A successful model still had a vulnerable prediction
The Dirac description of an electron in an ideal Coulomb field accounted for important features of hydrogen fine structure. It also made two particular excited states coincide in energy. That coincidence was a sharper test than the general statement that hydrogen has discrete levels.
Lamb and Retherford’s 1947 report distinguishes this relative placement from the separation between the two P states. Finding a discrepancy in one comparison did not mean all the fine-structure predictions had failed. The experiment asked which relation needed correction.
Read the evidence: W. E. Lamb Jr. and R. C. Retherford (1947), Fine Structure of the Hydrogen Atom by a Microwave MethodLamb and Retherford (1947), original paper scan hosted by the University of Maryland, pp. 241–243
Turn a short lifetime into a detectable loss
The apparatus produced a beam containing metastable excited hydrogen. Atoms that survived to a metal target generated an electron signal. A transition into a rapidly decaying state removed atoms from that detectable metastable population, reducing the downstream current.
Microwaves and a magnetic field made this loss a spectroscopic probe. The initial report describes holding microwave frequency fixed and recording the signal change against magnetic field. Calling that record a simple direct measurement of absorbed microwave power would misidentify the detector.
Read the evidence: Lamb and Retherford (1947), original paper scan hosted by the University of Maryland, pp. 241–243
Name the energy difference and the historical precision
The expression below defines the positive zero-field energy difference normally associated with this hydrogen Lamb shift, after separating the relevant structure. The letters identify the excited S and P levels with total electronic angular momentum one half. The microwave measurement accesses field-dependent transitions from which the relative placement is inferred.
The 1947 report placed the S level higher by about one gigahertz. Its figure caption explicitly says that shifting comparison curves by that amount was not a best fit. Replacing that historical statement with a later, more precise number would give the early experiment a precision it did not claim.
Read the evidence: Lamb and Retherford (1947), original paper scan hosted by the University of Maryland, pp. 241–243
Linewidth and calibration are part of the result
The original caption lists several contributors to the observed widths: the short lifetime of the P states, hyperfine structure, strong microwave excitation and magnetic-field inhomogeneity. It also notes limitations in field calibration. A broad dip can locate a resonance, but its center and uncertainty require an apparatus model.
A sharper display is not a correction to the atom. Improving resolution helps distinguish nearby features; a theoretical shift changes their expected positions. Those two operations should remain distinct when a modern teaching spectrum is compared with the historical argument.
Read the evidence: Lamb and Retherford (1947), original paper scan hosted by the University of Maryland, pp. 241–243
An anomaly is not an isolated proof of all QED
Bethe’s 1947 electromagnetic energy-level calculation supplied a theoretical response to the discrepancy. The comparison motivated treatment of the atom together with its electromagnetic interaction beyond the ideal Dirac–Coulomb problem. It should not be summarized as one unexplained number validating every later high-order calculation.
The hydrogen and coupled-spin Labs offer baseline levels and resolution comparisons. They do not compute the Lamb correction or reconstruct Lamb and Retherford’s beam. The fuller QED direction remains a separate topic; this article establishes how to read the experimental anomaly and its stated limits.
Read the evidence: H. A. Bethe (1947), The Electromagnetic Shift of Energy Levels, Phys. Rev. 72, 339Lamb and Retherford (1947), original paper scan hosted by the University of Maryland, pp. 241–243