Proper Time and Four-Velocity

The previous article built a scalar-manufacturing machine and issued a warning: quantities with a secretly transforming denominator are impostors. The most important casualty is ordinary velocity. The quadruple $(c,\,d\mathbf{x}/dt)$ looks ready-made for relativity, but $dt$ is one frame's coordinate time — divide a four-vector displacement by it and the result transforms like nothing useful at all. Before electrodynamics can put charges in motion covariantly, it needs a clock that all observers agree on. Relativity supplies exactly one candidate: the wristwatch of the moving object itself. That wristwatch reading is proper time $\tau$ — the arc length of a worldline, measured by the Minkowski metric, and as frame-invariant as any interval. This article makes it operational: how to compute $\tau$ along an arbitrary accelerated trajectory (chop the worldline into instants, borrow the momentarily comoving inertial frame for each, and integrate $dt/\gamma$ ), and what physical fine print the recipe carries — the clock hypothesis, tested by muons circulating at $10^{18}\,g$ in a storage ring and dilating by pure $\gamma$ anyway. With $\tau$ in hand, one division builds the first physical four-vector of the chapter: the four-velocity $u^\mu=dx^\mu/d\tau=\gamma(c,\mathbf{v})$ , the unit tangent to the worldline. Its mechanics-track companion (Four-Vectors and Minkowski Spacetime) owns the normalization theorem $u^\mu u_\mu=c^2$ ; our thread is the ladder principle — that dividing by $d\tau$ is the one legal way to differentiate along a trajectory, the move this chapter will repeat verbatim to build the four-momentum, the four-current of a charged flow, and eventually the covariant Lorentz force. Master the wristwatch and the tangent vector here; the rest of the ladder is assembly.

Key takeaways

Proper time is the invariant arc length of a worldline: $c^2d\tau^2=\eta_{\mu\nu}dx^\mu dx^\nu$ , so $d\tau=dt/\gamma(t)$ — every observer computes the same $\tau$ for a given traveler, using their own $t$ and $\gamma$ .
For accelerated motion, integrate: $\tau=\int dt/\gamma(t)$ . The clock hypothesis — that only instantaneous speed matters, not acceleration — is physics, not definition, and is confirmed by storage-ring muons at $10^{18}\,g$ .
The four-velocity $u^\mu=dx^\mu/d\tau=\gamma(c,\mathbf{v})$ is a certified four-vector: a four-vector displacement divided by an invariant tick.
Its norm is fixed: $u^\mu u_\mu=c^2$ always — a unit tangent in the metric's sense; four-velocities have no "magnitude," only direction in spacetime.
The ladder principle: $d/d\tau$ is the only covariant time derivative. Every relativistic quantity of this chapter is built as (scalar) × (four-vector) or $d(\text{four-vector})/d\tau$ .
At $v\ll c$ : $u^\mu\to(c,\mathbf{v})$ and $\tau\to t$ — Newtonian kinematics is the low-speed shadow of the four-velocity.
Key idea: the moving object's own wristwatch is the unique invariant clock, and differentiating the worldline by it produces the four-velocity — the template for every covariant "rate" in electrodynamics.

The core insight

Proper time — the invariant interval accumulated along a worldline, $d\tau=dt/\gamma$ — is the only frame-independent parameter a trajectory carries, and the four-velocity $u^\mu=dx^\mu/d\tau=\gamma(c,\mathbf{v})$ , with fixed norm $u^\mu u_\mu=c^2$ , is the covariant rate of motion built from it.

Where this lives today

Proper time is billable engineering. GPS satellite clocks run fast relative to ground clocks by about 38 μs/day net — a special-relativistic slowdown of $\sim$ 7 μs/day from orbital speed (this article's $\int dt/\gamma$ , verbatim) beaten by a general-relativistic speedup from altitude — and the constellation's firmware corrects for both; uncorrected, positions would drift by ten kilometers a day. On the ISS ( $v=7.66$ km/s), astronauts age about 5 ms less per six months than the ground — computed with (2) in one line. Meanwhile optical clocks have made the wristwatch effect a laboratory commodity: modern lattice clocks resolve the time dilation of walking pace and the gravitational shift of a few centimeters of height, turning $d\tau$ comparisons into a proposed geodesy instrument. In accelerator physics the four-velocity is the working coordinate. Beam-dynamics codes propagate particles in proper time or its accelerator cousins, radiative lifetimes of unstable secondaries set shielding and detector geometry through $\gamma\tau_0$ (a 40 GeV pion beam travels hundreds of meters; its rest-frame twin would die in nanometers), and the "magic momentum" $\gamma=29.3$ of muon $g-2$ rings is chosen so that electric-field corrections to the spin equation cancel — a condition written, of course, in four-velocity contractions. The ladder this article started climbs immediately: multiply $u^\mu$ by the rest mass and you have relativistic dynamics (next article); multiply by rest charge density and you have the covariant sources of the field tensor (three articles on). One derivative, two multiplications, and electrodynamics goes covariant.

