The Fruit

The Cavendish experimenters instead resolved the matter by pure experiment. At the end of February 1934, Cockcroft reported that they had at last detected fast protons from deuteron bombardment of copper, iron, gold, and yttrium, at energies down to 200 kV. Also, and the eventual key to the problem: "Oliphant is getting queer results with H² and H² ." Oliphant and Harteck had made up targets like NH₄ Cl containing heavy hydrogen. When bombarded with protons, the targets behaved as if constituted entirely of ordinary hydrogen. When shelled with deuterons, they gave off protons with a range of 14.3 cm, which the Cavendish researchers tacitly identified with Berkeley's 18-cm rays. And this irrespective of the deuterons' energy, as long as it exceeded 20 keV. At 100 keV the effect was too large for recording on the

sensitive detector of the Oliphant-Rutherford apparatus, so great in fact that Oliphant and Rutherford at first ascribed it to a burst of x rays. Analysis of range and energy, confirmed by cloud chamber pictures, showed that the protons arose from the reaction d+d ® H³ +p. They also found the neutrons that Lawrence had advertised. Rutherford discerned their origin in the competitive reaction, d+d ® He³ n.^[53] Oliphant broke the news gently. "We suggest very tentatively that your results may be explained as due to the bombardment of films of D and of D compounds. . . . I hope these results are of interest to you."^[54]

The puzzle Lawrence started had therefore the simple solution of contaminated apparatus: deuterons stuck in the walls of the cyclotron chamber and in the targets gave the plentiful d-d reaction when bombarded by deuterons; the fast protons, apparently from everything, which had prompted the disintegration hypothesis, did in fact come from deuterons, not from the target, but in consequence of fusion, not disintegration. The accelerator men at Caltech and Carnegie also offered plentiful evidence that Berkeley's results came from synthesis of deuterons with pervasive contaminants. Lauritsen and Crane showed that the fast protons from deuterons on carbon came from C¹² (d,p)C¹³ , an observation extended by Cockcroft and Walton to explain, as a consequence of oil films on metals, why deuterons appeared to knock protons from copper and tungsten. Tuve and Hafstad repeated all the Berkeley experiments with carefully controlled beams on immaculate targets and found that the Berkeley experimenters had got nothing right. With a chamber contaminated with deuterons, however, they had no trouble recovering Berkeley errors.^[55] The demonstration with the hydroxide targets, which Lawrence had thought so convincing, also failed through ubiquitous deuterons.

As Lauritsen and Oliphant independently explained it to Lawrence: the proton beam playing on Ca(DH)₂ liberated deuterons by collision or exchange; these deuterons then joined the bombardment and produced the 18-cm protons by d-d synthesis.^[56]

The hunters horsed to catch the hare let loose from Berkeley were formidable: the Cavendish from the top down and the experienced teams and big machines at two of the best-endowed physics research institutes in the United States. Resolution of the conflicting results required very careful experiment in previously unexplored regions of multiple, competing nuclear reactions; and it required, and abetted, discovery of the occult d-d fusion process and the isobars of mass three. That detection and analysis of deuterium fusion was not so easy as its plentiful yield might suggest appears from the struggles of Otto Frisch, who proposed to Bohr that they build a big source of neutrons utilizing the d-d reaction. He had a machine available in December 1934. As late as the following March he had not detected a single neutron, although he commanded 60 kV, far above the threshold reported by Oliphant, Harteck, and Rutherford. He did succeed eventually, as did G.P. Thomson, who also had trouble reproducing the Cambridge results.^[57]

In accordance with the social expectations of science, Lawrence confessed himself chagrined at the "stupidity" of his error, and Cockcroft softened the blow with the falsehood that "for a long time Rutherford and Chadwick were nearly convinced that you were right" and the truth that the experiments of Lawrence, Lewis, and Livingston had brought "an important enlargement of the field of nuclear research."^[58] Lawrence diagnosed his stupidity as a consequence of the great productivity of the cyclotron, which

discouraged "methodical, quantitative, measurements;" a healthy point of view that enabled him soon to dismiss the experiments that had prompted his error as "of the character of a preliminary survey."^[59] He had persisted in his errors, however, in the face of warnings from many sides of the likelihood of contamination.^[60]

This flouting of good advice and good procedure irritated some of Lawrence's closest associates. Tuve wrote Cockcroft: "From our own experiments we feel that the important issue is rather one of judgment and point of view rather than of the errors in technique which can give rise to such a situation." Lawrence had botched everything so badly that the corrective, as Tuve wrote in an internal report, "will be a difficult thing to present in public." Kurie wrote Cooksey: "The Englishmen are doing what seem at this distance to be clean experiments, but Ernest and Malcolm [Henderson] are too excited to go slowly." These expressions of disappointment appear to reflect a worry that Lawrence's way of doing science and his rising celebrity in the United States might compromise American physics just as it was assuming world leadership. Cockcroft agreed that Lawrence's style did not suit nuclear physics. "There is a real danger of the subject getting into a mess, and I feel that the only thing to do is to delay publication until we are reasonably sure."^[61]

