This in itself was a major contribution towards the success of D-Day. But it was not the only one to be made by Bletchley Park, which played a key role in one of the most remarkable espionage operations of all time, the Double Cross system. The main credit for this system is normally given to MI5 and to a lesser extent MI6. But the role played by the codebreakers, and particularly by Dilly Knox, was absolutely crucial.
The Double Cross system originated from an MI5 plan based on an operation carried out by the French Deuxième Bureau. Dick White, a future head of both MI5 and MI6, suggested that captured agents of the German military intelligence service, the Abwehr, should be left in place and ‘turned’ to work as double agents for British intelligence. MI5 would be able to keep complete control over all German espionage activities in Britain and, as a welcome side-effect, the information the agents asked for would tell the British what the Abwehr did and did not know. At this early stage, this was the full extent of MI5’s ambitions.
One of the earliest opportunities to turn a German agent came with the arrest of Arthur Owens, a Welsh businessman who travelled frequently to Germany and who had volunteered in 1936 to collect intelligence for MI6. But the intelligence he provided was of little use and he was soon dropped. He subsequently got back in touch with MI6 to inform them that he had managed to get himself recruited as an agent by the Abwehr, claiming to have done so in order to penetrate the German intelligence service on behalf of the British. But interception of his written correspondence with his German controller threw doubt on this, suggesting that he was playing the two services off against each other. In September 1938, he announced to MI5 that he had now been appointed the Abwehr’s chief agent in Britain and that he had been given a special German secret service code with which to encode his messages. On the outbreak of war, he was arrested at Waterloo station and agreed to work as a double agent under the cover name of Snow. His controller was Lieutenant-Colonel Tommy ‘Tar’ Robertson of MI5, a remarkable man who was to become the key British figure in the Double Cross system, setting up an MI5 section (B1A) to run them.
Snow had been given a radio transmitter by the Germans in January 1939, handing it over to MI5 immediately. He had also been given a very primitive cipher with which to contact the Germans (see Appendix I). This was used to send Snow’s ‘reports’ to his German controller, and was also sent to Bletchley Park for evaluation. An alert MI5 officer, monitoring Snow’s frequencies to ensure that he sent exactly what he was told, noticed that the control station appeared to be working to a number of other stations. GC&CS was sent copies of the messages the station was transmitting, which were in a different cipher to that given to Snow. But the codebreaker who looked at them expressed ‘considerable disbelief’ that they were of any importance, suggesting that they might be Russian telegrams originating from Shanghai.
Despite the scepticism displayed by Bletchley Park, the Abwehr radio nets were monitored by the Radio Security Service (RSS), which was run by Major E. W. Gill, a former member of the British Army’s signals intelligence organization in the First World War, and was based close to Bletchley at Hanslope Park. It had the services of Post Office intercept operators, plus a small army of volunteers, most of them radio ‘hams’, who scanned the shortwave frequencies looking for enemy wireless traffic.
Gill and a colleague, Captain Hugh Trevor-Roper (later Lord Dacre), who worked in the Radio Intelligence Service, the analysis section of the RSS based at Barnet, north London, broke one of the ciphers in use. They managed to show that the other messages on Snow’s allotted frequency were indeed Abwehr traffic. This appears to have been the source of considerable embarrassment at Bletchley and the row over the significance of the traffic went on for some weeks with Trevor-Roper becoming increasingly unpopular with the professional codebreakers. Eventually, a new section was set up at Bletchley, in Elmer’s School, to decipher the various messages on the network. It was headed by Oliver Strachey. The Abwehr’s cipher instructions given to Snow led to a number of ciphers being broken and the first decrypt was issued on 14 April 1940. Initially codenamed Pear, the decrypts became known as ISOS, standing either for Illicit (or Intelligence) Services (Oliver Strachey).
The ISOS decrypts enabled MI5 to keep track of the messages of the double agents and spot any other German spies arriving in the country. It also meant that the agents’ reports could be designed to allow the codebreakers to follow them through the Abwehr radio networks. Hopefully, this would help them break the keys for other ciphers that the German controllers were using to pass the reports on to Hamburg.
