Extensive travel in Europe and many contacts with the Jewish refugee problem had convinced me, over the 1930s, that something evil was taking place in Germany. In spite of all my theoretical persuasion that war is no way to solve the world's problems, I became even more convinced that the United States would have to become involved. So dominant and compelling was this judgment that I concluded, in the spring of 1940, that I must try to find some way of being personally useful. This was some eighteen months before the attack on Pearl Harbor; but I knew that a few national leaders in science, and notably Vannevar Bush, had already been thinking and planning about ways in which the scientific resources of the country could be brought to bear upon the problems of warfare.
The National Academy of Sciences was created in 1863, during the Civil War, to make available to the government the best scientific advice available. By the time of World War I, the Academy had quite naturally taken on a broad range of responsibilities not at all closely associated with warfare, so it was then necessary for President Wilson to request the President of the National Academy to set up a new affiliated organization, the National Research Council, to stimulate research and to aid in the application of all science to the problems of war. The NRC was the principal agency that brought science to bear on World War I.
History then repeated itself. By 1940 the National Research Council had become so deeply engaged in the programs it had developed over the preceding twenty-year period that clearly a new organization, flexible and uncommitted, must again be set up. Vannevar Bush was then chairman of the National Advisory Committee for Aeronautics, a group that had shown what could be done in a field strategic to national defense, and he was also president of the Carnegie Institution of Washington, an organization with varied scientific interests. He was a highly respected scientist-engineer, with wide acquaintance. He had energy, imagination, and courage. It was natural and, as it turned out, exceedingly fortunate that President Roosevelt asked the Council of National Defense to enlist Bush as the head of a National Defense Research Committee (a year later to be called the OSRD—the Office of Scientific Research and Development—this latter being a more accurate title, as it had become clear that the organization would be involved not only in research but also in the practical stages of construction and testing of pilot models).
As soon as I heard the news about this new organization, I wrote to Bush, telling him that I was anxious to be of some service and that I was prepared to take on a fulltime job. As a result of this offer, I was one of the first appointees to Bush's new organization. And I was asked to take charge of organizing and directing a section of the OSRD called the fire-control section.
As used by the military, "fire control" refers to all the devices and procedures used to assure that any "projectile" (a shell fired from an anti-aircraft gun, a bomb dropped from an airplane, or a torpedo launched against a ship) will in fact hit the desired target.
During the course of the war the fire-control division of OSRD was concerned with the design and test of devices for all of these varied uses, but a big part of our effort was concerned with "ground-to-air" fire; that is, the equipment controlling ground-based anti-aircraft guns, which may be anything from 50-caliber machine guns up to the 90millimeter rifles whose targets are high-flying aircraft.
To hit an enemy airplane you have to know where it is and how it is moving, so that you can calculate where it will be by the time the shell reaches it. The equipment that furnishes the data about position and motion of the target can be called the tracking mechanism. This mechanism utilizes both optical and electrical means (radar), the latter procedure involving the sending of a pulse of electrical energy which is reflected by the enemy plane back down to the tracking equipment, which then computes the distance to the plane from the time required for the pulse to make the round trip to the plane and back.
There must be sophisticated and rapid equipment usually called a "director" or a "predictor" which calculates where the target plane will be when the projectile gets there, and which also calculates just how the anti-aircraft gun should be pointed and when it should be fired.
This calculation must take into account not only the characteristics of the target's motion but also the ballistic path of the shells, as influenced by wind and air density at different levels. Since the guns are heavy, there must be powerful equipment, controlled by the predictor, which continuously adjusts the orientation of the gun.
When we came on the scene, the Navy had comparatively advanced devices to carry out the functions of fire control, their equipment involving gyroscopic stabilization to take care of the roll and pitch of the vessel, rapid computation, and automatic power control of the guns. The Army, however, had equipment that had hardly been improved since World War I and was useless in handling the heights and speeds of World War II targets.
It is not surprising that anti-aircraft fire, using the then existing equipment, was ineffective. The bombing of England began on July 10, 1940 (the very day on which I had the first conference about my new task with Vannevar Bush), and in October 1940 it was estimated that at least 10,000 rounds of fire from 3-inch anti-aircraft guns were expended for each plane shot down in the London area. Even that figure, moreover, is almost certainly too low.
