With an audacious scheme that paired Enigma decrypts with a network of turned enemy spies, Britain pitted Germany against itself throughout the war.
In early 1941, an astonishing idea began to dawn on John Masterman, a British official whose fluent knowledge of German, skill at interrogation of enemy prisoners, and more than a little Machiavellian craftiness had landed him the high-wire job of running double agents for Britain’s Secret Intelligence Service. “Dimly, very dimly,” as Masterman would later put it, the evidence was suggesting the seemingly unbelievable possibility that the entire German spy network in Britain consisted of double agents under his direct control. There were no real German agents left.
By July 1942, reports from a variety of inside sources—the most convincing of them the Germans’ own high-level intelligence reports, captured in decoded messages—had persuaded the British that the seemingly impossible was in fact true. And what had begun as a purely defensive effort to safeguard Britain from German espionage and detect Nazi plans was about to turn into one of Britain’s most effective secret weapons of the war.
For the duration of the war, brilliantly orchestrated false reports sent back to Germany by Masterman’s tame agents would lead the German high command into one blunder after another at the most crucial junctures—ensuring the success of the Allied landing at Normandy, helping to turn the tide in the war against U-boats in the Atlantic, even tricking the Germans into firing most of their V-2 rockets short of central London.
At the heart of these brazen ploys was what is today a well-known success story of the war: the Allies’ breaking of coded messages enciphered using the Germans’ supersecret Enigma machine. From 1940 on, thanks to an extensive harnessing of mathematical and scientific talent, British code breakers were regularly reading the orders and messages of the highest levels of the German command. The Enigma decrypts provided an incomparable window into German military plans and thinking, allowing the Allies to reroute Atlantic convoys around lurking enemy U-boats, catch and sink the German battleship Bismarck, and lay minefields and position artillery exactly where Field Marshal Erwin Rommel planned to attack during the tank battles that raged in the Egyptian desert.
But what Masterman realized was that with some fiendishly clever thinking, the Enigma decrypts could be something much more than just a source of intelligence about enemy plans: they were the key to manipulating the enemy. To put it simply, in providing a window into the Germans’ minds, the Enigma messages could also show the best way to mess with their minds. And as the man running the double agents who reported back to the Nazi spymasters, Masterman was just the right person to do that messing.
Masterman was nearly a neophyte as an intelligence officer, the kind of innovative outsider that the Second World War routinely thrust into unusual positions of responsibility. He had spent most of World War I interned in Germany as an enemy alien, which led to his fluency in the language. Unquestionably an intellectual— he was an outstanding teacher and historian at Oxford when he was drafted into the British Army’s Intelligence Corps in March 1940 at the age of 49—he was also the author of a murder mystery set in Oxford, as well as a Wimbledon tennis player, member of the Olympic field hockey team, noted cricket player, and a man known as “cool and calculating” under all manner of circumstances.
As he later recalled, at the start of the war the conventional view of double agents—spies who pretend to work for the enemy as spies—was that their main value was in counterespionage: penetrating the enemy’s secret service, learning about its methods of operation, and discovering its intentions. Several of the first double agents to come under British control had been playing both sides, on their own initiative, for several years before the war; they were of a type, Masterman explained, “who have a natural predilection to live in that curious world of…deceit, and who attach themselves with equal facility to one side or the other, so long as their craving for adventure of a rather macabre type is satisfied.”
One of the first of these was Arthur Owens, code-named “Snow,” a Welsh-born Canadian electrical engineer who had returned to Britain in the 1930s and began doing business in Germany. He was soon doing jobs for both the British Secret Intelligence Service and its German counterpart, the Abwehr.
His motivation was never entirely clear, but it seemed to be less ideological than mercenary, or rather libidinous: the Germans paid him off mostly by arranging assignations for him in Hamburg with attractive women. That may also have been his initial downfall. A week before the war broke out, Snow’s wife denounced him to the British police as a German agent, and he was arrested. Under interrogation he gave British intelligence authorities a reasonably full confession of his dalliances with the Abwehr and readily agreed to save his skin by accepting his interrogators’ proposition that he continue to work as a double agent under British control. Reunited in his prison cell with the radio transmitter his Abwehr handlers had supplied him, Owens made contact with Hamburg (tapping out his Morse code messages under the watchful eye of a warder who was an amateur radio operator), and soon was receiving orders and intelligence questionnaires back from the Abwehr. As Masterman, ever the Oxford don, would later write, Snow was “the fons et origo”—the source and origin—“of all of our activities for the next five years.”
