The Relativistic Lagrangian for a Free and Charged Particle

The substitution this article derives classically — $\boldsymbol{\pi} \to \boldsymbol{\pi} - q\mathbf{A}$ — is, after quantization, the construction principle of the Standard Model: every interaction of matter with the electromagnetic, weak, and strong forces enters through a covariant derivative that is minimal coupling with a richer gauge group. When lattice-QCD codes on exascale machines compute a proton mass, or a quantum-computing group engineers artificial gauge fields for neutral atoms, they are running descendants of equation (6). The Aharonov–Bohm effect, meanwhile, has become a laboratory workhorse — routinely imaged in electron holography and used to certify phase coherence in mesoscopic rings. The geometric reading pays operationally too. Every GPS fix solves for geodesics of a metric that differs from $\eta_{\mu\nu}$ by parts in $10^{10}$ — the "extremize aging" principle with Earth's gravity in the metric — and every storage ring tracks particles by symplectically integrating the Hamiltonian (9), preserving the canonical structure this action defines for $10^{10}$ turns. At the speculative frontier, the einbein trick's descendants (worldline formalism, reparametrization-invariant actions) remain the standard toolkit for computing quantum field theory amplitudes and the starting point of string theory, where the worldline becomes a worldsheet and the einbein a two-dimensional metric. Hamilton's principle, having survived relativity, turned out to be the part of mechanics the twentieth century kept.

Between two fixed events, every timelike path through spacetime is a possible history, and each is graded by exactly one invariant number: the proper time a clock would accumulate traveling it. Straight (unaccelerated) worldlines maximize that aging — wiggling costs proper time, as the traveling twin discovers. The action $S = -mc^2\int\mathrm{d}\tau$ simply converts "maximize aging" into "minimize action" via the minus sign, and Hamilton's principle from the W3 articles does the rest. A free particle's dynamics is not driven by forces; it is the geometry of doing nothing as slowly as possible, in the only invariant sense available.

Why must the Lagrangian abandon the $T - V$ template? Because $T - V$ secretly assumes absolute time — it grades a path by quantities one observer computes on their own simultaneity slices. Insisting instead that all observers agree on the grade leaves spacetime length as the only currency. The astonishment is how much falls out of so little: differentiate $-mc^2/\gamma$ and relativistic momentum appears; Legendre-transform it and the rest energy $mc^2$ — which Newtonian mechanics cannot see at all — arrives uninvited, sitting in the Hamiltonian at $v = 0$.

Minimal coupling has an equally geometric reading. The interaction $-q\int A_\mu\,\mathrm{d}x^\mu$ is the circulation of the four-potential along the worldline — a quantity whose variation feels only the antisymmetric combination $\partial_\mu A_\nu - \partial_\nu A_\mu = F_{\mu\nu}$. Gauge transformations shift the integrand by a total derivative, dying at the fixed endpoints: the potential is convention, the field tensor is physics, and the action principle enforces that division of labor automatically. The bookkeeping consequence — that the momentum conjugate to position is $\boldsymbol{\pi} = \mathbf{p} + q\mathbf{A}$, not the kinetic $\mathbf{p}$ — looks like a technicality here; promoted to quantum mechanics it becomes the covariant derivative, the Aharonov–Bohm effect, and the structural template of the Standard Model.

Metric signature $(+,-,-,-)$ and SI units throughout, matching the rest of this chapter. The invariant grade of a timelike worldline is its proper time, and the free action is that grade times the only available constant with dimensions of action:

where: $\tau$ — proper time; $\eta_{\mu\nu}$ — Minkowski metric; the square root is real precisely for timelike (sub-light) paths. Linearity in the worldline, correct dimensions, and Lorentz invariance leave essentially no other choice — the minus sign converts maximal aging into minimal action, and the factor $mc^2$ is fixed by the Newtonian limit below. Converting the invariant measure to any observer's clock via time dilation, $\mathrm{d}\tau = \mathrm{d}t/\gamma$, hands that observer a working Lagrangian:

where: $v$ — lab-frame speed; $\gamma$ — Lorentz factor. Note what this is not: not the kinetic energy $(\gamma - 1)mc^2$, not $T - V$, not positive. At low speed it expands to $-mc^2 + \tfrac12 mv^2 + \cdots$ — Newton's free Lagrangian riding on an inert constant whose sole (and crucial) job is to appear later in the energy. The Euler–Lagrange machinery of Hamilton's principle now runs unchanged (Derivation A). The conjugate momentum is

where: the square root differentiates into exactly the $\gamma$ of relativistic momentum — the same $\gamma m\mathbf{v}$ the four-momentum article obtained from spacetime geometry, now regenerated by the variational principle. Free motion conserves it, recovering Newton's first law relativistically. The Legendre transform then delivers the energy:

