16

Fortitude, FUSAG and France

‘The enemy must not know where I intend to give battle. For if he does not know where I intend to give battle, he must prepare in a great many places. And when he prepares in a great many places, those I have to fight in any one place will be few.’

Sun Tzu, The Art of War¹

Miss Jane Hughes described it as ‘one of the ugliest country houses ever to have been constructed in England’, but Bletchley Park was always affectionately ‘B.P.’ to those who worked there. Officially, they laboured at the Government Code and Cipher School (GC&CS), but informally, the curious blend of debs (high society debutantes) – of which Hughes was one – and boffins (scientists) called themselves the ‘Golf, Cheese and Chess Society’. She recalled fondly of the war years, ‘We were a mix of beautiful girls and very desirable young men, all in our twenties; class barriers didn’t exist at all. Working for the same goal, we women were treated as equals. They were great days.’²

Established in 1938, by the time Hughes arrived Bletchley had evolved into a production-line factory analysing the signals sent between German military units using their Polish-designed Enigma enciphering machine. She had been recruited by an old school friend in 1940 who wrote to her complaining of the ‘frightful’ workload: ‘We’re so overworked, so desperately busy. You must come and join us.’ Aged eighteen, Hughes passed her interview with the chess-playing Stuart Milner-Barry, one of the senior codebreakers, who was ‘desperately shy. He couldn’t think of a single thing to say and I couldn’t think of anything to say to him because I wasn’t supposed to say anything.’ She signed documents promising never to reveal her work and the ‘loose lips sink ships’ wartime secrecy stayed with her for fifty years. It was only in the 1990s that she came clean to her family: ‘It felt quite overwhelming. I’d never told a soul, not even my husband. My grandchildren were very surprised.’³

It was only in 1974 that the secret of Bletchley Park was disclosed to the world. This was where nearly nine thousand worked, decoding and analysing signals traffic sent by the supposedly secure German Enigma enciphering machine. ‘B.P.’ – as it was known to those who worked there – witnessed the birth of US–UK intelligence sharing. After much prevarication the government eventually surrendered the house and it has become a world-class museum devoted to this aspect of the war. (Author’s archives)

Milner-Barry was one of the many cryptographers, linguists and mathematicians seconded from Oxford and Cambridge as well as the armed forces who gravitated to Bletchley. By 1944, almost nine thousand personnel worked there in three daily shifts. Their contribution to the war – against Italy and Japan as well as Germany – cannot be understated. All Wehrmacht and SS headquarters, even the German police and railway services, used improved versions of Enigma, which was considered impregnable – when they felt security was compromised, the Reich looked for human traitors, rather than mechanical ones. Enigma transmissions were monitored and transcribed by around forty ‘Y’ (listening) stations around Britain, who pinpointed their source and sent the copied signals by despatch rider or teleprinter for processing at Bletchley, known as Station ‘X’. Leading Aircraftman Peter Read worked for the ‘Y’ service aboard Fighter Direction Tender-216, ‘the most ungainly vessel I had ever seen’, which sprouted a variety of radar aerials and antennae, and lay ‘off Dunoon, at the mouth of the Clyde. It was staffed by German-speaking RAF and WAAF personnel of all ranks who listened to Luftwaffe traffic.’⁴

Bletchley’s methodology of signals interception and analysis was assessed as being so sensitive as to be rated ‘Ultra Secret’, or code-named simply ‘Ultra’. Hughes, who had learned German at St Moritz before the war, embraced the station’s social life, joining the choral and Highland dancing societies. However, she recalled, ‘We were always working against time, there was always a crisis, a lot of stress and a lot of excitement.’ In May 1941, Hughes decoded messages during the hunt for the battleship Bismarck. When the BBC announced it had been located and sunk, she recalled ‘a great cheer went up in the dining room’, for this was the first time the codebreaking teams had seen tangible evidence of the importance of their work. In 1942 naval Enigma signals traffic between weather stations, submarines and Kriegsmarine HQ was first deciphered in Hut 8 and translated in Hut 4, where their significance was assessed, allowing U-boats to be pinpointed and destroyed, so turning the Battle of the Atlantic.

Prior to and during the Normandy campaign, German Army and Luftwaffe Enigma traffic was deciphered in Hut 6, where Hughes worked, then translated and assessed in Hut 3. The latter hut then distributed published extracts – with appropriate safeguards to protect the source (usually said to be the well-placed agent ‘Boniface’) – to field commands who needed to know their contents and could use them. Only Allied army group and army headquarters were generally permitted to access Ultra material, which amounted to ‘about twenty-five thousand signals sent out from Hut 3 to western commands between the opening of Eisenhower’s HQ in January 1944 and the end of hostilities’.⁵ While photo reconnaissance missions flown by the RAF and USAAF acted as the Allies’ eyes, Ultra performed the task of their ears, delivering accurate German orders of battle. For example, on 20 April 1944 Guderian’s HQ transmitted the itinerary of a tour of inspection of panzer formations in the West, offering ‘a splendid insight into the German distribution of armour a month before the landings’.⁶

Such was the sensitivity of the project that its very existence was only uncovered when Group Captain F. W. Winterbotham, who trained personnel to distribute Ultra material to senior officers, was allowed to publish his memoirs in 1974 as The Ultra Secret. He wrote from memory and without reference to archives, which the UK Public Record Office in Kew began to declassify soon after. They revealed how crucial Ultra was to the success of Normandy, something earlier memoirs – for example those of Eisenhower, Bradley, Churchill or Montgomery – and all other narrative histories were forbidden from mentioning. It is no understatement to claim that subsequent historians have had to completely revise their understanding of the way the Second World War was fought.

After its finest moments, Bletchley Park was much neglected following the war, partly because of the secrecy under which Ultra operated, and seemed ripe for demolition in 1991.⁷ Hughes, by then Mrs Fawcett, had forged a second career in the Victorian Society, founded to prevent the destruction of old buildings, and knew a thing or two about obstructing such processes. She was almost as proud of rescuing London’s St Pancras railway station – now the departure point for the Eurostar to Paris and Lille – as of helping to save Bletchley: frustrated officials knew her as ‘the Furious Mrs Fawcett’. Today the site, including the main house and the huts where penetrations of Enigma’s secrets occurred, through the lobbying of Hughes and many others who worked there, has been restored and given international recognition for the war-winning dramas that occurred there – various sources claim the Second World War was shortened by anything between six months and two years by the work done at B.P.

