by Jon Gertner
Scientists who worked on radar often quipped that radar won the war, whereas the atomic bomb merely ended it. This was not a minority view. The complexity of the military’s radar project ultimately rivaled that of the Manhattan Project, but with several exceptions. Notably, radar was a far larger investment on the part of the U.S. government, probably amounting to $3 billion as contrasted with $2 billion for the atomic bomb. In addition, radar wasn’t a single kind of device but multiple devices—there were dozens of different models—employing a similar technology that could be used on the ground, on water, or in the air. Perhaps most important, radar was both an offensive and a defensive weapon. It could be used to spot enemy aircraft, guide gunfire and bombs toward a target, identify enemy submarines, and land a plane at night or in thick fog. Its origins for domestic military uses dated back to the 1930s, when several scientists at the Naval Research Laboratory discovered that when they directed a radio pulse from a transmitter toward planes passing overhead some of the waves were reflected back. In 1937, the Navy approached Bell Labs to help it refine the technology. Those first applications were primitive and riddled with problems, but by the start of the war radar sets were being used in Great Britain, where an intricate network of coastal radar stations helped the British defend against the onslaught of German bombers. The U.S. military also used radar stations in the Pacific—indeed, the Japanese squadrons flying toward Pearl Harbor were picked up well before they arrived. The officers monitoring the stations disregarded their readings, thinking the blips to be friendly aircraft.15
In time difficulties like this were overcome—a method was devised to distinguish between friendly and enemy aircraft—but the rapid evolution of radar between 1937 and 1943, the same speedy development that had awed Kelly in assessing the innovative forces of war, was rife with frustrations. In early 1940, if one could by chance eavesdrop on a group of Bell Labs radio engineers discussing the ideal radar set, one would hear described a technology that with the help of vacuum tubes could transmit very brief electromagnetic pulses (perhaps a thousand per second) in a very focused beam of waves that measured perhaps ten or fifteen centimeters in length. This was a fraction of the length of regular radio waves, the ones that brought music and news, which were sometimes a hundred meters long. It was also an ideal and not a reality. At the start of the war in Europe, the vacuum tubes that powered the early radar sets mostly sent out longer waves measuring a meter or more in length; such waves were too diffuse to help “pilots home in on their quarry.”16 And when American scientists attempted to create sets that could emit shorter waves of thirty or forty centimeters, they discovered that their vacuum tubes lacked the power to send out a strong enough signal. “The big problem in radar is to generate enough power to get a detectable echo from a distant point,” Time magazine explained. “Of the total energy sent out in a radar beam scanning the skies, only a tiny fraction hits the target (e.g., a plane), and a much tinier echo gets back to the receiver. Engineers estimate that if the outgoing energy were represented by the sands of a beach, the returning echo would be just one grain of sand.”17
How to create a device to send out shorter waves, and with more power? There was a solution to this problem—it came by ship, in a veritable black box, by way of a secret mission to the United States by British scientists in the late summer of 1940. In the box was something called a cavity magnetron, a metal device invented by two physicists at the University of Birmingham that resembled a small fishing reel. “Unlike conventional vacuum tubes,” a Bell scientist explained, “with their components exposed in a glass envelope, the new tube was an inscrutable copper cylinder with cathode leads and a coaxial line emerging from it.” The magnetron whirled electrons inside its six or eight circular internal cavities to produce short waves of ten centimeters and transmissions of great power. It was brought to Kelly at the West Street labs not long after the British mission came to the United States—the idea was that Western Electric would be the logical company to mass-produce the device if Bell Labs could refine the technology and design. On October 6, Kelly watched as the magnetron was demonstrated for the first time in the United States at a small Bell Labs branch office in Whippany, New Jersey.18 The engineers in the room were dumbstruck by its power output. It wasn’t just an incremental improvement in radar engineering. Luis Alvarez, a physicist who wasn’t present that day but would later work directly on the magnetron designs, pointed out that the invention improved upon current technologies by a factor of three thousand. “If automobiles had been similarly improved,” he noted, “modern cars would cost about a dollar and go a thousand miles on a gallon of gas.”19
