The American Embassy quickly passed word that it could guarantee the Bohrs safe passage to the United States. Bohr again chose duty. His immediate concern was to burn the files of the refugee committee that had helped hundreds of emigres to escape to exile. “It was characteristic of Niels Bohr,” his collaborator, Stefan Rozental, writes, “that one of the first things he did was to contact the Chancellor of the University and other Danish authorities in order to protect those of the staff at the Institute whom the Germans might be expected to persecute.”1289 Those were Poles first of all, but Bohr also sought out government leaders to argue for concerted Danish resistance to any German attempt to install anti-Semitic laws in Denmark.
He even found time on the day of the occupation to worry about the large gold Nobel Prize medals that Max von Laue and James Franck had given him for safekeeping.1290 Exporting gold from Germany was a serious criminal offense and their names were engraved on the medals.1291, 1292 George de Hevesy devised an effective solution—literally: he dissolved the medals separately in acid. As solutions of black liquid in unmarked jars they sat out the war innocently on a laboratory shelf. Afterward the Nobel Foundation recast them and returned them to their owners.
Norsk Hydro was a prime German objective and there was heavy fighting around Rjukan, which held out until May 3, the last town in southern Norway to surrender. Then a management under duress reported to Paul Harteck that its heavy-water facility, the Vemork High Concentration Plant, could be expanded to increase production of the ideal neutron moderator to as much as 1.5 tons per year.
* * *
“What I should like,” Henry Tizard wrote Mark Oliphant after he had studied the Frisch-Peierls memoranda, “would be to have quite a small committee to sit soon to advise what ought to be done, who should do it, and where it should be done, and I suggest that you, Thomson, and say Blackett, would form a sufficient nucleus for such a committee.” Thomson was G. P. Thomson, J.J.’s son, the Imperial College physicist who had ordered up a ton of uranium oxide the previous year to study and felt ashamed at the absurdity. He had concluded after neutron-bombardment experiments that a chain reaction in natural uranium was unlikely and a war project therefore impractical. Tizard, who had been skeptical to begin with and had taken Thomson’s conclusions as support for his skepticism, appointed Thomson chairman of the small committee; James Chadwick, now at Liverpool, his assistant P. B. Moon and Rutherford protege John Douglas Cockcroft were added to the list. Blackett was busy with other war work, although he would join the committee later. The group met informally for the first time on April 10 in the Royal Society’s quarters at Burlington House.
It probably met as much to hear a visitor, the ubiquitous Jacques Allier of the Banque de Paris and the French Ministry of Armament, as to discuss the Frisch-Peierls work. Allier warned the British physicists about the German interest in heavy-water production and bid for collaboration on nuclear research between Britain and France. Only then, Thomson notes in the minutes he kept, did they consider “the possibility of separating isotopes . . . and it was agreed that the prospects were sufficiently good to justify small-scale experiments on uranium hexafluoride [a gaseous uranium compound].” They proposed rather ungenerously to remind Frisch to avoid “any possible leakage of news in view of the interest shown by the Germans.”1293 They were willing to inform him that his memorandum was being considered but not to supply details. (Peierls’ name seems not yet to have made an impression on Thomson, and Tizard apparently retained the second Frisch-Peierls memorandum in his files.) “We entered the project with more scepticism than belief,” the committee would report later, “though we felt it was a matter which had to be investigated.”1294 Thomson’s minutes make that skepticism evident. Tizard for his part wrote Lindemann’s brother Charles, a science adviser to the British Embassy in Paris, that he considered the French “unnecessarily excited” about the perils of German nuclear research.1295 “I still . . . think that [the] probability of anything of real military significance is very low,” he estimated in a note written the same week to the British War Cabinet staff.1296
It might have been as unpromising a start as the first meeting of the Briggs Uranium Committee had been, but the men on the Thomson committee were active, competent physicists, not military ordnance specialists, and whatever their initial skepticism they understood where the numbers Frisch and Peierls had used came from and what they might mean. At a second meeting on April 24 Thomson recorded laconically that “Dr. Frisch produced some notes to show that the uranium bomb was feasible.”1297 Many years later Oliphant recalled a more expansive response: “The Committee generally was electrified by the possibility.” Chadwick’s good opinion helped. He had just begun exploring fast-neutron fission himself with his new Liverpool cyclotron, the first in England, when he saw the Frisch-Peierls memorandum. At the April 24 meeting he awarded the emigres’ work chagrined confirmation: he “was embarrassed,” says Oliphant, “confessing that he had reached similar conclusions, but did not feel justified in reporting them until more was known about the neutron cross sections from experiments. Peierls and Frisch had used calculated values. However, this confirmatory evidence led the Committee to pay great attention to the development of techniques for . . . separation.”1298
Chadwick agreed to undertake the necessary studies. For several more weeks, until their protests through Oliphant registered with Thomson, Frisch and Peierls would be walled off from their own secrets. But work toward a bomb of chain-reacting uranium was now fairly begun, and this time it had found the right—fast—track.
