The Secret in Building 26

by Jim DeBrosse

  The Navy took over the 36,800-square-foot building at NCR and found it an ideal match for its purposes. Erected just four years before, Building 26 was one of the first structures in the country to use steel-reinforced concrete floors. It was strong enough to serve as a factory site for the five-thousand-pound Bombes and roomy enough with its twelve-foot-high ceilings and wide hallways for moving the massive machines, which stood seven feet high, eight feet long, and two feet wide—larger than three modern refrigerators side by side. The building was fireproof and also air-conditioned, thus protecting the Bombes’ sensitive electronic components. Not far behind the building ran the old Miami-Erie Canal and a spur of the Baltimore and Ohio railroad line, which quite conveniently led to Washington, D.C., in twelve hours.

  Its twenty-three classrooms provided privacy and security for each step of the manufacturing process, as did the glass-brick windows that let in plenty of daylight while blocking out potential snoopers. The project was broken up into units and, where necessary, assigned to different rooms, so that only the top managers knew enough to comprehend the purpose of the project. Each room had a thick wooden door that could be locked. Only if your name was posted on the door could you be admitted to that room. “And many people didn’t get in many rooms,” Kern said.

  The engineers sometimes felt the need to leaven the building’s strict need-to-know atmosphere. One day, just as a maintenance man finished drilling a hole from the hallway into one of the secure rooms, engineer Ralph Bruce stooped down and shouted through the hole in a booming voice: “Hey, do you have a pass to get in here?”

  Wenger soon got the organization for the project he had wanted, as free of bureaucratic constraints as possible. While the bean counters in the Navy’s Bureau of Ships at first controlled the money and were theoretically in charge of all the project’s technical details, the bureau yielded more and more autonomy to Wenger, Engstrom, and Desch. The end result was a new Navy entity, dubbed the U.S. Naval Computing Machine Laboratory, that would serve as an umbrella for all the Navy’s codebreaking-machine projects. In formal terms, the NCML was the boss of the Dayton operation, but by early 1943 it was really only a source of funding and support for the OP20G engineers and scientists. In essence, OP20G had its own machine-development program at what, in effect, was to become its own secret factory inside Building 26.

  It is unlikely that NCR executives at the time had any notion of how much of the company’s resources would be tied up by Navy demands as the Bombe project ballooned. In mid-1942, when Deeds ordered NCR to accept a “best efforts,” no-profit contract with the Navy, the agreement called for preliminary work on codebreaking machines, not a massive development and manufacturing effort. The effect of the open-ended contract, combined with the formation of the NCML, was to free Engstrom’s M section from the hassles of paperwork and bureaucratic interference. But at the same time, it bound NCR to the Navy: before the war was over, NCR executives began to chafe under the no-profit clause.

  The longtime cordial relations between Desch and the young men from MIT helped to create an open atmosphere in Dayton. And as luck would have it, the Bureau of Ships found someone they thought would be just the right man for managing a loosely controlled project of engineers and technicians who were unused to taking military orders: Commander Ralph Meader. Although over forty at the time, Meader was not a member of the bureau’s meddlesome old guard. As a Navy reservist, he had spent more than two decades working for the engineering divisions of America’s largest electrical corporations, including Western Union, often in a more freewheeling relationship as an inventor. He had earned several patents and a reputation as both a solid engineer and an even-tempered executive. He had learned how to work with, as well as to direct, creative people—perhaps, too, because he had been married to one for twenty years. He met his wife, Janet Clark McLaren—a graduate of the Royal Academy of Music in London—while she was traveling in the States with a theater company. She kept her show-business connection after their marriage and at one point brought her husband into the fold as a stage manager for a Broadway show.