Origins · 2 minPaul Langevin and how this was discovered

Paul Langevin

1872–1946 · Paris — who made proper time a story physics could not forget

Langevin owns the wristwatch. At the 1911 Bologna philosophy congress he dramatized the asymmetry of proper time as no equation had: a traveler leaves Earth in a projectile at near-light speed, explores for what his own clocks call two years, and returns to find Earth two hundred years older — "a concrete example of the relativity of time" that made worldline arc length viscerally real (and seeded a century of twin-paradox pedagogy). The choice of messenger was apt: Langevin was an electrodynamicist — his doctorate under Pierre Curie led to the theory of para- and diamagnetism via electron orbits (the Langevin function of the magnetism chapter is his), his analysis of ionized gases founded ion-mobility physics, and in WWI he built the first powered sonar from piezoelectric quartz. He grasped relativity's radicalism early and lectured it at the Collège de France when much of the French establishment was still skeptical; Einstein said of him that he would "certainly have developed special relativity had it not been done elsewhere." Politically fearless — Dreyfusard, antifascist, dismissed and confined by the Vichy regime — he lived to be restored to his chair in 1944. The traveler who returns young carries his name in the literature: the Langevin twin.

History

1905

— Einstein's kinematics contains time dilation for moving clocks; the accumulated-time-along-a-path reading is implicit

1908

— Minkowski's geometry identifies the invariant interval; "proper time" (Eigenzeit) is coined as the arc length of a worldline

1911

— Langevin's Bologna traveler dramatizes worldline-dependent aging; proper time enters physical folklore

1941

— Rossi and Hall measure atmospheric-muon survival: cosmic-ray clocks dilate by γ, the first quantitative wristwatch test

1971

— Hafele and Keating fly cesium clocks around the world; east–west asymmetry matches the proper-time integral (plus gravity)

1977

— CERN Muon Storage Ring: lifetimes at γ = 29.3 under 10¹⁸ g confirm pure-γ dilation — the clock hypothesis passes at parts per thousand

1978

— GPS prototype clocks fly with relativistic corrections designed in; proper time becomes navigation infrastructure

2010

— NIST optical clocks resolve time dilation at 10 m/s and gravitational shift at 33 cm: wristwatch physics on a tabletop

The physical picture

Picture a worldline in a spacetime diagram — a muon's, say, curving through the chapter's coordinate grid. Different inertial observers slice that picture with different simultaneity planes and disagree about how much coordinate time elapses between two marked events on the curve. But ride along the curve itself and there is a fact no slicing can touch: the muon's internal clock — any physical process it carries, its own decay probability included — ticks off a definite amount between those events. That reading is proper time. Geometrically it is the length of the worldline segment, measured with the Minkowski ruler: relativity turns "how much did you age?" into "how long is your path?", with one crucial inversion. Because the metric weighs time positively and space negatively, detours reduce arc length: among all worldlines connecting two events, the straight (inertial) one is the longest in proper time. Wandering costs youth — the reverse of Euclidean intuition, and (as the mechanics companion Relativistic Kinematics proves in its twin-asymmetry derivation) the entire resolution of the twin story. Why must the wristwatch be the chapter's official clock? Because it is the only one nature attaches to the trajectory itself. Coordinate time belongs to an observer, and there are infinitely many observers; $\tau$ belongs to the worldline, and there is exactly one worldline. When we need to differentiate along the motion — to ask how fast the position four-vector, or soon the momentum, changes — dividing by $d\tau$ is the unique choice that keeps four-vector status intact, because $d\tau$ is a scalar. Dividing by $dt$ ruins it, since $dt$ is one frame's opinion. The object this produces, $u^\mu=dx^\mu/d\tau$ , deserves its geometric reading before its components. It is the unit tangent vector to the worldline: at every event it points along the traveler's instantaneous direction through spacetime — mostly time for a dawdler, tilted hard toward space for a relativist — and its Minkowski length is always the same $c$ . A four-velocity cannot be "large" or "small"; it can only point. Speeding up does not stretch the tangent, it rotates it (hyperbolically) toward the light cone, which it may approach but never reach. All the drama of relativistic kinematics is in that rotation.

