Enrico Fermi and Ethical Problems
in Scientific Research


Richard L. Garwin

Senior Fellow for Science and Technology
Council on Foreign Relations, New York


IBM Fellow Emeritus
IBM Research Division
P.O. Box 218
Yorktown Heights, NY 10598

Tel: (914) 945-2555
fax: (914) 945-4419
Email: RLG2 at watson.ibm.com
Many papers at http://www.fas.org/rlg

Public Lecture in the Centennial Celebration
"Enrico Fermi and Modern Physics"


October 19, 2001

ABSTRACT. As part of the celebration in Pisa of the 100th anniversary of the birth of Enrico Fermi, I review in this public lecture my personal experiences with Fermi, his role in developing nuclear fission as a weapon, and the early history of the hydrogen bomb. I then discuss the efforts to control the hazards of nuclear weaponry and the other most serious problem facing humanity-- biological weapons in the hands of states or particularly terrorists. I discuss also the responsibilities of scientists in their own research and in societal decisions, the future of nuclear power and of nuclear weapons, and close with a plea in October 2001 for recognizing serious peril and for providing means for responding to and thus preventing the deaths of millions and hundreds of millions of people from biological terrorism.


We have had the pleasure of a two-day physics symposium celebrating the centennial of the birth of Enrico Fermi. The participants in the symposium are well aware of Fermi's contributions and the extent to which the world and physics are in his debt.

But the general public might benefit from hearing first some of the story of this great man, to introduce the topic on which I was invited to talk-- ethical problems in scientific research.

In preparing these remarks, I have been greatly aided by a number of works, which will repay the reader's time manyfold.(1),(2),(3),(4),(5),(6) After my comments on Fermi's youth and career, I discuss ethical problems in scientific research in which Enrico Fermi played a central part-- the possibility of nuclear weapons, making and using the first such, the control of nuclear weapons, and the decision to build the hydrogen bomb. Fermi's untimely death at the age of 53 precluded his involvement in most of the events of the last half century.

Archimedes and DaVinci must have faced ethical problems in the field of weapons. Fermi, Szilard, Hans Bethe, Edward Teller, and Andreii Sakharov provide varied examples with nuclear weapons in particular. There are also the more general questions of publication of "dangerous" knowledge-- in the nuclear, biological, and chemical field. These will provide ample substance for this discussion.

There are, of course, more routine ethical problems in the conduct of scientific research-- ranging from the treatment of colleagues, the allocation of credit, the discussion of errors, and the concealment of facts. Add to this, problems of "informed consent" in medical trials, and the like. But these questions would take us far afield from the intent of the organizers of this symposium, and I have no more competence to discuss them than would the next scientist. They are important, though, and deserve attention in the conduct of scientific research and the education of scientists and engineers. A little book by E. Bright Wilson, "An Introduction to Scientific Research" was required reading for each of my graduate students; in fact, I gave each a copy.

I was a colleague and, I hope, friend of Enrico Fermi, at the University of Chicago. I began working with him as an informal assistant in 1948; and I was fortunate to have him as my thesis advisor until I received my degree in December 1949. Then I was on the Physics faculty at the University of Chicago with him until I left for a position at the IBM Corporation in New York in December 1952. Fermi and I worked together several summers at the Los Alamos nuclear weapons laboratory, beginning in 1950, but he was not one to discuss other than the technical problems with which we were faced. And he did not leave a body of reflective work aside from his published scientific papers, so it is difficult to infer what Fermi thought about many of these matters.


Enrico Fermi was born in Rome September 29, 1901. His brother, Giulio, was a year older, and the two boys were raised by a wet nurse in the countryside, Enrico being returned to his family only at the age of 2,1/2. Initially a sickly child, he grew to be sturdy like his father and grandfather. He and Giulio were the closest of friends until Giulio's tragic death at the age of 15 under anesthesia for a minor operation on an abscess of the throat. The Fermi brothers had occupied themselves in large part with the investigation of nature, with the building and testing of electric motors, and the attempt to understand the world around them.

Enrico Fermi was devastated by his brother's death and took refuge in independent study. In this he was greatly aided by the friendship and advice from one of his father's colleagues, Ingegner Adolfo Amidei, who lent Fermi substantive, if old, books from his library.

Fermi's depth and talent was evident at this age. He read the books-- first mathematics, then physics-- and returned them, having worked all the problems. And he retained this knowledge to the end of his life. After Giulio's death, Enrico Fermi found a boy of similar interests, Enrico Persico; the two of them shared scientific books bought in Rome's Campo dei Fiori, and became fast friends-- a friendship that would continue for many years.

Upon Fermi's graduation from school, Amidei played a crucial role by recommending to him that he apply for the Reale Scuola Normale Superiore at Pisa, rather than attend the university in Rome, where the Fermis had moved before Enrico's birth. It was a difficult decision for the family, but Amidei and Enrico persuaded Fermi's mother and father. At the Scuola Normale, Fermi had no difficulty with his studies, and in fact taught his teachers a good deal.

After receiving his Doctor in Physics in July, 1922, Fermi returned to Rome, where he was close to Professor Orso Mario Corbino, the head of the Physics Department. Soon Fermi went to Goettingen to study with the well-known physicist Max Born. In his seven months in Goettingen, he got to know other modern physicists, including Werner Heisenberg. Fermi then returned to Rome to teach elementary mathematics at the university.

His next position was to teach mathematics and mechanics in Firenze after he had tried unsuccessfully for a post in Cheliari. But Corbino soon brought Fermi to Rome in the Fall of 1926 to hold a new chair as Professor of Theoretical Physics. Fermi had become fast friends at Pisa with Franco Rasetti, for whom physics was too simple; Rasetti was to become a phenomenal naturalist with a great specialty in trilobites. Corbino brought Rasetti to Rome so that there were now two young teachers in the building on the Via Panisperna.

Just before returning to Rome, Fermi had made a major contribution with his "Fermi statistics" (which he never in his life called by that term) a concept absolutely fundamental to the understanding of solids, especially metals, and nuclei, and the world in general. It is easy enough to accept that not all of the electrons in an atom can be in the one state of lowest energy; in fact, the Pauli Exclusion Principle explains that only two electrons of opposite spin can occupy the same state-- else there would be no chemistry, and we would not exist. Fermi extended this in a simple and elegant way to the understanding of vast assemblages of electrons.

Fermi's understanding of the world was astonishingly clear. When he learned something, he wanted to work it out from first principles and then to apply it to other phenomena. So it was with his understanding of the field theory of light, in which photons are created or destroyed. Applying these concepts, as never before, to four fields (and their particles)-- the proton, neutron, electron, and a hypothetical particle, the neutrino-- he created the theory of beta decay. This quantitatively accounted for that type of radioactivity (for instance of the naturally occurring element potassium) in which electrons emerge from the nucleus with high energy. The existence of a neutrino had been postulated to take up a good deal of the energy difference between the initial radioactive nucleus and the resulting stable nucleus. The two nuclei differ by only about one one-thousandth of a proton mass, but constitute different chemical elements in view of the fact that a neutron changes into a proton during the decay.

The alpha particles emitted from a particular heavy element such as uranium, or especially radium, have identical energy or come in a series of discrete energies; in contrast, the electrons in beta decay may have an energy from zero to a maximum. The remaining energy in each case is imagined to be taken up by a neutrino-- a particle of zero charge. It was Fermi's invention to couple these four fields in a rigorous theoretical framework from which the detailed shape of the electron spectrum could be deduced, as well as the lifetime of the radioactive material.