* * *
Szilard chafed. The months after the first Uranium Committee meeting became “the most curious period of my life.” No one called. “We heard nothing from Washington at all. . . . I had assumed that once we had demonstrated that in the fission of uranium neutrons are emitted, there would be no difficulty in getting people interested; but I was wrong.”1299 The Uranium Committee’s November 1 report had in fact been languishing in Roosevelt’s files; Watson finally decided on his own in early February 1940 to bring it up again.1300 He asked Lyman Briggs if he had anything to add. Briggs reported the transfer, finally, of the $6,000 for Fermi’s work on neutron absorption in graphite. That was “a crucial undertaking,” Briggs said; he imagined it would determine “whether or not the undertaking has a practical application.”1301 He proposed to wait for results.
Something other than Briggs’ penurious methodology triggered a new burst of activity from Szilard. He had spent the winter preparing a thorough theoretical study, “Divergent chain reactions in systems composed of uranium and carbon”—divergent in this case meaning chain reactions that continue to multiply once begun (the document’s first footnote, numbered zero, cited “H.1302 G. Wells, The World Set Free [1913]”). Early in the new year Joliot’s group reported a uranium-water experiment that “seemed to come so close to being chain-reacting,” says Szilard, “that if we improved the system somewhat by replacing water with graphite, in my opinion we should have gotten over the hump.” He arranged lunch with Fermi to discuss the French paper. “I asked him, ‘Did you read Joliot’s paper?’ He said he did. I asked him, ‘What did you think of it?’ and Fermi said, ‘Not much.’ ” Szilard was furious. “At which point I saw no reason to continue the conversation and went home.”1303
He traveled again to Princeton to see Einstein. They worked up another letter and sent it under Einstein’s signature to Sachs. It emphasized the secret German uranium research at the Kaiser Wilhelm Institutes, about which they had learned from the physical chemist Peter Debye, the 1936 Nobel laureate in chemistry and director of the physics institute at Dahlem, who had been expelled recently to the United States, ostensibly on leave of absence, when he refused to give up Dutch citizenship and join the Nazi Reich. Sachs sent the Einstein letter on to Pa Watson for FDR. But Watson thought it sensible to check first with the Uranium Committee. Adamson responded, echoing Briggs: everything depended on the graphite measurements at
Columbia. Watson proposed to wait for the official report. Sachs may have rebutted; Roosevelt wrote the gadfly economist on April 5 emphasizing that the Briggs committee was “the most practical method of continuing this research” but also calling for another committee meeting that Sachs might attend.1304 Briggs dutifully scheduled it for Saturday afternoon, April 27.