By the end of 1940, Robertson had a dozen double agents under his control. At the same time, MI6 was running a number of German spies abroad. ‘Basically, MI5 was responsible for security in the UK and MI6 operated overseas,’ said Hugh Astor, one of the agent runners. ‘Obviously there was a grey area as far as double agents were concerned because they were trained and recruited overseas and at that point were the concern of MI6, while once they arrived here they became the responsibility of MI5.’
A ‘Most Secret’ committee was set up to decide what information should be fed back to the Germans. Its small select membership included representatives of MI5, MI6, naval, military and air intelligence, HQ Home Forces and the Home Defence Executive, which was in charge of civil defence. The committee was called the XX Committee, although it swiftly became known as the Twenty Committee, or more colloquially, the Twenty Club, from the Roman numeral suggested by the double-cross sign. It met every Wednesday in the MI5 headquarters, initially in Wormwood Scrubs prison, but subsequently at 58 St James’s Street, London. ‘The XX Committee was chaired by J. C. Masterman,’ said Astor. ‘Tar Robertson, who ran BIA, really developed the whole thing. He was absolutely splendid, a marvellous man to work for. He and Dick White were the two outstanding people I suppose and Tar collected around him some very bright people who actually ran the agents for him.’
The Twenty Club’s job was to decide what information could be fed back to the Abwehr without damaging the British cause. Initially, with the threat of a German invasion dominating the atmosphere in London, it was decided that the ‘intelligence’ provided by the double agents should be used to give an impression of how strong Britain’s defences were. But by the beginning of 1941, it was clear that more could be done with the double agents. They could be used to deceive the Germans, to provide them with misleading information that would give Allied forces an advantage in the field.
The MI5 and MI6 officers handling the double agents needed to know what information they could give to their agents to build up their reputations with the Germans. Much of it was ‘chicken-feed’, unimportant information that would give the Abwehr a feel that its agents were doing something and had access to real intelligence, without telling them anything really harmful to the war effort. But mixed among this were key pieces of specious or misleading information, designed to build up a false picture of what the British were doing.
The committee’s task was to co-ordinate this work. They supervised the system but they did not run the individual agents. ‘They approved the overall plan,’ Astor said. ‘I was in touch with the Germans probably two or three times a day by radio and so I had to move fairly quickly. So the approving authorities were not the actual Twenty Committee because it only sat once a week. I would get approval from people who were on the committee and every week I and others who were actually active would prepare a short report for the committee saying what we were doing and what we had done.’
But while the system appeared to be working, the Twenty Committee and the agent handlers had a problem. They could not be sure the Germans were fooled. The Abwehr’s operations abroad seemed to be unbelievably incompetent. The agents were ‘too amateurish’ to be genuine. Capturing them and turning them around was so easy that the British suspected that it might be part of an elaborate Abwehr deception. Even if this were not the case, the lax system employed by the Germans, who ignored basic security procedures by p
utting the agents in touch with each other, warned the Twenty Committee when any new agents were sent to Britain. But it also meant that if the Abwehr realized that one of its agents was operating under British control, it would have to assume that they were all blown. ‘The position at the beginning was largely experimental as no one knew very much about the working of double agents or about the working and general incompetence of the Abwehr,’ wrote Ewen Montagu, the naval intelligence representative on the committee. ‘Later on after we had had experience of the German Intelligence Service, no incompetence would have surprised us.’
While the response of the Abwehr controllers to the double agents’ reports helped the Twenty Committee to work out where the gaps in the Germans’ knowledge lay, it did not tell them whether or not the misleading intelligence picture they were attempting to build up was believed in Berlin. The only way of finding this out was by deciphering the messages passed between the Abwehr outstations in Paris, Madrid, Lisbon and their headquarters. But these links all used the Abwehr’s Enigma machine, which was completely different to those used by the other German services.
