The Collected Works of Leo Szilard Scientific Papers Bernard T. Feld and Gertrud Weiss Szilard, Editors with Kathleen R. Winsor Foreword by Jacques Monod Introductory Essays Carl Eckart Bernard T. Feld Maurice Goldhaber Aaron Novick Julius Tabin The MIT Press London, England and Cambridge, Massachusetts Volume I All new material except that in the public domain Copyright © 1972 by The Massachusetts Institute of Technology This book was set in “Monophoto” Times New Roman, Printed on Warren’s 1854 Medium Offset by Halliday Lithograph Corp., and bound by Halliday Lithograph Corp. in the United States of America. All rights reserved. No part of this book may be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from the publisher. ISBN 0 262 06039 6 (hardcover) Library of Congress catalog card number: 79-151153 Curriculum Vitae by Leo Szilard* (Expanded by Gertrud Weiss Szilard) I was born in Budapest, Hungary, in 1898.1 went through officers’ school there during the First World War and studied engineering there. In 1920 I left Hungary to continue my engineering studies in Berlin. However, the attraction of physics proved to be too great. Einstein, Planck, Von Laue, Schroedinger, Nernst, Haber, and Franck were at that time all assembled in Berlin and attended a journal club in physics which was also open to students. I switched to physics and obtained a Doctor’s degree in physics at the University of Berlin under Von Laue in 1922. My thesis (Ref. 1, see attached list of publications) showed that the Second Law of Thermodynamics covers not only the mean values, as was up to then believed, but also determines the general form of the law that governs the fluctuations of the values. Subsequently, I was a research worker in one of the Kaiser Wilhelm Institutes in Berlin and later joined the teaching staff of the University of Berlin (as Privatdozent) where I remained until 1933. Of the papers (1 to 4) published during this period, some are experimental, and some are theoretical. The last one (4) established the connection between entropy and information which forms part of present day information theory. In 1933 I went to England. I considered at that time becoming a biologist, and A. V. Hill said that he would find a position for me as a demonstrator in physiology. It occurred to me, however, just then that a nuclear chain reaction might be possible if we could find an element that would emit neutrons when bombarded by neutrons. Artificial radioactivity was discovered a few months later by Joliot and seemed to provide an important new research tool in nuclear physics. This decided me to move into nuclear physics. In the summer of 1934 I started work as a guest in St. Bartholomew’s Hospital in London and this work resulted in the establishment of the Szilard-Chalmers Reaction (5) and the discovery that slow neutrons are emitted by beryllium if the beryllium is exposed to gamma rays of radium (6). In 1939, after the discovery of the fission of uranium, the use of these slow neutrons from beryllium made it possible to see that uranium emits neutrons when bombarded by neutrons; the fast neutrons emitted by uranium could be easily distinguished from the bombarding slow neutrons. In 1935, after a visit to New York, where I spent a few months as research associate at New York University, I accepted a position at the Clarendon Laboratory, Oxford University. During this period I worked in the field of nuclear physics (8 to 11). In 1938 I came to America under arrangement with Oxford University, which permitted me to spend half my time in the United States. I was in the United States during the time the Munich Agreement was negotiated. After Munich I decided to stay in the United States on a full-time basis, and I resigned at Oxford. * Written by Leo Szilard in a grant application to the General Medical Sciences Division of the National Institutes of Health, Bethesda, Maryland, June 1959. 4 Part I: Biographical Notes In January 1939 I learned of the discovery of fission. It seemed important to find out at once if neutrons are emitted in that process, for in that case a chain reaction in uranium had to be regarded as a serious possibility. I, therefore, asked the permis­ sion of Columbia University to work there as a guest and perform an experiment in order to settle this question. This experiment (jointly performed with Walter Zinn) led to the discovery of the neutron emission of uranium, upon which the chain reaction is based (12,13). The same discovery was made independently at about the same time by Fermi and his co-workers and by Joliot and his group. In July, 1939,1 recognized that a chain reaction might be set up in a system com­ posed of graphite and uranium. Because of the serious consequences of this possibility, it seemed that this was a matter in which the government ought to take an interest. I, therefore, went to see Professor Einstein to enlist his help in approaching the government. After several consultations, in which E. P. Wigner and Edward Teller