The Zeeman Effect

Switch on an electromagnet around a glowing gas and every spectral line splits: a single frequency fans into a cluster whose spacing is a direct read-out of the field. Quantum mechanics supplies the engine — the Zeeman Hamiltonian $\hat H_Z=(\mu_B B/\hbar)(\hat L_z+2\hat S_z)$ and its first-order perturbation treatment are built in aq-qm-zeeman-effect, while the weak-field Landé machinery that governs spin-carrying levels lives in aq-qm-lande-g-factor and aq-qm-weak-field-zeeman-effect. This article takes that operator as given and turns to what a spectroscopist actually does with it: the discovery, the polarization anatomy of the split components, the real wavelength splittings in picometres, and the instruments the effect powers. The transparent case is the one Lorentz solved with a classical electron on a spring, weeks after Zeeman's 1896 measurement: a spin-singlet line ( $S=0$ ) breaks into exactly three components — a $\pi$ line at the original frequency and two $\sigma$ lines displaced by $\pm\mu_B B/h=\pm14.0$ GHz per tesla — carrying a polarization signature that encodes the field's direction. That "normal" triplet is nature's exception, not its rule, but it is where the physics is clean and where the electron's charge-to-mass ratio was first read off a spectrum, months before Thomson's cathode rays. From there the effect is a magnetic-field ruler at every scale: kilogauss sunspot fields mapped from Zeeman-split iron lines, femtotesla brain fields tracked by optically pumped vapor, and the position-dependent shift that supplies the restoring force of a magneto-optical trap. Its companion, The Paschen-Back and Anomalous Zeeman Regimes, takes over the moment spin enters and the triplet fractures.

Key takeaways

A field shifts each $\sigma$ component from line center by $\Delta\nu=\mu_B B/h=14.0$ GHz per tesla, or in wavelength $\Delta\lambda=\lambda^2\mu_B B/(hc)$ — about $12$ pm at $500$ nm and $1$ T.
Only spin-singlet lines ( $S=0$ , effective $g=1$ ) give the clean "normal" triplet: a $\pi$ line ( $\Delta m=0$ ) plus two $\sigma^\pm$ lines ( $\Delta m=\pm1$ ), whose $\pi/\sigma$ polarizations reveal the field direction.
Lorentz's classical bound-electron oscillator predicts the same triplet with spacing $\omega_L=eB/2m_e$ and yields $e/m_e=1.76\times10^{11}$ C/kg straight from the splitting.
The cadmium $643.847$ nm line ( $^1D_2\to{}^1P_1$ ) is the textbook triplet: $\pm19.35$ pm at $1$ T, needing a resolving power $R=33\,270$ that a Fabry-Pérot étalon delivers.
Applications ride on the linear-in- $B$ shift: solar and stellar magnetograms, optically pumped magnetometers, the Zeeman slower, and the magneto-optical trap.
Key idea: a magnetic field converts the projection quantum number $m$ from bookkeeping into measurable energy; for a singlet the splitting is the universal $\mu_B B$ triplet, and its spacing and polarization together measure the strength and direction of the field.

The core insight

A magnetic field turns the orientation quantum number $m$ into measurable energy. For a spin-singlet line the result is a universal three-component pattern spaced by $\mu_B B$ — the "normal" Zeeman triplet — whose splitting measures the field's magnitude and whose $\pi/\sigma$ polarization measures its direction.

Where this lives today

The Zeeman effect is physics' magnetic-field ruler across every scale. Solar and stellar telescopes invert the polarization of Zeeman-split iron lines into vector magnetograms of entire stellar surfaces — Hale's 1908 trick run continuously by modern spectropolarimeters, resolving fields from a fraction of a tesla in sunspots down to a few gauss in the quiet Sun. Optically pumped magnetometers track the Larmor precession of Zeeman sublevels in alkali vapor to femtotesla sensitivity, mapping the human brain's magnetic fields without cryogenics. Every atomic clock spends much of its error budget on Zeeman management, running its clock transition between field-insensitive $m=0$ states while using the $m\neq0$ neighbors as built-in field gauges. And ultracold-atom physics rests on two Zeeman devices: the slower, whose tapered field keeps a decelerating beam resonant, and the magneto-optical trap, whose quadrupole field turns the position-dependent Zeeman shift into a restoring force. The effect Zeeman found in a flame is now the spring constant of nearly every cold-atom experiment on Earth.