Figure

Proper time is arc length — two worldlines between the same events, and the wristwatch reading each accumulates

The math: the invariant clock, then the tangent vector

Start where the metric left off. For two infinitesimally separated events on a timelike worldline (the only kind matter traces), the interval is positive, and we name its square root suggestively:

c^2\,d\tau^2\;=\;\eta_{\mu\nu}\,dx^\mu dx^\nu\;=\;c^2dt^2-|d\mathbf{x}|^2. \qquad (1)

where: $d\tau$ is by construction a Lorentz scalar (it is built from the invariant interval), with dimensions of time. Its physical identity is fixed by evaluating (1) in the momentarily comoving frame — the inertial frame in which the object is instantaneously at rest: there $d\mathbf{x}=0$ , so $d\tau=dt_{\text{rest}}$ . The invariant is the rest-frame tick: the wristwatch. Dividing (1) by $dt^2$ in an arbitrary frame gives the working formula

d\tau=dt\sqrt{1-\frac{v^2(t)}{c^2}}=\frac{dt}{\gamma(t)}, \qquad (2)

where: $v(t)$ is the object's speed in that frame at that instant, and $\gamma(t)$ the corresponding Lorentz factor. Read it in both directions: the moving clock accumulates less time than the coordinate clock ( $d\tau<dt$ whenever $v\ne0$ — time dilation, whose light-clock derivation lives with the mechanics companion Relativistic Kinematics); and, less famously but just as deep, any observer — using their own $t$ , their own measured $v(t)$ — computes the same $\tau$ for the traveler, because the left side is a scalar. Proper time is the traveler's private experience and public property simultaneously.

Derivation AProper time along an arbitrary accelerated worldline

The goal is to make (2) legitimate for real trajectories — curved, accelerated, circulating — not just uniform motion: the total proper time along a worldline is $\tau=\int_{t_1}^{t_2}dt/\gamma(t)$ , an invariant functional of the path, and evaluating it for a storage-ring muon gives a laboratory-tested number. The subtlety is honest physics: the interval formula (1) was established between events as seen from inertial frames, and an accelerating object is at rest in no inertial frame for longer than an instant. The strategy — the standard one, but worth executing carefully because the whole chapter's dynamics will lean on it — is to chop the worldline into infinitesimal segments, borrow a different momentarily comoving inertial frame for each, and add up the verdicts. What licenses the addition is an experimental input called the clock hypothesis, flagged explicitly in Step 3.

Step 1. Chop. Partition the trajectory between events $P$ and $Q$ into segments of coordinate duration $dt$ , short enough that within each the velocity $\mathbf{v}(t)$ is constant to first order. Each segment is a straight sliver of worldline.

Step 2. Borrow a frame per sliver. For the sliver at time $t$ , the momentarily comoving inertial frame (MCIF) is the inertial frame with velocity exactly $\mathbf{v}(t)$ . Within the sliver the object is at rest in its MCIF, so the sliver's invariant interval — computable in any frame, per (1) — equals the rest-frame tick:

d\tau\big|_{\text{sliver}}=\frac{dt}{\gamma(t)},\qquad \gamma(t)=\Big(1-\frac{v^2(t)}{c^2}\Big)^{-1/2}.

Why this is frame-honest: each sliver's $d\tau$ is a scalar, so the bookkeeping frame (the lab) and the borrowed frame (the MCIF) agree on it — the lab never has to accelerate, and no non-inertial frame is ever used for more than a labeling instant.

Step 3. Add, and declare the physics. Summing slivers:

\tau_{PQ}=\int_{t_1}^{t_2}\frac{dt}{\gamma(t)}\;\le\;t_2-t_1,

with equality only if the object never moves in this frame. The addition assumes the sliver ticks depend on nothing but the instantaneous velocity — that acceleration itself does not additionally slow or hasten a clock. This is the clock hypothesis: an assumption about real clocks, not a theorem of geometry. It is superbly supported: muons circulating in the CERN Muon Storage Ring, with proper (rest-frame) centripetal acceleration $\gamma^2v^2/r\approx1.1\times10^{19}\ \text{m/s}^2\approx10^{18}\,g$ , were measured to dilate by pure $\gamma$ — lifetimes agreed with $\gamma\tau_0$ at the part-per-thousand level, acceleration contributing nothing detectable. (A clock built badly enough — a pendulum, say — fails the hypothesis trivially; the statement is about clocks whose mechanism is fast and stiff compared to the motion, which includes atomic clocks and particle decay.)

