One day in mid-September, 1942, about a month and a half before the invasion of North Africa, I was offered an extremely attractive assignment overseas. At that time I had been on duty in Washington for over two years as Deputy Chief of Construction of the Army Corps of Engineers. Under the general supervision of my immediate superior, Major General T. M. Robins, I was in charge of all Army construction within the United States as well as in our off-shore bases. This included the building of camps, airfields, ordnance and chemical manufacturing plants, depots, port facilities and the like. But though the responsibility was great and the work essential, I was, like every other regular officer, extremely eager for service abroad as a commander of combat troops; and I now replied with deep pleasure that any c aty in an active theater of operations appealed to me. However, I added, I would have to secure the consent of Lieutenant General Brehon Somervell, Commanding General of the Army Services of Supply, before I could definitely say yes. I promised to see him and to give my answer by noon the next day.
The following morning, a few minutes after I had finished testifying before a Congressional committee on a military housing bill, I met Somervell outside the hearing room, and asked him whether he had any objection to my being relieved from my Army construction duties. To my great surprise, he told me that I could not leave Washington. He went on to say: “The Secretary of War has selected you for a very important assignment, and the President has approved the selection.”
“Where?”
“Washington.”
“I don’t want to stay in Washington.”
“If you do the job right,” Somervell said, “it will win the war.”
My spirits fell as I realized what he had in mind. “Oh, that thing,” I said. Somervell went on, “You can do it if it can be done. See Styer1 and he will give you the details.”
My initial reaction was one of extreme disappointment. I did not know the details of America’s atomic development program at that time, but, because of the nature of my responsibilities, as I shall explain, I knew of its existence and its general purpose—through the use of uranium to produce an atomic bomb which it was hoped might be of unprecedented power. Though a big project, it was not expected to involve as much as $100 million altogether. While this was more than the cost of almost any single job under my jurisdiction, it was much less than our total over-all spending in a normal week.2 Magnitude aside, what little I knew of the project had not particularly impressed me, and if I had known the complete picture I would have been still less impressed.
Later that morning, I saw Styer at his office in the Pentagon. He confirmed my worst premonitions by telling me that I was to be placed in charge of the Army’s part of the atomic effort. He outlined my mission, painting a very rosy picture for me: “The basic research and development are done. You just have to take the rough designs, put them into final shape, build some plants and organize an operating force and your job will be finished and the war will be over.” Naturally I was skeptical, but it took me several weeks to realize just how overoptimistic an outlook he had presented.
In the course of our discussion, we agreed that, because the Pentagon was so nearly finished and because I had had so much to do with it, I would continue to control its construction, despite my new assignment. There were two reasons for this. First, my sudden disappearance from the work on the Pentagon would attract much more notice than would my absence from my other Army construction activities. Second, because of the natural interest in the Pentagon displayed by a number of Congressmen, it would be better for me to continue to carry the responsibility for that job than to pass it on to someone else who was unfamiliar with its past problems and their many political ramifications. To avoid complications I would arrange this informally.
We also agreed upon the wording of the order, to be signed by Somervell, directing me to take complete charge of the atomic project.3
Before I left, Styer told me that General Marshall had directed that I be made a brigadier general, and that the list of new promotions would be out in a few days. I decided at once, and Styer agreed, that I should not take over the project officially until I could do so as a brigadier. I thought that there might be some problems in dealing with the many academic scientists involved in the project, and I felt that my position would be stronger if they thought of me from the first as a general instead of as a promoted colonel. My later experiences convinced me that this was a wise move; strangely enough, it often seemed to me that the prerogatives of rank were more important in the academic world than they are among soldiers.
At the time I was brought into the picture, research on the uses of atomic energy had been going on at a gradually accelerating pace since January, 1939, when Lise Meitner explained that the uranium atom could be split. The discovery opened up two divergent paths for its exploitation. Most physical scientists realized that nuclear energy, derived from the splitting, or fissioning, of the atom, might be used either to generate power for peaceful purposes or to create super-weapons. In general, however, it was the scientists who were personally acquainted with Hitler’s New Order who first became most interested in the possible military uses of atomic energy and its effect on the existing balance of political power. Most American-born physicists were not originally so concerned with this aspect, because they had not yet become accustomed to thinking of new scientific truths in terms of their military applications. And from the beginning, of course, some of the physicists, both European and American, who considered the potential of atomic weapons were horrified at the thought of the devastation they might spread throughout the world.