where: the rest energy emerges at $v = 0$ because of the constant $-mc^2$ in $L$ — the term that never touches the equations of motion determines the zero of energy. Eliminating $v$ between (3) and (4) reproduces the mass shell: the Lagrangian, four-vector, and collision bookkeeping of the last two articles are one framework. Charge enters through the unique Lorentz- and gauge-respecting scalar linear in both the worldline and the four-potential $A^\mu = (\phi/c, \mathbf{A})$:

where: $q$ — charge. Under a gauge transformation $A_\mu \to A_\mu + \partial_\mu\chi$ the new term changes by $-q[\chi]$ at the fixed endpoints only, so the stationary worldline — the physics — is untouched even though the Lagrangian is not. Expanding the line integral against lab time ($A_\mu\,\mathrm{d}x^\mu = \phi\,\mathrm{d}t - \mathbf{A}\cdot\mathrm{d}\mathbf{x}$ in this signature) gives the complete charged-particle Lagrangian:

where: $\phi$, $\mathbf{A}$ — scalar and vector potentials evaluated along the trajectory. Three terms, three roles: geometry, magnetic coupling, electric coupling. Every charged-particle problem in this course — and, after quantization, in quantum mechanics — starts from this line. Differentiating (6) splits momentum into two physically distinct objects:

where: $\boldsymbol{\pi}$ — canonical momentum (conjugate to $\mathbf{x}$, conserved when $L$ ignores a coordinate, promoted to $-i\hbar\nabla$ in quantum mechanics); $\mathbf{p}$ — kinetic momentum (the one in the mass shell and the Lorentz force). Conflating them is the classic exam disaster; the field carries the difference. The Euler–Lagrange equations of (6) — worked in Derivation B — reproduce exactly the force law the previous article wrote covariantly:

where: $\mathbf{E} = -\nabla\phi - \partial_t\mathbf{A}$, $\mathbf{B} = \nabla\times\mathbf{A}$. The four-force article postulated this law and verified its covariance; here it is derived from one scalar action — the logical closure of the chapter's relativity thread. Finally, the Legendre transform of (6) gives the relativistic Hamiltonian in canonical variables:

where: the recipe is visible in the syntax: take the free-particle energy, replace $\mathbf{p} \to \boldsymbol{\pi} - q\mathbf{A}$, add $q\phi$. That substitution is minimal coupling, and it survives verbatim into the Klein–Gordon and Dirac equations.

Minimal coupling and the quantum gauge principle. Canonical quantization promotes the canonical momentum: $\boldsymbol{\pi} \to -i\hbar\nabla$. The Hamiltonian (9) then forces the substitution $-i\hbar\nabla \to -i\hbar\nabla - q\mathbf{A}$ — the covariant derivative — into the Schrödinger, Klein–Gordon, and Dirac equations. The Aharonov–Bohm effect, flux quantization in superconductors, and ultimately the entire gauge structure of the Standard Model are this classical Lagrangian's split between $\boldsymbol{\pi}$ and $\mathbf{p}$, quantized.

Accelerator lattice design. Storage-ring tracking codes integrate the Hamiltonian (9) with symplectic integrators for billions of turns; the canonical structure inherited from this action is what keeps numerical orbits from spuriously damping or exploding. The conserved canonical momenta of cyclic coordinates dictate closed-orbit existence, and the $\gamma$ in $\omega = qB/\gamma m$ — derived here variationally — sets every ramp profile.

The geodesic bridge to general relativity. Reading $S = -mc^2\int\mathrm{d}\tau$ as "extremize aging" makes gravity's kinematics a one-line generalization: swap $\eta_{\mu\nu}$ for $g_{\mu\nu}(x)$ and the same variation yields geodesic motion — no force law needed. GPS satellite clocks, gravitational lensing, and the Mercury perihelion all sit on the far side of that one substitution, which is why this article is the traditional launch point for a general-relativity course.

Gateway to field theory. The einbein form's lesson — actions may carry gauge redundancies whose bookkeepers are auxiliary fields — scales directly up to the continuum articles this chapter turns to next: from worldlines $x^\mu(\lambda)$ to fields $\phi(x^\mu)$, from reparametrization invariance to gauge and diffeomorphism invariance, from $-mc^2\int\mathrm{d}\tau$ to $\int\mathcal{L}\,\mathrm{d}^4x$. Modern "worldline formalism" computations in quantum field theory literally use this point-particle action as their engine.

Example 1 — the Legendre transform, numerically (confidence builder). An electron ($mc^2 = 0.511$ MeV) travels at $v = 0.6c$. Evaluate $L$, $\mathbf{p}$, and the Legendre transform $E = \mathbf{p}\cdot\mathbf{v} - L$, and check against $\gamma mc^2$.

Step 1. Lorentz factor: $\gamma = 1/\sqrt{1 - 0.36} = 1/0.8 = 1.25$.

Step 2. Lagrangian: $L = -mc^2/\gamma = -0.511/1.25 = -0.4088$ MeV. Negative, as it must be — and nowhere near the kinetic energy $(\gamma - 1)mc^2 = 0.1278$ MeV.

