The Secret in Building 26

by Jim DeBrosse

  6

  The Turing Memo

  December 21, 1942—Dayton, Ohio

  WHEN ALAN TURING arrived in Dayton’s Union Terminal at 2:00 P.M.—his train from Washington, D.C., delayed six hours by a snowstorm—he must have felt himself literally a stranger in a strange land, having entered for the first time the heartland of a culture he found both abhorrent and fascinating. Then again, Turing was a misfit no matter where he found himself.

  The thirty-year-old mastermind of the British Bombe had been too much an individualist for the tradition-bound circles of Cambridge, where he first made his mark as a young theoretician. And he had been too much the English gentleman for the competitive world of Princeton, where he pursued his postdoctoral research among the likes of Einstein and John von Neumann. Further placing himself in his own category, Turing was not only gay but an avowed atheist who had devoted much of his adult life to pondering ways of turning human logic into machinery. Turing had laid down the principles of the modern, all-purpose computer, predicting, quite correctly, that one day thinking machines “would play a good game of chess.”

  Turing’s eccentricities were legend during his reign as the top codebreaker at Bletchley Park. He seemed almost to cultivate an image as a rumpled, absentminded scholar, chaining his tea mug to his radiator (mugs were at a premium in England during the war), wearing a gas mask around town to curb his allergies, and riding a broken-down bicycle with a chain that fell off after a certain number of revolutions, both to discourage thieves and because he was fascinated by its periodic pattern.

  The visit to NCR was only one part of Turing’s assigned tour of the States in the winter of 1942–1943. He had already spent much time with the young cryptanalysts in Engstrom’s research section in Washington. Turing had gone on to Dayton with the permission of Bletchley Park, which now had an intense interest in the Navy’s Bombe development and other codebreaking machines being explored at NCR. He was visiting not only to offer his advice but to report back to his superiors on the progress of the Americans. The British wanted to know how much they could rely upon the promised American Bombes as well as upon the American pledge to be only helpers in Britain’s intelligence battle against Germany. In November 1942, not long before Turing left for the States, his commander, Edward Travis, sent this memo to Wenger at OP20G: “Should be glad if Turing . . . could examine machinery. Make any use you like of him in connection with bombes. Have suggested he stay a week in Washington but if you should like him longer I should be quite willing.”

  The British had already provided the Americans with much technical detail on their existing Bombes and given the research group at OP20G an outline of their hopes for new four-wheel machines as well as modified three-wheel Bombes. GCCS, in fact, asked that OP20G send a third team to England to learn the details of their first four-wheel Bombes, which they mistakenly believed would be in operation before Turing’s return to England in early 1943.

  But the advice Turing was able to give the Americans was limited for a number of reasons. Although Desch was still in the beginning stages of his project and had not even a working prototype, the British sensed that OP20G wasn’t likely to be receptive to any radical changes in Desch’s design, given the Navy’s desire to produce machines by the first months of 1943. Turing, who was not known for his diplomatic skills, had probably been told by his superiors to be as discreet as possible and, specifically, not to exhibit any of Britain’s anger over the Americans’ increasingly independent Bombe policies. The British weren’t happy with the American decision to build a Bombe, but without a working four-wheel Bombe of its own GCCS had to accept OP20G’s venture. Two months before Turing’s visit, the two sides had reached a compromise, signed by a GCCS representative and the head of U.S. Naval Communications, captain Carl Holden, known as the Holden agreement: Bletchley would continue to be in charge of the European and U-boat codebreaking work, but the Americans would now be a full, if junior, partner in the operation.

  Knowledge of Turing’s visit to Dayton was restricted to a select few in Britain: that a math genius was traveling to a small midwestern city in America would certainly raise questions about the kind of work being done there. Joan Clarke, Turing’s fiancée at the time, knew of his travels along the East Coast during the war, but even decades later denied that Turing had ever been to Dayton.

