The Doomsday Machine

by Daniel Ellsberg

  As a consequence of his reaction, many accounts describe Fermi’s offer as “a joke, intended to relieve the tension.” It’s unclear how anyone’s tension could have been relieved by this particular jest. But as William Laurence of the New York Times, permitted to chronicle the whole process and the testing of the bomb, put it at the time and in a retrospective later, “many of the scientists did not believe he was joking.” Indeed not: numerous accounts mention how anxious many of the participants were that night, especially the younger ones. That probably included those who had hit upon the possible phenomenon themselves and whose concerns had been met simply with formulaic assurances.

  As Peter Goodchild recounts, Fermi’s expression of uncertainty about the occurrence of atmospheric ignition had been neither a joke nor a last-minute tremor:

  In the final weeks leading up to the test214 Teller’s group were drawn into the immediate preparations when the possibility of atmospheric ignition was revived by Enrico Fermi. His team went to work on the calculations, but, as with all such projects before the introduction of computers, these involved simplifying assumptions. Time after time they came up with negative results, but Fermi remained unhappy about their assumptions. He also worried whether there were undiscovered phenomena that, under the novel conditions of extreme heat, might lead to unexpected disaster.

  As the test approached, Teller himself, Goodchild reports, “searched for and tested out hypotheses about such phenomena on anyone who would listen.” He was still doing this with Oppenheimer’s aide Robert Serber in the evening hours before the test. (Serber advised him to deal with the possibilities by bringing along a bottle of whiskey.)

  In 1982, Thomas Powers reported an interview with Stan Ulam, who in 1951 was the progenitor of the H-bomb along with Teller, that finally gives a sense of the measure of Fermi’s uncertainty that night. According to Ulam:

  Before the Trinity test,215 the physicist George Breit was given the job of estimating the chances that a nuclear bomb would ignite the earth’s entire atmosphere. The chance of this was very small, but after all, said Ulam, “the stake is infinite.… Fermi did the same calculations too.” He wanted to be sure. Theoretically, if the temperatures created by a nuclear explosion were high enough, the nitrogen in the atmosphere might spontaneously ignite. Fermi confirmed Breit’s calculations: such temperatures don’t exist in nature. On the long drive to Alamogordo for the Trinity test, Fermi joked about his conclusions. “It would be a miracle if the atmosphere were ignited,” he said. “I reckon the chance of a miracle to be about ten percent.”

  As Sam Allison, a physicist who had been assigned to “ride herd” on the final stages of the project, was counting down the last seconds over a loudspeaker, “Ten, nine, eight …” Davis reports that another young physicist had the responsibility of deciding whether to push a button that would abort the process. In those last seconds, he turned to Oppenheimer and said, “What if I just say this can’t go on and stop it?”

  Oppenheimer looked at him coldly and said, “Are you all right?” As Allison continued his countdown, “… five, four …” he was thinking partly, he told Davis, of “Fermi’s qualms,” which he shared. It had been his job, assigned by Oppenheimer, for the past six months to make sure that the project was moving ahead on schedule. But now, “for him it was no justification to say he had done what someone had told him to do; what right had he to participate in an experiment that might kill off the human race?” Seconds later, as the great light was followed by a blast wave that shook the bunkers and eventually subsided, Allison was musing, “Still alive.… No atmospheric ignition.”216

  Others watching ten miles away from ground zero had the same feeling of relief, for the same reason, having spent seconds before that fearing the opposite. One of these was James Conant, the president of Harvard who had oversight over the Manhattan Project as chairman of the NDRC. As Allison’s final countdown was echoing over a loudspeaker Conant whispered to Groves that he “never imagined seconds could last so long.” In his words:

  Then came a burst of white light that seemed to fill the sky and seemed to last for seconds. I had expected a relatively quick and light flash. The enormity of the light quite stunned me. My instantaneous reaction was that something had gone wrong and that the thermal nuclear transformation of the atmosphere, once discussed as a possibility and jokingly referred to a few minutes earlier, had actually occurred.

