Operation Garbo

by Juan Pujol Garcia

  SIS developed a special organisation, designated Section V, to exploit ISOS and follow up any of its leads. For security reasons, Section V led a semi-independent existence and was headquartered in a country mansion near St Albans in Hertfordshire. By the end of 1941 SIS had posted specially briefed Section V officers to both Spain and Portugal. Kenneth Benton was sent to Madrid (in spite of the ambassador’s protests) and Ralph Jarvis, a merchant banker in peacetime and formerly a member of General Templer’s intelligence staff with the ill-fated British Expeditionary Force in France, went to Lisbon. There he established an office in the British Repatriation Office, so as to avoid being too closely associated with Johns’s passport control office.

  MI5 quickly realised that every care had to be taken to protect both the double-cross system and Bletchley’s ‘most secret source’. Accordingly, the circulation of the ISOS decrypts was rigidly controlled. Measures to conceal the existence of the system itself were rather more complex. In January 1941, with the appearance of TRICYCLE, yet another German spy – of the real variety, equipped with a transmitter – the Wireless Board authorised the formation of a subgroup of specialist intelligence officers to develop the double-cross system further and coordinate the real and notional activities of their charges.

  One problem associated with B1(a)’s work was the necessity to supply the enemy with a convincing volume of plausible intelligence. If MI5’s double agents only communicated banalities while working in harness the Germans might understandably lose interest, and a valuable conduit would be lost to the Allies. But spicing the harmless information with more attractive tidbits required considerable skill and judgment, and a precise knowledge of future plans. It would be entirely counterproductive to guess at proposed operations and hope that coincidence did not bring the notional idea into conflict with reality. In such circumstances a well-meaning case officer might invent what he believed to be a non-existent military operation, perhaps a raid on a particular target in France, and then discover later that his chosen target had indeed been selected for attack. It would be equally embarrassing to suggest the location of a heavily camouflaged ammunition dump, and then learn that there was indeed some sensitive site located in the neighbourhood. If such an incident ever occurred, the military authorities would justifiably cease cooperating and the informal flow exchange of information, upon which B1(a) relied, would be jeopardised.

  During the months of late 1940 B1(a)’s eight double agents were being serviced with a satisfactory amount of intelligence, but the routing required much more refinement. The Wireless Board itself met infrequently, owing to the elevated status of its members, who had many other demands on their time. The business of acquiring and collating information for the enemy was a trifle haphazard and was left largely to the initiative of the individual case officers. They would travel the country, acting out the role of their charges, and would note various items of interest. Before inserting such material into the texts of messages for delivery to the Abwehr, the case officers would secure an unofficial consent from the intelligence division of the interested authority. If, for example, TATE had been asked to confirm the location of a particular aerodrome in East Anglia, his case officer, Bill Luke, would ask Major Robertson to clear the information with the director of intelligence at the Air Ministry, Air Commodore Archie Boyle. As it happened, Boyle was one of Robertson’s most enthusiastic supporters and probably would have given every assistance, but his views were not always shared by other senior intelligence bureaucrats who were doubtful about the advantages of giving the enemy our secrets. Less constructive colleagues opted for the altogether safer choice of substituting entirely innocuous or erroneous information. Such manoeuvres were less controversial and therefore easier to defend at a later date, but they also weakened the credibility of the agent in whose name it would be transmitted. In the interests of watertight security, only a few officers in the service intelligence departments could be advised about B1(a)’s work, and those who had not been indoctrinated were reluctant to disclose operational details without formal sanction. The ideal solution to the continuing dilemma was the introduction of a new coordinating body, staffed at a lower level than the Wireless Board and chaired by a non-partisan figure who could command everyone’s respect. That man was J. C. Masterman, then a don at Christ Church, Oxford.

  The son of a naval officer, Masterman had been educated at Osborne and Dartmouth before going on to graduate from Worcester College, Oxford, and Freiburg University. He was an accomplished athlete and played cricket, hockey and tennis for England. During the Great War he had been interned at Ruhleben, the notorious civilian prison camp constructed on a racecourse near Berlin, where he gave memorable history lectures and captained the camp’s cricket team. His well attended courses were only interrupted by his brief absence early in August 1918, when he escaped. His four years in captivity made a lasting impression on Masterman who, along with many of his Oxford contemporaries, joined the intelligence corps in 1940. On 2 January 1941 Masterman took the chair of the Twenty Committee, so-called because of the two Roman numerals of double-cross.

  Although Masterman was a newcomer to B1(a)’s work, his agile intellect soon grasped the intricacies of the operation. He was also assisted by a permanent committee secretary, John Marriott, who was also Robertson’s deputy. The creation of the Twenty Committee, which coincided with the welcome arrival of TRICYCLE, marked a turning point for the Security Service. From this stage onward, the progress of each agent was monitored by the committee, with responsibility for the day-to-day running of each remaining in the hands of the individual MI5 case officer. The new scheme freed the case officers from the sometimes dangerous chore of collecting information. Instead, the Twenty Committee was required to obtain the cooperation of the services and supply the case officers with suitable material for the enemy’s consumption. To cut red tape and expedite B1(a)’s plans, each service intelligence department was invited to second a representative to the committee. These liaison officers were thus able to serve the committee’s interests and simultaneously allay the fears and suspicions of their masters.

