Page 163

Chapter 10: The Battle of Britain: The Preliminary Phase*, (July–August 1940)


AT THE beginning of July the Luftwaffe continued its policy of harassing the United Kingdom by means of light and widely-scattered night attacks. By flying higher than in June its bomber-crews escaped serious interference from the defences while causing a good deal of inconvenience. South Wales and the West of England were easily reached from bases in Normandy and Brittany by well-marked routes which skirted the more heavily defended areas, and therefore received most attention; but between the 2nd and the 10th of the month bombs fell on one or more nights in every seaside county south of the Tyne.1 Attempts to limit warnings still more stringently than in June, by confining them to districts where severe attack was likely, failed because the officers who had to give the warnings could not draw such a distinction without running undue risks. The heaviest casualties suffered during the first half of the month occurred at Aberdeen, where more than fifty people were killed or seriously injured on the night of the 12th in an attack delivered when the sirens had not sounded.

The beginning of the month was also notable for a new series of daylight raids, differing markedly from the occasional attacks on ports which had been delivered in the past. These raids were of two kinds. On the one hand bombers flying singly or in small formations, and relying on cloud-cover or evasive tactics, started to penetrate well inland, reaching places as far afield as the Thames Valley, Norfolk, North Wales and Glamorganshire. On the other, formations sometimes escorted by fighters began to attack ports more heavily than heretofore and to visit places hitherto immune. During the first nine days of July Falmouth, Plymouth, Portland, Weymouth and Dover were all bombed in daylight and seven attacks were made on Channel convoys.2 At least six of the raids were made by some fifteen

* For a summary of operations, see Appendix 10.

Page 164

or twenty bombers escorted by about the same number of fighters, and several brisk engagements were fought between the German aircraft and our own.

In consequence, the effort made by the British fighter force rose sharply at the beginning of the month. Between nine o’clock and six o’clock on 8th July, for example, Fighter Command flew well over three hundred sorties, or roughly the same number as it had put over Dunkirk in one day at the height of the withdrawal. At the same time the raids revealed a new and particularly awkward aspect of an abiding problem. Since the previous autumn protection of shipping off the East Coast had proved burdensome, but comparatively inexpensive; protection in the Channel, where the emphasis now seemed to be shifting, threatened to be still more burdensome and far more costly. Already, as a result of a few attacks by a small fraction of the German air force, the task imposed on Fighter Command had trebled almost overnight and had cost fifteen aircraft and twelve pilots in nine days.3 To improve the chances of intercepting aircraft attacking shipping between Lyme Bay and the Nore, Dowding ordered Park to move a number of squadrons to forward aerodromes in the Hornchurch, Biggin Hill and Middle Wallop sectors; at the same time he was able to bring into the line a few squadrons hitherto unfit. But interception was one thing, protection of shipping by escort or cover quite another. Dowding foresaw that any big increase in the scale of attack might put it out of his power to do what was expected of him. He hoped, indeed, that before long abandonment of the south-western ocean convoy-routes would so reduce the importance of traffic through the Channel that the Admiralty would be content with less protection there; even so he took the precaution of warning the Air Ministry that heavy attacks on inland targets might soon prevent him from escorting convoys unless he had more aircraft.4 Pointing out that recent attacks on shipping had not been made merely by one or two bombers, as in the past, but by substantial formations with fighter escort, he calculated that full protection for all shipping between Land’s End and the Humber would alone absorb some forty squadrons. As it happened, events soon answered his tacit question; for within a few weeks the pace grew so hot that strong and continuous escort for all Channel convoys was clearly not to be expected.


From the German viewpoint the preliminary phase of the air assault on the United Kingdom may be said to have begun soon after the middle of July, when forces assigned to the task were ordered to assume ‘full readiness’, and the assault proper four weeks later on

Page 165

Adler Tag or ‘Eagle Day’. But by the middle of the second week in the month pressure on Fighter Command was severe enough to make many people in this country think that the decisive struggle was at hand. Accordingly, the Battle of Britain is reckoned from the British standpoint to have begun on 10th July, so that those who lost their lives on or after that date are deemed to have fallen in the battle. In point of fact, the rise in tempo was so gradual that any boundary drawn between the prelude and the preliminary phase must be arbitrary. Hence a still earlier date might quite well have been chosen.

