3—
Reception

The leaders of British nuclear physics began to warm to the cyclotron in 1935. No doubt the increasing strength of its currents and yields played a major part in their revaluation. So did a worry that they might be left behind. "Everyone in America is building cyclotrons!" So Fowler overestimated the situation, while making clear his understanding of the reason for the stampede: Lawrence was getting "something like a beam," namely a microamp of 11 MeV alpha particles. Chadwick saw the cyclotron as an engine for strengthening his position at Liverpool while indulging his aesthetic sense: the "magnetic resonance accelerator," he wrote its inventor, "ranks with the expansion [cloud] chamber as the most beautiful piece of apparatus I know. . . . I must have a cyclotron apparatus. When I look at your cloud chamber photographs and see the enormous number of recoil tracks I realize what I am missing." Cockcroft declared his desire to see a cyclotron built in Britain, but did not yet—in the summer of 1935—know how or whether to commit the Cavendish to it. "The medical applications would probably provide an excuse."^[44]

The Cavendish soon committed itself. The initiative came from an unlikely source, the Soviet Union, which, by detaining its citizen Peter Kapitza during his visit home in 1934 had upset Cambridge physics. Kapitza held a special professorship, financed by the Royal Society, to preside over a laboratory for high magnetic fields and low temperatures established within the Cavendish through the generosity of the industrial chemist Ludwig Mond. To coax Kapitza to cooperate, the Soviet government proposed to reproduce the Mond Laboratory in Moscow. In the late fall of 1935, Cambridge agreed to accept 30,000 pounds from the USSR for the Mond's magnetic equipment, for duplicates of the liquefaction plants, and for other apparatus; and by the end of the following year Kapitza had his stuff and the Cavendish its cash.^[45]

Rutherford decided not to replace the big generator for producing the very high magnetic fields that were Kapitza's specialty, but inclined to buy a large electromagnet, for—among other things—accelerating ions. Cockcroft consulted Lawrence on the dimensions of the putative all-purpose magnet. The reply—"the bigger the magnet the better"— did not cause the Cavendish to commission construction. Lingering doubts may have been resolved by notification by Lawrence of the fine initial performance of Cooksey's first vacuum chamber. "If you are undertaking the construction of [a cyclotron], whoever is directly in charge of the work will probably derive some comfort from [Cooksey's experience], because in many quarters it is not realized that it is possible to build an accelerator with a predictably satisfactory performance." This reassurance was dated February 3, 1936; on February 22, Rutherford wrote that he had decided to proceed with the magnet, but not to dedicate it to resonance acceleration. It would be available "for general purposes, and also probably for use as a cyclotron." Cockcroft provided details: they thought to have a magnet with pole pieces 100 cm in diameter, capable of 17.5 kG, the pole faces mounted vertically rather than horizontally as at Berkeley in order that the instrument be adaptable to cosmic-ray work.^[46]

Meanwhile Chadwick was busy raising money for a Liverpudlian cyclotron. He approached A.P.M. Fleming, director of research of Metropolitan-Vickers, who went to inspect Berkeley's 27-inch in November 1935 and agreed to create something similar for 5,000 pounds. That was just twice what Princeton expected to pay and over twice the 2,000 pounds that Chadwick had in hand. Lawrence sent the advice of experience: commission Metro-Vick with the 2,000 pounds and goad or embarrass Fleming into providing the rest gratis. Fleming was interested in cancer, as Lawrence had learned during Fleming's visit to Berkeley. He would therefore be interested in the recent discovery at Berkeley that a certain mouse tumor is more strongly affected by neutron

beams than by x rays. "If malignant tumors in general are correspondingly more sensitive to neutron radiation, neutrons will supersede x rays in the treatment of cancer. . . . You might tell Dr Fleming that in view of this important possibility, we are definitely planning to go forward with the treatment of human cancer with our cyclotron."^[47]

Fleming decided to be generous—eventually he gave Chadwick oscillator tubes and other essential parts—and he, Chadwick, and Cockcroft joined forces to design magnets for three cyclotrons. Fleming wanted a small one, for a 1 MV or 2 MV machine; the Cavendish, rich in rubles, wanted "a very large magnet indeed;" "for my part [Chadwick wrote] I must build for ten million volts." All wanted to put the poles vertical for magnetic work and cosmic-ray studies. Lawrence discouraged their uprightness. The vacuum chamber had to be removable and accessible, and mounted horizontally on wheels to roll out on tracks for servicing and repair. When Lawrence rejected vertical poles and implied that a cyclotron required a dedicated magnet, the Cavendish had just come into additional wealth that allowed it to multiply magnets outside the Mond laboratory. On May 1, 1936, the vice chancellor announced that the automobile manufacturer Lord Austin had given the 250,000 pounds sought for the Cavendish Laboratory. That put an end to Rutherford's penny-pinching. "It amazed me [Pollard, of Yale] to see the free way in which money is passed around. The Cavendish . . . seems to be wallowing in cash." Cockcroft got rid of his Morris car and collected plans for magnets with horizontal poles.^[48]