To run the double agents, Masterman was placed in charge of a group given the somewhat arch name “Twenty Committee”: the number 20 in Roman numerals being “XX”—double cross. As the committee’s initial instructions explained, its job was “to keep our agents sufficiently well fed with accurate information so as not to lose the confidence of the enemy”; “to control as many of the agents in this country as we can, in order to make [the enemy] feel that the ground is covered and they need not send any more of whose arrival we might not be aware”; and, finally, “by careful maneouvering of these agents and a careful study of the questionnaires [submitted to them by their Abwehr handlers], to mislead the enemy on a big scale at the appropriate moment.”
Its mission, in other words, was to lull the Abwehr into complacency while preparing for the big sting.
From Snow’s radio contacts with his Abwehr controllers, the double-cross operation quickly discovered the characteristics of the radio signals the Abwehr was using to communicate with its agents from its base station in Hamburg—frequencies, times of day, and message headings. This allowed British code breakers to identify other radio traffic between the Abwehr and its spies throughout Europe. By April 1940 the code breakers at Britain’s secret establishment at Bletchley Park were beginning to read Abwehr traffic that was enciphered using various relatively simple pencil-and-paper code systems distributed to German agents.
Throughout 1941, broken Abwehr messages were regularly alerting the Twenty Committee to the impending arrival of new agents in Britain. Some were dropped by parachute; others came ashore on rubber rafts launched from U-boats; others, recruited from occupied countries such as Norway and Poland, blended into the stream of refugees making their way from neutral Spain and Portugal. Of 23 agents who were sent by the Germans into Britain throughout 1941, 7 were identified and captured as a direct result of deciphered radio messages. Others were simply caught, or turned themselves in.
Not every captured agent was a suitable candidate for being “turned.” For one thing, as Masterman noted, “it would have taxed even German credulity if all their agents” had arrived safely, evaded capture, and gone to work efficiently filing espionage reports. For another, it was vital that a spy be apprehended almost immediately after his arrival to be sure he hadn’t already communicated with Germany and possibly warned of his impending capture, and (especially in the case of the parachutists) that no one—or only a very few trustworthy people— had witnessed his landing and apprehension.
The other big problem that limited the usefulness of the double agents was the constant, nagging fear that any cooked-up reports sent back by the controlled agents would be contradicted by other German agents who had slipped in undetected, possibly endangering the whole scheme.
And one huge hole remained in the entire system: it was clear throughout 1940 and 1941 that a growing volume of the Abwehr’s radio signals were being encoded not by the easily broken handciphers, but with some variant of the much more secure Enigma cipher. At the time, its solution remained elusive.
The Bletchley Park code breakers had begun cracking at least some German army, navy, and air force Enigma signals in early 1940. The major Enigma effort at Bletchley relied on a highly complex mathematical analysis developed by the brilliant mathematician Alan Turing, the centerpiece of which was the use of a behemoth electromechanical calculator Turing had devised which could deduce the Enigma’s daily setting for coded signals.
The Enigma machine consisted of a typewriter keyboard and a corresponding panel of 26 small lamps, each marked with a letter of the alphabet. To encode a message, a cipher clerk would type it in and write down the letters that illuminated as each key was depressed. Decoding worked the same way—the enciphered message would be typed in and the plain text would be indicated, letter by letter, by the lamps. At the heart of the machine were three wired rotors that scrambled the electrical connections that ran between the keys and the lamps. Every time a key was depressed, the rotors would advance one position, roughly like a car’s odometer, so that the scrambling pattern changed with each successive letter of the message. The starting position of the three rotors was changed each day according to a list circulated to the users; discovering that start setting was the crucial step for the code breakers. That was what Turing’s calculators—dubbed “bombes”—were designed to do.
But the Abwehr networks seemed to be using a different Enigma machine that defied the same method of attack. Initial analysis revealed that in one way this variant was much simpler: it lacked a series of plugs and jacks that had added a nightmarish complication to the code breakers’ task, and which was the main reason Turing’s mechanical bombes were needed to solve the daily settings. However, it was also built so that the rotors moved in a much more complex pattern than the models used by the German military.