  His thought at that moment was, “The whole world has gone up in flames.”217a

  In short, the first Trinity test at Alamogordo constituted a conscious gamble by the senior scientists at Los Alamos and their immediate superiors: a gamble with the fate of every sentient being on the face of the planet and in the atmosphere and the depths of the oceans. It is noteworthy that it was the scientists alone who took on themselves the responsibility for this gamble. On the basis of any documentation that has survived or any recorded memories, there is no evidence that the possibility of atmospheric ignition was ever made known to the president or anyone else in Washington, D.C., outside the Manhattan Project, either in 1945 or in the three years since it had first been raised to Compton by Oppenheimer in July 1942.

  If it had been made known to top civilian officials—as it was exposed to Hitler, that same month in 1942, by Speer—how would they have reacted? Would President Roosevelt have sided with Compton’s first reaction: that “no chance” of this event was acceptable, no matter how small the probability? Or with his judgment shortly after that the risk was sufficiently small that continued development work was appropriate?

  Probably the latter, since only the research was in question at that point, and for the next several years. After all, in June 1942, the scientists had every reason to fear that the Germans might develop the bomb before we did, and for the policy makers there was still a lively concern that Germany might win the war without a bomb. But none of that was still true in July 1945, when the effort had come to the point of actually detonating a test device without having put entirely to rest the possibility of atmospheric ignition.

  Might President Truman or Secretary of War Henry Stimson—if they had been aware of the possible loss of all life on earth, forever!—have demanded odds better than three in a million, let alone Fermi’s “ten percent”? As it was, in ignorance of any reason for anxiety as they awaited reports at the Potsdam Conference in Germany, hoping for news that would strengthen their hand in negotiations with the Soviets, they learned of its success without Sam Allison’s sense of relief. And they were equally ignorant of some scientists’ continued cause for apprehension about the longer-term effects of this result, and still more about the experiments on humans that lay just ahead in Japan.

  Part of that concern—for some the smaller part—had to do with the people who would be killed by those further explosions. Allison had qualms about that prospect within minutes after the test, as soon as his fears of burning everyone on earth had dissipated. “Oh, Mr. Conant,” he said, in anguish, “They’re going to take this thing over218 and fry hundreds of Japanese.” His estimate was low by three orders of magnitude, a thousand times.

  At the May 31 meeting, Oppenheimer had estimated that the first bomb219 would kill about twenty thousand people. It immediately killed four times that many, but that was still fewer than the hundred thousand who had been burned alive in one night by the Tokyo firebombing. The readiness of the highest civilian and military officials to allow General Curtis LeMay to multiply that scale of civilian death several times over had been thoroughly tested in the months since. They had all passed that test. Likewise, by late July the scientists had demonstrated their own readiness to take a sufficiently small chance (for Fermi, not so small) of burning up all life on the planet.

  According to Albert Speer,220 this would not have surprised Adolf Hitler. In June 1942, Hitler occasionally “joked that the scientists in their unworldly urge to bare all the secrets under heaven might some day set the globe on fire. But undoubtedly a good deal of time w
ould pass before that came about, Hitler said; he would certainly not live to see it.” Actually he died, by his own hand, only ten weeks before the Trinity experiment.

  Those who undertook that gamble in July 1945 did not appear to fit the stereotype of the “mad scientist”: though in the light of this long-unknown history, the notion is not so far removed from reality. But though they did expect to win that particular bet, with high probability, they were also aware—again more so, it seems, than their civilian superiors—that they were simultaneously engaged in a longer-term gamble imperiling the survival of humanity.

