Meanwhile, MOONBEAM reported his own safe arrival and, with the collaboration of the Federal Bureau of Investigation, recruited a cousin who was resident just over the American border in Buffalo. MOONBEAM’s messages, which were relayed to the Abwehr via GARBO, caused some amusement in London and Berlin because of the agent’s mercenary nature. Every tiny item of expenditure was carefully recorded and claimed for. On one occasion, the Germans queried a small sum spent from petty cash for paying a labourer to clear snow from the path to his home, MOONBEAM responded by demanding to know how he could be expected to collect information if he couldn’t get out of the house! The Abwehr promptly radioed their approval.
WATCHDOG himself came to an end late in the summer of 1943, when he ran out of funds, and he was transferred to MI5’s custody in England. Nevertheless, Mills continued with MOONBEAM and his cousin in upstate New York, and offered a safe refuge to those of GARBO’s network who suddenly had to flee Britain. CHAMILLUS, for example, the wretched Gibraltarian, eventually deserted from his post in the NAAFI in October 1944 and made his way to Canada, where he was employed as MOONBEAM’s wireless operator. This final act proved to be extremely useful because CHAMILLUS was entrusted with a new, high-grade cipher, a copy of which was delivered by MI5 straight to GHQ at Bletchley.
By February 1944 GARBO and Tommy Harris had completed the final sector of DAGOBERT’s ring and had laid the foundations for a truly integrated deception operation. Twenty-four subagents and sources were spread evenly throughout the country and the entire machine had been cranked up ready for action. Early in the month GARBO himself went on a tour of the south and south-west of the country and handed Charles Haines a brief message, which was transmitted from Crespigny Road at 8.10 in the evening of 17 February:
An observation to which I give importance is that the distribution of forces to date is all along the length of the coast and there is no concentration at special points.
This typically bland appraisal was to mark the beginning of GARBO’s active involvement in OVERLORD’s cover plan, code-named FORTITUDE. The only unforeseen hitch, which meant little at the time of its receipt, was a curious signal, dated 15 December 1943, warning GARBO to leave London as soon as possible. It was only six months later, after the first buzz bombs had fallen on the capital, that the significance of the message was fully realised. But by that time FORTITUDE was in full swing.
9
OPERATION FORTITUDE
While Operation OVERLORD was being planned, the Allied chiefs of staff became increasingly aware of two crucial problems: that establishing a beachhead in the midst of the enemy’s defence was one thing; maintaining it in the face of a swift counter-attack would be quite another. Three separate needs were therefore identified. Firstly, a requirement to persuade the enemy to position his forces as far away as possible from the intended target area. This was done in the large scale by formulating a scheme to keep Axis troops committed in nonessential regions, such as the Balkans and the Mediterranean. In Europe itself the requirement was translated into providing plausible evidence that the coming invasion, which was recognised as virtually impossible to disguise, would be headed by a two-pronged attack on Norway and the Pas-de-Calais area. Secondly, the enemy had to be misled about the exact timing of the assault; and thirdly, the Germans had to be convinced that any landings along the French coast, apart from those in the Pas-de-Calais, were diversionary in nature. In other words, even after troops had begun to move ashore in Normandy, the Germans must be made to believe that this initial operation was but a feint designed to draw their forces away from the real target area, which was further north. To bring off this last trick was the tall order presented to the deception planners at the Supreme Headquarters Allied Expeditionary Force (SHAEF). Their eventual solution, known as FORTITUDE, was described by the historian Charles Cruikshank in his history of Deception in World War II (Oxford University Press, 1979, p. 170) as ‘the largest, most elaborate, most carefully planned, most vital and most successful of all the Allied deceptive operations’.
Assembling the largest invasion fleet in history could not be achieved overnight, and the concentrations of landing craft and other essential equipment would be difficult to hide from German reconnaissance aircraft, quite apart from the 5,000 ships estimated to be involved, so the decision was taken to execute a variety of reasonable security measures to prevent unauthorised leakages. As GARBO had anticipated (and circumvented), a ban on visitors was placed on a zone ten miles deep along the coast from Cornwall to Lincolnshire and, in an unprecedented proposal, the Foreign Office discussed the imposition of a total ban on all uncensored communications to and from neutral diplomats in London, from a date yet to be decided in mid-April 1944.