Following the discovery of artificial radioactivity by Irene and Pierre Joliot-Curie, Fermi began his own experimental studies of radioactivity induced by neutrons. He initiated this work alone in Rome with neutron sources he constructed himself-- small glass tubes containing radon and powered beryllium. Fermi published his work March 25, 1934 as "Radioactivity Induced by Neutron Bombardment."

In these experiments Fermi showed the qualities that had marked him since he was a student-- precision, insight, willingness (even alacrity) to do things himself; add to this both persistence and a commitment to the routine. The radon gave off gamma rays, and the neutrons from the radium-beryllium source in any case prevented sensitive counters from detecting the weak radioactivity induced by the neutrons. The radioactive source could not be turned off; instead, Fermi rapidly transported the sample he was irradiating from the irradiation room to a room at the other end of the corridor. There he had set up sensitive Geiger counters he had made himself from aluminum tubes which had held cigars. In this way, he was able to demonstrate a radioactivity induced in fluorine with a half-life of about ten seconds. Visitors to the laboratory in Rome and later to the laboratory at Columbia would be treated to the sight of a small man in a white lab coat bursting out of one room and racing down the corridor to enter another room, followed by silence. Often, two men would race down the corridor in Rome, disputing whether Enrico Fermi or Edoardo Amaldi was the more fleet of foot.

Fermi's work on fission, on the nuclear reactor and the production of plutonium, and on the building of the fission bombs is discussed at length later. His work on neutrons at Rome posed no ethical problems, nor did that on mesons and high-energy physics at Chicago after the war. Returning to Chicago on the first day of 1946 after a year and a half in Los Alamos, New Mexico, Fermi was the acknowledged leader of the brilliant group assembled there. At seminars and colloquia he would interpret the efforts of others, and clarify and often extend their remarks. He was always ready to calculate with his ever-present sliderule and to estimate with uncanny accuracy. He was ever helpful with the problems of others-seemingly having finished his own work in the hours before arriving at the laboratory at 7:30 am.

When I joined his laboratory in 1948, bored with my classroom studies, I found that Fermi had machine tools in his lab, including a lathe, which he enjoyed using. An outstanding and practical theorist, he was also of great experimental skill. For the Chicago synchrocyclotron, which was under construction in 1950, and which Fermi used to good effect in studies of the interaction of pi mesons with nuclei, Fermi realized that the external beams of particles from the cyclotron could be used far more effectively if the target to be struck by the accelerated protons could be moved to different places within the cyclotron. To this end be built with his own tools a "trolley" which made use of the grooved steel pole face of cyclotron, and of the large steady magnetic field. This four-wheeled cart had simple coils attached to one wheel (much as the brothers Giulio and Enrico might have designed when they were young teenagers in Rome). The experimenter could move the cart remotely in a precise fashion, as if it were geared to the pole, and thus tune the desired beam while the cyclotron was in operation. Fermi soon fitted his cart with a thermocouple to measure the heat produced by the protons passing through the target, thus to be able to monitor and improve the performance of the cyclotron itself.

Fermi was a great teacher both in large groups and one-on-one. His students tried to emulate him in "building physics from the ground up" and some of them succeeded.

Fermi was ever secure, but modest, in his outstanding intelligence and achievements, but about his physical endurance and eyesight, he would joke that they were superior to those of others because his own heart or eyes were "made to order."

During the summer of 1954, Fermi was in Europe where he had lectured at Les Houches and given a Fermi-quality course on pions and nucleons at the summer school of the Italian Physical Society in Varenna. But he was unable to eat and on his return to Chicago was diagnosed with inoperable stomach cancer. He died November 29, 1954. It was a tremendous loss to the world of science, and to his friends and colleagues. In a 1955 tribute, Edoardo Amaldi emphasized not only Fermi's extraordinary contributions as a physicist-- both experimental and theoretical-- but especially his qualities of simplicity and serenity.

I have often wondered, to this day, what use Enrico Fermi would have made of the marvelous advances since his death-- from the hand-held programmable calculator and the enormous power of the personal computer, to the revolutions in understanding that have marked the sciences. He would have been especially disposed to the accumulation of tools and knowledge in the form of computer programs-- an extension of the "artificial memory" which he began building in his student days as an array of indexed notebooks.


In 1932, Chadwick, working in Cambridge University, discovered the neutron. With the crude observational techniques of the time, he saw energetic protons arising in the vicinity of beryllium bombardment with alpha particles. Looking at all other possibilities, either there was an enormously energetic gamma ray (of the same nature as x rays), or-- almost as unthinkable-- a massive neutral particle. The latter hypothesis was abundantly verified.

In 1933, Leo Szilard, living in London imagined the consequences of neutron bombardment of material. Unlike the previous bombardments to create artificial radioactivity, using protons or the nuclei of heavy hydrogen (deuterons), the neutron is not repelled by the nucleus-- the core of another atom-- because the neutron carries no positive charge. So whatever the neutron energy, it can approach and enter another nucleus, and with considerable probability provoke a nuclear reaction that in many cases might transform that nucleus into another, Szilard thought, with the emission of one or two neutrons.

If only one neutron is emitted, then a single neutron entering under optimum conditions might barely reproduce itself until it left the mass of material. But if considerably more than one neutron is emitted in some of the nuclear reactions, there is the possibility of amassing an adequate amount of material, with relatively few absorbers of neutrons present, such that a single neutron will become two, then four, then eight. In 80 doublings, a neutron chain reaction could consume 50 or 100 grams of material with an energy release that would be larger than that of combustion of the same amount of material by a factor four million, or so. One kilogram of material might thus, within less than a millionth of a second, liberate the energy equivalent of a thousand tons of fuel, or 10,000 tons of high explosive.

If the reaction could be controlled to proceed at a modest and continuous rate, limited by the ability to remove heat, the chain reaction might serve to replace a boiler burning fuel, and so heat a conventional power plant.

Leo Szilard was moved to take a British patent on this concept of the neutron chain reaction (July 4, 1934), and began to explore all of the known elements in a sequential fashion. Unfortunately for him, he started at the low end of the periodic table of elements and ran out of money before he explored very many elements. Had he begun with the heaviest element-- uranium-- the neutron chain reaction might have been realized years earlier than was actually the case.

In a 1969 interview, Chadwick recalled,

"I remember the Spring of 1941 to this day. I realized then that a nuclear bomb was not only possible-- it was inevitable... and there was nobody to talk to about it. I had many sleepless nights. But I did realize how very serious it could be. And I had then to start taking sleeping pills. It was the only remedy. I've never stopped since then. It's 28 years, and I don't think I've missed a single night in all those 28 years."

Fermi had pioneered the field of nuclear transformations caused by neutrons, and he and his team in Rome had discovered the efficacy of "slow neutrons"-- those which had made many collisions with their surroundings and thus moved at one ten-thousandth the speed and ten billionths of the energy of the neutrons when they were born. Fermi regarded this as his most important scientific discovery. But in the irradiation of uranium, Fermi thought that he had discovered transuranic elements-- those heavier than uranium, which, however, appeared to come in a bewildering array.

In November 1938, Fermi received word that he would be awarded the Nobel Prize in Physics December 10, 1938. He and his wife, Laura, and their two children, Nella and Giulio left for Stockholm, with the hidden intention never to return Italy. In the summer of 1938, Mussolini, by then the slavish junior partner to Hitler, launched a campaign of anti-Semitism against the one Italian in 1000 who was a Jew. Laura Fermi's family was Jewish, and the Nobel Prize gave Fermi the opportunity to leave for the safety of New York, where he took up a post at Columbia University in January, 1939. There he quickly learned of the suggestion and the confirmation, rapid-fire, that some of the peculiar results observed in Rome were due not to elements heavier than uranium, but to those much lighter. The slow neutrons had broken the uranium nucleus into "fission fragments" with the release of enormous amounts of energy. Furthermore, it was conjectured and soon established that on the average more than two neutrons were emitted in the course of each fission.