Hundreds, and eventually thousands, of men and women at Bletchley Park and in Washington would be part of an almost industrial-scale operation that solved German Enigma traffic. The job of breaking the Abwehr Enigma was given to one man: Alfred Dillwyn Knox, known as “Dilly.”
Knox was a code breaker of the old school, more linguist than mathematician, more intuitive than scientific in his approach. The son of an Anglican bishop, Knox was almost a parody of the eccentric, irascible British scholar. He had spent seven years before the First World War piecing together the shattered fragments of a single Greek papyrus scroll. Then he had been recruited to work in the navy’s code-breaking group during World War I, and had stayed on in the tiny British government code-breaking establishment that survived the end of the war.
Knox had since developed a reputation for blazing brilliance— and intense professional jealousy and secrecy about his work, which he hated to share with anyone. As one of his (very few) assistants put it, Knox “disliked most of the men with whom he came in contact.” One assistant assigned to him recalled that Knox had given him some menial tasks for a few weeks, then subjected him to “some sort of test and appeared to be, if anything, annoyed that I passed.” One of Knox’s favorite questions for hazing new recruits was an Alice in Wonderland sort of riddle: “Which way does a clock go round?” Anyone foolish enough to answer “clockwise” would receive the withering retort, “Not if you’re the clock it doesn’t!”
Knox had been working on the Enigma problem since 1936, when the British first became aware of the Germans’ use of the machine. But the arrival of the first of the Turing bombes at Bletchley Park in March 1940 signaled the arrival of a new era that had little use for old-school cryptographers like Knox. Yet it would be Knox who would have the last laugh on the Abwehr problem; by the end of 1941, without ever actually seeing one of the machines, working only by pencil and paper, a huge amount of trial and error, and some inspired guesswork, he had surmised the wiring of the rotors and the way the rotors moved together, and had devised a method for extracting the day’s starting position. It was a tour de force of deductive reasoning, based on spotting subtle patterns in the coded signals.
But Knox was also helped by the continuing payoff of the double-cross system itself, for it became apparent that many of the Enigma decrypts consisted of messages from Abwehr stations to headquarters in Germany, repeating reports received from agents in the field. On more than one occasion, knowing what had been filed by a double agent in his hand-cipher paid off by providing uncoded text that could be matched with a subsequently intercepted Enigma transmission.
Over the next three and a half years, 140,000 Abwehr Enigma messages would be decrypted by Knox’s section and sent to Masterman to be exploited in the double-cross operation.
Dilly Knox’s Enigma breakthrough paid off almost immediately, delivering to the Twenty Committee its single greatest double agent of the war.
In February 1942, British intelligence officers found themselves puzzling over a series of agent reports “of superb inaccuracy” that began appearing in Enigma messages being sent to Berlin from the Abwehr station in Madrid. They purported to be from an agent in Britain, and dealt mainly with merchant shipping convoys from the British Isles to the Mediterranean. The only trouble was that none of the reported convoys corresponded with actual shipping movements.
There were some outright howlers in the agent’s reports, too. A dispatch supposedly sent from Glasgow explained his success in prizing out information from the locals by noting that “there are men here who will do anything for a liter of wine”—which would have been news indeed to anyone familiar with the drinking habits of Scotsmen in the 1940s.
The startling fact, though, was that these reports not only seemed to be taken with 100 percent seriousness by the spy’s Abwehr masters; they also aligned precisely with a seemingly fantastic story that had been told to a British intelligence officer in Spain a few months earlier by a Spaniard named Juan Pujol Garcia, who had approached him offering to become a British spy.
Garcia had first tried to become a British agent right after the war broke out and was rebuffed. He tried again in the fall of 1941, this time explaining that after he had been turned down, he had hit on the idea that maybe the British would find him more valuable if he first established himself with the Abwehr as a German agent, and then offered to double-cross them. That, he insisted, was exactly what he was doing.
Garcia’s story was this: After using some faked documents to persuade the Abwehr station in Madrid that he could infiltrate his way into Britain, he had traveled to Lisbon. There, armed with nothing more than a tourist guide of Britain, a Portuguese publication on the British fleet, and whatever technical journals he could find in the public library, he had invented a series of subagents and a raft of imaginative reports that he duly mailed to his Abwehr handlers in Madrid. To explain their Lisbon postmarks, he told his Abwehr control that he had recruited as a courier an airline employee who had agreed to take his dispatches from England and drop them in the post in Portugal during his regular flights there.