The physical picture

A bound electron carries a magnetic moment, and a magnetic field gives each allowed orientation of that moment a different energy — exactly as a compass needle stores least energy pointing north. A level that was degenerate in the field-free atom fans into equally spaced sublevels, and the light emitted between two such fans acquires sidebands. The whole subject is the anatomy of those sidebands: how many there are, how far apart, and how they are polarized. For a singlet the anatomy is universal. Both the upper and lower levels shift by the same $\mu_B B$ per unit of $m$ , so the emitted frequencies collapse onto just three values: the unshifted $\pi$ line and the two $\sigma^\pm$ lines. The polarization of these three is the gift that makes the effect a directional instrument — a fact worth seeing before any algebra.

Figure

Two views of the normal triplet: looking across the field all three components appear, but looking along the field the central π line vanishes and only the circularly polarized σ± survive

The classical roots of this picture matter because they came first. Lorentz modeled the radiating electron as a charge on a spring; a magnetic field makes the transverse oscillation precess, splitting one frequency into three. The spacing of that split is the Larmor frequency $\omega_L=eB/2m_e$ , and because it depends only on $e/m_e$ , Zeeman and Lorentz read the electron's charge-to-mass ratio off the width of a broadened sodium line before anyone had isolated the electron. The second figure is the quantity a working spectroscopist reaches for: the actual triplet of a real line, with the resolving power an instrument must deliver to see it.

Figure

The punchline: cadmium's 643.847 nm line splits into a symmetric triplet with σ components at ∓19.35 pm in a 1 tesla field, demanding a resolving power R = 33 270 that a Fabry-Pérot étalon supplies

The core equations

The Zeeman perturbation itself is recalled, not rebuilt: for a field along $\hat z$ , aq-qm-zeeman-effect establishes the operator whose diagonalization gives every splitting in this article. We quote it and immediately specialize to the transparent singlet case.

\hat H_Z = \frac{\mu_B B}{\hbar}\,(\hat L_z + 2\hat S_z) \;\xrightarrow{\ S=0\ }\; \hat H_Z = \frac{\mu_B B}{\hbar}\,\hat L_z \qquad (1)

where: $\mu_B=e\hbar/2m_e$ is the Bohr magneton; $\hat L_z,\hat S_z$ are the orbital and spin projections; the factor 2 is the spin $g$ -factor $g_s\approx2$ . Physical reading: the field pays orbit and spin different wages for alignment, but with no spin ( $S=0$ ) the operator is proportional to $\hat L_z$ alone and is already diagonal in the $m_\ell$ basis — no anomalous complications survive. The one number every practitioner memorizes is the frequency shift per unit field, the slope that turns a measured splitting into a field strength.

\Delta\nu = \frac{\mu_B B}{h} = 13.996\ \mathrm{GHz\,T^{-1}}\cdot B,\qquad \frac{\mu_B}{hc}=0.4669\ \mathrm{cm^{-1}\,T^{-1}} \qquad (2)

where: $\Delta\nu$ is the displacement of a $\sigma$ component from line center; $h$ is Planck's constant. Physical reading: at $1$ T the shift is a $14$ GHz nudge on an optical line near $5\times10^{14}$ Hz — a fractional effect of $3\times10^{-5}$ , tiny but cleanly resolvable. For an optical instrument the relevant quantity is the wavelength split, which scales as $\lambda^2$ and therefore grows toward the red.

\Delta\lambda = \frac{\lambda^2}{c}\,\Delta\nu = \frac{\lambda^2\,\mu_B B}{hc} \qquad (3)

where: $\lambda$ is the vacuum wavelength of the line; $c$ is the speed of light. Physical reading: because $\Delta\lambda\propto\lambda^2$ , a red line splits much more visibly than a blue one at the same field — one reason infrared magnetography is powerful, and why the cadmium red line was a favored triplet standard. The pattern and its polarizations follow from the electric-dipole selection rule applied to the singlet fans.

\nu \in \{\nu_0,\ \nu_0\pm\mu_B B/h\},\qquad \Delta m_\ell = 0\ (\pi),\quad \Delta m_\ell=\pm1\ (\sigma^\pm) \qquad (4)

where: $\nu_0$ is the zero-field line; $\pi$ components are linearly polarized parallel to $\mathbf B$ (dark when viewed along $\mathbf B$ ); $\sigma^\pm$ components are circularly polarized along $\mathbf B$ and linearly polarized perpendicular to it. Physical reading: however many sublevel pairs a transition offers, a common spacing collapses them to exactly three frequencies — the normal triplet. Historically the decisive relation is the classical one: the split measures the Larmor frequency, and the Larmor frequency is $e/m_e$ in disguise.

