The Newtonian Limit

The Newtonian limit is the calculation every relativistic theory of gravity must survive. Newton's gravity is one scalar potential Φ obeying Poisson's equation ∇²Φ = 4πGρ, with test masses accelerating as a = −∇Φ; three centuries of celestial mechanics rest on it. Einstein's theory is ten coupled nonlinear equations for the metric tensor. Showing that the ten collapse to the one — in the regime where Newton is known to be right — is a correspondence-principle requirement, and, crucially, the step that fixes the single dimensionless number linking geometry to matter. The reduction rests on three physically distinct smallness assumptions. First, the field is weak: the metric is a small perturbation of flat spacetime, $g_{\mu\nu}=\eta_{\mu\nu}+h_{\mu\nu}$ with $|h_{\mu\nu}|\ll1$ , so we keep only first order in $h$ . Second, motion is slow: velocities are far below $c$ , so time derivatives are negligible beside spatial gradients, $\partial_0\ll\partial_i$ . Third, the source is non-relativistic: the stress-energy tensor is dominated by its rest-energy component $T_{00}\approx\rho c^2$ , with pressure and momentum flux suppressed by $v^2/c^2$ . Under these three, two things must come out right at once. The geodesic equation — how a slow particle moves — must reduce to $\ddot{x}^i=-\partial_i\Phi$ , which forces $g_{00}=-(1+2\Phi/c^2)$ : the Newtonian potential lives in the time-time metric component. And the field equation must reduce to Poisson's equation for that same Φ. Demanding both determines the Einstein coupling $\kappa=8\pi G/c^4$ and certifies that general relativity contains Newton as a special case. Carrying the limit one term further exposes a spatial curvature that Newton never had — the piece that doubles the bending of starlight and turns a reassuring check into the first quantitative test of relativity.

Key takeaways

The Newtonian limit is the regime — weak field, slow motion, static non-relativistic source — where Einstein's field equations reduce to Poisson's equation ∇²Φ = 4πGρ for the gravitational potential Φ.
The potential sits in the metric as $g_{00}=-(1+2\Phi/c^2)$ , fixed by demanding the geodesic equation reproduce a = −∇Φ for a slow test particle.
Matching the reduced field equation to Poisson's equation fixes the Einstein coupling $\kappa=8\pi G/c^4\approx2.08\times10^{-43}\ \mathrm{s^2\,kg^{-1}\,m^{-1}}$ — a number dimensional analysis alone cannot supply.
The spatial metric picks up an equal-and-opposite correction, $h_{ij}=-2\Phi/c^2\,\delta_{ij}$ ; a slow particle never feels it, but light does, and it doubles the deflection to $4GM/c^2b = 1.75''$ at the solar limb.
The limit is superb because it is small: at Earth's surface $\Phi/c^2\approx-7\times10^{-10}$ , so spacetime is flat to nine decimals and Newton is essentially exact.
Key idea: linearize the metric around flat space, keep only the static weak-field slow-motion terms, and demand the geodesic equation give a = −∇Φ. The single surviving field equation is Poisson's equation, and matching to Newton fixes κ = 8πG/c⁴.

The core insight

Linearize the metric around flat space, keep only the leading static, weak-field, slow-motion terms, and demand that the geodesic equation reproduce a = −∇Φ. The single surviving field equation is Poisson's equation, and consistency with it fixes the Einstein coupling κ = 8πG/c⁴.

Where this lives today

The Newtonian limit is not a historical relic; it is the everyday interface between general relativity and applied physics. The relation $g_{00}=-(1+2\Phi/c^2)$ is what every satellite-navigation system computes second by second: the 45 μs/day gravitational advance of GPS clocks would accumulate into 10 km of positioning error per day if left uncorrected, so relativity is wired into the receiver in your phone. The next term past the limit — the first post-Newtonian (1PN) expansion — is the working language of precision gravity: the parametrized post-Newtonian (PPN) formalism casts every solar-system test (Cassini's Shapiro-delay measurement of the deflection parameter γ to one part in 10⁵, lunar laser ranging, Mercury's perihelion) as bounds on coefficients that all equal their general-relativistic values in the limit above. In gravitational-wave astronomy, LIGO–Virgo–KAGRA template banks are built on high-order post-Newtonian expansions, and the leading term of each is this Newtonian limit. And in cosmology the same reduction, with the $-\Lambda c^2$ term restored, underlies Newtonian N-body simulations of large-scale structure — billions of particles evolved under Poisson's equation because on sub-horizon scales that is all general relativity asks for.

