by Ray Monk
Moreover, for Oppenheimer, it involved maintaining smooth personal relations with, on the one hand, military and security men, and, on the other, prickly, obsessive scientists with large and easily bruised egos. Seth Neddermeyer was not an easy man, and he had developed an intense obsession with making his version of implosion work. In effectively side-lining him, while keeping him working on the programme under the leadership of someone else, Oppenheimer demonstrated rare tact, sensitivity and understanding.
In March 1944, still thinking of implosion as a means of making a uranium bomb with as little uranium as possible, Oppenheimer wrote to Groves that the ‘prime objective’ of the laboratory for the coming year was ‘to bring to a successful conclusion the development of the implosion unit with U-235’. The arrival of the British mission at the end of 1943 and the beginning of 1944 gave Oppenheimer the opportunity to pick yet more brains for hints on how to solve the problem of squeezing a ball of uranium without deforming it. Chief among these fresh brains was that of Rudolf Peierls, who, when implosion research received its new impetus, was still attached to the Columbia University group in New York, but who in February 1944 came to Los Alamos for a visit. ‘At that time,’ Peierls writes in his autobiography, ‘the laboratory was urgently required to obtain numerical solutions to the equations for the implosion.’ As it happened, the equation required ‘was of the same form as that for the blast wave in air, for which I had done my numerical experiments . . . I therefore came just at the right time to explain the step-by-step method by which the equation could be solved and the limits of the size of the steps.’ Doing it step-by-step was important, because these calculations were to be done on IBM punch-card machines, the ancestors of modern computers. After the February visit Oppenheimer wrote to Groves confirming the usefulness of his discussions with Peierls, with whom, he reported, he had gone ‘into the technical aspects of the British methods in considerable detail’. Oppenheimer was now ‘planning to attack the implosion problem along these lines with the highest possible urgency’.
Developing implosion was both a theoretical and an experimental matter. Indeed, as had been shown during von Neumann’s visit in the autumn of 1943, implosion raised theoretical questions interesting enough even for Edward Teller, who had steadfastly refused even to pretend to be stimulated by what he regarded as the merely engineering problems raised by the gun-assembly method. After his visit to Los Alamos, von Neumann kept up a correspondence with Teller, and in January 1944, Teller was appointed head of a small group of the Theoretical Division, devoted to solving the mathematical and theoretical problems raised by implosion. After Peierls’s visit in February, the work of this group was centred on using the mathematical techniques developed by the British.
The effort – indeed, the perceived need – to solve the problems of implosion was soon considered sufficiently important for Bethe, in March 1944, to reorganise the entire Theoretical Division to meet ‘the great and increased urgency of the implosion program’. In accordance with this reorganisation, Teller was put in charge of group T-1, the responsibilities of which were officially described as ‘Hydrodynamics of Implosion, Super’. Oppenheimer and Bethe both thought it was obvious that the first of these was the more important. Teller, however, whose interest in implosion in January seems to have waned by March, thought otherwise, and from the spring of 1944 onwards spent almost all his time working on the Super. The short section of his autobiography that he devotes to his abandonment of work on implosion presents a somewhat unclear picture of why, exactly, he lost interest in it. After describing ‘Johnny’ von Neumann’s visit to Los Alamos and his conversations with him about ‘fast’ implosion, Teller seems keen to convey the importance of those discussions and to highlight his own role in them:
The next morning, Johnny and I presented our findings to Oppenheimer. He immediately grasped their implications. Within a week, magnificent administrator that he was, he had turned the direction of the research around. From then on, our main efforts were no longer devoted to a gun-assembled weapon but rather to the implosion assembly.fn54
But, having stressed his role in providing the guiding ideas of the implosion problem, Teller recounts how, after ‘Johnny’ had left, Bethe called him into his office and told him: ‘I want you to take charge of solving the equations that will be needed to calculate implosion.’ This, Teller says, was a task he was reluctant to take on because it ‘seemed far too difficult’: ‘Not only were other people more capable than I of providing such work, but I also suspected that a job that formidable might not be completed in time to have any influence on a bomb that could be used during this war.’
