01

The Invisible Force

How can one piece of metal grab another — without touching it?

Hold two magnets close. Feel them snap together — or push apart — before they even touch. Something is reaching across empty space. That something is the magnetic field.

Humans have used lodestones (naturally magnetic rocks) for navigation for over 2,500 years. But we only understood what magnetism actually is in the 20th century, when quantum mechanics revealed it emerges from the spin of electrons deep inside atoms.

Drag the magnet anywhere. Watch the field lines follow.

Magnetic field lines around a bar magnet. They emerge from N (red) and re-enter at S (green) — always forming closed loops.

Those curved lines aren't drawn on anything. They're a map of direction: at every point in space, the line shows where a compass needle would point. Pack the lines closer together, and the field is stronger there.

Grade 4 version: The magnet shoots invisible arrows out in all directions. The arrows from the N (north) end fly through the air, loop around, and dive back into the S (south) end — like a river that always flows in circles and never runs out.

Why Some Things Are Magnetic Basics

Iron, nickel, cobalt — and their alloys like steel — are ferromagnetic. A plastic spoon won't stick to your fridge; an iron nail will. The difference is buried in atomic structure. We'll get there in Section 3.

Ferromagnetic

Iron, Nickel, Cobalt, Gadolinium. Atomic magnetic moments cooperate, lock into "domains," and stay aligned permanently.

Paramagnetic

Aluminium, platinum, liquid oxygen. Weakly attracted while in an external field — but lose all magnetism the moment you remove it.

Diamagnetic

Bismuth, graphite, water, frogs. Weakly repelled by strong fields. A live frog was levitated in 1997 using 16 T in a Bitter magnet.

Antiferromagnetic

Manganese oxide, chromium. Adjacent atoms anti-align — every atom is magnetic, but neighbors cancel each other exactly.

Earth's magnetic field is generated by convecting liquid iron in the outer core — a self-sustaining geodynamo. It has reversed polarity hundreds of times in Earth's history. The last reversal was ~780,000 years ago. We're overdue, and it appears to be weakening in patches (the South Atlantic Anomaly). A reversal takes thousands of years and doesn't flip instantly.
02

Field Lines: Maps of Invisible Space

Drag the poles. See the topology of the field.

Michael Faraday invented field lines in the 1840s — not because he had equations, but because he needed a picture. He had no formal mathematics training. Yet his insight — that electricity and magnetism could be understood as fields pervading space — became the conceptual foundation for Maxwell's equations and then Einstein's general relativity.

Drag N (red) and S (green) poles independently. Try placing them on top of each other — or switching to two N poles.

What Field Lines Mean — and Don't Mean

Field lines are mathematical constructs that encode real physics:

Gauss's Law for Magnetism ∇ · B = 0 The divergence of B is zero everywhere. No magnetic monopoles exist. Field lines always form closed loops — they cannot begin or end on any "magnetic charge."

This stands in stark contrast to electric fields, where ∇·E = ρ/ε₀: field lines start and end on charges. But magnetic field lines have no source point. They loop forever. Despite decades of searching in cosmic ray detectors, ancient rocks, and particle accelerators, no magnetic monopole has ever been detected.

Why monopoles matter: Paul Dirac showed in 1931 that if even one magnetic monopole exists anywhere in the universe, it would mathematically require electric charge to be quantized — explaining why all charges come in multiples of e. Grand Unified Theories predict monopoles should exist. Their absence (or extreme rarity) remains one of physics' deepest open puzzles.

Field Strength Falls as 1/r³ High School+

A bar magnet is a magnetic dipole. Far from it, its field drops much faster than gravity (1/r²) or a point charge (1/r²). This is why a magnet that lifts a paper clip from 1 cm barely affects it from 5 cm away.

Magnetic Dipole Field (along the axis) B = (μ₀ / 4π) · (2m / r³) m = magnetic dipole moment (A·m²), r = distance from center, μ₀ = 4π × 10⁻⁷ T·m/A (permeability of free space).
03

Atomic Origins

Magnetism lives inside the electron — in two distinct ways.

Zoom into any iron atom. Inside every atom, electrons do two things that each generate magnetism: they orbit the nucleus (a tiny current loop), and they possess an intrinsic "spin" — a quantum property with no classical analog, but one that acts exactly like a tiny bar magnet.

Electron orbiting the nucleus (blue arrow = orbital moment μ_L) and carrying spin (red arrow = spin moment μ_S). Both contribute to the atom's total magnetic moment.

Orbital Magnetic Moment Classical picture

A moving charge is a current. An orbiting electron is a tiny current loop. Any current loop creates a magnetic dipole moment proportional to its angular momentum:

Orbital Magnetic Moment μ_L = −(e / 2m_e) · L L = orbital angular momentum (quantized as ℏ√(l(l+1))), e = electron charge, m_e = electron mass. The ratio e/2m_e is the orbital gyromagnetic ratio.