No doubt the main force that drove the creation of cyclotrons at Cambridge and Liverpool was the desire of the directors of

both laboratories to have the complete equipment for nuclear research. But that did not drive Metro-Vick, whose attitude was decisive for cyclotroneering in Britain. In September 1935 an official of Metro-Vick, George McKerrow, had asked Cockcroft whether they should try to acquire rights under Fermi's patents or whether they should work up their own process.^[49] Cockcroft answered with Berkeley's latest neutron yields. They excited McKerrow greatly: "The business is just on the edge of practical possibility." Since Philips intended to work Fermi's patents via neutrons from high-tension machines, Metro-Vick's best bet appeared to be cyclotron production. At the end of February, just after the Cavendish had decided on a cyclotron, McKerrow asked Cockcroft for drawings of Lawrence's machine.^[50] After more careful consideration, the industrial possibilities, already circumscribed by the Research Corporation's patent on the cyclotron and Fermi's patents on nuclear activation, must have seemed less promising. Oliphant was no doubt right in placing the blame for the delay and inefficiency of the first British cyclotrons on the growing indifference of Metro-Vick.^[51]

Hardware

You cannot always get what you can pay for. Take the magnet, for example. Everyone at Berkeley knew the importance of having poles absolutely symmetrical and the gap between them big enough to admit shims, accommodate a proton chamber, and allow easy service. Lawrence suggested a gap of over 7 inches for magnets of pole pieces 36 inches (90 cm) in diameter, the size chosen at Liverpool, Cambridge, and Copenhagen; Joliot began with the hope of 100 cm, but the cost reduced him quickly to 80. The British adopted a wide gap (8 inches), as at Berkeley, despite the resultant sacrifice in magnetic intensity (a maximum of 18 kG); but the Continentals, reaching for 20 kG and not consulting Lawrence, narrowed the gap to 3.5 inches (9 cm) at Copenhagen

and Paris and landed in trouble.^[52]

All the magnets were built by large industrial contractors, in England by Brown-Firth of Sheffield and Metro-Vick; in Denmark by the Thrige works; in France—or rather in Switzerland, where Joliot had his magnet made by the only firm in Europe he thought capable of it—by Oerlikon of Zurich.^[53] In every case the construction of the mammoths—forty-six tons of steel and eight of copper for the robust British, thirty-five tons of steel and three tons of copper for the Danes, thirty tons for the delicate French—went slowly. In the spring of 1937 the Cavendish discovered that Metro-Vick was winding its magnet at the rate of one coil a week and would finish in eight months, well over a year after commissioning, if something were not done. And the Cavendish magnet had precedence over Liverpool's. After appeal to Metro-Vick, they expected delivery in August, then in October; but neither Cambridge nor Liverpool had its magnet for Christmas. As they realized their orders competed with commissions for armaments. This difficulty also stymied magnet making in Sweden, whose noble neutrality and Nobel industry earned it an enviable trade in arms.^[54] In Copenhagen they looked forward to having their donated magnet around Easter 1936, but it had not arrived by September; when it came it exceeded all expectations as to the magnitude of its field, but failed in homogeneity, which could not be corrected by shimming because of the narrow gap. That made a most awkward situation, since Thrige, which had built and donated the magnet, did not like to admit and rectify its

mistake.^[55] The deficiency had to be corrected by cutting slits in the pole stems to admit shims and by reducing the acceleration chamber to a brass ring sealed directly onto the pole faces.^[56]

The precise Swiss precisely calculated their delay in advance: seven and a half months, counting from November 1, 1936. They finished the following May and tested in June; the magnet proved "eminently satisfactory," magnifique , "beyond expectation," "much better [according to Wolfgang Gentner, who saw it at Oerlikon] than the American ones." It could give 22 kG, though, to be sure, at a great expenditure of power, 130 kW. Then it occurred to Joliot that the 9-cm gap that came with the big field might not accommodate the dees. He asked Lawrence. The answer—that the large capacitance between the dees and the tank consequent on their proximity would certainly make many difficulties in the construction of the oscillator circuit—was not reassuring.^[57]

The narrowness of the gap exactly matched the place that Joliot planned to put his cyclotron, a sub-basement at the Collège de France beneath the foundations of the chemistry building, enlarged from a cellar for storage of dangerous materials, a place condemned by government architects for lack of heat and ventilation. Into this hole Joliot expected to fit all the auxiliary equipment for his cyclotron and to hook it up to the vacuum chamber crammed with little working space into the 9-cm gap between the pole pieces of his magnificent magnet. An American visiting in 1939 saw a machine "built in a strangely cramped way by a Swiss firm and . . . stowed away in a strangely narrow subterranean chamber where working conditions . . . are not only uncomfortable but positively dangerous." Another inspector of Joliot's cyclotron cave judged "the French attack on nuclear physics [to be] about as adequate as their preparations to repel Hitler."^[58]