Step 4. Instantiate for circular motion — the case electrodynamics cares about, since magnetic fields steer charges in circles. At constant speed, $\gamma$ is constant even though the velocity vector rotates constantly:

\tau=\frac{t_{\text{lab}}}{\gamma}\qquad\text{exactly, for any number of turns.}

For the storage-ring muon at $\gamma=29.3$ : proper lifetime $\tau_0=2.197\ \mu$ s, laboratory lifetime $\gamma\tau_0=64.4\ \mu$ s, in which time it travels $19.3$ km — 439 laps of the $r=7$ m ring. One integral, three orders of magnitude of acceleration beyond anything engineered, and the wristwatch formula holds. Verified: $\gamma\tau_0=29.3\times2.1970\ \mu\text{s}=64.37\ \mu$ s, matching the CERN measurement of the dilated lifetime; path $=\beta c\,\gamma\tau_0=0.99942\times(3.00\times10^8)\times64.37\ \mu\text{s}=19.3$ km; laps $=19286/(2\pi\times7)=439$ ; proper acceleration $\gamma^2v^2/r=(29.3)^2(0.99942c)^2/7\ \text{m}=1.10\times10^{19}\ \text{m/s}^2=1.12\times10^{18}\,g$ . Matches Bailey et al. (1977) and Griffiths §12.1.2 conventions.

Now the division that starts the four-vector ladder. The displacement $dx^\mu$ is a certified four-vector; $d\tau$ is a certified scalar; their quotient is therefore a four-vector with no further checking — this is the entire construction:

u^\mu\equiv\frac{dx^\mu}{d\tau}=\gamma\frac{dx^\mu}{dt}=\gamma\,(c,\,\mathbf{v}). \qquad (3)

where: the chain rule $d/d\tau=\gamma\,d/dt$ converts to laboratory bookkeeping, and $\mathbf{v}=d\mathbf{x}/dt$ is the ordinary coordinate velocity. Note what the $\gamma$ does to each slot: the time component $u^0=\gamma c$ is the rate at which coordinate time passes per unit wristwatch time (times $c$ ) — dilation itself, packaged as a component; the spatial part $\gamma\mathbf{v}$ exceeds the coordinate velocity and is unbounded as $v\to c$ , which is the first hint of why momentum, built on $u^\mu$ in the next article, can grow without limit while speed cannot. The norm is fixed once and forever,

u^\mu u_\mu=\gamma^2(c^2-v^2)=c^2, \qquad (4)

where: this one-line check in the lab frame agrees, as it must, with the effortless rest-frame evaluation $u^\mu=(c,\mathbf{0})$ ; the full theorem — including its differential consequence, that the four-acceleration $a^\mu=du^\mu/d\tau$ is always Minkowski-orthogonal to $u^\mu$ — is owned by the mechanics companion Four-Vectors and Minkowski Spacetime, and we quote it rather than re-derive it. Geometrically (4) says $u^\mu$ is a unit tangent (in units of $c$ ): every worldline through spacetime is traversed at the same "four-speed," and kinematic variety lives entirely in the tangent's direction.

Go deeper · Derivation B · 2 minWhy d by d-tau is the only covariant rate — the ladder principle

Derivation BWhy d by d-tau is the only covariant rate — the ladder principle

The goal is the construction principle the rest of the chapter uses as a reflex: differentiation of any four-vector by proper time yields a four-vector, while differentiation by coordinate time does not — so covariant "rates" must be built with $d/d\tau$ . The strategy is to test both candidates against the transformation law directly; the failure of $d/dt$ is as instructive as the success of $d/d\tau$ , because it locates the exact spot where Galilean habits break. The result matters beyond kinematics: it is why the four-current will be $\rho_0u^\mu$ , why the equation of motion in Relativistic Dynamics of a Charged Particle will read $dp^\mu/d\tau=qF^{\mu\nu}u_\nu/c^2$ -type contractions, and why no honest relativistic law ever contains a bare $d/dt$ .

Step 1. The successful candidate. Let $V^\mu$ be any four-vector along the worldline: $V'^\mu=\Lambda^\mu{}_\nu V^\nu$ with $\Lambda$ constant. Differentiating by the scalar $\tau$ :

\frac{dV'^\mu}{d\tau}=\Lambda^\mu{}_\nu\frac{dV^\nu}{d\tau},

since $d\tau$ is the same in both frames. The derivative transforms like $V^\mu$ itself: four-vector in, four-vector out. Iterating gives the four-acceleration $a^\mu=du^\mu/d\tau$ , and later $dp^\mu/d\tau$ — all certified by this one line.