The first day of the battle, according to the British reckoning, began as usual with widespread weather- and shipping-reconnaissance nights by the German air force. As a rule, the weather-aircraft which ranged daily over the North Sea and the Atlantic kept well clear of the British coast. They were often tracked for part of their course by the radar chain, and sometimes one would pass over an outlying corner of the kingdom; but in general they gave few chances to the defences. Similarly, long-range aircraft in search of convoys west of Ireland were seldom within reach of land-based fighters. Aircraft searching for coastwise shipping or reconnoitring harbours, on the other hand, were always liable to interception. Even so their discovery was seldom easy, for often early-morning haze or patches of sea-fog helped them to escape unseen. And even in clear weather the limitations of the radar chain made the interception of single aircraft or small formations far from certain.

Not long after sunrise on the 10th a Spitfire from the Coltishall sector overcame these handicaps and engaged a German bomber-reconnaissance machine near Yarmouth. Three hours later a section of Hurricanes, patrolling a southbound convoy off Lowestoft, saw and were probably responsible for driving off two bombers which left the convoy unmolested; but an unescorted convoy in the same area was not so lucky, losing one ship when attacked by a couple of bombers about noon.5 Further south an attack on yet another convoy by two more bombers, this time accompanied by single-engined fighters, led to a skirmish near Margate, in which some twenty British fighters from the Biggin Hill and Hornchurch sectors were involved.

Thus the first half of the day passed almost uneventfully. The first hint of anything unusual came a little before half-past one, when radar stations in south-east England saw signs of a substantial muster behind Calais. A westbound convoy was off Dover at the time, and six Hurricanes from Biggin Hill had been ordered to keep guard above it. While they were doing so about twenty German bombers arrived over the Straits escorted by some forty single-engined and twin-engined fighters. Within the next half-hour the Hurricanes from

Page 166

Biggin Hill were joined by elements of four more squadrons from neighbouring sectors; and a lively action followed. At the cost of three or four of their own number the German fighters protected the bombers from heavy loss, but the attack on the convoy was not very successful, only one small ship being sunk.6 Meanwhile, further to the west a single aircraft from Brittany bombed Falmouth, where one ship of six thousand tons was sunk and two others of about the same size were set on fire.

If 10th July was the opening day of the Battle of Britain, then the action off Dover was the first considerable engagement of the battle. On the whole its outcome was not unsatisfactory. Given such warning as the radar chain could reasonably be expected to provide, the fighter force had shown itself capable of dealing with quite large numbers of the enemy in circumstances which called for swift cooperation between aircraft drawn from several sectors. Even so the intercepting fighters had not arrived in such good time as to suggest that escort for convoys could be abolished, off this particular stretch of coast at any rate, unless the traffic were reckoned so unimportant that lost ships did not matter.

Next day the main interest shifted westward. Besides the usual weather-flights the day began with reconnaissance sorties by German aircraft over the Channel and Thames Estuary. Most of the aircraft over the Channel in the early hours were detected only on their way home; hence no fighters were sent to intercept them, although a convoy was passing eastwards across Lyme Bay.7 But a bigger threat arose soon after half-past seven, when the radar chain detected two formations moving north from the neighbourhood of Cherbourg. Thereupon No. 11 Group ordered six Spitfires from Warmwell, in the Middle Wallop sector, to patrol the convoy and sent forward three Hurricanes from the same base to meet the enemy. A little before eight o’clock the Hurricanes made contact with a greatly superior force, comprising some nine or ten Junkers 87 dive-bombers loosely escorted by about twenty single-engined fighters, and lost one aircraft. About five minutes later the Spitfire pilots saw the Junkers 87’s diving to attack the convoy. Half of them made ready to engage the dive-bombers while the rest protected their rear, but the German fighters broke through our rearguard and shot down two Spitfires. Nevertheless the attack on the convoy failed, not one ship being sunk.