In the forthcoming intelligence war, Ultra would not only pry into the Reich’s confidences, but also confirm the success of the Allied strategic-level misinformation campaign, labelled Operation Fortitude, guarding the secret of when and where the invasion of France would take place. When Morgan’s COSSAC had been set up in April 1943, a group of officers within the G-3 staff, known as Ops (B), was given the twin tasks of overseeing physical deception and espionage missions implemented by the ‘controlled leakage of information’. These evolved into Fortitude, which was formally agreed by the Combined Chiefs of Staff on 20 January 1944 and sketched out in a secret directive of 3 February, addressed to Bedell Smith, SHAEF chief of staff.⁸ The deception was designed ‘to induce the enemy to make faulty strategic dispositions in relation to operations by the United Nations [as they were then known] against Germany’.⁹

It had many subsidiary aspects, six of which were known as Operations Quicksilver I–VI. Quicksilver I was the basic distraction of a First US Army Group (FUSAG) based in south-east England ready to invade Calais. For the benefit of the German Army’s signals intelligence units, Quicksilver II consisted of FUSAG’s entirely bogus signal traffic. Quicksilver III was the display of dummy landing craft and other military equipment that would allegedly launch FUSAG. The air plan supporting a FUSAG assault was the central theme of Quicksilver IV, with much bombing of the Pas-de-Calais beach defences prior to D-Day. Increased activity around Dover to suggest embarkation preparations formed the logic behind Quicksilver V, while Quicksilver VI would simulate activity at night where the imitation landing fleet was anchored.

In fact, there were two Fortitudes: the main effort was Fortitude South, which acknowledged that the Allies couldn’t conceal the presence of Montgomery’s real 21st Army Group, but set out to persuade the Germans that a second, fictitious, First US Army Group (FUSAG) – comprising the equally notional Fourteenth US and Fourth British Armies – was waiting in England to invade German-held Europe. It was made more convincing by having a real commander – the most able Allied tactician, in German eyes – George S. Patton. The Germans swallowed FUSAG whole: in late 1943, Ultra picked up a German Foreign Armies West intelligence assessment (more of them later) that there were thirty-four divisions in southern England. However, of the itemised formations, the Allies knew eleven were fake FUSAG units.¹⁰

Fortitude North also came into play, similarly equipped with the ruse of non-existent formations waiting in Scotland, Northern Ireland and northern England to make a smaller landing in Norway. By January 1944, the Germans estimated a total of fifty-five divisions in the British Isles – when in reality there were thirty-seven – rising on 27 May to fifty-seven British, Canadian, Free French and Polish divisions, twenty-two US formations, plus a further ten of armoured and airborne troops – totalling eighty-nine divisions, plus twenty-two other brigades – poised to attack.¹¹ The actual Allied strength at that moment in 1944 was precisely half that, at forty-five divisions.¹²

The Quicksilver I detail was incredible: headquartered in Edinburgh Castle, the fake Fourth British Army was overseen by the personable General Andrew Thorne, who had once been military attaché in Berlin. Its phoney VII Corps in Dundee contained the equally false 5th British Armoured Division, comprised of the bogus 37th Armoured and 43rd Infantry Brigades. Based in Lincolnshire, the British 2nd Airborne Division included the 11th and 12th Parachute and 13th Air-Landing Brigades – not a word of it true. The notional Fourteenth US Army, led by real John P. Lucas, who had recently commanded an army corps at Anzio, was comprised of the mythical XXXIII and XXXVII Corps, based in Bury St Edmunds and Chelmsford, respectively. The US Army’s heraldry department even devised unit shoulder patches for their phantom army, its corps and divisions – some were featured in US magazines available in neutral Spain, Portugal, Sweden and Switzerland – which Quicksilver operatives wore conspicuously in London and elsewhere.¹³

Time and resources were invested on a huge scale in the strategic deception surrounding the Patton Army Group (FUSAG) to make it convincing. As far back as a year before D-Day, details of the notional Fourteenth US Army were published. None of these units existed, but magazine features in the international press, available in neutral Sweden, Switzerland, Spain and Portugal – and double agents – persuaded the Germans otherwise. (US Army National Archives)

Quicksilver I’s success was confirmed when ‘a German map of the British order of battle as on 15 May 1944, which was later captured in Italy, showed how completely our imaginary order of battle had been accepted’. It clearly showed the FUSAG formations in south-east England. Later, ‘a recognition booklet captured in France, which would have been issued to field commanders, included coloured drawings of our notional divisional signs’.¹⁴ Allegedly, the Fourteenth Army’s ranks were filled with bloodthirsty convicts who were released from prisons to serve in the military – information which reappeared in interrogation of nervous German POWs by the British Combined Services Detailed Interrogation Centre. This organisation collected and assessed material from prisoners, and was, in fact, nine ‘cages’ established across the United Kingdom where captured soldiers were questioned by the Prisoner of War Interrogation Section of the Intelligence Corps’ Field Security Police, and agencies such as MI5 and MI9, before being sent to POW camps. One Russian-speaking interrogator consistently produced interesting nuggets of intelligence by wearing Soviet uniform, which frightened those who’d served on the Eastern Front into revealing all they knew. These organisations formed an important, and often overlooked, adjunct to Ultra in checking the effectiveness of the Fortitude ruses.

The convict legend was widely circulated throughout the German Army, surfacing, for example, in the memoirs of Gunner Hein Severloh, with the artillery regiment of the 352nd Infantry Division behind Omaha Beach. On being taken prisoner, he noted one of his captors was ‘an unattractive, tough-looking type with a fearsome Iroquois haircut, and I could see conspicuous prison numbers tattooed on the left side of his skull – the Americans having deployed many hardened criminals in front of the attacking troops’.¹⁵ The tattoos and haircut would have been a voluntary indulgence by a unit before combat, for no federal penitentiary since the nineteenth century has tattooed their convicts, let alone released them for suicidal combat duty – though this was, tellingly, a German tradition, hence being widely believed. Waiting impatiently to cross the Channel, a British soldier remembered his North American neighbours shaving ‘their heads, leaving tufts of hair as diamonds or squares, or sometimes they’d give themselves a Mohican cut’.¹⁶

In addition to the Quicksilver units, several real formations waiting to cross the Channel in follow-on waves were also nominally part of FUSAG, making it a mixture of truth and falsehood. In reality, no more than a few hundred signals troops were actually assigned to FUSAG, who began transmitting wireless details of their fantasy activities on 26 April.¹⁷ Corporal Les Phillips was one of them. ‘You would be given a map reference, take the truck to that spot, and at a certain time broadcast according to a script you were given. It might be asking for reinforcements, ammunition supplies, all sorts of things. You’d pretend to be a whole corps. The timing was very important because it had to seem as if one tank was answering another. You had the audio sound of the tanks, to hear them, too. It was all very secret.’¹⁸ Phillips worked alongside the US Army’s 3103rd Signals Service Battalion.

