Newton's Law of Gravitation: Gravitational Field, Potential, and the Shell Theorem

Intuition Beginner

Drop an apple. It accelerates downward at about 10 metres per second squared. Newton's insight was that this same acceleration — the same force, scaled by mass — also pulls the Moon toward the Earth. The only difference is distance: the Moon is about 60 Earth-radii away, and the force weakens as the square of that distance ^{[source pending]}.

This is Newton's law of gravitation. Every pair of masses attracts. The force between two point masses is proportional to the product of their masses and inversely proportional to the square of the distance between them. Double the distance, quarter the force. In symbols, the magnitude is $F=G{m}_1{m}_2/r^2$, where $G$ is the gravitational constant, about $6.674\times 10^{-11}$ newton-metres-squared per kilogram-squared ^{[source pending]}.

Now imagine one very large mass, like the Earth, and one small test mass. The large mass produces a gravitational field $\mathbf{g}$ at every point in space. The field tells you the acceleration any test mass would experience if placed there. Near the Earth's surface the field points straight down and has magnitude about 9.8 m/s${}^2$. Far from Earth it points toward Earth's centre and weakens as $1/r^2$.

You can draw field lines to visualise this. Field lines point in the direction of $\mathbf{g}$. They radiate inward toward the mass from all directions. Where the lines are crowded together (close to the mass) the field is strong; where they spread apart (far away) it is weak. The density of field lines crossing a surface falls as $1/r^2$, matching the inverse-square law.

There is also a gravitational potential $\Phi$, a single number at each point in space. The field is the slope of this potential: the steeper the slope, the stronger the field. The field always points from high potential toward low potential. For a point mass, the potential at distance $r$ is $\Phi =-GM/r$. The minus sign reflects that gravity is attractive: masses "roll downhill" toward lower (more negative) potential ^{[source pending]}.

Why introduce a potential at all? The field is a vector (three numbers at each point); the potential is a scalar (one number). Computing one number is easier than computing three. And the potential satisfies simple equations — Laplace's equation in vacuum, Poisson's equation inside matter — that are easier to solve than the vector field equations directly.

One beautiful result ties all this together: the shell theorem. Take a hollow spherical shell of uniform density. Outside the shell, the gravitational field is exactly as though all the shell's mass were concentrated at its centre. Inside the shell, the gravitational field is exactly zero — the contributions from all parts of the shell cancel perfectly ^{[source pending]}.

This theorem explains why you can treat planets as point masses when computing orbits, and why the gravitational field inside a uniform solid sphere grows linearly with distance from the centre rather than following the inverse-square law.

Visual Beginner

Worked example Beginner

A uniform solid sphere has mass $M$ and radius $R$. What is the gravitational field at a distance $r$ from the centre, both inside ($r<R$) and outside ($r>R$)?

Outside ($r>R$): By the shell theorem, the entire sphere acts as a point mass at its centre. The field points inward with magnitude $g=GM/r^2$. This is the familiar inverse-square result.

Inside ($r<R$): Split the sphere into an inner solid sphere of radius $r$ and an outer spherical shell from $r$ to $R$. The outer shell contributes zero field at the point (shell theorem — you are inside it). The inner sphere has mass $M_{\text{inner}}=M(r/R{)}^3$. The field from this inner sphere is:

$$g=\frac{GM(r/R{)}^3}{r^2}=\frac{GMr}{R^3}$$

So inside a uniform sphere the field grows linearly with distance from the centre. At the centre itself the field is zero. At the surface ($r=R$) the inside formula gives $GM/R^2$, matching the outside formula — the field is continuous.

A numerical check: for Earth ($M\approx 6\times 10^{24}$ kg, $R\approx 6.4\times 10^6$ m), the surface field is $g=GM/R^2\approx 9.8$ m/s${}^2$. Halfway to the centre ($r=R/2$), the field is about 4.9 m/s${}^2$ — half the surface value.