Thereafter cloudy weather kept activity to a minimum until the middle of the forenoon, when a rather disturbing incident occurred near Portland. The protagonists on the British side were a flight of Hurricanes from Tangmere, originally ordered to intercept a German aircraft believed to be on its way back from a protracted reconnaissance of Wales, but later sent south to deal with a raid apparently

Page 167

making for Lyme Bay from the neighbourhood of Cherbourg. ‘Raid’ was a technical term, which might denote any number of aircraft from one upwards; but in most cases radar stations could give an early estimate of size, which frequently had to be amended later. In the present case only a single aircraft was believed to be involved. The six Hurricane pilots were therefore surprised to meet, not a single bomber or reconnaissance aircraft, but some fifteen dive-bombers escorted by thirty or forty twin-engined fighters.

As soon as the true position became known at Uxbridge and Rudloe both groups sent up more fighters, but none arrived in time to do anything useful before the dive-bombers reached their target. Credit for a bold move which did much to retrieve the error of the radar chain goes, therefore, to the original six pilots of No. 601 (County of London) Squadron. Being up-sun from the enemy and at a greater height, they exploited these advantages by diving on the Junkers 87’s and shooting down two before the German fighters could intervene.8 Their prompt action may well have averted serious damage to Portland and its shipping; as it was, the harbour escaped unharmed and only one merchant ship was hit. The bombing was followed by a stiff action between our own single-engined and the enemy’s twin-engined fighters.

The last important engagement on the nth followed in the late afternoon. A little before six o’clock the radar chain again gave warning of a formation flying north from Cherbourg. Once more No. 601 Squadron were in the forefront. Sent forward by No. 11 Group to meet the enemy over the Channel, they came upon a dozen Heinkel 111 bombers and the same number of twin-engined fighters approaching the South Coast. The squadron split into two flights, one of which engaged the bombers while the other climbed to attack the fighter escort; but the bombers succeeded in reaching Portsmouth and dropping about twenty bombs there. Afterwards No. 601 were joined by another squadron from the Tangmere sector, and both squadrons took part in a running fight across the Channel.9 Several pilots claimed successes, but on the whole the verdict on this action must be that our fighters were too few and too late.


The fighting on 10th and 11th July was the stiffest yet experienced by Fighter Command on its own side of the Channel, but still only a foretaste of what must be expected. Even so the fighter force had flown more than six hundred daylight sorties on the 10th and roughly two-thirds of that number on the 11th. Its experiences went far to confirm the Commander-in-Chief’s view that Nos. 10 and 11 Groups

Page 168

would need most of their resources for interception and so have few aircraft left for guarding convoys.

There remained the question whether timely interception could be expected. On the assumption that the function of the defences was not merely to inflict losses but also to hamper bombing, approaching forces clearly ought to be met, if possible, before they reached their targets. So far that had not been always done. In Kent the radar stations had proved capable of giving some twenty minutes’ warning that attack was likely; further west they had been less successful and at least one fairly large raid had come as a surprise. And even when good warning was received the response of the fighter groups had not always been impeccable. Group commanders and their deputies had done their best, but as yet they had little experience of such attacks, and their difficulties were great. In the most favourable conditions the warning was shorter than it seemed, for about four minutes intervened between any observations made by a radar operator and the appearance of the corresponding plot on the operations table. In those four minutes an approaching raid might cover three-quarters of the distance from Cap Gris Nez to Dover. In addition, a Spitfire squadron ordered up in response to the warning must be allowed some thirteen minutes to climb to 20,000 feet, a Hurricane squadron about three minutes longer; and heights of that order were not at all uncommon. Thus the time left for the group commander, or his deputy the group controller, to weigh up the situation and frame his orders was very short indeed. Yet in that brief space he had to make a decision which might be crucial; for to despatch too many aircraft in response to a vague threat was quite as dangerous as to send too few in answer to a real need. If he guessed wrong, a big raid following a small one after a well-judged interval might find him with most of his aircraft running short of fuel. Not surprisingly, therefore, while they were feeling their way commanders and controllers mostly erred on the side of caution. Reluctance to put many squadrons into the air because a big formation seemed to be assembling over France was natural enough, since no one could be certain when it would cross the Channel, or that the manoeuvre was not expressly designed to draw up our forces and exhaust them in preparation for an attack by another formation not yet visible.

