by John Keegan
Hitler was only partially deluded. On 4 and 20 March and 6 April he alluded to the likelihood of a Normandy landing. ‘I am for bringing all our strength in here,’ he said on 6 April, and on 6 May he had Jodl telephone Günther Blumentritt, Rundstedt’s chief of staff, to warn that he ‘attached particular importance to Normandy’. However, apart from allocating Panzer Lehr and 116th Panzer Divisions to Normandy in the early spring, he made no decisive alteration of OB West’s dispositions; indeed until he allowed divisions to cross the Seine into Normandy from the Pas de Calais at the very end of July, he himself remained prisoner to the delusion of a ‘second’ invasion throughout the crucial weeks of the Overlord battle.
His concern to back both horses nevertheless compromised Rommel’s urge to disperse the mist of deception by direct assault. Rommel’s argument was that it was better to have some armour on the right beach, even if the rest was wrongly disposed, than to keep armour in central reserve and then fail to move it when Allied airpower descended. At the end of January 1944 he was translated from the post of inspector of the Atlantic Wall to commander of Army Group B (Seventh and Fifteenth Armies), as Rundstedt’s direct subordinate for defence of the invasion zone. Almost at once he fell into dispute with his chief. Rundstedt had never experienced a battle in which the Luftwaffe was not dominant. He therefore believed that there would be time, even after the enemy landing craft had arrived, to make a deliberate assessment of the military situation and then commit reserves to a counter-attack. Rommel knew that an unhurried counter-attack would be destroyed by enemy aircraft. From personal experience in Egypt and Tunisia he knew how great was the power of the Allied air forces and was convinced that only by holding armour ‘forward’ and committing it immediately could the invasion be met and defeated.
The Rommel-Rundstedt dispute, in which personal experience favoured one general, conventional military wisdom the other, eventually reached the ears of Hitler. He resolved it on his own terms, to neither subordinate’s liking, when the two visited him at Berchtesgaden on 19 March 1944. Panzer Group West, which oversaw the six armoured divisions of Army Group B, was split; three of its divisions were allocated to Rommel, three to Rundstedt – but with the proviso that Rundstedt’s divisions (21st, 116th and 2nd) were not to be committed without the direct approval of Hitler’s operations staff at OKW, with the attendant risk of even greater delay than Rommel had feared in the first place.
As the 21st Panzer Division was the only armoured division close to the beaches chosen by the Overlord planners, Rommel’s intention to launch a quick counter-attack was thus compromised from the start. Montgomery, his old desert opponent, had warned on 15 May in his pre-invasion assessment:
[Rommel] will do his level best to ‘Dunkirk’ us – not to fight the armoured battle on ground of his choosing but to avoid it altogether and prevent our tanks landing by using his own tanks well forward. On D-Day he will try (a) to force us from the beaches; (b) to secure Caen, Bayeux, Carentan. . . . We must blast our way onshore and get a good lodgement before he can bring up sufficient reserves to turn us out. . . . While we are engaged in doing this, the air must hold the ring and must make very difficult the movement of enemy reserves by train or road towards the lodgement areas.
Had Montgomery known, at the time he wrote this assessment, how grievously the Rommel-Rundstedt-Hitler dispute on armoured deployment had harmed the Westheer’s prospect of defeating the landing force, his fears for the successful outcome of D-Day would have been greatly relieved.
Montgomery was appointed to the command of the landing force only on 2 January 1944. Until the Stalin-Roosevelt-Churchill conference at Tehran in November 1943 no commander for Overlord had been nominated at all. Both the American and British chiefs of staff, General George Marshall and General Sir Alan Brooke, had been promised the appointment by their heads of government, though since August Brooke had known that for reasons of international politics it must go to an American. However, it was only at Tehran that the issue of nomination had been brought to a head. Stalin had there made it the test of Anglo-American dedication to the alliance’s Second Front. ‘Do the British really believe in “Overlord”,’ he had asked, ‘or are they only saying so to reassure the Soviet Union?’ In the face of Churchill’s protestations of commitment, he demanded that a commander be nominated not later than one week after the conference ended. Churchill acquiesced and Roosevelt agreed to make the choice. On 5 December, however, at the end of the time limit, Roosevelt recognised that he could not spare his helpmate, Marshall, from Washington, and told him so; the Supreme Command of the Allied Expeditionary Force would therefore go to Eisenhower. Because Eisenhower’s talents were strategic rather than tactical, however, operational authority would be vested in a ground commander, Montgomery, until the ‘foothold’ on the soil of France had been consolidated into a ‘lodgement’ from which the Wehrmacht could not displace the Allied liberation army.