With the growth of Hitlerism, many scientists gradually came to doubt the wisdom of communicating the results of their work to scientists in the enemy’s camp. Again, this was particularly true of those who had come to America to escape Nazi persecution. They had great respect for the capabilities of German scientists and were familiar with their efforts in the field of atomic energy. They were also aware of the pressures that certainly would be brought to bear on the German scientists to ensure their utmost support of their country’s military program. The American-born scientists, in the main, did not have so acute an appreciation of the danger of the situation as did their foreign-born colleagues.
Nevertheless, the Americans and British made a concerted attempt to achieve voluntary international censorship of information relating to atomic energy. The effectiveness of this arrangement was hindered at the start by the refusal of Joliot-Curie to participate; however, it provided a sound foundation for the strict governmental measures that followed the outbreak of war in Europe.
At the same time, the group of refugee scientists in America became a focal point of the attempts to apprise officials in the federal government of the dangers and the prospects that atomic physics held for the United States. Discussions of developments in this field took place between representatives of the Navy Department and Dr. George B. Pegram and Dr. Enrico Fermi, of Columbia, as early as March, 1939. However, Fermi expressed some skepticism at this meeting and the United States Government did not become seriously interested until October of that year when Alexander Sachs, a Wall Street economist and a personal friend and adviser of President Roosevelt, called upon him to obtain his support of the scientific research then under way. Sachs had been following the possibilities of atomic energy for some time and felt that the government should actively aid in the development. To achieve this he discussed the matter with the Columbia group and with Einstein. The latter agreed to sign a suitable letter to the President if Sachs would prepare it. This he did, had it signed and carried it to the White House. This letter emphasized the seriousness of the problem. The President, impressed by Sachs’ arguments, appointed an Advisory Committee on Uranium to report to him on the situation.
The Uranium Committee, as it was called, consisted of representatives of the Bureau of Standards, the Army and the Navy. It met from time to time with various physicists and chemists to discuss the development of both atomic power and atomic weapons. On the basis of these discussions, the committee recommended that the Army and Navy make available a modest sum for the purchase of research materials. The work of the committee gained momentum when it was learned, in April of 1940, that the Kaiser Wilhelm Institute in Berlin had undertaken an extensive research program involving uranium.
When the National Defense Research Committee (NDRC) was created in June of 1940, under the chairmanship of Dr. Vannevar Bush, the Uranium Committee was made one of its subcommittees and embarked upon an ambitious research program. The work was carried out through contracts with universities and private and public institutions, using at first funds transferred from the Army and Navy, and later, NDRC-appropriated monies. November, 1941, saw sixteen projects, totaling about $300,000, under way.
During the previous spring and summer, the entire nuclear program had been carefully reviewed from both the scientific and the engineering points of view, with emphasis on the military feasibility of the project. From this review, after noting the optimism of British scientists engaged in the same field, Bush concluded that the United States’ efforts to develop military uses of atomic energy should be expanded. He discussed the problem with the President and got his approval to enlarge the program, improve its organization, obtain special funds and interchange information with the British. Mr. Roosevelt created at that time a Top Policy Group, consisting of himself, Vice President Wallace, Secretary of War Stimson, the Chief of Staff, General Marshall, Dr. Bush and Dr. James B. Conant.
In November of 1941, Bush decided that the uranium project was growing to be of such importance that it should be outside of NDRC control. Accordingly, it was placed directly under the Office of Scientific Research and Development (OSRD), of which NDRC was a part. Bush, by then the head of OSRD, set up at the same time a planning board to initiate the engineering of pilot plants and full-scale production facilities.
Virtually all laboratory research until this time had been aimed at achieving a controlled chain reaction, using U-235, a rare isotope of uranium which comprises less than one percent of the metal in its natural state. This isotope has the property of fissioning readily—a property which the far more abundant form of uranium, U-238, does not display. But it soon became apparent that unless unprecedented quantities of this material could be produced in a much purer state, a U-235 chain reaction would be impossible. The basic problem was to arrive at an industrial process that would produce kilograms of a substance that had never been isolated before in greater than sub-microscopic quantities. The processes then being considered were all designed to take advantage of the very minor physical difference between U-235 and U-238.
