The Verge Escapement is the oldest mechanical clock escapement, using a vertical staff carrying two flag-shaped pallets that engage alternately with the axially-pointing teeth of a crown wheel. You see it in surviving 14th-century tower clocks like the Salisbury Cathedral clock and in pre-1700s verge pocket watches. Its job is to release the going train one tooth at a time while feeding impulses to a foliot or balance, converting steady gear-train torque into the timed beats that drive the hands. The outcome is a self-regulating clock — crude by modern standards at roughly ±15 minutes per day, but reliable enough to run continuously for 600 years.
Verge Escapement Interactive Calculator
Vary pallet angle, crown-wheel tooth count, and wheel speed to see beat rate, impulse arc, and lock-up margin.
Equation Used
The calculator uses the article's key verge-escapement geometry: a 90 deg pallet separation gives a short symmetric impulse, about 100 deg gives stronger action, and more than about 105 deg risks lock-up. Beat rate is estimated by releasing one crown-wheel tooth per beat.
- One crown-wheel tooth is released per beat.
- Pallet angle guidance follows the article: 95 to 100 deg preferred, lock-up above about 105 deg.
- Impulse arc is treated as the geometric opening beyond the 90 deg symmetric pallet condition.
- This is a simplified escapement geometry calculator, not a full clock rate or friction model.
The Verge Escapement in Action
The Verge Escapement, also called the Crown Tooth Escapement because of the way the escape-wheel teeth point along the axis like points on a crown, works by trapping rotational energy from the going train and releasing it in discrete beats. The crown wheel sits horizontally in a tower clock or vertically in a watch, and its teeth are cut on the face — not the rim. A vertical shaft, the verge or pallet staff, carries two small flag-like pallets set roughly 90° to 105° apart around the staff. As the crown wheel turns, one pallet catches a tooth, stops the wheel, then gets pushed aside as the staff rocks — which swings the second pallet directly into the path of the next tooth on the opposite side of the wheel. That oscillation is what you hear as the tick.
The pallet angle is the critical spec. If the two pallets sit at exactly 90° the action is symmetric but the impulse arc is short, so the foliot or balance gets only a weak push. Open the angle to about 100° and you get a longer impulse arc and a stronger swing — but go past 105° and the second pallet starts contacting the next tooth before the first has fully released, and the clock locks up. The classic Verge escapement (crown-wheel) tolerance from English clockmakers like Thomas Tompion's workshop ran 95° to 100°, no looser.
This is a recoil escapement, meaning the escape wheel actually rotates backwards a small amount on every beat as the foliot's inertia pushes the pallet back into the tooth. You can see it on the second hand of a verge clock — it visibly stutters backwards before each tick. That recoil is the verge's defining weakness: it couples the foliot's swing amplitude directly back into the going train, so any change in driving torque (a fouled fusee, a dry pivot, a dirty pinion) immediately changes the rate. Common failure modes are pallet wear at the leading edge — a verge pallet typically loses 0.1 to 0.3 mm of working face over a century — and crown-wheel tooth tip rounding, which together open the drop angle and slow the clock by several minutes per day.
Key Components
- Crown Wheel (Escape Wheel): The driven wheel with teeth cut axially on its face, resembling a crown. Tooth count is almost always odd — 11, 13, 15, or 17 — because the two pallets must engage alternate sides of the wheel and an even count would let both pallets meet teeth simultaneously. Tooth pitch typically runs 8°–12° at the working radius.
- Verge (Pallet Staff): A vertical steel shaft carrying the two pallets and, at one end, the foliot bar or balance wheel. Staff diameter in a watch verge runs 0.6–1.0 mm; in a tower clock it can be 15–25 mm. Pivots must be hardened and polished to Ra ≤ 0.4 µm or pivot drag dominates the energy budget.
- Pallets (Flags): Two small flat flags, forged or screwed onto the verge, set 95°–100° apart around the staff axis. Working faces are hardened steel ground flat to within 0.05 mm. Pallet wear is the single biggest cause of long-term rate drift.
- Foliot or Balance: The timekeeping element fixed to the top of the verge. Early tower clocks used a foliot — a horizontal bar with sliding weights to tune the period. Later verge watches used a circular balance wheel with a pre-hairspring inertia of around 5–15 g·mm². Period is set by moving the weights inward (faster) or outward (slower).
- Contrate Wheel (in watches): A wheel with teeth cut at 90° on its face, used in pocket watches to transfer drive from the horizontal third wheel to the vertical crown wheel. Not present in tower clocks where the crown wheel is driven directly by a horizontal arbor.
Who Uses the Verge Escapement
The Verge Escapement dominated mechanical timekeeping from roughly 1275 to 1675, and persisted in cheap pocket watches until about 1850. You will not specify one for a new design today, but if you restore antique clocks, build educational kinetic pieces, or work in a museum conservation lab, you need to know how it behaves. Below are the real machines and contexts where you will encounter it.