So the Twenty Club’s confidence in their double agents was considerably enhanced in December 1941 when Dilly Knox, who was terminally ill with cancer and working from home, broke the Abwehr Enigma. This followed six months of research during which he was assisted by a young Mavis Lever (now Batey) and Margaret Rock. It was to be the last of Knox’s remarkable achievements. Just over a year later, in February 1943, and after a long struggle against cancer, he died.
The first of the messages, known as ISK for Illicit (or Intelligence) Services Knox, was issued on Christmas Day 1941. They were invaluable to the Twenty Committee, revealing that the Germans believed the false intelligence the Twenty Committee was feeding them and showing whether or not individual double agents were trusted or under suspicion, in which case steps could be taken to remedy the situation. Two months later, Mavis Lever solved a separate Abwehr Enigma machine, known as GGG, which was used near the Spanish border. By the spring of 1942, the information collected from the Bletchley Park decrypts had built up such a good picture of Abwehr operations in Britain that Robertson was able to state categorically that MI5 now controlled all the German agents operating in Britain. The Twenty Committee was able to watch the Germans making arrangements to send agents to Britain and discussing the value of their reports, Robertson wrote. ‘In two or three cases we have been able to observe the action (which has been rapid and extensive) taken by the Germans upon the basis of these agents’ reports.’
Nevertheless, the breaking of the Enigma cipher had brought a new problem for the committee. The release of any material from Bletchley Park was controlled extremely strictly by MI6 in order to safeguard the Ultra secret. The fact that the ‘unbreakable’ Enigma ciphers had been broken had to be protected at all costs. The MI6 representative on the committee was Felix Cowgill, the head of Section V, the MI6 counter-espionage division. A former Indian Police officer, Cowgill was a shy, slightly built man in his mid-thirties. ‘His face gives the impression of intensity coupled with a great weariness,’ said Kim Philby, the MI6 officer and KGB spy, in one of his reports to Moscow. ‘Although normally quiet in manner, due to shyness, he is combative in his work, always prepared to challenge an office ruling.’
Cowgill defended the Ultra decrypts vigorously, to the extent of refusing to allow the Home Forces and Home Defence Executive representatives on the Twenty Committee to see them at all, while anything that referred to MI6 agents was held back even from MI5. ‘Cowgill was so imbued with the idea of security that when he was put in charge for C of this material, he was quite willing to try entirely to prevent its use as intelligence lest it be compromised,’ Montagu said. ‘These views inevitably caused friction.’
While there was no doubt that some within MI5 were paying scant regard to the necessary restrictions on the Bletchley Park decrypts, Cowgill’s attitude made the Twenty Committee’s operations almost impossible. Some members were not privy to vital information about the agents on which the others were basing their decisions. The result was potentially far more detrimental to security than the widespread dissemination that Cowgill was trying to prevent. His controls were soon being ignored on a wholesale basis. ‘A good deal of bootlegging of information had to take place,’ said Montagu. ‘Many undesirable “off the record” and “under the table” practices were essential unless work was to stop entirely.’
The situation came to a head over the case of a man who was to become the most valuable of all the Double Cross agents – Juan Pujol Garcia, better known by his codename: Garbo. The Bletchley decrypts had revealed an Abwehr agent who claimed to be running a network of agents in Britain. His reports were ridiculously inaccurate. He was clearly a fraud, reporting ‘drunken orgies and slack morals in amusement centres’ in Liverpool, and Glasgow dockers who were ‘prepared to do anything for a litre of wine’. It ought to have been obvious to the Germans that, not only had he never met a Glasgow docker in his life, he had never been to Britain. Yet they believed him wholeheartedly. MI6 became concerned that his false reports might damage the Twenty Club’s own plans. Then in early February 1940, the MI6 head of station in Lisbon reported that he had been approached by a Spaniard, claiming to be a top Abwehr secret agent. He said he had been disaffected by the Spanish Civil War and was keen to help Britain to fight the Germans. Having been turned down by the MI6 station in Madrid, he had gone to the Abwehr equivalent, persuading the officers there that he was a Spanish intelligence officer who had been posted to Britain and offering to act as a German spy.