The way for a major breakthrough was open as a result of studies that suggested the theoretical feasibility of transmuting U-238 into a highly fissionable new element, plutonium, which might then be separated from the parent uranium by chemical means. The hope was that this would be easier to do than to isolate or concentrate the rare U-235 by physical means. The group headed by Dr. Glenn T. Seaborg at the University of California undertook to prepare extremely small amounts of plutonium, and in March of 1941 succeeded in creating the first submicroscopic amounts of Plutonium-239. Later that month the California group confirmed the theory that under neutron bombardment plutonium atoms fissioned as readily as atoms of U-235.
The desirability of investigating more carefully the military possibilities of plutonium was emphasized by Dr. Ernest O. Lawrence, of the University of California, and the following December, after further studies had been made, the Uranium Committee of the OSRD seconded his proposal. This decision was supported by calculations of the amounts of plutonium required, estimates of its military effectiveness, and roughly drawn timetables of how long it would take to produce it, assuming that the proposed production process was feasible. Later that same month, an intensive program of research on plutonium, supported and sponsored by the OSRD, was begun at the Metallurgical Laboratory4 at the University of Chicago, under the direction of Nobel Prize winner Dr. Arthur H. Compton, who had achieved eminence by his discovery of the Compton Effect and in pioneer studies of cosmic rays. The purpose of the research was to develop the knowledge needed to design, build and operate a plant for the conversion of uranium into plutonium.5
In the meantime, other laboratories in colleges, universities and a few industrial plants were trying to find some method of physically separating U-235 from U-238 that would be practical from the standpoints of both economy and time. The entry of the United States into World War II caused the abandonment of all projects aimed at developing atomic energy as a source of power and gave added impetus to the efforts to build an atomic bomb. At the same time, Bush and a number of others in policy-making positions began to realize that vital as continued laboratory investigations were, even more pressing problems were developing in the fields of engineering and construction. Although they had created a planning board to cope with problems of this nature, it was fast becoming apparent that a much more powerful organization would be required. It is to their everlasting credit that Bush and his colleagues had the discernment to recognize the limitations of their own organization as well as the moral fortitude to admit them in the national interest. Very few men, confronted with a similar situation, would have done so.
Consequently, when the Top Policy Group met on December 16, 1941, Bush recommended that the Army Corps of Engineers carry out the construction work, and asked that a competent Army officer become thoroughly familiar with the project. General Marshall assigned this task to Major General W. D. Styer.
The following March, in reporting favorably to the President on the plutonium project, Bush recommended again that the Corps of Engineers handle the construction that it would entail. Shortly afterward, Dr. Conant reviewed the entire nuclear program. In his report to Bush, he expressed the prevailing opinion that there were five basic production methods, each of which held out equal chances of success. U-235 could be separated by means of the centrifuge, diffusion and electromagnetic processes; while plutonium could be obtained from either the uranium-graphite pile or the uranium-heavy-water pile. All these processes appeared to be nearly ready for pilot plant construction and possibly for the preliminary design of production plants.
On June 17, 1942, Bush forwarded a detailed report on the status of the program to the President. In it he pointed out that it was possible to make a nuclear weapon that could be employed decisively in combat. He went on to discuss means of producing it and ended by saying that, under ideal conditions, it might be made ready in time to influence the outcome of the present war. President Roosevelt approved Bush’s report.
The next day, Colonel James C. Marshall was directed by Styer to form a new engineer district to carry out the Army’s responsibilities in the development of atomic energy. Marshall was told that he could count on the full support of the War Department and that his needs would be considered paramount to those of any other program.
The coming of autumn in 1942 brought a general reorientation of the project. No longer would it be conducted in the laboratory on a purely theoretical basis, for our scientists had now accumulated sufficient theoretical knowledge to permit the preliminary engineering of possible production processes. Of course, nobody knew whether any of the methods they were considering would work, but, assuming that at least one of them would, they could now make some guess as to the major items that would be needed.
The basic American military requirements were twofold: to provide our armed forces with a weapon that would end the war and to do it before our enemies could use it against us. To fulfill these needs we would have to move ahead with the utmost speed.