Most of the details were worked out by Frederick Morgan’s staff at COSSAC, prior to their absorption into SHAEF, who foresaw the need for a large investment of time and resources to craft a convincing and foolproof deception. For Quicksilver III, due to the acute shortage of real landing craft, fake stand-ins, code-named ‘Bigbobs’, were placed in likely south-east England anchorages. Lieutenant Colonel White of the Royal Engineers recalled driving past some at anchor in Hampshire one day: ‘They looked very realistic and at half a mile away I certainly couldn’t tell them from the real thing. We’d leave them for two or three days then move them elsewhere, so it looked as if the troops were moving around.’¹⁹ They were made of canvas stretched over a steel frame, floating on an array of 45-gallon oil drums; building them was very labour intensive, as each kit filled six or seven three-ton trucks, and took twenty men six hours to assemble. Ray Marshall, Normandy bound with the 5th Royal Horse Artillery, recalled embarking with his 25-pounder Sexton self-propelled guns in LCTs and finding ‘the Thames Estuary was full of dummy boats’.²⁰

Meanwhile, inflatable tank and truck kits made of rubber were assembled by Montgomery’s 1,200-strong ‘R’ Force, responsible for tactical deception, and five hundred men of the US 603rd Camouflage Engineers, many of them artists drafted from schools in New York.²¹ Made by the US Rubber Company of Rhode Island, Goodyear and others, the tanks weighed in at ninety-three pounds, and were lifted easily by four men. ‘R’ Force soldier Arthur Merchant remembered being confronted with several mysterious cases of different sizes at an ordnance depot in Essex. ‘One was a machine with two handles that was a sort of pump. We plugged it in and started blowing this rubber thing up. It looked like a giant sausage at first, but it suddenly turned into a 75mm gun barrel. Then we got some more mysterious boxes and had a go at them, and a Sherman tank gradually flipped out and took shape.’²²

Hundreds of other vehicles, artillery pieces and even planes were made by movie prop men at Shepperton Film Studios in west London. Together with the 406th Engineer Combat Company, who handled security and construction, they made the fakes look realistic, providing ‘poor’ camouflage netting, and tent cities around them, with vehicle and tank tracks, using special rolling tools. The camoufleurs remember the plan was nearly unhinged by an irate bull, which suddenly lunged at a column of Sherman tanks parked in ‘his field’. Instead of the thud of horns against armour, the puzzled bovine was confronted by the hiss of air as a tank deflated into a puddle of olive-drab rubber: bull 1, tank 0.

The Quicksilver deceptions were designed to delay the move west to Normandy of Salmuth’s Fifteenth Army for up to seven weeks after D-Day. Once the invasion was under way, the Allies would feed a battle plan piecemeal to the Germans that Normandy was only a feint, and a further assault would take place later on the Pas-de-Calais, with a smaller operation against Norway – Fortitude North. We met John Emery earlier, the signaller on HMS Largs; he spent a while driving up and down the British east coast sending his dummy messages to reinforce this northern ruse.²³

Also part of the Fortitude plan was Operation Copperhead, initiated when British intelligence discovered a lieutenant in the Royal Army Pay Corps who bore an uncanny resemblance to Montgomery. Fortunately, Lieutenant Meyrick Clifton James had been an actor before the war and, duly briefed by MI5 and having studied his subject’s mannerisms and voice, was deployed to Gibraltar and Algiers – at Montgomery’s insistence on a full general’s pay for the duration of his impersonation – in an attempt to distract German attention away from the obvious target of northern France.²⁴ The real George S. Patton, itching to invade the Continent, reciprocated in the world’s newsreels and newspapers, by inspecting troops in south-east England.

The traditional narrative is that Fortitude put the Germans off their guard in Normandy, as confirmed by Ultra. However, as more documents come to light from the war years it becomes apparent that the Germans knew far more about the Allies’ preparations than the traditional Overlord historiography would have us believe.

Headquarters, Signals Intelligence Regiment No. 5 (Kommandeur der Nachrichtenaufklärung.5; KONA 5) does not usually rate an appearance in the accounts of 1944, yet it was the eyes and ears of Rommel’s Army Group ‘B’. It consisted of long- and close-range mobile intelligence-gathering companies which intercepted and decoded Allied transmissions, passing that material back to their regimental signal intelligence evaluation centre. Under its commander, Oberst Walter Kopp, KONA 5’s headquarters then undertook evaluation and traffic analysis, cryptanalysis, collation and dissemination of all intelligence passed to them. Kopp had two jobs, being also attached to Rundstedt’s HQ at OB West, with responsibility for all signal intelligence matters, also copying his findings to OKW in Berlin.²⁵

Kopp’s unit was not intended to solve complex Allied cryptography, along the lines of the Allied effort associated with Bletchley Park, but concentrated on exploiting lower-level ciphers, which yielded plenty to keep them busy. For example, much appears to have come via the M-209-A cipher machine, manufactured by Corona typewriters. Used extensively by US air and ground forces throughout 1943–5, the portable, rotor-based device was not as strong as its Enigma counterpart, and could be read in as little as four hours by the Germans. Between February and May, intercepts of nearly two thousand administrative signals sent by M-209 revealed details of the Allied order of battle. KONA 5 also managed to break some of the Royal Navy’s ‘small ship’s basic code’ used by landing vessels, which confirmed general movement patterns of craft from the Mediterranean and Scotland to southern and eastern England.²⁶

They also listened to British railway signals – just as Bletchley was monitoring Deutsche Reichsbahn Enigma traffic – and claimed ‘a ninety-eight percent success rate in reading over four thousand signals from late November 1943 to February 1944’, which pointed to a general build-up in the south of troops and supplies.²⁷ Another indicator was the absence of test radio transmissions and practice wireless traffic, which ceased altogether from March 1944 – suggesting that offensive operations, rather than exercises, were imminent. However, in the second half of May exercise radio traffic between aircraft control stations aboard warships and personnel was intercepted for the first time; their locations were confirmed by direction finding as Plymouth and Southampton.

As a result, on 1 June 1944, the head of intelligence for Army Group ‘B’, Oberst Anton Staubwasser, noted that the German High Command became convinced that Normandy would be the site of Allied landings.²⁸ KONA 5 also worked out the tactical field code book of the US 29th Infantry Division, stationed primarily in Devon and Cornwall, and were decoding much of their training and administrative traffic in May and June.²⁹ One routine message assessed as highly significant noted the receipt of shipments of whole blood, which was known to have a shelf life of a month.

On another level, General Walter Schellenberg of the Sicherheitsdienst (SD, or security service of the SS) claimed in his memoirs, ‘Early in 1944 we hit a bull’s eye by tapping a telephone conversation between Roosevelt and Churchill which was overheard and deciphered by the giant German listening post in Holland. Though the conversation was scrambled, we unscrambled it by means of a highly complicated apparatus. It lasted almost five minutes, and disclosed a crescendo of military activity in Britain, thereby corroborating the many reports of impending invasion.’³⁰ He was referring to the Allies’ use of the Bell A-3 speech scrambler, used on the Washington–London telephone link during the war. Although the two leaders talked person-to-person, their speech was meant to be ‘guarded’, and monitored in Washington and Whitehall by operators who had, on occasion, to intervene and remind the great men not to divulge specific military details or dates, even over this ‘secure’ line. At most, intercepting scrambled calls would merely have confirmed general Allied intentions.

