The Secret in Building 26

by Jim DeBrosse

  Soon after the war, Desch learned that nearly every member of the Army ordnance unit in which he had been commissioned, the 512th, had been either killed or wounded in action. In his darkest moments, Anderson said, her father used to say, “I would much rather have been with them.” Desch had known many of the men in the unit from his twelve years in the Army Reserve Corps. But in October 1940, he had been forced to resign his commission in order to devote himself to his work for the National Defense Research Committee, a duty to which he had been recruited by Vannevar Bush. Desch’s letter of resignation to the corps showed his deep ambivalence about leaving the Army for civilian service. In the letter, he spoke of his pride at having served as an officer for twelve years. But despite Desch’s feelings of personal loss, he was convinced that the resignation would allow him to perform a more valuable service to his country as a researcher.

  At his cottage that rain-soaked summer night in 1945, Desch had stayed in the background while the WAVES and their young friends celebrated the final victory of the war. Vogel said she didn’t give much thought that evening to what might have been weighing on Desch’s mind. All that mattered to her and to most of the WAVES was that the fighting had finally ended and that they would soon be going home to their loved ones. “We danced on the front lawn in the rain until midnight,” she said. “And, yes, we did forget our manners.”

  After the war, Meader commanded fewer than one hundred WAVES and Navy engineers to carry on what they knew was top secret and vital work at Building 26, all of which is still classified. Finally closing shop and moving to Washington, D.C., in 1946, Meader left his beat-up Nash Rambler in Desch’s driveway. Desch had to send several terse letters to naval authorities before they disposed of it.

  But the postwar fate of America’s frontline anti-Enigma machines—the 120 Bombes built at NCR—is less clear.

  Epilogue:

  Burying the Past

  BY V-J DAY, most of the Bombes in Washington were already being retired and prepared for a secret burial at sea. Whether any Bombes were left at NCR and what happened to them remains a deeper mystery. Most were smashed with sledgehammers and axes. The remaining Bombe parts and the machine tools for making them were removed from Building 26 under cover of night and buried deep below a parking lot, perhaps across West Stewart Street, the busy four-lane road just north of Building 26.

  Officials in the National Security Agency, the organizational heir to OP20G and the Army’s codebreaking operations, say all but one of the working machines that reached Washington were destroyed. But Phil Bochicchio, the Navy technician who had been in charge of setting up and maintaining the machines in Washington, said he was ordered to ship thirty Bombes to a Navy warehouse in Mechanicsburg, Pennsylvania, soon after the war.

  We know that many secret German transmissions were decrypted after the fighting ceased, as shown by dates on messages. That suggests that some machines were held in reserve. The British have let it slip that they kept sixteen of the Bombes they built during the war.

  Both Allies had good reason to retain at least some of the Bombes: many European nations, including the Soviet Union and eastern-bloc states, confiscated thousands of Enigma machines from the Nazis after the war and likely put them to use. That raises the question of whether the Soviets had been privy to the Ultra secret during World War II. At least one of the WAVES, Mary Lorraine Johnson, who later became a Utah state legislator, recalled Russian generals being shown the Bombes during a tour of OP20G toward the end of the war. If so, the Soviets may have used America’s expertise to spy on its eastern-bloc satellites. However, no Navy officials or government documents have yet corroborated Johnson’s story.

  NSA officials insist they know of no other existing Bombes except the one now on display behind a glass case at the agency’s National Cryptologic Museum at Fort Meade. The machines sent to Mechanicsburg “were probably just excessed and destroyed,” said Jack Ingram, curator of the museum. “We don’t have any records on it.”

