CHAPTER THIRTEEN

SPLITTING THE ATOM

EXACTLY TWO WEEKS AFTER HIS ARRIVAL IN NEW YORK, ON January 16, 1939, Fermi and his wife traveled back to the Manhattan piers, this time to meet Niels Bohr, who was arriving on the Swedish-American Line’s SS Drottningholm. Also there to meet the great man was John Wheeler, a twenty-seven-year-old Princeton physicist whose job it was to take Bohr and his son Erik and fellow passenger, Belgian physicist Leon Rosenfeld, to the Princeton campus.

Bohr was in town for two reasons. He wanted to visit Einstein at Princeton to discuss physics and the increasingly dire political situation in Europe. He also wanted to attend the fifth annual Washington Conference on Theoretical Physics, cosponsored by the Carnegie Institution, represented by Merle Tuve, and George Washington University, represented by Fermi’s old friend Edward Teller and Russian émigré George Gamow. The conference was scheduled for January 26, 1939.

The Fermis persuaded Bohr and his son to spend the day with them before heading to Princeton. Bohr, keeping his promise to Frisch, did not mention the Hahn-Strassmann-Meitner results to Fermi. He may also have wanted to spare Fermi the inevitable embarrassment the news would cause. Fermi would find out soon enough. Why should Bohr be the one to tell him? As it was, Fermi and Bohr spoke of other things and the Fermis were glad to see their Danish friend, a friendly and familiar face from across the Atlantic.

In the meantime, Wheeler whisked Rosenfeld off to Princeton. On the train ride, probably unaware of Bohr’s promise to Frisch, Rosenfeld broke the news to an astonished Wheeler. The next day, on January 17, 1939, Rosenfeld gave an impromptu talk on fission to the weekly physics department meeting, at Wheeler’s urging. I. I. Rabi, visiting from Columbia, was in attendance but, for reasons that are not entirely clear, did not bring the news of fission back to Columbia. On Friday, another Columbia physicist, Willis Lamb, visited Princeton and brought the news to Fermi the next morning.

There is no direct record of how Fermi reacted to the news. Lamb reported with presumed understatement that Fermi received the information with great interest. Fermi must have been mortified. His embarrassment that he might have been the one to discover fission some five years earlier was compounded by the fact that the Nobel Prize citation mentioned the now discredited discovery of transuranic elements. Additionally, Fermi must have felt that, given the depth of his understanding of nuclear physics, he should have been the one to have had the Meitner insight. It would have been natural for the man to kick himself. As the years passed, he developed a sense of humor about his mistake. After the war, when reviewing the construction plans for a new nuclear science institute at the University of Chicago, he and his colleagues were speculating on the identity of a vaguely defined human figure shown in bas-relief above the entrance. Fermi quipped he was “probably a scientist not discovering fission.”

On the other hand, none of the great scientists who followed his lead in March 1934, when he began the neutron experiments, came up with fission as a possibility, and these included truly brilliant experimentalists like Rutherford and Joliot-Curie. Fermi later told his wife that the Via Panisperna team was not good enough in chemistry, and he may have privately blamed D’Agostino, the team’s chemist. Such blame would be unfair, considering that the Dahlem team was arguably the best in the world in radiochemistry and that it took them four years of intensive work to unlock the secret. Confronted with the puzzling results of the Dahlem team as early as November 1938, even Bohr was unable to process it correctly. No one had seen the solution until it was too obvious to ignore. Segrè lays part of the blame for the Rome team’s failure on their experimental design. Shielding the uranium the way they did prevented them from seeing the ionizing pulse of energy, which would have naturally caused them to look for explanations leading to the discovery of fission.

In retrospect the true culprit was what psychologists call “cognitive dissonance.” First identified by academic psychologist Leon Festinger, cognitive dissonance occurs when we are confronted with empirical data at odds with the way we “know” the world to work. To resolve this discrepancy, we choose to ignore data or try to fit the data into our preconceived belief structure. Sometimes, there is a crisis and the belief structure eventually crumbles. Hahn, Rutherford, Joliot-Curie, Bohr, Fermi, and other major scientists involved in the analysis of uranium bombardment believed that particles in the nucleus were like bricks and mortar, rigid and incapable of deforming into the shape required for fission to occur. It was only when confronted repeatedly with evidence to the contrary that they realized they had all been wrong. The discovery of fission is a classic case of cognitive dissonance in action.