Further breaches of supposedly secure Allied communications were revealed during a post-war debriefing at Berchtesgaden on 18 May 1945, when Oberstleutnant Friedrich, a senior German signals intelligence officer, claimed that his cryptographers in Paris ‘had intercepted and broken US non-Morse teleprinter traffic between Washington and Europe. This success was maintained throughout ‘owing to the lamentable insecurity of the operators’. The Germans therefore had always known all the details of ferry flights, strength of the US air forces in Europe, and a good deal about training and replacements.’ Pencilled in the margin of the report, an Allied interrogator wrote: ‘I don’t believe this came from M-228 traffic’ – a reference to the SIGCUM, a rotor cipher machine used to encrypt US teleprinter traffic from April 1943. Whatever the source, although Overlord was not directly threatened, the Allied aerial preparations for the invasion would have been clear to German signals analysts.³¹

Schellenberg, head of the SD foreign intelligence department from 1941, confirmed in his memoirs that the SS were running their own signals intelligence service, which had been founded by Reinhard Heydrich in 1931 – in parallel to that of other agencies.³² The German chief air historian observed in 1990 that ‘about twelve major intelligence services existed in Germany. They were run by the army, navy and air force, by the Nazi Party and some ministries, but there was no real evaluation centre or clearing office for all.’³³ Within the Luftwaffe, for example, there were no less than eight intelligence collecting agencies, who rarely cooperated and thus never compared their findings. This was the nub of the German problem: sometimes the quality of their intelligence products was excellent – but they competed Darwinian-style against one another, at the expense of the regime. Schellenberg’s retrospective view, for example, was that their failures were ‘not due to a lack of ability but rather to a lack of historical integration of intelligence into the command structure’.³⁴

The SD had long loathed the Abwehr, the principal German military intelligence service founded in 1929, which they wanted to control. Eventually they were able to prove the anti-Nazi credentials of its director, Admiral Wilhelm Canaris, and his deputy, General Hans Oster, and on 18 February 1944, they were sacked; however, it took until May for the Abwehr to be formally abolished, and for Schellenberg to persuade the regime to let him take over all its functions.³⁵ Many of Canaris’s staff were inherited by Schellenberg, including Oberst Georg Hansen, who became his new second in command. However, embarrassingly for Schellenberg, Hansen was arrested and executed in the aftermath of the 20 July 1944 Stauffenberg Plot, along with Canaris and Oster.

The Abwehr–SD clash was partly a political one, the former playing a long, professional intelligence game against their opponents, the latter being driven by short-term demands for results from Berlin. They additionally trod on each other’s toes, duplicating effort and wasting manpower and resources. Each agency also attracted different kinds of people: the Abwehr being a military organisation, it was led by career soldiers with codes of conduct, who tended to avoid physical violence, preferring psychological ploys, making deals and recruiting double agents – although fewer, and with less success than the British. The Sicherheitsdienst, on the other hand, were less scrupulous, recruiting criminals and freely employing extortion, bribes – and torture.

The Abwehr occasionally used this to their advantage by frightening suspects into talking merely by threatening to send them to the SD. Yet historians have not been kind to Canaris or his Abwehr, noting that it failed to predict the massive mobilisation capabilities of the Soviet Union, the United Kingdom or America. Neither did it anticipate Torch or Husky, or (prior to the Abwehr’s abolition) warn of Overlord looming.³⁶ The Admiral himself is often assessed as being more interested in anti-Nazi intrigue than in his official duties. However, much German intelligence traffic remained opaque to their Allied opposite numbers, as the Enigma key of the SD, though not that of the Abwehr, remained unbroken during the war. This was compounded by the one weakness of Bletchley Park: it was only able to intercept messages sent by radio – in Western Europe the Abwehr, SD and German police forces tended to rely on landlines.

The February 1944 abolition of the Abwehr was also to the general detriment of the Wehrmacht, who had to rely more than ever on their own signals intelligence service, run by the German Army’s chief signals officer. This was Generaloberst Erich Fellgiebel, another anti-Nazi executed after the 20 July Stauffenberg Plot. This had the effect of discrediting his command, whose professional advice was subsequently ignored at the expense of agencies like the SD, who knew how to present to Hitler’s circle a picture they liked to hear, whatever the truth on the ground.³⁷

A senior NATO officer observed of German wartime intelligence,

In NATO the J-3 (Operations) and J-2 (Intelligence) staff organisations are on a superior level, set apart from all other functions, and working as co-equals. However, during the Third Reich, the commander and his chief of staff prepared the situation assessment, with minimal input from the intelligence officer – if he was asked at all. Amidst the rivalry and friction of various headquarters, information was sometimes deliberately withheld, because to know more than others gave one more influence. Intelligence officers were also called ‘defeatists’ if their estimates were too unfavourable, and their products considered as ‘lies’. Self-delusion became the counter to intelligence.³⁸

In Hitler’s mind, because of ideology and the euphoria of the early victories gained in 1939–41, there was a ‘conviction of the superiority of the German warrior over the Anglo-American tradesman’, which eschewed the contribution of intelligence.³⁹

Confusion was increased by the reporting activities of Oberst Alexis, Freiherr von Roenne, the head of the army’s Fremde Heere West (Foreign Armies West), the intelligence department responsible for evaluating intelligence relating to the British Isles – and therefore Overlord. He noticed when his Allied troop estimates were submitted to OKW and Hitler’s headquarters, they were reduced by half to make them more palatable to Hitler. Roenne and his deputy then started doubling their numbers before submitting them, thus playing into the hands of the Operation Fortitude deceivers.⁴⁰ After the war, surviving German officers would make claims about what they, or their departments, knew about the Allied build-up. Such assertions were also used to illustrate the gladiatorial nature of the Third Reich command systems, where chaos and confrontation was the order of the day, rather than collaboration. The implication was usually ‘if they had listened to me and not those other bastards’, the outcome would have been different.⁴¹

The French department of the SD, based at 82–84 avenue Foch in Paris, and directed from 1 June 1942 by the SS and police leader for France, General Karl Oberg, had also penetrated several Resistance movements.⁴² French memoirs and Hollywood often refer to the Gestapo’s sinister activities in France and elsewhere. The moniker was a postal clerk’s unofficial abbreviation of Geheime Staatspolizei (Secret State Police), which had begun life as part of the Prussian police under Hermann Göring, and was expanded to a national agency in 1936 with the task of investigating treason, espionage, sabotage and criminal attacks within the Reich. In 1944 the Gestapo’s D-4 department dealt with such matters in France, whilst its Amt (Department) E-3 was responsible for counter-intelligence in the west.