The physical picture

Picture spacetime as a nearly flat sheet. In the everyday world the only curvature that matters is a tiny sag in the time direction: clocks tick very slightly slower deeper in a gravitational well. A slow particle spends almost all of its four-velocity marching forward through time, barely moving in space, so it is the time-warp it responds to — and it drifts toward wherever proper time runs slowest, which is toward larger $|\Phi|$ . That drift is Newtonian gravity. The whole content of $g_{00}=-(1+2\Phi/c^2)$ is that the rate of a clock encodes the potential, and the gradient of that rate is the force.

Figure

A slow test particle rides the g₀₀ time-warp: it accelerates toward the deepest potential, where clocks run slowest, reproducing a = −∇Φ. At Earth's surface Φ/c² ≈ −7×10⁻¹⁰, so the metric departs from flat only in the tenth decimal — which is exactly why Newton's law works so superbly.

Why does only $T_{00}$ act as the source? In ordinary matter the rest-energy density $\rho c^2$ dwarfs the kinetic-energy density ( $\sim\rho v^2$ ) and the pressure ( $\sim\rho v^2$ for a gas) by the enormous factor $c^2/v^2$ . That $c^2$ enhancement of rest mass is precisely why Newton could couple gravity to mass and ignore everything relativistic. And the slow-motion assumption is what kills the time derivatives: since $\partial_t\sim v\,\partial_x\ll c\,\partial_x$ , the spatial Laplacian $\nabla^2$ dominates the wave operator $\Box=-c^{-2}\partial_t^2+\nabla^2$ . Drop the time derivative and the relativistic wave equation for the potential collapses to Poisson's instantaneous equation — gravity appears to act at a distance only because we have thrown away the finite propagation speed. The second figure shows the payoff of tracking the spatial curvature too: the factor of 2 that separates Einstein from Newton in the bending of light.

Figure

The deflection of starlight grazing the Sun. Time curvature (the g₀₀ warp) alone gives the Newtonian value 0.87″; the spatial curvature adds an equal 0.87″, doubling the total to the Einstein value 1.75″ that Eddington confirmed at the 1919 eclipse and ruled out the Newton-only prediction.

The core equations

The target of the whole calculation is Newton's field equation — one scalar equation relating the potential to the mass density. Every reduction below is aimed at recovering it from the metric.

\nabla^2 \Phi = 4\pi G \rho \qquad (1)

where: $\Phi$ — Newtonian gravitational potential (energy per unit mass), units $\mathrm{J/kg}=\mathrm{m^2/s^2}$ ; $\rho$ — mass density of the source, $\mathrm{kg/m^3}$ ; $G$ — Newton's constant, $6.674\times10^{-11}\ \mathrm{m^3\,kg^{-1}\,s^{-2}}$ ; $\nabla^2$ — flat-space Laplacian. Reading: the potential curves in response to local mass, and a slow particle accelerates down its gradient, $\mathbf{a}=-\nabla\Phi$ . The starting point on the relativity side is that the metric differs only slightly from Minkowski. We split it into flat background plus small perturbation and keep first order.

g_{\mu\nu} = \eta_{\mu\nu} + h_{\mu\nu}, \qquad |h_{\mu\nu}| \ll 1 \qquad (2)

where: $\eta_{\mu\nu}$ — Minkowski metric, signature $(-,+,+,+)$ ; $h_{\mu\nu}$ — small, dimensionless perturbation. Reading: linearizing in $h$ is what turns the nonlinear Einstein equations into something Newton-shaped; the split is exact, the smallness is the approximation. The single most important structural result is that the Newtonian potential is not free to sit anywhere — the geodesic equation pins it to the time-time component.

g_{00} = -\left(1 + \frac{2\Phi}{c^2}\right), \qquad h_{00} = -\frac{2\Phi}{c^2} \qquad (3)

where: $g_{00}$ — time-time metric component; $\Phi/c^2$ — dimensionless potential, $\approx-7\times10^{-10}$ at Earth's surface, $\approx-2\times10^{-6}$ at the Sun's. Reading: a clock at potential $\Phi$ runs at rate $\mathrm{d}\tau\approx(1+\Phi/c^2)\,\mathrm{d}t$ — slower where the well is deeper. Derivation A fixes the factor and sign. The field equation, once its coupling is known, is Einstein's equation. The number $8\pi G/c^4$ is the output of the Newtonian match, not an input.