Check your understanding Beginner

Formal definition Intermediate+

Newton's law of gravitation. Two point masses $m_1$ and $m_2$ separated by a distance $r$ attract each other with force of magnitude

$$F=\frac{G{m}_1{m}_2}{r^2},$$

directed along the line joining them ^{[source pending]}. In vector form, the force on $m_2$ due to $m_1$ is

$${\mathbf{F}}_{12}=-\frac{G{m}_1{m}_2}{|{\mathbf{r}}_2-{\mathbf{r}}_1{|}^3}({\mathbf{r}}_2-{\mathbf{r}}_1).$$

The gravitational constant $G\approx 6.674\times 10^{-11}\mathrm{N}{\mathrm{m}}^2\mathrm{k}{\mathrm{g}}^{-2}$ ^{[source pending]}.

Gravitational field. The gravitational field produced by a mass distribution is the force per unit mass on a test particle:

$$\boxed{\mathbf{g}(\mathbf{r})=-G\int \frac{\rho ({\mathbf{r}}^{\prime })}{|\mathbf{r}-{\mathbf{r}}^{\prime }{|}^3}(\mathbf{r}-{\mathbf{r}}^{\prime })d^3{\mathbf{r}}^{\prime },}$$

where $\rho ({\mathbf{r}}^{\prime })$ is the mass density. For a single point mass $M$ at the origin, $\mathbf{g}=-GM\hat{\mathbf{r}}/r^2$.

Gravitational potential. Because the gravitational force is conservative (the work done around any closed path is zero), there exists a scalar potential $\Phi$ such that $\mathbf{g}=-\nabla \Phi$ ^{[source pending]}. For a point mass $M$:

$$\Phi (r)=-\frac{GM}{r},$$

with the convention $\Phi \to 0$ as $r\to \infty$. For a continuous distribution, $\Phi (\mathbf{r})=-G\int \rho ({\mathbf{r}}^{\prime })/|\mathbf{r}-{\mathbf{r}}^{\prime }|d^3{\mathbf{r}}^{\prime }$.

Gauss's law for gravity. Taking the divergence of $\mathbf{g}$ and using ${\nabla }^2(1/r)=-4\pi \delta (\mathbf{r})$ yields the differential form ^{[source pending]}:

$$\boxed{\nabla \cdot \mathbf{g}=-4\pi G\rho ,}$$

or equivalently ${\nabla }^2\Phi =4\pi G\rho$ (Poisson's equation). The integral form is

$${\oint }_S\mathbf{g}\cdot d\mathbf{A}=-4\pi G{M}_{\mathrm{enc}},$$

where $M_{\mathrm{enc}}$ is the mass enclosed by the surface $S$. In vacuum ($\rho =0$), ${\nabla }^2\Phi =0$ (Laplace's equation).

Superposition principle. The gravitational field of a collection of masses is the vector sum of the fields of each mass individually. Equivalently, the potential is the sum of the individual potentials. This linearity follows from the linearity of Poisson's equation: if ${\Phi }_1$ is the potential of ${\rho }_1$ and ${\Phi }_2$ is the potential of ${\rho }_2$, then ${\Phi }_1+{\Phi }_2$ is the potential of ${\rho }_1+{\rho }_2$.

Potential of extended bodies. For a bounded mass distribution $\rho ({\mathbf{r}}^{\prime })$ confined to $|{\mathbf{r}}^{\prime }|<a$, the potential at an exterior point $|\mathbf{r}|>a$ admits the multipole expansion:

$$\Phi (\mathbf{r})=-\frac{GM}{r}-\frac{G\mathbf{p}\cdot \hat{\mathbf{r}}}{r^2}-\frac{G{Q}_{ij}{x}^i{x}^j}{2r^5}+\cdots ,$$

where $M=\int \rho d^3{\mathbf{r}}^{\prime }$ is the total mass, $\mathbf{p}=\int \rho {\mathbf{r}}^{\prime }{d}^3{\mathbf{r}}^{\prime }$ is the dipole moment (zero when the origin is the centre of mass), and $Q_{ij}=\int \rho (3x_i^{\prime }{x}_j^{\prime }-r^{\prime 2}{\delta }_{ij})d^3{\mathbf{r}}^{\prime }$ is the quadrupole moment tensor ^{[source pending]}. The monopole term dominates at large distances; the quadrupole correction is important for non-spherical bodies like the oblate Earth.