My first move, after a long conference with Bush in which he outlined the array of problems and opportunities as he saw them, was the selection of a group of key associates, and in that selection I was fortunate indeed—or expertly advised, or both. I have never enjoyed working with a large "committee," and I originally chose only three persons as the core of our organization. They were—alphabetically listed for I could not possibly assign any priorities— Samuel H. Caldwell, then a professor of electrical engineering at MIT who had been associated with Bush in the development of electrical and mechanical computing devices; Thornton Carl Fry, the head of the mathematics group at the Bell Telephone Laboratories and an extremely clearheaded person with imagination as well as knowledge of the analytical theory and the practical construction of all types of electrical devices; and Edward J. Poitras. Ed was —and is—a man of very special personal and professional qualifications who at that time had most recently been engaged in designing the automatic control system for the 200-inch telescope on Mount Palomar—a pointing problem which, except for the vast differences in angular velocities, has many similarities with gun-pointing problems. Ed knew gyroscope theory and practice, and was extraordinarily gifted and ingenious in his knowledge of all sorts of control procedures.
To our central committee we soon added Duncan J. Stewart, then the chief engineer of the Barber-Coleman Company of Rockford, Illinois, a concern with special skill in the design and construction of intricate precision devices as well as in control engineering. This control committee we supplemented, as the work progressed, with a considerable group of full-time technical aides, each especially chosen for a particular line, with an extensive group of consultants, largely drawn from engineering industries. The actual projects were ordinarily carried out under contract with engineering firms.
It is a tragic shame that peace seems to offer almost no challenge that will evoke the best from men, but war, horrid and cruel as it is, often brings out the highest qualities of everyone.
We met the most dedicated cooperation in universities and in many technical industries. In the case of the various branches of the armed services we were initially—and quite understandably—met by the attitude, "NDRC, what is that and who are you?" After all, we were raw amateurs, meeting men who were lifetime professionals.
There were numerous groups and military and industrial agencies with which we had to establish working relations, from whom we had to learn what needed to be done, and with whom we had to explore the possibilities where we might be useful.
Certain major facts emerged fairly promptly: that for the Army the problem of small-arms fire was both more difficult and less pressing than the problem of furnishing good fire control for a weapon capable of shooting down high planes—say the 90-millimeter rifle which was by that time scheduled to replace the older 3-inch weapon; that the Air Corps was pretty content with its Norden bomb sight, and was not yet ready to worry much about plane-to-plane fire. Indeed, about the only way at that time for a fighter plane to shoot down an enemy plane was to outfly and outmaneuver him, getting so close on his tail that all the pilot had to do was pull the trigger. Plane-to-ground problems, and the defense of big bombers against fighters, were to come later.
We carried out various analytical studies and practical tests which showed that the mechanical methods (gears and cams) that had previously been used in directors or predictors would be neither rapid enough nor accurate enough to cope with the fast targets of World War II. At just the moment when it was obvious that a brand-new approach would be necessary, the Bell Telephone Laboratories came forward with the proposal that they develop an electrical gun director. The original concept of this novel device had occurred to one of their engineers, D. B. Parkinson, in a dream!
Mervin L. Kelly, director of research at the Bell Telephone Laboratories, had proposed the development of this instrument to the Ordnance Department but had been rebuffed. He asked Fry whether the OSRD would consider backing the idea. It was, at that moment, not at all clear that the proposed electrical director would be either as rapid or as accurate as the existing mechanical directors; for its computation process relied on various electrical devices of novel and unproven design. But the electrical instrument could be produced in large numbers by relatively unskilled labor, whereas the Sperry type required precision machine tools and machine-tool skills. Thornton Fry therefore decided to recommend the project to our group.
On October 24, 1940, we had our first conference at the Bell Telephone Laboratories, and their group outlined their thinking about an electrical gun director or predictor.
On November 4 I had a meeting with the officers of the Technical Staff, Army Ordnance. It was my task to sell the idea of an electrical predictor to a group unacquainted with the techniques involved and traditionally skeptical about the use under battle conditions of anything "electrical." However, the meeting concluded with a decision to request the NDRC to proceed, to be responsible for financing, and to take full charge of all technical supervision and direction.
From that moment on we kept in closest touch with the BTL group. We assisted their contacts with various service groups, obtained data for them, and held frequent conferences with them at which all the aspects of their design and plans were thoroughly discussed.
Throughout 1941 we had long monthly conferences at the Bell Telephone Laboratories. We acted, on a continuous schedule, as the link between the Bell Labs group and the U.S. and British service authorities, as well as all the other scientists, engineers, and industrial groups interested in the director problem.
By September 1941, a first model was being assembled. On November 3 an "open house" was held at the Bell Labs.
To call it "open" is a misnomer; for the development was of course carried out under closest secrecy. The select group invited was told: "Overalls will be worn. Bring your own connecting clips. Lubricating oil will be served at 4:30 p.m." On November 29 the electrical gun director went to the Coast Artillery Board at Fort Monroe for test.