G_{\mu\nu} = R_{\mu\nu} - \tfrac12 R\,g_{\mu\nu} = \frac{8\pi G}{c^4}\,T_{\mu\nu} \qquad (4)

where: $G_{\mu\nu}$ — Einstein tensor; $R_{\mu\nu},R$ — Ricci tensor and scalar; $T_{\mu\nu}$ — stress-energy, $\mathrm{J/m^3}$ ; $\kappa=8\pi G/c^4\approx2.08\times10^{-43}\ \mathrm{s^2\,kg^{-1}\,m^{-1}}$ . Reading: geometry on the left, matter on the right, joined by a coupling so tiny that it takes a star's worth of mass to bend spacetime measurably. Two ingredients feed the match: a non-relativistic source, and the static weak-field value of the time-time Ricci tensor.

T_{00} \approx \rho c^2, \qquad R_{00} \approx -\tfrac12\nabla^2 h_{00} = \frac{\nabla^2\Phi}{c^2} \qquad (5)

where: $T_{00}$ — energy density, dominated by rest mass $\rho c^2$ ; $R_{00}$ — time-time Ricci, which in the static weak field is just the Laplacian of $h_{00}$ . Reading: the geometric object that becomes $\nabla^2\Phi$ is $R_{00}$ , and the matter object that becomes $\rho$ is $T_{00}$ — Poisson's equation is the $00$ component of Einstein's, nothing more. Carrying the spatial equations along too yields the full Newtonian-limit line element, whose spatial part is the term Newton never wrote.

\mathrm{d}s^2 = -\left(1+\frac{2\Phi}{c^2}\right)c^2\,\mathrm{d}t^2 + \left(1-\frac{2\Phi}{c^2}\right)\delta_{ij}\,\mathrm{d}x^i\,\mathrm{d}x^j \qquad (6)

where: $\Phi<0$ near masses; $\delta_{ij}$ — flat spatial metric. Reading: the time part and the space part carry $\Phi$ with equal magnitude and opposite sign. A slow particle samples only the $\mathrm{d}t^2$ term; light, moving at $c$ , samples both equally — and that is the entire origin of the factor of 2 in light bending (Derivation C).

Derivation AWhy the Newtonian potential lives in the time-time metric

The goal is to show that the potential is not placed in $g_{00}$ by hand but forced there by demanding that a freely falling slow particle obey Newton's second law. We start from the exact geodesic equation, impose slow motion and a static weak field to strip it down to a single term, evaluate the surviving Christoffel symbol, and match the result to $\ddot{x}^i=-\partial_i\Phi$ . The matching fixes both the numerical factor 2 and the sign, and it is the reason the identification $g_{00}=-(1+2\Phi/c^2)$ is unique rather than conventional.

Step 1. Write the exact geodesic equation, $\mathrm{d}^2x^\mu/\mathrm{d}\tau^2+\Gamma^\mu_{\alpha\beta}(\mathrm{d}x^\alpha/\mathrm{d}\tau)(\mathrm{d}x^\beta/\mathrm{d}\tau)=0$ . This is free fall: the particle follows a straight line in curved spacetime, and all the gravity is in $\Gamma^\mu_{\alpha\beta}$ . Why: Newton's "force" must emerge from these connection coefficients, since a geodesic has no force term at all.

Step 2. Impose slow motion: the spatial velocity components obey $\mathrm{d}x^i/\mathrm{d}\tau\ll c\,\mathrm{d}x^0/\mathrm{d}\tau$ , so in the double sum over $\alpha,\beta$ only the $\alpha=\beta=0$ term is not negligible. The equation reduces to $\ddot{x}^\mu+\Gamma^\mu_{00}(\dot{x}^0)^2=0$ . Why: keeping only the largest term is exactly the slow-motion approximation, and it is what removes velocity-dependent (magnetic-type gravitational) forces that Newton does not have.

Step 3. Evaluate the connection in the static weak field. To first order $\Gamma^i_{00}=\tfrac12\,\eta^{ij}(2\partial_0 h_{j0}-\partial_j h_{00})$ , and for a static field $\partial_0 h_{j0}=0$ , leaving $\Gamma^i_{00}=-\tfrac12\,\eta^{ij}\partial_j h_{00}=\tfrac12\,\partial_i h_{00}$ . Only the spatial gradient of $h_{00}$ survives. Why: dropping the time derivative is the static assumption, and it is what makes the resulting force conservative — a gradient of a single potential.