\omega_L = \frac{eB}{2m_e} = \frac{\mu_B B}{\hbar},\qquad \frac{e}{m_e} = \frac{2\omega_L}{B} = 1.7588\times10^{11}\ \mathrm{C\,kg^{-1}} \qquad (5)

where: $\omega_L$ is the classical precession (angular) frequency of the electron cloud about $\mathbf B$ . Physical reading: measuring $\omega_L$ (from the observed split) and $B$ gives $e/m_e$ directly — the "weighing of the electron by light" that Zeeman and Lorentz performed in 1896–97.

Derivation ALorentz's classical oscillator gives the triplet and weighs the electron

The goal is to reproduce, from nothing but Newton's law and the Lorentz force, the entire normal-Zeeman phenomenology: a line splitting into a central component at the original frequency and two flanks displaced by a spacing that depends only on the charge-to-mass ratio of the radiating particle. This is the calculation Lorentz completed within weeks of Zeeman's measurement, genuinely optics-owned classical electron theory, so we build it here rather than recall it. The payoff is twofold: the three-line pattern with its polarizations, and a laboratory value of $e/m_e$ from an optical splitting.

Step 1. Model the radiating electron (charge $-e$ , mass $m_e$ ) as bound by a restoring force $-m_e\omega_0^2\mathbf r$ and place it in $\mathbf B=B\hat z$ . Newton's law with the Lorentz force $-e\,\dot{\mathbf r}\times\mathbf B$ gives $m_e\ddot x=-m_e\omega_0^2x-eB\dot y$ , $m_e\ddot y=-m_e\omega_0^2y+eB\dot x$ , and $m_e\ddot z=-m_e\omega_0^2z$ . Why the axis decouples: the field only twists the transverse motion; the $z$ -oscillation is untouched and radiates at the unshifted frequency $\omega_0$ — the future $\pi$ component.

Step 2. Combine the transverse equations into the circular coordinate $u=x+iy$ . The pair collapses to the single complex equation $m_e\ddot u - ieB\dot u + m_e\omega_0^2 u=0$ . Why circular coordinates: a magnetic field is rotationally symmetric about $\hat z$ , so its natural normal modes are left- and right-circular, not Cartesian — the algebra rewards choosing them.

Step 3. Seek $u=e^{-i\omega t}$ ; the characteristic equation is $m_e\omega^2+eB\omega-m_e\omega_0^2=0$ , with roots $\omega=\pm\omega_0-\dfrac{eB}{2m_e}$ to first order in the small ratio $eB/m_e\omega_0$ . The two circular modes therefore radiate at $\omega_0\pm\omega_L$ with $\omega_L=eB/2m_e$ — the future $\sigma^\pm$ components. Why only a linear shift survives: the quadratic (diamagnetic) correction is smaller by $eB/m_e\omega_0\sim10^{-6}$ at laboratory fields, invisible against the linear split.

Step 4. Read off the physics. Three frequencies $\{\omega_0,\omega_0\pm\omega_L\}$ ; the central one polarized along $\mathbf B$ , the flanks circular about it — the polarization map confirmed by Zeeman. Inverting the spacing, $e/m_e=2\omega_L/B=1.7588\times10^{11}$ C/kg, a particle thousands of times lighter than the hydrogen ion and negative in charge. Why this was revolutionary: the number arrived from an optical line width in 1896, months before Thomson's cathode-ray deflections, and it agreed. *Verified in verify_aq-optics-zeeman-effect.py: the characteristic polynomial roots $\pm\omega_0-eB/2m_e$ by sympy; $\omega_L=\mu_B B/\hbar$ ; $e/2m_e=8.794\times10^{10}$ rad s $^{-1}$ T $^{-1}$ ; $e/m_e=1.7588\times10^{11}$ C/kg.*

Go deeper · Derivation B · 2 minThe quantum normal triplet and its polarization fingerprint

Derivation BThe quantum normal triplet and its polarization fingerprint

The aim is to show that the quantum treatment of a singlet-to-singlet transition reproduces Lorentz's three lines exactly, pinning each polarization to a selection rule rather than a classical hand-wave. We recall the Zeeman operator by name from aq-qm-zeeman-effect, evaluate it on singlet states where it reduces to $\mu_B B\,m_\ell$ , and let the electric-dipole rule $\Delta m_\ell=0,\pm1$ prune the frequencies. What is owned here is the spectroscopic bookkeeping: which components appear, where, and how they are polarized in each viewing geometry.

