The competence of the original 184-inch cyclotron to make mesons had been disputed. When it was designed in 1939 theorists knew too little about the process of creation to determine the threshold of energy; and even had they been wiser, experimental uncertainty about the mass of the cosmic-ray meson would have left the matter doubtful.

Artist's conception of the 184-inch cyclotron with a fanciful beam emerging toward the observer.

In arguing the need for the 184-inch cyclotron to the Rockefeller Foundation in 1940, Lawrence gave high priority to the study of nuclear forces, in which mesons had been implicated. Taking the meson mass to be 150 me (electron masses), he assured the Foundation that he would make the elusive particle in the proposed machine, which would certainly give alpha particles of 150 MeV. And that, he said, was the likely production threshold. He reasoned that the energy supplied to a heavy nucleus by an incoming alpha particle or deuteron would be quickly shared among the target nucleons; to have much chance of creating a meson in such a process, the incident particle would need an energy very much larger than his estimate of the rest mass, 75 MeV, equivalent to 150 me. A light target nucleus, with fewer opportunities for sharing, appeared to be more promising, although the exigencies of momentum conservation then doubled the ante. The threshold Lawrence proposed to cross in 1940 was therefore 2mc2 (m the meson mass). He wrote the Rockefeller Foundation: "We are planning everything with the idea that we shall certainly be able to produce alpha particles at least 150 million volts energy."

Lawrence's sanguine estimate of 1940 referred to a machine in potentia. After the war, with the thing in hand, he hedged his bet. Writing Groves about the $170,000 that the MED had given to complete the synchrocyclotron, he declined to say what might happen when it went on. "The fundamental interactions of heavy particles are not well enough known to predict what may happen (for example, the important question of the production of mesotrons is unpredictable at present)." Apparently he then thought prospects sufficiently remote that no special hunt was made for mesons until the end of 1947, a year after the great machine first delivered alpha particles at just short of 400 MeV. By then theorists had made it appear likely that the "4000 ton atom smasher" -- as journalists liked to call the synchrocyclotron, creating the impression that it split nuclei by falling on them -- could indeed make mesons, and with something to spare.

At the University of Chicago, Edward Teller and W. G. McMillan observed that at the very high energies in question bombarded nuclei become almost transparent. When a collision does occur, it involves only one nucleon from the target, whose kinetic energy might contribute to the stock necessary for meson production. For the most favorable case, McMillan and Teller found that the threshold energy might be as little as 95 MeV. As they said, "this result is radically different" from the 206 MeV that would be required by using the latest value of the cosmic-ray meson's mass (as determined by W. B. Fretter of Berkeley's physics department) and estimating in Lawrence's manner. Two of Oppenheimer's students improved the calculation of the collision cross-section and concluded that neutrons stripped from 195 MeV deuterons might do the trick. If the kinetic energy of the incident nucleon within the deuteron or alpha particle were also taken into account, the threshold might sink below the admittedly crude estimate from Chicago.

On the strength of these considerations, the group responsible for the detection of nuclear particles by photographic emulsions undertook to look for meson tracks on plates bombarded with deuterons and 380 MeV-alphas (95 MeV/nucleon). The group had some idea of the likely appearance of the tracks from the beautiful photomicrographs made by Powell and Occhialini. But their results also raised the threshold for meson production. The Berkeley materialists could not make the cosmic-ray particle of 202 me (the µ or muon) without first making its parent ¼ particle (pion), the mass of which they estimated at above 250 me, perhaps more than 350 me. They required a minimum incident energy possibly as high as the 150 MeV per nucleon originally estimated by Lawrence and possibly beyond the reach of the synchrocyclotron.

The leader of the Laboratory's film group was Eugene Gardner, who had taken his Ph.D. under Lawrence in 1943 with a thesis on calutron ion sources and had then worked at improving the alpha process of electromagnetic separation at Oak Ridge. Gardner's job was to adapt Powell's technique to the abundant flux from the synchrocyclotron. He and his group worked closely with Eastman Kodak, to which Lawrence had become consultant late in 1945, on the development of a domestic film to compete with the nuclear emulsion created by Ilford, Ltd. for Powell's group in Bristol. For 18 months Kodak supplied new test plates and new methods of development while Gardner's group tried to find the best position and exposure time for the plates within the cyclotron tank. Their notebooks record constant disappointment and, occasionally, exhilaration. A note of November 30, 1946 by one of the group reads: "Please don't take this plate out of the microscope without calling me. This is a beautiful plate ... Can't go home -- searching this plate is too exciting." The beautiful plate had been made by Ilford.

Schematic of the arrangement within the tank of the 184-inch synchrocyclotron devised by Gardner et al. to detect negative mesons, the paths of which are bent around the shielding and into the plate by the accelerator's field, 1948.

Identification of a track rests on its length and the number and distribution of its grains. These parameters, which characterize the charge, mass, and velocity of the responsible particle, may also be affected by irregularities in the plate or in its development. Interpretation of a unique event may plausibly be disputed. During the first six months of operation of the big synchrocyclotron, the film group found a few odd tracks unlike those for protons, alpha particles, and deuterons, and unlike one another, which they took to be the work of mesons. On careful examination they appeared to be artifacts of Kodak film. "Every time we run a set of plates in the cyclotron we look for meson tracks," Gardner wrote James Chadwick. "But so far we haven't found any."