In the meantime another development intervened. Alfred Nier at the University of Minnesota had gone to work, after Fermi wrote urging him again to do so, to prepare to separate measurable samples of U235 and U238. John Dunning sent him uranium hexafluoride, a highly corrosive compound that is a white solid at room temperature but volatilizes to a gas when heated to 140°F. “I worked with this for a couple of months in late 1939,” Nier remembers.1305 Unfortunately the gas was too volatile; it dispersed through Nier’s three-foot glass spectrometer tube despite the best efforts of his vacuum pump to clear it and contaminated the collector plates:
Finally I said, “This won’t do.” A new instrument was built in about 10 days in February, 1940. Our glass blower bent the horseshoe-shaped mass spectrometer tube for me; I made the metal parts myself. As a source of uranium, I used the less volatile uranium tetrachloride and tetrabromide left over from [his earlier] Harvard experiments. The first separation of U-235 and U-238 was actually accomplished on February 28 and 29, 1940. It was a leap year, and on Friday afternoon, February 29, I pasted the little samples [collected on nickel foil] on the margin of a handwritten letter and delivered them to the Minneapolis Post Office at about six o’clock. The letter was sent by airmail special delivery and arrived at Columbia University on Saturday. I was aroused early Sunday morning by a long-distance telephone call from John Dunning [who had worked through the night bombarding the samples with neutrons from the Columbia cyclotron]. The Columbia test of the samples clearly showed that U-235 was responsible for the slow neutron fission of uranium.
The demonstration vindicated Bohr’s hypothesis, but it also led Briggs to even greater suspicion of the value of natural uranium; it was “very doubtful,” he reported to Watson on April 9 “whether a chain reaction can be established without separating 235 from the rest of the uranium.”1306 Nier, Dunning and their collaborators Eugene T. Booth and Aristide von Grosse had written much the same thing in the Physical Review on March 15: “These experiments emphasize the importance of uranium isotope separation on a larger scale for the investigation of chain reaction possibilities in uranium.”1307 But isotope separation was Dunning’s approach to the problem in the first place and his enthusiasm as well; the slow-neutron finding hardly ruled out the Fermi-Szilard system. More misleading may have been the measurements Nier and the Columbia team published on April 15 using larger (but still microscopic) samples: “Furthermore, the number of fissions/microgram of U238 observed under these neutron intensity conditions, is sufficient to account for practically all the fast neutron fission observed in unseparated U.”1308 The statement was correct within the limits of measurement for such small samples, but its wording seems to deprecate U235 fast-neutron fission. In fact, Nier had not collected enough U235 to allow Columbia to measure that possibility. All anyone knew by then was that the U235 cross section for fast-neutron fission was less than the isotope’s cross section for slow-neutron fission. But that cross section, as the first Nier/Columbia paper reported, was a whopping 400 to 500 × 10−24cm2.1309
Predictably, then, when the Uranium Committee met on April 27, with Sachs, Pegram, Fermi, Szilard and Wigner in attendance, it listened to the renewed debate, squared its shoulders at Sachs’ exhortation to plunge ahead—and never wavered in its adamant conviction that a large-scale uranium-graphite experiment should await the outcome of Fermi’s graphite measurements.
* * *
Now that the $6,000 had been paid, Columbia was able to buy the graphite Szilard had tracked down for Fermi’s use. “Cartons of carefully-wrapped graphite bricks began to arrive at the Pupin Laboratory,” Herbert Anderson remembers, four tons in all. “Fermi returned to the chain reaction problem with enthusiasm. This was the kind of physics he liked best. Together we stacked the graphite bricks in a neat pile. We cut narrow slots in some of the bricks for the rhodium foil detectors we wanted to insert, and soon we were ready to make measurements.”1310
“So the physicists on the seventh floor of Pupin Laboratories started looking like coal miners,” adds Fermi, “and the wives to whom these physicists came back tired at night were wondering what was happening.”1311
The arrangement was designed to determine how far neutrons from a radon-beryllium source set in paraffin on the floor under the graphite column would diffuse up the column through the graphite after first slowing down in scattering collisions: the farther the neutrons traveled, the smaller was carbon’s absorption cross section and therefore the better moderator it would be. The Pupin seventh floor became a racetrack like the second floor of the institute in Rome. Anderson describes the scene:
A precise schedule was followed for each measurement. With the rhodium in place in the graphite, the source was inserted in its position inside the pile and removed after a one-minute exposure. To get the rhodium foil under the Geiger counter in the allotted 20 seconds [because its induced half-life is only 44 seconds] took coordination and some fast legwork. The division of labor was typical. I removed the source on signal; Fermi, stopwatch in hand, grabbed the rhodium and raced down the hall at top speed. He had just enough time to place the foil carefully into position, close the lead shield and, at the prescribed moment, start the count. Then with obvious satisfaction at seeing everything go right, he would watch the flashing lights on the scaler, tapping his fingers on the bench in time with the clicking of the register. Such a display of the phenomenon of radioactivity never failed to delight him.1312
The absorption cross section, as Fermi and Anderson subsequently calculated it, proved usefully small: 3 × 10−27cm2.1313 And could be made smaller still, they thought, with purer graphite. The measurement strongly supported Fermi’s and Szilard’s plan to attempt to induce a slow-neutron chain reaction in natural uranium.