  First, some of them (not all) were convinced that even a unilateral U.S. test, still more the unwarned use of the bomb on cities in wartime—in the absence of collaboration with the Soviets and of international controls—virtually assured a desperate postwar nuclear arms race with the Soviets. Second, nearly all understood that such a race would probably lead in a few years’ time to the production of thermonuclear weapons on both sides. Bombs with a million times the explosive power of the largest blockbusters of World War II; thousands of them. These two developments together—the latter recognized in July 1942 at the same moment as the possibility of atmospheric ignition—foretold the distinct possibility of destroying the whole of human civilization. Total incineration of the world of cities of the last four thousand years. And with a probability of a lot more than three in a million.

  Back in Washington, James Conant wrote up his notes on the Trinity test for his boss, Vannevar Bush. Conant concluded by suggesting that his first few seconds’ sense that they had participated in the destruction of humanity might have been prescient. “My first impression remains the most vivid, a cosmic phenomenon like an eclipse. The whole sky suddenly full of white light like the end of the world. Perhaps my impression was only premature on a timescale of years!”

  George Kistiakowsky’s reaction to the flash was much the same as Conant’s. He told the New York Times reporter who had witnessed the spectacle from ten miles farther away that it was “the nearest thing to Doomsday221 one could possibly imagine.”

  That was mistaken. More than three years earlier, Enrico Fermi had stirred Edward Teller to imagine—and for the next nine years, obsessively to pursue—an explosion a thousand times nearer to Doomsday than the one they witnessed at Alamogordo.b

  a I was told by the daughter of a scientist who had experienced the test lying near James Conant that when the preternaturally intense white light first enveloped them, Conant’s first thought, he said later, was, “Fermi was right.” It was my hearing this comment at a reception at the University of Colorado Boulder in 1982 that first attracted my attention to this issue and led to my talk with David Hawkins a few days later.

  b The yield of the first droppable H-bomb tested by the United States in 1954 was fifteen megatons. That is a million times more explosive power than the largest blockbusters in World War II. The largest warhead ever tested, fifty-eight megatons, was detonated by the Soviets in 1961. The yield was 250 percent greater than the largest yield that had been predicted for it, six megatons, resulting—along with an unexpected shift in wind—in heavy radioactive fallout contaminating inhabitants of the Marshall Islands and the crew of the distant Japanese fishing boat Lucky Dragon, one of whom died. The reason for the great underestimate of yield, with its serious human consequences, was precisely the kind of scientific error or unforeseen reactivity that Fermi had feared in connection with the possibility of atmospheric ignition from the Trinity test. Los Alamos bomb designers had neglected (or greatly underestimated) the contribution to the production of neutrons and to the yield from one of the isotopes included in the hydrogen fuel, lithium-7, which had been thought to be relatively inert but proved not to be under the unprecedented conditions of the dry-fuel thermonuclear detonation. (See Alex Wellerstein, “Castle Bravo Revisited,” Restricted Data, June 21, 2013, and comments: blog.nuclearsecrecy.com/2013/06/21/castle-bravo-revisited/.)

  CHAPTER 18

  Risking Doomsday II

  The Hell Bomb

  In July 1942, on the way to the U.C. Berkeley conference that preceded the formal launch of the Manhattan Project, Edward Teller shared a train compartment with his close friend Hans Bethe. During the trip, he told Bethe that “the fission bomb was all well and good222 and, essentially, was now a sure thing. He said that what we really should think about was the possibility of igniting deuterium by a fission weapon—the hydrogen bomb.” This was the idea that Teller would lay out on the blackboard in Berkeley’s Le Conte Hall—simultaneously introducing the possibility of atmospheric ignition. The theorists present spent most of the remaining four weeks of the conference discussing the concept of Teller’s “Super,” some of them with a sense of foreboding.

  Goodchild recounts:

  Hans Bethe recalled talking to his wife223 [also a physicist, though not cleared for this], who knew in broad terms what they were discussing, “and on a walk in the mountains in Yosemite National Park she asked me to consider carefully whether I really wanted to continue to work on this. Finally, I decided to do it.” For Bethe, the Super was a terrible thing, but its development was inextricably linked to the German threat, and to the fission bomb. This was, after all, to be the indispensable trigger for a thermonuclear reaction and, because of the Germans, they were committed to developing a fission weapon anyway. So for the time being, any moral dilemma associated with the Super itself could be held in abeyance.