It all sounded ludicrous to the British, and they had again sent him packing. But the Enigma decrypts now changed everything. They confirmed every key point of Garcia’s story. Moreover, they showed, incredibly, that the Germans had complete, unquestioning confidence in him. On April 2, 1942, Bletchley Park broke a signal on the Abwehr Enigma circuit relaying Garcia’s latest report of a nonexistent convoy, about to sail from Liverpool to Malta. Almost immediately, a flurry of orders went out on naval Enigma circuits directing German units to intercept the convoy. Now fully convinced of his value as a double agent, the British smuggled Garcia into England on April 24. With the code name “Garbo,” he would proceed to tie the highest levels of the German command in knots on more than one occasion.
Under Masterman’s direction, Garbo continued to expand his network of imaginary subagents (and sub-subagents). Soon there were 30 of them, almost as many as the total number of real double agents that the British would run in the entire course of the war. Their identities were carefully crafted to be plausible to the Germans, though they were as colorful a collection of characters as the network invented by the hero of Graham Greene’s novel Our Man in Havana. Garbo’s “recruits” included a garrulous Royal Air Force officer, a Ministry of Information official with extreme left-wing views, a Venezuelan businessman in Glasgow, a communist Greek sailor in eastern Scotland, a Gibraltese waiter in a service canteen, an Anglophobic American sergeant, an Indian poet in Brighton—all of them complete fabrications. Their fictitious reports ran the gamut from the morale of British troops in North Africa to a supposedly large and fanatical pro-fascist fifth column within Britain, to the existence of a vast underground network of tunnels in the London area used to supply ammunition to airfields and antiaircraft batteries around the capital.
The Enigma decrypts confirmed that the Abwehr was eating all of it up. They also provided a constant check on whether the other turned agents were still playing their part—and not trying to switch sides yet again. And they provided an ever-more-solid reassurance that every new German attempt to penetrate Britain would be known to the British ahead of time.
The Twenty Committee’s almost giddy success created its own problems, some deadly serious, some almost comical. One serious problem was that Churchill felt that more of the captured spies should be shot, as an example to others. An intense bureaucratic wrangle ensued. The political authorities suggested that “any spy or enemy agent whom we no longer require” should be put on trial. The intelligence agencies indignantly replied that “intelligence should have precedence over blood-letting,” and that any public trial would risk exposing the double-cross operations and undermining their ability to recruit double agents by promising them they could save their lives by cooperating. Nine spies were executed from December 1940 to December 1941, before cooler heads got the upper hand.
One of the comical problems was the discovery that the radio gear the Abwehr had supplied to its agents was incapable of reliably reaching Germany. British technical experts were called in to carefully tune up the equipment and install antennas in high, unobstructed locations to improve communications, which helped some. But when it came time for the pièce de résistance—a series of reports from Garbo that would prove crucial to cementing the massive Allied deception effort in the run-up to D-Day—they decided they could take no chances that his transmissions would not be received by the Germans, and substituted his 3-watt German-supplied set with a state-of-the-art 600-watt American-made military transmitter, which was eventually installed atop the London headquarters of the British counterintelligence service.
Another absurd snag occurred when the Twenty Committee hatched a scheme to deliberately blow one of their own double agents. The idea was to have one well-established agent make some obvious gaffes that would give him away as a double agent; the Germans would accordingly form a low opinion of the sophistication of the British at running this kind of operation, thereby deflecting suspicion from the remaining double agents. The only problem was that no matter how hard they tried, they could not make the Germans doubt “their” man.
And it kept happening. The ideal deception, which the Twenty Committee managed to pull off a few times, left the Germans never knowing they had been fooled. More realistically, the committee was prepared to sacrifice an agent’s future usefulness for one big sting. Yet even when it should have seemed glaringly obvious that they had been the victims of a ploy, the Abwehr would never believe they had been outwitted.