As a reason for concentrating on the Super – which stood no chance whatsoever of being completed in time to influence the outcome of the war – this seems strikingly unconvincing. Besides, if Teller had originated the central ideas that inspired the renewed interest in implosion, who did he think could possibly be more capable of seeing it through to completion?
Interspersed with his account of how he helped to establish and then backed away from the rejuvenated implosion programme, Teller provides some personal reflections that, one suspects, are more to the point in explaining why he abandoned it. Though he enjoyed Bethe’s company, Teller says, ‘as physicists we approach problems differently’. Whereas Bethe was a brick builder, Teller was a brick layer; he liked to build things, rather than provide other people with the tools: ‘I much prefer (and am better at) exploring the various structures that can be made from brick, and seeing how the bricks stack up.’ Then, apparently apropos of nothing, Teller confesses that when Oppenheimer told him that he had appointed Bethe as head of the Theoretical Division, ‘I was a little hurt.’ ‘I had worked on the atom bomb project longer than Bethe. I had worked hard and fairly effectively on recruiting, and on helping Oppie organize the lab during the first chaotic weeks.’
In other words, it seems to be implied, Teller quit working on implosion because he could not bear the thought of being consigned to working on a mere brick-building task, working under another brick builder. He wanted to be working not on small mathematical tasks in the service of a goal set by someone else, but on large tasks in which he could pursue his own vision. He wanted to be the boss, not the under-labourer – especially if it turned out that the particular under-labouring task he had been set was one that someone else, namely Peierls, was better at than he.
On 1 May 1944, Oppenheimer wrote to Groves, asking if, as a matter of the ‘greatest urgency’, Teller (‘who is, in my opinion and Bethe’s, quite unsuited for this responsibility’) could be relieved of his role in the implosion programme and replaced by Peierls. But even now, and despite the tone of his letter to Groves, Oppenheimer did not completely fall out with Teller. On the contrary, he actively encouraged him to devote his energies to the Super and even – at a time when he and the entire laboratory were stretched to the limit – made time to see him for an hour’s discussion every week.
Peierls came to Los Alamos to take up leadership of the implosion theory group on 3 June 1944. Soon afterwards Oppenheimer gave a party for Lord Cherwell, Churchill’s scientific advisor, but somehow forgot to invite Peierls. When he came to Peierls’s office the next day to offer abject apologies for this unintended slight, Oppenheimer told him: ‘There is an element of comfort in this situation: it might have happened with Edward Teller.’
Shortly before Peierls’s arrival, a major breakthrough in the theory of implosion was made by another member of the British mission, James Tuck, who had worked previously on armour-piercing shells. Drawing on that work, he came up with a means of solving the problem of creating the smooth, symmetrical, inward-moving spherical shock wave that implosion required, which had, in practice, until this point proved impossible to create. The best one could do was simulate such a spherical shock wave by placing jets of explosive energy in a spherical arrangement. This, however, created a series of diverging and converging shock waves, the physics and mathematics of whic
h were at this time only imperfectly understood. What Tuck, building on von Neumann’s suggestion of shaped charges, envisaged was a way of arranging the explosive charges so that the waves of energy produced by them converged. The arrangement Tuck had in mind called for a series of ‘lenses’, analogous to optical lenses. Just as an optical lens forces waves of light to converge on a target, so Tuck’s lenses would force waves of explosive energy to converge, thus increasing their force and enabling the ‘fast implosion’ envisaged by von Neumann to be realised. Such an arrangement was, however, as David Hawkins puts it in his official history of Los Alamos, ‘a completely untried and undeveloped method, which no one wished to employ unless it became absolutely necessary to do so’. After Segrè’s shattering news about spontaneous fission in reactor plutonium, it looked as if it might indeed be necessary to employ this untried method. The laboratory needed to take risks. After all, in the summer of 1944, a mere twelve months before the deadline for producing the bomb, there was, as Hawkins puts it, ‘not a single experimental result that gave good reason to believe that a plutonium bomb could be made at all’.