THE DEVELOPMENT WORK on magnetrons—the work that preceded their manufacture—was done in concert at MIT and Bell Labs. At MIT, an ad hoc group of scientists and engineers, eventually numbering several thousand, worked in a secure campus building with blacked-out windows known as the Rad Lab. At Bell Labs, many of the scientists on the radar project worked under Jim Fisk, who had already won favor for his work with Bill Shockley on uranium. Kelly directed him to set up a workshop in a building near the West Street offices that had once been a biscuit factory. The horses for the New York Police Department were housed nearby; Fisk would later recall that his senses were often sharpened by the “man-eating flies” that shuttled between his offices and the stables. At first Bell Labs’ technicians took X-ray photographs of the British magnetron so they could create blueprints; in those photographs the device looks like a small, circular metal plate that contains a linked chain of eight smaller circles around its inner edge. It resembled an old-style film projector reel. But it still wasn’t immediately clear how to manufacture devices for production. Engineers at the Labs knew that the gulf between an invention and a mass-produced product could in some cases be extraordinary, even insurmountable. “Could the new magnetron be reproduced quickly and in quantity?” Fisk wondered. “Was its operating life satisfactory? Could its efficiency and output power be substantially increased? Could one construct similar magnetrons at [wavelengths of] forty centimeters, at three centimeters, even at one centimeter?”20
In some respects, the work ahead of him was, by necessity, at odds with Kelly’s philosophical preferences. Instead of using scientific research for development so that he could then make a device, Fisk was reverse-engineering—analyzing an existing device so he could work out a research plan he would then use for development. He quickly saw that he couldn’t just make a magnetron bigger or smaller to get different outputs in microwave power and wavelength; engineering the device for ships and planes and other uses involved a careful consideration of everything about the magnetron—the size and number of its internal cavities, the shape of those cavities, its input voltage, and so on. Fisk carried around with him a notebook he filled with sketches and ideas. Often he consulted with Clinton Davisson, the thin, quiet researcher who stayed far away from development but was always willing to ponder a difficult problem, especially if it had to do with electrons.
Having just turned thirty years old, Fisk was now charged with perhaps the most important scientific project in the United States. Only a couple of years before, when he was finishing graduate school and beginning to look for a job, he had visited the research department at West Street to see his old MIT friend Bill Shockley. After he left, Shockley turned to a colleague and said, “Remember I told you. If that man gets hired, we’ll all be working for him in ten years.” Fisk himself had no managerial aspirations when he began at Bell Labs, but the war began to shape the course of his career in unexpected ways. “When we got all through,” he said of the various researchers doing wartime magnetron work, “some of us emerged as something else.”21 Kelly had nonetheless seen executive potential in Fisk from the start, when he’d offered him a job the first time they met over lunch at a midtown Manhattan hotel. “Jim was Mervin’s protégé,” Kelly’s son-in-law recalls, noting that Fisk was soon a frequent guest at Kelly’s house in Short Hills, New Jersey, joining other regula
rs in the Kelly home like Davisson, Buckley, and, during his occasional visits to the United States, the physicist Niels Bohr.22 In some ways the two had much in common—agile minds and a natural talent for decision-making. Both loved golf and gardening, though Kelly grew flowers and Fisk grew vegetables. Fisk—slender, patrician, polite, unflappable—was a more polished version of his mentor. Always with a cigar in hand, he had the soft touch and twinkle that Kelly lacked. On long train rides with colleagues (frequently they would go from New York to Western Electric factories in Chicago), Fisk would sometimes produce a bottle of Southern Comfort and pass it around.23 He was fond of putting his colleagues on mailing lists of doctors peddling dubious tonics. When a friend of his went away for a week to a New Hampshire resort, Fisk sent to him a series of telegrams that resembled the instructions of an underworld gangster to one of his cronies.24 At the magnetron workshop in the old biscuit factory, Fisk sometimes wore a striped train engineer’s cap and, on occasion, striped overalls to meetings. “After all, we’re engineers,” he would say. His friend Larry Walker recalled that at the biscuit factory work was done in a large open first-floor room divided by waist-high partitions. Sitting in an office, Walker remembered, “one saw only the upper half of the people as they passed.” It was Fisk’s occasional habit, “if he caught someone’s eye as he was leaving, to continue walking but gradually [bend] his knees farther and farther. Few who saw it can have forgotten the sight of Mr. F. apparently disappearing in a hole in the ground, eyes firmly ahead, chin up.”