Counterexamples and cautions

• The shell theorem applies to spherically symmetric distributions. An ellipsoidal mass distribution does not produce a pure inverse-square field outside it; the external potential contains quadrupole and higher multipole terms.
• The potential $\Phi =-GM/r$ is defined up to an additive constant. The convention $\Phi (\infty )=0$ is standard in astrophysics but not universal; in geophysics one sometimes sets $\Phi =0$ at the surface.
• Gauss's law for gravity has the opposite sign to Gauss's law in electrostatics because mass is always positive (gravity is always attractive), whereas electric charge comes in both signs.

Key result: Shell theorem Intermediate+

Theorem (Shell theorem). A uniform spherical shell of mass $M$ and radius $R$ produces: (i) outside the shell, a gravitational field identical to that of a point mass $M$ at the centre; (ii) inside the shell, a gravitational field that vanishes identically ^{[source pending]}.

Proof (direct integration). Consider a thin uniform spherical shell of surface mass density $\sigma$ and radius $R$. Place a test mass at distance $r$ from the centre along the $z$-axis. Partition the shell into thin rings of angular half-width $d\theta$ centred on the axis. A ring at polar angle $\theta$ has radius $R\sin \theta$, area $2\pi R^2\sin \theta d\theta$, and mass $dm=\sigma \cdot 2\pi R^2\sin \theta d\theta$.

Every point on the ring is at distance $s=\sqrt{r^2+R^2-2rR\cos \theta }$ from the test mass. By symmetry the net force points along the $z$-axis. The $z$-component of force from the ring is

$$d{F}_z=-\frac{Gmdm}{s^2}\cos \alpha ,$$

where $\alpha$ is the angle between the line from the ring to the test mass and the $z$-axis, satisfying $s\cos \alpha =r-R\cos \theta$.

Substituting and using $sds=rR\sin \theta d\theta$:

$$d{F}_z=-\frac{Gm\sigma \cdot 2\pi R}{2r^2}\left(1-\frac{R^2-r^2}{s^2}\right)ds.$$

Case (i): $r>R$ (outside). The variable $s$ ranges from $r-R$ to $r+R$. Integrating:

$$F_z=-\frac{Gm\sigma \cdot 2\pi R}{2r^2}{\left[s+\frac{R^2-r^2}{s}\right]}_{s=r-R}^{s=r+R}.$$

At the upper limit: $(r+R)+(R^2-r^2)/(r+R)=2R$. At the lower limit: $(r-R)+(R^2-r^2)/(r-R)=-2R$. The difference is $4R$, giving

$$F_z=-\frac{Gm\sigma \cdot 2\pi R\cdot 4R}{2r^2}=-\frac{GMm}{r^2}.$$

This is exactly the force of a point mass $M=4\pi R^2\sigma$ at the origin.

Case (ii): $r<R$ (inside). Now $s$ ranges from $R-r$ to $R+r$. At the upper limit the bracketed expression evaluates to $2R$. At the lower limit it also evaluates to $2R$. The difference is $2R-2R=0$, so $F_z=0$.

The field vanishes identically inside a uniform shell. $\square$

Proof (Gauss's law). Take a spherical Gaussian surface of radius $r$ concentric with the shell. By symmetry, $\mathbf{g}$ is radial and constant on the surface. For $r<R$: $M_{\text{enc}}=0$, so $4\pi r^{2}g=0$, giving $g=0$. For $r>R$: $M_{\text{enc}}=M$, so $4\pi r^{2}g=-4\pi GM$, giving $g=-GM/r^2$. This derivation shows the shell theorem is a consequence of the inverse-square law plus spherical symmetry ^{[source pending]}.

Worked example: field inside and outside a uniform solid sphere

For a solid sphere of uniform density $\rho$ and radius $R$, the mass enclosed within radius $r$ is $M(r)=(4/3)\pi \rho r^3=M(r/R{)}^3$. By Gauss's law:

$$4\pi r^{2}g(r)=-4\pi GM(r).$$

Outside ($r>R$): $g=-GM/r^2$. Inside ($r<R$): $g=-GMr/R^3$. The field is continuous at $r=R$ and linear inside.

Potential inside and outside a uniform solid sphere

Outside: $\Phi (r)=-GM/r$.