In August 1944, then, the entire laboratory was reorganised to reflect the central importance of the effort to solve the many problems that stood in the way of making an implosion bomb. Stanislaw (‘Stan’) Ulam, the Polish mathematician and friend of Teller’s who arrived in Los Alamos at the end of 1943, has recalled with amusement the ‘fascination with organizational charts’ that he found at Los Alamos:
At meetings, theoretical talks were interesting enough, but whenever an organizational chart was displayed, I could feel the whole audience come to life with pleasure at seeing something concrete and definite (‘Who is responsible to whom,’ etc.).
As a result of the August 1944 reorganisation, those charts became more complicated as divisions multiplied, new groups were added and hundreds more men were recruited. Oppenheimer calculated that, if the laboratory were to solve the problems of implosion in time to have a plutonium bomb ready for the summer of 1945, he would need an additional 600 men. These, of course, would not all have to be distinguished physicists, chemists or ballistics experts. What was needed was relatively unskilled labour to carry out the experiments and make the observations and measurements that were required to solve the many scientific problems that remained unsolved. Why did Oppenheimer need so many men? The answer lies in the sheer number of questions raised by implosion. For example, in order to understand the nature of implosion itself, and to know what was needed to achieve the smooth symmetrical shock wave it required, literally thousands of experiments – analogous to, but far more sophisticated than, those conducted by Neddermeyer and his team on stove pipes – needed to be carried out.
In September 1944, the theoretician Robert Christy suggested that the metal to be imploded should be a solid ball, rather than the hollow sphere envisaged by Neddermeyer. After this had been accepted, the problem of implosion was tackled by experiment after experiment in which the implosion of a solid sphere of metal, usually cadmium, was attempted and measurements taken to see how close each successive attempt had come to producing that elusive uniform, symmetrical shock wave. The methods used to make these measurements were many and varied, and some were invented at Los Alamos. For example, Robert Serber came up with a novel idea that became the basis of what was called the ‘RaLa method’. The idea was to put a radioactive substance – the one chosen was an isotope of lanthanum (La-140), called Radiolanthanum (hence ‘RaLa’) – at the centre of the metal sphere being imploded. The gamma rays emitted by this radioactive source would be absorbed by the metal in proportion to its density, so the density changes in the metal as it was imploded could be measured by recording the intensity of the gamma radiation before, during and after the detonation. In this way, they could see how close they were to achieving a uniform, symmetrical shock wave. Other methods of measuring what happens when a lump of metal is imploded were devised using X-rays and photographs.
In addition to these meticulous experiments and observations concerned with the nature of implosion, much experimental work had to be carried out in order to design and manufacture the shaped explosive charges, providing answers to such questions as: what material should the explosive charge be made of? how, exactly, should it be shaped? and how should the shaped charges, the ‘lenses’, be arranged around the plutonium core? As a result of these experiments, a design was arrived at that was far more complex than the gun-assembly bomb originally planned. The weapon that was now envisaged looked like an enormous football, at the centre of which was a 3½-inch solid sphere of plutonium with a hole for the initiator, around which was a uranium tamper with a diameter of nine inches, in turn surrounded by thirty-two explosive charges, all carefully shaped into ‘lenses’, concentrating the shock wave at the centre of the sphere. In all, ‘Fat Man’ would be fifty-four inches wide and weigh nearly 5,500 pounds.
This design – which, astonishingly, was arrived at in February 1945, just six months after the major reorganisation – was the outcome of observing innumerable detonations, requiring tens of thousands of charges. To supply these, the laboratory had its own workshop, which became in effect a factory, employing dozens of young men. These men, like most of the 600 extra people that Oppenheimer had estimated he needed, were ‘SEDs’, members of the US Army’s Special Engineering Detachment. ‘They were kids mostly,’ Kistiakowsky said, ‘with partial college education, but there were even a few PhDs.’ The SEDs had a difficult time at Los Alamos, where they worked the long hours of the scientists, while keeping up the disciplined and tightly structured life of soldiers. They lived not in the kind of houses or flats provided for the scientists, but in crowded barracks, within which they had just a tiny space of their own. ‘We had reveille at six,’ one of them recalled, ‘we had drill and exercises at six-thirty and the fatigues until eight a.m. – and sometimes, working on something in the workshops, we had not gotten to bed until two or three in the morning.’