Step 4. Convert proper time to coordinate time. Since $\dot{x}^0=c\,\mathrm{d}t/\mathrm{d}\tau\approx c$ to leading order, the spatial equation becomes $\ddot{x}^i=-\tfrac12 c^2\,\partial_i h_{00}$ , where dots now denote $\mathrm{d}/\mathrm{d}t$ . Why: Newton's law is written in coordinate time, so we must trade $\tau$ for $t$ before comparing.

Step 5. Match to Newton. Setting $\ddot{x}^i=-\tfrac12 c^2\,\partial_i h_{00}\stackrel{!}{=}-\partial_i\Phi$ forces $h_{00}=2\Phi/c^2$ up to a constant, and with the signature $(-,+,+,+)$ this means $g_{00}=\eta_{00}+h_{00}=-(1+2\Phi/c^2)$ . Why: this is the load-bearing identification of the whole subject — the potential is the time-time metric perturbation, factor and sign fixed by dynamics, not by choice.

Go deeper · Derivation B · 1 minReducing the field equations to Poisson and fixing the coupling constant

Derivation BReducing the field equations to Poisson and fixing the coupling constant

The goal is to turn Einstein's equation into Poisson's equation and, in doing so, pin down the coupling $\kappa$ that Einstein's equation leaves as an unknown constant $G_{\mu\nu}=\kappa T_{\mu\nu}$ . The clean route is the trace-reversed form, where the time-time component isolates $\nabla^2\Phi$ without the trace subtraction getting in the way. We trace-reverse, evaluate the non-relativistic source trace, assemble the time-time source, insert the geometric side from equation (5), and compare with Newton — which leaves exactly one value of $\kappa$ standing.

Step 1. Trace-reverse. Taking the trace of $G_{\mu\nu}=\kappa T_{\mu\nu}$ with $g^{\mu\nu}g_{\mu\nu}=4$ gives $-R=\kappa T$ ; substituting back yields $R_{\mu\nu}=\kappa\!\left(T_{\mu\nu}-\tfrac12 T g_{\mu\nu}\right)$ . Why: this form puts a bare Ricci tensor on the left, and $R_{00}$ is precisely what became $\nabla^2\Phi/c^2$ in equation (5).

Step 2. Compute the source trace for non-relativistic matter. With only $T_{00}\approx\rho c^2$ significant and $g^{00}\approx\eta^{00}=-1$ , the trace is $T=g^{\mu\nu}T_{\mu\nu}\approx-\rho c^2$ . Why: pressure and momentum flux enter $T$ at order $v^2/c^2$ , so dropping them is the same non-relativistic assumption used everywhere in this limit.

Step 3. Assemble the time-time source term: $T_{00}-\tfrac12 T g_{00}\approx\rho c^2-\tfrac12(-\rho c^2)(-1)=\tfrac12\rho c^2$ , using $g_{00}\approx-1$ to leading order. Why: the trace reversal halves the effective source, and this factor of one-half is exactly what will conspire with the $8\pi$ to reproduce the $4\pi$ in Poisson's equation.

Step 4. Insert the geometric side. Equation (5) gives $R_{00}\approx\nabla^2\Phi/c^2$ , so the $00$ component of the trace-reversed equation reads $\nabla^2\Phi/c^2=\kappa\cdot\tfrac12\rho c^2$ , i.e. $\nabla^2\Phi=\tfrac12\kappa c^4\rho$ . Why: this is Poisson's equation already, but with an undetermined coefficient in front of $\rho$ — the coefficient is the whole prize.

Step 5. Match to Newton. Comparing $\nabla^2\Phi=\tfrac12\kappa c^4\rho$ with $\nabla^2\Phi=4\pi G\rho$ forces $\tfrac12\kappa c^4=4\pi G$ , hence $\kappa=8\pi G/c^4$ . Numerically $\kappa=8\pi(6.674\times10^{-11})/(2.998\times10^8)^4=2.08\times10^{-43}\ \mathrm{s^2\,kg^{-1}\,m^{-1}}$ . Why: dimensional analysis fixes only the powers of $G$ and $c$ ; the pure number $8\pi$ can come from nowhere but this correspondence with Newtonian gravity.