In the autumn of 1947 the group began to search for mesons directly. A thin target of carbon was placed within the cyclotron tank to intercept the beam of alpha particles; adjacent to the target stood a stack of photographic plates so positioned that negatively charged mesons created in the carbon and projected in the direction of the beam would be directed into them by the field of the big magnet. In December, after three months of development and exposures, the group had found no mesons. There was nothing wrong with their experimental setup. They lacked some relevant experience.

Immediately after celebrating the new year, G. C. M. (Giulio) Lattes wrote Lawrence for permission to work at the Laboratory. He would come on Rockefeller Fellowship and with the approval of the AEC to teach the film group what he had learned during two years' collaboration with Powell. He arrived in February 1948, preceded by a package of Ilford plates. They were exposed in Gardner's apparatus and developed according to Lattes's recipe which differed from Berkeley practice. Then Lattes who knew what to look for, discovered what the Berkeley group had not been able to find. As Gardner reported the result: almost immediately after his arrival, Lattes had "made the Bristol technique successful in detecting for the first time man made mesons."

Photomicrograph of the track of one of the first mesons found by Gardener and Lattes, 1948.

The plates became prolific, and the team practiced; a good scanner might find as many as fifty meson on a plate in an hour. They found muons as well as pions. They had more raw data than they could handle. While developing the emulsion technique Gardner had prepared plates for analysis by physicists outside Berkeley, and he continued the practice even as it became likely that the emulsions contained something interesting. Powell, a frequent recipient, later called attention to this uncommon liberality: "Many laboratories all over the world are greatly indebted to our American colleagues of Berkeley, California, for the very generous way in which, promptly and without conditions, they have exposed photographic plates to the particles provided by their machines."

Among the results of Gardner's group was the answer to the question that had perplexed the meson experiment from the start. They found no pions when they sent 165 MeV protons against their carbon target; the yield at 200 MeV was but one percent, and that at 300 MeV less than half of the yield at 345 MeV. For alpha particles, the number of pions observed fell by two thirds as the energy of projection declined from 380 to 342 MeV, and to fewer than seven events (less than one percent of the maximum) at 260 MeV. The 184-inch machine as originally designed probably could not have made mesons.

The University and the AEC officially announced the artificial production of mesons on March 9, 1948. Typically, Lawrence used the opportunity to argue the case for the next generation of accelerators. "To exploit fully the knowledge which the meson may provide, it will be necessary to construct super-giant cyclotrons." Time reported the discovery and hinted that the study of mesons might "lead in the direction of a vastly better source of atomic energy than the fission of uranium." The University of California explained that the work of Gardner and Lattes opened the newest new age since the discovery of fission. "It might roughly be compared with the discovery of America."

Gardner mastered the photographic technique and oversaw its adaptation to the cyclotron despite a progressive enervation that contrasted sadly with the robust machine he served. He died in November 1950 of berylliosis contracted from dust that he had inhaled while drilling beryllium for calutron test fittings at Oak Ridge.

The detection of the charged pion encouraged the search for evidence for a neutral one, which, according to a theory published by Oppenheimer and his students in 1948, might be the source of the electromagnetic radiation associated with the soft component of cosmic rays. Neither the hypothetical meson nor the two photons into which theory required it to decay would leave tracks in emulsions or cloud chambers. Experimental detection of neutral mesons accordingly would depend on subtle analysis and circumstantial evidence. Two different groups at the Laboratory, proceeding in two different ways, nonetheless managed to clinch the case.

Robert Serber, Laboratory theorist, writing for a photographer shortly after the announcement of the discovery of machine-made mesons by Gardner and Lattes in February, 1948.

In September, 1949 R. Bjorkland, W. E. Crandell, B. J. Moyer, and H. York observed pairs of electrons made by photons from a target struck by protons from the 184-inch synchrocyclotron. Their main experimental findings were a rapid increase of pairs beyond proton energies of 175 MeV and a telling relationship between two curves. The curves gave the yield of pairs in the detector as a function of the energy of the photons producing them for two different cases: photons leaving the target in the direction of the proton beam and (after reversal of the synchrocyclotron fields) photons leaving against the direction of the beam. The two plots could be transformed into one another on the assumption that all the photons came from particles traveling along the beam at one third the velocity of light and emitting equal numbers of photons in opposite directions. The possibility of this Doppler superposition and the threshold energy of around 175 MeV, as well as behavior incompatible with other possible sources of the photons, made a strong circumstantial case that the radiator was a decaying neutral pion.

Further evidence -- coincidences between photons liberated in opposite directions -- could not then be obtained at the synchrocyclotron because its concrete shielding blocked the necessary observations. In experiments concluded in April 1950, J. Steinberger, W. Panofsky, and J. Steller avoided this difficulty by procuring their photons from a target struck by a collimated beam of x rays from McMillan's electron synchrotron. The photons from the pion decay were examined at right angles to the x-ray beam by coincidence counters. The energy distribution of the electron pairs made by the photons and the yield of coincidences at various angular separations of the counters agreed with the assumption that the photons arose from the decay in flight of a relativistic neutral pion. Thus was confirmed the existence of the first particle detected at an accelerator, and the analysis of the unmarked graves where nonionizing neutral particles disintegrate into nonionizing radiation.