Preparation can change which absorption is possible
Atoms in different magnetic sublevels need not absorb the same polarized light. Selective absorption followed by spontaneous emission can repeatedly transfer population between sublevels. Optical pumping uses this cycle to make the distribution depart from its thermal pattern, often in a long-lived ground state.
The point is not simply that light excites atoms. It is that repeated cycles plus selection rules create an imbalance that changes later absorption or fluorescence. The atom, the polarized beam and the radiation leaving the atom must all be included when accounting for angular momentum.
Read the evidence: A. Kastler (1957), Optical Methods of Atomic Orientation and of Magnetic Resonance, JOSA 47, 460–465 — author abstractR. A. Bernheim (1965), Optical Pumping: An Introduction, chapters 2–3 — Berkeley-hosted excerpt
An ideal dark state is a conditional statement
In a suitable transition and polarization scheme, population can accumulate in a sublevel that the pumping light no longer excites. The sample can then become less absorbing even though the atoms have not left it. This is an optical signature of redistribution, provided other causes of transmission changes are controlled.
The absorbing and dark states depend on the complete transition structure. Different hyperfine branches, imperfect polarization, off-resonant excitation, collisions and wall encounters can return atoms to absorbing states. One cannot infer perfect polarization merely from a brighter transmitted beam.
Read the evidence: A. Kastler (1957), Optical Methods of Atomic Orientation and of Magnetic Resonance, JOSA 47, 460–465 — author abstractR. A. Bernheim (1965), Optical Pumping: An Introduction, chapters 2–3 — Berkeley-hosted excerpt
A minimal rate model makes the competition explicit
The illustrative equation below describes the fraction in one effective dark population. Pumping transfers the remaining bright fraction into it at an effective rate; relaxation returns the dark fraction toward a specified equilibrium value. Both rates are nonnegative and the fractions lie between zero and one.
It predicts competition between preparation and relaxation, not coherent evolution or an isotope-specific pumping efficiency. It omits explicit excited-state populations, branching ratios, optical coherence, radiation trapping and spatial transport. Turning off pumping restores equilibrium in this model; increasing pumping alone does not calculate a real apparatus’s optimum.
Read the evidence: R. A. Bernheim (1965), Optical Pumping: An Introduction, chapters 2–3 — Berkeley-hosted excerpt
Drive one transition, observe another
In double resonance, a radiofrequency field drives transitions between magnetic sublevels while an optical channel detects their effect. Brossel and Bitter’s 1952 paper describes polarized excitation of mercury, unequal excited-state populations and a change in emitted-light polarization at magnetic resonance.
Ground-state optical pumping is a related extension, not the same mercury measurement. Kastler’s 1957 summary distinguishes optical detection in excited states from population concentration in ground-state sublevels, including sodium. A radiofrequency resonance can disrupt a prepared distribution and thereby change an optical signal without producing a visible-frequency radio photon.
Read the evidence: J. Brossel and F. Bitter (1952), A New Double Resonance Method for Investigating Atomic Energy Levels, Phys. Rev. 86, 308–316 — abstractA. Kastler (1957), Optical Methods of Atomic Orientation and of Magnetic Resonance, JOSA 47, 460–465 — author abstract
The chronology separates a proposal from an observation
The laboratory’s history places Kastler’s ground-state pumping proposal in 1950 and the sodium atomic-beam observation by Brossel in 1952. It also records earlier unsuccessful attempts and the importance of irradiation and relaxation. The later development of vapor cells was an experimental advance, not an automatic consequence of a level diagram.
Kastler’s 1966 award concerns optical methods for studying atomic radiofrequency resonances. It should not be reduced to a laser-cooling claim, and it does not mean every optical signal measured a single atom. Preparation, optical detection and suppression of relaxation were distinct parts of the method.
Read the evidence: Laboratoire Kastler Brossel, Birth of Double Resonance and Optical Pumping — institutional historyAlfred Kastler — official 1966 award facts
Compare selection, cooling and fluorescence carefully
Optical pumping can prepare internal populations without cooling the center-of-mass motion. Laser cooling must also account for momentum transfer and diffusion. Fluorescence readout can reveal a bright or dark condition, but its meaning depends on the actual level scheme and detector; a dark interval is not a universal state label.
The linked coupled-spin Lab explores an effective transition-selection model, and the quantum-jump Lab explores a separate fluorescence record. Neither integrates the pumping-and-relaxation model above or recreates Kastler’s sodium apparatus. Their comparison shows which additional preparation and detection mechanisms a historical reconstruction would need.
Read the evidence: A. Kastler (1957), Optical Methods of Atomic Orientation and of Magnetic Resonance, JOSA 47, 460–465 — author abstractJ. Brossel and F. Bitter (1952), A New Double Resonance Method for Investigating Atomic Energy Levels, Phys. Rev. 86, 308–316 — abstract