Step 3. Momentum: $pc = \gamma mc^2\,(v/c) = 1.25 \times 0.511 \times 0.6 = 0.3833$ MeV.

Step 4. Legendre transform: $E = \mathbf{p}\cdot\mathbf{v} - L = (pc)(v/c) + mc^2/\gamma = 0.3833 \times 0.6 + 0.4088 = 0.2300 + 0.4088 = 0.6388$ MeV.

Step 5. Direct check: $\gamma mc^2 = 1.25 \times 0.511 = 0.6388$ MeV. ✓ Mass shell: $(pc)^2 + (mc^2)^2 = 0.1469 + 0.2611 = 0.4080 = 0.6388^2$. ✓ $L = -0.409$ MeV, $pc = 0.383$ MeV, $E = 0.639$ MeV — the "wrong-looking" Lagrangian delivers every right answer.

Example 2 — acceleration and bending, variationally (exam level). An electron starts from rest and falls through a $U = 1.0$ MV electrostatic gap ($\mathbf{A} = 0$), then enters a uniform field $B = 0.5$ T perpendicular to its velocity. Find its final speed and the bending radius — and show what Newtonian mechanics would have predicted for the speed.

Step 1. Conservation from the Lagrangian: with $\partial L/\partial t = 0$, the energy $H = \gamma mc^2 + q\phi$ is constant. Between the plates, $\Delta(\gamma mc^2) = qU = 1.0$ MeV.

Step 2. Final Lorentz factor: $\gamma = 1 + qU/mc^2 = 1 + 1.0/0.511 = 2.957$.

Step 3. Final speed: $v/c = \sqrt{1 - 1/\gamma^2} = \sqrt{1 - 0.1144} = \sqrt{0.8856} = 0.941$, so $v = 0.941c$.

Step 4. The Newtonian forecast $\tfrac12 mv^2 = qU$ gives $v = c\sqrt{2 \times 1.0/0.511} = 1.98\,c$ — faster than light, flatly absurd. The relativistic Lagrangian keeps the books honest.

Step 5. Momentum entering the magnet: $pc = \sqrt{E^2 - (mc^2)^2} = \sqrt{(1.511)^2 - (0.511)^2} = \sqrt{2.283 - 0.261} = \sqrt{2.022} = 1.422$ MeV.

Step 6. Bending radius from $r = p/qB$: $p = 1.422\ \mathrm{MeV}/c = \dfrac{1.422 \times 10^6 \times 1.602 \times 10^{-19}}{3.00 \times 10^8} = 7.59 \times 10^{-22}\ \mathrm{kg\,m\,s^{-1}}$, so $r = \dfrac{7.59 \times 10^{-22}}{1.602 \times 10^{-19} \times 0.5} = 9.5 \times 10^{-3}\ \mathrm{m}$. The electron exits at $0.941c$ and bends on a 9.5 mm radius — pocket-sized, which is why MeV-scale electron spectrometers fit on a bench.

References. Goldstein, Classical Mechanics, 3rd ed., §7.9–7.10 · Landau & Lifshitz, The Classical Theory of Fields, Vol. 2, §8–§16 · Morin, Introduction to Classical Mechanics, Ch. 13 · Taylor, Classical Mechanics, Ch. 15 · Jackson, Classical Electrodynamics, 3rd ed., §12.1 Key self-consistency checks: $L=-mc^2/\gamma$ yields $\partial L/\partial\mathbf{v}=\gamma m\mathbf{v}$ and $H=\gamma mc^2$; the Lorentz-force derivation turns on the vector identity $\nabla(\mathbf{A}\cdot\mathbf{v})-(\mathbf{v}\cdot\nabla)\mathbf{A}=\mathbf{v}\times(\nabla\times\mathbf{A})$; the einbein equivalence follows by eliminating $e$ and confirming reparametrization invariance $\lambda\to f(\lambda)$, $e\to e/f'$, with signs adapted to $(+,-,-,-)$; the low-speed expansion reads $-mc^2\sqrt{1-v^2/c^2}=-mc^2+\tfrac12 mv^2+\tfrac18 mv^4/c^2+\cdots$; and the worked examples give $\gamma=1.25$ for the Legendre check and $\gamma=2.957$, $v=0.941c$, $pc=1.422$ MeV, $r=9.5$ mm at 0.5 T.

Scope note. This is the mechanics-course introduction, and the payoff of this chapter's relativity thread. The full story — covariant Hamiltonian dynamics, spinning particles, radiation reaction from effective actions, the worldline path integral — lives in the Special Relativity track (migrates later as its own subject).

Builds on: Hamilton's Principle · The Lagrangian Function · Special Relativity Postulates and the Lorentz Transformation · Four-Velocity, Four-Momentum, and Mass-Energy Equivalence · Relativistic Dynamics and the Four-Force Leads to: Lagrangian Density and Continuum Lagrangian Formalism · Hamiltonian Density and Canonical Field Momentum · Euler–Lagrange Equations for Fields