Fisk was liked the same way that Kelly was feared. Soon enough, he became known for “Fiskian” aphorisms: “When you don’t know what to do,” he would say, “do something.” Or: “We have now successfully passed all our deadlines without meeting any of them.” The magnetron project was starting to look like that, as the men worked out the fiendish difficulties, often during all-night sessions, of not only putting the original model in production but modifying the device for new applications. The low point for Fisk’s team was the day after Pearl Harbor, as the men sat among stacks of nonworking magnetrons (they had poor vacuum seals, apparently) and listened to the grim news in the Pacific.25 At one point during the work, Fisk—perhaps seeking a diversion from the grinding six-days-a-week schedule—came up with the notion that the propulsion system of sharks was worth studying for applications to naval warfare. He decided to try to requisition $50,000 for a swimming pool, to be constructed in the basement of the biscuit building, from Bell Labs management. It was a joke with an appearance of plausibility (in fact, many years later at Bell Labs, biological systems would become an acceptable research pursuit). Fisk’s request rose through the executive ranks, receiving several green lights, until it got to Kelly, who instantly rejected it.26
. . .
FISK AND HIS STAFF eventually worked out fifteen different and successful magnetron designs for new radar instruments. His team passed its finished work on to Western Electric, which ultimately manufactured more than half of the radar sets used in World War II. Bill Shockley had been involved with the radar work at the start, but for the star Bell scientist, the war brought a profusion of other opportunities. In early 1942, Shockley received an offer from his friend Philip Morse, an MIT professor who was now in Washington organizing something called the Anti-Submarine Warfare Operations Research Group, commonly referred to as ASWORG. Morse asked Shockley to become the group’s research director, and Kelly agreed to give his prize physicist a leave of absence from Bell Labs. Essentially a small think tank, the ASWORG group was staffed by statisticians and physicists and even a chess grandmaster. Together, they used sophisticated probability calculations to solve military problems. At the start, they were charged with figuring how to better detect and destroy the elusive German U-boats that were making passage through the Atlantic deadly. The work—conceptual rather than experimental—suited Shockley well: He and his colleagues looked at statistical information to figure out more effective methods of combat. An early problem: Why did bombs dropped from Allied airplanes have no effect on German submarines that had come to the surface? Shockley realized that the bombs were set to explode at a depth of seventy-five feet—too deep to sink a sub that was on the surface. He suggested changing the depth charge to thirty feet. Morse noted that within two months the change increased by a factor of five the number of subs sunk by Allied bombs.
Shockley and his staff at ASWORG rarely went out into the field. But the military men grudgingly admitted the effectiveness of the insights coming from these desk-bound civilian scientists. With the help of an IBM data processing system, the team assumed record-keeping duties for the entire U.S. antisubmarine effort. They brought in a computer expert and several insurance actuaries to analyze data on “hits” and “misses.”27 Shockley became immersed in these efforts. As his Washington schedule became increasingly demanding, in fact, Shockley moved to the capital, eventually taking a room at the University Club and coming home by train on occasional weekends to see his family in Madison, New Jersey. His marriage was not going well. He barely saw his daughter or newborn son. A tendency to push himself to exhaustion, an old habit, returned. His new habit was to try and organize a life that was so absurdly busy it couldn’t possibly be organized. From the war onward until the end of his life, Shockley began to keep several different sets of calendars and diaries, some for his military work, some for his scientific work, and some for his home life. The war diaries suggest that his days were a blur of ideas and appointments and phone calls, all tucked between exercise regimens, doctor’s appointments, train trips, and lunch and dinner meetings.28 It hardly helped that his travel schedule was frenetic, too. As a testament to his high security clearance, he was authorized by the secretary of war to ride any commercial airplane in the country at any time. Often his schedule took him around the country as well as to Europe.29
The work, the travel, the wartime pressure—all of it bore down on him. In addition, Shockley was struggling to manage a few dozen men, some of whom in his opinion were not as bright or committed as he regarded himself. And here, perhaps for the first time, a clear sign of Shockley’s limitations emerged. Whatever his friend Jim Fisk found easy and natural—relaxing a roomful of scientists with some inspired slapstick, for instance, or giving men freedom to do their work as they chose—Shockley found difficult. He simply could not get the hang of managing people. Some fifty years later, Shockley’s biographer, Joel Shurkin, found among his private papers a sealed envelope from this period containing a note informing his wife that he had just attempted suicide. He had played Russian roulette with a revolver. “There was just one chance in six that the loaded chamber would be under the firing pin,” he wrote, before adding, with characteristic precision, that “there was some chance of a misfire even then.” He apologized to her that he hadn’t found a better way of solving their domestic problems.30 Shockley never gave her the note and returned to his work on the war.