The remedy lay partly in technical improvements to the radar stations, but mainly in growing skill on the part of radar operators, group commanders and controllers. Reliable estimation of the size of approaching or assembling raids depended almost wholly on the ability of radar operators to match their observations against previous experience. Accurate assessment of height, although governed to some extent by the extent to which radar stations could be spared from active use while their equipment was being calibrated, was also

Plate 11

Plate 11. Spitfires of a Fighter Command Squadron.

Plate 12

Plate 12. Air attack on a British Convoy in the English Channel, 14th July 1940.

Plate 13

Plate 13. Air Chief Marshal Sir Hugh Dowding, Air Officer Commanding-in-Chief, Fighter Command 1936–1940.

Plate 14

Plate 14. Air Vice-Marshal K. R. Park, Air Officer Commanding, No. 11 Group, Fighter Command, April-December 1940.

Page 169

determined partly by the personal element. Consequently performance improved markedly in both respects as the newer operators recruited during the recent expansion of the system grew in knowledge. And group commanders and controllers, however learned in peacetime theory, were still more dependent on the daily lessons of the battle for a working knowledge of their job.

Otherwise no very striking lessons emerged at such an early stage. As compared with ten of our own aircraft lost, the enemy’s combat losses were correctly estimated at twenty-eight, a too-liberal reckoning of his casualties in the bigger engagements being a too-conservative one where minor combats were concerned.*10 The twin-engined Messerschmitt 110 was clearly seen to be no match for our Hurricanes and Spitfires. On the other hand, the single-engined Messerschmitt 109 was, equally clearly, a tough opponent; but that was known already from experience at Dunkirk and elsewhere. Since May, Dowding had felt that the two-seater Defiant was of doubtful value against single-seater fighters; but he was not yet ready to exclude it from the most active sectors, although he did so later.

In the technical field the chief needs of the fighter force were constant-speed airscrews, which in fact were being gradually fitted to all its aircraft, and well-protected fuel tanks to give its pilots a better chance of survival under fire. In the second respect the Germans, although at first slow to provide armour for their bombers, had taken the lead by fitting both bombers and fighters with excellent self-sealing tanks. Our own designers, seeking a tank which was required to be substantially crash-proof, as well as bullet-proof, had been slower to adopt the self-sealing principle.11 The best having proved the enemy of the good, the wrapping of wing-tanks in a layer of self-sealing fabric had now begun. The reserve tank carried in the fuselage of the Hurricane was more difficult to deal with. As it was believed to be well protected by the armour already fitted it was left untreated. Subsequent events were to show that this belief was not well founded. Later in the battle, after a number of pilots had been badly burned by sheets of flame which filled the cockpit before they could escape by parachute, the reserve tank was covered and the cockpit shielded from it by a metal bulkhead.

In other respects, too, the importance of safeguarding pilots whose machines were hit was keenly felt. Apart from the more obvious aspects of the question, a pilot who baled out was not lost to the battle if he could be quickly brought back to his unit: hence one result of early combats over the Channel was to draw attention to the need for a means of rescuing those who came down in the sea. The German air force already possessed an organisation for the purpose and had

* In addition, the Luftwaffe lost five aircraft from other causes.

Page 170

equipped it with both marine craft and aircraft.12 Although the British Government did not agree with the German claim that aircraft so used should be treated as flying ambulances and allowed free access to territorial waters, the Air Ministry were not slow to follow the German lead. Within a few days of the opening of the battle they arranged with the Admiralty that small craft should patrol inshore when heavy air fighting was in progress. Soon a number of high-speed launches under Coastal Command were working regularly from bases between the Solent and the Nore. Later some Lysander aircraft were given the task of ‘spotting’ for the launches, and ultimately a full-blown ‘air-sea rescue service’, with a variegated establishment of amphibian and other aircraft, was brought into being.