Step 2. The failed candidate, dissected. Try $d/dt$ instead:

\frac{dV'^\mu}{dt'}=\Lambda^\mu{}_\nu\frac{dV^\nu}{dt'}=\Lambda^\mu{}_\nu\frac{dV^\nu}{dt}\frac{dt}{dt'}.

The stray factor $dt/dt'$ is not a constant — for a moving object it depends on the velocity through the simultaneity mixing of the boost — so $dV^\mu/dt$ does not transform with $\Lambda$ alone. Concretely for $V^\mu=x^\mu$ : the quadruple $(c,\mathbf{v})$ built with $d/dt$ transforms with an extra velocity-dependent scale factor, which is precisely why relativistic velocity addition (owned by Lorentz Transformations) is a quotient of transformed differentials rather than a matrix product. Coordinate velocity is the ratio of two four-vector pieces measured in one frame's slicing; it belongs to the observer, not the worldline.

Step 3. The general recipe. Since scalars pass freely through $\Lambda$ , the theorem generalizes: (invariant scalar) × (four-vector) is a four-vector, and $d/d\tau$ of a four-vector is a four-vector. Every construction in the chapter's ladder is one of these two moves:

u^\mu=\frac{dx^\mu}{d\tau},\qquad p^\mu=m\,u^\mu,\qquad J^\mu=\rho_0\,u^\mu,\qquad \frac{dp^\mu}{d\tau}=\text{(force four-vector)},

with $m$ the rest mass and $\rho_0$ the rest-frame charge density — both invariant by definition of the quantity in its own rest frame, the pattern to notice now and reuse.

Step 4. The low-speed audit. As $v/c\to0$ : $\gamma\to1+v^2/2c^2$ , so $u^\mu\to(c,\mathbf{v})$ and $d/d\tau\to d/dt$ — the ladder collapses onto Newtonian kinematics smoothly, with first corrections at order $v^2/c^2$ . Relativity does not replace the old kinematics; it identifies it as the tangent vector's shadow at small tilt. Verified: Step 1–2 logic matches Griffiths §12.1.2 (his "proper velocity" discussion) and Zangwill §22.4; the $v=0.8c$ numerical audit — $u^\mu=\tfrac{5}{3}(c,\,0.8c)=(1.667c,\,1.333c)$ , norm $(1.667^2-1.333^2)c^2=c^2$ ; boosting with $\beta=0.8$ gives $u'^\mu=(c,0)$ exactly — confirms both the transformation law and the fixed norm.

Figure

The four-velocity is a unit tangent — speeding up rotates it toward the light cone, never stretches it

One bookkeeping convention completes the toolkit. Because $u^\mu u_\mu=c^2$ identically, the four-velocity has only three independent components — the constraint is the price of the invariant clock, and it is why differentiating (4) forces $a^\mu u_\mu=0$ (the mechanics companion's orthogonality theorem). When a calculation seems to need "the fourth component of the velocity," equation (4) is the answer: solve the constraint, never treat $u^0$ as free data.

Limiting cases

case	result	meaning
$v\ll c$	$\tau\to t$ , $u^\mu\to(c,\mathbf{v})$ , corrections $O(v^2/c^2)$	Newtonian kinematics as the flat shadow of the tangent vector
$v\to c$	$d\tau/dt\to0$ , spatial $u^i\to\infty$ , norm still $c^2$	the light cone as unreachable asymptote; wristwatch nearly freezes
exactly lightlike	$d\tau=0$ : proper time undefined	photons have worldlines but no wristwatch; parametrize by anything but $\tau$ ( $k^\mu$ , not $u^\mu$ )
uniform circular motion	$\tau=t/\gamma$ exactly, indefinitely	constant speed means constant $\gamma$ even with perpetual acceleration — the storage-ring case
inertial straight line	$\tau$ maximal among all worldlines $P\to Q$	detours cost youth; twin asymmetry (owned by the mechanics companion)
clock hypothesis regime	$\tau=\int dt/\gamma$ independent of $a$	verified to $\sim10^{-3}$ at $10^{18}\,g$ (CERN muons); assumption flagged, tested, kept

Common trapGamma of what, measured by whom — the single bookkeeping error behind most proper-time nonsense