Spin Magnetic Moment Quantum

Spin is not the electron physically rotating — that picture breaks down (an electron would have to spin faster than light to produce its observed magnetic moment). It's an intrinsically quantum mechanical property. Yet it couples to magnetic fields exactly as a classical bar magnet would:

Spin Magnetic Moment μ_S = −g_s · (e / 2m_e) · S S = spin angular momentum = ℏ/2. g_s ≈ 2.00232 — the "anomalous" g-factor. Dirac's 1928 equation predicted g = 2 exactly; the extra 0.00232 comes from virtual photon interactions (QED).
Why g ≈ 2 is profound: Dirac wrote a relativistic equation for the electron and did not put spin in by hand — it popped out automatically as a mathematical consequence of combining quantum mechanics with special relativity. The factor of 2 was a shock. The extra 0.00232... is the most precisely tested prediction in science, confirmed by QED to 12 decimal places.

Magnetic Domains: Why Most Iron Isn't Magnetic Basics

In an unmagnetized iron bar, atoms are grouped into domains — regions where spins are aligned. But different domains point in different directions. They cancel out. Net magnetism: zero.

Apply an external field: domains aligned with the field grow at the expense of others. Eventually most moments point the same way. The bar is magnetized.

Magnetic domains. Each arrow = net spin direction of a region. "Apply Field" watches them align. "Heat" randomizes them above the Curie temperature.

Curie Temperature: Above 770°C for iron, thermal energy overwhelms the alignment tendency. The material abruptly becomes paramagnetic. This phase transition — where long-range order vanishes — is a textbook example of spontaneous symmetry breaking.

Why Only Some Atoms Are Magnetic Quantum Mechanics

Pauli's exclusion principle forces electrons into pairs with opposite spins in each orbital. Paired electrons cancel. An atom is only magnetic if it has unpaired electrons — found in partially filled d or f shells.

Iron (Fe, Z=26) has the electron configuration [Ar] 3d⁶ 4s². By Hund's rules, the six 3d electrons distribute as 5 spin-up + 1 spin-down, giving 4 unpaired electrons and a large magnetic moment of ~2.2 Bohr magnetons per atom. This is why iron, not copper or silver, is the archetypal magnet.

04

Electricity's Hidden Twin

Moving charges create magnetic fields. Changing fields create currents. They are inseparable.

Before the 1820s, electricity and magnetism were thought to be completely separate. Then Hans Christian Ørsted noticed something odd: when he ran current through a wire near a compass during a lecture, the needle swung. He had discovered that every moving charge creates a magnetic field.

Solenoid cross-section. Each circle = one loop of wire, with current into (X) or out of (dot) the page. Their combined field is nearly uniform inside and weak outside.

Solenoid Interior Field B = μ₀ · n · I n = turns per meter, I = current in amperes. An ideal solenoid has perfectly uniform B inside and zero outside. MRI machines use superconducting solenoids at ~1.5–3 T.

Biot-Savart Law Undergrad

For any current-carrying wire, the field contribution from an infinitesimal segment is:

Biot-Savart Law dB = (μ₀ / 4π) · (I dl × r̂) / r² The × is the cross product — B is perpendicular to both the current direction dl and the displacement r to the field point. Right-hand rule: curl fingers from dl toward r̂; thumb points along dB.

Faraday's Law: The Reverse Key insight

Change the magnetic flux through a loop, and a voltage appears. Every generator, transformer, wireless charger, and guitar pickup runs on this:

Faraday's Law of Induction EMF = −dΦ_B / dt Φ_B = ∫B · dA is the magnetic flux. The negative sign is Lenz's law: the induced current always opposes the change that created it. Faraday discovered this empirically in 1831; Maxwell gave it its final mathematical form.

The Lorentz Force Why particles spiral

A charge moving through a magnetic field feels a force perpendicular to its velocity — so it curves but doesn't speed up. In a uniform field, it moves in a perfect helix. This is why:

Lorentz Force F = q (E + v × B) For charge q at velocity v in fields E and B. The magnetic term v × B is always perpendicular to v — so the magnetic force does no work on the particle. It changes direction but not speed.
MRI in 60 words: A 3 T superconducting solenoid aligns hydrogen nuclei in your body. Radiofrequency pulses at the Larmor frequency (ω = γB) tip them out of alignment. As they precess back, they emit RF signals. Different tissues relax at different rates (T1, T2 relaxation times). A 3D Fourier transform converts those signals into your anatomy.
05

The Deep Mathematics

Maxwell, Einstein, Dirac — three revolutions that unified it all.

Maxwell's Equations

In 1865, James Clerk Maxwell combined Faraday's insights with Ampere's law and added one critical term (the displacement current ε₀ ∂E/∂t). The result was four equations that completely describe classical electromagnetism — and predicted the existence of electromagnetic waves traveling at speed c, which Hertz confirmed experimentally 22 years later.