What an opportunity for further research!

But Leo Szilard, who had preceded Fermi to New York, swept from England by the winds of war, joined him at Columbia University in a most peculiar kind of collaboration. Szilard was consumed with the thought of nuclear energy and especially of a nuclear weapon. With the idea of the chain reaction in the air, Szilard suggested (in January 1939) that there be a voluntary imposition of secrecy in publishing, and American physicists seem to have adopted such a code, which was not, however, to be the case with those in Europe-- France especially.

But Hitler had set forth on his campaign of world domination and extermination. He was making great progress in Europe, and those in Britain who were alert to the prospect of nuclear energy, were under bombardment and threat of conquest.

When Szilard urged I.I. Rabi at Columbia University to persuade Fermi of the need for secrecy in these matters. Fermi answered "Nuts!" Although Fermi strictly obeyed the regulations that were to come, in 1939 he was far from convinced that the neutron chain reaction was possible. Nor was he to have a high opinion of the ability of the U.S. military and their technical support structure: at Chicago in 1942, according to Szilard, Fermi remarked, "If we brought the bomb to them already-made on a silver platter, there would still be a 50-50 chance that they would mess it up." (Rhodes, p.423).

In October 1941, with the war in Europe in full swing, Werner Heisenberg, who was in charge of the German uranium project, attended a scientific meeting in occupied Copenhagen, and had a long evening walk with Niels Bohr-- an incident now famous through the play, "Copenhagen", by Michael Frayn. According to Rhodes (p. 34), "Heisenberg remembers asking Bohr if it was right for physicists to work on 'the uranium problem' in wartime when there was a possibility that such work would lead to 'grave consequences in the technique of war.'" Heisenberg reports that Bohr asked whether a bomb really was possible, and "Heisenberg says he answered that a 'terrific technical effort' would be necessary, which he hoped could not be realized in the present war." Exactly what happened in this discussion remains unclear, together with the motivation and understanding of the two men.

By early 1940, Otto Frisch and Rudolf Peierls in England had already come to the conclusion that the fission of U-235 with fast neutrons would produce a nuclear explosive, if the U-235 could be separated from the 140-times more abundant U-238. And the amount of U-235 needed would be on the order of a few kilograms. This information was not made available to the Americans until the summer of 1941, and the concept of nuclear explosion and the essential use of fast neutrons were not nearly so well understood in the United States as was the case in Britain. It is of interest that Chadwick in 1941 felt himself unable to discuss his concerns about nuclear weapons with Frisch or Peierls, in Chadwick's own laboratory: they were not British citizens.

Jumping ahead in the story, after the end of the European war, Heisenberg and his colleagues of the German uranium effort were interned in England, at Farm Hall. Their conversations were monitored and recorded surreptitiously, and transcribed each night-- sometimes in German-- sometimes directly in English-- and published only recently as the Farm Hall Papers.(7) When the German scientists learned of the destruction of Hiroshima, Heisenberg was asked by his colleagues how the Americans could have built the claimed fission weapon-- what had been Heisenberg's estimate of the critical mass? Heisenberg was skeptical that this was indeed a nuclear weapon and replied that the critical mass of U-235 was "about a ton"-- a thousand kilograms. He then explained that for a fast-neutron chain reaction to be possible, a neutron would be required to remain in the ball of fissionable material for the 80 doublings of the population that would be needed, and since the neutron would be undergoing a "random walk", the number of steps in the radial direction that it would diffuse during this time would be the square root of 80 or about 9. So that the linear dimension of the ball would need to be nine times the distance between scatterings, and that would amount to about a ton of U-235.

This was a bizarre error, which would not have been possible had Heisenberg ever previously calculated correctly the critical mass.

About a week later, Heisenberg did provide his colleagues with an analysis which no longer made the same mistake.

It is my personal opinion that it was not a knowing Heisenberg who intentionally slowed or did not pursue an effective uranium program in Germany. It was a leader who had made a simple mistake, never corrected, and hence regarded nuclear fission and the chain reaction as a potential source of energy and not really a source of effective weapons.


This is a fascinating story, superbly told by Richard Rhodes. By 1940, German attacks on Poland in September 1939, and the intentional bombing of civilian populations in cities, and the brutal treatment of Jews and other identifiable minorities, left no doubt that the western allies were in a fight for their very survival. Britain and then the United States would engage in city bombing as well.

In early 1939, the mechanism of fission was finally understood, and the fact that it was U-235 that could undergo fission to a significant extent. Already in Britain (as in Germany) analyses had been made of possible means for separating U-235, but the conclusions were daunting. In the United States, two Columbia University Physicists Eugene T. Booth and John R. Dunning reviewed isotopic separation methods and chose as the most promising the phenomenon of gaseous diffusion. A gaseous molecule containing uranium (UF6) is about 1% lighter if it contains an atom of U-235 rather than U-238. It would diffuse through a tiny pore 0.5% more rapidly, so that several thousand such barriers, with recirculation and recompression of the gas for each stage, would provide a beginning at effective separation. Booth and Dunning began to make barriers by dissolving the zinc from brass, leaving small pores, and much more effective barriers were later used in the production plant at Oak Ridge, Tennessee.

At Columbia University, Harold Urey, who had received the Nobel Prize in Chemistry for the discovery of heavy hydrogen led this uranium enrichment effort. Fermi concentrated on the nuclear reactor, which would produce transuranic elements, on which Fermi's Rome colleague, Emilio Segre was already working in Berkeley, California. By 1942, Fermi was to move to the University of Chicago, to join the "Metallurgical Laboratory" which would concentrate on the production of plutonium, which was expected to be as fissionable as U-235, and which could be obtained without the enormously costly and difficult process of isotopic separation. Simple chemistry would separate this new element if it could be produced in sufficient quantity in a nuclear reactor.

The program to build a nuclear weapon was extremely secret in the United States-- known to only a very few of the military.

Laura Fermi writes that "The head of the Metallurgical Laboratory, Arthur Compton, gave a series of parties to welcome new arrivals to the Met Lab. Compton arranged uniquely to have his wife, Betty, cleared". That is, Betty Compton was officially permitted to know of the secret work being conducted to build a nuclear weapon. But Laura Fermi found out, as did many other wives, about her husband's work only in August, 1945.

On December 2, 1942, the work of Fermi and Szilard and their colleagues bore fruit, with the first self-sustaining nuclear chain reaction conducted under the West Stands of the football field at the University of Chicago. The "pile" of natural uranium and graphite was operated at a power up to one-half watt. Szilard writes, "There was a crowd there and then Fermi and I stayed there alone. I shook hands with Fermi and I said I thought this day would go down as a black day in the history of mankind." One of the Fermi team which built the pile and the bomb is with us today-- Harold Agnew.

The Met Lab took on the design of production reactors at Hanford, in the state of Washington, the first of which operated at a fission rate corresponding to the creation of 200 million watts of heat. Theoretical physicists and engineers worked together. Fermi and Eugene P. Wigner played crucial roles in creating the theory of nuclear reactors and in solving practical design problems.