Soon after his visit to Princeton, Bohr was back in New York. Wandering around Pupin Hall, he ran into a young Columbia physics graduate student by the name of Herbert Anderson. Squirreled away in his lab, Anderson had missed the commotion over fission, and Bohr eagerly filled him in. Anderson, who had been looking for a way to introduce himself to Fermi, now saw his chance. When Bohr left, Anderson immediately went to Fermi’s seventh-floor office and introduced himself, mentioning the conversation with Bohr about fission. Characteristically, Fermi jumped at the opportunity to teach this young stranger something: “Let me tell you about fission!” He outlined the theory for Anderson, walked him through some ideas for experiments, and by the end of the conversation Anderson was offering Fermi the modest equipment he had in his lab. Anderson had figured out a way to connect an ionization chamber, capable of detecting an ionizing pulse, to an oscilloscope, invented in 1932 in Britain to display electrical signals on a small cathode-ray screen, similar to a television. Because uranium fission was accompanied by an ionizing pulse of some two hundred million electron volts, Anderson’s equipment seemed perfect for the job.

Fermi had found an eager collaborator and, even more important, someone he could teach. Anderson had found a mentor, and would play an important role in the Americanization of Fermi. The partnership would last for the next fifteen years.

ANDERSON WAS A SLIM, ATHLETIC, ATTRACTIVE TWENTY-FIVE-YEAR-old graduate student who had been helping John Dunning, a physics professor and head of the team, build Columbia’s first cyclotron, working out some of the kinks in the project. Anderson’s previous degree in electrical engineering enabled him to solve some thorny engineering and design problems that arose as they built the new machine. He was also ambitious and understood that befriending the legendary physicist could prove beneficial to his own career. He was right.

Fermi was happy to throw himself back into work after the revelation of the Hahn-Strassmann-Meitner breakthrough, eager to repeat the neutron bombardment experiment with proper instrumentation to see the ionizing pulse himself. With Anderson, Dunning, and several other Columbia physicists, Fermi set up the experiment, using two sources for neutrons, the Columbia cyclotron and a Rome-type radon-beryllium glass bulb, and Anderson’s ionizing chamber/oscilloscope for detection. Fermi had to leave for Washington before the results of the experiment came in, but Anderson assured him they would relay the results to him at the conference.

The Washington Conference was already quite important. At the previous one in 1938, Hans Bethe was so stimulated by the discussions during the conference that he used his train ride back from Washington to Cornell to work out the fusion cycle of the sun, effectively demonstrating why and how stars shine and create new elements in the process, work that won him a Nobel Prize some thirty years later. So it is not surprising that the attendees of the January 26, 1939, Washington Conference were a who’s who of contemporary theoretical physics in the United States and abroad: aside from Bohr and Fermi, the fifty-one participants included current and future luminaries such as Bethe, Gregory Breit, George Gamow, Maria Goeppert, I. I. Rabi, Edward Teller, Merle Tuve, George Uhlenbeck, and nuclear chemist Harold Urey.

The conference was supposed to discuss low-temperature physics. From the beginning, however, Bohr and Fermi stole the show. Bohr took the podium to announce that the uranium atom had been split. He then turned it over to Fermi, who discussed the theory behind the phenomenon, using Meitner’s work and the liquid drop analogy. This left the conference in an uproar. The Columbia experimental results came through during the conference and were duly reported by Fermi. The team in New York measured the ionization at 90 MeV, smaller than theoretically expected but within the error range of the equipment they were using. They had seen fission. Shortly after the conference, teams at Copenhagen, Johns Hopkins, the Carnegie Institution, Berkeley, and elsewhere climbed on to the experimental bandwagon, each one repeating the experiment and detecting the strong ionization pulse.

FIGURE 13.1. The Fifth Washington Conference on Theoretical Physics, January 1939. Fermi, who has just explained fission to the group, is smiling in the front row, second from the left. Bohr is fourth from the left. Hans Bethe is directly above Bohr. Maria Mayer and Edward Teller are sitting together, two seats to Bethe’s right. Courtesy of the Department of Physics, George Washington University.

If his colleagues had lost any respect at all for Fermi due to his 1934 oversight—and they had little reason to do so, considering that all of them were guilty of the same error—they quickly regained it through his frankness about it and his enthusiastic acceptance of the Hahn-Strassmann-Meitner results. At the Washington Conference he took center stage, a place where he naturally belonged. His colleagues agreed. In the group photo at the end of the conference, Fermi sat in the front row, beaming.

