Cockpit check-thumbs up-switches on-press the starter button. A few turns of the two-bladed airscrew, blue-grey smoke puffs from the exhausts and the Merlin roars into life … At take-off power the Hurricane needs a fair bit of right rudder, then, almost unexpectedly, she leaps eagerly off the grass and flies. Unconsciously moving the stick when reaching for the undercarriage lever, I immediately have to pick up the nose and port wing-God! but these controls are sensitive. But what a beautiful aeroplane-instant obedience to the controls, superb view, and what power. So much in fact that one’s leg aches holding her in a prolonged climb.- Graham Leggett, No 46 Squadron, 1940
For Air Component squadrons, the first ten days of the German offensive had been a harsh introduction to the realities of modern war. Losses had been heavy, but lessons had been learned and a more experienced and flexible force was beginning to emerge. Fighters were still flying in tight ‘vic’ formation, but they were beginning to fly in greater numbers. Some squadrons were beginning to use flights at higher altitude to cover those flying lower. The pilots in France were relearning the lessons of the First World War.
By the morning of 16 May, half of the 200 Hurricanes sent to France had been put out of action. Over fifty pilots had been killed, were wounded, or were missing. More reinforcements sent out on the 16th and 17th and the fighter squadrons operating from Britain and refuelling in France helped make up for these losses. There seemed to be some justification that basing more squadrons permanently in France would add to the confusion. There were some dangerously overcrowded airfields. On the 18th, bombers and strafing Bf 109s and Bf 110s destroyed seven Hurricanes from the six squadrons operating from Vitry. There was no shortage of space in the swathe of territory north of the breakthrough, but there was a reluctance to disperse the fighters and operate them from more basic airstrips.
The newly arrived pilots had to start learning the tactical skills their comrades in France had been acquiring. Often the Hurricanes they flew had no rear armour, and some still had fixed two-blade wooden airscrews. Even the twin-engine Bf 110 posed problems for the struggling Hurricane pilots. Five out of nine patrolling Hurricanes were shot down on the 18th in a clash with I/ZG 26; three of the casualties were newly arrived Canadians. Nevertheless, if they could avoid the Messerschmitts, the Hurricanes could still be very effective. On 17 May, No. 151 Squadron, operating from Abbeville for the day, shot down seven Ju 87s from III/StG 51. The Component Hurricanes were a very real presence; on the 19th, they managed around 250 fighter sorties. In particularly fierce clashes over Lille, KG 54 lost fourteen He 111s to Hurricanes. For those who had time to gain experience, the Bf 110 and even the Bf 109 were not posing quite the same degree of difficulty.
The Air Component was becoming more flexible. Despite the desperate shortage of fighters, reconnaissance Blenheim and Lysander sorties were now sometimes getting a close escort. If Hurricanes were going to accompany the Blenheim or Lysander anyway, there seemed to be a case for the fighter carrying out the mission. They were not equipped with cameras; the only information they could provide was what the untrained eye of the pilot might see, but that was often enough. On the morning of the 20th, it was Hurricanes sent out on reconnaissance missions that reported enemy columns were advancing towards Arras from Cambrai. These reports led to another step on the path towards a more versatile tactical air force. The Merville ground controllers, whose normal task was to direct the fighters towards bombers, instructed Hurricanes already on patrol to strafe these columns. Other Hurricanes were ordered into the air to join them. Squadrons on escort or fighter patrols began to use any ammunition they still had at the end of their mission on any suitable ground targets. After years of being frowned upon by Fighter Command, ground strafing had suddenly become an accepted part of the fighter’s role.
The fighter pilots were scarcely prepared for this new role. They flew ground-staffing sorties as they flew any other fighter mission—in tight formations. These proved as unsuitable for ground strafing as they did for air combat. Sqn Ldr Kayll of No. 615 Squadron led his twelve Hurricanes in four sections of three and lost three to ground fire. Of the fifty-odd ground-strafing sorties flown on the 20th, six were shot down by flak, with three pilots lost. In a day of intensive interception, escort, reconnaissance, and ground-attack missions, mostly in the Arras region, thirteen Hurricanes were lost, but forty enemy planes were claimed as destroyed or damaged and several enemy columns were shot up. In the heat of battle, the versatility the single-seater fighter had displayed in the First World War was being rediscovered.
It was not a versatility Dowding wanted rediscovered. Escort, reconnaissance, and ground strafing were not what his fighters were supposed to be doing. He did not even want them protecting the Army. On 16 May, he demanded to know how low the Air Staff was willing to let Britain’s fighter defences run down before turning off the ‘Hurricane tap’. Operations in France meant his remaining thirty-six home-based squadrons were already seriously understrength. Remarkably, Dowding was already contemplating French defeat; Britain had to look forward and consider how she could survive alone, he argued. As long as Fighter Command and the Royal Navy remained capable, defeat in France would not bring the defeat of Britain. If, however, Fighter Command was consumed in an effort to keep France in the war, ‘defeat in France [would] involve the final, complete and irremediable defeat of this country’. The Army was not mentioned. Apparently, ground forces were not essential to Britain’s survival.