In the summer of 1942, six months before the success of the West Stands, Robert Oppenheimer headed a theoretical physics summer study on the actual use of U-235 or plutonium to make a nuclear weapon. Ironically, the assembly of U-235 metal from two sub-critical masses into a single supercritical mass (by the use of smokeless powder as in a military gun) seemed so simple that after a short while the leading physicists, including Hans Bethe and Edward Teller, turned their attention to a suggestion that Fermi had made to Teller in September 1941-- that an atomic bomb might be used to heat a mass of heavy hydrogen (deuterium) to a temperature at which it would undergo thermonuclear fission. At the 1942 summer study, the consequences (but not the mechanism) of such a system were well understood, so that the chairman of the NDRC (National Defense Research Council) noted

"If you use two or three tons of liquid deuterium and 30 kg U-235, this would be equivalent to 100 million tons of TNT. Estimate devastation area of 1000 sq km. Radioactivity lethal over same area for a few days."

And the plan was to have 100 kg of deuterium available by the Fall of 1943.

The actual assembly of U-235 or plutonium into nuclear weapons was to take place at a new site at Los Alamos, New Mexico beginning March 1943. By a stroke of genius, the army General in charge of the entire nuclear weapon program, Leslie R. Groves, has selected J. Robert Oppenheimer to head the laboratory. Known for his quickness of wit, his facility in calculation, and his sharp tongue, Oppenheimer proved to be a superb leader who knew every detail and was able to direct the work of the laboratory with a sure hand. Busy with the design and construction of the Hanford plutonium production facilities and with experiments with the new-found neutron source-- an operating pile-- Fermi did not move to Los Alamos until September 1944. At Los Alamos, Fermi was puzzled by the conviction he found there, according to Oppenheimer, "After he had sat in on one of his first conferences here, he turned to me and said 'I believe your people actually want to make a bomb.' I remember his voice sounded surprised." (Rhodes, p. 468). I believe that Fermi gave the program his full effort, and it was enormously valuable. But he was more interested in physics.

In an interesting book, Sylvan Schweber(8) compares the lives of Oppenheimer and Bethe. He emphasizes that Bethe to the present day, at the age of 95, has remained grounded in physics. The revered spirit of physics at Cornell University since 1935, Bethe led the theoretical effort at Los Alamos to produce the atomic bomb and later in similar fashion there to develop the hydrogen bomb.

In contrast, Oppenheimer did no physics at Los Alamos or after, until his death in 1967. But he did play an important official role in nuclear policy until his security clearance was revoked in 1953.

Fermi never abandoned his dedication to physics. Only reluctantly did he accept the presidency of the American Physical Society for a year, but he then took it as a serious responsibility.

By the beginning of 1945 it was clear that the war in Europe would be won without the atomic bomb, which in any case was not yet ready. But hundreds of thousands of additional American deaths were foreseen in bringing the war in the Pacific to a close-- and far more Japanese deaths. The work of the Met Lab in Chicago was essentially done, and Leo Szilard, to whom the program owed so much from his earliest thoughts in 1934 through his essential work in pushing Fermi to build the first reactor, was restless.


Szilard had been excluded from Los Alamos because of General Grove's fear of his political unreliability, and he was frustrated with his inability in this time of peril to speak with those in Washington. Szilard wrote a memorandum for President Franklin Roosevelt and obtained a letter of introduction from Albert Einstein. Szilard took the initiative to see the President's wife, Eleanor Roosevelt, on May 8, 1945. Ever conscious of the aphorism that it is easier to get forgiveness than permission, only then did he tell Arthur Compton what he was doing. Compton unexpectedly cheered him on.

Szilard returned to his office and learned within minutes that President Roosevelt had died. The new President, Harry Truman, diverted Szilard, Harold Urey, and a colleague to South Carolina to talk with a Truman confidant, Jimmy Byrnes. Byrnes read the Szilard letter that "... the greatest immediate danger which faces us is the possibility that our 'demonstration' of atomic bombs will precipitate a race in the production of these devices between the United State and Russia." This would eliminate the invulnerability of the continental United States and Szilard emphasized "These decisions ought to be based not on the present evidence relating to atomic bombs, but rather on the situation which can expected to confront us in this respect a few years from now." (Rhodes, p. 637). But Szilard did not manage to see President Truman. Note that Szilard judged that these dire consequences would arise not only from the use of the bomb in war but even from a demonstration.

Fermi was involved specifically in a committee decision to use the first nuclear weapon. Enrico Fermi, Arthur H. Compton, Ernest O. Lawrence, and J. Robert Oppenheimer explained their choice as follows:

"The opinions of our scientific colleagues on the initial use of these weapons... range from the proposal of a purely technical demonstration to that of the military application best designed to induce surrender. Those who advocate a purely technical demonstration would wish to outlaw the use of atomic weapons and have feared that if we use the weapons now our position in future negotiations will be prejudiced. Others emphasize the opportunity of saving American lives by immediate military use, and believe that such use will improve the international prospects, in that they are more concerned with the prevention of war than the elimination of this special weapon. We find ourselves closer to these latter views; we can propose no technical demonstration likely to bring an end to the war; we can see no alternative to direct military use."(9)

As von Hippel notes "Thus, the relative priority of the importance of abolishing nuclear weapons and using them to deter war (was) being debated before the rest of the world even knew of the existence of the bomb."

In fact, Oppenheimer told his secretary that it took him all night to convince Fermi to go along with the conclusion unanimously expressed by the committee of Fermi, Compton, Lawrence, and Oppenheimer. He wanted to keep the bomb secret and out of circulation as long as possible and was against both a demonstration and military use.(10)

I, myself, believe that the correct decision was made, to use the nuclear weapon on Hiroshima. Whether it was necessary to use it against Nagasaki is another question. I don't believe that the bomb could have been "kept secret", in view of the fact that the Soviet Union already had detailed knowledge from a Soviet spy at Los Alamos, Klaus Fuchs, and others, and was, in any case, perfectly capable of developing the fission weapon on its own.

And I don't believe that a demonstration would have brought an end to the Japan war without the actual use of an atomic bomb. But I could be wrong.


Many of the scientists who built the atomic bomb in the United States were very much concerned that it should not be simply another military weapon, and they spent years from their scientific careers lobbying Congress and informing the public to this effect. Prominent among such was John A. Simpson, University of Chicago, who was later to do important work in cosmic ray physics, and who died just this year. Some scientists believed in a world government, among them, in 1945, Edward Teller. According to Laura Fermi,

"Enrico did not think that in 1945 mankind was ripe for world government. For these reasons he did not join the Association of Los Alamos Scientists."(11)

The Association of Los Alamos Scientists was succeeded by the Federation of Atomic Scientists, itself the forerunner of the present Federation of American Scientists.

Although I had been involved technically with the development and testing of nuclear weapons since 1950, it was only after about 1955 that I had anything to do with nuclear weapons policy. I have long argued that arms control was an important tool of defense policy, and that the national security of the United States was more involved with the survival of the United States than with the destruction of another power. However, one of the key and general tools for preserving a society is to threaten unacceptable damage to the forces or even to the survival of another society.

Evidently, such a posture of "deterrence by assured destruction," when both sides have nuclear weapons, may result in a posture of "mutual assured destruction" characterized as MAD. I believe that if one has two antagonistic democracies, each well armed with nuclear weapons, MAD is in fact both rational and morally justified. According to a formulation of the Golden Rule, by which one would not do unto another what one does not want done to oneself, I fully endorse security obtained by MAD, if it is achievable in no other way.

In fact I argued in the highest councils of government in the 1960s and 1970s, in regard to the U.S.-Soviet confrontation that one really need not worry about the prospect of the Soviet Union being able to destroy U.S. retaliatory capability (by the use of many accurate missiles armed with nuclear warheads), because one could have a reliable system for launching our land-based nuclear weapons while they were under attack. It would not be in the Soviet interest to attempt to destroy them, knowing that the attempt would fail and would result in the immediate destruction of the Soviet Union-- as surely as if their own weapons had been launched directly against their own territory.