  Although there had been hints in OP20G histories that Turing had traveled to NCR and that he had criticized Desch’s efforts, the visit was not confirmed until 1999 when federal archivist Lee A. Gladwin found proof among a newly released collection of documents that had been sealed for decades in the Navy’s Crane, Indiana, archive. A nine-page memo written by Turing for his superiors in En-gland made the facts indisputably clear from the title: “Visit to the National Cash Register Corporation of Dayton, Ohio.”

  The memo began with an apology—or perhaps an implicit criticism of American trains—for the lateness of his arrival that first afternoon at “the works . . . where the Bombes are being made.” As a result, he wrote, “we did not have quite so long there as we might have had, but probably sufficient.”

  Turing did not complain about his less-than-elegant lodgings in Dayton, although he may have had reason to. Debbie Anderson said her father mentioned before his death in 1987 that Turing had spent the night on their living-room floor. While there is no proof of this, the Desch home may have been the only residence in Dayton under the protection of plainclothes guards at the time. Perhaps, too, under Dayton’s crowded wartime conditions the Desches’ offer of lodgings seemed entirely hospitable. The four OP20G officers from Washington who had accompanied Turing on his visit—among them Joe Eachus, who had worked with Turing at Bletchley the summer before—ended up sleeping on a hallway floor in a downtown hotel.

  Despite their differences in background, as well as Desch’s fears that Turing was there to undermine his work, the two men may have struck a chord with each other, at least intellectually. Desch shared Turing’s interests in the concept of real numbers as well as in mathematically based cryptography, Anderson said. Both men claimed to have invented an unbreakable code at one time in their lives. One can only wonder if the two men talked about their ideas over martinis before a crackling fire in the Desches’ living room. Then again, maybe they did not. Turing had an aversion to conversation with most Americans, as he had made clear while he was at Princeton, in a letter to a friend in England:

  These Americans have various peculiarities in conversation which catch the ear somehow. Whenever you thank them for anything, they say, ‘You’re welcome.’ I rather liked it at first, thinking I was welcome, but now I find it comes back like a ball thrown against a wall, and become positively apprehensive. Another habit they have is to make the sound described by authors as ‘Aha.’ They use it when they have no suitable reply to a remark, but think that silence would be rude.

  There’s no evidence that Turing was outwardly rude to anyone during his brief visit to Dayton, although Bob Mumma once recalled an argument that Turing and Desch had over the wire brushes to be used in the NCR machines. It’s more likely Turing was oblivious to those he met or frightened of them and found his refuge in the self-consistent, self-contained world of pure math. Even in the company of mathematicians, Turing was less than outgoing. At a later stop in his trip to the States, new acquaintances there “complained of Alan giving no sign of recognition or greeting when he passed them in halls; instead, he seemed to ‘look straight through them.’ ”

  Turing’s withdrawn and somewhat cold approach to people wasn’t limited to Americans, according to his biographer Andrew Hodges. At Bletchley Park, Alex Fowler, who was just over forty, took Turing to task about his lack of social skills. Turing was stung and felt the need to explain to Fowler why he struggled with so many aspects of life and social relations. Fowler recalled what the younger man had told him: “ ‘You know at Cambridge,’ he said, ‘you come out in the morning and it’s redundant to keep saying hallo, hallo, hallo.’ �
�� Hodges concluded that Turing “was too conscious of what he was doing, to slip into conventions without thinking.”

  Turing’s problems with social relationships and, later, his entanglement in then-illegal homosexual affairs, led to his being marginalized at Bletchley during the last years of the war. In 1952, the same year he was awarded Britain’s highest civilian honor for his wartime service, the Order of the British Empire, he was arrested, put on trial for homosexuality, and, as a result, placed on probation. Two years later—and two weeks shy of his forty-second birthday—he committed suicide. In a scenario reminiscent of one of his favorite movies, Walt Disney’s Snow White and the Seven Dwarfs, he dipped an apple in a mixture of jam and cyanide. Then, without writing a note to explain his actions, he took several bites and passed away.

  HOW MANY TIMES Turing came to NCR remains uncertain. Anderson believes he made more than one trip to Dayton, based in part on her father’s comments and those of his assistant, Bob Mumma, years later to a Smithsonian historian. If Turing had found time to tour the city, he would have encountered a community that, although not threatened with bombing as were the cities in England, was nonetheless stressed by war.