The oscillator provided as fine an opportunity as the magnet for delay and frustration. Joliot entrusted his difficult radiofrequency system to Culmann et compagnie, specialists in a French specialty, electric furnaces, who offered in June 1936 to install a modification of Livingston's original design, adjustable down to a wavelength of 17 meters, for 45,000 francs without the tubes. The plan was scrapped a year later in favor of an ambitious design with several stages and regulators. The price rose faster than the power, reaching 359,282 francs by June 1938, some two years after the original contract; and that omitted 449,800 francs for equipment that Joliot, who had ordered economies on wiring, regulators, and safety devices, decided he did not need. But with all this the furnace maker never got his oscillators going properly and Joliot had eventually to ask the Rockefeller Foundation for salaries for two specialists in high-frequency construction to correct the installation.^[59] The other would-be cyclotron laboratories would not meet the price of commercially built oscillating systems (30,000 DKr from Philips, too much from Metro-Vick) and decided to make them.^[60] This decision did not result in disaster because each laboratory had acquired the most important instrument for the efficient construction of a cyclotron: a man trained in Berkeley.

Software

Already in the winter of 1935/36 Lawrence was offering students as well as advice and blueprints to Chadwick and Cockcroft.^[61] He recommended Bernard Kinsey, who in three years at Berkeley as a Commonwealth Fellow had become "a thorough master of the art of high frequency oscillators," to the Cavendish. But it was Chadwick who acquired him and made him responsible

not only for the oscillators, which he completed in the summer of 1937, but for the entire installation of the Liverpool cyclotron. A visitor with Berkeley experience, James Cork, judged the oscillators to be "stupendous." Cambridge chose something less spectacular, a simplification of an oscillator circuit devised by the BBC.^[62] In the spring of 1937 Lawrence proposed to Chadwick and to Cockcroft that they find a stipend for Harold Walke, another Commonwealth Fellow nearing the end of his tenure, who wanted only "enough to barely live on . . . , to go on with research for a while." The Cavendish did not supply the pittance, since they were getting a Berkeley man for free, Donald Hurst, who had an 1851 Exhibition fellowship. With his help ("Hurst is a great acquisition"), Cambridge got the first faint evidence of a beam in August 1938. Unfortunately for Walke, Chadwick found him a place, which he took up in October 1937. Walke and Kinsey succeeded in getting the Liverpool machine going in 1939. Later in the year, while replacing the grid resistance of Kinsey's stupendous oscillator, Walke touched a 230-volt line, which was enough to kill him.^[63]

Early in 1937 Joliot realized that he needed Berkeley experience. He obtained a fellowship from the Rockefeller Foundation for Nahmias to go to the United States and arranged with Bohr to divide the services of one of Lawrence's students for a year, again with Rockefeller money. The grant for a travelling student was made, but Lawrence declined to supply one on these terms: Joliot and Bohr must each have a man for a year. As Tisdale reported Lawrence's argument, "the job of designing, installing, adjusting, and testing a cyclotron is considerably longer, more arduous, more tricky [!], and more difficult than other people expect." As Nahmias misreported it, all the cyclotroneers building outside Berkeley had had "experience of at least two years with the monster." The

Rockefeller Foundation doubled its grant accordingly. The Berkeley experts—Paxton (Paris) and Laslett (Copenhagen)—arrived at their stations in July and September 1937, respectively.^[64]

Laslett needed all his talents to bring about the reconstruction of the Copenhagen magnet and to design and help make the thousands of components deemed too expensive to buy. He built and wired the switchboard that controlled the measurement instruments and safety equipment. By October he had proved himself to be of great use; in the end he was "indispensable," and Bohr arranged to keep him on with a grant from the Rask-Oersted Foundation, which recirculated dollars from Denmark's sale of the Virgin Islands to the United States in 1917.^[65] That the Copenhagen machine could give a tentative beam in November 1938 (or, rather, a few flashes on a fluorescent screen, "which can hardly have been anything but positive ions"), and so became the second European cyclotron to operate, was "to a large extent due to his efforts."^[66]

Paxton shouldered a more formidable task. Joliot had left his magnet in Zurich while awaiting completion of its subterranean den. It was buried there in January 1938 with a vacuum chamber designed by Paxton and built by Oerlikon between its teeth. Swiss efficiency again threw British bumbling into relief: it took Metro-Vick over a year to finish Liverpool's vacuum tank and perhaps as long to do Cambridge's, although both were built almost exactly to plans furnished by Lawrence.^[67] (Cooksey had designed and built

his tank in a month or two.) But Joliot's high-frequency system would not work, or put enough voltage on the dees if it did; leaks and sparks still plagued the vacuum system and insulators when Paxton left Paris in September 1938 to help start up the Columbia cyclotron. By the end of January 1939, Nahmias, who had returned with such Berkeley experience as he had permitted himself to acquire, had cured some of these ills and sought a better vacuum, more dee voltage, and a beam. He was at last requited, on March 3, 1939. He immediately called Tisdale. "Nothing will do but that I must come over and see the phenomenon—which I did. Needless to say that there is great elation."^[68]