I argued also that the United States should not have nuclear weapons so accurate that they could destroy Soviet missile silos. In this way, the United States would not force the Soviet Union to a posture of launch under attack or launch on warning, which the Soviet Union had adopted-- and which Russia maintains to this day-- with increasing hazard both to itself and to the rest of the world.

Of course, in a democracy it is ultimately the responsibility of the citizenry to maintain control for the nuclear weapons, just as it is over any other important aspect of a society. From this point of view, the golden rule would permit and even encourage maintaining international security by means of MAD, if it works. The golden rule would not support a democracy maintaining its own security by an assured destruction capability against a dictatorship, where the people who would bear the brunt of a retaliatory attack have no influence on the government. Under those circumstances, efforts are typically made to ensure that it is the leadership and the people in authority who suffer, but such destruction may extend to the entire society. That is not, however, its purpose. People have long struggled with the moral aspects of this problem. The U.S. Roman Catholic bishops judged that deterrence was provisionally acceptable only in the context of major efforts to reduce and eliminate the nuclear threat. In their words, "'sufficiency' to deter is an adequate strategy; the quest for superiority must be rejected."

But not all threats can be deterred, and this has been recognized for a long time. If Hitler had succeeded in obtaining a nuclear bomb, it is likely that he would have used it at the time of his own suicide. And that is why it is very much in the interests of all countries, including those with nuclear weapons, to ensure that the destructive potential is as little as can be achieved in maintaining an uneasy peace under the present system of national governments.


I know the United States's program far better than that of any other nation. Nuclear weapons, of course, are intended to have the potential for destruction, whenever they are called upon. At the same time, like any other military system, the side which possesses and may use them tries to make them adequately safe, considering the destruction that an accident with a nuclear weapon might wreak upon its own forces and territory. To this end, the early U.S. nuclear weapons maintained the fissile material separate from the explosive system. It was installed, typically, in flight.

Very soon, however, military readiness was achieved by mechanical "in-flight insertion devices" that would ensure that no nuclear explosion could ensue, no matter what the accident, until the weapon's fissile core was inserted.

Later, U.S. nuclear weapons either had mechanical "safing" devices or were "inherently one-point safe" against the detonation of the high explosive at the worst possible single point. The incoming detonation wave in the high explosive normally requires the ignition of explosive lenses or trains at two or more points-- dozens in the case of the early U.S. implosion weapons.

Even more important than the requirement of one-point safety is that the weapons be safe against unintended detonation. This requires that the electrical firing system not be subject to ignition by lightning or electromagnetic interference, and much effort has gone into that program.

Nevertheless, current U.S. nuclear weapons (although all one-point safe) do not have the ultimate in safety configuration. In addition to the Enhanced Nuclear Detonation Safety system (ENDS) and the virtue of one-point safety, it would be desirable that weapons containing a plutonium core not distribute the plutonium in case of a fire. A fire will typically burn the explosive but not detonate it, and plutonium is likely to be vaporized in the fire and distributed over the local area. The hazard is an increased incidence of cancer caused by the ingestion of radioactive plutonium.

It would be preferable, of course, to have the plutonium in a fire-proof casing, and that is done in certain U.S. nuclear weapons. On the other hand, this is no small modification of an existing weapon, and it would require nuclear explosion tests to provide fire-resistant pits for the weapon types that do not have them.

Even with a fire-resistant pit, a one-point detonation, although it would not result in a nuclear yield, would distribute plutonium, since the pit would be destroyed by the explosion.

Given the opportunity in 1991 to conduct over several years 15 additional nuclear tests to maximize the safety U.S. nuclear weapons, the armed services and the Department of Energy decided not to do this.

In my opinion, this is entirely justified, and it would be a tragedy if nuclear testing resumed in the name of making already safe enough weapons, safer still. The number of people (Americans, very likely) who would die from the dissemination of plutonium in such an accident would typically be a few, as estimated from experiments in which plutonium was intentionally liberated, and also the accidents at Palomares, Spain and Thule, Greenland.(12)

Arguments against the Comprehensive Test Ban Treaty that U.S. nuclear weapons do not contain the latest safeguards are a distortion of reality. I judge that U.S. nuclear weapons are adequately safe. The problem of U.S. nuclear weapons is not that they are unsafe; it arises from the possibility of their intended use.


As I have indicated, in the summer study at Berkeley in 1942, much theoretical attention was devoted to the Super-- the hydrogen bomb. In this approach, a fission explosion would be used to heat a portion of a large mass of liquid deuterium, which would then undergo fusion of deuterium nuclei at a sufficient rate to burn much of this deuterium to helium before the system disperses and thus reduces its density so that the fusion reaction is quenched. It was clear from the first, however, that an atomic bomb capable of initiating a fusion reaction, with the understanding of the day, would destroy a city in its own right. When Oppenheimer, the leader of the summer study, became the Director of the Los Alamos Laboratory in March, 1943, he limited the work on the Super to Edward Teller and a couple of other people. Teller chafed at the lack of resources, but there was really a lack of ideas as to how actually to carry out the program, and there was the general point that a fission bomb would in itself be an adequate strategic weapon. In fact the development of the gun-type weapon proceeded in a straight line, awaiting only the U-235 separated by gaseous diffusion and by electromagnetic process at Oak Ridge.

In 1945, after the two fission bombs destroyed Hiroshima and Nagasaki, Teller felt that the new Director of the Los Alamos Laboratory, Norris Bradbury, would now emphasize work on the Super, but he did not. Nor had Robert Oppenheimer, in his last days at Los Alamos, increased the effort as Teller thought it should be enlarged.

On September 23, 1949, the Soviet Union detonated its first nuclear bomb, which gave the advocates of the Super a telling argument in favor of proceeding. By that time, nuclear weapons were the responsibility of the Atomic Energy Commission, which produced them for the military. The AEC had a high-level General Advisory Committee (GAC) which Oppenheimer chaired, and of which Rabi and Fermi (but not Teller) were members. The GAC considered the question of a commitment to build the hydrogen bomb, and recommended against it. In a strongly worded minority report, Fermi and Rabi went further, writing,

"The fact that no limits exist to the destructiveness of this weapon makes its very existence and the knowledge of its construction a danger to humanity as a whole. It is necessarily an evil thing considered in any light"

Nevertheless, President Truman made the decision to build the Super, and Los Alamos had the major responsibility for this.

Hans Bethe, who had lobbied against building the hydrogen bomb now agreed to work with this program, once it had been committed. In fact, he was to head the theoretical work on the Super program just as he had headed the theoretical work at Los Alamos on the fission bomb during World War II. But neither Hans Bethe nor anyone else knew how to build a Super.

In 1950, during my first summer at Los Alamos, I shared an office with Enrico Fermi. Part of the time, he was involved in computations together with the mathematician, Stanislaw Ulam, on the classical Super. They were investigating with so-called hand calculations, the burning of a long cylinder of deuterium, ignited in some fashion by nuclear heat at one end. The temperatures involved, of course, are greater than those at the center of some stars, since the star can burn for billions of years, and the Super has only a fraction of a microsecond before the high pressure of the materials causes it to disassemble.

Fermi would use an accountant's spreadsheet, filling in the top row with various initial conditions and carrying out the calculation by using a sliderule and a desk-top mechanical calculator (a Marchant, as I recall) to advance each time step. After seeing that the calculation progressed properly, Fermi and Ulam would call in their computer (a young woman named Miriam Caldwell) who would work overnight on the problem, bringing it back in the morning. Fermi and Ulam would graph the results, decide to change some input condition or assumed cross-section or other parameter, and go on to other things, while Mrs. Caldwell worked on the next day's calculation. The results were discouraging. And the reason was clear. The parameter space, if any, in which the system would work was very small.

