This is a measurement from Ludwig-Maximilians-Universität München and the Max Planck Institute of Quantum Optics, by Lothar Maisenbacher, Vitaly Wirthl, Arthur Matveev, Alexey Grinin, Randolf Pohl, Theodor W. Hänsch and Thomas Udem, published 11 February 2026 in Nature, volume 650. Using 410 nm laser light, they performed Doppler-free one-photon spectroscopy on cryogenic atomic hydrogen and measured the frequency of the 2S–6P transition: 730,690,248,610.79(48) kHz. The relative precision is about 0.7 parts per trillion, roughly 2.5 times better than earlier determinations in atomic hydrogen. The data were taken by 2019; years of evaluating interfering effects followed.

The Standard Model predicts the same frequency: 730,690,248,610.79(23) kHz. The difference is 0.00(53) kHz, that is, coincidence within the measurement uncertainty. The prediction contains the successive quantum-electrodynamic corrections, and the paper says the measurement tests those corrections to 0.5 parts per million, their most precise test so far. The Max Planck press release adds that the measurement saw hadronic contributions and, through vacuum polarization, effects of muons inside ordinary hydrogen, for the first time. The same measurement yields a proton charge radius of 0.8406(15) fm, in agreement with the muonic hydrogen value and 5.5σ away from older hydrogen measurements.

This is the answering face of the fourth ray, the Beauty of Physics. The form goes first: a theory, by its own rules alone, writes down a frequency to twelve digits. Then the world is brought to that place, and a laser answers at the same spot. What a judge cannot not acknowledge here is that the answer lands exactly where the form could not have been otherwise. The digits run to the twelfth or thirteenth place and nothing yet fails to fit. It is not beautiful because high precision is useful (usefulness is another dimension), nor because the number is impressive; if nobody ever needed this frequency, form and world meeting here, with the world not saying no, would still stand. And it is not anyone's taste: it is the part different judges can jointly acknowledge.

It is worth saying what it is not. It is not “the Standard Model is proven right”: the same kind of measurement could at any time have answered otherwise, and that would be physical beauty too (Kepler refusing to let an 8-arcminute error go is the refusing face). 0.00(53) kHz only says that to this precision no crack has appeared. The beauty is not in being right; it is that the world really had the chance to say no, and was asked at the finest place yet.

Its remainder is stated in the paper and the press release: why earlier hydrogen measurements gave a different proton radius is still unexplained; this result only pushes it back. A second remainder belongs to the ray itself: the 0.5 kHz in the uncertainty is not an end, and the next finer measurement will ask whether “nowhere a crack” can hold one digit further. There is no highest instance.

It is worth seeing now because this is the quietest moment on the whole line: theory and experiment meet on one number, and neither side has bent itself to suit the other. The original paper:

Han Qin (秦汉) · @hqinjarsy · SAE Judgment & Aesthetics · Ray 4