participated, Einstein wrote a letter to President Roosevelt; and in response to this letter, the President appointed a committee under the chairmanship of the Director of the National Bureau of Standards. In February 19401 described the chain-reacting uranium-graphite system in a paper I sent to the Physical Review (February 1940). For reasons of secrecy, this paper was not published. In November of 1940 a government contract was given to Columbia University for the development of the graphite-uranium system, and I became a member of Columbia University’s National Defense Research Staff. Early in 1942 our group was moved to the University of Chicago; and on December 2, 1942, the chain reaction system was put into action. Recently a patent was granted to the Atomic Energy Commission on the chain­ reacting graphite-uranium system, jointly in the names of Enrico Fermi and myself. In 1943 I became a naturalized citizen of the United States. In October, 1946, I joined the staff of the University of Chicago as Professor of Biophysics in the Institute of Radiobiology and Biophysics. This institute never grew as originally intended; it had a succession of directors, and it was recently dissolved. I remained on the staff of the University of Chicago as Professor of Biophysics and was transferred to the Enrico Fermi Institute for Nuclear Studies. When in 19461 was faced with the task of converting myself into a biologist, I teamed up with Dr. Aaron Novick, a physical chemist. I had known him from his work in the uranium project. We both got our training in biology through summer courses, such as Dr. Delbriick’s course in Cold Spring Harbor in bacterial viruses and Dr. Van Niel’s course in bacterial biochemistry at Pacific Grove. Dr. Novick and I worked as a team until the Institute of Radiobiology and Biophysics was dissolved. A list of publications is attached, containing a short description of each paper. When we started out, we tried to understand a striking phenomenon just then dis­ covered by A. Kelner, who showed that bacteria killed by ultraviolet light can be reactivated by shining visible light at them (17). A detailed analysis of the phenomenon enabled us to interpret it in terms of a “poison” that is produced by ultraviolet light and is decomposed by visible light. This interpretation was at first controversial due Part 1: Biographical Notes 5 to Dulbecco’s work on light reactivation of ultraviolet killed bacterial viruses but has in the meantime become widely accepted. My own interest in the subject waned when I could not convince myself that we were dealing with a phenomenon that serves a useful biological purpose in the life of the bacteria. Next, we turned our attention to the study of bacterial viruses in the assumption that viruses may prove to be much simpler than bacteria. We obtained some very inter­ esting results (18) but decided to shift after a while to the study of bacteria. The two phenomena in which we were particularly interested were (a) mutations and (b) the formation of adaptive enzymes which promised to provide a tool for the study of protein synthesis. We were dissatisfied, however, with the methods that were available for the study of these phenomena. It seemed to us necessary to study bacterial populations in the growing condition in a stationary state, i.e., we thought we ought to use a continuous flow device. We developed such a device, which we called a “Chemostat.” In this particular device the rate of growth of the bacteria can be changed by changing the concentration of one of the growth factors of our choosing which we make the con­ trolling growth factor. We started out by using the “Chemostat” for the study of mutations and obtained quite unexpected results at the very outset. It turned out, for instance, that the rate at which certain mutations occur does not change when we change the rate at which the bacteria divide; we could vary the rate of growth within a wide range without changing the rate at which these mutations occurred. We found one family of com­ pounds—purines—which may cause an about tenfold increase in the mutation rate of bacteria without any appreciable killing. And we also found antimutagens, which in very small concentrations will fully counteract the effect of purine-type mutagens. In a bacterial population maintained in the “Chemostat” there occur evolutionary changes (19), and one strain of bacteria is replaced by a mutant strain, which can grow faster in the conditions prevailing in the growth tube of the “Chemostat.” We observed successive evolutionary steps of this sort in each experiment of sufficiently long dura­ tion and were able to analyze the phenomenon. After the dissolution of the Institute of Radiobiology and Biophysics I did not maintain a laboratory. In the last few years my interests centered mainly on quanti­ tative studies of general biological phenomena, with strong emphasis on molecular biology. The paper I published most recently (25) attempts to give a quantitative theory of the process of aging which should be applicable to mammals. Partial Bibliography of Leo Szilard [with his own annotations] A. Physics (1) Zeitschriftfur Physik, 32, p. 753,1925. This paper extends the application of ther­ modynamics to the derivation of the laws of thermodynamical fluctuations. It was accepted as dissertation by the University of Berlin. 