Too small a diameter of the deuterium cylinder, and energy would escape rather than being deposited as a result of fusion. Too large a diameter, and the thermal radiation itself would not escape, so that the electromagnetic radiation bath would absorb most of the fusion heat, without allowing the temperature to rise far enough for the fusion reaction to work properly.

And other groups were not having better results.

The situation changed entirely just before March 1, 1951, when Stanislaw Ulam suggested to Edward Teller that instead of using a cylinder of deuterium at normal liquid density, use be made of an atomic bomb for compressing the deuterium manyfold. Presumably it would then be heated and would propagate.

According to his informal "testament" of 1979(13) Teller replied that it would do no good to compress the deuterium, because he had a theorem that said that if it wouldn't work at normal density, it wouldn't work at one thousand time normal density, since all interactions were "bi-molecular" or two-body. It would be too much of a distraction to explain the error, but it is enough to record that in a joint paper of March 9, 1951, Teller and Ulam described two approaches to such a use of an auxiliary bomb, clear from the title, "Hydrodynamic Lenses and Radiation Mirrors". By May, 1951, when I arrived again in Los Alamos for my second summer, the entire Laboratory was convinced that this new approach had merit, and I was able in a week or so in July of that year to put together the ideas that were current and to make certain decisions, as to how a nuclear explosive could be built using liquid deuterium and the Teller-Ulam "radiation implosion" approach. The Los Alamos Laboratory joined forces, to the extent that the system was built and successfully tested in the Pacific November 1, 1952. It exploded with an energy release of ten megatons-- some 500 or more times that of the weapons which had destroyed Hiroshima or Nagasaki.

In his 1989 book, "The Advisors: Oppenheimer, Teller, and the Super Bomb" Herbert York argues that the United States should have tried to obtain an agreement with the Soviet Union not to develop hydrogen bombs. Even if the Soviet Union did not abide by such an agreement, he suggests, the United States could rely on deterrence with its fission bombs until it could catch up. In principle, this is probably so, but in the politics of the era of Senator Joseph McCarthy, it would not have been feasible and might have destroyed the scientific community and impaired world security.


There were at one time about 70,000 nuclear weapons in the world. There are probably 30,000 still in existence-- 12,000 for the U.S. and 18,000 for Russia. The lesser official nuclear powers-- Britain, China, and France have a few hundred each, and Israel, India, and Pakistan add little to the total. It would have been inconceivable in 1945 to imagine such numbers of nuclear weapons-thousands of them with explosive yield 30 times that of the first fission bombs.

That these weapons have not been used in war since 1945 and none has created a nuclear explosion by accident is the result both of hard work and a lot of good luck. The luck will not last forever, and much of the hard work is no longer being conducted.

Although the Bush Administration committed itself during the campaign and through many comments since taking office that it would engage in massive reductions in nuclear weaponry, no such plan has been announced, or is any sign imminent. Rather, it may be that such reductions will be held hostage to approval of the building of a defense system against ballistic missiles-- which is evident from events since September 11 should not be the highest priority for U.S. national security.

Since 1996 the major nuclear powers and now almost all the nations of world have signed a Comprehensive Nuclear Test Ban Treaty-- with the notable exceptions of Israel, India, Pakistan, and a few others.

From the point of view of most nations, this reflects a view that their own acquisition of nuclear weapons would do far less for their security than it would be impaired by the acquisition of nuclear weapons by their neighbors. For the United States, and especially for me, this reflects the judgment that further advances in U.S. nuclear weaponry, for which nuclear testing will be required, would contribute little to our security compared with the acquisition of far more primitive nuclear weapons by other states. The Comprehensive Test Ban Treaty is only one tool in reducing the danger of nuclear weapons to the world and to our nations. Others include truly massive reductions in nuclear weaponry-- in the United States to 1000 or less, and ultimately perhaps to the level of a few hundred for the entire world-- with the possibility eventually of elimination of nuclear weaponry.

Essential also would be extending whatever security benefits nuclear weapons provide to those states which have renounced the possession of nuclear weaponry.

Since its founding in 1980, I have been a member of the Committee on International Security and Arms Control of the U.S. National Academy of Sciences. As its name indicates, the Committee does much of its work in the exploration of cooperative means, scientifically informed, for the reduction of hazard posed by weapons-- especially nuclear, but also biological. Since a meeting convened by CISAC in Washington in 1985, the Accademia dei Lincei and other academies have joined with CISAC in the annual Almadi conferences-- named after Edoardo Amaldi, who was the prime mover in accepting that academies did have a role in informing their national governments in affairs of international security.

In addition to the stalled reductions in U.S. and Russian nuclear weapons, there is especially since the stocks stopped growing and the demise of the Soviet Union, the hazard posed by excess nuclear weapon materials-- particularly in these two countries, and especially in Russia.

In particular, by the very active dismantlement of nuclear weapons in support of existing arms control agreements between the two states, more than 100 tons of weapon plutonium will be made excess and much is already excess over the next few years. Since U.S. nuclear weapons were initially built with six kilograms of plutonium and now on the average use considerably less, 100 tons of plutonium correspond to about 25,000 nuclear weapons, each of which could considerably exceed the explosive power of the bomb that destroyed Nagasaki, killing about 100,000 people.

With the diffusion of knowledge and technology since 1945, it is far easier for a group or a state to build a nuclear weapon, if they have access to the plutonium or to highly enriched uranium.

This weapon uranium is also present in great surplus. In fact, the United States has purchased and received from Russia more than 100 tons of excess weapon uranium, blended down with natural or low-enrichment uranium to form fuel for nuclear reactors. This low-enriched uranium is not directly usable in nuclear weapons, without difficult isotopic separation, done at present only in a few large installations around the world. With a nominal 60^kg of HEU required to make the weapon that destroyed Hiroshima, 1000 tons of excess HEU corresponds to 16,000 nuclear weapons.

A uranium weapon is far simpler to make than is a plutonium weapon-- in part because uranium does not pose a radiological hazard.

The scientific research is long done on this problem of the hazard of excess weapon uranium and plutonium. It is incredible that the scientists of the world and their governments-- particularly in the G7-- do not join to provide resources and leadership to enable Russia to blend down in advance all of the excess high enriched uranium, to prevent its theft or direct sale posing the nuclear weapon threat to the world.

And although CISAC and other institutions have had discussions and similar governmental and independent groups even reached agreement as to what needs to be done with excess weapon plutonium (for which there is no economic benefit as there is with the burning of excess weapon uranium in nuclear reactors), what little progress has been made is being reversed. Russia, where the peril exists because of the degraded environment for securing nuclear materials, persists in the misguided notion that plutonium has an economic benefit when burned as fuel in a nuclear reactor-- despite the fact that the fabrication of reactor fuel free plutonium metal costs more than the mining, purification, enrichment, and fuel fabrication of fresh uranium. The world ignores this problem at its peril-- not only of accident but of blackmail.


When I was on the President's Science Advisory Committee in 1970, it was clear that the worldwide battle under the World Health Organization to eradicate smallpox would soon be won. The United States was considering a formal decision to eliminate smallpox vaccination, and thus the protection of a population, because the disease would no longer occur. I argued strongly against this.

I made the analogy to a nuclear reactor, which, never having been started, as the fuel is loaded has so-called control rods and safety rods inserted. After all of the fuel has been installed, as was laboriously the case with the graphite-uranium reactor built by Fermi at the football field at the University of Chicago, the control rods are gradually removed, and a nuclear reaction carefully started. Because of the 0.7% of the fission neutrons which are delayed by some tens of seconds (and which play no role in a nuclear weapon) a reactor is very easy to control. The Fermi reactor operated at a power level up to 2^watts-- that of the bulb in your flashlight.