The SHAEF planners, huddled over their maps in Norfolk House, St James’s Square, knew they had two useful advantages they could exploit. The Allied chiefs of staff had decided not to try and capture a Channel port intact, which was the most straightforward answer to the logistical problem of supplying such a huge body of men and armour. Instead, they had opted for prefabricated harbours, consisting of huge concrete caissons, which were to be floated into position off the beaches. Once the caissons had been constructed, they were to be left submerged in the Solent and elsewhere until the moment came for them to be towed across the Channel. Provided the Germans remained ignorant of the true purpose of the floating harbours, code-named MULBERRY, there was a good chance that the Abwehr would recommend a concentration of defences around the Channel ports between Ostend and Brest. The second advantage lay in the Twenty Committee’s confidence that every enemy agent in Britain was working under MI5’s control. Although other double agents were to be employed in FORTITUDE, much of the burden would fall on GARBO and his network because the ISOS intercepts had demonstrated that he enjoyed the highest standing in Berlin.
The Twenty Committee, therefore, resolved to convey the elements deemed essential to OVERLORD’s success: that the invasion could not be launched until at least July 1944; that it would be preceded by an attack on Norway; and that the eventual thrust into France would centre on the Pas-de-Calais, after some initial feints elsewhere. It was learned after the war that this strategy fitted neatly into current German thinking, which took a characteristically rational approach to their dilemma. Field Marshal von Rundstedt, for example, took four major factors into consideration when he decided that the Allies would opt for the shortest route across the Channel: the Allies would need at least one major port; they would require constant protection from the air, so the nearer the beachhead the better; Calais offered the most direct path to Germany; and the capture of the V-weapon launch sites would become an important political consideration once the bombardment of London had begun. In fact, the projected V-1 offensive did not occur until after D-Day, so SHAEF was never in a position to waste time debating the last factor. SHAEF never knew that FORTITUDE was actually confirming the conclusions that German high command had already reached, albeit tentatively.
The cover deception plan itself was divided into FORTITUDE NORTH, which concentrated on the notional build-up of troops in Scotland for the attack on Norway, and FORTITUDE SOUTH, which promoted the ‘shortest route’ assault. Just as the fictitious British Fourth Army was to play a vital part in FORTITUDE NORTH, it was Roger Hesketh’s creation of an entirely bogus First United States Army Group (FUSAG) which was to be the key to FORTITUDE SOUTH. But in February 1944, just when SHAEF and the Twenty Committee had settled the final details of the operation, Tommy Harris suddenly stepped in with a bombshell: GARBO and his network should be withdrawn immediately.
The last-minute crisis was precipitated by another double agent code-named ARTIST. In fact, ARTIST was a young Abwehr officer stationed in Lisbon named Johann Jebsen who, since the summer of 1943, had been actively collaborating with a network of Yugoslav double agents. Jebsen, who was supposed to be virulently anti-Nazi, cut a colourful figure in wartime Portugal, driving between Lisbon and his villa in Estoril in a Rolls-
Royce. He claimed to be an Anglophile, and his first link with the British came via two Yugoslav brothers, Dusko and Ivo Popov, who were code-named TRICYCLE and DREADNOUGHT respectively by MI5. Jebsen’s excuse for his contact with the Popovs was that he was negotiating an underground escape route through Europe for evading Yugoslav airmen. At the end of January Dusko Popov returned to London, after a two-month stay in Portugal where he had held several meetings with Jebsen, and gave the Twenty Committee a lengthy summary of his conversations. Popov had known Jebsen for some years and trusted him implicitly, but Tommy Harris was aghast by one particular section of Popov’s report, in which Jebsen attempted to prove his bona fides to the Allies by naming some of the Abwehr’s chief agents run from Lisbon. Heading the list was the name of Juan Pujol, code-named ARABEL.