Tuve did not hide his irritation from Lawrence. Their different institutional settings and personal ambitions had become more important in determining their scientific ethos than their similar backgrounds and almost identical scientific interests. In the spring of 1934 Tuve advised Lawrence to withdraw his faulty claims formally, a step not only ethically sound but also useful to Tuve, who would be spared the obligation of a public exposé. Lawrence replied petulantly. In the summer of 1934 Tuve carried the attack

to Berkeley, to a meeting of the American Physical Society. Either the hospitality or the prevailing genius of the place caused the secretaries who reported the discussions to nod: they attributed to Tuve the irenic solution that his and Lawrence's results could be brought into harmony by considering the different apparatus and energy with which they worked. Not a mention of contamination or haste. Tuve could not allow that to pass, since it implied that not Lawrence, but he, had been wrong. He therefore wrote to Science , which had published the irenic report, to point out that the Berkeley group had abandoned their major claims and withdrawn many of their results, "which have been sought for but not verified," and which, by confusing the subject, had "delayed some-what the examination of these questions."^[62]

To recover his standing, Lawrence thought to make a few "precise and trustworthy measurements." The deuteron business brought home forcefully, he told Cockcroft, that most of the Laboratory's fast data had no value. Accordingly, Lawrence made an elaborate study of the excitation and decay products of the useful isotope radiosodium, a novelty whose discovery will occupy us presently. Up to the maximum energies he used (1.9 MeV), deuterons seemed to induce transformations in sodium in the amounts predicted by Gamow's theory. But continuation of the experiments to aluminum and to higher energies in collaboration with a new recruit to the Laboratory, a postdoc from Princeton named Edwin McMillan, who knew how to work exactly, showed excitation larger than Gamow allowed. The departure of their measurements from a theory that worked well for protons and alpha particles interested Oppenheimer and his student Melba Phillips, who had just finished failing to explain the curious rise of absorption beyond the series limit in Lawrence's old reliable measurements on the photoeffect.^[63]

When a fast deuteron enters a heavy atom, Coulomb repulsion slows its proton and so increases the relative velocity of its constituents. Under these circumstances, the nucleus may capture the

neutron. The released proton then makes an appearance in the world. Encouraged by this attractive possibility, which occurred to several physicists, Lawrence, McMillan, and another new postdoc, Robert Thornton from McGill, pushed on to silicon and copper. Their results agreed exactly with the calculations of Oppenheimer and Phillips for a binding energy of the deuteron—about 2 MeV—that agreed very well with values obtained from mass-energy balances in nuclear transformations as measured by others.^[64] Here was a point for the underdog. "I am amazed at the agreement," Lawrence wrote Rutherford. And also incurable: with the new mechanism, deuterons with available energies might provoke nuclear reactions "much further up the periodic table than one could ever have hoped for."^[65]

There was another point for Lawrence's side of the story. Maurice Goldhaber, an émigré research student at the Cavendish, suggested to Chadwick an idea for an experiment that he had brought from Berlin: the photodisintegration of the deuteron. Goldhaber recalled that Chadwick showed little interest in the matter until told that if successful it would give data from which the neutron's mass could be deduced.^[66] Telltale protons announced the disintegration of deuterons in heavy water irradiated by the hardest photons then available, the gamma rays of 2.6 MeV from ThC", of which the Cavendish had a good supply. With a preliminary determination of the kinetic energy of the proton, Chadwick and Goldhaber had a value of the neutron's mass without making assumptions about the mass of complex nuclei:

m_n = m_d – m_p + E_g – 2E_p» 1.0081.

Their value, almost midway between Chadwick's and the Joliots', is close to the modern one (1.0085). Since the mass of the proton was known to be 1.0078, it followed that Chadwick's neutron was radioactive. Its decay was first observed in 1948.^[67] The discovery

of the photodisintegration of the deuteron gave rise to a new branch of nuclear investigation, to which the Berkeley Laboratory, which preferred to shoot particles it could accelerate, contributed very little. For a time the (g ,n) reaction appeared to be a promising source of neutrons.^[68]

And finally, a point to be dilated later, the discovery of the d-d reaction, although instigated by cyclotron experiments, did not much recommend cyclotrons to nuclear physicists. Certainly it helped confirm the opinion of Rutherford, who had outplayed Lawrence with beams ten or even a hundred times less energetic than Berkeley's, that the Cavendish had no need for a cyclotron. More generally, it made possible a cheap substitute for the energetic neutron beams that, for a brief time, had been the peculiar preserve of large accelerators. It was now only necessary to have a supply of heavy water and a Cockcroft-Walton or Van de Graaff that could develop 100 or 200 kV, and one had a copious current of fast neutrons almost equal in energy courtesy of the peculiar sociability of deuterons.^[69]