Go deeper · Derivation C · 1 minThe spatial curvature and the doubling of light deflection

Derivation CThe spatial curvature and the doubling of light deflection

The goal is to expose the one piece of the Newtonian-limit metric that Newton's scalar theory has no room for — the spatial perturbation $h_{ij}$ — and to show it doubles the gravitational deflection of light. A slow massive particle is blind to $h_{ij}$ because it barely moves in space, but a photon travels at $c$ and threads equal path through space and time, so it feels both metric warps. We solve the spatial Einstein equation for $h_{ij}$ , then compare the deflection with and without it.

Step 1. Solve the spatial field equation. For pressureless dust $T_{ij}\approx0$ , so the trace-reversed spatial equation is $R_{ij}\approx-\tfrac12\kappa T g_{ij}\approx\tfrac12\kappa\rho c^2\,\delta_{ij}$ , the same magnitude of source as the $00$ equation. This yields $h_{ij}=-\dfrac{2\Phi}{c^2}\,\delta_{ij}$ . Why: the spatial curvature is locked to the same potential $\Phi$ but with opposite sign, giving the symmetric line element of equation (6) — this symmetry is a genuine prediction absent from any pure time-warp theory.

Step 2. Compute the deflection from the time warp alone. Treating light in only the $g_{00}$ potential (a "Newtonian photon") gives a bending angle $\Delta\phi_{\text{time}}=2GM/c^2 b$ for a ray of impact parameter $b$ past a point mass $M$ . Why: this is the historical Soldner–Cavendish value, and it is what a scalar gravity that curves only time would predict.

Step 3. Add the space warp. Because a null ray satisfies $\mathrm{d}s^2=0$ , the spatial term $h_{ij}$ contributes to its trajectory on exactly equal footing with $h_{00}$ , adding a second, equal $2GM/c^2 b$ . The total is $\Delta\phi=2\times\dfrac{2GM}{c^2 b}=\dfrac{4GM}{c^2 b}$ . Why: the equality of the two contributions is the physical meaning of the factor of 2 — light "sees twice the gravity" a slow particle does, purely because it moves at $c$ .

Step 4. Put in the Sun. With $M=M_\odot$ and $b=R_\odot$ , $\Delta\phi=4GM_\odot/(c^2R_\odot)=8.49\times10^{-6}\ \mathrm{rad}=1.75''$ , twice the Newtonian $0.87''$ . Why: this number, measured by Eddington at the 1919 eclipse, is the first quantitative confirmation that the spatial curvature is real — the Newtonian limit taken one term past Newton.

Limiting cases

The Newtonian limit is one corner of a hierarchy; relaxing each of its three assumptions opens a richer regime.

Case	Result	Meaning
Keep time derivatives (still weak field)	$\Box\,\bar{h}_{\mu\nu}=-\dfrac{16\pi G}{c^4}T_{\mu\nu}$	Restoring $\partial_t$ turns Poisson into a wave equation: gravitational waves and retarded potentials. This is linearized gravity, not the Newtonian limit.
Restore pressure and velocities (post-Newtonian)	$\nabla^2\Phi=4\pi G\!\left(\rho+\dfrac{3p}{c^2}+\dots\right)$	The 1PN expansion brings back pressure, kinetic energy, and the spatial metric; it governs Mercury's perihelion advance and binary-pulsar timing.
Include the cosmological constant	$\nabla^2\Phi=4\pi G\rho-\Lambda c^2$	A nonzero $\Lambda$ adds a repulsive harmonic term $\Phi_\Lambda=-\tfrac16\Lambda c^2 r^2$ ; with $\Lambda\approx1.1\times10^{-52}\ \mathrm{m^{-2}}$ it is utterly negligible in the solar system.
Strong field (limit breaks down)	$\|\Phi\|/c^2\sim1$	Near the Schwarzschild radius $r_s=2GM/c^2$ the expansion in $h$ fails entirely; the full nonlinear Einstein equations are required.

Common trapWeak field is not the same as Newtonian

It is tempting to treat "weak field" and "Newtonian limit" as synonyms — they are not. Weak-field linearized gravity keeps the time derivatives and therefore still describes gravitational waves, retardation, and relativistic motion. The Newtonian limit is more restrictive: on top of weak fields it demands slow motion (so $\Box\to\nabla^2$ ) and a non-relativistic source (so $T_{00}\approx\rho c^2$ dominates). A second, subtler trap: the spatial perturbation $h_{ij}=-2\Phi/c^2\,\delta_{ij}$ is a real part of the Newtonian-limit metric even though a slow particle never feels it. Forgetting it makes light bending come out at half its true value — the mistake that separated Newtonian gravity from Einstein's at the 1919 eclipse.