Equation (2) contains one $t$ and one $v$ , and both belong to the same arbitrary inertial frame: $\tau=\int dt/\gamma(t)$ uses the lab's clock and the lab's measurement of the traveler's speed. The classic blunders all violate this pairing. Blunder one: using the traveler's "felt" acceleration or rocket-frame time in the integral — the formula never asks for non-inertial bookkeeping; the MCIF construction (Derivation A) exists precisely so you never integrate in an accelerated frame. Blunder two: symmetry paralysis — "each twin sees the other dilated, so the integral must be ambiguous." It is not: pick ONE inertial frame, any one, and integrate $dt/\gamma(t)$ for EACH worldline in that frame; the traveler's $\gamma$ exceeds 1 during the trip while the homebody's equals 1, and every choice of bookkeeping frame yields the same two $\tau$ 's, because each is an invariant of its own worldline. Blunder three: assigning a proper time to light — $d\tau=0$ on the light cone, and "a photon's point of view" is not a frame; nothing in this article (or this subject) is licensed there.

Worked examples

1 · A four-velocity, certified and boosted (confidence builder)

Setup. A particle moves at $v=0.8c$ along $+x$ in the lab. Construct $u^\mu$ , verify its norm, and boost to the particle's rest frame.

Step 1. The Lorentz factor: $\gamma=(1-0.64)^{-1/2}=5/3$ .

Step 2. Components, per (3): $u^\mu=\gamma(c,\,v,\,0,\,0)=\left(\tfrac{5}{3}c,\;\tfrac{4}{3}c,\;0,\;0\right)\approx(1.667c,\,1.333c,\,0,\,0)$ . Note $u^1=\gamma v>v$ : the spatial part of the four-velocity already exceeds the coordinate velocity by 67% at this speed.

Step 3. The norm, per (4): $u^\mu u_\mu=\left(\tfrac{5}{3}\right)^2c^2-\left(\tfrac{4}{3}\right)^2c^2=\tfrac{25-16}{9}c^2=\mathbf{c^2}$ . ✓

Step 4. Boost with $\beta=0.8$ (into the rest frame), using the boost matrix from the previous article: $u'^0=\gamma_b(u^0-\beta u^1)=\tfrac{5}{3}\left(\tfrac{5}{3}-0.8\cdot\tfrac{4}{3}\right)c=\tfrac{5}{3}\cdot\tfrac{1.8}{3}\,c=\mathbf{c}$ , and $u'^1=\gamma_b(u^1-\beta u^0)=\tfrac{5}{3}\left(\tfrac{4}{3}-0.8\cdot\tfrac{5}{3}\right)c=\mathbf{0}$ . The rest-frame tangent $(c,\mathbf{0})$ emerges exactly — the four-velocity of "sitting still," pointing purely along time.

The lesson: build with $\gamma(c,\mathbf{v})$ , audit with the norm, and trust the transformation — the same matrix that boosts events boosts tangents.

2 · The muon storage ring — wristwatch physics at 10¹⁸ g (exam level)

Setup. In the CERN Muon Storage Ring, muons ( $\tau_0=2.197\ \mu$ s at rest) circulate at $\gamma=29.3$ in a ring of radius $r=7.0$ m. Find the laboratory lifetime, the distance and laps covered in one lifetime, the proper acceleration, and state what the experiment tested beyond time dilation. (This is real electrodynamics: the muons are steered by the ring's magnetic field via the Lorentz force, and the same $\gamma$ bookkeeping calibrates every collider's expectations.)

Step 1. Lab lifetime. Constant speed means constant $\gamma$ , so Derivation A's integral collapses: $t_{\text{lab}}=\gamma\tau_0=29.3\times2.197\ \mu\text{s}=\mathbf{64.4\ \mu s}$ — a 29-fold stay of execution, exactly as measured.

Step 2. Distance and laps. $\beta=\sqrt{1-1/\gamma^2}=0.99942$ , so path $=\beta c\,t_{\text{lab}}=0.99942\times2.998\times10^8\times64.37\times10^{-6}=\mathbf{19.3\ km}$ , i.e. $19286/(2\pi\times7.0)=\mathbf{439}$ laps. Without dilation a muon would cover $\beta c\tau_0=659$ m — it would not complete 15 laps, and no storage-ring measurement would exist.

Step 3. Proper acceleration. In the momentarily comoving frame the centripetal acceleration is boosted twice: $a'=\gamma^2v^2/r=(29.3)^2(0.99942\times2.998\times10^8)^2/7.0=\mathbf{1.1\times10^{19}\ m/s^2}\approx1.1\times10^{18}\,g$ . No engineered system subjects a clock to anything remotely comparable.

