by Nigel West
Garby-Czerniawski reported his new appointment at FUSAG on 18 May, explaining that his assignment was to act in a liaison capacity with local Polish communities in areas recently liberated, a clear indication that he would be heading for the Pas-de-Calais as the largest Polish population in France was located in and around Lille. A week later the Abwehr responded, ‘All my congratulations on your post. It is a pity that you will not be able to stay in London.’
On 4 June, by which time BRUTUS had disclosed the component parts of FUSAG, the Abwehr requested further information: ‘Give details of the composition of the 21st Army Group and the location of the various headquarters.’
The decisions to isolate the Polish airman from FUSAG’s headquarters, and deny him instant access to his transmitter, were intended to help avoid the pressure of responding immediately to awkward demands from the enemy, and that policy remained in force until after the invasion.
Of all the BODYGUARD spies, the lion’s share of the burden fell on to Juan Pujol, a Spanish businessman code-named GARBO by MI5, who had been brought to London in April 1942 and thereafter had established himself, and his network of twenty-four notional sub-agents, as the Abwehr’s star source, code-named ARABEL. Pujol was a natural for FORTITUDE because of the sheer scale of his contact with Madrid. He transmitted from central London for up to three hours a night, sent mail to cover addresses in Madrid and used an air courier to deliver letters and packages in Lisbon. He was also well paid, having devised an ingenious means of receiving cash from fruit merchants in London who had been credited in Spain by the Abwehr. Furthermore, the members of his imaginary network were distributed right across the country, allowing coverage virtually everywhere. On 10 May, GARBO described a conversation with an unnamed American logistics NCO based at an US Army Communications Zone (COMZ) headquarters in London whom he had recruited in early November the previous year and who had said:
that the second front would open soon as the two Army groups destined for operation were ready. One of these, the 21 Army Group, is under Montgomery. The other, the First Army Group, is provisionally under the order of Bradley. The American troops which are expected here will enter the latter Army Group. He assured me that Eisenhower would give a very important task to the American Army Group.
This exceptionally well-informed American sergeant had been invented by MI5 to give GARBO access to the vital American dimension to the FORTITUDE SOUTH deception. According to the bogus narrative prepared by GARBO and his MI5 case officer, Tommy Harris, the NCO worked for the US Army Services of Supply (SOS) and had met ‘Fred’, one of GARBO’s sub-agents who supposedly was CHAMILLUS, a Gibraltarian waiter in a Soho restaurant. The background to the relationship with the boastful clerk was typically ingenious, and not uncomplicated, but fulfilled the objective of acquiring a relatively low-level, plausible source who could be expected to have sight of documents usually restricted to much more senior personnel. As a filing clerk, the sergeant routinely dealt with papers far above his rank, which made him an ideal candidate for Ops (B). His introduction to the Abwehr by GARBO was through the expedient of passing information to Madrid that the Germans already knew to be accurate. Accordingly, the sergeant confided to Fred in early December that some landing craft had recently returned to British ports from the Mediterranean, the implication being that a major amphibious assault was intended. This information was true, and ISOS had shown that an Abwehr observation post in Algeciras had spotted the vessels on ships transiting the Straits of Gibraltar and reported the news to the KO in Madrid, which had relayed the message to Berlin. When CHAMILLUS reported the sergeant’s indiscretion, Berlin was satisfied that his gossip was true, thereby appearing to confirm his bona fides. The fact that he did not work at FUSAG meant he would not have to answer every question put to him by GARBO, who was very aware that discretion was required because the NCO was an unconscious source, and emphatically not a recruited agent.
On 17 May GARBO received a pressing questionnaire:
Where is the headquarters of the 21 Army Group, English? The numbering of the armies within the said Army Group, and their headquarters. How many and which divisions are within each of the armies of the said Army Group indicating where possible, which divisions are armoured and which are infantry.
Ten days later, on 27 May, GARBO responded, after supposedly having had a conversation with his American NCO friend:
I questioned him but it seems that he does not know much about this formation. He was only able to say that there are few American troops in it, but that the Americans are mainly in the First Army Group.