At its operating power of 200 megawatts, the Hanford, Washington, reactor experienced six billion fissions per microsecond. Each microsecond, more than ten billion neutrons would be born and die. With the control rods inserted, so that the boron or cadmium they contained would absorb a number of neutrons comparable of that of the uranium, there would be essentially no fissions or neutrons present.

To return to smallpox, the elimination of vaccination in the absence of the disease was analogous to having a reactor already built and removing the control rods. After all, there were no neutrons or fissions so why have the control rods-- they could be sold for a small amount of money, and had no function.

Yet with the control rods out, a single cosmic ray neutron would initiate the chain reaction, which in an ordinary reactor would simply lead to heat and perhaps to overheat, but in an assembly of uranium metal would lead to massive nuclear explosion.

So it is with smallpox, with the difference that smallpox probably would take a long time to reemerge naturally as a mutation. But the society would be vulnerable to the intentional or accidental introduction of the disease.

The argument in favor of ceasing vaccination was that for each million people vaccinated, a few would die from side effects. For myself and my own family I willingly accept this risk, in comparison with what I thought was a substantial probability that smallpox would reemerge either as a weapon of war or as a terrorist plague. And I believe that the old, by current standards, unsafe, vaccine was adequate and adequately safe.

Note that my views in this regard are the same as my views with the possession of U.S. nuclear weapons that do not, every one of them, have the ultimate in safety-- for instance, lacking fire-resistant pits. Naturally, if three million Americans per year would be vaccinated, and ten would die, it would be worth spending a few million dollars a year on the development of a safer vaccine. But it makes no sense to imperil the health and survival of the majority by concern for this tiny, tiny number of people who would be injured by the process. Certainly, vaccination should have been retained as an option, even if not mandatory.


Ethical problems arise in many cases not out of scientific research, but out of advocacy by scientists. Here is a clear case. In 1991, Edward Teller published an article advocating the development of mini-nukes and micro-nukes.(14) He argued that we needed to develop these weapons in order to see whether they were feasible, and that after they were tested and we learned how to build them, and we had established their effectiveness, then we could decide whether to manufacture some numbers of them and plan for their use.

This was, however, after the 1991 Gulf War, which saw substantial use of laser-guided bombs. In fact, 25,000 such weapons had been used in the years 1969-1974, in Vietnam, and had proved their effectiveness in striking designated targets. So it really made no sense to have a 50-pound or 500-pound bomb with a power of 5 or 20 tons of high explosives, to compensate missing the armored vehicle, when, in fact, it was perfectly feasible to strike the armored vehicle, and a high-explosive weapon was substantially cheaper and would not pose problems of escalation that a nuclear weapon would. So strictly from a U.S. military point of view, the conventional munition was better.(15)

From the point of view of one who wished to explore technology to the limit before deciding upon its use, what we know how to do is rarely adequate. But in this case that approach would reduce U.S. military effectiveness by spending more money, by impeding the acquisition of weapons which worked perfectly well, and, especially, by breaking the barrier between conventional and nuclear weapon use. The peril such weapons pose to society far exceeds any benefit they may provide on the battlefield. Why?

Any nuclear weapon needs to assemble a "critical mass" of fissionable material-- plutonium or U-235. For a "gun" type nuclear weapon, this is on the order of 60 kg of U-235 (plutonium cannot be used in gun-type weapons). The weapon which destroyed Hiroshima had a yield of 13 kilotons; what is important is not its yield, but that it killed almost 100,000 people. And that is the problem-- a nuclear weapon procured in sufficient numbers to kill many armored vehicles one at a time would be effective in destroying that many small cities. It would be a tremendous force for destruction in the world, and the fissionable materials could be used far more effectively for that purpose rather than for an anti-tank design yield of perhaps 10 tons.

Things are not much better with plutonium, although the amount of plutonium that is required for a nuclear weapon is considerably less than the amount of U-235, first because the nuclear properties of plutonium are in that respect better (instead of a bare-sphere critical mass almost 60 kg, the plutonium bare-sphere critical mass is about 10 kg.

But the critical mass is much reduced by compression. If the force of high explosives can be so arranged as to double the density of plutonium over that of the metal, only 25% as much plutonium is needed to make a critical mass. Still, thousands of plutonium weapons capable of destroying armored vehicles one at a time would destroy substantial portions of that many cities. It is a bad bargain to choose a nuclear weapon to destroy an armored vehicle, when the opponent or some totally independent country in some totally independent conflict might use a similar weapon to destroy a city.

It cannot be said that the deterrent effect on the opposing military of threatening to destroy their armored vehicles one at a time with nuclear weapons is any greater than the deterrent effect of a force-in-being capable of doing that with laser-guided bombs or other advanced homing weapon-- the ability of which is demonstrated everyday in practice.


In 1969, I served on a panel of the President's Science Advisory Committee under Richard Nixon, which led to his Executive Order banning work on biological weapons in the United States. This soon led to the 1972 Biological Weapons Convention, now almost universally signed and ratified, by which states undertake not to use and not even to possess biological warfare agents except in amounts necessary for vaccination, research, and protection.

Unlike the Geneva Convention of 1923, the BWC makes no provision for the possession or use of BW agents for retaliation in case of BW attack.

But it is clear that several states are not honoring their commitment to the BWC. Russia admits that it flagrantly violated its undertakings, and as the Soviet Union and even as Russia maintained a major program in BW, producing and militarizing many tons of anthrax and other biological agents.

Several initiative are required here. First, all nations should make clear in the United Nations and other fora that it is unacceptable for any nation of the world, having adhered to the BWC, not to carry out its obligations in good faith. To this end, individual work on BW should be criminalized, as are seven other crimes including attacks on diplomatic personnel, or hijacking; and any individual ordering or participating activities in contravention of the BWC should be tried in any competent jurisdiction.

Beyond that, there is the prospect for the emergence not only of new diseases, but very specific and powerful directed means for turning the marvelous advances in biology to the detriment of humanity.

Such potentials are evident in the research on toxins (poisonous material manufactured by biological systems), in order that they be carried to the site of cancer, to provide better tools for fighting that dread disease.

Matthew S. Meselson of Harvard University has been especially prominent in urging the criminalization of work in violation of a country's adherence to the BWC. John D. Steinbruner, of the University of Maryland, has taken the initiative in recognizing the potential threat caused by research in modern biology. In no way does he want to slow the pace of this work, but he does recognize the potential peril. Steinbruner proposes to combine a campaign for the eradication of many diseases (in the same way that smallpox was eradicated) with enhanced and distributed supervision of the work that goes on in the field. This is a complex undertaking, just at its beginning, but there is an important potential hazard from scientific and industrial research, and a new moral and ethical imperative for scientists to understand this and to create norms of responsible behavior.


Granting that the world's energy supply must shift to one that emits less than the current emission of 8 gigatons of carbon to the atmosphere each year (in order to avoid unacceptable global warming from the enhanced greenhouse effect of the carbon dioxide in the air), energy, and nuclear power are the two leading contenders for the replacement of coal, gas, and oil. I thoroughly support research and investments in solar energy, but it is more expensive than nuclear reactors, where nuclear power is accepted. In fact, combustion of fossil fuel with sequestering of carbon in exhausted oil fields or in the deep ocean is another possibility.

In a book published this month with Georges Charpak, we look at the history, technology, and future of nuclear weapons and of nuclear power.(16) I have already discussed nuclear weaponry in this talk, and now want to give my judgment on nuclear power.