  By 1943, the war economy dominated the working and family life of industrial cities like Dayton to a degree that is hard to imagine today. In addition to the rationing of meat, gasoline, sugar, and even coffee, some sixty war-production industries employed nearly half the city’s population—about 115,000 workers—and turned out products that were sent to nearly every Allied front. Another 45,000 men and women were employed at the three military air bases in the area: Wright, Patterson, and McCook fields.

  Starting in 1941, the peacetime production of Dayton’s factories—cash registers, automotive parts, refrigerators, and tires—had been quickly converted to meet the needs of fighting men. The Inland Division of General Motors was turning out the M-30, considered one of the best rifles of the war, as well as a one-pound pistol called the “Little Monster,” to be air-dropped to resistance fighters in Europe. GM’s Frigidaire division switched from building refrigerators to turning out giant propellers for the B-29 and .50-caliber machine guns for defending the B-17 Flying Fortresses. NCR built the carburetors for both the B-24 Liberator and B-29s, as well as shell fuses, computing gun sights, a rocket motor, and a variety of gun parts and shells. GM’s Delco division cranked out tank treads, special shock absorbers for Army trucks and tanks, bomb fuses, and the Allison liquid-cooled engines used in Bell Airacobras, Curtiss P-40s, and Lockheed P-38 interceptors.

  By 1943, the forty-eight-hour workweek was standard in Dayton and many other industrial cities, and most factory hands worked much longer. The long hours meant that many workers had little time to shop for necessities. In an era when family life was still valued above convenience, downtown stores usually closed at 5:45 P.M. and all day Sunday. Dayton retailers may have been the first in the country to extend their hours until 8:30 P.M. on Monday and Wednesday nights. Even banks began to keep their doors open on Friday nights so that factory workers could cash their checks.

  With thousands of men off to war, labor shortages were critical. To ease the shortage, some plants created “buddy shifts” in which high school students filled in after school starting at 7:00 P.M., allowing full-time workers to break for dinner before returning to their jobs at 10:00 P.M.

  The almost insatiable needs of the war plants triggered unprecedented efforts to expand the Dayton labor force. More than ninety-three thousand registration cards were mailed, one to every home in Dayton and Montgomery counties, urging women to join the workforce. The number of women in the Dayton workforce nearly doubled between 1942 and 1943 alone. Recruiters were dispatched to surrounding rural areas and as far south as the hardscrabble hills and isolated hollows of Appalachia, where they touted the wages and benefits to be earned in Dayton plants. From a population of 211,000 in 1940—then the fortieth largest city in America—Dayton swelled by another 43,000 by the end of the war.

  The city’s infrastructure showed the strain. Hotels needed a week’s notice for reservations, traffic snarled, trailer camps ballooned, and more than 2,500 temporary housing units went up all over town. And with so many fathers overseas and mothers working outside the home, juvenile-delinquency rates jumped almost overnight. From January through March 1943, the number of first-time offenders brought into Montgomery County juvenile court—407 cases in all—nearly doubled over the same period the year before. As one distraught judge observed, most of the growth in new cases was occurring among eleven- to fourteen-year-olds. To address the problem for the duration of the war, city commissioners signed a petition in February 1943 calling for a voluntary curfew of 10:00 P.M. for children ages twelve and under and of midnight for high schoolers.

  Given Turing’s inward-looking temperament, it’s doubtful he noticed how much the war effort was transforming America. Eachus said he remembered little of Turing’s reactions, except that the mathematician had noted with ironic interest that the NCR site housing the U.S. Bombe project “was Building 26, and that there are, of course, twenty-six letters in the alphabet.” Eachus said he never found Turing to be arrogant in his dealings with Americans.

  But some in Dayton at the time felt otherwise. Phil Bochicchio, a Navy mechanic who did much of the troubleshooting maintenance on the Dayton Bombes, said British visitors in general “acted like they knew everything more than we did. Maybe they did. But they were as peculiar as hell, and, like God, they were way up there looking down on us.”