Worked examples

1 · The bending of starlight by the Sun (fully worked)

A ray of starlight grazes the solar limb, impact parameter $b=R_\odot=6.96\times10^8\ \mathrm{m}$ , past $M_\odot=1.99\times10^{30}\ \mathrm{kg}$ . Find the general-relativistic deflection, and the value a time-warp-only (Newtonian) theory would give.

Step 1. Assemble the constant $GM_\odot/c^2$ , the Sun's gravitational radius: $GM_\odot=(6.674\times10^{-11})(1.99\times10^{30})=1.328\times10^{20}\ \mathrm{m^3/s^2}$ , so $GM_\odot/c^2=1.328\times10^{20}/8.99\times10^{16}=\mathbf{1.48\times10^{3}\ m}\approx1.48\ \mathrm{km}$ .

Step 2. The full deflection is $\Delta\phi=4GM_\odot/(c^2 b)=4(1.48\times10^3)/(6.96\times10^8)=8.49\times10^{-6}\ \mathrm{rad}$ . Converting with $1\ \mathrm{rad}=206265''$ gives $\Delta\phi=8.49\times10^{-6}\times206265=\mathbf{1.75''}$ .

Step 3. The Newtonian (time-only) prediction is exactly half, because it omits the spatial curvature $h_{ij}$ : $\Delta\phi_{\text{Newton}}=2GM_\odot/(c^2 b)=\mathbf{0.87''}$ . The 1919 eclipse measurements landed near $1.75''$ and excluded $0.87''$ — a direct detection of the spatial part of the metric. GR: 1.75″; Newtonian: 0.87″; the difference is the space curvature.

2 · GPS clocks and the g₀₀ time-warp (fully worked)

A GPS satellite orbits at radius $r_\mathrm{s}=2.656\times10^7\ \mathrm{m}$ from Earth's centre; a ground clock sits at $R_\oplus=6.371\times10^6\ \mathrm{m}$ . Using only $g_{00}=-(1+2\Phi/c^2)$ , find the gravitational rate difference between the two clocks.

Step 1. Proper time runs as $\mathrm{d}\tau\approx(1+\Phi/c^2)\,\mathrm{d}t$ with $\Phi=-GM_\oplus/r$ and $GM_\oplus=3.986\times10^{14}\ \mathrm{m^3/s^2}$ . Ground: $\Phi_\oplus/c^2=-3.986\times10^{14}/(6.371\times10^6\cdot8.99\times10^{16})=\mathbf{-6.96\times10^{-10}}$ .

Step 2. Satellite (higher up, shallower well): $\Phi_\mathrm{s}/c^2=-3.986\times10^{14}/(2.656\times10^7\cdot8.99\times10^{16})=\mathbf{-1.67\times10^{-10}}$ .

Step 3. The fractional rate difference is $\Delta\Phi/c^2=\Phi_\mathrm{s}/c^2-\Phi_\oplus/c^2=(-1.67+6.96)\times10^{-10}=\mathbf{+5.29\times10^{-10}}$ : the satellite clock runs fast. Over one day, $5.29\times10^{-10}\times86400\ \mathrm{s}=\mathbf{+45.7\ \mu s/day}$ . This is the gravitational piece the GPS system must correct continuously; special-relativistic time dilation from orbital speed subtracts about $7\ \mu\mathrm{s/day}$ , leaving a net $\approx+38\ \mu\mathrm{s/day}$ . The g₀₀ term alone advances a GPS clock by ≈ 46 μs each day.

3 · The Einstein coupling as a number

Evaluate $\kappa=8\pi G/c^4$ and confirm it reproduces $4\pi G$ in Poisson's equation.

Step 1. Numerator: $8\pi G=8\pi(6.674\times10^{-11})=1.677\times10^{-9}$ . Denominator: $c^4=(2.998\times10^8)^4=8.08\times10^{33}$ . So $\kappa=1.677\times10^{-9}/8.08\times10^{33}=\mathbf{2.08\times10^{-43}\ s^2\,kg^{-1}\,m^{-1}}$ .

Step 2. Check the reduction: $\tfrac12\kappa c^4=\tfrac12(8\pi G/c^4)c^4=4\pi G$ , exactly the coefficient in $\nabla^2\Phi=4\pi G\rho$ . The staggering smallness of κ — 43 orders below unity — is why gravity is the weakest force and why spacetime is nearly flat everywhere we live.

