skip to primary navigationskip to content

Nuclear Physics in the Modern Cavendish

Prof. O.R. Frisch

When I took up my appointment at the Cavendish Laboratory in 1947 as successor to Sir John Cockroft, it was still pioneering days for nuclear physics. Scintillation counters were just coming into general use, whereby gamma rays could be recorded with much greater efficiency and more precise timing. The primitive pulse height analysers ("kick sorters") which I had developed during the War were soon enormously improved by George Hutchinson, Denys Wilkinson and others, and so were many other techniques, created during the work on the atom bomb.

Through these techniques it became possible to explore the angular distribution of the particles emitted from atomic nuclei under bombardment, and also the angular correlation of two particles emitted in succession by the same nucleus. A great deal was learned in that way about the excited states (eg their spins and parities) of light nuclei, with Sam Devons as the driving spirit. Tony French and Denys Wilkinson pursued similar work, and it is sad to record that all three have since left Cambridge; but it is, after all, the job of the Cavendish to provide the rest of the world with good physicists.

There was a cyclotron built about ten years previously by a team including Albert Kempton, then the only one of Lord Rutherford's collaborators still at the Cavendish. It was still a useful machine, accelerating deuterons to about 10 MeV (million electron volts), and interesting work was being done by Kempton's students; but better cyclotrons had been built since, and a new tank was designed and built which would bring it once more into the front line. But the cost of a new building was too high, and the intensity of the radiation would have required more stringent shielding. The metallurgists, our next-door neighbours, opposed the project, particularly when it was discovered that one of our shielding tanks had leaked! After refilling it we installed indicators to show our neighbours that the tanks were full and giving protection, but our image had suffered.

Two more machines (1 MeV and 2 MeV) were operating, commercial versions (made by Philips of Eindhoven) of the Cockroft-Walton machine which had revolutionised nuclear physics in 1932. They were housed in a specially-built hall, nicknamed the cinema because it had no windows; they were well beloved by journalists and TV producers, being their idea of the shape of science to come, with their tall columns of polished metal electrodes and the crashing sparks they could be provoked to generate. But it was already known that higher voltages of greater stability could be produced by a fast-moving electrified belt inside a large steel tank filled with compressed gas, a design by Van de Graaff. Such a machine was being built by Edward Shire and came into operation at about 3 MeV a few years after my arrival.

All three machines were used for studying the energy levels of light nuclei until the late 1950's when the decision was taken to get rid of these machines, which were no longer competitive with bigger and better ones elsewhere. The two Philips machines were shipped to South Africa where a previous student of ours had become professor in Johannesburg, but the Van de Graaff machine nobody wanted and, sadly, it had to be scrapped; so too was the octagonal tower (originally housing a Link trainer for airmen) which had for so long been a landmark, looking a bit like a gasholder on the museum site. The "cinema" got windows and floors; it later housed practical classes and, on the ground floor, the new lecture theatre.

The cyclotron was offered back to the government department that had paid for it, the Department of Scientific and Industrial Research (DSIR, later renamed SRC, Science Research Council). They gave it to another Cavendish man, Prof. William Burcham at Birmingham University, who gave it a new lease of life. It is now a strong-focusing machine (with spiral grooves in its pole pieces) with much higher energy and intensity, and with a source of polarised protons and deuterons which allows new types of research to be done with it.

By the 1950s nuclear physics had become a mopping-up operation; while it was still possible to attack a great many questions, and to pin down more precise knowledge was important, there was no longer the feeling of worlds to be conquered. That feeling then became attached to high-energy physics, and when machines in the GeV (thousand million electron volt) region were being built elsewhere in the early 1950's, we considered our chances of catching up in this field. The construction of a linear electron accelerator was begun by Devons and Hereward and continued by Burcham and Kempton. The planned energy of 0.4 GeV would have been enough to make pi-mesons. But when (Sir) Nevill Mott succeeded Sir Lawrence Bragg as Cavendish Professor in 1953, one of his first actions after consulting various people concerned was to terminate the project, having concluded that it was "too little and too late".

A few years later a much more ambitious project of building an electron synchrotron was promoted by Sir John Cockroft, and I had the vainglorious notion of leading that project. So I went for a couple of weeks to Ann Arbor (Michigan) where various bold schemes for strong focusing - a new idea then - were discussed under the leadership of Don Kerst, inventor of the betatron. The enthusiasm of the working party was only slightly affected by the summer heat; we worked in a new building, with its air-conditioning not yet working.

On my return I learned that the project had not been approved; yet my visit had not been in vain. I had met Don Glaser, the inventor of the bubble chamber, who had told me that his invention was becoming a practical instrument, and the tracks of the particles resulting from high-energy collisions could now be photographed in very large numbers. He stressed that there would soon be a frightful bottleneck unless semi-automatic equipment was quickly developed to cope with the expected flow of film.

So I forgot all about strong focusing and concentrated on the development of track-measuring devices. The first machine that was built, largely with the help of Alan Oxley, was not very different from machines built in the USA, Switzerland and other places; manually operated, slow and inaccurate by present standards. But it gave us an entrance to the international brotherhood of bubble-chamber physicists; we obtained film in return for sending people to CERN in Geneva, and from time to time played host to groups from Germany or Italy with whom we had shared the analysis of some batch of film, to compare results. New machines were designed and some of them built. Not all of them worked, and we gradually became experts. (A perfect expert is one who has made every possible mistake.)

While showing one of the failures at the Physical Society Exhibition in Manchester in 1964 I had plenty of time - few of the visitors were interested - for reflecting on how to build a better machine; the result has become known as Sweepnik (because it sweeps up data at the speed of a sputnik!) Its success was due to a number of fortunate factors: commercial lasers were just becoming available, providing the intense and very small spot of light that was needed; computers were getting cheap enough for one to be bought and put on line to serve as the "brain" of the machine, and two brilliant students turned up at just the right time. Julian Davies, who wrote most of the required software, moved to the Dept. of Machine Intelligence at Edinburgh University; Graham Street, who designed most of the electronics and contributed many original ideas went on to manage the firm (Laser Scan) that manufactured Sweepniks and other devices, and which kept me busy after I retired from my chair.

One contributory factor was the failure of the SRC to find the money for buying a commercial machine, after urging us to do that rather than trying to build our own! The Cavendish now has two Sweepniks as well as other advanced equipment for track measurement, and our research group, led by John Rushbrooke, is making valuable contributions to the knowledge of subatomic particles.