6 Part I: Biographical Notes (2) Zeitschrift fur Physik, 33, p. 688, 1925—jointly with H. Mark. This paper reports experiments which revealed anomalous scattering of x rays. (3) Zeitschrift fur Physik, 35, p. 743, 1926—jointly with H. Mark. This paper reports experiments on polarizing x rays by reflection on crystals. (4) Zeitschrift fur Physik, 53, p. 840,1929. This paper evaluates the increase of entropy which is connected with operations of an intelligent being on a thermodynamical sys­ tem if these operations are controlled by measurement of variables which are subject to thermodynamical fluctuations. This paper was accepted as Habilitationsschrift by the University of Berlin. (5) “Chemical Separation of the Radioactive Element from its Bombarded Isotope in the Fermi Effect”—jointly with Chalmers. Nature, 134, p. 462, 1934. This paper demonstrates a generally applicable process (Szilard-Chalmers reaction) for the con­ centration of a radioactive element produced by neutrons if the element has to be sep­ arated from a mass of a stable element with which it is chemically isotopic. (6) “Detecting Neutrons Liberated from Beryllium by Gamma Rays,” Nature, 134, p. 494,1934. This paper describes the discovery of radium-beryllium photo neutrons which, being of low energy, represent a useful tool in nuclear research. They were universally used later in the discovery and investigation of neutron emission of uran­ ium on which a chain reaction is based. (7) “Liberation of Neutrons from Beryllium by X-Rays”—jointly with a group of six others. Nature, 134, p. 880, 1934. Using x rays in place of gamma rays, the threshold for the emission of photo neutrons from beryllium is determined by varying the voltage of an x-ray tube and is found to be somewhat above 1.5, and well below 2 Mev. (8) “Radioactivity Induced by Neutrons”—jointly with Chalmers. Nature, 135, p. 98, 1935. In this paper a neutron-induced radioactive period of about 3^ hours is reported in indium which does not fit in with the explanations found for other radioactive periods. In a later paper it is shown that it is due to an excited indium nucleus which is isomeric with stable indium nucleus 115. (9) “Absorption of Residual Neutrons,” Nature, 136, p. 950,1935. This paper reports the discovery of neutron resonances at low energies, gives an estimate of their energies, and states that the energies can be measured by observing the absorption of the resid­ ual neutrons in boron or lithium. (10) “Gamma Rays Excited by Capture of Neutrons”—jointly with Griffiths. Nature, 139, p. 323, 1937. This paper reports on the observation of gamma rays emitted by a Part I: Biographical Notes 7 number of odd elements which are strong neutron absorbers. The counts observed per absorbed neutron were found to be within 15 % identical for all these elements. (11) “Radioactivity Induced by Nuclear Excitation”—jointly with Goldhaber and Hill. Phys. Rev., 55, p. 47,1939. In this paper the previously reported period in indium is investigated and the conclusion is reached that it is due to nuclear excitation of the stable indium isotope 115. (12) “Instantaneous Emission of Fast Neutrons in the Interaction of Slow Neutrons with Uranium”—jointly with Zinn. Phys. Rev., 55, p. 799,1939. In this paper the dis­ covery of the neutron emission of uranium is reported. It is estimated that two neutrons are emitted per fission. The neutrons from uranium are made visible on an oscil­ lograph screen. As primary neutrons, radium-beryllium photo neutrons were used which, because they are slow, can be easily distinguished from the fast neutrons emitted by uranium. This discovery which was made independently by Fermi in the same year indicated the feasibility of a sustaining nuclear chain reaction. (13) “Emission of Neutrons by Uranium”—jointly with Zinn. Phys. Rev., 56, p. 619, 1939. Detailed report of above mentioned experiments, number of neutrons per fission measured as 2.3. (14) “Neutron Production and Absorption in Uranium”—jointly with Anderson and Fermi. Phys. Rev., 56, p. 284, 1939. This paper reports an investigation on the chain reaction qualities of a uranium-water system. It is estimated that 1.5 neutrons are emitted for every thermal neutron which is absorbed by uranium* B. Biology (17) A. Novick and Leo Szilard, “Experiments on Light-Reactivation of Ultra-Violet Inactivated Bacteria,” Proceedings of the National Academy of Sciences, 35, No. 10, pp. 591-600, 1949. (18) Aaron Novick and Leo Szilard, “Virus Strains of Identical Phenotype but Dif­ ferent Genotype,” Science, 113, No. 2924, pp. 34-35, January 12, 1951. (19) Aaron Novick and Leo Szilard, “Experiments with the Chemostat on Spontan­ eous Mutations of Bacteria,” Proceedings of the National Academy of Sciences, 36, No. 12, pp. 708-719, December 1950. * [References 15 and 16, left blank in this physics listing of Szilard’s, presumably would include Report A-55, see Part III this volume. Dr. Szilard’s part in bringing about of the first nuclear chain reaction, in the design of the first nuclear reactor (atomic pile) were first described, insofar as these matters could then be nade public, in Henry D. Smythe, Official Report: Atomic Energy for Military Purposes, Princeton Univer­ sity Press; 1945, pages 34, 47, etc.