Inside: $d\Phi /dr=GMr/R^3$, so $\Phi (r)=GM{r}^2/(2R^3)+C$. Matching at $r=R$: $-GM/R=GM/(2R)+C$, giving $C=-3GM/(2R)$. Therefore

$$\Phi (r)=-\frac{GM(3R^2-r^2)}{2R^3},r<R.$$

The potential is parabolic inside and hyperbolic outside, continuous at $r=R$, and equal to $-3GM/(2R)$ at the centre.

Escape velocity

The escape velocity from the surface of a spherical body of mass $M$ and radius $R$ is the minimum speed needed to reach infinity with zero residual speed. By energy conservation: $\frac{1}{2}m{v}_{\text{esc}}^2-GMm/R=0$, so

$$v_{\text{esc}}=\sqrt{\frac{2GM}{R}}.$$

For Earth, $v_{\text{esc}}\approx 11.2$ km/s. This result is independent of direction (provided the trajectory does not intersect the surface) and of the mass of the escaping body — a direct consequence of the equivalence of gravitational and inertial mass.

Exercises Intermediate+

Lean formalization Intermediate+

lean_status: none. Mathlib has the divergence theorem for vector fields on ${\mathbb{R}}^n$ (via the general Stokes theorem), the gravitational constant $G$, and basic SI dimension machinery. It does not formalise the gravitational field as a vector-valued integral over a mass distribution, the distributional identity ${\nabla }^2(1/r)=-4\pi \delta (\mathbf{r})$, Poisson's equation ${\nabla }^2\Phi =4\pi G\rho$ for the gravitational potential, or the shell theorem as a named result about spherical shells. The missing pieces are: (1) a definition of the Newtonian potential as a convolution of $\rho$ with $1/r$; (2) integration over $S^2$ with the inverse-square kernel; (3) assembly of these into a gravitational-mechanics framework. Each is within reach of Mathlib's measure-theory and manifold integration machinery but has not been assembled.

Advanced results Master

Potential theory and harmonic functions

The gravitational potential $\Phi$ satisfies Laplace's equation ${\nabla }^2\Phi =0$ in vacuum. Solutions — harmonic functions — satisfy the mean-value property: a function harmonic on a ball equals its average over any smaller concentric sphere. This underlies the uniqueness of the external gravitational field: given the mass distribution inside, the external potential is uniquely determined ^{[source pending]}.

Proposition (Uniqueness of the exterior potential). If $\Phi$ is harmonic on the annulus $a<r<b$ and continuous on the closed annulus, then the values of $\Phi$ on the inner sphere $r=a$ and outer sphere $r=b$ uniquely determine $\Phi$ throughout the annulus.

Proof sketch. The difference of two solutions is harmonic and vanishes on both boundaries. By the maximum principle for harmonic functions, the difference vanishes everywhere. $\square$

This uniqueness result guarantees that the inverse-square field is the only spherically symmetric solution of Laplace's equation in three dimensions. In $n$ dimensions the corresponding field goes as $r^{1-n}$ and the shell theorem generalises: a spherically symmetric shell produces zero interior field if and only if the force law is $F\propto r^{1-n}$ ^{[source pending]}.

Tidal forces and the gradient of the field

The gravitational field $\mathbf{g}$ is not uniform — it varies from point to point. Two nearby test masses at positions $\mathbf{r}$ and $\mathbf{r}+\delta \mathbf{r}$ experience different accelerations. The tidal acceleration is the difference:

$$\delta \mathbf{g}\approx (\delta \mathbf{r}\cdot \nabla )\mathbf{g}.$$

For the field of a point mass $M$, the tidal tensor is

$$T_{ij}=\frac{\partial g_i}{\partial x_j}=GM\left(\frac{3x_i{x}_j}{r^5}-\frac{{\delta }_{ij}}{r^3}\right).$$

This tensor is traceless ($T_{ii}=0$) in vacuum, reflecting $\nabla \cdot \mathbf{g}=0$. The tidal force stretches along the radial direction and compresses in the transverse directions — this is what raises ocean tides under the Moon's differential gravity, and what tears apart satellites that venture inside the Roche limit ^{[source pending]}.