While Fermi was in Washington, Leo Szilard was battling a cold.

ON JANUARY 16, 1939, ONE WEEK BEFORE THE WASHINGTON Conference, on the same day that Bohr arrived in New York, Szilard took the train from New York to Princeton to visit his old friend Eugene Wigner, who was in the hospital suffering from a serious case of jaundice. The courtly and quiet Wigner, whom Szilard had known since their childhood in Budapest, was already one of the most respected quantum theorists in the world. After looking in on his friend, Szilard settled himself in Wigner’s apartment.

Wigner learned of Rosenfeld’s presentation at Princeton while still in the hospital and mentioned it to Szilard, who immediately grasped its significance. If, during the fission of uranium, neutrons were emitted, uranium could be the basis of the chain reaction that Szilard first envisioned in 1933.

While in Princeton, Szilard left Wigner’s apartment poorly protected during a heavy rainstorm and came down with a nasty cold. It is unclear whether Szilard had been invited to the Washington Conference—it was supposedly on low-temperature physics, something he knew a bit about. Szilard remained cooped up in Wigner’s apartment in Princeton for about ten days owing to his cold and spent the time alone obsessing on the chain reaction idea while the Washington Conference was underway without him.

By the time Szilard managed to get himself back to the King’s Crown hotel in Manhattan, the only thing he wanted to do was find Fermi, who would understand how fission and the chain reaction were related and who would appreciate the importance of the moment. Distressed by the news that Bohr and Fermi had already briefed some fifty physicists in Washington about fission, he also wanted to plead the case for secrecy over all future work on uranium. “Since it was a private meeting,” Szilard later recalled, “the cat was not entirely out of the bag, but its tail was sticking out.” Hahn had done his work in Berlin, and so Hitler now had access to the most dangerous information on the planet. No sane person would willingly make it easier for the Nazis to create a fission bomb. If Hitler succeeded in making a fission weapon, the war in Europe, now seemingly unavoidable, would be over before it even began.

Szilard visited Pupin Hall looking for Fermi but could not find him. He ran into Rabi and asked him to pass along his plea for secrecy to Fermi. Rabi agreed. The next day, still on the hunt for Fermi, Szilard looked in on Rabi and asked if he had spoken to Fermi. Rabi told him he had. Szilard asked Rabi for Fermi’s reaction, and Rabi reported that Fermi, proudly showing off his newly acquired American slang, had said “Nuts!” Szilard asked Rabi to elaborate, but Rabi declined, suggesting instead that they should walk over to Fermi’s office and hear it from the man himself.

When the two of them confronted Fermi, he seemed unsympathetic to Szilard’s insistence on secrecy. He explained that he understood the concept of the chain reaction quite well, but in his view secrecy was not necessary because the probability of a successful chain reaction occurring in uranium was too remote to be of any practical concern.

Rabi pressed him to define “remote possibility” to which Fermi, consistent with his view of the world, replied, “Well, ten percent.”

Rabi and Szilard were astonished. Rabi shot back, “Ten percent is not a remote possibility if it means we may die of it. If I have pneumonia and a doctor tells me that there is a remote possibility that I might die of it, and it’s ten percent, I get excited about it.”

When Fermi said he thought 10 percent was remote, he meant it. Some years later he was in conversation with a number of physicists when the subject of faster-than-light travel came up. Someone asked what the odds were that physicists would discover that the speed of light could be exceeded. Teller estimated one in ten million. Fermi estimated it at 10 percent. In his experience, 10 percent probabilities never happened, reflecting a deep-seated and somewhat unique view of how the world worked and what constituted a rare event.

Perhaps Fermi also dismissed the possibility of a chain reaction because he understood what it might lead to and was reluctant to start on the road toward a nuclear weapon. He may also have judged the project too difficult for Germany to pursue successfully. In any case, Fermi’s position was totally at odds with the usually sober Rabi, not to mention the usually volatile Szilard. Szilard later recalled this moment:

From the very beginning the line was drawn; the difference between Fermi’s position throughout this [period] and mine was marked on the first day we talked about it. We both wanted to be conservative but Fermi thought that the conservative thing was to play down the possibility that this may happen and I thought the conservative thing was to assume that it would happen and take all the necessary precautions.… I personally felt that these things should be discussed privately among the physicists of England, France, and America, and that there should be no publication on this topic if it should turn out that neutrons are, in fact, emitted and that a chain reaction might be possible.

