|---|
| Typical accuracy (drift per day) | ±15 minutes | ±10 seconds | ±5 seconds |
| Tick period range | 0.5-4 s | 0.5-2 s | 0.1-0.4 s |
| Sensitivity to driving-weight variation | High — rate changes with weight | Low — pendulum is isochronous | Very low — spring-regulated |
| Pivot wear interval before re-bushing | 3-12 months continuous | 5-20 years | 5-10 years |
| Mechanical complexity | Lowest — 4 moving parts | Moderate — pendulum, anchor, crutch | Highest — escape wheel, lever, balance, hairspring |
| Suitable for portable use | No — gravity-independent but motion-sensitive | No — needs vertical mount | Yes — wristwatches and marine chronometers |
| Period authenticity for pre-1660 restoration | Required | Anachronistic | Anachronistic |
Frequently Asked Questions About Foliot
The foliot has no isochronism — its rate is directly tied to the impulse torque from the crown wheel. A freshly wound rope sits at the largest effective drum diameter and delivers slightly more torque, which kicks the foliot through its swing faster. As the rope unwinds, drum diameter decreases and friction in the going train accumulates, so impulse drops and the swing slows.
The standard fix on medieval clocks was to add a stackfreed or, later, a fusee — neither of which are period-correct for a 14th-century build. If you must keep it authentic, accept the 5-10% rate drift across the wind cycle and rewind every 12 hours rather than every 24.
Run a quick impulse-vs-inertia check. If the foliot completes its swing in less than about 0.4 s the bar slaps its bearing stops because the impulse overpowers the inertia, and you'll hear a metallic clack rather than a clean tick. The foliot must be long enough — or the end weights heavy enough — that the period at maximum-inward weight position is at least 0.5 s.
Rule of thumb for a tower-scale build: with a crown wheel impulse torque around 0.25 N·m, you want the bar at least 400 mm tip-to-tip carrying at least 0.5 kg per end.
If the clock is going behind glass in a museum and accuracy matters, retrofit a short pendulum with an anchor escapement — you'll cut daily drift from 15 minutes to under 30 seconds. If the clock is for interpretation, demonstration, or any context where period authenticity is the point, the foliot is the only correct answer. A pendulum on a clock dated before 1657 is an anachronism a knowledgeable visitor will spot immediately.
One compromise: build the foliot but design the verge so a pendulum can be swapped in for long-term display, with the foliot reinstalled for demonstrations.
Asymmetric swing almost always means the two pallets are not at exactly 90° to each other on the verge, or the crown wheel axis is not square to the verge axis. On hand-forged movements the pallets were filed individually and small angle errors of 3-5° are common. The pallet that catches the tooth at a steeper angle gets a stronger impulse and drives a wider swing on that side.
Diagnose by counting the swing arc on each side with a protractor card behind the foliot. If one side is more than 10° wider than the other, reset the lower pallet — the upper one is usually less worn and a better reference.
No — and this is the most common mistake first-time foliot builders make. Heavier driving weight increases impulse torque, which makes the foliot swing faster, not more accurately. The foliot's drift comes from the absence of a natural resonant period, not from insufficient drive.
The only way to improve foliot rate stability is to keep the impulse torque as constant as possible across the wind cycle (good rope, clean train, low pivot friction) and to minimise pivot wear. Even a perfectly maintained foliot will not beat ±5 minutes per day — that's the physics, not the build quality.
The classic period-correct pairing is hardened steel verge pivots running in a wrought-iron frame, which wears noticeably within months under continuous running because the verge reverses direction roughly 60,000 times per day. For a working museum installation that needs to run year-round, brass bushings pressed into the iron frame multiply life by 5-10×, and the substitution is invisible from the visitor side.
If you measure the pivot diameter and find it has worn out-of-round by more than 0.1 mm, the pallet engagement depth has shifted enough to change the rate by several minutes per hour — re-bush before that point.
References & Further Reading
- Wikipedia contributors. Verge escapement. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