Step 1. For $S=0$ the states $|n\,\ell\,m_\ell\rangle$ already diagonalize $\hat H_Z=(\mu_B B/\hbar)\hat L_z$ , so the upper level shifts by $\mu_B B\,m'_\ell$ and the lower by $\mu_B B\,m''_\ell$ , each fanning into equally spaced rungs with the same spacing $\mu_B B$ . Why the spacing is common: with $g=1$ for every singlet, upper and lower fans share one slope — the equality that will collapse the differences.

Step 2. The photon energy is $h\nu=h\nu_0+\mu_B B(m'_\ell-m''_\ell)$ , and the dipole rule allows only $\Delta m_\ell=m'_\ell-m''_\ell\in\{0,\pm1\}$ . The set of emitted frequencies is therefore exactly $\{\nu_0,\nu_0\pm\mu_B B/h\}$ — three members, no matter how many sublevel pairs the two manifolds contain. Why three and not more: a $d\to p$ line connects fifteen sublevel pairs, but the common spacing forces every allowed difference into one of three buckets.

Step 3. Assign polarizations. The $\Delta m_\ell=0$ ( $\pi$ ) transition radiates like a dipole oscillating along $\mathbf B$ : linearly polarized $\parallel\mathbf B$ , and dark when viewed down the field axis (a dipole does not radiate along itself). The $\Delta m_\ell=\pm1$ ( $\sigma^\pm$ ) transitions radiate like charge circling in the transverse plane: circular along $\mathbf B$ , linear $\perp\mathbf B$ when viewed across it. Why this is a directional instrument: the splitting encodes $|\mathbf B|$ while the polarization pattern encodes the orientation of $\mathbf B$ — two dials in one measurement.

Step 4. Attach the scale. At $1$ T the $\sigma$ components sit $\mu_B B/h=14.0$ GHz off center; at $500$ nm that is $\Delta\lambda=\lambda^2\mu_B B/hc=11.7$ pm, and at the cadmium red line $643.847$ nm it is $19.35$ pm. Why cadmium became the standard: $^1D_2\to{}^1P_1$ is a genuine singlet–singlet transition, so it gives an unusually clean triplet that early interferometry could calibrate against. *Verified in verify_aq-optics-zeeman-effect.py: the three-frequency set $\{0,\pm\mu_B B\}$ ; $\mu_B B/h=13.996$ GHz/T; $\Delta\lambda(500\ \mathrm{nm},1\ \mathrm{T})=11.67$ pm; Cd $643.847$ nm split $19.35$ pm at $1$ T.*

Go deeper · Derivation C · 1 minThe Zeeman slower turns the effect into a brake for atoms

Derivation CThe Zeeman slower turns the effect into a brake for atoms

The goal is to derive the operating principle of the Zeeman slower — the device that decelerates a thermal atomic beam to loading speeds for a magneto-optical trap — as a direct application of the splitting formula. The challenge: a laser tuned to decelerate a fast atom falls out of resonance as the atom slows and its Doppler shift shrinks, and the fix is a tailored magnetic field whose Zeeman shift tracks that vanishing Doppler shift. We compute the field required to compensate a sodium atom's Doppler shift — an owned application number that anchors the cold-atom enterprise.

Step 1. A counter-propagating laser resonant with a moving atom is Doppler-shifted by $\Delta\nu_D=v/\lambda$ . For sodium at $v=1000$ m/s on the $589$ nm $D$ line this is $\Delta\nu_D=1.698$ GHz — a large detuning that must be matched by the laser or by shifting the atom's transition. Why the beam must be slowed at all: a $1000$ m/s beam is far too fast to capture; a trap loads at tens of m/s, so the beam must lose most of its speed while staying resonant.

Step 2. The Zeeman effect shifts the atomic transition by $\Delta\nu_Z=\mu_B B/h$ (with an effective $g\approx1$ for the cycling transition). Setting the Zeeman shift to compensate the Doppler shift, $\mu_B B/h=v/\lambda$ , gives the field $B=hv/(\mu_B\lambda)=0.121$ T for $v=1000$ m/s. Why a spatially varying field: as the atom decelerates along the beam its Doppler shift shrinks, so $B(z)$ must taper to keep $\mu_B B(z)/h+ \text{detuning}$ locked on resonance — the signature decreasing-field profile of a slower.