Newall agreed. The Chief of Air Staff could not see how a few more squadrons could make any difference in France. Newall presented these arguments, with Dowding’s note, to the Chiefs of Staff Committee. They in turn recommended that no more fighter squadrons should go to France. On 19 May, Churchill agreed. This was not good news for the troops fighting in France, and what followed was even worse. Panic set in.
Churchill was not contemplating pulling the Army or the RAF out of France at this point. On the 20th, he asked Newall to make preparations in case the RAF had to be pulled out, but that was all.8 He still envisaged the BEF breaking out to the southwest to rejoin the French Army, as Ironside wanted and Gamelin was planning. On the evening of the 19th, he dispatched Ironside and Slessor to France to make this clear to Gort. The BEF commander, however, was already planning to retreat northwards to Dunkirk. Indeed, Blount, with Gort’s approval, was already sending his Air Component reconnaissance squadrons back to Britain. It was Gort who set in motion the evacuation of the RAF from France, not the Air Ministry. As Slessor left Dover for France, retreating Air Component squadrons were already landing at airfields a few miles away. By the time Slessor reached Gort’s headquarters, all the Blenheim and Lysander squadrons—apart from one squadron and a flight of Lysander at Gort’s headquarters—had left.
Churchill’s decision that no more fighters should go to France was interpreted rather liberally by the Air Staff. Squadrons operating from Britain stopped using bases in France to refuel, and the half-squadron reinforcements sent on the 16th and 17th were also pulled back to Britain. This was perhaps not quite what Churchill meant. Slessor, however, took the policy a stage further by taking it upon himself to organise the withdrawal of all Air Component fighter squadrons. At 4 p.m. on 20 May, the Air Component flew its last fighter patrols from French bases. By the evening, the only RAF fighter squadrons in France were the three with the AASF south of the breakthrough.
There was an air of haste about the decision and panic about its implementation. Masses of equipment was abandoned, unserviceable aircraft were destroyed, and brand new aircraft suffered a similar fate if no pilots were around to fly them away. Hurricanes at storage parks were hauled into groups of three, nose-to-nose, and set alight. Others were lined up and destroyed by anti-aircraft guns. Elsewhere, transport planes were flying over to France to collect pilots with no planes to fly. If air transport was not available, RAF personnel were told to make their way back to Britain as best they could. Some fled as far as Cherbourg, nearly 250 miles behind the French front line, to find a boat back to Britain. On hearing that tanks were approaching, the staff of the Merville operations room clambered through the toilet window at the rear of the building and fled across the fields. The nearest German forces were over 20 miles away at the time, and they were not heading for Merville. The staff made their way to Boulogne and back to England. They were actually fleeing towards the German forces; Boulogne would fall into enemy hands several days before Merville.
The evacuation was premature to say the least. Over forty Allied divisions were north of the German breakthrough in a pocket that extended from Abbeville in France to the Scheldt estuary. There was scarcely a shortage of space to deploy an air force. Blount and the Air Ministry seemed to be in rather a hurry to get the RAF back to Britain and Gort did not seem to mind; indeed, he seemed to believe that the sooner the BEF followed them, the better. Between them, British Army and Air Force commanders had scattered the RAF far more effectively than the Luftwaffe.
The attitude of Pownall, Gort’s chief of staff, is particularly intriguing. He had spearheaded the pre-war Army drive to acquire more effective air support, but he seemed to raise no objections to the Air Component returning to Britain. Even if Gort’s preferred option was to retreat to the coast, his Army would still need air support. It may be that Pownall and Gort were persuaded that the RAF could operate just as effectively from airfields in Britain, or it might just have been a way of pre-empting any decision to keep the British Army on the continent.
Part of the problem was that neither Pownall nor Gort seemed to understand what was happening around them. To some extent, the BEF was in the eye of the storm. British troops were only facing German infantry; they had not yet experienced the full ferocity of the air/tank blitzkrieg combination. Pownall could not understand why Dutch, Belgian, and French forces were faring so badly. He was convinced the French in the south had collapsed in the face of a few small-scale raiding parties. He did not think the main German Army had even been committed yet; he was so consumed by contempt for Britain’s allies that he could not see that the way they were being defeated proved he had been right all along about the importance of air power.