There were, of course, preliminary difficulties when the instrument was first tested. This was a brand-new device, involving many novel parts. Minor errors turned up in the tests and had to be corrected. A major problem was the design of suitable "smoothing circuits" to average out the inevitable fluctuations in the input data. But in February 1942, the instrument was accepted by the Army, and its experimental designation as T-10 was changed to the production label M-9.
This was by no means the only director project which Section D-2 set up and supervised. A number of novel ideas came to the surface and were tried out. At the Bell Labs itself another design, using a different coordinate system and capable of predicting along curved flight paths, was given a great deal of attention. All this was necessary insurance, although the alternative designs were never used.
To go on with the M-9 story: On June 12, 1944, the first German "buzz bombs," the V-1's, began to rain down terror on London. These were small pilot-less aircraft— or as we say now, missiles—preset when launched to descend and detonate in the London area. By this time three American developments, all carried out under the OSRD, had progressed to the point of practical service. A number of U.S. 90-millimeter anti-aircraft batteries with automatic power controls were installed on the east coast of England. Radar equipment developed by another section of the OSRD carried out automatic tracking of the enemy "planes" (in this case, the buzz bombs), and furnished the data about location and flight path. The BTL director M-9 accepted these data and computed the predicted position. The fuse-time settings did not have to be utilized, for by then still another OSRD development was available for use, the proximity fuse. This device, developed under the inspired direction of the physicist Merle Tuve (his section of the OSRD was so clearly a personal affair that it was designated as Section "T"), was an electrical marvel that replaced the time fuse in the nose of the shells and automatically detonated them when they came into the near neighborhood of the target.
The success of these three interdependent devices, the radar set, the electrical director, and the proximity fuse, is a matter of record. As of August 2, 1944, before the U.S. batteries became operative, the cumulative averages on destructions of V-1 were that, of those launched, anti-aircraft fire destroyed 10 per cent. After the five battalions of U.S. 90-millimeter guns were installed, equipped with automatic radar tracking, the M-9 director, and proximity fuses, the percentage of buzz bombs destroyed by A.A. fire rose by a factor of five, to 50 per cent. Actually the shooting record was better than this average figure indicates. Of the V-1's actually engaged by the heavy A.A., the percentage of kills was about 80. For example, the U.S. 124th and 125th Battalions, during one period, destroyed thirty-one of forty that were launched over their area. One battery, observed by Clarence A. Lovell of Bell Telephone Laboratories, who made a trip there, engaged nine buzz bombs, all of which were shot down.
Roughly 300 M-9's were landed in France on D-Day, being floated ashore in waterproof packages. One M-9 floated out to sea and had to be sunk by our own naval shellfire. In operation on the beach, on D + 1, a battery engaged three targets and shot down two. The performance in France set completely new records for anti-aircraft fire.
On August 12, 1944, General Sir Frederick A. Pile, in charge of the British anti-aircraft command, wrote to General George Marshall a letter referring to the destruction of buzz bombs (our division was sent a copy) which said in part, "The equipment you have sent us is absolutely first class, and every day we are getting better results with it.... We are employing the SCR 584* with the BTL predictor. This predictor is also an outstanding job.... Finally there is the fuse which is so secret that I can only describe it by its nickname in this country, 'Bonzo'... Our percentage of 'kills' is not high enough, but the curve is going up at a nice pace.... As the troops get more expert with the equipment I have no doubt very few bombs will reach London.... All this is due in the first place to you for sending us the equipment, and then to the extraordinarily skilled designers and the many fine workmen who had a hand in producing three of the most outstanding A.A. equipments of the war."
I have devoted considerable space to the account of the design and development of a successful electrical anti-aircraft predictor, because this was one of the largest and most useful of the projects sponsored by the fire-control section of OSRD.
However, we were involved in dozens of other projects with special emphasis on the sighting systems used to direct the guns of an airplane against enemy aircraft, and on equipment which made possible accurate and realistic testing of fire-control equipment. We had rather special success with the design of a bomb sight for use in low-level attacks on submarines. A thousand of these sights were built, of which half went to the Atlantic Fleet and half to the British.
* The official designation of the automatic radar tracker.
Long after the war, Poitras, in France on a business trip, was talking to an official of Air France. Reminiscing about war days, the Frenchman explained that he had served with the Free French forces, and as a bombing pilot had operated out of Oran and Dakar with American B-24 planes, hunting subs. "Did you have any kills?" Poitras asked. "But yes, m'sieur, with the magnificent sight I had, it was impossible to miss." Poitras went on to ask about this sight—did it have this, did it operate thus? The Frenchman, more and more amazed, finally asked how his American friend could possibly know so much about the sight. "Well, you see," said Poitras, "I designed it."