Step 4. What was actually tested. Had clocks cared about acceleration — an extra term $f(a)$ in $d\tau$ — this experiment was built to see it: the measured dilated lifetime agreed with pure $\gamma\tau_0$ at the $\sim10^{-3}$ level despite the $10^{18}\,g$ environment. That is the clock hypothesis, promoted from assumption to data. The same experiment measured the muon's anomalous magnetic moment via spin precession in the field — wristwatch bookkeeping and covariant electrodynamics in one apparatus, which is why its modern successor (the Fermilab $g-2$ ring, same magic $\gamma=29.3$ ) still runs.

The lesson: $\tau=t/\gamma$ for circular motion is exact, enormous accelerations included — and "how long does a relativistic particle live in my machine" is always $\gamma\tau_0$ , an equation colliders bill by.

Going further

References. Griffiths, Introduction to Electrodynamics, 4th ed., §12.1.2 · Jackson, Classical Electrodynamics, 3rd ed., §11.4–11.5 · Zangwill, Modern Electrodynamics, §22.4 · Taylor & Wheeler, Spacetime Physics, 2nd ed., Ch. 1

Provenance. Definitions and conventions follow Griffiths §12.1.2 (proper time, "proper velocity" $\gamma\mathbf{v}$ as the spatial part of $\eta^\mu$ , here written $u^\mu$ with $u^0=\gamma c$ ) and Jackson §11.4. A verifier should recompute: (1) $d\tau=dt/\gamma$ from the invariant interval; (2) the norm $u^\mu u_\mu=\gamma^2(c^2-v^2)=c^2$ ; (3) the low-speed expansions $\gamma\approx1+v^2/2c^2$ , $u^\mu\to(c,\mathbf{v})$ ; (4) the muon numbers $\gamma\tau_0$ at $\gamma=29.3$ . Addendum for this edition: per the chapter's division of labor, the light-clock derivation of time dilation, length contraction, and the twin asymmetry are owned by the mechanics companion Relativistic Kinematics (cross-referenced by name); the four-velocity normalization theorem with the four-acceleration orthogonality corollary is owned by Four-Vectors and Minkowski Spacetime (mechanics companion) and quoted here as equation (4)'s one-line lab-frame check; the $d\tau=dt/\gamma$ equation-context also appears in this track's Spacetime and Minkowski Geometry. This edition owns the arbitrary-worldline proper-time integral with the clock hypothesis flagged and tested (Derivation A, storage-ring instantiation) and the ladder principle — covariance of $d/d\tau$ and failure of $d/dt$ (Derivation B) — which the four-current, four-momentum, and charged-particle-dynamics articles consume. Worked numbers for this edition (muon: $64.4\ \mu$ s, 19.3 km, 439 laps, $a'=1.1\times10^{19}$ m/s² $=1.1\times10^{18}g$ ; four-velocity at $0.8c$ : $(1.667c,1.333c)$ , norm $c^2$ , boost to $(c,0)$ ; ISS 5 ms/6 months; LHC proton $\gamma=6928$ , 5.2 s proper per 10 lab hours) recomputed and verified for this edition.

Builds on: Postulates of Special Relativity · Lorentz Transformations · Spacetime and Minkowski Geometry · Four-Vectors and the Metric Tensor Leads to: Relativistic Energy and Four-Momentum · The Four-Current Density · The Four-Potential · The Electromagnetic Field Tensor

Is proper time just time dilation with different words?

They are one formula viewed from two ends, and conflating them costs understanding. Time dilation is a two-frame comparison statement: a clock moving relative to me ticks slower by $\gamma$ , and the statement is reciprocal (each sees the other slowed) because it compares different simultaneity slicings. Proper time is a one-worldline invariant: the arc length a specific path accumulates between two events, about which all observers agree — nothing reciprocal about it. The dilation formula $d\tau=dt/\gamma$ is the conversion between an observer's bookkeeping and the worldline's invariant. Confusions like the twin "paradox" arise precisely from using the reciprocal two-frame statement where the one-worldline invariant is called for; computed as $\int dt/\gamma$ per worldline in any single frame, there is nothing paradoxical to resolve.

The muons in the ring are accelerating violently — why are we allowed to use special relativity at all?