  Parts of the NCR machines may still be classified. If true, NSA officials won’t say why, but it could be that some of the same methods were used against the Soviets or their allies during the cold war. Likewise, NCR’s defense role after the war remains unclear. “The Cash” had had its fill of unprofitable military work during the war and, in its aftermath, was faced with hundreds of thousands of back orders from a global clientele impatient to return to peacetime commerce. Even before the end of the war, Wenger and Engstrom had known that NCR and other industrial firms were eager to drop their government work as they geared up to meet the nation’s long-suppressed consumer demands. As early as September 1944, NCR management had told the Navy in very direct terms that it wanted no more open-ended demands, no more hard-to-manage military personnel, and no more essentially profitless contracts. Negotiations and calls on patriotism smoothed over the situation, and before the war’s end NCR did accept two more large contracts with the Navy. But discontent at NCR remained and, just days after Japan’s surrender, executives told the Navy to get its men and women out of NCR’s factory.

  Even so, NCR continued to be a contractor for the government after the war, although on a much-reduced scale, while Desch retained a seat on the scientific advisory board of the NSA until at least 1961, when his correspondence with the agency came to an end. No documents have ever been released that confirm NCR’s cold-war involvement. Patrick Wheadon, an NSA public-affairs officer, said, “We can’t comment. The closer you get to present day, the less we can discuss.”

  Rumors abound as to the whereabouts of Adam and Eve and Cain and Abel, the first four prototype machines, which were never sent to Washington. Several engineers who worked on the project say the prototypes were dumped from a railroad overpass into an old canal bed not far from Building 26, where they remain buried on land now occupied by the Montgomery County Fairgrounds. Others involved in the project say the machines lie underneath a parking lot at Building 26, which has undergone numerous expansions and renovations since the war. Desch himself hinted to his daughter that his war work was buried deep in the water table below the pavement of Stewart Street.

  There are no known records to support any of these claims.

  WHEN THE WAR had at last ended, Joe Desch was ready to bury it, too, in his past. He had made firm decisions about his future, and it didn’t involve the U.S. Navy. He told Howard Engstrom and Joseph Wenger that he would not join the new company they wanted to establish to fill the gap left by NCR’s withdrawal from top secret codebreaking work—even though the idea had been hatched over cocktails in Desch’s living room in late 1944. He likewise declined an offer to move to Washington to become the director of research at OP20G.

  In the years after the war, Wenger and the new breed of OP20G engineers and codebreakers continued to guide Navy communications intelligence and the U.S. military toward a totally electronic era in cryptanalysis. As early as 1943, they were planning for the day when they could move beyond the crisis design and engineering that had dominated and constrained their work to the very last days of the war. And they vowed to take steps so that America’s military would never again be caught by surprise, as they were so tragically in the early-morning hours of December 7, 1941. Nor, Wenger vowed, would OP20G ever find itself desperately behind the technological curve, as it had in its battle against the Enigma. He knew all too well that much had been left unrealized during the war, including the dream, perhaps still unrealized, of codebreaking that relied neither on cribs nor enemy error.

  In contrast to Desch, Wenger faced many uncertainties in the summer of 1945. He wanted OP20G to have a permanent research group such as the M section Howard Engstrom had built during the war, and he wanted OP20G to have the engineers and industrial resources that Ralph Meader had commanded in Building 26. He knew that most of his highly valued researchers wanted to return to their universities and corporations and that those who might stay with him would probably be rejected for Navy service because of their age, educat
ion, and high-paying rank.

  Wenger had nowhere to turn for help. NCR’s withdrawal from defense work left OP20G without an industrial partner willing and able to do secret work. No other major corporation or even any other Navy agency, such as the Naval Research Laboratory, seemed willing to concentrate on OP20G’s mission. And Wenger feared that his request for higher-paid civilian experts within OP20G would not be approved. In fact, he could not be sure that OP20G itself would exist much longer. The Army, in alliance with the White House, was pushing hard for a unification of all the intelligence services, and “G” might well disappear into a new intelligence group where the Navy played a subordinate role.