Montgomery, arriving in England direct from Italy where he had been commanding the Eighth Army, threw himself into the rationalisation of the Overlord plan with an energy, familiar to his staff in the Mediterranean, that left the COSSAC headquarters breathless. General Sir Frederick Morgan, Chief of Staff to the Supreme Allied Commander (Designate), had been putting together a scheme for a landing in north-west Europe since the Churchill-Roosevelt meeting at Casablanca in January 1943. COSSAC’s proceedings had not been dilatory; but they had been deliberate. Morgan had set himself the task of presenting the Supreme Commander, when nominated, with a flawless military appreciation. His Anglo-American staff, proceeding from first principles, had first of all identified where landings would be possible. The operational radius of a Spitfire, the most numerous Allied fighter, was used to delimit the zone in which the Allies would enjoy unchallenged air superiority. It reached from the Pas de Calais to the Cotentin peninsula in Normandy; the coast east and west of those places could be eliminated. Within the zone, however, long stretches of coastline were topographically unsuitable: the chalk cliffs of the Pays de Caux were too steep, the mouth of the Seine estuary was too indented, the Cotentin itself was too easily sealed off at its base. By reduction, therefore, only two coastal stretches recommended themselves: the Pas de Calais, with its gently shelving, sandy beaches, and the Normandy coast between the Seine and the Cotentin. The Pas de Calais had the attraction of proximity both to the English coast and to the ‘short route’ into Germany; but for those reasons it could be judged the sector where the Germans would expect to be attacked and would defend most heavily. COSSAC therefore plumped for Normandy.
Because the chosen stretch of Normandy had no ports, but also because the Germans could be counted on to fight to deny nearby Cherbourg and Le Havre to the enemy, it was decided to construct two artificial floating harbours (‘Mulberries’) and tow them to the beaches once they had been seized. The initial landing would be made by three divisions, disembarked from landing craft under heavy air and naval bombardment; airborne troops would be dropped at either end of the chosen bridgehead to secure ‘blocking positions’ on the flanks. As soon as the bridgehead was consolidated, seaborne reinforcements would be poured in to transform it into a ‘lodgement area’ from which a break-out into Brittany and then the west of France would be mounted. Eventually a hundred divisions would pass through Normandy; the main strength of the American army, which would supply the majority of divisions, would be shipped directly from the United States.
Success depended, however, on minimising the strength the Germans could oppose to the landing. Although an intelligence blackout over the invasion fleet itself could be guaranteed, and German air and naval interference be discounted, COSSAC agreed it was vital that near Caen, the Schwerpünkte of the invasion zone, there should be ‘no more than three [German divisions] on D-Day, five by D plus 2 and nine by D plus 8’. The first week of the landings, in short, would be a race between the Allied and German armies’ capacity to build up forces in and against the bridgehead. The Germans could not
prevent the Allied build-up; the Allies could, by contrast, prevent the German. A crucial element of the invasion effort, therefore, would be the bringing to bear of Allied airpower against the roads, railways and bridges by which Rundstedt’s sixty divisions would march to the battlefield. The greater the devastation Allied airpower could inflict on the infrastructure of the French transport system – at whatever subsequent cost to the Allies’ own capacity to supply its armies in mainland France – the more certainly would the seaborne divisions survive the landing and the shock of initial combat in the lodgement area.