When the Corps of Engineers started its work, its job was simply to build and operate the production plants. The problems involved in the development of the bomb and its delivery were for the time being largely ignored.
Nor was the full magnitude of the project generally appreciated. No one thought of it as entailing expenditures running into the billions of dollars. Not until later would it be recognized that chances would have to be taken that in more normal times would be considered reckless in the extreme. Not until later would it become accepted practice to proceed vigorously on major phases of the work despite large gaps in basic knowledge. Not until later would every other consideration, whether the advancement of knowledge for the sake of knowledge or the maintenance of friendly diplomatic relations with other powers, be subordinated to achieving the project’s single overriding aim. Not until later would all concerned grow accustomed to the idea that, while normally haste makes waste, in this case haste was essential.
For the time being, work on the project continued in an orderly, straightforward, thoroughly conventional manner.
On June 18, immediately after receiving orders to set up his special district, Colonel Marshall had informed General Eugene Reybold, the Chief of Engineers; General Robins, the Deputy Chief of Engineers and Chief of Construction; and me, the Deputy Chief of Construction, of his new assignment. He did not reveal many of the details of his work, but he did ask that we give him every possible assistance. When he first told us of his carte blanche instructions, Robins and I were rather skeptical about how long it would be before his priority would be overridden by some other super-important project. We had seen this happen before. Yet, in spite of our mental reservations, we spared no effort in supporting him fully.
Marshall selected Lieutenant Colonel K. D. Nichols as his deputy and together they began to organize the district. After visiting Dr. Bush to learn the general background of the project, Marshall went over with me his urgent need for competent assistance in solving the many engineering problems that confronted him. He said that his work would be of considerable magnitude and would involve extremely complex problems of design. After some discussion, I told him that Stone and Webster appeared to be the engineering firm best fitted to meet his requirements: they were accustomed to working with scientific people—far more than most engineering firms; they were a large firm, capable in both engineering and construction; and they were performing well on all their contracts with the Corps of Engineers. Marshall agreed and promptly began to lay the groundwork for bringing Stone and Webster into the project to take over a major portion of the engineering studies and the preparation of plans, as well as to carry out much of the construction that would follow.
On June 25, Marshall, Nichols and Styer met with the Uranium Committee6 of the OSRD. In the course of this meeting, it became clear that neither the centrifuge nor the diffusion methods for the separation of Uranium-235 were ready for development and that it might be some time before any plants to utilize these processes could be designed. They estimated that the total power requirements of the plants to be constructed in the immediate future would be 108,000 kw and, based on this estimate, they planned for a firm power demand for 150,000 kw by the end of 1943.7
Many of the OSRD’s contracts with the laboratories were being delayed for lack of money, so Colonel Marshall arranged for the transfer of $15 million of Corps of Engineers funds to the OSRD, in order that existing research projects could continue uninterrupted after June 30, the end of the fiscal year. His immediate money worries taken care of, Marshall’s most pressing difficulties involved the procurement of scarce materials and equipment required by the laboratories. Shortages were growing more critical daily and, unless relief were obtained promptly, all research would soon come to a standstill.
Then the question of a name for the project came up. Toward the end of June, Reybold called Robins, Marshall, Nichols and me to his office and told us that he had conferred with Somervell and Styer, and that they had agreed that the new establishment—that is, the various plants where the bomb’s ingredients would be produced—was to be called the “Laboratory for the Development of Substitute Materials” or “DSM.” I demurred on the grounds of security, feeling that the name was bound to arouse curiosity. Several other names were suggested, but Reybold decided that, since Somervell had already agreed to it, no change would be made. Before we parted, it was decided definitely to hire Stone and Webster as the over-all architect-engineer-managers, and to use the Corps of Engineers to acquire whatever land might be needed by the project.
On June 29, officials of Stone and Webster met with Robins, Marshall, Nichols and me. Robins and I questioned them in great detail to determine whether they could place a force of adequate quality and sufficient strength on the project without impairing their performance on a number of very important Corps of Engineers and other war contracts which they had already undertaken. (I wanted to do all I could to help Colonel Marshall, but at the same time I did not want his project to impede any of the construction work for which I was responsible.) Their answers were satisfactory in every respect, and they were hired on the spot.