  Like Engstrom, Meader had worked his way into the nation’s elite academic circles from a relatively modest background. His father was a shoemaker in Lynn, Massachusetts, who lost his shop in a swindle, according to Meader’s son, Bruce. Meader’s academic abilities gained him admission to the prestigious Boston Latin High School and later, on scholarship, to Dartmouth, where he played tennis and took all the toughest math courses. During World War I, he enlisted in the Navy as an ensign and, afterward, took graduate courses at Harvard and Columbia. Using the money he later earned from his engineering patents, Meader tried a variety of business ventures between the wars, most of them failures, including an import-export business with a Hungarian friend of his wife and an investment in a Florida quarry. He went bankrupt before World War II defending his patents on a repeating pen-and-pencil mechanism against a suit filed by EverSharp.

  ALTHOUGH RELATIONS WITH Meader and the Navy began harmoniously enough, Joe Desch and NCR began to feel put-upon after only a few months. It didn’t help matters that when Desch’s civilian team was transferred from NCR’s payroll to the Navy’s, they had to wait three weeks for their first paycheck. “For a young family back then, that was a real hardship,” recalled Carmelita Ford Bruce, widow of NCR engineer Ralph Bruce.

  Hundreds of sailors and WAVES, the Navy’s women’s auxiliary, would have to be assimilated into NCR’s operations to help Desch’s team with the assembly and maintenance of the machines. “The sailors were the hardest to control,” Jack Kern said,

  because they wouldn’t take orders from a civilian, and yet they were given to us to use. This made for a rather sticky situation for us. There were no naval officers in charge—they were all upstairs running Bombes. They were problem solving, testing, whatever one does behind closed doors in their situation. I think that only Joe Desch and perhaps [his assistant and business manager] Bob Mumma ever got in there. The Navy had direct wire connections to Washington and were actually transacting business.

  The converse relationship also got sticky—civilian employees having to take orders from the Navy. Once NCML and Meader took over Building 26 in late 1942, pressure was put on the NCR staff to work overtime, including weekends, even though only Desch and possibly Mumma knew the importance and urgency of the project. By May 1943, Desch’s team of engineers was placed on a wartime work schedule of fifty-four hours per week, according to payroll records, and many were putting in additional hours even at that. “I know of more than one person who worked twenty-four hours on Sunday and got paid double-time,” Kern said. “Saturday and Sunday were supposed to be off, yes, but we usually worked on Saturday.”

  The Navy’s heavy-handed attempts to increase the pace of work weren’t always successful, Kern remembered: “If you’re doing something mental, trying to overcome a problem or something, you can’t work fast, like you could if you were simply putting something together.”

  Meader, to his credit, eventually recognized that engineers couldn’t be managed like factory hands. In a report filed with Wenger after the war, he wrote:

  It must . . . be borne in mind that work being done by engineers in research and development cannot be passed on from shift to shift. Engineers must personally see their work thru [sic] to a conclusion and obviously no engineer could work 24 hours a day nor seven days a week. The advantage of [NCML’s] full-time operation was really effective only in the production of parts and the assembly of components.

  In that same report, Meader noted that the Navy ran into early problems with NCR when executives tried to sidestep Deeds and the Navy’s no-profit contract by letting out most of the Navy work to subcontractors. That approach soon ended, however, when the Navy brass wrote directly to Deeds in January 1943 and then summoned NCR’s top executives to Washington for a tongue-lashing on wartime priorities.

  “The contract written with NCR was . . . definitely the wrong thing to do,” Mea
der wrote later.

  The contractors, as their lucrative fixed-price war contracts increased, were very reluctant to use their personnel and facilities for no-profit work at the expense of increasing their profitable work. We reached a stage where 93 percent of all our specially designed parts were let out to subcontractors. Many of these contractors were incapable of executing precision work and their equipment was such that delays which were intolerable resulted. . . . It was necessary to direct the contractor to prosecute whenever possible the work in his own plant under threat of the Navy exercising its war-time prerogative of taking over.