Repeatedly, the Twenty Committee cooked up reports of an impending invasion of Norway from Britain, fully expecting that the agent who delivered the information would be blown once the invasion never materialized; repeatedly the German high command responded by redeploying forces against the threat; and repeatedly the Abwehr explained away their agent’s mistake—he had been innocently taken in by a well-crafted British cover plan, or the British had for some reason decided to drop the plan at the last minute, or he had been misled or had exaggerated what he had actually seen. “In short,” reported Masterman, “it was extremely, almost fantastically, difficult to ‘blow’ a well-established agent.”
One of the masterminds of the Allied deception strategy, Lt. Col. Dudley Clarke, observed in the summer of 1942 that there were two core principles for succeeding at this shadowy game. First, it was essential to know what the enemy was already inclined to believe; the most effective deceptions were built on a preexisting “foundation of fear.” Second, the purpose of deception was not just to make the enemy think something, but make him do something that would help the Allies or hurt the Axis.
On both counts, Enigma was a tool of incomparable value— in laying bare what the Germans were inclined to believe and how they were likely to react.
In the lead-up to D-Day, a huge deception effort had been launched to make the Germans believe that the attack would come at Calais rather than Normandy. Now the Twenty Committee, and Garbo, were to put the icing on the cake. In the early morning hours of June 6, 1944, Garbo sent an urgent radio message to his Abwehr control: the invasion was imminent. The timing was masterful—arriving just too late to be helpful to the German defenders but early enough to send Garbo’s already high stock soaring with German intelligence.
Then three days later came the masterstroke the entire operation had been building up to. Garbo, having “met” again with his network of agents, reported urgently that the Normandy landing was merely a feint: the real attack was still to come at Calais. Immediately orders began to fly from the German high command. Two panzer divisions were ordered to head for Calais; the 85th Infantry Division, which was already on its way to the Normandy beachhead, was recalled. For two full weeks, seven German divisions that had been expected to be sent to the Normandy landing areas were held back from the fight.
In late June, Garbo was informed by his German masters that he was being awarded the Iron Cross for his heroic work. (Incredibly, they apparently never did suspect that Garbo was a plant: even after the war, German generals concluded that the Allies had canceled the “real” landing at Calais only because of the unexpected success of the “feint” at Normandy.)
In the Battle of the Atlantic, the double-cross effort scored a series of coups. In the spring of 1943, Enigma intercepts revealed that Adm. Karl Dönitz, commander of the German U-boat force, suspected that his mounting losses were due to the use by the British of some new infrared detection device. The Twenty Committee immediately fanned those fears with double-agent reports that confirmed Dönitz’s supposition. In fact, the British were using a new secret microwave radar—and as it happened, the infrared-masking paint that Dönitz ordered applied to the U-boats’ conning towers actually increased the boats’ visibility to the radar.
In 1945, when the Germans began equipping their U-boats with the schnorkel—a breathing tube that allowed them to run the diesel engines underwater, thereby recharging their electric batteries without surfacing, permitting the U-boats to remain hidden for extended periods—the double-cross system came through with an urgent report that 3,600 square miles of water along the western approaches to Britain had been heavily seeded with antisub mines. In fact, the British were drastically short of mines at the time; but again the Germans fell for it hook, line, and sinker and ordered all of their subs out of the area.
As the Germans’ V-weapons began to rain down on London in the final year of the war, the double-cross system scored one final coup. The Abwehr sent its agents urgent requests to report the time and location of the weapons’ impact. Masterman’s committee saw an opportunity at once, and after consulting with British scientific experts, began ever-so-artfully skewing the data the agents sent back to make the Germans believe they were overshooting their target of central London. In response, the Germans kept shortening the range of the weapons, and the mean impact point of the V-2s began moving eastward at a rate of two miles a week. By mid-February 1945, most of the rockets were falling well outside metropolitan London.
All in all a remarkable body of achievement for what the intelligence historian Nigel West termed a “relatively small group of amateur security officers,” who found themselves in charge of one of the most daring operations in the history of secret warfare. But as Masterman himself noted, the greatest credit had to go to Dilly Knox—who died of cancer in 1943— and his fellow Bletchley Park code breakers. “None of this would have been possible” without the Enigma decrypts, concluded Masterman in his final report as chairman of the Twenty Committee in May 1945. “It was vital to double-cross work.”
Originally published in the May 2009 issue of World War II. To subscribe, click here.