Of course, with such a massive influx of personnel, it was inevitable that standards of security would decline. It was clearly impossible to watch the movements of every one of the hundreds of people who came to Los Alamos in the summer of 1944, and in any case it seems not to have occurred to Groves, de Silva or Lansdale that among the young men of the Special Engineering Detachment there might be some who would be willing to spy for the Soviet Union. And yet there was at least one such person, David Greenglass, whose willingness to pass information to the Soviets, though it may not have been very significant in increasing Soviet knowledge of the bomb, would turn out to have drastic consequences for his sister and her husband, Ethel and Julius Rosenberg.
Greenglass had studied mechanical engineering before, at the age of twenty-one, being inducted into the US army in 1943. He and his wife, Ruth, were at that time both members of the Communist Party. After working for a while at Oak Ridge, Greenglass was sent to Los Alamos in August 1944, as part of the Special Engineering Detachment. During his time at Oak Ridge, and for the first three months of working at Los Alamos, Greenglass did not know that he was working on the atomic-bomb project. At Los Alamos he was employed in the workshop making moulds for the shaped explosives that were needed for the implosion experiments. He had no idea what these ‘lenses’ were for, and no understanding of the implosion process of which they were such an essential part.
Greenglass learned that he was working on the atomic-bomb project in November 1944, when he was told by his wife, Ruth, who in turn had been told by the Rosenbergs. At the Rosenbergs’ expense, Ruth (who lived in New York) had gone to New Mexico to visit her husband, whom she had not seen since his transfer to Los Alamos. They spent a few days together in Albuquerque, where Ruth relayed the information to Greenglass. The Rosenbergs had been communists since the early 1930s, and Julius Rosenberg was the hub of a spy network that sought to gather information about secret military and industrial projects. Before she left New York, the Rosenbergs gave Ruth Greenglass a list of q
uestions to ask her husband about the layout and the personnel at Los Alamos.
A few months later, in January 1945, David Greenglass came to New York on leave and provided the Rosenbergs with a description and a drawing of the lens moulds he was making. He also, then or later, provided the Rosenbergs with a rough, and presumably not very useful, drawing of the ‘Fat Man’ bomb. After Greenglass’s return to Los Alamos, Julius Rosenberg arranged for him to meet a Soviet contact periodically to pass on anything else he had managed to find out. That contact was none other than Harry Gold, the man Fuchs had been meeting to share information about gaseous diffusion. Indeed, it would be via the connections with Fuchs and Gold that Greenglass, and therefore the Rosenbergs, were exposed as spies in 1950. At the Rosenbergs’ trial, Greenglass testified against his sister, leading her to be executed alongside her husband on 19 June 1953. Greenglass himself was sentenced to fifteen years in prison. In 1960, after serving ten years, he was released and rejoined his wife in New York. Nearly forty years later, he told a reporter from the New York Times that his testimony against his sister had been a lie told in order to protect his wife, who was never charged in connection with the affair.
The Greenglass affair clearly upset John Lansdale. Speaking in 1954, at Oppenheimer’s security hearing, he referred twice to ‘the inexcusable Greenglass case’. The word ‘inexcusable’ seems to have been chosen by Lansdale to refer not to Greenglass’s behaviour (though no doubt he would have thought it applicable to that), but to his own lapse in failing to catch Greenglass. ‘He is certainly an example of one we missed,’ he remarked ruefully. When he was asked to confirm that Oppenheimer had no responsibility for Greenglass ‘in any way, shape or form’, Lansdale replied: ‘I don’t believe so. I will take full responsibility for that one. That was the outstanding blunder of the century.’