  The Navy’s reluctance to reveal to the staff in Dayton all that OP20G knew about the British machines and methods probably caused some of the resentment at NCR toward the British. Although Desch left no written comments on his encounters with the British or with Turing, he did complain in a 1973 interview with Smithsonian historian Henry Tropp that “it was a one-way street. The British came over and visited me and looked at everything I was doing, but I could never see anything they were doing. And when you mention Turing, he was one of them that came over here very frequently.”

  Anderson discovered among her father’s wartime memorabilia a photo he had taken at a July 1943 outdoor gathering at NCR, where a young man who looks remarkably like Turing is standing with his arms akimbo in a dark tweed jacket among U.S. naval officers in dress whites. Desch left no identification of anyone in the photo, though Ralph Meader is easily recognized in the crowd.

  Turing’s memo to GCCS about his visit to Dayton seems rather evenhanded, especially given his somewhat indelicate writing style as well as the tardy and meandering development of the American project. Turing provided valuable insights for Desch and his team, but by swaying Desch’s superiors at OP20G to make last-minute changes in their approach he also complicated Desch’s already daunting assignment. Desch and John Howard, now the Navy’s supervising engineer on the project, scrambled to include the changes in a second Bombe design even as they rushed to construct the first prototype.

  Turing began his critique of the December 1942 American Bombe by arguing against one of Desch’s most prized features: the automatic rewind. Turing didn’t think the feature extravagant or superfluous, but he thought a faster, more ambitious design would eliminate having to stop the machine at all. The British, in fact, had hoped to avoid the cumbersome stops for the Diagonal Board test in their own high-speed version of the four-wheel Bombe. In this context, Turing’s doubts about Desch’s rewind feature seem reasonable.

  We were given a demonstration of how the motor was able to reverse and be going full speed in the reverse direction in a fraction of a second, with the full load: however, this seems to me hardly to prove that all will be well when one tries to reverse the Bombe itself, e.g. the gears might get distorted under the strain. They say that the whole machine is being built sufficiently strongly [sic] to withstand such strain. Possibly the real objection to this method is that the time taken over each stop is fairly considerable, viz 15 seconds, and of course it seems a pity for them to go ou
t of their way to build the machine to do all this stopping if it is not necessary.

  Engstrom originally had liked the rewind feature, but apparently Turing persuaded him to lobby Desch to drop it. Desch refused—and, as it turned out, he was right. His mechanism performed admirably throughout the war, while the British failed in several attempts to produce a working nonstop Bombe.

  Turing expressed a more valid criticism about Desch’s idea of using a commutator wheel of different size to simulate the first and fastest wheel on the Enigma. Desch worried that he might not be able to drive the fastest of the wheels at the necessary speeds (three thousand rpm) unless it was smaller than the rest and could therefore withstand the additional torque and centrifugal force. To head off the anticipated problems with overheating and distortion, Desch had wanted to use two smaller wheels instead of one large wheel. Smaller wheels were less likely to warp and chip as they whirled at high speeds.

  But Turing realized that different-size wheels made sense only if the Americans went through with their plans to build 336 fixed-wheel Bombes. If not, Turing’s experience told him that having to change two different kinds of wheels would be a logistic and operational nightmare for hard-pressed Bombe clerks. Adding another sixteen wheels to the sixty-four needed for each setup would significantly increase the changeover time between runs. In fact, Turing became a bit sharp in his comments about the idea: it made him “smile inwardly at the conception of [British] Bombe hut routine implied by this programme.”

  Turing’s point scored well with Engstrom, who had already realized that different wheel sizes would add considerably to the cost of the Bombe program. Those brass-and-copper wheels, sandwiched between heat-resistant Bakelite, were hard and expensive to manufacture. Worse, their 104 contacts had to be wired by hand and by trained military personnel who could be trusted not to talk about what they were doing. Providing the eleven thousand or so commutators OP20G required during the war did, as Engstrom feared, prove to be one of the greatest manufacturing and security challenges of the Bombe project.