  If nothing else, Wenger wanted Navy codebreakers to get a share of the diplomatic codebreaking after the war—a role that had traditionally been left to the Army’s cryptanalytic staff. Despite the success of its Bombes, OP20G was still chafing under its dependence on the British for analyzing a variety of European radio traffic and was looking ahead to the days when it could focus on Russian codes. The Navy already was intercepting Soviet traffic at the time and had, by 1945, an agreement with Britain to share Soviet intelligence and processing capabilities. It was none too soon: by 1948, America and Britain would discover that the Soviets were far more adept at code warfare than the Germans.

  Despite frictions over Britain’s colonialist policies in other parts of the world, it was agreed that the Ultra partnership should continue after the end of the war. Liaison groups were established, and sharing of intercepts, methods, and technologies deepened as the transatlantic alliance became permanent. Although postwar Britain was financially bankrupt, it provided America valuable expertise, experience, and facilities.

  Finding corporate partners for the Navy’s postwar codebreaking ventures would prove far more difficult. By February 1945, the idea of creating a new independent company for OP20G’s work had turned into something like a business plan. Knowing that the Navy would be unlikely to keep even the core of the mathematical and engineering groups he had put together during the war, Engstrom envisioned a new company that would hire his researchers and allow them to devote their talents to advanced scientific and secret projects.

  The plan was circulated through the Navy hierarchy, and soon it was passed on to some of the nation’s most influential bankers and industrialists in hopes that funds would be found for the new company, tentatively titled the National Electronics Laboratory. Engstrom and Meader wrote to banks and investment houses and even approached the nonprofit Rockefeller Foundation. Every response came back no.

  But soon after the war, the intertwinings of what would someday be called the military-industrial complex came into play. Through a mutual acquaintance, Ralph Meader was introduced to Nelson Talbott, who was part of Dayton’s old aviation elite going back to the Wright brothers and Colonel Deeds. As president of Trans World Airlines and brother to a future secretary of the Air Force, Talbott was a well-connected businessman who had done much planning and contracting for the Army Air Forces during the war and perhaps even lent the family estate for top secret Air Force research. Talbott, in turn, introduced the Meader-Engstrom group to an investor who seemed ideal, John Parker.

  Parker was not only a reputable investment banker but a graduate of the U.S. Naval Academy with an interest in computing devices and experience in how to deal with government agencies. During the war, he had built a glider factory and was looking for a way to put his now-idle facilities to use. By the end of 1945 it seemed that the renamed Engineering Research Associates (ERA) might have a future.

  In the meantime, Joe Wenger had received some good news. OP20G’s postwar budget, though short of his request, was better than he had expected, with authorization to create slots for civilian researchers. With that, Wenger was able to persuade some of the younger men in Engstrom’s research and engineering group to remain at OP20G. Among them were men who became the nucleus of the nation’s postwar codebreaking efforts: Joe Eachus, Howard Campaigne, Louis Tordella, and James Pendergrass.

  Wenger also received a go-ahead for his postwar computer research plans. While the funding was not as great as during the war, it was appreciable. The 1947 budget included $800,000 to acquire new codebreaking equipment and another $1.2 million for machine development, $500,000 of which had been left over from the previous year. Still, all together, those funds could have paid for only one third of the Bombes built at NCR during the war.

  DESCH TURNED DOWN nonmilitary opportunities after the war as well, including an offer from Xerox to be its chief engineer. “I was worn out,” he said years later in explaining his decision to Carl Rench, his young protégé and confidant who eventually became NCR’s vice president for research. Instead, Desch chose to hitch his fate to NCR and his hometown of Dayton.

  In his research after the war, Desch returned to the passion that had originally brought him to NCR in 1938: gas-filled tubes. Even though transistors and solid-state technology were to replace the unpredictable and energy-hungry tubes as the basis of modern computing, those developments were still a decade away. In the meantime, Desch reported each day to his own private laboratory in Building 20, climbing a steep, back stairway into a secluded loft. There, under a dense tangle of glass piping and beakers that towered above his workbench almost to the ceiling, Desch was allowed to experiment for the next twenty-five years with what eventually became an outdated technology. His private lab was closed and the room sealed off by bricks in 1973.