Earlier, at Marshall’s request, I had initiated within the Office of the Chief of Engineers a survey of the United States to assist in the selection of a suitable site for his production plants. The plan then was to concentrate all facilities required for the production of fissionable materials at a single site. We assumed that a very large tract would be required. Consequently, we sought out a relatively undeveloped area where land was still available at a reasonable price. It had to be located well away from both coasts, so that there would be no threat of direct enemy interference. We knew that there would be a heavy requirement for electric power. Travel between the site and the Washington, New York and Chicago areas had to be reasonably convenient and not too time-consuming. There must be an assured and quite considerable supply of water. As was the case with the acreage and the power, we could only guess at the total water requirement, but we thought that it would be large. It was essential that the site should have a climate that would permit heavy construction to be carried on throughout the year. And, finally, it was most desirable that an adequate labor force, both for construction and for operation, be available.
On the basis of these criteria, I soon concluded that the best location would be somewhere near Knoxville, Tennessee. That area seemed to meet all the requirements, and beyond that, it gave promise of being a pleasant place for the permanent operating force to live. This was most fortunate because we anticipated that many of the operating personnel would have to be imported, since the specialized skills we would need would be difficult to recruit locally. Marshall agreed with this view and, on July 1, he and Nichols, together with their assistants and representatives of Stone and Webster and the Tennessee Valley Authority, began investigating sites in the vicinity of Knoxville. Colonel T. B. Parker, the Chief Engineer of TVA, informed them at that time that the Authority could supply the power needs of the project, provided it obtained priority assistance in procuring heavy generating equipment.
A few days later, meetings were held at the Metallurgical Laboratory to give the responsible Stone and Webster people a thorough understanding of the plutonium process. Compton was concerned over a possible shortage of uranium oxide, so Stone and Webster was directed to purchase 350 tons of this material immediately. At the same time, efforts were made to select a site for a pilot reactor plant, for it had been decided that a full-scale production plant should not be built until an experimental reactor had been previously placed into operation and this had been followed by a semi-works.
The site chosen for the pilot reactor was quite close to Chicago, in the Argonne Forest, which was under the control of the Forest Preserve of Cook County. The Forest Commission, headed by Mr. Clayton F. Smith, was most co-operative, and great pains were taken to insure that the plant would be designed in such a manner that it would not interfere with any of the Commission’s plans for the future. Under the terms of the lease for the land, both the site and any buildings erected were to be returned to the Forest Preserve after the war.
On July 9, Marshall, Nichols and the Stone and Webster people met with the S-1 Committee of the OSRD. Among other things, it was agreed that Stone and Webster would subcontract the design and fabrication of the heavy-water plant to be built at Trail, British Columbia, to E. B. Badger and Sons of Boston, and the erection and operation of it to the Consolidated Mining and Smelting Company of Canada, on whose property it would be situated.8
Out of this same meeting a tentative construction program emerged. It called for starting the construction of the plutonium reactor piles by October 1, 1942; of the centrifuge process by January 1, 1943; of the gaseous diffusion process by March 1, 1943; and of the electromagnetic separation process by November 1, 1942. It soon became apparent that these target dates were wholly unrealistic, for basic research had not yet progressed to the point where work on even the most general design criteria could be started. I was disturbed by the vagueness of the starting dates for the several elements of the project, and urged Marshall to insist upon a detailed schedule, to be agreed to by all concerned, which he could then apply himself to meeting.
Marshall was still having financial problems. He estimated that he would need $85 million, yet every attempt he had made to obtain it had been turned down. We agreed that if he continued to meet refusals he should, acting through Styer, ask Bush to appeal directly to the President for help.
On July 13, the project received an AA-3 priority from the War Production Board, with the promise of AAA for special items if any difficulties were encountered. This priority was completely inadequate for the expeditious completion of a venture as complicated as this one, and serious priority troubles soon arose. I began to feel that the doubts that Robins and I had originally entertained about the support the project would receive were being confirmed.
Toward the end of the month the estimates started to come in on the most likely Tennessee site. The 83,000-acre tract which was under consideration would cost slightly over $4 million, and would require the relocation of some four hundred families. Faced with this prospect, Marshall felt that he should delay any site acquisition proceedings, pending further developments in research on the plutonium process. It must be remembered that at this time the pile in Chicago, which was to prove that a chain reaction was possible, was not yet in operation—indeed, it did not have its first test until December.