Step 3. The resonance condition along the tube, $\delta_L + kv(z) = \mu_B B(z)/\hbar$ , with constant deceleration $v(z)=\sqrt{v_0^2-2az}$ produces the ideal tapered profile $B(z)=B_0\sqrt{1-z/L_0}$ ; a field of order a tenth of a tesla holds resonance across the whole deceleration. Why the numbers are modest: $0.12$ T is an ordinary solenoid, which is exactly why the Zeeman slower became the standard front end of atomic-beam experiments. *Verified in verify_aq-optics-zeeman-effect.py: $\Delta\nu_D(1000\ \mathrm{m/s},589\ \mathrm{nm})=1.698$ GHz; compensation field $B=h v/(\mu_B\lambda)=0.1213$ T.*

Limiting cases

case	result	meaning
$B\to0$	line restored, degeneracy back	with no axis, $m$ is unmeasurable; the field is what makes orientation cost energy
singlet, $S=0$	triplet at $\nu_0,\nu_0\pm\mu_B B/h$	the normal Zeeman effect: universal, $g=1$ , the only case Lorentz's classical theory reproduces exactly
spin present, $S\neq0$	anomalous multiplets, $g_J\neq1$	the triplet fractures into four, six, or more lines — developed in The Paschen-Back and Anomalous Zeeman Regimes
$\mu_B B\sim\Delta E_{fs}$	weak-field picture fails	the field rivals fine structure; the crossover is owned by aq-qm-paschen-back-effect
$\mu_B B\sim\Delta E_{hfs}$	nuclear decoupling	field also unlocks $\mathbf I$ from $\mathbf J$ (Back-Goudsmit); teased by title in Hyperfine Structure

Common trapEvery spectral line splits into a clean Zeeman triplet

It does not. The three-line "normal" pattern requires a spin-singlet transition ( $S=0$ , effective $g=1$ ) so that the upper and lower levels fan with identical spacing. That is the exception: any level with unpaired spin has $g_J\neq1$ because the spin $g$ -factor is $\approx2$ , and its line splits into an "anomalous" multiplet of four, six, or more components with unequal spacings. One-electron atoms and the alkalis — hydrogen, sodium, cesium — are always in this camp; the tidy cadmium and zinc triplets come from paired-spin two-electron configurations. The name "normal" is a fossil of the years before electron spin, when the triplet was all that classical theory could explain and everything else looked pathological. Reach for the triplet only after confirming $S=0$ ; otherwise the Landé $g_J$ machinery of aq-qm-lande-g-factor governs the pattern.

Worked examples

Example 1 — reading a sunspot's magnetic field (Hale's 1908 discovery). A magnetically sensitive iron line, Fe I at $630.25$ nm with effective Landé factor $g_{\mathrm{eff}}=2.5$ , is observed in a sunspot where $B=0.3$ T. What splitting does the Zeeman effect produce, and can a solar spectrograph see it?

Step 1. Frequency shift of each $\sigma$ component: $\Delta\nu=g_{\mathrm{eff}}\,\mu_B B/h=2.5\times13.996\ \mathrm{GHz/T}\times0.3\ \mathrm{T}=10.50$ GHz.

Step 2. Convert to wavelength with equation (3): $\Delta\lambda=\lambda^2\Delta\nu/c=(630.25\ \mathrm{nm})^2\times10.50\ \mathrm{GHz}/c=13.9$ pm, so the two $\sigma$ lines straddle line center by $\pm13.9$ pm and are separated by $27.8$ pm.

Step 3. Resolving power: $R=\lambda/\Delta\lambda=630.25/0.0139\approx4.5\times10^4$ — within reach of a grating spectrograph, easier still through the circular-polarization signature. A $0.3$ T sunspot splits the Fe I 630.25 nm line by 13.9 pm per σ component (27.8 pm tip-to-tip): a kilogauss field made visible — the measurement with which Hale first proved sunspots are magnetic in 1908.

Example 2 — the cadmium triplet as a field gauge at half a tesla. The Cd $643.847$ nm line ( $^1D_2\to{}^1P_1$ ) is observed transverse to $\mathbf B=0.5$ T. Predict the pattern and the étalon it needs.