Montgomery, on his arrival in London in January 1944, dissented from none of COSSAC’s broad criteria. However, he and Eisenhower, who was eventually to succeed him in command on the ground, had both briefly seen the operational plan when en route to England via Marrakesh (where Churchill was recovering from pneumonia), and they jointly judged that the attack would have to be launched ‘in greater weight and on a broader front’. In brief, they wanted the American landing to be separated from the British, both to be made in heavier weight, and the airborne contribution to be much increased. Montgomery warned that, as things stood, ‘[German] reserve formations might succeed in containing us within a shallow covering position with our beaches under continual covering fire.’ He remembered Salerno, where a well-planned assault had almost come to naught because of the rapidity of the German reaction.
By 21 January, therefore, he had proposed a major amplification of the landing. It was to be mounted by five seaborne divisions abreast, two American to the west, two British and a Canadian to the east; the original ‘two airborne brigades’ were to be increased to two American airborne divisions, dropped astride the river Vire at the base of Cotentin peninsula, and the British 6th Airborne Division, dropped astride the river Orne between Caen and the sea. The creation of airheads on the Vire and the Orne would prevent the Germans from ‘rolling up’ the amphibious bridgehead in between; within it the five seaborne divisions, reinforced by two others pre-loaded in landing craft, would win ground for the post-invasion reinforcements to be landed and deployed. Specialist armour, including ‘swimming’ Sherman tanks, would accompany the assault infantry to their debarkation; the 79th (British) Armoured Division, composed of obstacle-clearing tanks, would open the way out of the beaches for the assault battalions to move inland.
Eisenhower, as Supreme Commander, at once endorsed these proposals. The only difficulty that remained was how to accumulate the craft necessary for the enlarged landing. Admiral King, Chief of (US) Naval Operations, both an Anglophobe and a devotee of the amphibious war in the Pacific, directly controlled the lion’s share of Allied landing-craft production, since the vast majority were launched from American yards (82,000 were built in the USA throughout the war). A near-doubling of the D-Day assault divisions required a proportionate accretion of vessels in which to deliver and support them. These included the Landing Ship Tank (LST), Landing Craft Tank (LCT), Landing Craft Infantry (LCI), Landing Craft Mechanised (LCM) and Landing Craft Vehicle and Personnel (LCVP) as well as the versatile amphibious truck (DUKW or ‘Duck’). King had a surplus of such vessels, particularly the crucial Landing Ship Tank, in the Pacific, but proved unwilling either to transfer any from one ocean to the other or to make available craft no longer needed in the Mediterranean. As a result SHAEF, the Supreme Headquarters Allied Expeditionary Force (as COSSAC was renamed after Eisenhower’s nomination), was obliged to accept a postponement of Overlord from May to June, while its staff scrambled to find landing craft where they could. In addition, Operation Anvil, the landing in the south of France originally scheduled to coincide with Overlord, was set back a month further.
Subsequent investigation has revealed that the shortage of landing craft was illusory rather than objective. By 1943 the output of LSTs from British yards alone already sufficed to land the D-Day divisions; American LSTs were a bonus. The COSSAC staff had convinced itself that the US Navy’s anti-Japanese imperative was depriving it of its just allocation; but the truth seems to be that the shortage was the result of faulty allocation in Europe, not of deliberate starvation by SHAEF’s Pacific rivals. The postponement of Anvil, moreover, though undoubtedly caused by the lack of landing craft, may actually have helped rather than hindered the success of Overlord. Although it was initially conceived as the answering blow to the northern operation (hence ‘Anvil’) which would crush the Westheer by concentric action, the force dedicated to Anvil – four French and three American divisions – was not strong enough to mount a major attack on the rear of the Westheer and, because of the conflicting demands of the Italian campaign, could not have been increased, however many landing craft might have been assembled in the Mediterranean. Anvil’s real value proved to be diversionary; as we shall see, the mere menace of a ‘third’ landing, like that of a ‘second’ in the Pas de Calais, succeeded in retaining German divisions in Provence throughout the weeks when they were desperately needed for the north to fight the real landing in Normandy.
