In his New York hotel room, Szilard drafted a paper revealing just how chain reactions could be made to work: “Divergent Chain Reactions in Systems Composed of Uranium and Carbon.” Joliot’s reactions had been “convergent,” as they shut down for lack of neutrons to extend the chain. Szilard’s would be “divergent” by multiplying once the chain reaction began. He mailed his paper on this topic to the Physical Review on February 6 and followed it with a thirty-nine-page article on the fourteenth. The first source Szilard cited in the longer article was H. G. Wells, who had written about atomic bombs as early as 1913. Both papers, Szilard asked, should be withheld from publication until he had clearance from the government.5
Early in 1940, Szilard had new reasons for advocating secrecy among the atomic scientists; a warning about Germany’s threat came when Peter Debye, a Dutch physical chemist and Nobel laureate, visited Columbia. Debye had headed the Institute of Physics at the Kaiser Wilhelm Institute in Berlin but was being forced out because he had refused to accept German citizenship under the Nazis. At Columbia, Debye first called on Fermi, who seemed unconcerned about reports of German A-bomb work. The German scientists were not at one location, Fermi said, and by being scattered around the country they would not be able to make a concerted effort. Szilard was not so assured by Debye’s news; he was alarmed.6
In their February Princeton meeting, Szilard and Einstein agreed that a second letter to the president was needed to carry out their threat of publishing the chain-reaction manuscript. Accordingly, Szilard drafted an Einstein letter to Alexander Sachs, bearer of the first Einstein letter to Roosevelt, which relayed news that “since the outbreak of the war, interest in uranium has intensified in Germany [and] . . . that research there is being carried out in great secrecy and . . . has been extended to another of the Kaiser Wilhelm Institutes, the Institute of Physics.” The research, Einstein warned in a March 7 letter, had been taken over by the German government. He mentioned that Szilard’s manuscript “will appear in print” unless “held up” by a change in administration policy. Einstein promised a memo by Szilard on “the progress made since last October” on chain-reaction research. A March 15 letter by Sachs to Roosevelt alerted the president to Germany’s advances and to Szilard’s improvements on the French chain reaction. This prompted Roosevelt to ask his secretary, “Pa” Watson, to convene a meeting in Washington of Einstein, Advisory Committee chairman Lyman J. Briggs, and some military men.7
But more weeks passed with no results. Einstein declined to attend a meeting, and it was late April before Szilard had completed for Sachs the promised chain-reaction memo. In that memo (later called the “A-55 Report”), Szilard cited three chain-reaction “applications” that relate to national defense: power production for warships; a radiation device that might kill “human beings who are exposed to it within a radius of one kilometer”; and bombs whose “explosions of extraordinary intensity” may create tidal waves if detonated near ports. The bombs “would not be too heavy to be carried by small boats, but could hardly be carried by existing airplanes.”8
In this memo Szilard also calculated relative speeds and ranges for oil-and nuclear-powered ships, based on advice from an Annapolis graduate and a quick read of Jane’s Fighting Ships. “I should imagine,” he wrote in April 1940, in his typical blend of geopolitics and physics, “that the combination of high speed [from lighter nuclear fuel] and a greatly increased cruising radius might be of decisive importance in case of a war with Japan.”9 Pearl Harbor was still twenty months away.10
Apparently, Einstein’s second letter to the White House prompted release of the $6,000 promised to Fermi and Szilard, and they bought the graphite and uranium that Szilard had ordered from suppliers. Fermi and research assistant Herbert Anderson began measuring the way graphite absorbs neutrons and soon discovered that the pure material Szilard had specified absorbed very few: It would be a good “moderator,” able to slow down neutrons for capture by uranium atoms without itself absorbing them. By contrast, in January 1940 the Germans had used impure graphite for similar measurements and had concluded that it absorbed too many neutrons to sustain a chain reaction.