Throughout the war, the Gestapo mostly comprised German policemen, relatively few of whom were Nazi Party members; they wore plain clothes (unlike Hollywood’s insistence of a fetching two-piece black uniform), possessed police detective ranks, and were charged with making investigations – in France into the Resistance. They usually worked on the same cases as the SD, though the latter were creatures from a different stable. The SD were not policemen, known in Gestapo circles as ‘bad detectives but good Nazis’, and wore grey SS uniform and bore membership of the Nazi Party. The SD and Gestapo were separate agencies but both came under direct command of Himmler, chief of all SS and police in the Reich; as the war ground on, SD and Gestapo personnel were loaned and seconded to each other, the SD effectively becoming the intelligence agency for local Gestapo detachments – and it became often difficult to distinguish between the two – although their combined numbers in France totalled no more than three thousand.⁴³

In Paris, the Gestapo logically took over the headquarters of the French secret police, the Sûreté Nationale, at 11 rue des Saussaies, which was bursting with files on communists, refugees and foreign agents.⁴⁴ However ‘Gestapo headquarters’ is historical shorthand, for both it and avenue Foch housed offices of every branch of the German intelligence services and police. By 1944, the SD and Gestapo wielded the real power in occupied France, cooperating with frightening efficiency – in contrast to the previous SD–Abwehr antipathy – and were represented by the armed sentries outside their many buildings, fleets of black Citroëns at their disposal that flashed through the streets and lanes, spreading terror, and thirty-two thousand informers at all levels of French society. In the eyes of the Resistance and associated Allied saboteurs and agents, the SD and Gestapo were one and the same, and since the war the term Gestapo has been used widely – and incorrectly – to signify both.

All German personnel were instructed to recognise the Resistance as terrorists, rather than legitimate warriors, and termed anti-partisan operations as Bandenbekämpfung – combating banditry.⁴⁵ Thus, with an invasion looming, German reaction to Maquis activity was to be swift and brutal. As General Carl-Heinrich von Stülpnagel, Military Commander of France (destined to be another victim of the post-Stauffenberg hunt for conspirators), ordered on 12 February 1944, ‘The main task in the coming weeks and months is fully to re-pacify the areas which are contaminated by bandits and to break up the secret resistance organisations and to seize their weapons. In areas where gangs form, these must be combated with a concentrated use of all available forces. The objective must be to break up all terrorist and resistance groups before the enemy landing.’⁴⁶

More harsh still was the angry directive issued at the same time by Rundstedt’s deputy, Luftwaffe Field Marshal Hugo Sperrle: ‘We are not in the occupied Western territories to allow our troops to be shot at and abducted by saboteurs who go unpunished. If troops are attacked, all civilians in the locality, regardless of rank or person, are to be taken into custody. Houses from which shots have been fired are to be burnt down. Measures regarded subsequently as too severe cannot, in view of the present situation, provide reason for punishment.’⁴⁷ This encouraged all those in German uniform, not just the SD/Gestapo and wider SS, to employ extreme measures against the Maquis, often in contravention of Hague and Geneva conventions which recognised the laws and customs of war, and to which Germany was a signatory. The immediate consequence was several large-scale punitive operations undertaken by Wehrmacht units before the invasion. These were Operations Corporal, Haute-Savoie, Brehmer and Spring of February–April 1944, which had the dual aim of destroying the Maquis and terrifying civilians from cooperating with partisans in the future.⁴⁸

As the Resistance knew they could expect no quarter from their oppressors, it merely made them more determined. An SD report dated 20 March 1944 observed, ‘continuing observation of enemy air attacks, agent’s activities, and agent’s wireless networks in the occupied areas of the West unanimously and clearly show [partisan] concentrations in the areas Pas-de-Calais, Paris, Tours, Loire estuary and the South coast of France’.⁴⁹ On 1 June, the last Foreign Armies West intelligence summary issued before D-Day – intercepted, but not decoded by Bletchley for ten days – noted, ‘Considerable increase in parachuting of weapons since the full moon of 28 May. Officers in uniform have been dropped in small groups. Since they can hardly stay underground for long, there is reason to regard the period beginning 12 June – the moon’s last quarter – as dangerous.’⁵⁰

The SD were also aware that Resistance members would be alerted to the imminence of invasion in order to enact several missions. Tortue (Tortoise), as its name implied, required the slowing down of German reinforcements with the sabotage of roads, bridges and viaducts. Plan Vert (Green) would see railway stations, switches and lines simultaneously damaged en masse. Plan Bleu (Blue) was concerned with destroying electrical facilities and power lines, whilst Violet involved the cutting of underground long-distance telephone cables.⁵¹

From his office on the second floor of avenue Foch, Dr Josef Götz, head of the SD’s Section IV (Signals Intelligence), and Oberstleutnant Oscar Reile, head of Abwehr Section IIIF (Counterintelligence in France), operating out of the elegant Hôtel Lutétia at 43 boulevard Raspail, had discovered that some Maquis partisans had been instructed to listen to the BBC French language service on the 1st, 2nd, 15th and 16th of each month for coded messages relating to the invasion. They understood that on hearing the first three lines of Paul Verlaine’s 1886 poem ‘Chanson d’automne’ (Autumn Song), partisans should ready themselves for landings within the next two weeks, and monitor the BBC continuously for a second trio of Verlaine’s lines. When this was also received, it meant simply that the assault would take place within forty-eight hours.⁵² However, these much-quoted lines of poetry were not the only indications to prepare. The first coded message of Thursday, 1 June 1944 preceding the Verlaine, warned quite explicitly ‘L’heure des combats viendra’ (The hour of combat is at hand), though other signals that night were more in keeping with the spring: ‘Les fleurs sont très rouges’ (The flowers are very red).

Since September 1941, it was an open secret that the BBC had been broadcasting coded messages to the French Resistance at the behest of Colonel Maurice Buckmaster, the spymaster running the French section of the Special Operations Executive (SOE). Götz and Reile reported on 2 June 1944 that the BBC in London had broadcast the previous evening some ‘125 phrases of which twenty-eight were recognised as advance-warning codes’, indicating their level of penetration of the Resistance.⁵³ In fact they miscounted, for preceded by the four opening notes of Beethoven’s Fifth Symphony (which also spelt out the Morse code letter ‘V’ for victory) and fifteen minutes of war news, the BBC’s French service, Ici Londres! Les Français parlent aux Français, put out 162 messages d’alerte that night.⁵⁴

Hubert Verneret, a nineteen-year-old maquisard in Burgundy, committed the offence of listening to the BBC French service. ‘Although banned by the occupying enemy, the magic of wireless is such that all we have to do is turn a small black Bakelite knob to the long wave frequency of two hundred kilohertz (fifteen hundred metres) to be reunited as brothers, in perfect communion. There is the usual jamming and the words are practically inaudible at times, but we cling to this unclear noise, as if to a friend’s heartbeat, for it brings us every day, from London, a little hope.’⁵⁵

Every evening at 2015 hours (2115 hours in France), after the news in French, Franck Bauer was one of those who took his position for fifteen minutes behind the microphone in one of Bush House’s airless cubbyhole studios to read out and repeat coded phrases that had an air of Alice in Wonderland about them. The twenty-one-year-old had fled his home town of Troyes by bicycle in 1940, and was elated to be waging his war against Germany and Vichy with a Radio Londres microphone. Bauer recalled, ‘The BBC was trusted throughout France for its broadcasts; we helped create Général de Gaulle and the notion of Free France. We had a slogan: Radio Paris ment, Radio Paris est allemand (Radio Paris is lying, Radio Paris is German).’⁵⁶ Bauer and his better-known colleague Pierre Holmes (son of a French mother and English father) did not always read poetry.