  The Twenty Committee met on a weekly basis during the course of the following four and a half years (until 10 May 1945) – a total of 226 times, usually on Wednesday afternoons. Although its composition was to vary, the longest serving members were Robertson, Bill Luke (who succeeded John Marriott as committee secretary), Martin Lloyd (the SIS representative), John Drew from the Home Defence Executive and Ewen Montague from the Naval Intelligence Division. Whenever necessity demanded it, individual B1(a) case officers attended the committee’s meetings, and from the spring of 1942, Tomás Harris frequently dominated the proceedings on behalf of GARBO. From 1943 onward, when the American Office of Strategic Services created a counter-espionage branch, Norman Holmes Pearson was delegated from their London office to liaise with MI5. This crippled academic, then professor of literature at Yale University, was granted the unique privilege of sharing Masterman’s office in St James’s Street, and was therefore privy to all the Twenty Committee’s secrets.

  As the double-cross system grew more sophisticated, and the ISOS decoders gained in experience, so the demand increased for some plausible ‘covers’ to mislead the enemy. One vital objective was to convince the enemy that British Intelligence was no match for the Abwehr, and in September 1941 an opportunity presented itself for MI5 to demonstrate some deliberate ineptitude. The purpose of the exercise, which bore a close resemblance to the wretched experience of Carl Muller in the Great War, was to persuade the enemy that MI5 were novices in the art of running double agents, and an uncooperative double agent code-named SCRUFFY was the chosen vehicle.

  SCRUFFY was a genuine German spy and his real name was Alphons Timmerman; he was a twenty-eight-year-old Belgian ship’s steward. He had presented himself at the frontier at Gibraltar, claiming to have trekked across Europe. He had been given a passage to Holland, but by the time his ship had docked he had been betrayed by a
n ISOS text from Spain reporting the successful conclusion of the first part of his mission. The reference to Timmerman had been linked to a further ISOS message reporting that a Belgian recently recruited by the Abwehr was to have his pay sent straight to his mother, whose address was provided. Having landed at Glasgow, Timmerman underwent a routine examination at the Royal Victorian Patriotic School, the reception centre where all new arrivals were accommodated before being cleared for official entry into the United Kingdom. Headed by Major Ronald Hayler of MI5’s B1(d), this huge establishment in Battersea processed many thousands of genuine refugees, and a small number of enemy agents. During his RVPS examination Timmerman was found to be carrying an unusually large sum of money and the ingredients for making secret ink. He was promptly transferred to the harsher regime at Camp 020, where skilled interrogators extracted an admission that his mother was living in the same Belgian village that had been mentioned in the ISOS decrypt. A confession soon followed, and Timmerman was removed to Wandsworth Prison for trial and eventual execution. In the meantime, an MI5 officer corresponded, somewhat ineptly, with Timmerman’s German controller, using his secret ink and his post office box number in a neutral country. The idea was to persuade the Germans that their agent was still at liberty. As soon as the details of his arrest were made public, it was hoped that the Abwehr would realise that Timmerman’s correspondence had been faked by MI5. According to the theory, the Abwehr would then congratulate themselves on MI5’s incompetence, demonstrated by their poor choice of agent. The Belgian was found guilty under the Treachery Act and sentenced to death, and the execution was duly carried out on 7 July 1942. As was customary in those days, a brief public statement was released the following day and carried in most newspapers. MI5 had calculated that immediately the official announcement had been spotted the Abwehr would realise that all the letters purporting to have been written by their agent had been forgeries. A detailed review of the Timmerman letters would have confirmed the deception and revealed a number of deliberate errors. All had been thoughtfully constructed by B1(a) and then inserted into the covert texts but, much to MI5’s chagrin, the Abwehr appeared to ignore Timmerman’s death notice and the deliberate mistakes. Instead of suddenly breaking off contact, as had been expected, the Abwehr continued the traffic as if nothing was amiss, and the Twenty Committee decided to abandon the exercise before it got completely out of hand. The Abwehr could hardly have failed to spot Timmerman’s death notice, yet they seemed willing to continue with the bogus correspondence. Evidently, the Abwehr considered MI5 to be even less sophisticated than MI5 had anticipated or wanted!

  Although this particular ploy failed, it is an illustration of the extraordinary lengths MI5 were prepared to go to in order to develop the double-cross system. On this occasion, a real German spy, SCRUFFY, had been hanged simply to promote MI5’s interests. As it turned out, the execution had failed in its prime intelligence purpose, to demonstrate MI5’s inefficiency. If anything, the episode, and certainly the Abwehr’s apparent willingness to remain in contact with an agent they knew to be dead, illustrated how easily the enemy could be taken in.