Harris accepted that Popov’s judgement of Jebsen was probably the correct one and that he could be trusted, but he pointed out that the implication of not taking action against any of the spies on Jebsen’s list was the equivalent of providing him with confirmation that they were all being run by the Allies. If Jebsen came to this conclusion, which was inescapable, he would inevitably realise that most of the intelligence they were supplying was nothing more than a clever fabrication. This information might indeed be safe with Jebsen, but supposing his circumstances changed: would his loyalties also change again? Another cause for concern was Jebsen’s admission that his contacts with the Allies had the approval of the Abwehr controller in Lisbon, who had told him to develop some cover as a discontented German out of favour with the regime. On the present showing, argued Harris, OVERLORD’s fate was held in the hands of a single man who happened to be an enemy intelligence officer. If Jebsen knew GARBO was controlled by MI5, he would guess that FORTITUDE was a fake and so jeopardise OVERLORD. It was not the kind of news any self-respecting security official cared to tell an Allied supreme commander.
The Twenty Committee deliberated over the ARTIST crisis and considered all the alternatives. The easiest option, which was to pull Jebsen out of Portugal, thus isolating him from the Abwehr, was ruled out because if the Germans learned that he had defected, they would automatically assume that all the agents that Jebsen had had access to would be compromised. Another possibility was asking the Secret Intelligence Station in Lisbon to assassinate Jebsen, but it was pointed out that such a move would be bound to attract a major German investigation into all the circumstances surrounding the incident, and that too might uncover Jebsen’s illicit contacts with the British. Eventually, the Twenty Committee reached the only possible decision: they would do nothing, but ask SIS to keep an eye on him.
Understandably, Tommy Harris was greatly agitated about this decision, and it later seemed that his worst fears were about to be realised. But before describing the dramatic events of May 1944, we should return to the first stages of FORTITUDE, which were proceeding with GARBO’s help in spite of Harris’s objections.
The initial part was the pretence that the Allies were not yet in a position to launch a major offensive across the Channel and had, therefore, postponed their original invasion plans until much later in the year.
In the middle of February 1944 GARBO pretended to have undertaken a tour of the south coast of England. On 19 February, from Portland in Dorset, GARBO posted his fifteenth letter and mentioned having spotted some American soldiers in the neighbourhood of the town. He also described their shoulder flashes: ‘the number “1” in red on a khaki ground’. This was a calculated reference to a genuine unit, the 1st US Infantry Division, which had in fact been brought back to England from the Italian front some months earlier, but SHAEF was anxious to conceal the return of these battle-seasoned troops until the last possible moment before the real invasion. GARBO’s intervention was designed to win him credit for passing on some legitimate news. It had been argued that so many Africa Star campaign medals were being worn on uniforms in England, with its distinctive red and yellow ribbon, that it would be impossible to keep the division’s arrival secret for much longer. Other agents supported GARBO’s observation by stating that these hardened troops were engaged in training other troops, thereby promoting the idea that Allied preparations for the invasion were not particularly advanced.
GARBO explained that ‘balanced forces are being held in readiness in England to occupy any part of north-west Europe against the contingency of a German withdrawal or collapse’. In his first message concerning SHAEF’s strategy, which effectively marked the opening of Operation FORTITUDE, GARBO remarked that ‘it does not require a very wise man to deduce that should the way be left free they would not hesitate to take advantage of it’. As additional evidence of the Allies’ plan, he described how he had seen a pile of newly printed leaflets in the office of his contact in the Ministry of Information and had purloined a copy. Entitled Avis à la Population, it was despatched to Lisbon. The forgery, the only one of its kind ever printed, was marked for distribution to French civilians following a German withdrawal and urged them to cooperate with their new occupiers. As a foundation for FORTITUDE NORTH, GARBO reported that he had ordered his Glasgow-based deputy, BENEDICT, to monitor the growing number of naval exercises in the Clyde, and the Greek deserter had been instructed to find lodgings at Methil, on the east coast of Scotland. This he did on 1 April 1944, booking a room for the next six weeks. As confirmation that the British Fourth Army was still operational (and, indeed, the key to the northern campaign), BENEDICT reported to GARBO on 28 March 1944 that he had just returned from a trip to Dundee, where he had spotted the 52nd Lowland Division and a unit bearing shoulder flashes of a shell on a dark background. When relaying this message in a four-minute wireless transmission at 1920 hours that evening, GARBO commented: ‘This insignia is completely unknown to me.’ This was not entirely surprising, considering that Roger Hesketh had only just invented it!