  Sailors who refused orders from civilians “continued to be a bone of contention with the management of the contractor,” Meader wrote in his report. But he fired back with his own complaint against NCR’s employees for not being more grateful that they had been given draft deferments

  solely at the request of the Navy and solely for the purpose of working under this overall project. . . . This situation became critical when certain civilians refused to work overtime even when such overtime was deemed absolutely essential to the successful prosecution of the war. However, nothing actually was done except to handle each case individually and to continually pacify the civilians as much as possible.

  The mix of civilians and military personnel in Building 26 led to some intense encounters at times. Marines armed with shotguns and pistols patrolled the hallways and actually lived in the building, doing their bathing, shaving, and laundering in the restrooms and hanging their uniforms to dry on the roof. The leatherneck guards were hardened cases even by Marine standards—all were veterans who had been wounded early in the war and reassigned to Dayton. Some were prone to skittishness, as Lou Sandor, one of the engineers, discovered. “I had a habit of slamming the door to the restroom whenever I went in,” Sandor recalled. One day, his habit startled an off-duty Marine, who was shaving at a sink. “All of a sudden, I was staring into the barrel of a forty-five. I don’t know who was scared more—me or the Marine who almost shot me.”

  The local draft board, unaware of the vital nature of the work being done at NCR, could be a major nuisance for its personnel. More than once, Desch had to retrieve one of his charges at the Dayton train station, just moments before the engineer or technician was shipped out for military duty. One NCR engineer languished for three days in the local jail on charges of draft dodging before Desch was able to persuade the authorities to release him.

  NO ONE AT NCR, of course, took more orders from the Navy than Desch. When he first became involved in the Bombe project in the summer of 1942, he was told to drop all his other research projects then in progress for the NDRC, the Aberdeen Proving Ground, and the Army. And his personal life came under intense scrutiny for the duration of the war. Even after he had signed a pledge of secrecy the year before as an investigator for the NDRC, binding him “not to disclose any confidential information regarding this research except to others engaged in work on this specific problem,” Desch was asked to undergo a second and more grueling security clearance before he could be told about Ultra.

  In May 1942, he spent three days in Washington under relentless interrogation by Navy inquisitors who hurled insults and accusations, trying to break him down. The questioning eventually became so abusive that Desch told the Navy he didn’t want the job—at which point they relented and told him he was cleared. He was to return to Dayton immediately to begin his new assignment. “Dad told me once he had had it up to here with their bull,” Debbie Anderson said. “But he must have calmed down, because he got the job.”

  With it was to come more stress and more intrusion into his personal life than Desch could have possibly imagined at the time. It meant working fourteen- and sixteen-hour days under mounting pressure from Navy officials and living under twenty-four-hour surveillance by Navy guards, even in his own home.

  “You could see it was having an effect on him,” Kern said. “He would more frequently get mad in meetings. Of course, getting mad was his way of getting things done . . . of increasing your enthusiasm for you to get your work done. We all noticed that he complained more about things.”

  And yet because of the top secret nature of the project, Desch could never really explain to his subordinates the pressures he was facing, or even what the machines they were laboring so mightily to perfect would be used for. “We knew that [the machine] was for a decoding activity, but we just didn’t know the details of it,” Sandor said. Even Desch’s on-the-job vocabulary was restricted. “We never heard of . . . Bombe, Enigma, or NCR Navy [sic] Computing Machine Laboratory during the war,” Kern said.

  Some men might have turned to their families for solace and support, but again Desch was blocked. For the duration of the war, he would be forced to sever relations with most of his German relatives. The Navy was taking no chances with a man who could speak fluent German, who was an avid ham-radio operator and communications expert, and whose mother had been born and reared in Germany.

  Desch’s mother, Augusta Stoermer, had emigrated at age thirteen from Germany to Liverpool, England, then to an uncle’s home in Pittsburgh, where she had rolled cigars for a living, and finally to Dayton, where she met Edward Frank Desch. Homesick even in her later years, Desch’s mother had kept in touch with relatives in Europe and had gone back to visit in Germany before the war. Edward Desch died in 1937.

