When physics learned to photograph the invisible

The bubble chamber was invented in 1952 by Donald Glaser, a young physicist at the University of Michigan. The story that he was inspired by bubbles rising in a glass of beer is almost certainly too good to be true, but the principle is close. Heat a liquid above its boiling point while keeping it under pressure, then release the pressure. For a moment the liquid is superheated and unstable. A charged particle passing through will leave a trail of tiny bubbles along its path, which can be photographed before the whole chamber boils.

Place the chamber in a magnetic field and the tracks curve: positive particles one way, negative the other, slow ones in tight spirals. From the curvature, length and branching of the tracks, physicists could reconstruct a particle’s charge, momentum and fate. A trained eye could read a good photograph the way a tracker reads footprints in snow.

Liquid hydrogen and an industrial scale

Glaser’s first chambers were small enough to hold in one hand. Luis Alvarez’s group at Berkeley scaled the idea up dramatically, building chambers filled with liquid hydrogen, whose single-proton nuclei made collisions easier to interpret. Liquid hydrogen in large quantities, surrounded by electrical equipment, was also a serious explosion hazard, and building these machines came to resemble a chemical-plant project more than a tabletop experiment. The 72-inch chamber, completed in 1959, produced photographs by the million.

The physics changed along with the scale. The photographs revealed dozens of short-lived particles, called resonances, and the sheer number of them forced theorists to look for order underneath. In 1962, Murray Gell-Mann and Yuval Ne’eman’s classification scheme predicted a particle no one had seen, the omega-minus. In early 1964, a team at Brookhaven found it, in a bubble chamber photograph, a single decisive event from among roughly a hundred thousand pictures. It remains one of the great examples of a theory predicted, tested and confirmed by looking at a picture.

Glaser received the Nobel Prize in 1960, Alvarez in 1968. At CERN, a heavy-liquid chamber called Gargamelle found the first evidence of weak neutral currents in 1973, a key confirmation of the theory that unified the electromagnetic and weak forces.

Every discovery passed first through human eyes.

The scanners

Every one of those photographs had to be examined. Laboratories employed teams of scanners, many of them women hired without physics degrees, who searched the images for interesting events, traced tracks on projection tables and recorded their coordinates for computer analysis. The best scanners developed an eye for rare topologies that physicists relied on. Every discovery passed first through human eyes.

Their work is rarely mentioned in histories of the period, but it echoes the experience of the women computers at Harvard who cataloged photographic plates of the sky half a century earlier: indispensable, skilled, poorly paid and mostly uncredited. In both cases, the arrangement was presented as a division between routine labor and real science, and in both cases the routine labor required judgment that could not easily be specified. Alvarez’s group tried hard to mechanize it, with semi-automatic measuring machines and computer programs, and those efforts helped push physics toward large-scale computing.

Two ways of knowing

The historian Peter Galison has argued that twentieth-century experimental physics was divided between two traditions. The image tradition, of cloud chambers, emulsions and bubble chambers, sought the single telling picture, a golden event that could be pointed at. The logic tradition, of counters and electronics, built its arguments from statistics over very large numbers of events that no one ever looked at individually.

For a long time, each side had what the other lacked. A photograph gave you the whole event, every track, with no prior decision about what mattered. But it could not be selective. Electronic detectors could be triggered to record only certain kinds of collisions, at the cost of throwing away everything the trigger had not been told to look for.

The end of the photograph

The bubble chamber’s limit was fundamental. It could take only a few pictures per second, and it could not be triggered at all, since the liquid had to be prepared before the particles arrived. As accelerators produced collisions at ever higher rates, physicists turned to electronic detectors. By the 1980s the great chambers had been retired. The scale of the shift is hard to overstate: the Large Hadron Collider produces bunch crossings 40 million times per second, tens of millions of times the rate of a bubble chamber’s camera, and only a tiny selected fraction of them is ever stored.

Something was lost along the way. A bubble chamber photograph showed an event whole, in a form a person could look at and understand. Today the decision about what to keep is made in microseconds by hardware and software, and increasingly by machine learning, trained on what physicists have already seen. The old question has returned in a new form: what does a selection process quietly leave out? The scanners at least could notice something odd and call someone over.

The images that remain, in archives and on the walls of physics departments, are some of the most beautiful records of the invisible that science has made.

Sources

• Donald Glaser, Nobel lecture, 1960
• Luis Alvarez, Nobel lecture, 1968
• Barnes et al., “Observation of a Hyperon with Strangeness Minus Three”, Physical Review Letters, 1964
• Hasert et al., “Observation of neutrino-like interactions without muon or electron in the Gargamelle neutrino experiment”, Physics Letters B, 1973