By 1942 it had become evident that the activities of the fire-control section were developing along two rather different directions, one concerned with the design and production of "hardware"—i.e., actual operating devices of various sorts, and the other concerned primarily with the mathematical analysis of problems, this involving paper rather than hardware.
The problems we worked on sometimes related to, and were preliminary to, the design of devices; often they related to the optimum employment of devices; and sometimes they were of still broader character, concerned with tactical or even strategic plans.
As the war went on, the emphasis on the design and production of hardware necessarily tapered off somewhat, for the practical reason that by then a brand-new device simply could not be conceived of, designed, built in pilot model, tested, improved, standardized, and put into service in time to affect the conduct of the war. On the other hand, the demands to carry out analytical studies kept increasing rapidly.
In the summer of 1942 we had had to recruit more and more mathematicians in order to keep up with the demands on us. And by the late fall of 1942, Dr. Bush and his colleagues heading up the OSRD carried out a reorganization which shifted the fire-control problems to a new Division 7 and created a new OSRD agency called AMP, or Applied Mathematics Panel. This panel was asked to be of general assistance in connection with analytical and mathematical problems, not only for Division 7 but for all the other divisions of OSRD as well—even more broadly, for the services and the war effort. I was continued as a member of the new Division 7, and was there placed in charge of the analytical aspects of fire-control problems; and I was made the chief of the Applied Mathematics Panel. The new Division 7 was headed by Harold L. Hazen, then head of Electrical Engineering at MIT and later Dean of the Graduate School.
The Applied Mathematics Panel was relatively small, consisting of the mathematicians Richard Courant, Griffith Conrad Evans, Thornton Carl Fry, Lawrence Murray Graves, Harold Marston Morse, Oswald Veblen, and Samuel Stanley Wilks. They were officially known as the "Committee Advisory to the Scientific Officer," that being myself. We also had the incomparably fine assistance of Dr. Mina Rees as chief technical aide. We sponsored and directed the work of several hundred other mathematicians. Many of these were men whose primary interests were in the purest of pure mathematics but who were unselfishly willing to devote themselves, during the war, to very specific applied problems. The roster included a good number of the ablest mathematicians of our country. We also were fortunate in recruiting a number of men highly skilled in statistical techniques although not professionally classified as mathematicians. Notable among this latter group were W. Allen Wallis, now president of the University of Rochester, and Milton Friedman, now holding a distinguished professorship of economics at the University of Chicago.
To those unfamiliar with the power of mathematical analysis, it may seem strange that there were so many demands on the Applied Mathematics Panel from the different branches of the armed services.
Some of our studies were of the "operations research" type, furnishing a guide as to how certain military actions could most effectively be carried out—for example, what kind of bombing attack had the best chance of clearing a safe passage through a mined area; how a multiple salvo of torpedos should be aimed to have the highest probability of hitting an enemy vessel carrying out evasive action; what flight pattern for a group of bombers would minimize the probability that shots fired at attacking enemy fighter planes would inflict damage to our own planes.
Under the auspices of the Applied Mathematics Panel were developed powerful new statistical techniques which improved the efficiency and lowered the cost of testing our own war matériel.
In just one such instance, involving improved testing of the propellant for rockets, the financial saving—not to mention the improvement in the matériel —was so great that within a few months it was sufficient to pay the cost of the total program of the Applied Mathematics Panel throughout the war.
The whole kaleidoscopic pattern of activity of the Applied Mathematics Panel does not easily lend itself to condensed description. We set up, carried through, and reported on a total of 194 studies. They were summarized, after the war, in four published volumes.
I have postponed to the end of this chapter the report of a period that chronologically belongs near the beginning. Early in 1941 it became clear to the officials of Dr. Bush's OSRD that close contact must be established with the British military experts and with the English scientists, so many of whom were by then devoting their energies and abilities to the war. President Roosevelt accordingly appointed an official scientific mission. The first group to go, on February 15, was headed by James B. Conant and included Carroll Wilson (for a time the general manager of the Atomic Energy Commission and now a professor at MIT) and Frederick L. Hovde (now president of Purdue University). When Conant and Wilson returned to the United States, Hovde stayed on as the permanent secretary of the mission.
The second group to go, following soon after the first, consisted of Dr. Kenneth T. Bainbridge, physicist at Harvard, Ed Poitras, and myself.