Physics of the portrait

The Newtonian limit is where two of physics' greatest edifices are made to touch. Newton supplied the target equation and two centuries of confirming orbits; Einstein supplied the ten-equation theory that had to reproduce them — and it was Einstein himself who performed the limit, in November 1915, to check his equations and to fix their coupling.

Origins · 1 minAlbert Einstein and how this was discovered

Albert Einstein

1879–1955 · general relativity and the Newtonian correspondence

Albert Einstein did not arrive at his field equations and then discover they contained Newton; he used the Newtonian correspondence as a rudder while searching for them. Through 1913–1915 he demanded that any acceptable field equation reduce, for weak static fields and slow matter, to Poisson's $\nabla^2\Phi=4\pi G\rho$ — a constraint that repeatedly killed candidate equations, including his own earlier "Entwurf" theory. When he presented the final equations to the Prussian Academy in November 1915, the weak-field reduction was not an afterthought but part of the proof: it is what fixed the coupling as $8\pi G/c^4$ rather than any other multiple, and what let him compute Mercury's perihelion advance of 43″ per century that same month. The identification $g_{00}=-(1+2\Phi/c^2)$ had appeared even earlier, in his 1907–1911 work on gravitational time dilation and light bending, where he first found the (then still Newtonian-sized) $0.87''$ deflection — later doubled once the full theory supplied the spatial curvature. In Einstein's hands the Newtonian limit was both a consistency check and a discovery tool: the demand that a new theory contain the old one, made quantitative, is a template that has guided theoretical physics ever since.

History

1687

— Newton's Principia states universal gravitation and the inverse-square law

1813

— Poisson recasts gravity in field form as ∇²Φ = 4πGρ

1911

— Einstein predicts a Newtonian-sized 0.87″ solar light deflection from time dilation alone

1915

— Einstein publishes the field equations and recovers Newton in the weak-field slow-motion limit, doubling the deflection to 1.75″

1919

— Eddington's eclipse expeditions measure ≈1.75″, confirming the spatial curvature

1959

— Pound and Rebka measure the g₀₀ gravitational redshift in a Harvard tower

Going further

References. Misner, Thorne & Wheeler, Gravitation (1973), §17.4 and Box 17.2 (Newtonian correspondence), §40.3 (light deflection) · Wald, General Relativity (1984), §4.4 · Carroll, Spacetime and Geometry (2004), §4.1 (the Newtonian limit) · Hartle, Gravity (2003), Ch. 6 · Schutz, A First Course in General Relativity, 2nd ed. (2009), §8.4.

Provenance. The derivations follow the standard treatments in Carroll §4.1, Wald §4.4, and MTW §17.4. A verifier should recompute: (1) the linearized $\Gamma^i_{00}=\tfrac12\partial_i h_{00}$ and the geodesic reduction giving $h_{00}=2\Phi/c^2$ (Derivation A); (2) the trace-reverse yielding $\nabla^2\Phi=\tfrac12\kappa c^4\rho$ and hence $\kappa=8\pi G/c^4\approx2.08\times10^{-43}\ \mathrm{s^2\,kg^{-1}\,m^{-1}}$ (Derivation B); (3) the spatial solution $h_{ij}=-2\Phi/c^2\,\delta_{ij}$ and the factor-of-2 deflection $\Delta\phi=4GM/c^2 b$ (Derivation C). Numbers verified independently: $\Phi_\oplus/c^2=-6.96\times10^{-10}$ , GPS gravitational rate $+45.7\ \mu\mathrm{s/day}$ , solar deflection $1.75''$ (GR) vs $0.87''$ (Newton), $\Lambda\approx1.1\times10^{-52}\ \mathrm{m^{-2}}$ — all textbook standard.

Prerequisites. The Einstein Field Equations · The Weak-Field Approximation · The Stress-Energy Tensor · The Einstein Tensor. Leads to. Perfect-Fluid Stress-Energy · The Cosmological Constant · The Einstein–Hilbert Action.

Ask the tutor: Ask me to walk through any one of the three derivations line by line — the geodesic match that puts Φ in g₀₀, the trace-reverse that delivers 8πG/c⁴, or the spatial-curvature step that doubles the bending of starlight.

What is the difference between the weak-field approximation and the Newtonian limit?