While the Air Component fled northwards, the Army it was supposed to support was preparing to strike southwards. The last squadron of Lysanders departed on the morning of the 21st, leaving just a single flight. Having gone to so much trouble to design an army-cooperation plane that could operate from any convenient open space close to the front line, most of his Lysanders would now be operating from airfields on the other side of the English Channel. The transformation could not have been more striking. The previous day, the skies above Arras had been the scene of furious battles between the RAF and the Luftwaffe; now there was not an RAF plane in sight.
Gort was about to discover the limitations of air support from 100 miles away. The ‘Back Component’ (as the Air Component was renamed) was no longer solely an Army force—its Blenheims and Lysanders were now controlled jointly by the War Office and Air Ministry. The Hurricane squadrons went straight back to Fighter Command. From now on, they would be used as Dowding saw fit. The retreat of the Component was a disaster for the BEF.
Dowding, however, saw it as a disaster averted. Indeed, to reduce losses further, he wondered if the three remaining Hurricane squadrons with the AASF could also return to the UK. For Dowding, the problem was not how to protect the British Army; in fact, the problem was the British Army. He did not measure the success of his force by how effectively it enabled the Army to operate—he measured it by how many enemy planes were shot down, and he believed that his fighters could shoot down more bombers and suffer lower losses when operating over Britain. Fighters that crashed could be repaired, not abandoned; pilots who baled out would land in friendly territory; and, most importantly of all, radar could direct the fighters to the bombers. He was sure that the best place to fight the war was over Britain. However, this analysis ignored the fact that the war was being fought in France. Fighters might well be more efficient fighting a defensive war over friendly territory, but it was scarcely worth losing allies, abandoning armies, and taking your country to the brink of defeat to gain this tactical advantage.
Remarkably, Dowding’s greatest fear was that the Allied armies surrounded in the north might successfully counterattack and re-establish contact with the bulk of the Allied forces in the south. If this were to happen, they would inevitably start asking for fighters again. As he put it: ‘They will be unable to continue the battle without wrecking the Home Defence Units.’ As far as Dowding was concerned, the sooner the Allied armies in France were defeated, the better it would be. If losing France as an ally kept his Fighter Command intact, it was a price that he was willing to pay. It was the tunnel vision of a commander who could not see beyond his assigned mission of defending British airspace.
Dowding left Sinclair in no doubt about what was at stake:
I earnestly beg, therefore, that my commitments be limited as far as possible, unless it is the intention of the Government to surrender the country in the event of a decisive defeat in France.
For good measure, he also ‘earnestly recommended’ that Bomber Command be allowed to focus on oil plants and aircraft factories rather than wasting its time on communication targets that could be ‘very quickly repaired’. Dowding desperately needed the Luftwaffe to attack London. It was an interesting role reversal for Bomber Command; it had been set up to deter attack, but Dowding wanted to use it to encourage one.
Unfortunately, the German High Command was stubbornly refusing to comply with Dowding’s vision of how the war should be fought. In five nights, the RAF had launched around 250 sorties against oil installations, yet still the Luftwaffe showed no inclination to retaliate. There was so little news emerging from Germany about the bombing, Duff Cooper, the Minister of Information, decided that the German authorities must be going to extraordinary lengths to ‘hush up’ the raids. Their single-minded determination not to be goaded into retaliating was impressive.
In fact, the Germans faced no such dilemma. The bombers used in the initial strike on 15–16 May might have delivered more bombs than were dropped on Rotterdam, but the results could not have been more different. The Rotterdam raid had shocked the world, whereas Bomber Command’s effort against the Ruhr merely sparked curiosity among the residents. The night of the 15th–16th had been no different to the previous four nights. A few bombs had fallen in built-up areas, but they were so scattered that it was impossible to work out what the intended targets might have been. There was speculation that the British raids were navigational training exercises flown by inexperienced crews, with a few bombs dropped randomly as an afterthought. On the first night of the strategic air offensive one person was killed and at least seven were wounded. It never occurred to the Germans that a full-scale air offensive had been launched. The need for retaliation was even further from their thoughts.
On the night of 17–18 May, the offensive was extended to oil targets outside the Ruhr. It was at least clear which towns were the targets. A fertiliser factory in Hamburg was gutted and 160 buildings damaged; thirty-four people were killed, and another thirteen died in Bremen. In terms of number of bombs landing in a built-up area, this was a more successful effort, but oil production was totally unaffected. The Germans were not even aware that oil plants were the target. The distress that the civilian casualties generated at a personal level did not affect a country enthralled by the staggering successes of its Army. To the Germans, Bomber Command’s best efforts appeared to be indiscriminate terror raids that were so poorly executed they did not merit retaliation. Dowding would have to wait.