Meanwhile, GARBO himself had been on a tour to check on some of DAGOBERT’s informants. During a transmission at six in the evening of 7 March, he reported:
I was able to confirm last Sunday the accuracy of the recent report sent by DONNY from Dover. I am, therefore, able to classify him in future as a good reporter.
In the same message he also authenticated messages received from DAGOBERT’s other agents, including DICK the Indian fanatic: ‘With regard to the military report, it is completely accurate so that we can catalogue this collaborator as being good.’
On 13 April he sent a radio message giving a good opinion of DORICK, one of the Welsh Fascists, who had reported seeing ‘a lot of troops and vehicles of the ninth Division’ passing through Norwich: ‘I consider this first report of this collaborator fairly good as he tries to get details, from which one is able to appreciate the interest he takes in explaining what he has seen.’
On 22 April GARBO received a long letter from the Abwehr, dated 3 April 1944, Lisbon, which showed that they were particularly impressed with DAGOBERT’s subagents:
I have taken note with great interest of what you have told me in your letters about the amplification of your network, and the numerous messages which you have sent during the last few weeks have demonstrated to me that you have been absolutely right in your idea of nominating the old collaborators as subagents of their networks. In particular, the network of DAGOBERT appears to be the one which is giving the best results.
This message caused some celebration at Crespigny Road and St James’s Street, for it was the first real proof that FORTITUDE was off to a good start. GARBO responded by relaying a message on 30 April from CHAMILLUS, the NAAFI waiter, who, on 27 April, had volunteered to leave the depot at Chislehurst for a secret invasion embarkation point. On 29 April CHAMILLUS telephoned GARBO to arrange a rendezvous the following day at Winchester railway station. At their meeting CHAMILLUS had disclosed that his new base was Hiltingbury Camp, near Otterbourne in Hampshire. GARBO reported that, according to CHAMILLUS,
all the 3rd Infantry Division are concentrated here ready to embark. There are other camps full of troo
ps ready for attack. Have identified the 47th London Division in a camp to the south of mine … it is extremely difficult to leave the camp. They are preparing cold rations for two days, also vomit bags and lifebelts for troops’ sea voyage.
The purpose of this particular scheme was to place CHAMILLUS in a position where he might reasonably be expected to let GARBO know when the invasion fleet had put to sea. Both GARBO and Harris were adamant that advance notice of the invasion of just one hour would be enough to enhance his reputation with the Abwehr. The Twenty Committee accepted this idea, with some dissent, but unfortunately the atmosphere soured early the next month and the committee’s attention was temporarily diverted. News arrived from Portugal which suggested that ARTIST’s (Jebsen) case had taken a turn for the worse. On 28 April Jebsen had kept a rendezvous with Graham Maingot in Lisbon, at which Jebsen had disclosed that the Abwehr had become suspicious of TRICYCLE (Dusko Popov) and were convinced that even if he was not now operating as a controlled led agent for the British he probably had been in touch with them at some stage. The relevant part of the Secret Intelligence Service’s summary of Jebsen’s debriefing is reproduced here as it sheds important light on the kind of agonising a case officer had to endure:
Abteilung III has announced that in their opinion TRICYCLE’s material is not controlled by the British. This judgment is entirely on the basis of TRICYCLE’s latest report, and the Abteilung hold to their opinion that in one period of his career TRICYCLE was under Allied control. They explain the change as follows: the Allies provided poor material; ARTIST complained to TRICYCLE that he was not earning his keep; TRICYCLE then decided to collect material himself, and the result has been his last report – so good that Abteilung III is in entire agreement with the general staff that it is inconceivable that the British should have deliberately fed it. From ARTIST’s point of view the outcome is a complete triumph, and he is sure that whatever happens now, these two departments whose confidence it is hardest to win will never reverse decisions so categorically expressed. To crown it all, ARTIST has been awarded the ‘Kriegsverdienstkreuz, 1st Class’, an honour shared by no one in Lisbon.
Operation Garbo Page 14