The Lamb
Willis Lamb walked into the wrong seminar in 1947, heard a rumor, measured a number, and accidentally launched quantum electrodynamics.
Lamb was at Yale, a young experimentalist with a reputation for building instruments that measured things nobody thought measurable. He was studying hydrogen — the simplest atom, two particles, proton and electron, held together by electromagnetism. In Dirac's theory, the energy levels of hydrogen were determined by a single quantum number, n. The 2S₁/₂ and 2P₁/₂ levels should be degenerate — the same energy. Identical. No splitting.
Dirac's theory was, at that point, the best theory of the atom. It combined quantum mechanics with special relativity. It predicted the fine structure of hydrogen to astonishing accuracy. And it predicted that 2S₁/₂ and 2P₁/₂ should have exactly the same energy.
Lamb wasn't planning to test this. He was supposed to be studying hyperfine structure — the splitting caused by the interaction between the electron's spin and the proton's spin. But in April 1947, he attended a seminar at Princeton where Polkar and Kusch discussed some preliminary results that suggested the 2S₁/₂ and 2P₁/₂ levels might not be degenerate. There might be a splitting. A tiny one. But there.
Lamb went back to Yale and built an apparatus that measured the difference. It was radiofrequency spectroscopy — a technique that had been developed during the war for radar research, applied by Willis Lamb to the simplest atom in the universe. He excited hydrogen atoms with an electric discharge, used a filter to select atoms in the 2S state, and then hit them with a radiofrequency field. If the frequency matched the energy difference between 2S₁/₂ and 2P₁/₂, the atoms would transition, and Lamb would see it in the detector.
He measured a splitting of about 1000 megacycles per second — roughly one thousandth of an electron volt. The levels were not degenerate. Dirac's theory was wrong.
This was a small number. A tiny number. A number that might have been lost in experimental error, buried under the noise floor of any other measurement. But Lamb had built the instrument, and the instrument said: the splitting is real. The levels are not degenerate. Something is lifting the 2S₁/₂ level slightly above the 2P₁/₂ level.
The discovery sent a shock wave through theoretical physics. Feynman was at Cornell, Schwinger at Columbia, Bethe at Columbia, Oppenheimer at Princeton — the entire west coast and east coast of American physics converged on this single, tiny number. The theory had to account for it. If the 2S₁/₂ and 2P₁/₂ levels were split, the reason had to be in the interaction between the electron and the vacuum.
Bethe was the first to calculate it. In a single afternoon — the kind of afternoon that changes physics — he sat down with pencil and paper and computed the Lamb shift using non-relativistic quantum mechanics and a cutoff on the virtual photon energies. He got 1040 megacycles. Lamb's measurement was about 1000 megacycles. The agreement was close enough to convince everybody that they were on the right track.
The physical picture is this: the electron in the hydrogen atom is constantly emitting and reabsorbing virtual photons. These virtual photons jitter the electron's position. The electron is "smearing out" over a small region of space, and because the Coulomb potential isn't perfectly linear, this smearing changes the energy. The S state, which has a non-zero probability of finding the electron at the nucleus, is affected more than the P state, which has a node at the origin. Hence the S level is shifted up relative to the P level.
But Bethe's calculation was approximate. It used a non-relativistic treatment and an ad hoc cutoff. A full treatment required relativistic quantum field theory — a theory that didn't exist yet in a complete form.
What happened next is one of the great collaborative dramas in the history of physics. Feynman, Schwinger, and Sin-Itiro Tomonaga independently developed the full theory of QED. Freeman Dyson proved that their approaches were equivalent. The theory predicted the Lamb shift exactly — not approximately, not with a cutoff, but from first principles.
And it worked.
Willis Lamb received the Nobel Prize in Physics in 1955 "for his observations on the fine structure of the hydrogen spectrum." The observation was a tiny shift in a spectral line, measured with an instrument of his own design, triggered by a rumor heard at the wrong seminar.
The Lamb shift was the crack in Dirac's theory that let the light of QED in. It was the experimental anomaly that demanded a new theory. And the new theory didn't just explain it — it predicted every subsequent measurement to a precision that borders on the miraculous.
The story of the Lamb shift is the story of how an experimentalist, following a hunch, measured a number so small that it should have been invisible, and changed physics forever.
The Lamb shift is 1057 megacycles per second. If you divide by Planck's constant, it's about 4.37 × 10⁻⁶ electron volts. A whisper. A breath. A number so small that if it were a distance, it would be the width of a hydrogen atom measured in Planck lengths.
And it was the number that broke Dirac.