While Dowding marshalled his resources for the air attack that would never come, Gort was trying to deal with a German offensive on land that showed no signs of relenting. On 20 May, General Weygand replaced Gamelin as overall Allied commander and began organising the Allied breakout. The British would attack south from the Arras region and the French from Douai towards Cambrai. Weygand hoped to have forces for the main push ready for the 26th; in the meantime, preliminary raids were to prepare the way and hopefully knock the Germans off-balance. Gort was already planning a counterattack as a defensive measure, to slow the German drive past Arras. This now became the first instalment of Weygand’s counterattack. Major-General Franklyn would command the British force, which would be spearheaded by two battalions of tanks, with seventy-four Matildas, led by Major-General Martel. Two battalions of infantry would accompany the tanks. This small force would need all the air support it could get.
It was unlikely to get any if Dowding and Portal got their way. Portal, however, was being forced to back-pedal. On the evening of the 19th, with the Air Ministry decision to cut back tactical air support still making its way to BAFF HQ, Barratt contacted Bomber Command about future bomber support. He was horrified to hear that there would not be any by day and only Blenheims by night. Furious, he immediately contacted Douglas; within hours, Newall’s deleted caveat, guaranteeing Barratt’s demands for direct support would be met, had been issued.
Georges was soon making it clear what he wanted. At midday on the 20th, the French requested maximum RAF effort by day and night to halt the tanks threatening Arras, with the focus in the Cambrai-Arras-Peronne triangle. General Dill, Ironside’s deputy, backed the French request. Barratt and the French made ‘tentative enquiries’ about using the Wellingtons and Hampdens by day. In their reply, the Air Ministry went out of their way to explain they could not use the aircraft unless they had an escort. It was a somewhat defensive explanation; the Wellington and Hampden were day bombers, and there was no reason why they could not have an escort—apart from the Air Staff’s firm conviction that this was a misuse of fighters.
Slessor was busy organising the evacuation of the Air Component from France, but even he could see the need for more bomber support, insisting that No. 2 Group return to day operations. Portal had no choice but to sanction the use of Blenheim by day, but they would now get a close escort. There would be no more talk of clearing the air ahead of the bombers or meeting the escorts over the target; the Blenheims would fly to the fighter airfield and wait for the escort, after which the whole formation would fly to France, with the fighters 1 mile behind and 1,000 feet higher. The Blenheims flew around seventy sorties on the 20th, mostly south and east of Arras, as Dill and Georges had requested, and no aircraft were lost. It was a remarkable reversal of fortune. Cynics in the Air Ministry pointed out that no enemy fighters had been encountered. Instead of taking satisfaction on a mission successfully completed, it was felt that the escorting fighters had simply wasted their time. Perhaps a more useful conclusion would have been that it was time for escorted Wellington and Hampden missions.
On the night of 20–21 May, the night before the British attack, Bomber Command was ordered to suspend the oil offensive and strike much further west, in the Cambrai-Hirson-Vervins region. French bombers would operate in the St Quentin-Bapaume-Arras triangle. The French did their best to fly as many sorties as possible, with some planes flying two missions during the short summer nights. Fifty-nine sorties were squeezed out of the thirty-seven available planes. In its assigned area, Bomber Command used just seventy ‘heavies’ out of an available 250; Portal insisted that any more was impossible in such a small area. Eighteen Blenheims, flying their first nocturnal missions, attacked targets west of Brussels. Thirty-eight Battles continued their nightly offensive against the Meuse crossing points. At such a crucial time, the Allies could not afford to spread their meagre resources so thinly.
For Martel’s counterattack, little if any reconnaissance had taken place since the Hurricane sorties the previous morning. Lysanders based in Britain were sent off to reconnoitre a swathe of territory between the coast and Arras to establish how far the German advance had reached. The flight of Lysanders left at St Omer flew eight missions along the southern flank of the BEF. Their reports include tanks spotted south of Boulogne, which suggests that these too were being used for strategic rather than tactical reconnaissance. There do not appear to have been any sorties in the Arras region. Gort believed the counterattack would be striking in the gap between the armoured spearheads and the supporting infantry; as it turned out, the British forces would run into elements of Rommel’s 7th Panzer Division preparing to strike to the north-west of Arras.
On the morning of the 21st, the Air Ministry instructed Fighter Command to supply three Hurricane squadrons for escort duties and three for sweeps in the area Arras-Cambrai-Le Cateau. This was unnecessarily too far east for fighters that would struggle to patrol the Arras region for long from airfields in Britain. Again, the Blenheim escorts did their job—just two bombers being lost in fifty-eight sorties. The targets, however, were enemy columns advancing up the coast, towards Boulogne. As dangerous as these advances were, a choice had to be made between halting the advance in the rear or making it irrelevant by severing the German spearhead. Weygand had decided on the latter, but the RAF was attempting the former. The temporary disruption they caused might have been of more value south of Arras, where British troops could take advantage.