This NCO, described by GARBO as ‘sociable, jocular and fairly talkative’, was also ‘anti-Communist and, to a lesser degree, anti-English imperialism, following in part the ideas of Randolph Hearst, sustaining an admiration for Franco as Catholic crusader and first leader in the struggle against the Bolshevik.’ However, because he was an unwitting source, GARBO had to exercise discretion in the questions he put to him.
When in April 1942 SIS rather grudgingly handed control of Pujol to MI5’s Spanish section, designated B1(g), thus changing his code name from BOVRIL to GARBO, he came fully equipped, carrying copies of the thirty-eight letters he had written to the Abwehr since July 1941, and with a ready-made ring of sub-agents, although one, a Swiss named William Gerbers, would have to succumb to a sudden illness because of his inconvenient vantage point in Bootle just when convoys were assembling in the Mersey for Operation TORCH, the invasion of North Africa. As a consolation, his widow was granted a pension and allowed to help Pujol as his cipher clerk.
Supposedly, GARBO had found work in London in the Spanish section of the Ministry of Information (MoI), where he drafted propaganda material and cultivated a colleague, Anthony Gordon-Bright, who supposedly believed him to be a republican sympathiser. On that basis he was entrusted with access to confidential documents and was introduced to a wide circle of friends who occasionally indulged in some indiscretion. From MI5’s perspective, GARBO’s relationship with Gordon-Bright, who really existed and on two occasions in 1944 had travelled to Madrid to stay at the Palace Hotel, so could be expected to be known to the Abwehr, was a calculated risk, but he was never indoctrinated into the secret or given any hint that his name and role had been ‘borrowed’ for nefarious purposes.
Usually GARBO confined his activities to organising his network, making personal observations of troop movements, and distilling his own opinions. However, in February 1944 he stole the draft of a pamphlet urging co-operation with the Allies that he had helped prepare that was to be dropped to the French population in the event of a German withdrawal.
Similarly, on 13 June he stole an MoI document that he had been asked to burn with some other classified papers. Instead of placing the material in the office furnace, GARBO had removed one item, which purported to be the minutes of a meeting held a month earlier, on 10 May, by the War Cabinet’s OVERLORD planning sub-committee attended by the MoI’s director-general, Sir Stephen Tallents. The pages purloined by GARBO were entitled ‘Railway Loading in Connection with PLAN MARS’ and described the Second Front, OVERLORD, as having been divided into two operations, code-named NEPTUNE and MARS, which would both have a major American component. However, a crisis was alleged to have arisen because of the burden imposed on the rail system by the logistical requirement to support these two large concentrations, while simultaneously maintaining supplies to the capital. The unspoken implication was that England’s transport network was finding it hard to cope with feeding London while competing with the needs of tens of thousands of troops concentrated in the same region, the south-east of England. This tale was typical of the methodology adopted by Ops (B) to feed the enemy analysts’ diverse items that, when connected, added up to a thoroughly misleading picture.
‡
Rather less satisfactory was Natalie Sergueiev, code-named TREASURE by MI5 and TRAMP by the Abwehr, a mercurial, temperamental French journalist of Russian parentage who arrived from
Lisbon in December 1943, having been recruited by Emil Kliemann of Eins-Luft in Paris. In March she returned to Portugal to attend a rendezvous with Kliemann, who gave her a detailed questionnaire relating to D-Day preparations, and a wireless transmitter. She then flew back to England, to her lodgings in Kensington, but her role thereafter was limited because she blamed MI5 for the death of her beloved dog, Babs, who had been quarantined in Gibraltar. Having threatened to reveal her duplicity to the enemy, she fell largely silent during May 1944, making only a couple of heavily supervised transmissions to her controller in Paris, and none after 17 May. On that Wednesday she had reported, while returning from the weekend in Bristol, a sighting of the 50th Infantry Division just north of Southampton, following its transfer from Scotland. This snippet would be circulated on 22 May by FHW with the comment that the source was ‘trustworthy Abwehr messages which have been confirmed by “Y”.’ Although not formally terminated as a deception channel in December, this was TREASURE’s last significant contribution to FORTITUDE, the genuine excuse being her hospitalisation with a diagnosis of kidney stones.