Roche limit. A satellite of density ${\rho }_s$ held together by self-gravity, orbiting a primary of density ${\rho }_p$ and radius $R_p$, is disrupted when the tidal force exceeds its self-gravity. The Roche limit is approximately

$$d_{\text{Roche}}\approx 2.44R_p{\left(\frac{{\rho }_p}{{\rho }_s}\right)}^{1/3}.$$

The numerical coefficient depends on the satellite's rigidity and rotation; 2.44 applies to a fluid satellite ^{[source pending]}.

Lagrangian points

In the restricted three-body problem, two massive bodies orbit their common centre of mass. There exist five Lagrange points where a test particle can remain stationary in the rotating frame. The three collinear points ($L_1$, $L_2$, $L_3$) lie on the line joining the two masses; they are unstable saddles of the effective potential. The two Trojan points ($L_4$, $L_5$) form equilateral triangles with the two masses and are stable when the mass ratio $m_2/m_1\lesssim 0.0385$ (Routh's criterion) ^{[source pending]}.

The effective potential in the rotating frame is

$${\Phi }_{\text{eff}}=-\frac{G{M}_1}{r_1}-\frac{G{M}_2}{r_2}-\frac{1}{2}{\omega }^2{r}^2,$$

where $r_1$, $r_2$ are distances from the two masses and $\omega$ is the orbital angular velocity. The centrifugal term $-\frac{1}{2}{\omega }^2{r}^2$ is the Coriolis-modified contribution from the non-inertial frame ^{[source pending]}. The Lagrange points are critical points of ${\Phi }_{\text{eff}}$.

Connection to general relativity

Newtonian gravity is the weak-field, low-velocity limit of general relativity. The metric of a weak gravitational field can be written as $g_{00}\approx -(1+2\Phi /c^2)$, $g_{ij}\approx {\delta }_{ij}(1-2\Phi /c^2)$, where $\Phi$ is the Newtonian potential. The geodesic equation reduces to $\ddot{\mathbf{r}}=-\nabla \Phi$ in the slow-motion limit. Poisson's equation ${\nabla }^2\Phi =4\pi G\rho$ becomes the time-time component of the Einstein field equations $G_{00}=8\pi G{T}_{00}/c^4$ in the weak-field limit ^{[source pending]}.

Birkhoff's theorem is the general-relativistic analogue of the shell theorem: the exterior spacetime of any spherically symmetric mass distribution is exactly the Schwarzschild solution, independent of the interior structure. This directly parallels the Newtonian result that the exterior field of any spherically symmetric body equals that of a point mass ^{[source pending]}.

Self-gravitating bodies and the Lane-Emden equation

For a self-gravitating polytropic gas $P=K{\rho }^{1+1/n}$, hydrostatic equilibrium $dP/dr=-\rho d\Phi /dr$ combined with Poisson's equation gives the Lane-Emden equation:

$$\frac{1}{{\xi }^2}\frac{d}{d\xi }\left({\xi }^2\frac{d\theta }{d\xi }\right)+{\theta }^n=0,$$

with $\theta (0)=1$, ${\theta }^{\prime }(0)=0$. The first zero of $\theta (\xi )$ gives the stellar radius. Analytic solutions exist for $n=0$ (uniform density), $n=1$, and $n=5$ (infinite-radius isothermal sphere). The cases $n=3/2$ (fully convective) and $n=3$ (radiative) require numerical integration and are central to stellar structure theory ^{[source pending]}.

Synthesis. The shell theorem is the foundational reason that Newtonian gravity outside a spherical body reduces to a one-body central-force problem; the gravitational potential and Gauss's law generalise this insight to arbitrary mass distributions via Poisson's equation. The central insight is that the inverse-square law is uniquely compatible with the vanishing of the interior field, and this is exactly the property that makes Newtonian gravity linear and superposable. The potential-theoretic framework unifies the point-mass field, the shell theorem, Gauss's law, and the multipole expansion into a single coherent structure governed by Laplace's and Poisson's equations. The tidal tensor and Roche limit extend the theory to differential gravitational effects, the Lagrange points connect to the restricted three-body problem, and Birkhoff's theorem in general relativity shows that the same spherical-symmetry principle governs curved spacetime. The generalisation to continuum mechanics and self-gravitating systems completes the bridge from the two-body Kepler problem 09.01.04 to the physics of stars 28.03.01.