  Despite his obsession with electronic tubes, Desch did not become a technological throwback. He kept abreast of the more promising developments that were leading to the modern computer and continued to fret over NCR’s hesitant entry into that business. By the 1950s, Desch had regained enough of his earlier drive to head the company’s advanced research effort into small computer systems, though it was a low priority for the company at the time. By then, his style of management had softened, Rench said. “He perhaps felt that he had had to drive his people too hard during the war and, therefore, he was trying to be more rational in his later life.”

  Desch soon became head of NCR’s military project and research division. Though NCR named him a vice president, much of his advice in the 1950s about the future of computers fell on deaf executive ears. Then–NCR chief Stanley Allyn refused Desch’s request to launch a subsidiary of the company devoted to electronic machines and computers. It was a refusal NCR was to regret for years to come.

  NCR missed its chance to get a jump on the digital age. By pulling back from its commitment to cutting-edge electronic research and ignoring the rise of data processing and programmable computing, it continued an errant course that some say began as far back as 1914 when one of its forward-thinking young executives, Thomas Watson, left the company after a personal dispute with John H. Patterson. With a then-generous severance pay of fifty thousand dollars from NCR and the moral support of an attractive socialite wife whose introduction he owed to Patterson, Watson started International Business Machines. IBM, already a leader in electromechanics, rose to world dominance in the electronic computer field during the 1960s and 1970s and plagued NCR as one of its chief competitors into the next millennium.

  NCR went back to business as usual after the war, still clinging to its heavy-handed centralized organization and its mechanical line of products, even into the early 1970s. Labor strife, high costs, outmoded technology, and a sales force entrenched in pitching the old line of business machines eventually led the company to financial disaster. After posting nearly seventy million dollars in losses in 1972, NCR called home William S. Anderson of its Japan division. As the new president, Anderson began the painful and necessary process of decentralizing and downsizing the company. In the space of two years, NCR’s workforce in Dayton declined from 22,000 employees to a mere 5,500, with the bulk of its manufacturing moved overseas and to antiunion states. NCR and Dayton would never be the same.

  LIKE DESCH, MANY of the engineers who had been recruited from the major corpo
rations to work on military codebreaking projects during the war became part of a “crypto-reserve” for America’s top secret agencies, especially in the ensuing cold-war struggle with the Soviet Union. Joe Eachus, John Howard, Howard Engstrom, and William Norris, a Navy pioneer in radio-direction finding, became prominent figures in the computing industry. Andrew Gleason and other academic mathematicians who became major names in their fields continued to advise OP20G and, later, the NSA. Engstrom even returned to the NSA for a few years in the 1950s to head its research and development program.

  Wenger had valued Desch’s work during the war and continued to seek his advice throughout the 1950s and 1960s. Wenger had pushed for recognition of Desch’s contribution to the war, leading to an honor with an ironic twist. In a secret ceremony at the Navy Department in 1947, Desch was awarded the National Medal of Merit—the highest civilian honor for wartime service—for his work in developing the Bombe. But Desch could tell no one of the nation’s gratitude, not even his daughter, Debbie, who for much of her life puzzled over the significance of her father’s award and why he had earned it. He seems never to have complained, either, that he had to pay his own way to Washington for the ceremony. The medal hung in the study of his home, without an explanation, until his death in 1987, when it was packed up along with the many other mysterious artifacts of his life and moved into the family basement.

  Two years later, Debbie Anderson went searching through her father’s things to help her ten-year-old son with an English assignment that he wanted to write about his grandfather. She happened upon two thick transcripts she had retrieved from the piles of papers on her father’s desk after his death but had never bothered to read. They were Desch’s interviews with Henry Tropp, an historian from the Smithsonian Institution, dated January 17 and 18, 1973.