team found 3.5 secondary neutrons per fission. “The interest of the phenomenon discussed here as a means of producing a chain of nuclear reactions,” the three men wrote, “was already mentioned in our previous letter.” Now they concluded that if a sufficient amount of uranium were immersed in a suitable moderator, “the fission chain will perpetuate itself and break up only after reaching the walls limiting the medium. Our experimental results show that this condition will most probably be satisfied.” That is, uranium would most probably chain-react.
Joliot's was an authoritative voice. G. P. Thomson, J.J.'s son, who was professor of physics at Imperial College, London, heard it. “I began to consider carrying out certain experiments with uranium,” he told a correspondent later. “What I had in mind was something rather more than a piece of pure research, for at the back of my thoughts there lay the possibility of a weapon.” He applied forthwith to the British Air Ministry for a ton of uranium oxide, “ashamed of putting forward a proposal apparently so absurd.”
More ominously, two initiatives originated simultaneously in Germany as a result of the French report. A physicist at Gottingen alerted the Reich Ministry of Education. That led to a secret conference in Berlin on April 29, which led in turn to a research program, a ban on uranium exports and provision for supplies of radium from the Czechoslovakian mines at Joachimsthal. (Otto Hahn was invited to the conference but arranged to be elsewhere.) The same week a young physicist working at Hamburg, Paul Harteck, wrote a letter jointly with his assistant to the German War Office:
We take the liberty of calling to your attention the newest development in nuclear physics, which, in our opinion, will probably make it possible to produce an explosive many orders of magnitude more powerful than the conventional ones… That country which first makes use of it has an unsurpassable advantage over the others.
The Harteck letter reached Kurt Diebner, a competent nuclear physicist stuck unhappily in the Wehrmacht's ordnance department studying high explosives. Diebner carried it to Hans Geiger. Geiger recommended pursuing the research. The War Office agreed.
A public debate in Washington on April 29 paralleled the secret conference in Berlin. The
Tempers and temperatures increased visibly today among members of the American Physical Society as they closed their Spring meeting with arguments over the probability of some scientist blowing up a sizable portion of the earth with a tiny bit of uranium, the element which produces radium.
Dr. Niels Bohr of Copenhagen, a colleague of Dr. Albert Einstein at the Institute for Advanced Study, Princeton, N.J., declared that bombardment of a small amount of the pure Isotope U235 of uranium with slow neutron particles of atoms would start a “chain reaction” or atomic explosion sufficiently great to blow up a laboratory and the surrounding country for many miles.
Many physicists declared, however, that it would be difficult, if not impossible, to separate Isotope 235 from the more abundant Isotope 238. The Isotope 235 is only 1 per cent of the uranium element.
Dr. L. Onsager of Yale University described, however, a new apparatus in which, according to his calculations, the isotopes of elements can be separated in gaseous form in tubes which are cooled on one side and heated to high temperatures on the other.
Other physicists argued that such a process would be almost prohibitively expensive and that the yield of Isotope 235 would be infinitesimally small. Nevertheless, they pointed out that, if Dr. Onsager's process of separation should work, the creation of a nuclear explosion which would wreck as large an area as New York City would be comparatively easy. A single neutron particle, striking the nucleus of a uranium atom, they declared, would be sufficient to set off a chain reaction of millions of other atoms.
The
There is one line of attack that deserves strong effort, and that is where we need your cooperation… It is of the utmost importance to get some uranium isotopes separated in enough quantities for a real test. If you could separate effectively even tiny amounts of the two main isotopes [a third isotope, U234, is present in natural uranium to the trace extent of one part in 17,000], there is a good chance that [Eugene T.] Booth and I could demonstrate, by bombarding them with the cyclotron, which isotope is responsible. There is no other way to settle this business. If we could all cooperate and you aid us by separating some samples, then we could, by combining forces, settle the whole matter.
The important point for Dunning, the reason for his passion, was that if U235 was responsible for slow- neutron fission, then its fission cross section must be 139 times as large as the slow-neutron fission cross section of natural uranium, since it was present in the natural substance to the extent of only one part in 140. “By separating the 235 isotope,” Herbert Anderson emphasizes in a memoir, “it would be much easier to obtain the chain reaction. More than this, with the separated isotope the prospect for a bomb with unprecedented explosive power would be very great.”
Fermi urged Nier in similar terms; Nier recalls that he “went back and figured out how we might soup up our apparatus some in order to increase the output… I did work on the problem, but at first it seemed like such a farfetched thing that I didn't work on it as hard as I might have. It was just one of a number of things I was trying to do.”
Fermi in any case was more interested in pursuing a chain reaction in natural uranium than in attempting to separate isotopes. “He was not discouraged by the small cross-section for fission in the natural [element],” comments Anderson. “‘Stay with me,’ he advised, ‘we'll work with natural uranium. You'll see. We'll be the first to make the chain reaction.’ I stuck with Fermi.”
By mid-April Szilard had managed to borrow about five hundred pounds of black, grimy uranium oxide free of charge from the Eldorado Radium Corporation, an organization owned by the Russian-born Pregel brothers, Boris and Alexander. Boris had studied at the Radium Institute in Paris; Eldorado speculated in rare minerals and owned important uranium deposits at Great Bear Lake in the Northwest Territories of Canada.
Like Fermi's and Anderson's previous experiment, the new project involved measuring neutron production in a tank of liquid. For a more accurate reading the experimenters needed a longer exposure time than their customary rhodium foils activated to 44-second half-life would allow. They planned instead simply to fill the tank with a 10 percent solution of manganese, an ironlike metal that gives amethyst its purple color and that activates upon neutron bombardment to an isotope with a nearly 3-hour half-life. “The [radio]activity induced in manganese,” they explained afterward in their report, “is proportional to the number of [slow] neutrons present.” So the hydrogen in the water would serve to slow both the primary neutrons from the central neutron source and any secondary neutrons from fission, and the manganese in the water would serve to measure them — a nice economy of design.
Atoms on the surface of a mass of uranium are exposed to neutrons more efficiently than atoms deeper inside. Fermi and Szilard therefore decided not to bulk their five hundred pounds of uranium oxide into one large container but to distribute it throughout the tank by packing it into fifty-two cans as tall and narrow as lengths of pipe — two inches in diameter and two feet long.
Packing cans and mixing manganese solutions, which had to be changed and the manganese concentrated after each experimental run, was work. So was staying up half the night taking readings of manganese radioactivity. Fermi accepted the challenge with gusto. “He liked to work harder than anyone else,” Anderson notes, “but everyone worked very hard.” Except Szilard. “Szilard thought he ought to spend his time thinking.” Fermi was insulted. “Szilard made a mortal sin,” Segre remembers, echoing Fermi. “He said, ‘Oh, I don't want to work and