Step 1. Singlet check: both terms have $S=0$ , so a true normal triplet forms at any laboratory field.

Step 2. Shift: $\Delta\lambda=\lambda^2\mu_B B/hc=9.68$ pm, so the components sit at $643.847$ nm and $643.847\pm0.00968$ nm; transverse viewing shows an unshifted $\pi$ line flanked by two $\sigma$ lines.

Step 3. Resolving power: $R=\lambda/\Delta\lambda=643.847/0.00968=66\,540$ — double the $33\,270$ needed at $1$ T, but still comfortably met by a Fabry-Pérot étalon. A symmetric triplet at $643.847\pm0.00968$ nm, its $\pi/\sigma$ polarization fixing the field direction; halving the field halves the split and doubles the required resolving power.

Physics of the portrait

The Zeeman effect is one of the few discoveries where the experiment, the theory, and the first great application all arrived within a decade, and where a spectral line width quietly weighed a particle no one had yet seen.

Origins · 2 minPieter Zeeman and how this was discovered

Pieter Zeeman

1865–1943 · Leiden → Amsterdam · the electron weighed by light

Zeeman found the effect that bears his name with apparatus every well-equipped laboratory already owned — a sodium flame, a Ruhmkorff electromagnet, a Rowland concave grating — and the stubbornness to look for a change others had reasoned away (Faraday tried and failed for want of resolution in 1862). In the autumn of 1896, working at Leiden alongside Hendrik Lorentz, Zeeman watched the yellow sodium $D$ lines broaden the instant the magnet fired; a durable legend says he ran the experiment against his supervisor's instructions and was reprimanded, a story he never confirmed. Lorentz supplied the theory almost immediately: a bound "ion" oscillating in a field must emit a triplet with precisely the polarizations Zeeman then verified line by line, and the width of the split delivered the charge-to-mass ratio of the vibrating particle — about $10^{11}$ C/kg, a thousandfold lighter than a hydrogen ion and negative in charge. They had weighed the electron with light, months before J. J. Thomson's cathode-ray measurement. The 1902 Nobel Prize in Physics — only the second ever awarded — went jointly to Zeeman and Lorentz. At Amsterdam, Zeeman built a laboratory whose great electromagnet stood on a vibration-isolated concrete island; there he pursued the four- and six-component "anomalous" patterns he had glimpsed right after the triplets, which classical theory could not touch — the opening problem of the companion article.

History

1896

— Zeeman observes the magnetic broadening of the sodium D lines at Leiden

1897

— Lorentz's electron theory predicts the triplet and its π/σ polarizations, confirmed on the blue cadmium line; e/m extracted from the splitting

1898

— Preston photographs anomalous quartets and sextets the triplet theory cannot produce

1902

— Nobel Prize in Physics to Zeeman and Lorentz, the second ever awarded

1908

— Hale reads kilogauss magnetic fields in sunspots from Zeeman-split iron lines

1912

— Paschen and Back reach the strong-field regime where anomalous patterns collapse toward triplets

1982

— Phillips and Metcalf demonstrate the Zeeman slower for a sodium beam

1987

— Raab, Prentiss, Cable, Chu and Pritchard build the first magneto-optical trap

Going further

Texts. Foot, Atomic Physics, Ch. 1 and Ch. 5 (Zeeman effect, normal and anomalous) · Demtröder, Laser Spectroscopy, Vol. 1, §on Zeeman spectroscopy and polarization · Bransden & Joachain, Physics of Atoms and Molecules, Ch. on atoms in external fields · Hecht, Optics, §on the Zeeman effect · Foukal, Solar Astrophysics, on Zeeman magnetometry.

Provenance. The Zeeman Hamiltonian $\hat H_Z=(\mu_B/\hbar)(\mathbf L+g_s\mathbf S)\cdot\mathbf B$ and its perturbation treatment are recalled by name from aq-qm-zeeman-effect; the weak-field Landé $g_J$ machinery that governs spin-carrying levels is owned by aq-qm-lande-g-factor and aq-qm-weak-field-zeeman-effect. A verifier should confirm: (1) $\mu_B=9.274\times10^{-24}$ J/T $=5.788\times10^{-5}$ eV/T and $\mu_B/h=13.996$ GHz/T (CODATA); (2) the classical triplet spacing $\omega_L=eB/2m_e$ and $e/m_e=1.7588\times10^{11}$ C/kg; (3) the Cd 643.847 nm split $19.35$ pm at $1$ T and the required $R=33\,270$ ; (4) the Zeeman-slower compensation field $0.121$ T at $1000$ m/s; (5) the selection rules $\Delta m_\ell=0,\pm1$ and the $\pi/\sigma$ polarization assignments. No non-standard numbers are introduced.