  But in June 1945, a month before the Trinity test in New Mexico, for some of the scientists who understood (unlike any of the decision makers in Washington) that what was about to be tested was a potential trigger to a hydrogen bomb—and trigger to a thermonuclear arms race with the Soviet Union—facing up to that moral dilemma could no longer be postponed. With most of those at Los Alamos preoccupied with the last-minute technical issues of producing and testing the fission bombs, some scientists in the Chicago lab of the Manhattan Project focused then, belatedly, on the long-run implications of nuclear weapons in a committee chaired by James Franck and strongly influenced by Leo Szilard.

  They concluded, in a report that never reached the president, that using the bomb against Japan, especially without warning and without direct Soviet participation in the testing, would make international control of the weapon very unlikely. In turn, that would make inevitable a desperate arms competition, which would before long expose the United States to uncontrolled possession by adversaries of thermonuclear weapons. As a result, some of them said in a prescient pre-attack petition to President Truman, “the cities of the United States224 as well as the cities of other nations will be in continuous danger of sudden annihilation.”

  Szilard was the driving force behind this petition. Its many signers in the project tried to caution the president—on both moral grounds and considerations of the long-run survival of civilization—against beginning this process by using the bomb against Japan even if its use might shorten the war and save the lives of American troops.

  But their petition was sent “through channels” and was deliberately held back by General Leslie Groves, director of the Manhattan Project. It never got to the president, or even to Secretary of War Henry Stimson, until after the bomb had been dropped. There is no record that the scientists’ concerns about the future impact of nuclear attacks on Japan were ever made known to President Truman before or after his decisions. Still less were they made known to the American public.

  At the end of the war the petitions and their reasoning were reclassified secret to keep them from public knowledge, and their existence was unknown for more than a decade. Several project scientists later expressed regret that they had earlier deferred to the demands of the secrecy managers—for fear of losing their clearances and positions, and perhaps facing prosecution—and had collaborated in maintaining public ignorance on this most vital of issues.

  One of them, Eugene Rabinowitch—a physicist who had been rapporteur for the Franck Committee and who after the war f
ounded and edited the Bulletin of the Atomic Scientists (with its Doomsday Clock)—had in fact, after the German surrender in May, actively considered breaking ranks and alerting the American public to the existence of the bomb, the plans for using it against Japan, and the scientists’ views both of the moral issues and the long-term dangers of doing so.

  Rabinowitch first reported this in a letter to the New York Times published on June 28, 1971. It was the day I submitted to arrest at the federal courthouse in Boston, so I didn’t see it that day or for many years afterward. For thirteen days before it was published, my wife and I had been underground, eluding the FBI while distributing the Pentagon Papers to seventeen newspapers after injunctions had halted publication in the New York Times and the Washington Post.

  The Rabinowitch letter began by saying it was “the revelation by the Times of the Pentagon history of U.S. intervention in Vietnam, despite its classification as ‘secret’ ” that led him now to disclose the following for the first time:

  Before the atom bomb-drops on Hiroshima and Nagasaki, I had spent sleepless nights thinking that I should reveal to the American people, perhaps through a reputable news organ, the fateful act—the first introduction of atomic weapons—which the U.S. Government planned to carry out without consultation with its people. Twenty-five years later, I feel I would have been right if I had done so.

  Rereading this, still with some astonishment, I agree with him. He was right to consider it, and he would have been right if he had done it. He would have faced prosecution and prison (as I did, at the time his letter was published), but he would have been more than justified, as a citizen and as a human being, in informing the American public and burdening them with shared responsibility for the fateful decision (even though, as he said later, he had no expectation that they would have demanded a different decision).