I — Gauss's Law (Electric) ∇ · E = ρ / ε₀ Electric field lines diverge from charges. ρ is the free charge density.
II — Gauss's Law (Magnetic) — no monopoles ∇ · B = 0 Magnetic field lines never diverge. No magnetic monopoles. The field always loops.
III — Faraday's Law ∇ × E = −∂B / ∂t A changing B field curls the E field. This is how transformers and generators work.
IV — Ampere-Maxwell Law ∇ × B = μ₀J + μ₀ε₀ ∂E / ∂t Current AND changing electric flux create a curling B field. Maxwell's ∂E/∂t term was the key addition — without it, the equations were inconsistent.

Take the curl of III, substitute IV, and the wave equation drops out:

Electromagnetic Wave Equation ∇²E = μ₀ε₀ · ∂²E / ∂t² Wave speed: c = 1 / √(μ₀ε₀) = 2.998 × 10⁸ m/s. Maxwell wrote in 1865: "We can scarcely avoid the inference that light consists in the transverse undulations of the same medium which is the cause of electric and magnetic phenomena."

Magnetism is Relativistic Electricity Special Relativity

This may be the most surprising fact about magnetism: it is not a separate force from electricity. It is what electricity looks like to a moving observer.

Consider a wire with current (electrons drifting left). A positive charge at rest nearby feels no force — the wire is electrically neutral. Now set that charge in motion to the right. Length contraction (special relativity) causes the negative charges in the wire to appear Lorentz-contracted — denser — so the wire looks negatively charged. The charge feels attraction. That "electric force" in the moving frame is what we call "magnetic force" in the lab frame.

In relativity, E and B are not separate. They are components of a single antisymmetric rank-2 tensor, the electromagnetic field tensor:

Electromagnetic Field Tensor F^μν = ∂^μA^ν − ∂^νA^μ A^μ = (φ/c, A) is the 4-potential (scalar + vector potential combined). A Lorentz boost mixes the E and B components — what one observer calls purely magnetic, another calls partly electric. They are projections of the same geometric object.
In this language, Maxwell's four equations collapse into just two:
∂_ν F^μν = μ₀ J^μ  and  ∂_[μ F_νλ] = 0
The full power and elegance of classical electromagnetism, in two tensorial lines.

The Dirac Equation: Spin Emerges Automatically QFT

In 1928, Paul Dirac sought a relativistic quantum equation for the electron. He didn't put spin in by hand. It emerged automatically from the mathematics of combining quantum mechanics and special relativity. And so did the electron's magnetic moment, with g = 2 exactly.

Dirac Equation (iγ^μ ∂_μ − m) ψ = 0 γ^μ are four 4×4 Dirac matrices. ψ is a 4-component spinor. Two solutions describe the electron (spin up/down); the other two describe a positively-charged partner — the positron, which Dirac predicted in 1928 and was discovered in 1932.

QED: The Most Precise Theory Ever Built Quantum Field Theory

Dirac's prediction g = 2.000... was later refined by quantum electrodynamics. Julian Schwinger computed the first correction in 1948:

Anomalous Magnetic Moment (Schwinger term) g/2 = 1 + α/(2π) + O(α²) + ... α ≈ 1/137.036 is the fine-structure constant. The full QED series computed to 10th order in α gives g/2 = 1.00115965218059... — matching experiment to 12 significant figures. The most precisely tested prediction in all of science.
Muon g-2 (2025 frontier): The muon's g-factor is harder to calculate (heavier particle = more QCD hadronic corrections). The Fermilab g-2 experiment currently shows a ~4σ deviation from the Standard Model prediction. If confirmed, this would be the first crack in the Standard Model — pointing to new particles or forces we haven't discovered yet.

Topology Enters: Beyond Equations Modern Physics

In condensed matter, magnetic fields and quantum mechanics together produce phases of matter characterized not by symmetry but by topology — mathematical properties that survive continuous deformation:

Quantum Hall Effect

2D electrons in strong B form Landau levels. Hall conductance is quantized at σ = νe²/h — a topological invariant (the Chern number). Won the 1985 Nobel Prize.

Magnetic Skyrmions

Topological "knots" in spin textures. Like a knot in a rope, they can't be smoothly untied. Proposed as ultra-dense, low-power magnetic memory bits.

Weyl Semimetals

Materials with band crossings that act like magnetic monopoles in momentum space — not real space. The chiral magnetic anomaly is macroscopically observable.

Berry Phase

A quantum phase acquired when a spin traces a loop in parameter space. Equivalent to a magnetic field in momentum space (the Berry curvature). Foundation for modern topological band theory.

The full picture: The thing you stick to your fridge is the macroscopic trace of quantum mechanics — electron spin, Pauli exclusion, exchange interactions, and spontaneous symmetry breaking in iron's 3d shell, all cooperating across 10²³ atoms to produce a force you can feel with your fingers. Every fridge magnet is a relativistic, topological, quantum mechanical object. It just doesn't look like one.