Connections Master

• 09.01.01 Kinematics defines position, velocity, and acceleration — the language in which gravitational motion is described. The gravitational field $\mathbf{g}$ is an acceleration field; Newton's second law connects it to the trajectory of a falling body.
• 09.01.02 Newton's laws of motion provide the framework $m\mathbf{a}=\mathbf{F}$ into which the gravitational force law $F=G{m}_1{m}_2/r^2$ fits. The equivalence of gravitational and inertial mass is assumed throughout.
• 09.01.04 The two-body Kepler problem exploits the shell theorem to reduce the orbital dynamics to motion in a central $-GM/r$ potential; the conservation laws that follow are consequences of the conservative nature of gravity established here.
• 09.02.01 The action principle introduces the gravitational potential energy $-GMm/r$ as a term in the Lagrangian; the Euler-Lagrange equations for this Lagrangian reproduce the Kepler orbits, and the shell theorem justifies treating planets as point masses in the Lagrangian framework.
• 09.07.01 Continuum mechanics generalises the gravitational field and potential to mass distributions described by density fields; Gauss's law in its differential form becomes the Poisson equation for the potential, which is the starting point for the theory of self-gravitating fluids.
• 28.03.01 Stellar astrophysics applies the gravitational self-energy formula $-3G{M}^2/(5R)$ and the Lane-Emden equation to model the internal structure of main-sequence stars, white dwarfs, and neutron stars.

Historical and philosophical context Master

Newton's Principia (1687) contains the first proof of the shell theorem in Propositions LXX and LXXI of Book I ^{[source pending]}. The proof is entirely geometric, using Euclidean constructions rather than calculus — which Newton had developed but chose not to publish in this form. The geometric argument compares the force contributions from thin conical solid angles subtended at the test mass by pairs of surface elements on opposite sides of the shell. Newton shows that the inverse-square law produces exact cancellation: the ratio of the areas of the two surface elements (which grows with the square of the distance) exactly compensates the inverse-square weakening of the force. The proof occupied Newton for years and was, by his own account, one of the results that delayed the Principia. Proposition LXXII extends the result from a thin shell to a solid sphere by decomposing it into concentric shells.

Henry Cavendish (1798) performed the first laboratory measurement of $G$ using a torsion balance, confirming the quantitative prediction of the inverse-square law at laboratory scales ^{[source pending]}. Before Cavendish, only the product $G{M}_{\oplus }$ could be determined from orbital mechanics (via Kepler's third law); Cavendish's experiment separated $G$ from $M_{\oplus }$ for the first time, yielding the mass of the Earth. His apparatus — a torsion wire suspending a beam with two small lead spheres, deflected by two large lead spheres brought nearby — measured forces of order $10^{-7}$ newtons and determined $G$ to within about 1%.

Poisson (1813) formulated the differential equation ${\nabla }^2\Phi =4\pi G\rho$, placing Newtonian gravity on the same mathematical footing as electrostatics ^{[source pending]}. Laplace had already established ${\nabla }^2\Phi =0$ in vacuum (1782); Poisson's contribution was to incorporate sources. This potential-theoretic framework was later essential for general relativity, where the Newtonian potential is replaced by the metric tensor and Poisson's equation becomes the Einstein field equations.

Modern precision gravimetry uses free-fall absolute gravimeters (measuring the acceleration of a falling corner-cube reflector in a Michelson interferometer) and superconducting gravimeters (measuring the force needed to levitate a superconducting sphere in a magnetic field) to determine $g$ to parts per billion ^{[source pending]}. These measurements detect tidal variations of order $10^{-6}$ m/s${}^2$, crustal motions, and even the gravitational signal from water-table changes, confirming Newtonian gravity to extraordinary precision at laboratory scales.

A philosophical point: the shell theorem illustrates a general pattern in physics where symmetries of the source produce simplifications in the field. The spherical symmetry of the shell leads to the complete cancellation of the interior field. This pattern repeats in electrostatics (Gauss's law), magnetostatics (Ampere's law), and general relativity (Birkhoff's theorem). The inverse-square law is uniquely compatible with this pattern in three dimensions — a fact that Newton recognised and exploited.