After the news on 5 June – which announced the fall of Rome to Mark Clark’s Fifth Army – the local Resistance leader in Bayeux, Guillaume Mercader, was overcome with emotion on hearing ‘Il fait chaud à Suez’ (It is hot in Suez), which he knew meant to begin Plan Tortue (Tortoise), the sabotage of road networks.⁵⁷ Jean Château, an electricity board inspector who organised agents in the Caen area, and his wife were sitting down to an omelette supper when they heard ‘Les dés sont sur le table’ (The dice are on the table). ‘Why – I think this is it,’ Château exclaimed to his wife, knowing he had just been ordered to initiate Plan Vert (Green) – the destruction of railway lines.⁵⁸

Other seemingly random phrases also portended the invasion – in Caen a young teacher with the Organisation Civil et Militaire (OCM) Resistance group, André Heintz, anxiously listened for ‘Ne faites pas de plaisanteries’ (Don’t make jokes). His eyes still well up at the recollection of that prelude to his own liberation, whilst having to evade the attention of local SD chief, Hauptscharführer Harald Heyns headquartered at 44 rue des Jacobins.⁵⁹ Following the BBC broadcast, another maquisard recorded outside his home, ‘Suddenly the village exploded into unusual activity. Despite the black-out, lights suddenly burst on in the windows of apartments and houses. Someone knocked on the door. The order for mobilisation had been given. We were to report immediately to HQ. Right under the noses of the Germans, everyone was alerted. In the darkened streets, silhouettes flitted from door to door, and shadows moved silently down walls and streets.’⁶⁰

Elsewhere on 5 June, ‘Le chapeau de Napoléon est dans l’arène’ (Napoléon’s hat is in the ring), ‘La Guerre de Troie n’aura pas lieu’ (The Trojan War will not take place), ‘La flêche ne percera pas’ (The arrow does not pierce), ‘Messieurs, faites vos jeux’ (Gentlemen, place your bets), ‘Méfie toi du toréador’ (Beware of the toreador) or ‘Les carottes sont cuites’ (The carrots are cooked) each told different Resistance circuits (or réseaux to their French membership) that now was the moment to pull on their boots, grab a weapon or some explosives, slip through the darkness, quietly wake their neighbours to tell them ‘The Hour’ had come, and go about their respective missions.⁶¹

Jean Dacier, a résistant in southern France, recollected, ‘On the evening of 5 June we were all gathered round our little radio, which we had christened “Biscuit” because of the tin in which it was hidden. The Captain had the headphones on and I was almost lying on top of him trying to hear what was being said. Suddenly London announced D-Day was happening. We all jumped around delirious with joy.’⁶² The final message of 5 June, ‘Les enfants s’ennuient sur dimanche’ (The children get bored on Sunday), was addressed to all; it meant get ready, distribute arms, contact your leader and report every move the Germans make: it was a general call to arms.⁶³

However, as the Maquis maintained their own security and discipline, not all their followers were immediately aware of the wider picture, as Jacques Lazare Julius – a young Jewish lad who had joined a Resistance group near Lyon – recalled. He remembered what he did on 6 June, then aged twenty-two – though he had no idea that this was supporting the invasion. ‘One day we were ordered to cause as much disruption as possible – to cut telephone lines, destroy bridges, railway tunnels and railway lines, as well as blocking roads by blowing them up. Up until then we operated under cover of darkness; this was unusual because we were ordered to act in broad daylight – it was only that night the BBC told us that the Allies landed in Normandy.’⁶⁴

Their chiefs had been trained in England. The numerous requisitioned country houses used by SOE for the purpose give rise to the joke that its initials stood for ‘Stately Omes of England’, after Noël Coward’s popular divertissement of the time. Founded by Churchill on 22 July 1940 with instructions to ‘go and set Europe ablaze’, the French sections of SOE (there were two – Gaullist and Communist) supported the many Resistance réseaux to the tune of depositing nearly three hundred agents and four thousand tons of war materiel into wartime France between the springs of 1941 and 1944, but were not the only Allied players in the field.

De Gaulle’s own intelligence services – the 2ème Bureau, subsequently the Bureau Central de Renseignements et d’Action (BCRA), run by Colonel Passy, the undercover name for André Dewavrin, based in London – had an interest. So, too, did those of the Polish government-in-exile; Britain’s MI6 (also known as the Secret Intelligence Service, SIS), controlled by the Foreign Office; MI9 (the wartime service that helped POWs and airmen behind enemy lines), administered by the War Office; uniformed SAS units; Jedburgh, Sussex and Proust teams (which comprised multinational US, British and French two- or three-person intelligence-gathering teams); as well as the American Office of Strategic Services (OSS – which became the post-war CIA): all operated personnel in France and not infrequently trod on each other’s toes.

Despite this assistance some maquisards remained a law unto themselves – and working not just for the good of France. Such duplication of effort in any case led to poor security, which occasionally allowed the Germans to penetrate groups. In the end – despite the protests of an incandescent de Gaulle – Churchill decreed in January 1944 that all réseaux must use British or American radio equipment and recognised codes.