Szilard went to Fermi’s office and said that the neutron-absorption value should not be published. Fermi had honored Szilard’s request for secrecy in 1939. But “this time Fermi really lost his temper; he really thought this was absurd” Szilard recalled. Szilard did not press his point but later asked, in a memo to physics department chairman George Pegram and Fermi, whether the “value for absorption cross sections of graphite” should be discussed within the lab, or should questions be evaded?11 This led Pegram to pressure Fermi with the same question, and he reluctantly agreed to censor his work.
“From that point, the secrecy was on,” Szilard recalled. And just in time. Had the Germans realized that their January calculations were off and that in graphite they had an abundant and inexpensive moderator, they might have pursued this research to make a reactor. Instead, acting on their erroneous conclusions, they used heavy water as a moderator—a choice that would doom their chances of making an A-bomb during the war.12
This second clash over censorship strained an already testy collaboration between Szilard and Fermi, a relationship that focused their different professional and personal styles. Ever the outsider, Szilard challenged conventions and authority with bursts of original ideas, while Fermi, a team player, enjoyed collaborating step-by-step on systematic research into the fundamental questions of nature. Szilard made intuitive leaps, while Fermi framed and resolved theoretical questions before designing and conducting experiments to test his ideas. Szilard’s deep and concentrated thoughts led him to see through problems to sometimes outrageous, sometimes prescient, results, while Fermi’s rigorous grasp of both theory and practice kept him centered on the task at hand. In personal style, too, Szilard was assertive, bumptious, and direct; Fermi was polite, quiet, and controlled.
“On matters scientific or technical there was rarely any disagreement between Fermi and myself or the rest of us whose opinions counted in those early days,” Szilard recalled. “Nor was there much difference in our attitude toward those who ‘directed’ our work from Washington.” Nevertheless, “Fermi thought we took too great chances in the [uranium] project,” and “on this we thoroughly disagreed.”
That fundamental disagreement Szilard traced to a very different view of the world and their lives as scientists.
Fermi is a scientist pure and simple. This position is unassailable because it is all of one piece. . . . I doubt that he ever understood that some people live in two worlds like I do. A world, and science is a part of this one in which we have to predict what is going to happen, and another world in which we try to forget these predictions in order to be able to fight for what we would want to [have] happen. But then how many people are able to understand the coexistence of these two separate worlds? I certainly would not understand it were it not for certain accidents of my education. Fermi and I had disagreed from the very start of our collaboration about every issue that involved not science but principles of action in the face of the approaching war. If the nation owes us gratitude—and it may not—it does so for having stuck it out together as long as it was necessary.
Reflecting on their years of work together, Szilard once wrote that “of all the many occasions which I had to observe Fermi I liked him best on the rather rare occasions when he got mad (except, of course, when he got mad at me).” Szilard recalled that when a young physicist visited New York one summer and asked Fermi’s help to use the library at Columbia, he personally talked to the librarian. She said the request must be decided by the central library. “If we admit every one,” they told Fermi, “our libraries will be overcrowded.” Fermi saw only two or three students loitering in the physics library. “Fermi got mad,” Szilard recalled. “He got mad for the right reasons and in the right degree and in the right manner. . . . Whatever grudges I might still have against Fermi I am willing [to] forgive them f
or the sake of this one heartwarming memory.”13
For his part, Fermi’s “grudges” against Szilard were reciprocal, but so were his “heartwarming memories.” A methodical worker, Fermi often rose before dawn to think and plan out his calculations for the day’s experiments: reviewing the work that had come before and anticipating what might be learned from the tasks ahead. Szilard often slept late, then soaked in his bathtub for inspiration. So it angered Fermi when Szilard appeared at his office door, suggested new experiments or lines of inquiry, then sauntered on down the hall. Fermi had the status of a group leader among his collaborators in Rome, and he quickly attracted a circle of dedicated graduate students and colleagues at Columbia. Szilard’s technique was to challenge every hierarchy he encountered.