Bibliography Master

• ^{[source pending]} Newton, I., Philosophiae Naturalis Principia Mathematica (1687). Book I, Prop. LXX–LXXII. The original geometric proof of the shell theorem, comparing force contributions from surface elements on opposite sides of a spherical shell. Still read for its elegance; see Chandrasekhar's Newton's Principia for the Common Reader (1995) for a modern commentary.

• ^{[source pending]} Taylor, J. R., Classical Mechanics (University Science Books, 2005). Ch. 8. The standard undergraduate treatment of gravitational field, potential, shell theorem via direct integration, and Gauss's law for gravity. Accessible and thorough.

• ^{[source pending]} Goldstein, H., Poole, C. P. & Safko, J. L., Classical Mechanics, 3rd ed. (Pearson, 2002). Ch. 3. The central-force problem, reduction to one body, effective potential, Kepler's laws derived from the equations of motion.

• ^{[source pending]} Landau, L. D. & Lifshitz, E. M., Mechanics, 3rd ed. (Pergamon, 1976). §35–36. Concise treatment of central-field motion and the Kepler problem, including the hidden $SO(4)$ symmetry.

• ^{[source pending]} Chandrasekhar, S., Ellipsoidal Figures of Equilibrium (Yale University Press, 1969). The definitive treatment of self-gravitating figures, including the bifurcation sequence from Maclaurin spheroids to Jacobi ellipsoids and the onset of secular instability.

• ^{[source pending]} Murray, C. D. & Dermott, S. F., Solar System Dynamics (Cambridge University Press, 1999). Ch. 3. The restricted three-body problem, Lagrange points, stability analysis, and Trojan dynamics. The standard reference for orbital mechanics beyond the two-body problem.

• ^{[source pending]} Misner, C. W., Thorne, K. S. & Wheeler, J. A., Gravitation (Freeman, 1973). §17. The Newtonian limit of general relativity, Birkhoff's theorem as the relativistic shell theorem, and the connection between Poisson's equation and the Einstein field equations.

Meta summary Master

This unit establishes the gravitational field $\mathbf{g}=-GM\hat{\mathbf{r}}/r^2$, the gravitational potential $\Phi =-GM/r$, Gauss's law $\nabla \cdot \mathbf{g}=-4\pi G\rho$, and the shell theorem as the core results of Newtonian gravitational theory. The shell theorem — the exact cancellation of the gravitational field inside a uniform spherical shell — is proved by direct integration and by Gauss's law, and its consequences are developed: the field inside a uniform solid sphere is linear, the exterior field of any spherically symmetric body is that of a point mass, and the escape velocity is $v_{\text{esc}}=\sqrt{2GM/R}$.

The intermediate tier builds the full potential-theoretic machinery: Poisson's equation, the multipole expansion, the potential inside and outside a uniform sphere, and applications to the gravitational self-energy and the cavity problem. Eight exercises span easy to hard, covering the gradient of the potential, Laplace's equation in the distributional sense, Gauss's law applied to non-spherical geometries, the superposition method for cavities, the potential of an extended body (rod), the gravitational self-energy of a sphere, the harmonic-oscillator tunnel result, and the quadrupole expansion.

The master tier extends the theory in four directions. First, potential theory: harmonic functions, the maximum principle, uniqueness of the exterior solution, and the $n$-dimensional generalisation of the shell theorem. Second, differential gravity: the tidal tensor, its tracelessness in vacuum, the Roche limit for tidal disruption of satellites, and the Lagrange points of the restricted three-body problem. Third, the connection to general relativity: the weak-field metric, the reduction of the geodesic equation to Newton's law, and Birkhoff's theorem as the relativistic shell theorem. Fourth, self-gravitating bodies: the Lane-Emden equation and its role in stellar structure.

The historical thread runs from Newton's geometric proof in the Principia (Prop. LXX–LXXII, 1687) through Cavendish's measurement of $G$ (1798), Poisson's formulation of the differential field equation (1813), and modern precision gravimetry. The philosophical thread highlights the pattern whereby symmetries of the source produce simplifications in the field — a pattern that repeats across electrostatics, magnetostatics, and general relativity.