While Fermi considered Szilard’s plea, the Hungarian also wrote to Joliot-Curie in Paris with the same request, but the French physicist was less accommodating than Fermi and flatly refused to be bound by secrecy.

FERMI AND SZILARD NOW SET OUT INDEPENDENTLY TO CONDUCT experiments to determine whether neutrons were emitted during uranium fission, and if so, how many. The results of these experiments would determine whether a chain reaction was possible. If on average more than one neutron was emitted each time a uranium nucleus split, then the type of cascade that Szilard had dreamed of for five years would be possible. If the average was either one neutron or even less, then the cascade would not occur and an explosion would not result.

Fermi brought Anderson under his wing. They placed a neutron source in a glass bulb and immersed it in a cylindrical tank of water, three feet high and three feet in diameter, measuring the induced radioactivity in a rhodium foil placed in the tank at various distances from the bulb. Then they put uranium oxide in the bulb along with the neutron source and compared the results with those of the bulb without the uranium oxide. They found a 6 percent increase in the radioactivity of the rhodium when the uranium was present—suggesting that the fission of uranium had resulted in neutron emission. By his calculations, Fermi believed that if the neutron emissions were from fission alone, this would suggest that two neutrons were produced for each fission reaction. A chain reaction was thus theoretically possible.

In parallel to Fermi, Szilard decided to do a similar experiment with a slightly different neutron source, producing only slow neutrons, thus eliminating the possibility that fast neutrons had merely knocked neutrons off the uranium nuclei without splitting them. He negotiated to “rent” a gram of radium from the Radium Chemical Company with a loan of $2,000 from a businessman friend and a letter of reference from Wigner himself, who assured the company that the experiments would be conducted under the auspices of a university. Not deterred by his lack of a faculty position at Columbia—or anywhere else, for that matter—the irrepressible Szilard then persuaded Pegram to authorize him to conduct experiments over a three-month period and recruited an eager young Columbia physics faculty member, Walter Zinn, to work with him. Zinn, a thirty-three-year-old Canadian who joined the Columbia faculty as an instructor in the early 1930s was happy to help, although this was the last time he was to work directly with Szilard, probably because of a difference in their experimental styles. He would soon become one of the core members of Fermi’s first nuclear reactor team.

Szilard’s experiment, a bit more complex than Fermi’s, used a target with a series of carefully nested boxes containing uranium and various moderating media such as paraffin and demonstrated neutron emission from fission. With this result in hand, Szilard offered up his source to Fermi and Anderson. When they substituted it into their experiment, there was a 30 percent increase in the detected radioactivity. In the paper he published with Anderson, Fermi was careful to note that direct comparisons could not be made between the two sources because the geometric configurations were so different, but once again he concluded it was probable that at least two neutrons were emitted for each fission reaction.

The results of these two parallel, relatively primitive experiments, completed in early March 1939, were not enough to determine the practicality of a chain reaction, but they were sufficient to persuade Szilard that the US government needed to be alerted. A recent Nobel laureate, an exceptionally clear lecturer, and the acknowledged expert on all things nuclear, Fermi was the natural person to present the discovery to the government, but sensing Fermi’s reluctance to enter the breach, Szilard arranged a meeting in Pegram’s office of Pegram, Fermi, and himself. He asked Wigner from Princeton to join them as well.

The discussion began with the enormous implications of the experimental results. Fermi was probably still of the view that an actual weapon would be impractical—no calculations, for example, had been done on how much uranium would be needed for a weapon—but up against a three-on-one assault, Fermi could not resist. Neither Wigner nor Pegram were excitable men, so their palpable alarm about the possibility of a German nuclear breakthrough must have impressed Fermi. As a foreign national, Fermi was naturally anxious about briefing the US military on such a sensitive subject only three months after arriving in New York. Szilard could be ignored, but Pegram and Wigner simply could not.

The conversation ended with a discussion of secrecy. Wigner joined Szilard in pressing for secrecy about work on the chain reaction. Fermi respected Wigner as much as he respected anyone, but he remained opposed to any restrictions on publication, deferring to Pegram to make the final decision. Pegram opted for sending papers on the two experiments to the Physical Review, but with the extraordinary request that their publication be delayed while the community considered the implications of the secrecy debate. In particular, Joliot-Curie in Paris was resisting any attempts to keep chain reaction research under wraps.

The next day Enrico Fermi met the US Navy.