The Making of the Atomic Bomb

national goal. Science — a fragile, nascent political system of limited but increasing franchise — would have to wait until the war was won. Or so it seemed. But a few among the men and women gathered at Los Alamos — certainly Robert Oppenheimer — sniffed a paradox. They proposed in fact to win the war with an application of their science. They dreamed further that by that same application they might forestall the next war, might even end war as a means of settling differences between nations. Which must in the long run have decisive consequences, one way or the other, for nationalism.

By the time Robert Serber finished his orientation lectures at Los Alamos in mid-April most of the scientific and technical staff was on hand, many lodged temporarily in the surviving buildings of the Ranch School. Now began a second phase of the conference, to plan the laboratory's work. “If there were any ground-breaking ceremonies at Los Alamos like champagne or cutting ribbons,” John Manley comments, “I was unaware of them. Most of us who were there felt that the conference in April, 1943, was really the ground-breaking ceremony.” Rabi, Fermi and Samuel Allison arrived from Cambridge and Chicago to serve as senior consultants. Groves appointed a review committee — W. K. Lewis again, an engineer named E. L. Rose who was thoroughly experienced in ordnance design, Van Vleck, Tolman and one other expert — to follow planning and advise. Groves despite his formidable competence as an organizer and administrator was intellectually insecure around so many distinguished scientists, as who would not be?

They laid their plans, often during hikes into the uninhabited wild surroundings of the mesa. They had to rely heavily on theoretical anticipations of the effects they wanted to study; that was their basic constraint. Any experimental device that demonstrated a fast-neutron chain reaction to completion would use up at least one critical mass: there could be no controlled, laboratory-scale bomb tests, no squash-court demonstrations. They decided they had to analyze the explosion theoretically and work out ways to calculate the stages of its development. They needed to understand how neutrons would diffuse through the core and the tamper. They needed a theory of the explosion's hydrodynamics — the complex dynamic motions of its fluids, which the core and tamper would almost instantly become as their metals heated from solid to liquid to gas.

They needed detailed experiments to observe bomb-related nuclear phenomena and they needed integral experiments to duplicate as much as possible the full-scale operation of the bomb. They had to develop an initiator to start the chain reaction. They had to devise technology for reducing uranium and plutonium to metal, for casting and shaping that metal, possibly for alloying it to improve its properties. Particularly with plutonium, they had to discover and measure those properties in the first place and do so quickly when more than microgram quantities began to arrive. As a sideline, because they agreed that work on the Super should continue at second priority, they wanted to construct and operate a plant for liquefying deuterium at — 429 T — the cryogenics plant to be built near the south rim of the mesa.

Ordnance work was crucial. From the April discussions came immediate breakthroughs. An Oppenheimer recruit from the National Bureau of Standards who had been a protdgd at Caltech, a tall, thin, thirty-six-year-old experimental physicist named Seth Neddermeyer, imagined an entirely different strategy of assembly. Neddermeyer could not quite remember after the war the complex integrations by which he came to it. An ordnance expert had been lecturing. The expert had quibbled at the physicists' use of the word “explosion” to describe firing the bomb parts together. The proper word, the expert said, was “implosion.” During Serber's lectures Neddermeyer had already been thinking about what must happen when a heavy cylinder of metal is fired into a blind hole in an even heavier metal sphere. Spheres and shock waves made him think about spherically symmetrical shock waves, whatever those might be. “I remember thinking of trying to push in a shell of material against a plastic flow,” Neddermeyer told an interviewer later, “and I calculated the minimum pressures that would have to be applied. Then I happened to recall a crazy thing somebody had published about firing bullets against each other. It may have had a photograph of two bullets liquefied on impact. That is what I was thinking when the ballistics man mentioned implosion.”