—Eds.] 8 Part I: Biographical Notes (20) Aaron Novick and Leo Szilard, “Description of the Chemostat,” Science, 112, No. 2920, pp. 715-716, December 15, 1950. (21) Aaron Novick and Leo Szilard, “Experiments on Spontaneous and Chemically Induced Mutations of Bacteria Growing in the Chemostat,” Cold Spring Harbor Symposia on Quantitative Biology, Vol. XVI, 1951. (22) Aaron Novick and Leo Szilard, “Anti-Mutagens,” Nature, 170, p. 926, November 29, 1952. (23) Aaron Novick and Leo Szilard, “Experiments with the Chemostat on the Rates of Amino Acid Synthesis in Bacteria,” Dynamics of Growth■ Processes, Princeton University Press, pp. 21-32, 1954. (24) Maurice S. Fox and Leo Szilard, “A Device for Growing Bacterial Populations under Steady State Conditions,” Journal of General Physiology, 39, pp. 261-266,1955. (25) Leo Szilard, “On the Nature of the Aging Process,” Proceedings of the National Academy of Sciences, 45, pp. 30-45, 1959. The first of these papers (17) investigates a phenomenon discovered by A. Kelner after the war, who showed that bacteria “killed” by ultraviolet light can be revived by shining visible light on them. Experiments designed to analyze the phenomenon are described in this paper; they lead to the conclusion that the ultraviolet light produces a “poison” which can be inactivated by light and that this “poison,” if present when, subsequent to irradiation, the bacteria divide, will cause both death and mutations. The second paper (18) describes the discovery that, when a bacterium is infected simultaneously with two related viruses which differ from each other both in genotype and phenotype, the virus population emerging from the bacterium contains a class of viruses which have the genotype of one and the phenotype of the other. The Papers 19 to 23 describe a new way of studying bacteria by maintaining a bacterial population in a stationary (exponentially growing) state indefinitely and controlling the growth rate by controlling the rate of supply of an essential growth factor. An apparatus is described in these papers which will conveniently accomplish this and which is designated as the “Chemostat.” In studying mutations in bacteria or the formation of adaptive enzymes in bacteria, inaccurate and, therefore, misleading results are frequently obtained by studying bacterial cultures in flasks in which the number of bacteria increases exponentially, and today the use of the Chemostat appears to be indispensable. In the Papers 19 to 22, the Chemostat is used in the study of mutations. It turns out that the rate at which mutations occur in a growing bacterial population under the conditions studied is not proportional to the rate at which cell division occurs; rather- Part I: Biographical Notes 9 the mutation rate is constant per unit time independent of the rate at which the culture is growing. There is found one group of compounds, all purine derivatives, of which caffeine is one, which greatly increases the mutation rate without having an appreci­ able killing effect on the bacteria. There is another group of compounds described in these papers, all of them ribo­ sides of purines which in small quantities will completely counteract the action of the above-mentioned purine type mutagens and also reduce the rate of spontaneous mutations. In Paper 23, the Chemostat is used to study the biosynthesis of amino acids in bacteria and the regulatory mechanisms that are involved in it. The bio-synthetic apparatus of the bacteria responds to amino acid concentrations in the medium, which are exceedingly low. For instance, a bacterium that can make arginine and will do so if there is no arginine in the medium, will stop making arginine if an arginine concentration of 10~9 gm/cc is maintained in the medium in the Chemostat. (Novick and Szilard, unpublished paper.) One way of studying such regulatory mechanisms is based on the use of a mutant which is blocked in the synthesis of an amino acid—in our case tryptophane—and which pours out into the medium a “precursor” of that amino acid. Paper 23 utilizes such a mutant. In the absence of tryptophane in the medium, a precursor of tryp­ tophane is poured out by the mutant into the medium at a rate which is independent of the growth rate of the bacteria. In the presence of tryptophane this “precursor” is not poured out by the bacteria. It is conceivable that this indicates a general phenomenon of regulation through a negative feedback