The weak-field approximation is the single assumption $|h_{\mu\nu}|\ll1$ : the metric is close to flat, so Einstein's equations can be linearized. That alone still permits fast-moving sources and time-varying fields, and it describes gravitational waves and retarded potentials. The Newtonian limit is the weak-field approximation plus two more assumptions: slow motion, which lets you drop time derivatives so the wave operator $\Box$ becomes the Laplacian $\nabla^2$ ; and a non-relativistic source, so that $T_{00}\approx\rho c^2$ dominates the stress-energy. Only with all three does Einstein's equation collapse to the instantaneous Poisson equation. In short, every Newtonian limit is weak-field, but not every weak-field situation is Newtonian.

Why does the potential appear in g₀₀ and nowhere else at leading order?

Because it is forced there by how a slow particle moves. A slow test mass has almost all of its four-velocity in the time direction, so the geodesic equation retains only the $\Gamma^i_{00}$ connection coefficient, which depends solely on the gradient of $h_{00}$ . Demanding that the result equal Newton's $\ddot{x}^i=-\partial_i\Phi$ then pins $h_{00}=2\Phi/c^2$ . The spatial components $h_{ij}$ are also nonzero and equally large, but a slow particle simply does not sample them — they multiply the tiny spatial velocities. So "the potential lives in $g_{00}$ " is precisely a statement about slow-particle dynamics; for light, which is not slow, the spatial part matters just as much.

Where does the factor of 8π in the Einstein coupling come from?

It comes entirely from the Newtonian match — dimensional analysis cannot supply it. The powers of $G$ and $c$ in $\kappa=8\pi G/c^4$ are fixed by units, but the pure number is not. Tracing it through Derivation B: the trace-reversed field equation puts a factor $\tfrac12$ in the source (from $T_{00}-\tfrac12 T g_{00}=\tfrac12\rho c^2$ ), so the $00$ components give $\nabla^2\Phi=\tfrac12\kappa c^4\rho$ . Comparing with Poisson's $\nabla^2\Phi=4\pi G\rho$ demands $\tfrac12\kappa c^4=4\pi G$ , and the factor of 2 combines with the $4\pi$ to give $8\pi$ . The $4\pi$ itself descends from the surface area of a sphere in Gauss's law for gravity.

Why does a slow particle ignore the spatial curvature while light responds to it?

Both feel the full metric, but they weight its pieces differently. The line element is $\mathrm{d}s^2=g_{00}c^2\mathrm{d}t^2+g_{ij}\mathrm{d}x^i\mathrm{d}x^j$ . For a slow particle $\mathrm{d}x^i\ll c\,\mathrm{d}t$ , so the spatial term is smaller than the time term by $(v/c)^2$ and is entirely negligible — the particle's trajectory is set by $g_{00}$ alone. For light $\mathrm{d}s^2=0$ with $|\mathrm{d}x^i|=c\,\mathrm{d}t$ , so the spatial and temporal terms are the same size, and $h_{ij}$ contributes to the photon's path on equal footing with $h_{00}$ . That equality is the whole reason light bends by $4GM/c^2b$ while a "Newtonian photon," feeling only $g_{00}$ , would bend by half as much.

Does the Newtonian limit mean gravity really acts instantaneously?

No — it means the finite propagation speed is invisible here, not absent. Poisson's equation $\nabla^2\Phi=4\pi G\rho$ has no time derivative, so it responds to the source everywhere at once, which looks like action at a distance. But that instantaneity is an artifact of dropping the $\partial_t^2/c^2$ term, legitimate only because the source moves slowly compared with $c$ . Restore that term and you recover the wave equation $\Box\bar{h}_{\mu\nu}=-16\pi G T_{\mu\nu}/c^4$ , whose solutions are retarded potentials propagating at exactly $c$ — gravitational waves. General relativity is fully causal; the Newtonian limit merely describes systems where the light-crossing time is too short to notice.

When does the Newtonian limit break down, and by how much?

It breaks down when any of its three smallness parameters stops being small. The controlling ratio is $|\Phi|/c^2\sim GM/c^2 r$ , which is $10^{-9}$ at Earth's surface and $10^{-6}$ at the Sun's, but climbs to $\sim0.1$ at a neutron star and to $\tfrac12$ at a black-hole horizon ( $r=r_s=2GM/c^2$ ); once it is not tiny, first order in $h$ fails and the full nonlinear equations are needed. Likewise, a relativistic source — $v\sim c$ in a binary inspiral, or $p\sim\rho c^2$ in a neutron-star core — defeats the slow-motion and non-relativistic-source assumptions even where the field is weak. Post-Newtonian expansions bridge the gap in powers of $\Phi/c^2$ and $v^2/c^2$ .