  By the time Pujol emerged on the scene the Allied intelligence Machine had accommodated the Twenty Committee and had given due recognition to its achievements. The double-cross system now embraced the entire Abwehr effort in Britain, and was poised on the brink of much greater successes. B1(a)’s ever-expanding stable of real and notional agents had completely eliminated every independent German spy and had enabled the port security staff to prepare reception committees for new arrivals. RSS cryptanalysts were supplying GCHQ with valuable clues to the construction of the enemy’s latest ciphers, and their study of the opposition personalities had enabled them to build an accurate order-of-battle for both the Abwehr and the Sicherheitsdienst (SD). The B1(d) investigators at the Royal Victoria Patriotic School also benefited from the system because they frequently received advance warning of suspects. Clues from B1(a) also helped the B1(e) interrogators at Camp 020 to extract damaging admissions from even the most well-trained of spies. Most were genuinely astonished at the extraordinary depth of MI5’s knowledge about the Abwehr’s operations and intentions which, of course, had been obtained from the invaluable ISOS.

  By the spring of 1942 the British intelligence machine had gained sufficient experience to carry the double-cross system into the dangerous area of large-scale strategic deception. The necessary foundation had been laid; all that was needed was a suitable agent.

  6

  GIBRALTAR

  No one can be perfectly happy till all are happy.

  Herbert Spencer

  I had smuggled the $3,000 that Federico had given me in Madrid into Portugal without any trouble. I had rolled the money up tightly and slid it into a rubber sheath, then I had cut a tube of toothpaste open at the bottom end, emptied out all the toothpaste, inserted the notes and rolled the bottom of the tube up to look as if it were half empty. I couldn’t get all the money in one tube, so I had put the rest of the 100-denomination notes into a tube of shaving cream using the same method. It was perfect camouflage and I had no problems at the Fuentes de Oñoro checkpoint at the border.

  With this money I now bought a Baedeker tourist guide to England, Bradshaw’s railway timetable and a large map of Great Britain, and retired to Cascais to study them in detail. Then, in October 1941, I sent my first message to the Germans from Cascais, although, as far as Federico was concerned, I was already in England; it was quite a long message.

  In invisible ink, I told them that before leaving Portugal I had posted the key to a safety box in the Espírito Santo bank to the German embassy in Lisbon, with instructions that they should send it to Federico at the German embassy in Madrid. I went on to explain that on arriving in Britain, I had got talking to the KLM pilot who flew us in to London and had become very friendly with him, introducing myself as a Catalan political exile who had had to flee because of my political views. I had then persuaded the KLM pilot to take my Spanish mail to the Espírito Santo bank in Lisbon on a regular basis. In my letter I also said that he had not at first been too keen to help, perhaps because he suspected something, but that I had assured him that none of my envelopes would be stuck down so he could always see for himself what the contents were; my letters were meant, I had explained to him, to give my fellow Catalan patriots information about other Catalans in exile in England. I had also told the pilot, I said to the Germans, that this, my first letter, would merely be informing my fellow Catalans in Spain of my arrival in the British Isles. Finally, I suggested to the Germans that I was thinking of going to live near Lake Windermere in the centre of Britain because I had heard that there were a fairly large number of troops stationed up there.

  I had had to devise this ploy about the KLM pilot because, as I was still living in Portugal, there was absolutely no possibility of my letters being franked by the British Post Office. So this, my first letter, consisted, on the surface in ordinary visible ink, of an enthusiastic account of England by a passionate Catalan democrat with all the information for the Germans written in between the lines in invisible ink.

  I had become a real German spy.

  I sent three messages to the Germans from inside Portugal (purporting to come from England), all of which were worded with the care and attention worthy of the most adroit German field agent. I tried hard to introduce new information gradually and to be cautious when I mentioned the new contacts I had recruited to help me. In my first message I told them that I had found three people who would continue to supply me with further information, whom I had made my subagents: one in Glasgow, one in Liverpool and one from the West Country. The naivety with which I told them the facts I had discovered probably contributed to the conviction that I really was in London.

  In the second message I said that I’d been offered a job at the BBC and was about to accept. I also said that I’d heard that the navy was carrying out landing-craft manoeuvres on Lake Windermere and described in de
tail how I had grappled with a whole string of obstacles.

  The third message, which as always was a complete invention, had a unique impact, although it was not until much later that I learned of the stir it caused in the British Secret Service. In this third message I said that a convoy of five ships had left Liverpool for Malta, although neither the date nor the number of ships actually tallied exactly with my message. But the coincidence was sufficiently close for the British to think that a German agent was loose in England.

  This worried the British enormously, especially when they were able to confirm that the Germans had carried out an aerial reconnaissance of the projected route and of Malta’s Valletta Harbour. Who was this agent? Where was he getting all this information from which endangered British security?

  The British were going crazy looking for me as they had no idea where I was and, indeed, whether I existed at all. Cyril Mills told me later that he was in Portugal at the time and that he had mentioned to the intelligence gatherers in Lisbon and to those in Madrid that the German agent ARABEL could have come from either of those cities. If so, how had he got into Britain? Everyone in Britain was hunting for me and trying to discover how I was getting information through to Madrid.

  So much of this reads like a fairy story that it will not come as a surprise to readers to learn that it was this third message which led to the British accepting me, and which eventually enabled me to become both the German’s top spy ARABEL and MI5’s counterspy GARBO.