of the final product at one of the early steps of the metabolic pathway leading to tryptophane. In Paper 24, there is described a device called a breeder. In this device bacteria may be grown in a continuous flow of nutrient. The flow of the nutrient is controlled by the turbidity of the bacterial culture and the growth is not limited by a growth factor, as is the case in the Chemostat. This device was developed in order to study mutations in bacteria under conditions of growth at the maximal rate, and such a study was carried out by Maurice S. Fox. Paper 25 develops a theory of the basic process of aging. According to the theory, the elementary step in the process of aging consists in the random inactivation of whole chromosomes. The differences of longevity of individuals are attributed to the difference of the number of defective “vegetative genes” they have inherited. The preceding pages were written by Szilard as part of an “Application for a Re­ search Grant” which he filed with the General Medical Sciences Division of the National Institutes of Health in June 1959. His Curriculum Vitae written for this purpose therefore contains only information pertinent to his scientific career and omits other activities of great importance in his life. In 1933, after Szilard left Germany for England, he established jointly with Sir William Beveridge, Esther Simpson, and others the “Academic Assistance Council,” an organization to relocate scholars displaced from Nazi Germany. While Szilard was instrumental in launching the atomic bomb project under the assumption—later proved to be erroneous—that there was a race against Nazi 10 Part I: Biographical Notes Germany, it became clear to him in the Spring of 1945 that Germany had been defeated, and he made desperate efforts to prevent the bomb from being used. There­ after, a major part of his life was devoted to the prevention of nuclear warfare. In July 1945 he circulated a petition in the Uranium Project, addressed to the President of the United States, asking that the bomb not be dropped. Within the year following the war he fought successfully for civilian control of atomic energy. In the late 1940s, through the Emergency Committee of Atomic Scientists and numerous publications in the Bulletin of the Atomic Scientists and elsewhere, he launched a perpetual campaign to make people aware of the dangers of nuclear war­ fare and the arms race. He was a dedicated participant of the international Pugwash Movement, attending all their conferences during his lifetime. In 1962 he founded the Council for Abolishing War, now called the Council for a Livable World, an organization aimed at influencing the decisions and actions of the U.S. Senate in foreign policy issues. His writings on these topics will be included in succeeding volumes. In the Grant Application, mentioned earlier, the title of the project was “Quan­ titative Studies of General Biological Phenomena” (RG 6876). Because the proposed research plan is unusual indeed, parts of it are excerpted here: Research Plan and Supporting Data The purpose of the proposed study is to gain insight into certain general biological phenomena rather than to try to understand the functioning of specialized biological structures (such as, for instance, of the nerve fiber, of the muscle fiber, or of the special­ ized sense organs). I am particularly interested in those general biological phenomena where it may now be possible to gain insight into quantitative relationships which can be checked against data obtained from available observations or experiments as yet to be made. The proposed work would take as its starting point preliminary theoretical studies which I carried out in the past three years. At the University of Chicago, I am holding a Research Professorship. I have neither any teaching duties nor any fixed obligations to be in Chicago at certain fixed periods of time. This freedom has enabled me in the past three years to spend considerable time at various laboratories away from Chicago. It is my understanding that, under the grant here requested, I would have full freedom to move about wherever my research interests may take me. It is anticipated that I may spend nine months of the year away from my home, at various laboratories where work may be pursued in fields in which I am interested. I anticipate that my work will have a strong theoretical orientation. But in order to be able to function as a theoretical biologist, it is necessary to have intimate knowl­ edge of experiments relating to a variety of biological materials and involving diverse techniques. Biology has not quite reached the stage which was attained by physics half a century ago when enough relevant facts were established to permit a theoretical physicist to come up with significant insights. Yet in biology we might be very well on the verge of a similar situation, and a few scientists who are so inclined may now perhaps attempt to function as theoretical biologists. Accordingly, these days it might be as