Finally, the other two BODYGUARD spies, based in Reykjavik, were COBWEB and BEETLE, who were to support FORTITUDE NORTH, the threat of a planned invasion of Norway. COBWEB was Ib Riis, landed from U-252 in April 1942, and BEETLE was Peter Tomsen, delivered by U-279 in September 1943, but they were unaware of each other and acted independently, managed by SIS. As purveyors of strategic deception, they reported on the non-existent US 55th Infantry Division, and in April 1944 COBWEB declared that this American unit had been placed under the command of the British 7th Corps, headquartered in Dundee. Although he had not said so explicitly, FHW interpreted this news as an indication that the division had actually moved to Scotland, and circulated this development on a map attached to an assessment dated 15 May, albeit tagged with a question mark. On that day the OKW released an assessment, numbered 27, which concluded that ‘… the Schwerpunkt [focus] of the enemy concentrations has shown more and more clearly to be in the south and south-east of the British Isles …’
A fortnight later, on 31 May, the OKW produced a further review, numbered 30, highlighting the shift:
Further transfer of formations to the south and south-east of the British Isles again emphasises that the main point of enemy concentrations is in this area.
The main thrust of the FORTITUDE campaign would be delivered by the agents equipped with transmitters, but there were others who worked in supporting roles and relied on the regular mail. The first objective, of course, was to provide the enemy with an order of battle, and this had added significance when eight Allied divisions were shipped back to England from Italy. The US 1st and 9th Infantry Divisions, the 2nd Armoured and 82nd Airborne Divisions, the British 50th Infantry, 7th Armoured and 1st Airborne Divisions slipped into the country under radio silence, although by January the disappearance of the 50th British Infantry and the 51st Highland Division had been detected by the OKW and prompted a question sent to BRUTUS and others on 11 January 1944: ‘What are the British divisions which have been withdrawn from Italy and have come back to Britain?’
The response was that only selected personnel had been transferred from Italy for training purposes, and this explanation was accepted by the OKW, which circulated on 21 January:
Independent reports about the transfer of parts of the 8th British Army from the Mediterranean to England continue. The possibility must therefore be considered that individual formations which have not appeared in Italy and North Africa for some considerable time have been transferred wholly or in part to the British Isles.
By 7 March FHW was still speculating about the whereabouts of the British 7th Armoured Division, but the issue was settled by GARBO, who referred in a letter dated 16 February to the fact that he had seen soldiers of the US 1st Infantry Division in Portland harbour in Dorset.7 This sighting resulted, twelve days later, in a comment by FHW: ‘The reported appearance of traces of the 1st American Infantry Division, hitherto assumed to be in the western Mediterranean, still lacks confirmation.’
Then, on 2 March, referring to the same item, OKW commented that a:
well-regarded source reports the appearance of weak forces of the 1st American Infantry Division in Great Britain. This division was hitherto assumed to be in the Western Mediterranean. It may be that these are remnants since the 1st Infantry Division was in Great Britain before being sent to the Mediterranean. Since there is no further reliable evidence for the transfer of this division from the Mediterranean to Great Britain we must for the present accept it in an unknown location.
The Abwehr’s ‘well-regarded source’ was GARBO, who had described his recent visit to the south-west of England in a letter dated 16 February, in which he had mentioned the American unit:
In Portland harbor I saw some American soldiers with the number one in red on khaki ground in the neighbourhood of the town. I learned that the insignia belongs to the 1st American Division.