Even more annoying to Fermi was Szilard’s refusal to work with his hands. As the graphite began to arrive at Columbia in the spring of 1940, wrapped in paper packages the size of bricks, Fermi and his colleagues Anderson and George Weil unwrapped and stacked them in four-foot-square piles. Inside the piles they embedded powdered uranium in tin cans. Then they measured how many neutrons from the uranium reached different parts of the piles. Soon the physicists “started looking like coal miners” from the graphite dust, Fermi recalled. Szilard often peered in on the large rooms on the Pupin Laboratory’s seventh floor, suggested new calculations and stacking methods, and then strolled off. This angered Fermi at first, especially when Szilard hired a burly undergraduate to do his share of the graphite-brick stacking. But Fermi soon mellowed and later admitted to colleagues that this was the best arrangement; because he lacked manual dexterity, the last thing Szilard should be doing is stacking graphite bricks.
Indeed, as more and more piles were created and activated, Pegram arranged to have the graphite stacked by members of Columbia’s football team. Fermi later praised Szilard’s “decisive and strong steps” to organize the pure graphite and uranium supplies that made the difference in their experimental success.
Reminiscing during a visit to Columbia in 1954, a few months before he died, Fermi remembered Szilard as “a very peculiar man, very intelligent.” His audience—mostly physicists—roared with laughter. “I see that is an understatement,” Fermi added. Again wild laughter. Fermi continued. “He is extremely brilliant, and he seems somewhat to enjoy, at least that is the impression that he gives to me, he seems to enjoy startling people.”14
But Szilard himself was startled in May 1940 when he opened a letter from Louis A. Turner, a theoretical physicist at Princeton whom he had met casually. Before American scientists reimposed self-censorship on their publications, Turner had surveyed nearly a hundred articles written about uranium fission and in the Reviews of Modern Physics concluded that a “reasonable possibility” existed for “utilizing the enormous nuclear energy of heavy atoms” and that “the practical difficulties can undoubtedly be overcome in time.”15 Now Turner had a more pressing concern. Physicists knew by this time that only the rare isotope uranium235 could be split or fissioned by slow neutrons. At first, that made uranium238, which is 140 times more abundant in nature, seem worthless to the chain-reaction enterprise. That spring, Szilard and Fermi had discovered in their experiments that when a uranium235 atom fissioned in natural uranium, at least one of the two escaping neutrons was absorbed by the nonfissionable uranium238. This “neutron absorption” they first considered a nuisance, until Turner’s letter gave it a frightening new importance.
Turner asked if he should publish an enclosed letter to the Physical Review, a letter that speculated how neutrons absorbed by uranium238 might transform it into a heavier element (later named plutonium239) that might itself be fissionable. Since uranium238 was so much more plentiful than uranium235, Turner saw this transformation as a way to create a new element for chain reactions.
Turner further suggested that a uranium-to-plutonium transformation might be truly beneficial if more plutonium were produced than uranium consumed. Turner asked if this notion should be withheld from publication “because of its possible military value.” He thought not, for “it seems as if it was wild enough speculation so that it could do no possible harm. . . .”16 But to Szilard, himself a master at wild speculation, the implications were stunning. “With this remark of Turner,” Szilard said later, “a whole landscape of the future of atomic energy rose before our eyes in the spring of 1940, and from then on the struggle with ideas ceased and the struggle with the inertia of Man began.”17
This letter, Szilard wrote to Turner, “will have to be delayed indefinitely in the same way as that of my own last paper,” the A-55 Report on uranium fission. Szilard hoped to meet Turner to discuss his ideas, which “might eventually turn out to be a very important contribution.” But Turner, still wondering why his “wild” speculation might need to be censored, challenged Szilard’s decision to withhold publication, seeking a more orderly system for review than “the accident of our being acquainted. . . .”18 In fact, Szilard now suspected, chain reactions—and perhaps bombs—might be made more easily with plutonium than with uranium, an intuition that proved correct.