Prerequisites. Electron Spin and the Spin Magnetic Moment · Fine Structure and Spin-Orbit Coupling · Quantum Numbers and Orbital Angular Momentum. Leads to. The Paschen-Back and Anomalous Zeeman Regimes · Hyperfine Structure · The Stark Effect.

Why does the magnetic field pick out a z-axis, and why does m suddenly matter?

An isolated atom has no preferred direction — states of different $m$ are related by rotations and must share an energy, which is exactly why $m$ is pure bookkeeping in a field-free atom. Switching on $\mathbf B=B\hat z$ creates a physical axis against which orientation can be measured, so the degeneracy finally lifts and different $m$ costs different energy. That is the whole content of "the field makes $m$ observable": the splitting you see is the energy the atom now pays for pointing its moment one way rather than another. Every Zeeman quantum number, $m$ or later $m_J$ , is a projection along the field the experimenter chose.

How could a line width possibly reveal the electron's charge-to-mass ratio?

Because the classical split has a clean formula: $\Delta\omega=\omega_L=eB/2m_e$ . The spacing between the shifted components depends on the applied field $B$ (known) and on $e/m_e$ and nothing else — no atomic constants, no spectroscopic model. Zeeman measured the broadening, Lorentz supplied $\omega_L=eB/2m_e$ , and inverting gave $e/m_e\approx1.8\times10^{11}$ C/kg — enormous next to the electrolytic $e/m$ of a hydrogen ion, implying a particle roughly two thousand times lighter, with the polarization handedness fixing the charge as negative. A complete measurement of the electron's specific charge, done in 1896 with a magnet and a grating.

If the normal triplet needs S = 0, why did Zeeman see a clean effect in sodium, which has spin?

He did not, at first — he saw only a broadening, not resolved components. The 1896 resolution could not expose sodium's underlying structure, which actually splits the $D_1$ and $D_2$ lines into four and six components respectively. Early confirmations of the clean triplet came from lines that approximate a triplet at the available resolution and, decisively, from genuine singlet lines like cadmium's red $643.847$ nm. Within a year Preston's sharper photographs revealed the sodium quartets and sextets, and the "anomalous" Zeeman effect became the standing puzzle of pre-spin atomic physics — the subject of the companion article.

How do astronomers measure fields when the Zeeman split is far below the linewidth?

By using polarization instead of raw splitting. The two $\sigma$ components carry opposite circular handedness, so subtracting the right- and left-circularly polarized spectra (the Stokes $V$ profile) isolates a signal proportional to the line-of-sight field $B_\parallel$ even when the split is a small fraction of the Doppler-broadened width — converting a resolution problem into a far more tractable signal-to-noise one. This is why solar magnetograms can map fields of a few gauss in the quiet Sun despite thermal broadening thousands of times larger than the Zeeman shift.

Does the Zeeman slower need the atom's exact Landé g-factor?

For the cycling transition used to slow and trap, an effective $g\approx1$ captures the compensation to the accuracy that matters, which is why $B=hv/(\mu_B\lambda)=0.12$ T at $1000$ m/s is a good design number. In detail the relevant shift is the difference of $g_J\mu_B B m_J$ between the specific sublevels of the cycling transition, and the field profile is tuned experimentally to hold resonance as the atom decelerates. The physics is entirely the Zeeman shift of this article; the anomalous $g_J$ arithmetic that fixes the exact factor is developed in The Paschen-Back and Anomalous Zeeman Regimes.

Is there a magnetic analogue of this article for electric fields?

Yes — the Stark effect, in which an electric field rather than a magnetic one lifts the $m$ -degeneracy and shifts spectral lines. The structures differ: a magnetic field couples to the magnetic moment and gives shifts linear in $B$ with the clean $\pi/\sigma$ polarization pattern, whereas an electric field couples to the electric dipole and gives shifts that are typically quadratic in $E$ (linear only for hydrogen's degenerate levels). The two are siblings; the electric case is developed in The Stark Effect. What they share is the idea organizing this article: an external field converts an orientation quantum number into measurable energy.