  Desch was permitted only limited visits with his mother and two younger sisters, providing the entire family avoided all contact with Desch’s half cousin in Dayton, Augusta “Gusty” Burger. Gusty still had a father in Germany who was a member of the Nazi Party, and her husband, Albert Burger, had been known to tune in Hitler’s broadcasts on his shortwave radio before the war. *22 Unable to explain to his family the reason for his infrequent visits and for avoiding Gusty, Desch found his relations with his mother and sisters became strained during the war.

  Throughout the war, Desch was never out of sight of his Navy shadows—plainclothes guards who sat in parked cars outside his home and office, waiting to tail him around town. Years later, Desch gleefully told his daughter what fun it had been one night after work to lead the guards on a wild-goose chase through the countryside south of Dayton, before doubling back to his own driveway.

  Hardest of all for Desch to swallow was having Ralph Meader quartered in his own home to keep an eye on him. Desch had built the two-bedroom Tudor cottage just two years earlier as a gift to his wife, Dorothy, on a small lot at 413 Greenmount Avenue in the charming streetcar suburb of Oakwood. Even by 1940s standards, it was a cozy home for a young couple, but no doubt a bit too cozy with Meader in the spare bedroom across the hallway. Meader “practically slept in his bed,” said Bob Mumma. “I’ll tell you, day and night, he couldn’t get rid of him. Joe just about lost his mind.” Years after the war, and even after his only daughter had taken occupancy of the room, Desch disdainfully referred to “Meader’s room,” never bothering to remove the large oak desk his boss had used there.

  Desch’s loathing of Meader had many possible sources, but certainly one was his Navy-bred style of management. Realizing that Desch was the best man to motivate his own employees, Meader applied the Navy’s pressure most directly on him and expected it to ratchet down the civilian chain of command. He quickly discovered what worked most effectively on Desch: guilt. Meader told Desch more than once that he was going to be responsible for the deaths of countless American boys if he didn’t get the job done.

  Anderson believes Meader never suspected the full impact of his guilt-lashings. Her father was a devout Catholic who had been raised to believe in the concept of mortal sin and the eternal damnation of hell for those who failed to live up to their earthly responsibilities. Some of the men who would die in the war included his former ROTC buddies. Desch himself had been excused from the unit because the Navy had needed his talents at home—all the more reason to blame himself as the death toll rose in the Atlantic. “We were losing ships like mad,” Mumma said. Desch “felt g
uilty, I think, because he wasn’t active. That’s what bothered him.”

  Desch had been commissioned as an officer in an Army ordnance unit, but when the Navy learned it might lose his technical expertise to the front (and worse, perhaps, to the Army), it pulled his orders, Anderson said. Far from bringing relief, the dispensation induced bouts of guilt in Desch for the rest of the war, and the rest of his life. “Now I know why Dad was really conflicted and angry most of his life,” Anderson said, who didn’t learn of her father’s role in the Bombe project until the mid-1990s, almost a decade after his death. “The Joe Desch I knew carried a lot of anger around. I know it wasn’t about me and Mom. And I wasn’t sure it was about NCR, either.”

  IN THE EARLY months of 1943, tension infused the Bombe project from the highest levels down. Desch smoked his Chesterfields, two packs a day, and when the long workday was done, he liked his Scotch and water.

  Despite the Navy’s constant pressures and Desch’s ramrodding of his staff, progress was painfully slow. The design for the pilot model of the Bombe was submitted to the Navy in January. But with spring and an expected U-boat offensive fast approaching, Desch was far from producing a working machine.

  The British, wondering whether to rely upon the American effort because of their own problems developing a four-wheel Bombe, sent Alan Turing to Dayton to check on the NCR work. Turing’s visit was to provide some valuable insights for the Americans, but his recommendations to OP20G would also add to Desch’s mounting burden of stress.