The preliminary phase of the Battle of Britain lasted from 10th July until 12th August. Throughout the greater part of that time events conformed closely to the pattern set on the first two days. Ports and shipping were the targets for nearly all daylight attacks of any size; yet, notwithstanding what has been said about the strategic objects of the German High Command, the purpose of the attackers was probably not so much to damage ports or sink ships as to wear down the defences in preparation for the main assault. They failed to achieve that purpose largely because their operations were neither planned nor carried out in such a way as to make the most of Fighter Command’s weak spots. Knowing that British fighter pilots got their orders from the ground, German staff officers believed them to be rigidly tied to the immediate vicinity of their bases, and those who gave the orders to be debarred by their position from distinguishing between large raids and small.13 In fact, the fighter system was not wholly free from such defects, but they were neither so widespread nor so fundamental as was believed by the German Air Staff. The Germans overlooked recent improvements in radio equipment and the ability of radar operators and others to profit from experience. Hence the attackers were ill served by a policy which gave the defenders every chance of learning from their mistakes, instead of overwhelming them by a series of well-concerted blows delivered without prolonged rehearsal.

To a limited extent the failure of the preliminary offensive to achieve anything of value to the Germans can be measured by the crude yardstick of statistics. Between dawn on 10th July and nightfall on 12th August German aircraft attacked merchant shipping in the Channel almost daily. Yet in those five weeks only some 30,000 tons of shipping were sunk by aircraft between Land’s End and the Nore, out of a volume of coastwise traffic amounting to nearly a million tons a week.14 On the thirty-four days Fighter Command flew more than 18,000 daylight sorties, or a daily average of about 530. The

Page 171

====> reference location unknown15

number of sorties flown by the Luftwaffe is not known and cannot be estimated with even approximate accuracy; but the German effort is likely to have been smaller than our own, which included a large number of routine and precautionary patrols. Even so our squadrons were often locally outnumbered. Yet in all their daylight combats Fighter Command lost only 148 aircraft, nearly half of them on three days during the second week in August. Another two were lost at night. The Luftwaffe lost 286 aircraft in operations against the United Kingdom, of which all but a very few succumbed in daylight battles. Of that number 105 were single-engined or twin-engined fighters. On the three days in August which cost us 73 aircraft the Luftwaffe lost a hundred. Thus over the whole period the Luftwaffe lost nearly twice as many aircraft of all classes as Fighter Command lost fighters, for a very small return in merchant shipping sunk. In terms which cannot be measured by statistics the preliminary phase was still less profitable to the attackers, for, as we have seen, it taught Fighter Command some useful lessons without advancing the German strategy in any discoverable way.

What those lessons were has been suggested in earlier paragraphs, where the conclusions drawn from two days’ fighting were discussed. On the whole the events of the next few weeks confirmed them without adding a great deal that was novel. Apart from certain technical shortcomings, the main weaknesses of the defensive system continued to be the partial inability of radar operators to give reliable estimates of height and strength and the occasional failure of groups to oppose raids early enough or with large enough formations, in most cases because the radar picture was confused or incomplete. Despite the improvements already sketched, attempts to gauge height were always likely to be defeated by the time-lag between an observation and the appearance of the corresponding plot in the operations room. Moreover, approaching aircraft could climb so quickly as they crossed the Channel that a group commander or controller could never be sure that a formation was where it appeared to be, even if the latest radar estimate was well founded. On the whole, his safest course was to order his fighters to fly substantially higher than the enemy’s estimated height, thus lessening the risk of their being pounced upon; but in cloudy weather there was always the fear that they might miss the enemy altogether, especially if—as sometimes happened—the fighter leader exercised a similar discretion by flying higher still. The remedy lay in a careful study of the enemy’s habits as revealed by reports from squadrons; in intelligent anticipation; and in the growth of mutual confidence between pilots and those from whom they received their orders which usually followed a few successful interceptions.

Even so the frequency with which our squadrons were outnumbered was disturbing, especially as the improvement to be

Page 172

expected from added experience was limited. The Germans could assemble large formations beyond or below the limit of radar cover and were in fact accustomed to use the Gruppe of thirty aircraft as a tactical unit. Fighter Command’s normal tactical unit was the squadron of twelve aircraft. Over Dunkirk our fighters had often flown in wings of two, three, or four squadrons; but in present conditions little time could be spared to assemble wings if German formations were to be met before they reached their targets. Hence our squadrons often went into combat singly, and in consequence sometimes found themselves outmatched.16 That a good toll was nevertheless taken of both bombers and escorting fighters shows how well our pilots faced their task.