With the Resistance given their specific marching orders, what to do about the rest of the people? Jean-Louis Crémieux-Brilhac with the Free French Forces in London recalled the heated debate about how far to incite the population into immediate acts against the Germans on D-Day. ‘The Communist party – whose Francs-Tireurs et Partisans (FTP) were extremely influential – wanted an immediate general insurrection on D-Day, with workers going on strike and calls to arms across the country. It would have been a very stupid mistake. The Germans would have taken massive reprisals – as indeed they did. The policy we decided on was of a gradual, phased insurrection, developing in accordance with the advance of Allied forces. Nevertheless, we advised that all French men and women should consider themselves engaged in the total war against the invader in order to liberate their homeland. We told the people of Normandy that every minute lost to the Germans was a minute gained by the Allies. A car stuck on the road could delay traffic for ten minutes – and blocking an enemy transport for ten minutes might ensure the success of an Allied operation.’⁶⁵

The 1962 movie The Longest Day made much of Oberstleutnant Helmuth Meyer in the Fifteenth Army’s signals bunker at Tourcoing excitedly noting the lines of Verlaine on 1 June: ‘Les sanglots longs / Des violons / De l’automne’ (The long sighs of the violins of autumn). On 5 June, Meyer heard: ‘Bercent mon coeur / D’une langueur / Monotone’ (Lull my heart with a monotonous languor).⁶⁶ He understood its significance, though not for whom. In fact it was a specific message to Pierre de Vomécourt’s Ventriloque (Ventriloquist) Resistance circuit to commence attacking railways around Tours, rather than all résistants. The Fifteenth Army’s war diary recorded that Meyer immediately alerted Salmuth’s chief of staff, Rommel’s Army Group ‘B’ headquarters, Rundstedt at OB West and Generaloberst Jodl at OKW in Berlin.

This is where the popular understanding of London’s signals to the Resistance is in error; according to the BBC archives at Caversham, the Verlaine lines were merely one of 187 coded messages transmitted on 5 June warning of the invasion within forty-eight hours – more than on any previous evening – and something which Walter Schellenberg in Berlin, Helmuth Mayer in Tourcoing, Josef Götz at avenue Foch and others listening in immediately spotted. If the quantity was not enough of a concern, the Germans recognised that at least fifteen of these confirmed the imminence of invasion.

The Fifteenth Army was placed at readiness; Generalleutnant Hans Speidel, deputising for the absent Rommel, did not immediately issue an alert – although he later relented – and thus had not warned his other formation, Dollmann’s Seventh Army. Rundstedt did nothing, assuming Speidel had the matter in hand. Jodl did not issue an alert, assuming Rundstedt had done so – which highlighted the individualistic approach of different commanders to their intelligence services.⁶⁷

Schellenberg also sent a warning up through the SS chain of command – but why did Army Group ‘B’ not react? Because OKW had already issued an invasion alert on 18 May: ‘Main effort in Normandy, with secondary attack in Brittany. It is thought that the enemy will attack land targets with very heavy bombs, concentrating on small areas, and will try to silence coastal defences in the same way, combined with heavy naval bombardment and attacks from the sea. The use of new weapons cannot be excluded. Large-scale paratroop drops may be made after nightfall.’⁶⁸

The earlier alert, made on the Kriegsmarine’s advice, came to nothing, and infuriated the Desert Fox, who vowed that no more false alarms would waste his time. May had also seen the SD intercept BBC coded phrases addressed to the Armée Secrète in Le Havre, and other groups in Brittany and the Lille–Amiens area, ordering them on full alert in readiness for the invasion – again illustrating the SD’s success in penetrating the Maquis – though these later proved to be an error made by the BBC.⁶⁹

Meanwhile, Buckmaster had a final trick up his sleeve: not all the messages personnels transmitted by the BBC to France were real. His French operatives had been encouraged to let slip that any broadcasts containing references to soup indicated an Allied interest in the Calais region. As Buckmaster recorded in his autobiography, ‘we broadcast endless sentences of the order of Monsieur Gerald aime le potage (Mr Gerald loves his soup), Caroline demande bouillon (Caroline asks for broth), and so forth’. It left the Germans scratching their heads and wondering whether to shuffle more troops towards Calais.⁷⁰

Militarily, Cherbourg seemed to beckon to the Germans as a likely option for invasion, because of its transatlantic port facilities, and by May 1944 some attention had shifted away from Calais towards the Normandy–Brittany area. Assuming the construction of their extensive fortifications in the Calais area – with substantial military reinforcements behind them – would act as a deterrent, Wehrmacht staff work had shifted by early 1944 to scoping alternative landing sites. Attention focused on ports like Le Havre, Cherbourg, Saint-Malo and Brest, capable of handling the logistics needs of an invading army – though the idea of an artificial, towable harbour was never seriously entertained. By June 1944, all likely anchorages were heavily defended by fortress troops, and a spider’s web of artillery housed in concrete bunkers, which were daily being increased – for by 6 June, the Atlantikwall was merely ‘work in progress’.

At his debriefing in July 1945, General Walter Warlimont, deputy chief of the Armed Forces Operations Staff under Jodl, claimed the late transfer of German units to Normandy and Brittany was a response to intelligence reports showing the concentration of Allied forces in south-east and south-west Britain.⁷¹ He was referring to the move in mid-May of Falley’s 91st Luftlande Division, with Heydte’s 6th Fallschirmjäger Regiment and 100th Panzer Battalion (armed with captured French light tanks) and other units, which caused huge upset to Allied planners when they discovered it.

However, as we have seen, this was in fact a response to Generalleutnant von Schlieben’s request for reinforcements and Rommel’s own appreciation of his defences, and not due to intelligence reports. In another post-war interrogation, Generaloberst Alfred Jodl observed that ‘signals intelligence provided very little definite information regarding the invasion before D-Day. They had not much idea where the Schwerpunkt [main target] was going to be, but thought it would probably be at Cherbourg, with a second attack in the Pas-de-Calais’⁷² – which appeared to reflect Rommel’s own anxieties.

Meanwhile, in early 1943 the very efficient Luftwaffe signals intelligence service at Asnières-sur-Seine (north-west of Paris)⁷³ monitored the pre-invasion Exercise Spartan and, in 1944, Tiger, both compromised by clear-speech conversations. Later in 1943, they worked out that call signs used by Brereton’s US Ninth Air Force had been altered to correspond to those used by Coningham’s British Second Tactical Air Force, suggesting the two organisations had merged (which they had, as the Allied Expeditionary Air Forces, under Leigh-Mallory).

By the middle of May 1944, Luftwaffe signals monitors had concluded that both Elwood ‘Pete’ Quesada’s IX (comprising the P-47s and P-51s that would support Bradley’s US First Army in Normandy) and Otto P. Weyland’s XIX (similarly assisting Patton’s Third Army) Air Support Commands had been redesignated Tactical Air Commands and changed their frequencies to those used by the Ninth Air Force. Traffic analysis showed a large concentration of squadrons in the Middle Wallop area of southern England, and a smaller one in south-east England. Similarly, Luftwaffe traffic analysis and direction finding indicated in April and May 1944 that most units of the British 2nd TAF were transferring to the Portsmouth–Tangmere and Reading–Odiham areas of southern England.⁷⁴