Marshall devoted considerable attention during this period to supply contracts, among the most important of which were one with Mallinckrodt for the purification of uranium oxide, and one with Metal Hydrides and one with Westinghouse for converting uranium oxide into the uranium metal needed for the plutonium process. At the same time, aided by Nichols, he had been striving to perfect his organization and to define the responsibilities and interrelationships of its several elements, to recruit competent personnel, find suitable office space, and in general to take care of myriad other necessary details.
Marshall brought in, largely on a temporary basis, a number of regular officers to get things under way while he was recruiting the key nonscientific personnel. Most of the latter were men who had been civilian employees of the Corps of Engineers for years; they were familiar with government operations and their records and abilities were well known. These men were given reserve commissions in the Army, and formed a trained and competent nucleus that proved to be of tremendous value throughout the entire project.
On August 11, Marshall handed me the draft of a general order to be published that day announcing the formation of the new district. In it he used the designation “DSM.” I again objected to this term because I felt that it would arouse the curiosity of all who heard it. After some discussion, during which we considered the possibility of using “Knoxville,” we decided upon “Manhattan,” since Marshall’s main office would at first be in New York City. Our choice was approved by Reybold, and so the Manhattan Engineer District, or MED, came into being.
During this month improvements to existing facilities in the laboratories at Berkeley, where research on the electromagnetic separation process was being carried on, and at Chicago were initiated. The design of the heavy-water plant at Trail got under way, but because this involved a great deal of difficult copper work, it immediately ran into conflicts with the synthetic rubber program, which carried a higher priority.
The situation during the summer of 1942 was not promising for any operation that did not hold the very highest priority. Stone and Webster estimated at the time that the construction of a pilot plant for the electromagnetic separation process would take eleven months under an AA-3 priority, the highest then available to the MED; yet the same job could be completed in eight months if an AA-1 rating were obtained.
Toward the latter part of August, considerable doubt began to crop up concerning the Tennessee site. For the most part, this indecision stemmed from the desires of the leaders of several laboratories to have the production plant situated conveniently close to their own facilities.
Toward the middle of September, Marshall met once more with the S-1 Committee in the Bohemian Grove, just north of San Francisco. A number of decisions were made at that time, the most important of which was to proceed with the Tennessee site, in spite of the contrary opinions. Almost equally important was the decision to design and to procure materials for a small electromagnetic separation plant, at an estimated cost of some $30 million, subject to cancellation at any time prior to January 1, 1943.
On September 16, I agreed to an AAA rating for the copper needed at Trail. The metal thus made available to the project would have to be taken from the Corps of Engineers’ quota and would upset a number of other projects of the Corps, but I was anxious to emphasize to Marshall that he had our full support and to encourage his people to push their work to the utmost.
Little did I anticipate then that within twenty-four hours I would be the one who needed support.
1 Major General W. D. Styer, Chief of Staff, Army Services of Supply.
2 The Army Corps of Engineers, with almost one million people engaged, was then completing about $600 million worth of work each month.
3 See Appendix I, page 417.
4 Metallurgical Laboratory was a code name chosen to conceal the nature of the work being done there.
5 See Appendix II, page 418.
6 By then reconstituted and called the S-1 Committee.
7 By 1945 our power requirements totaled almost 1,000,000 kw.
8 This program was brought into being to create a reserve of heavy water for use as a moderator in the nuclear reactors in case, for some unpredictable reason, the use of graphite proved to be impossible. The problem was that no one knew how much heavy water such a plant would be able to produce, or how much we would need—if we needed it at all. The requirements depended upon the final choice of a production method for plutonium, and this had not been settled at the time the decision about heavy water had to be made. If it were needed, however, we wanted to be ready, and so decided to provide facilities, in addition to the small installation at Trail, that would yield enough for our initial requirements. Fortunately, graphite worked and the heavy water was not needed. Our supply of it was then loaned to Canada for use in an experimental reactor.
The expansion of the heavy-water program was handled by du Pont’s Ammonia Department, which at a cost of some twenty million dollars built three plan’s as adjuncts to existing Army Ordnance plants.