We sailed on March 3 on the Sibony, a rickety and dirty old wreck that was leased to the American Export Lines by the "Cuba Mail," the designation assigned to the Ward Line after the Morro Castle tragedy had removed whatever luster the original company name had previously possessed. The ship had been decommissioned, presumably as unfit; but under the demands of war she was again, albeit somewhat dubiously, in service. The trip to Lisbon, with a stopover at Bermuda, was tedious, uncomfortable, and unpleasant. Ed and I threatened to move from our cabin to the nearby men's toilet, which was warmer and smelled better.
We arrived in Lisbon on March 14, and there had a three-day delay before flying to England.
From the very first, our stay in England was delightful and rewarding, for we were so warmly accepted and assisted by British scientists, many of whom I had come to know on my Rockefeller Foundation visits to England, and because we were given every conceivable assistance by top military authorities. We were able to meet and discuss problems with all the military, technical, and scientific personnel responsible for anti-aircraft and other fire-control problems. We were at Portsmouth, Plymouth, Exeter, and other frequently attacked locations on the vulnerable south coast. We went to Aberporth in Wales, where rocket research was going forward, and also to Swansea. We made short visits to Cambridge and Oxford. At Cambridge, where few incendiaries fell, one landed on the roof of the famed Cavendish Laboratory, fell through, and landed plop in a sink where it harmlessly burned out! We had an impressive session with the Ordnance Board, an organization with representation from the Army, Navy, and Air, originally set up in the fifteenth century.
We had one night with an anti-aircraft battery on the outskirts of London. And on the night of April 16, 1941, we had the stimulating but somewhat dubious pleasure of sitting out the worst night of bombing that London experienced throughout the war. The raid started before nine in the evening and lasted until after five in the morning. It was estimated that between 500 and 1,000 German bombers were over London that night, and about fifty bombs fell in the close vicinity of Grosvenor House, where we were staying. From the roof of our hotel one saw a ring of huge fires, over toward the City, the Cheapside area, and the East India Docks.
But the next morning everyone was calmly and efficiently at work, bringing all the fires under control so that they would not furnish directing targets for the next night, and re-establishing electric, water, and phone service.
The total picture, as eventually recalled, is made up of a large number of small vignettes, each itself unimportant. There was, for example, the night at dinner when I asked the waiter for some mustard with my lamb. It was a tense night, but his traditions did not waver. Very politely but firmly he said, "It is, sir, very unorthodox with lamb, but I will bring it."
One night we were staying—the only Americans—in a country inn not far from London. We arrived after "closing time" but an exception was made and the Americans were given a drink. We were at once accepted by all the persons in the lounge, and Anglo-American relations were toasted all around with vigor and friendly warmth. Soon one of the wives, who had been upstairs tucking in her five-year-old son, rejoined the group and said, "I think you should know what just happened. I told my small son there were two Americans downstairs, and he said, 'Mummy, did you remember to thank the American gentlemen for the big bombers?'" The next morning we even had five prunes apiece and a small piece of bacon with our egg.
One day, at lunch in a pub with officers from a nearby military establishment, one of them told us of a toy automobile-rocket contest some officers recently ran. Each team of two was allowed to spend up to a shilling for the automobile, to which they would lash a small toy rocket. All the contestants would be lined up on the side of a tennis court, the rockets lit, and the first toy auto across the court won the grand prize. He and his companion went to the big toy store on Regent Street—Hamley's—and asked to see toy automobiles. The first one the girl showed them cost thirtyfive guineas! They gradually backed her down to cheaper and cheaper ones, and finally one of the officers smiled and said to the girl, "But Mummy said I wasn't to spend over a shilling." They finally got a cheap enough model, and spent the rest of the afternoon trying it out in Regent's Park, to the considerable irritation of the park policemen.
On the plane down to Portugal, starting our way back to our warm, safe homes, I wrote in my diary: "I am moved to pay my tribute to the qualities of the average working-class Englishman. The morning after our big blitz our waiter came in, with our rolls and coffee, cheerful as ever; and when we inevitably turned to talking of the raid it developed that from twelve to three in the morning he had been up on the roof 'spotting,' fully exposed to the whole direct terror of that flaming night. We later got out of him that his home had been bombed some months ago, that a shell splinter had gone right through the only suits he owned, and that the roof caved in and it rained steadily for the three days which intervened before the authorities dared allow him to re-enter his house. But for all these things he had only a shrug and a smile. He was touchingly appreciative when we gave him a suit, an overcoat, and a pair of shoes; and at eight o'clock that night he was still on duty and still smiling. There are millions of such people in England today, carrying great burdens with modest gallantry."