The 300+ large nuclear power plants in the world produce almost 20% of the world's consumption of electrical energy. Annually, each plant consumes enriched uranium from about 200 tons of natural uranium. In 50 years, existing plants would consume all three million tons of the stock of assured terrestrial resources. If 100% of the world's electricity were provided in this way, and if the world made and used twice as much electrical power in order to raise the standard of living in the less developed countries, all this uranium would be consumed in a ridiculous five years.

Such uranium fuel "scarcity" was recognized from the very beginning, with emphasis on breeder reactors, for which a given uranium supply would last almost 100 times longer. But there is a lot more uranium in the rocks of the world, available perhaps at ten times the current price of $30 per kilogram. In our book we emphasize the four billion tons of uranium present in seawater-- more than a thousand times the assured terrestrial resource.

For conventional reactors, half of the ocean's uranium would thus last 2000 years if an expanded electrical sector would rely totally on nuclear power. And with breeder reactors ultimately brought into service when economical, the ocean's uranium would last 200,000 years. I hope that humanity will live productively on this earth for far more than 2000 or even 200,000 years, but even a few hundred years should be enough for us to master solar energy and even render nuclear fusion an economical source of power.

As for seawater uranium, the most advanced Japanese work shows that 20 kilograms of plastic felted cloth left to soak for a month in the warm ocean currents off Japan acquires in this way 3 grams of uranium. We do not yet know whether this will lead to uranium at $100 per kg or $1000 per kg. Either would be quite affordable for use in current reactors, and would result in fuel of negligible cost for a population of breeder reactors.

While it is not important for industrial firms now to know whether this uneconomic uranium is available (in the presence of uranium that can be mined at an average cost of $30 per kg), it is of the utmost importance to the future of the world (and hence to present governments) to do the research necessary to understand now what this cost is.

In the absence of this understanding, plutonium is being recycled in France into light-water reactors, at a cost which is equivalent to that of raw uranium at $700 per kg-- probably more than what would be required to obtain uranium from seawater. In any case, such recycling is economically disadvantageous, and if continued ought to be done with a full understanding of its economic costs offsetting whatever benefits it may offer. In fact, using the plutonium in this way to effect a 20% saving in uranium denies the future use of this plutonium (later to be extracted from the nuclear fuel) in the fast thorium cycle.

We need not and should not introduce breeder reactors now-- not until they are cheaper than the ordinary reactors using low-enrichment uranium, and after thorough investigation leads to a reactor of adequate safety.

Conventional reactors of existing type could perfectly well be built-- preferably underground-- and new type of reactors with improved safety should be built if, as has been claimed by their proponents, they are cheaper than the light-water reactors of the existing type.

Described favorably in our book, are the high-temperature gas-turbine reactors under development by General Atomics Corporation and by ESKOM in South Africa. These two programs use graphite and uranium-- the first with large prismatic fuel elements; the second with tennis-ball-size "pebbles". In both cases, the actual fuel is in the form of pellets less than a millimeter in diameter, of compounds of uranium and carbon, which constitute tiny pressure vessels to contain the fission products resulting the production of nuclear heat.

Another reactor type is that on which much work has been done by Carlo Rubbia and his colleagues-- a thorium core constituting a fast reactor cooled by a vast well of lead. This is, almost, a fast breeder reactor, which could perfectly well be operated without the particle accelerator that is required to make up the neutron deficiency in this reaction. Instead, the neutron deficit would be supplied by the consumption of excess weapon plutonium or excess plutonium waste from the existing population of light-water reactors. Only later would the particle accelerator need to be installed in new-construction models of such "energy amplifiers".


I began this talk with the consideration of some ethical questions in scientific research typified by the much discussed decision to build and use the atomic bomb, to build the hydrogen bomb, and the control of threats to humanity.

But I moved from there to some topics over which we still have some control, and for which the imperative is actually to take action-- to do the analyses in support of vaccination against diseases that may have been eradicated; to move aggressively in the community of nations, spearheaded by scientists, to ensure that the Biological Weapons Convention is truly honored.

Then I discussed the effort by John Steinbruner to control the fruits of research in modern biology, which offers so much opportunity for the control and even eradication of any diseases.

My final example was that of nuclear power, which can perfectly well supply as much nuclear power as the world can consume, if one taps uranium from seawater and ultimately, reactors.


I believe that time is running out. Our societies have already wasted much of the past 10 or 20 years, when we should have been concentrating on blunting the potential of bioterrorism by restoring smallpox vaccination-- even if there is not a single person in the world ill with smallpox. Having signed and ratified the Biological Weapons Convention, our states have been remiss in not insisting that all adherents strictly respect their own commitment not to possess biological warfare agents or the means for their dissemination-whether these actions are totally verifiable or not; this step would be aided by but need not await the universal criminalization of individual acts contrary to the biological warfare convention.

And we must take seriously the hazard of nuclear terrorism-- all nations-- with urgent attention to diluting weapon-usable uranium to below 20% U-235 content, where it is no longer directly usable in a fission weapon.

We should have taken these steps on the basis of prudence and analysis, without 3000(17) people being killed in a single hour of terrorist strikes September 11, 2001. If we wait longer, we risk hundreds of thousands dying in a single anthrax attack (millions in case of smallpox) or the use of a single crude nuclear weapon. We should take these measures now, without awaiting an initiative of the government of the United States.

  1. Laura Fermi, "Atoms in the Family," The University of Chicago Press, 1954.
  2. Emilio Segre, "Enrico Fermi, Physicist," The University of Chicago Press, 1970.
  3. Francesco Calogero, "Responsibility of Scientists and Hopes for Future Peace in the World," Accademia Nazionale dei Lincei (1993).
  4. Marvin L. Goldberger, "Enrico Fermi (1901-1954): The Complete Physicist," Physics in Perspective, 1 (1999), pp. 328-336.
  5. Frank N. von Hippel, "Where Fermi Stood," Bulletin of the Atomic Scientists, 57, September/October, 2001, pp. 26-29.
  6. Richard Rhodes, "The Making of the Atomic Bomb," Simon & Schuster, 1988.
  7. Jeremy Bernstein, "Hitler's Uranium Club: the secret recordings at Farm Hall," American Institute of Physics, 1996, pp. 139ff.
  8. Sylvan S. Schweber, "In the Shadow of the Bomb: Oppenheimer, Bethe, and the Moral Responsibility of the Scientist," Princeton University Press, 2000.
  9. Alice Kimball Smith, "A Peril and a Hope: The Scientists' Movement in America, 1945-1947" (MIT Press) 1965, p. 50.
  10. Peter Wyden, "Day One: Before Hiroshima and After," Simon & Schuster, 1984, p. 171.
  11. Fermi, "Atoms in the Family," p. 246.
  12. R.L. Garwin and G. Charpak, "Megawatts and Megatons: A Turning Point in the Nuclear Age?" Alfred A. Knopf, NY (October 2001), pp.^340-342. Also published as "Megawatts and Megatons: The Future of Nuclear Weapons and Nuclear Power," University of Chicago Press (January, 2003), same pagination.
  13. Edward Teller, transcript of taped conversation between Edward Teller and Jay Keyworth concerning an important question in the wartime history of Los Alamos-- "How the idea of the implosion emerged," September 20, 1979.
  14. E. Teller, "Guest Comment: Military applications of technology,-,A new turn," American Journal of Physics, Vol. 59, No. 10, p. 873, Oct. 1991,
  15. R.L. Garwin, "Guest Comment: Science, technology, and national security in an era of democracy and human rights," American Journal of Physics, Vol. 60, No. 5, pp. 395-396 (May, 1992).
  16. R.L. Garwin and G. Charpak, op. cit.
  17. (Estimated in October 2001 as 6000 but later revised).