Known as ‘the Big Red One’, the 1st Infantry had participated in the TORCH and HUSKY invasions, and had returned to England in November 1943 to spearhead the landing on OMAHA Beach. When the 1st Infantry had arrived in England in August 1942 it had stayed at the small Dorset town of Beaminster, and fifteen months later had been billeted in nearby Dorchester. The effect of this preparatory work was to make three divisions, the British 7th Armoured and 1st Airborne, and the US 82nd Airborne Divisions, simply disappear from German assessments, while the other five were mentioned variously by FREAK, GARBO and TREASURE, thereby creating the ‘mosaic’ effect, with multiple sources contributing individual, but apparently unco-ordinated, pieces of a jigsaw puzzle, ending up with the desired picture. In this case, it was the confirmation of five specific divisions, and the removal of three.
‡
Throughout the conflict the Allied deception planners were aware, through the ISOS and BJ diplomatic decrypts, that they did not enjoy a complete monopoly on the information reaching FHW, and there were some sources that could not be controlled. Some of this material came from captured Allied PoWs, who unwittingly disclosed data while under field interrogation; some intelligence was gleaned from captured documents, but the major risk was represented by uncontrolled agents. In particular, there were two networks managed from Portugal that gave considerable cause for concern. One was a multi-lingual Czech journalist, Paul Fidrmuc, code-named OSTRO, who was strongly suspected by MI5 of being a fabricator.8 Fluent in German, Spanish, Danish, Italian, Portuguese and English, he may have been an unscrupulous peddler of invented intelligence but, as GARBO had proved in the period before he had come under British supervision, such individuals may occasionally hit a raw nerve and make an embarrassingly accurate guess. For example, in early June 1944 he told the Abwehr that the Allied invasion would take place on the Cherbourg peninsula, a prediction that was greeted with dismay in London. However, not all his guesswork was unhelpful. On 6 July 1944 he reported:
All indications point to the fact that the trans-shipment to Normandy of divisions already assembled will proceed, other landings having been abandoned. Large-scale movements to South-East England have taken place, as transports by day however so that they were obviously intended as deception manoeuvres, especially as some of the troops have already been transported back to embarkation points area Bognor Regis–Weymouth.
While OSTRO was alleging that other future attacks had been abandoned, which rather contradicted FORTITUDE SOUTH, he was at least conforming to the fiction of major troop concentrations in the south-east. He did much the same again on 19 July when he reported:
General Patton’s headquarters was transferred to Dorchester, King’s Arms Hotel, on 11th July. Simultaneously, United States troops have moved from South-Eastern to Southern England. Patton’s Army has, since the beginning of July, been regularly supplying troops to the First U.S. Army in Normandy.
This second message was almost wholly inaccurate, as General Patton’s pretend command was FUSAG, and
not the US First Infantry Division, which had embarked at Weymouth the previous month for OMAHA Beach. Although there had been some 80,000 American troops massing in Dorset prior to D-Day, not many were left in mid-July, and Patton was in London, not in Dorchester’s famously historic hotel, which was built in 1720. Yet, despite these telltale errors, OSTRO was in effect supporting the proposition that major forces remained in southern England, even if FUSAG had not been committed to Normandy, as asserted. To that extent, OSTRO was not undermining FORTITUDE, as perhaps he might have done, but his industry was a continuing anxiety for Ops (B). Between January and June 1944 twelve of his messages reached the OKW’s intelligence chief Colonel Friedrich-Adolf Krummacher, and in the month after D-Day there were a further nine.
Born in 1898 in Lundenburg, 50 miles north of Vienna, Fidrmuc had served in the First World War as a reserve officer and then started a university course in Vienna studying Philology before a family financial loss forced him to join a business exporting steel in Lubeck. He then married a Danish woman and in 1935 he was working as a journalist in Berlin, where he was arrested on a charge of espionage. He was released, but later boasted to have acquired advance information about Nazi intentions to annex the Sudetenland and Austria. In May 1939 he became a member of the Nazi Party and the following year moved to Copenhagen just before the German occupation, and then to Lisbon, under the Abwehr’s sponsorship. Analysis of OSTRO’s reporting in the ISOS traffic suggested that he was a mercenary who compiled his material from open sources, such as newspapers, and sheer guesswork. The fear was that he might contradict a controlled agent, and might even hit on the truth, as he did on 31 March 1944 when he predicted that the invasion was scheduled for 6 June.