Statistical analysis, codebreaking and direction finding by Asnières also contributed to breaking the simple codes used by the US IX Troop Carrier Command’s C-47 fleet, which revealed the formation of three subcommands (the 50th, 52nd and 53rd Troop Carrier Wings), bunching in the three locations of Grantham–Cottesmore (in fact, the concentration area of the US 82nd Airborne Division), Aldermaston (British 6th Airborne) and Exeter (where the 101st Airborne were gathering). They also concluded that the C-47 fleet was over one thousand aircraft strong, which could enable a mass airborne drop of eighteen thousand fully-kitted paratroops in a single lift up to six hundred miles distant. This was an accurate assessment, for the Allies had available 1,200 C-47s and 1,400 gliders. Based on the known range – 1,200 miles – of a fully-loaded C-47, which was capable of carrying eighteen fully laden men, and using maps and dividers, Luftwaffe analysts began to triangulate likely areas for paratroop operations, of which the Cotentin peninsula was the most obvious, with Caen, Dieppe, Calais, Antwerp and Paris assessed as also at risk.⁷⁵

Some RAF codes had been compromised, too. Through cryptanalysis of the codes belonging to Air Vice-Marshal Leslie Norman Hollinghurst’s No. 38 Group (which operated troop-carrying aircraft and gliders), traffic analysis and direction finding, the Luftwaffe’s signal intelligence service was further able to surmise that group’s known activities – taking agents and SAS operatives into occupied Europe and airdropping weapons and equipment to Resistance movements – had increased markedly from February 1944, and that they were additionally training in glider operations (also revealed by clear speech transmissions such as ‘Have you the glider in tow?’) and the dropping of paratroops.

They were traced as having moved to the rough area of West Oxfordshire, West Berkshire and South Gloucestershire (the locations of their bases at RAFs Brize Norton, Fairford, Harwell, Keevil and Tarrant Rushton).⁷⁶ The conclusion of all these indicators was that the Luftwaffe SIGINT (Signals Intelligence) headquarters at Asnières issued an invasion warning at the end of May: ‘All preparations by the British and American Air Forces are complete. Two British and two American close support corps for the support of four armies are available. The beginning of a large scale landing must now be reckoned with any day.’⁷⁷

However, while they clearly identified the allied assembly locales, did the Germans manage in any way to identify the invasion target area? Both Rundstedt and Rommel felt the key ‘combat indicator’ of where and when would be found in England’s south coast harbours. Luftwaffe aerial photo reconnaissance was rarely able to penetrate the thorough defences around Britain. During the months of April and May the Luftwaffe had managed to fly only 120 reconnaissance missions over Britain, recording little of interest except on 24 April 1944, when a successful sortie by Aufklärungsgruppen (Long-Range Reconnaissance Groups) 121 and 122 equipped with Me 410s flying from bases near Paris, detected ‘234 LCTs, 254 small and 170 auxiliary landing boats and fifteen transports in Portsmouth, Southampton and Selsey Bill’, judged ‘capable of transporting seventy thousand men’.⁷⁸ Frustratingly, a subsequent Messerschmitt Bf 109 sortie managed to take pictures of Portsmouth harbour shortly before D-Day but the aircraft was damaged beyond repair on landing in France.

However, a further mission was able to photograph the south coast, using one of three captured P-47 Thunderbolts of the Luftwaffe’s élite Kampfgeschwader (Bomber Wing) KG-200, who also operated a pair of captured B-17s. Collectively the imagery seemed to suggest a coming offensive against Cherbourg rather than Calais.⁷⁹ Luftwaffe reconnaissance flights made at night, however, told a different story. Pilots reported ‘a great deal of road traffic in south west England, apparent because some drivers were careless about lights’, though confirmatory daytime flights were impossible. ‘Air reconnaissance has provided no information,’ complained the Army Group ‘B’ weekly summary dated 21 May 1944, and one of 5 June pleaded, ‘Survey urgently needed of harbour moorings on the entire English south coast by air reconnaissance’ – but on the key date, bad weather confined the Luftwaffe to its airfields.⁸⁰

On 2 May 1944, Rundstedt had received a message from Hitler which concluded: ‘The Führer thinks the attack will first be on Normandy, and then on Brittany.’⁸¹ However, Rundstedt knew that Hitler was mercurial and disdained intelligence. In his view, the Führer’s conclusions were based on nothing more concrete than a gut instinct. Personally convinced the Allies would attack Calais, the OB West decided to overlook this latest whim of his boss. Less easy to ignore were two Luftwaffe signals intelligence appreciations of 8 May (both intercepted by Ultra with the first decoded immediately, the second a week later), which were based on solid military detective work. ‘Based on Allied air attacks on railroads and waterways,’ read the first with deadly accuracy, ‘our assessment that the landing is planned in the area Le Havre–Cherbourg, is confirmed once more.’⁸²

The second document of the same date was more pithy and summarised the recent 2–4 May ‘Anglo-American landing exercise Fabius in the area of the Isle of Wight’ [probably Fabius II, and clearly the subject of careful monitoring by English-speaking linguists], where ‘the enemy, in view of the outer beach obstacles known to him, is attempting to achieve a modified landing and battle technique for his foremost landing wave’. Despite the best efforts of SHAEF, the Germans knew of Fabius in detail, though, curiously, reaction to the 28 April E-boat attack during Exercise Tiger never surfaced in their signals traffic, and its significance as a major pre-invasion practice seems to have completely escaped the Wehrmacht. Yet security surrounding the Allied plan was far from watertight – in fact, it had been all but compromised.

In depth, the 8 May report continued,

Invasion preparations by the Anglo-Americans in the English Motherland are completed. The observed concentrations of landing shipping space, especially in the area north of the Isle of Wight, Portsmouth–Southampton, give a clear picture of a main concentration defining itself in that area. Tonnage of shipping space for landings which has so far been observed can be assumed to be sufficient for twelve to thirteen divisions. The point of main effort appears to be roughly from Boulogne as far as Normandy inclusive. In this connection, the enemy’s chief concern must be to gain possession of large harbours. Of primary importance would be Le Havre and Cherbourg, and of secondary importance Boulogne and Brest. The enemy landing exercise which took place most recently indicates that the enemy attaches special importance to recognising and clearing the outer-beach obstacles at low water, with attacks coming not only on the incoming tide before dawn, but also later.⁸³

In fact, the shipping space was wildly overestimated, as fewer than half this number of divisions were preparing to land, and the follow-on troops were anchored elsewhere.

Thus, from many trusted intelligence sources, it was obvious that a major landing operation would come soon in the west. Hein Severloh, behind the future Omaha Beach, wrote to his sister on 28 April 1944, ‘All leaves are suspended as of yesterday. Something is going to happen here.’ He later recorded seeing two top secret signals addressed to his battery commander on 30 May and 1 June. The first read, ‘Ship concentrations in southern English harbours’, the second, ‘The ships in the English harbours are being loaded.’⁸⁴ Subsidiary operations were anticipated in Norway and southern France, but the main effort was expected along the northern coast of France, from Le Havre to Cherbourg. However, the real question would turn out to be whether the Germans trusted what all their intelligence agencies were clearly telling them – that it would be Normandy.