In fact, Pujol went to Lisbon, where, armed with a Blue Guide to Britain, a Portuguese book on the Royal Navy and an Anglo-French vocabulary of military terms, he produced a series of highly imaginative reports built on his alleged network of agents. Pujol was vehemently anti-Nazi and his reports were apparently designed to disrupt the German intelligence service – he was in effect a freelance double-cross operation in miniature. Cowgill kept him secret from MI5, on the basis that although sending reports ostensibly from British territory and therefore notionally under MI5 jurisdiction, he was actually abroad and the responsibility of MI6.
The fact that an important German agent was sending uncontrolled reports about Britain, however inaccurate, could have caused immense damage if it had not been taken into account in the overall deception plan. So when, at the end of February, senior officers in MI5 discovered that his existence had been hidden from them, they were furious. A few weeks later, they discovered that Cowgill had also been holding back ISOS messages thought to refer to MI6 agents, placing them in a separate series known as ISBA, which was not being circulated to either MI5 or the service intelligence departments.
This was the final straw and Sir David Petrie, the head of MI5, used the row to lobby for MI5 to take over control of Section V. He added all the arguments over the distribution of deciphered intercepts, MI5 criticism of the apparent lack of basic knowledge about Germany among a number of Section V officers, and the fact that it was based in St Albans, too far away from London to make liaison with MI5 as easy as it should have been. Stewart Menzies proposed a compromise. He would set up a new department within Section V called VX to deal exclusively with the Double Cross system. It would be based in London to allow easy liaison with MI5 over the work of the double agents and would be headed by a man with unrivalled knowledge of Germany, Frank Foley, who had been the MI6 head of station in Berlin throughout the 1920s and 1930s.
Foley, now better known for his role in helping tens of thousands of Jews to get to Palestine in contravention of the British rules, replaced Cowgill as the MI6 representative on the Twenty Committee. ‘There was an obvious qualitative difference in the way in which the committee worked from then on,’ one former MI5 officer said. ‘For the first time, the MI6 representative was speaking authoritatively because he was a real operational officer. He knew what he was talking about and it showed.’ Masterman, who as secretary
of the Twenty Committee was in the perfect position to know, also pointed to the mid-1942 changes as the moment that the Double Cross system really began to take off. ‘Broadly speaking, bad men make good institutions bad and good men make bad institutions good,’ he said. ‘It cannot be denied that there was some friction between MI5 and MI6 in the early days, but this disappeared when the MI6 representative on the committee was changed.’
One of the major problems faced by those running the Double Cross system was that when the British tried to work out what the Germans would do next, they based their judgements on what they would have done in the same situation. But their opponents, and in particular Hitler, had a different view of things. ‘It is necessary for the deception staff to think as the enemy thinks and to divorce themselves entirely from being influenced by what we would do if placed in what we imagine to be the enemy’s position,’ one former deception officer said. ‘Again and again, what the deceivers suggested was plausible to the enemy. But both our operations and intelligence staffs maintained that it was not because they were governed strictly by their appreciation of what we would think was plausible in the enemy’s place.’
Masterman had been a prisoner of war in Germany during the First World War and he had a better grasp of the German way of thinking than most of the committee, but even he could not match the new MI6 representative. ‘Foley knew the Germans backwards,’ one former MI6 officer said. ‘So if people wanted to know how the Germans would react to any particular deception plan, they would naturally ask him.’ Foley swiftly succeeded in turning around the committee’s attitude to MI6, which had been so heavily tarnished by Cowgill’s restrictions. ‘He was not a member of the establishment clique,’ said another of Foley’s former colleagues. ‘But he was a pretty serious chap, feet on the ground, solid, very much the elder statesman, giving useful advice whenever called upon. His exceptional knowledge of the workings of, and personalities in, the Abwehr, acquired during years of service in Berlin, made him a tower of strength.’
The Bletchley Park Codebreakers Page 32