BELL LABS VICE PRESIDENT Oliver Buckley wrote a form letter in January 1944 to the Bell Labs staff members on leave with the military services. The war hadn’t yet turned decisively in the Allies’ favor—D-Day and the invasion of Europe were still six months away—but the news was increasingly encouraging. “We hear a lot these days about ‘post war planning,’ and hear of some others doing research and development for the post-war period,” Buckley wrote. “As far as we are concerned, that is pretty well out. We want to see this war won before we start working on what is to follow.”31
It was not quite true. Nearly a year before, for instance, Mervin Kelly began writing a twenty-nine-page memorandum—“A First Record of Thoughts Concerning an Important Post-War Problem of the Bell Telephone Laboratories and Western Electric Company”—that was circulated to Buckley and a number of other executives at the Labs in May 1943. Kelly’s point was not to outline precisely what new products the company should make at the close of war. Rather, he felt compelled to outline his vision of how Bell Labs, once it regrouped, would fare within a
n electronics industry that was sure to grow exponentially after the war’s conclusion. “It is reasonable to expect that in the decade following the war there may well be changes of even greater total significance than those of the past thirty years,” Kelly wrote. Radar, he added, had already opened up vast new opportunities in the business of radio wave and microwave devices. Kelly also thought it likely that the telecommunications industry was destined to resemble, in its nature as well as its products, sister industries like radio and television. Before the war this had not been the case: Bell Labs researched and designed equipment for the highly specialized nature and problems of telephone service. But to Kelly, the era at hand would require different approaches. Deep within the long memo, he noted, “We have been a conservative and non-competitive organization. We engineer for high quality service, with long life, low maintenance costs, [and a] high factor of reliability as basic elements in our philosophy of design and manufacture. But our basic technology is becoming increasingly similar to that of a high volume, annual model, highly competitive, young, vigorous and growing industry.”32 In other words, there would soon be a revolution in electronics. And as he saw it, Bell Labs would need to lead it rather than join it.
Kelly wanted his old team back—the team he had handpicked in the late 1930s. First, he pursued Bill Shockley. In January 1944 Shockley had left his position with the Navy for a job at the War Department. Now working with a group considering radar problems on B-29 bombers, he had maintained his frenetic schedule, beginning with a world tour of B-29 bases overseas, “going by way of England and Italy, and visiting American and British bombing establishments en route,” the Bell Laboratories Record noted. “After six weeks at the B-29 base in India, he proceeded via Australia to 21st Bomber Command headquarters in the Marianas, then across the Pacific to the United States, where he arrived in January 1945.” Not long after, Kelly invited Shockley to Bell Labs for a series of meetings. The two men discussed Shockley’s possible return to industrial research, and Shockley thereafter agreed to leave Washington and return to Kelly’s shop. At first he began working part-time, visiting the Labs frequently in the late winter and spring of 1945. The war was finally winding down—Germany surrendered to the Allied armies that May. It had been several years since Shockley had done any work in physics, and the meetings and discussions at Bell Labs reawakened in him an old interest in some of the technologies he had been pursuing—the solid-state amplifier, for instance—just before the war.33 He returned to full-time work at the Labs immediately after the atomic bombs were dropped on Hiroshima and Nagasaki. Kelly had told him he had a new project in mind.