Two bullets fired against each other recall the double-gun model of the Los Alamos Primer. There were other clues to Neddermeyer's new strategy placed evocatively in the Primer as well. That document notes that when the surface of the bomb core blows off, it “expands into the tamper material, starting a shock wave which compresses the tamper material six-teenfold.” The Primer emphasizes more than once that the expansion of the core would be the greatest obstacle to an efficient explosion. It may have occurred to Neddermeyer that if a tamper merely by its inertia — by its tendency to stay where it is when the swelling core begins to push out against it — could resist the core's expansion and thereby increase the efficiency of the explosion, a tamper that somehow pushed back against the core might do even better. The compressing of the boron bubbles in the autocatalytic bomb may also have been suggestive. Finally, the Primer offered the interesting model of four apple-quarter wedges of core/tamper fired together by an encompassing explosive ring. “At this point,” says Neddermeyer, “I raised my hand.”

He proposed packing a spherical layer of high explosives around a spherical assembly of tamper and a hollow but thick-walled spherical core. Detonated at many points simultaneously, the HE would blow inward. The shock wave from that explosion would squeeze the tamper from all sides, which in turn would squeeze the core. Squeezing the core would change its geometry from hollow shell to solid ball. What had been subcritical because of its geometry would be squeezed critical far faster and more efficiently than any mere gun could fire. “The gun will compress in one dimension,” Manley remembers Neddermeyer telling them. “Two dimensions would be better. Three dimensions would be better still.”

A three-dimensional squeeze inward was implosion. Neddermeyer had just defined a possible new way to fire an atomic bomb. The idea had been suggested previously, but no one had carried it beyond conversation. “At a meeting on ordnance problems late in April,” records the Los Alamos technical history, “Neddermeyer presented the first serious theoretical analysis of the implosion. His arguments showed that the compression of a… sphere by detonation of a surrounding high-explosive layer was feasible, and that it would be superior to the gun method both in its high velocity and shorter path of assembly.”

The response at the time was not encouraging. “Neddermeyer faced stiff opposition from Oppenheimer and, I think, Fermi and Bethe,” Manley says. How do you make a shock wave spherically symmetrical? How do you keep tamper and core from squirting out in every direction as water does when squeezed between cupped hands? “Nobody… really took [implosion] very seriously,” Manley adds. But Oppenheimer had been wrong before — even about the possibility of fission when Luis Alvarez dropped by to report it in 1939, wrong for the fifteen minutes it took him to think past the stubbornness with which he rejected any possibility he had not himself foreseen. Apparently he was learning to steer by that grudging incredulity as Bohr steered by the madness of a truly original idea. “This will have to be looked into,” he told Neddermeyer in private conference after the dismissive public debate. He took his revenge for the trouble Neddermeyer was causing him by appointing that thoroughgoing loner to the newly invented post of group leader in the Ordnance Division for implosion experimentation.

The other fresh insight remembered from the April conference corrected an error that everyone wondered afterward how anyone could have overlooked. The error is perhaps a measure of how unfamiliar the physicists were with ordnance. E. L. Rose, the research engineer on Groves' review committee, woke up one day to realize that the Army cannon the physicists were basing their estimates on weighed five tons only because it had to be sturdy enough for repeated firing. A gun that wore an atomic bomb welded to its muzzle could be flimsier: it would be fired only once, after which it would vaporize and drift away. That specification cut its weight drastically and promised a practical, flyable bomb.

Fermi, superb experimentalist that he was, contributed valuably to the program of experimental studies, denning with clarity problems that needed to be examined. For him the war work was duty, however, and the eager conviction he found on the Hill puzzled him. “After he had sat in on one of his first conferences here,” Oppenheimer recalls, “he turned to me and said, ‘I believe your people actually want to make a bomb.’ I remember his voice sounded surprised.”

The leaders attended a party one night that April at Oppenheimer's house, the log-and-stucco former residence of the Ranch School headmaster. Edward Condon, whose father had been a builder of railroads in the West, who had worked as a newspaper reporter in tough Oakland, found occasion at Oppenheimer's party to satirize Los Alamos' Panglossian mood. He was an exceptional theoretician; he and Oppenheimer had boarded together at Gottingen; Condon thought they were fast friends. He would soon clash bitterly with Groves over compartmentalization and find that his friend the director had higher priorities than backing him up. Now, sitting in a corner at the director's house, Condon pulled from a bookshelf a copy of Shakespeare's The