But while such a plan might demonstrate a potential future source of power, the American scientists and administrators who were advising Briggs could not yet identify any military use. In April the British Thomson committee asked A. V. Hill, a scientific adviser to the British Embassy in Washington, to find out what the Americans were doing about fission. According to the official history of the British atomic energy program, Hill talked to unidentified “scientists of the Carnegie Institution,” whose opinions he reported pungently:1314
It is not inconceivable that practical engineering applications and war use may emerge in the end. But I am assured by American colleagues that there is no sign of them at present and that it would be a sheer waste of time for people busy with urgent matters in England to turn to uranium as a war investigation. If anything likely to be of war value emerges they will certainly give us a hint of it in good time. A large number of American physicists are working on or interested in the subject; they have excellent facilities and equipment: they are extremely well disposed towards us: and they feel that it is much better that they should be pressing on with this than that our people should be wasting their time on what is scientifically very interesting, but for present practical needs probably a wild goose chase.1315
The opinion from the Carnegie may have been hardheaded, but it was based on more than prejudice. Roberts, Hafstad and fellow DTM physicist Norman P. Heydenburg had improved their measurements of cross sections for fast-neutron fission, scattering and capture in natural uranium. Using their numbers, Edward Teller in one of the many calculations he made during this period arrived at a critical mass in excess of thirty tons, the same order of magnitude as Perrin and Peierls had calculated before him.1316 With only slightly more pessimistic assumptions Roberts concluded that “the cross-section for capture [in natural uranium] is sufficiently large that it now seems impossible for a fast-neutron chain reaction to
occur, even in an infinitely large block of pure uranium.”1317 By the spring of 1940 experiments at Columbia and the DTM had thus ruled out both slow- and significant fast-neutron fission in U238 and ruled in slow-neutron fission in U235. The asymmetry might have been a clue. No one picked it up.
* * *
Since at least the time of Einstein’s first letter to FDR, Edward Teller had debated within himself the morality of weapons work. His life had twice been cruelly uprooted by totalitarianism. He understood Germany’s frightening technological advantages at the outset of the war. “I came to the United States in 1935,” he notes. “ . . . The handwriting was on the wall. At that time, I believed that Hitler would conquer the world unless a miracle happened.”1318 But pure science still pacified him. “To deflect my attention from physics, my full-time job which I liked, to work on weapons, was not an easy matter. And for quite a time I did not make up my mind.”1319
The accidental juxtaposition of two events led him to decision. “In the spring of 1940 it was announced that President Roosevelt would speak to a Pan American Scientific Congress in Washington, and as one of the professors of George Washington University I was invited. I did not intend to go.”1320 The other event of that crucial day, May 10, 1940, reversed his intention: the phony war abruptly ended. With seventy-seven divisions and 3,500 aircraft Germany without declaration or warning invaded Belgium, the Netherlands and Luxembourg to make way for the invasion of France. Teller thought Roosevelt might speak to that outrage. In his voluntary prewar isolation he had never bothered, Teller says, to visit the Capitol or listen to one of FDR’s radio talks or otherwise involve himself in the political life of his adopted country, but he wanted now to see the President of the United States in person.1321