- Heritage tower clock restoration: The Salisbury Cathedral clock (c. 1386) and the Wells Cathedral clock both run on iron foliot-and-verge mechanisms — restorers commission verge components from specialists like Smith of Derby for in-situ rebuilds.
- Antique watch restoration: English verge fusee pocket watches by makers like Thomas Mudge, John Arnold's early shop output, and Liverpool trade-watch builders from 1780°—1840 still come through restoration benches needing crown-wheel and pallet renewal.
- Museum kinetic exhibits: The Science Museum London and Musée des Arts et Métiers in Paris display working foliot-verge demonstration clocks built specifically to show the mechanism's recoil behaviour to visitors.
- Horological education: BHI (British Horological Institute) and WOSTEP courses include hand-filing a verge and crown-wheel pair as a foundational skill exercise — students build a working assembly to ±5 minutes/day to pass the module.
- Replica and reproduction clockmaking: Companies like Erwin Sattler and small workshops producing De Vick-style replica Gothic chamber clocks build new verge escapements to original 14th-century geometry.
- Kinetic sculpture and art clocks: Artists building visible-mechanism wall clocks deliberately specify a verge for the audible tick and visible recoil — the foliot's slow side-to-side swing is what makes the piece read as a clock from across a gallery.
The Formula Behind the Verge Escapement
The single most useful number when you sit down with a verge clock is the beat period — how long one tick-tock cycle takes. That tells you what rate the going train should produce and lets you check whether the foliot weights are positioned correctly. The period depends on the foliot's moment of inertia, the impulse energy delivered per beat, and the swing amplitude. At the low end of the typical operating range — a heavy tower-clock foliot with weights set fully outboard — you'll see periods of 4 to 6 seconds per beat. At the nominal middle — a domestic chamber clock foliot — periods sit around 1 to 2 seconds. At the high end — a small verge watch balance — you're down to around 0.25 seconds per beat. The sweet spot for a foliot clock is wherever the swing amplitude lands between 40° and 60° of arc, because below 40° the impulse barely overcomes pivot friction and above 60° the pallets start banking on the tooth flanks instead of the tips.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| T | Period of one full oscillation (tick + tock) | s | s |
| I | Moment of inertia of the foliot or balance about the verge axis | kg·m² | lb·in² |
| M | Combined mass of the foliot weights | kg | lb |
| g | Gravitational acceleration (only relevant when foliot is horizontal-bar style and impulse comes from train torque, not from a restoring gravity term — for a true torque-driven foliot this term is replaced by the train torque τ; the form shown is for hairspring-balance verges where gravity term becomes spring constant k) | m/s² | ft/s² |
| r | Effective radius from verge axis to weight centre of mass | m | in |
Worked Example: Verge Escapement in a 1450-style replica Gothic chamber clock
A small instrument shop in Vienna is building a working replica of a mid-15th-century Gothic chamber clock for a private collector. The foliot is a 320 mm steel bar with two 180 g sliding weights, and the crown wheel has 13 teeth. The driving weight produces a steady torque of 0.04 N·m at the verge. The customer wants the clock to beat once per second so it can be set against a modern reference. We need to find the foliot weight position that gives a 1.0 s beat at nominal swing, and then check what happens at the extremes of the foliot adjustment range.
Given
- M = 0.36 (2 × 0.180) kg
- τ = 0.04 N·m
- Crown wheel teeth = 13 —
- Target beat period = 1.0 s
- Swing amplitude (nominal) = 50 °
Solution
Step 1 — for a torque-driven foliot, the period approximates T = 2π × √(I / τ × θ), where θ is the swing amplitude in radians. Convert 50° to radians:
Step 2 — solve for the required moment of inertia at nominal 1.0 s period:
Step 3 — for two point-mass weights of 0.180 kg each on a bar, I = 2 × m × r2. Solve for the weight radius:
Step 4 — at the low end of the adjustment range, the customer slides the weights all the way in to r = 25 mm. Period drops to:
That's a fast tick — twice per second — and you'll hear the clock running roughly 4 hours fast per day if left there. At the high end of adjustment, weights at r = 150 mm (near the bar ends):
Three seconds per beat is the slow, lazy swing you see on Salisbury-style tower clocks — the foliot drifts visibly across the field of view. For our 1.0 s target, the 50 mm position is correct, and the customer has roughly ±20 mm of practical adjustment range before the swing amplitude itself starts collapsing or running away.
Result
The nominal weight position is 50 mm from the verge centre to give a 1. 0 s beat. That feels right on a domestic chamber clock — a clean, audible tick once per second, with the foliot sweeping a visible 50° arc you can watch from across the room. At 25 mm the clock runs at twice the speed (0.5 s beat) and at 150 mm it crawls at one beat every 3 seconds, so the practical adjustment band sits roughly between 35 mm and 75 mm before the impulse-amplitude relationship breaks down. If you build this and measure 1.3 s instead of the predicted 1.0 s, the most likely causes are: (1) crown-wheel pivots running dry — verge pivots without oil can add 30–40% to the effective inertia through stiction, (2) foliot bar not perpendicular to the verge axis, which couples gravity into the swing and slows it asymmetrically, or (3) pallet angle filed wider than 100°, lengthening the impulse arc but burning extra energy in drop. Check pivot oil first, then square the bar with an engineer's square against the verge.