Because special relativity handles accelerated objects perfectly well; what it does not provide is accelerated reference frames (gravitation and genuinely non-inertial bookkeeping are general relativity's department). The worldline of an accelerating muon is a curve in flat spacetime, and every statement we make about it — its interval, its proper time — is evaluated in inertial frames: either the lab or the momentarily comoving frames used sliver-by-sliver in Derivation A. What must be added to the geometry is the clock hypothesis: that a real clock's tick rate depends only on its instantaneous velocity, not its acceleration. That is a statement about clock construction (a pendulum clock fails it; nuclear decay and atomic transitions pass), and the storage-ring result — pure- $\gamma$ dilation at $10^{18}\,g$ — is its cleanest test. So: geometry from special relativity, one experimentally certified assumption about hardware, and no general relativity needed anywhere in flat spacetime.

Why do people say the four-velocity has no magnitude, only direction?

Because its Minkowski length is frozen at $c$ by identity (4): $u^\mu u_\mu=c^2$ for every particle, dawdling or ultrarelativistic. All kinematic information lives in where the tangent points — nearly along the time axis for slow motion, tilted toward the light cone as $v\to c$ , with boosts acting as hyperbolic rotations of that direction. This is more than aesthetics; it is a constraint with teeth. It reduces the four independent-looking components to three (so never treat $u^0$ as free data), it forces the four-acceleration to be Minkowski-orthogonal to the four-velocity (differentiate the constant norm — the mechanics companion's theorem), and in the next article it will make the fixed norm of $p^\mu=mu^\mu$ into the mass-shell relation. When a relativistic calculation feels underdetermined, the frozen norm is usually the missing equation.

What does the four-velocity of a charge have to do with electromagnetism specifically?

It is the hinge on which the whole covariant formulation turns. Three consecutive constructions in this chapter consume it directly. The four-current of a moving charge distribution is $J^\mu=\rho_0u^\mu$ — rest-frame charge density times four-velocity — and that single expression contains the "moving charge is a current" fact that made the magnetism chapters work. The covariant Lorentz force law reads $dp^\mu/d\tau=(q/c)\,F^{\mu\nu}u_\nu$ -type contractions (executed in Relativistic Dynamics of a Charged Particle): the field tensor eats the four-velocity and returns the rate of four-momentum change — E-field forces and v×B forces unified as components of one contraction. And the fields of a uniformly moving charge (this chapter's transformation articles, cross-checking the E9 Liénard–Wiechert route) are most cleanly written with $u^\mu$ as the organizing vector. Electrodynamics without $u^\mu$ is possible — the last nine chapters prove it — but every velocity-dependent formula in them was this object's components in disguise.

Does anything special happen to proper time in a gravitational field, and should I worry about it here?

Yes and no, respectively. In general relativity the same formula $c^2d\tau^2=g_{\mu\nu}dx^\mu dx^\nu$ holds with the flat metric $\eta$ promoted to the curved $g_{\mu\nu}(x)$ : clocks deeper in a gravitational potential accumulate less proper time (gravitational redshift), an effect added to the velocity effect of this article. GPS is the canonical ledger: orbital speed slows satellite clocks by $\sim$ 7 μs/day (our $\int dt/\gamma$ ), altitude speeds them by $\sim$ 45 μs/day (the curved-metric term), net $+38$ μs/day, corrected in firmware. For everything in this chapter — storage rings, colliders, laboratory electrodynamics — the gravitational term is negligible ( $gh/c^2\sim10^{-16}$ per meter of height) and flat-spacetime proper time is exact at any realistic precision. The conceptual continuity is worth keeping, though: the wristwatch-as-arc-length idea survives unchanged into curved spacetime; only the ruler gets upgraded.

If the spatial part of the four-velocity can exceed c, doesn't that violate relativity?

No — nothing physical moves at $\gamma v$ . The bound "nothing exceeds $c$ " applies to coordinate velocity, $d\mathbf{x}/dt$ : distance per time measured in one inertial frame, the rate at which signals or objects actually traverse space. The four-velocity's spatial part $\gamma\mathbf{v}=d\mathbf{x}/d\tau$ has mixed parentage — lab distance over wristwatch time — and its unboundedness merely restates that the wristwatch nearly freezes at high speed. (Interstellar daydream: at $\gamma v=10c$ , you cross ten light-years in one shipboard year, coordinate speed still safely below $c$ .) Physically, $\gamma\mathbf{v}$ 's divergence is a feature doing real work in the next article: momentum $p=m\gamma v$ can grow without bound as accelerators push $v$ asymptotically toward $c$ — which is why "the LHC accelerates protons to $0.999999991c$ " and "the LHC raises proton momentum by a factor of thousands" are both true, and why no amount of pushing crosses the light cone.