The situation confronting Dowding towards the end of the preliminary phase was, then, that so far the burden had fallen chiefly on No. 11 Group and especially on the coastal sectors from Middle Wallop to North Weald. In terms of casualties inflicted and suffered the battle had gone well, but some formations had been intercepted only by small forces and after they had bombed their targets. Characteristically, he nevertheless resisted the temptation to strengthen the south-eastern sectors at the expense of others, foreseeing that to do so would invite a flank attack which he would be ill prepared to meet.17 Throughout the anxious opening phase of a contest whose issues were clarified by no rules derived from well-thumbed textbooks he maintained his opening dispositions almost unchanged; and such small changes as he did make were chiefly designed to strengthen the West Country and No. 11 Group’s right flank rather than its more obviously threatened centre. Thus on 12th July he moved the Spitfires of No. 152 Squadron from Acklington in No. 13 Group to the Middle Wallop sector. Six days later he moved the single flight of No. 247 Squadron, equipped with the only Gladiators still in the Command, from Sumburgh in the Shetlands to a small aerodrome at Roborough, in No. 10 Group. Roborough was unsuitable for Hurricanes or Spitfires, but could accommodate the Gladiators, whose new task was the local defence of Plymouth. To replace them at Sumburgh a flight of Hurricanes moved there from Wick. On 20th July the Hurricanes of No. 245 Squadron went from Turnhouse to Aldergrove, in Northern Ireland; and next day the Defiants of No. 141 Squadron, which had been outfought over Dover on the 19th after moving south from Turnhouse, were withdrawn from the busy Biggin Hill sector to Prestwick, in Ayrshire, where they might still do good service against unescorted bombers. The other Defiant squadron, No. 264, moved temporarily to Kirton-in-Lindsey, whence a flight was detached to Ringway for the defence of Manchester, but later went south again to the Hornchurch sector and for some days were in the thick of the day fighting. Other changes in July

Page 173

include a strengthening of the anti-aircraft defences of some western ports.

From this account may perhaps emerge the picture of a rigid system of sector-stations, each with its fixed quota of fighter squadrons. The reality was otherwise, for in each sector there were a number of other aerodromes serving a variety of purposes. Some, like Croydon and Martlesham, were quasi-permanent bases for squadrons detached, on tactical or administrative grounds, from the headquarters of the sector. Others, like Hawkinge and West Mailing, were more often used as temporary bases or forward landing-grounds, although they too might serve as more permanent bases if the need arose. Administratively these bases fell into a number of fixed categories; tactically the use that could best be made of them depended on a variety of factors, some of them not easily assessed except by those with local knowledge. While, therefore, the disposition of squadrons within a given sector was a matter on which much discretion ought clearly to be allowed to local commanders, it was also one which gave much room for differences of opinion. This became quite clear on 29th July when, after ships off Dover and in harbour there had been repeatedly attacked by German aircraft, the Air Staff urged the Commander-in-Chief to make more use of stations near the coast for the purpose of meeting the enemy with ‘superior forces and large formations’.18 Their diagnosis was sound, but their remedy was open to question, first because a more generous use of forward bases would not necessarily enable No. 11 Group to do what the Air Staff wanted, secondly because the stations in question were already getting more use than they seem to have supposed. In the two sectors fronting the Straits there were six forward aerodromes besides the sector-stations at Biggin Hill and Hornchurch.19 Of these six, three at Gravesend, Rochford and West Mailing were some way from the coast. A fourth, Lympne, was useful only in emergencies. The remaining two, at Manston and Hawkinge, were so exposed that, although in fact one squadron was based at Manston, the practice followed by No. 11 Group and endorsed by Dowding was to use them mainly as daytime points of departure and return for squadrons whose ground organisation and reserves remained as far as possible in safer quarters. On these terms both were in constant use as forward landing-grounds.20 Had excessive deference to the Air Staff’s wishes led the Commander-in-Chief to reverse his policy by ordering the Group Commander to move several squadrons permanently forward he might have had cause to regret it, for the stations were soon to be viciously attacked. As it was, no material change in the disposition of the squadrons in the south-east followed the Air Ministry’s démarche,

Page 174

although the point at issue was not overlooked. Ultimately the difficulty of meeting the enemy forward of his targets in sufficient strength compelled No. 11 Group to resort to different measures; but by that time the focus of attack had shifted inland.