Verge Escapement vs Alternatives
The Verge Escapement is the baseline against which every later escapement is measured. It works, it's simple to make with hand tools, and it tolerates dirt better than any precision escapement — but its accuracy is poor and its recoil wastes energy. Here's how it stacks up against the two escapements that replaced it.
| Property | Verge Escapement | Anchor Escapement | Deadbeat Escapement |
|---|---|---|---|
| Typical accuracy (per day) | ±15 minutes (foliot) to ±2 minutes (balance-spring verge) | ±10–30 seconds | ±1–5 seconds |
| Recoil behaviour | Heavy recoil — escape wheel reverses every beat | Light recoil | No recoil (deadbeat) |
| Beat rate range | 0.25–6 s per beat | 0.5–2 s per beat | 0.5–2 s per beat |
| Manufacturing complexity | Low — hand-filable | Medium — anchor pallets need careful angle | High — pallet faces need precision lapping |
| Tolerance to dirt and wear | Excellent — runs even when filthy | Moderate | Poor — sensitive to pallet condition |
| Power consumption from train | High (recoil dissipates energy) | Medium | Low |
| Typical lifespan before pallet refacing | 80–150 years | 30–60 years | 20–50 years |
| Best fit application | Heritage restoration, replicas, kinetic art | Domestic longcase and wall clocks | Regulators, precision clocks |
Frequently Asked Questions About Verge Escapement
This is position error, and it's a verge-specific problem. The verge runs vertically when the watch is in the pocket, so gravity loads the lower verge pivot harder than the upper one. That asymmetric pivot friction couples into the impulse and shortens the effective swing on one side, which speeds the rate. On the bench laid flat, both pivots share load evenly.
The fix is to check the lower pivot's endshake — it should be 0.02–0.05 mm. Anything tighter and the pivot binds under gravity load. Many restorers also lightly polish the lower pivot end to reduce thrust friction.
Use the verge if the original was verge — anything before roughly 1670 was verge by definition because the anchor wasn't invented yet. A William Clement-style anchor in a Gothic chamber clock replica is a historical error a knowledgeable buyer will spot immediately.
If accuracy matters more than authenticity (a working museum piece used as a reference rather than a display), the anchor will hold ±30 seconds per day vs the verge's ±15 minutes. But you lose the visible recoil and the slow, characteristic foliot swing that makes the piece read as 14th-century.
Geometry. The two pallets on the verge engage alternate sides of the crown wheel, separated by roughly 100° around the verge axis. If the tooth count is even, the two pallets can find a position where both contact a tooth simultaneously — the wheel locks and the clock stops dead.
With an odd count (11, 13, 15, 17 are standard), the teeth are always offset across the wheel diameter, so when one pallet is engaged the other is guaranteed to be in clear air. If you're cutting a replacement crown wheel and the original was lost, default to 13 teeth for tower work and 15 for watches unless you have a documented reason to deviate.
Very. The verge is a recoil escapement, so the foliot's swing amplitude is set directly by the impulse it receives, which is set by the train torque, which is set by the driving weight. Increase the weight by 20% and the swing amplitude rises, the foliot reaches further on each side, and the period actually shortens — the clock speeds up. This is the opposite of how a pendulum clock behaves and it catches restorers out.
Rule of thumb: a verge clock's rate changes by roughly 1–2 minutes per day per 10% change in driving weight. Always note the exact weight value when you set the rate, and never substitute a different weight without re-timing.
Almost always pallet-to-tooth drop angle that's too tight. As the verge warms slightly from running friction, the pallet steel expands by a few microns. If the original drop angle was below about 1.5° you've now got zero drop, the next tooth catches the pallet face before it has cleared, and the clock stalls.
Diagnostic check: stop the clock cold, manually rock the verge through a full beat, and watch the gap between the trailing edge of the released tooth and the leading edge of the next pallet. You want to see daylight — at least 1.5° of free rotation. If the gap is invisible, the pallets need re-shaping or the crown wheel needs depthing further from the verge.
Mechanically yes, but historically and practically it's a bad idea. The foliot has no restoring force of its own — its period is set by inertia and impulse alone. Adding a balance spring turns it into a sprung balance verge, which improves accuracy from ±15 min/day to perhaps ±2 min/day, but you've now built something that isn't authentic to any period and the foliot weights become almost decorative.
If you need the accuracy, replace the entire escapement with an anchor — it's the historically correct upgrade path because that's what clockmakers actually did in the 1670s. If authenticity matters, leave the foliot alone and accept the rate.
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.