Gregory Breit, the editor of the Physical Review, also wrote to Szilard, asking about Turner’s letter and seeking some “official channels” to decide on withholding publication. But the issue was quickly resolved. On June 7, Lyman Briggs, director of the National Bureau of Standards and chairman of the Advisory Committee on Uranium, invited Szilard to become a member of a new “Advisory Committee on Nuclear Physics,” with Nobel chemist Harold Urey, Breit, Pegram, the Department of Terrestrial Magnetism’s Merle Tuve, Fermi, and Szilard’s Hungarian friends Eugene Wigner and Edward Teller. Briggs offered Szilard five dollars a day for expenses, the first government support that he would personally receive for his chain-reaction work.19
Interest in chain reactions also gained attention within the government, largely because of concerns raised by Vannevar Bush, president of the Carnegie Institution of Washington and chairman of a policy panel at the National Academy of Sciences (NAS). On June 12, 1940, Bush and presidential adviser Harry Hopkins met with Roosevelt, who approved creation of a National Defense Research Committee (NDRC) that would direct scientific work for military uses. Bush’s new committee quickly absorbed Briggs’s Advisory Committee on Uranium, and with it the pioneering Columbia team of Fermi and Szilard.20
The day after Bush and Hopkins saw Roosevelt, Briggs’s new nuclear physics committee met in Washington and agreed that papers on uranium should be subject to “censorship.”21 At a later meeting, the committee also decided that research on a uranium-carbon experiment should follow two lines: further measurement of the radioactive elements’ nuclear constants and a search for a self-sustaining chain reaction.22
Back at Columbia, Szilard called on Pegram and urged that a “semi-large-scale experiment” using five tons of uranium metal should have priority over all other work. They also discussed Szilard’s role in conducting research at Columbia. Szilard considered approaching Fermi to propose a range of experiments “under joint direction,” and in a letter Szilard wrote to Fermi but did not send, he also took pains to review their uneasy relationship, concluding that if they had each worked alone with the necessary equipment, either one of them might have created a chain reaction by now.
“For us to work jointly in this matter has both its advantages and disadvantages, and we may at this juncture leave the question open whether the advantages outweigh the disadvantages from the point of view of obtaining speedy results,” Szilard concluded. He sought “a satisfactory arrangement” because accepting one “that I would inwardly, rightly or wrongly, not consider as fair and just in the circumstances . . . would put a strain on our collaboration.” Finally, Szilard said, if their effort taxed the Columbia department, then another laboratory—or work by other researchers—might be necessary. Clearly, Szilard was eager to create a chain reaction promptly and—though cautious—to enlist Fermi’s mind and methods in this quest.23
When Szilard pro
posed collaborating, in a July 4 letter to Fermi, he was very tentative and left the decision to Pegram.24 The compromise that evolved suited their different styles: Fermi took charge of all experimental work at Columbia, while Szilard generated fresh ideas and worked tirelessly to buy pure graphite and uranium supplies.25 By the late summer, when Fermi, Zinn, Weil, and others were toiling in the heat to construct small graphite and uranium piles, Szilard appeared on campus every day or so to urge them on, convinced that they were racing with Germany for an A-bomb. One Friday, Szilard walked in and, hearing that his colleagues planned to take off for a holiday weekend, berated them for all the work that remained. Reluctantly, they agreed to cancel their plans and spend the weekend on campus, continuing their arduous calculations and construction.
“Good for you,” Szilard said.
“Will you be around campus if we need your help?” asked one physicist.
“Oh, no,” answered Szilard, turning for the door. “I’m spending the weekend in the country.”26
In other ways, Szilard worked as hard, and as creatively, as anyone, although many of his insights came in unscientific settings. It was not in the lab but over lunch, at the Columbia Men’s Faculty Club with Fermi and two men from the National Carbon Company, that Szilard first made one decisive discovery. He pushed his guests for more details about the impurities in commercial-grade graphite as, one by one, he named elements and compounds that might absorb neutrons. Then, “half jokingly,” he mentioned an element that he knew gobbled neutrons.
Genius in the Shadows Page 31