Allied strength, German weakness
To the invasion army assembling in southern England during the spring of 1944, the notion that it might lack for anything would have defied all appearances. The great natural anchorages in which the Channel coast abounds – Chichester, Portsmouth, Southampton, Poole, Portland, Plymouth, Falmouth – were filling with warships and transports. So vast was the gathering armada – which could only have been assembled off Normandy where the Channel is widest – that two of the seven seaborne forces into which it was divided had to be harboured as far away as South Wales and East Anglia. These, Forces B and L, were to sail the day before invasion and join the other five under cover of darkness on the night of D-Day in the mid-Channel ‘Area Z’ from which, through channels cleared by a vanguard of minesweepers, they were to proceed in parallel columns to the five beaches on which the assaulting infantry and swimming tanks would debark; the beaches were codenamed from west to east Utah, Omaha, Gold, Juno and Sword. Operation Neptune, the Naval plan, provided for 6483 vessels to make the voyage, including 4000 landing craft, hundreds of ‘attack transports’, and a bombardment force of 7 battleships, 2 monitors, 23 cruisers and 104 destroyers. Their role was to engage and destroy the coastal batteries of the Atlantic Wall. Close fire support was to be provided by squadrons of rocket-firing landing craft, while others embarked self-propelled artillery which would ‘shoot itself in’ against the German shore positions before rolling up the beach to follow the seaborne infantry. Behind the bombardment and amphibious squadrons sailed the craft bringing the infrastructure required by the assault waves – the ‘beach parties’ which would set up traffic control and signal stations, organise obstacle clearance and evacuate casualties. Assault engineers, manning amphibious bulldozers, demolition tanks and fabric road layers, were also to follow the assault waves at close interval. And in the very forefront would land forward air controllers, to call in rocket, bomb and machine-gun strikes from the fighters and ground-attack aircraft among the 12,000-strong British and American air forces that were to support the landings.
Of these 12,000, over 5000 were fighters; to oppose them General Hugo Sperrle’s Third Air Fleet had only 169 available on the Channel coast on 6 June 1944. A thousand Dakotas were to fly the parachute battalions of the three airborne divisions to their destinations, and hundreds of other transport aircraft were to tow gliders filled with airlanding infantry, artillery and engineers. The mightiest element of the air forces, however, was provided by RAF Bomber Command and the US Eighth Air Force, temporarily diverted from the strategic campaign against Germany to prepare and support the invasion. In the weeks beforehand the ‘heavies’ – Lancasters and Fortresses – with the medium bombers of the British Second and US Ninth Air Forces had largely destroyed the French northern railway system. On the night and morning of D-Day, Bomber Command and Eighth Air Force, each dropping the unprecedented weight of 5000 tons of bombs – the short haul allowed them to substitute bombs for fuel – were targeted against
German defences in the immediate vicinity of the beaches.
The Allies’ overwhelming air superiority guaranteed not only fire support at the moment of assault but security from surveillance beforehand. In the first six months of 1944 only thirty-two Luftwaffe daytime flights over England were recorded; there was only one in the first week of June – on 7 June, a day too late – and this at a time when Allied intrusions into French air space were as common as the flight of swallows. Ultra was meanwhile monitoring the movement of units to and within France on an hourly basis, while the Abwehr had no access whatsoever to the meaning of Allied signals; the volume of such signals, however, was carefully controlled to disguise the presence of the invasion army in the west of England and enhance belief in the fictitious existence of FUSAG in Kent. The Abwehr could, in compensation, draw on the reports of its network of agents in Britain, and these were eagerly assessed for indications of the strength, timing and above all objectives of the invasion. However, since every single one of the agents apparently at liberty had in fact been ‘turned’ by British counter-espionage (the ‘Double-Cross System’), their reports were not only valueless but actively misleading. The British entertained fears that agents outside their control in Lisbon and Ankara might succeed in hitting on the truth by speculation, but none did so; the only serious leak of secrets, sold to the Abwehr out of the ambassador’s safe in Ankara by his Turkish valet, contained references to an ‘Operation Overlord’ but was bereft of details (this was the much misunderstood and over-inflated ‘Cicero’ affair).
