Operation Goodwood: The Guns of Cagny

Operation Goodwood, in July 1944, was possibly the most controversial Allied offensive of the Normandy campaign. Launched with an entirely unprecedented level of air support, it was ultimately halted by determined German resistance south and southeast of Caen, not least from units in and around the village of Cagny.

Operation Goodwood was launched on 18 July 1944. Although northern Caen had fallen to the Allies during Operation Charnwood, earlier in the month, southern Caen had still not been taken. To the south of the city, the Germans had established four defensive belts ten miles deep. These comprised (1) infantry, (2) panzer grenadiers, (3) fortified villages with anti-tank guns and mortars, and (4) defensive positions running across higher ground from the Bourguébus Ridge in the southwest to Troarn in the east. Beyond the fourth belt was an armoured reserve – Panther and Tiger tanks of the 1st and 12th SS Panzer divisions

Montgomery was rapidly losing the confidence of Eisenhower and his deputy, Tedder, but he was not quick to embrace the prospect of another offensive. By 1944, the British army was confronted by the looming prospect of a manpower crisis, and there had already been extremely heavy casualties among the infantry since D-Day. Montgomery was reluctant to worsen an already difficult situation.

However, since the landings, British armour had been building up steadily in the eastern sector of the beachhead. Lieutenant General Sir Miles Dempsey of Second British Army now came to Montgomery’s aid by devising a plan to use it in a new offensive – Operation Goodwood. Dempsey believed that the country south of Caen would be particularly suited to armoured warfare. He had been favourably impressed by the bombing of Caen during Charnwood, and felt that similar techniques might be applied to clear a path for his armour to advance down a mile-wide corridor east of Caen into the Caen-Falaise plain to the south. However, at the time, Allied intelligence did not appreciate the depth of German defences to the south of Caen.

Montgomery accepted Dempsey’s plan and quickly won Eisenhower’s support for it. Yet there was always a lack of clarity over the precise objectives of Operation Goodwood. Eisenhower and Tedder thought Montgomery was proposing a breakout from the beachhead; this was implicit in Dempsey’s operational order, which referred to Falaise as an objective. Montgomery also told the Chief of the Imperial General Staff that Falaise was his goal. But he then changed the directive to the principal corps commander, Lieutenant General Richard O’Connor (commanding 8 Corps). The revised directive proposed the establishment of the three armoured divisions of 8 Corps in the northern part of Caen-Falaise plain, but not a breakout to Falaise.

Tap/click to enlarge.

In planning the bombing for Operation Goodwood, the Allies sought to learn lessons from Charnwood, when craters and rubble had impeded ground offensive. The bombs and fuzes used on the main Goodwood axis of advance were chosen to avoid cratering. Additionally, the interval between the bombing and the start of the ground assault was reduced to take maximum advantage of the disorientation bombing was expected to cause among enemy troops. Nevertheless, for reasons discussed later, it proved very difficult to move the British armour forward rapidly after the bombing, and in sufficient strength.

It is also important to remember that Goodwood was a very different operation from Charnwood. In Charnwood, Bomber Command targeted two clearly defined boxes on the northern outskirts of Caen. In Goodwood, the bombing was undertaken by three forces – Bomber Command, the heavy bombers of the US Eighth Air Force, and the medium bombers of the US Ninth Air Force. Moreover, 8 Corps devised an integrated air and artillery bombardment concept in which the aerial bombing was spread around no fewer than eleven separate zones – the two flanks, the northern section of the corridor, and the southern end of the corridor. With hindsight, there seems little doubt that this plan was far too complex.

At its deepest, the bombing area was about six miles; at its broadest it was the same distance. Although a much greater weight of bombs was used than at Caen, about 70 per cent of the planned tonnage was aimed at the flanks against identified German concentrations or to impede counter-attacks; the remaining 30 per cent was divided between the northern and southern sectors of the corridor. The density of bombing planned for the corridor was between a quarter and a third of that planned for Operation Charnwood and was only about half that planned for Operation Cobra – in the American sector – a few days later. In the northern area of the corridor, this was not particularly significant because the aerial bombardment was supplemented by a massive artillery barrage, but difficulties arose further south, beyond the artillery’s range, owing to the depth and dispersal of German units.

After Goodwood, some Army officers argued that the weight of the attack should have been shifted from the flanks to the southern sector of the corridor, but this view was based on the assumption that all the bombers available had been used in Goodwood. In fact, both Bomber Command and the Eighth Air Force could have sent more aircraft to participate in the operation; only about half the Eighth Air Force’s bombers took part. More aircraft would almost certainly have been made available if Dempsey had asked for them. The truth is that neither the Army nor the Allied air forces had a clear grasp of the weight or density of bombing that would be necessary to suppress the German defences southeast of Caen. There was hardly any past experience by which to measure the effort that would be required.

The scale of the Operation Goodwood air attack was unprecedented; never before had so many heavy and medium bombers been assigned to the support of ground troops. Some 1,512 heavy bombers and 343 mediums took part, delivering a total of 6,000 1,000lb bombs and 9,600 500lb bombs, and also many smaller general purpose and fragmentation bombs. Bomber Command struck first. Weather conditions were clear and the bombing was very accurate by the standards of the day: about half the planned density per acre was achieved. Enemy forces on the two flanks were neutralised, and the third zone, around Cagny, was also bombed accurately. Nevertheless, as we shall see, German defences in this area played a key role in thwarting Montgomery’s plans.

Cagny, a Bomber Command target, was accurately
attacked; an aircraft is visible in the top right corner

Cagny after the bombing on 18 July 1944

Sannerville, on the eastern flank of the armoured
corridor, was devastated by the preliminary bombing

The US bombers were next. By the time they arrived, visibility had deteriorated, and many aiming points were obscured by smoke and dust from the British bombing. Nearly one quarter of the medium bombers did not attack, along with 72 of the 643 heavy bombers, and the medium bombers were unable to bomb accurately: they achieved only about one eighth of their planned mean bomb density in the northern corridor. The heavy bombers did better, but still only achieved about one fifth of their planned mean bomb density. At the southern end of the corridor, American aircrews delayed bomb release because of their concern to avoid friendly casualties on the ground, and many bombs fell long of the target areas – some by as much as four miles.

This is not to say that shortcomings in the aerial bombardment were primarily responsible for Goodwood’s failure. More accurate bombing in the corridor, particularly in the south, would probably have helped 8 Corps, but Dempsey’s basic concept of using heavy bombing in support of the offensive in this sector was fatally flawed. The German defences covered too large an area, and were too dispersed for bombing to be able to punch a hole right through them. The bombing plan did not address the problem of the German armoured reserves at all.

Several obstacles confronted the armoured assault. The launch area – the small British bridgehead east of the Orne – became very congested, and the corridor itself was narrow. There were also minefields to negotiate and a number of railway embankments. The armoured divisions – subdivided into their component brigades and battalions – had to move forward one at a time, and it proved to be a slow and laborious process.

Time was of the essence. It was vital to exploit the psychological effects of bombing and the artillery barrage on the defenders rapidly, before they wore off. Yet it took too long to move 8 Corps south in sufficient strength. In the southern area of the corridor, where the bombing had been least accurate, the Germans began to recover, and 8 Corps’ lead formation, 11th Armoured Division, were confronted by strengthening opposition.

After a substantial proportion of their motorised infantry became embroiled in clearing defenders from two villages along the corridor, 11th Armoured Division’s tanks pressed further south. Increasingly potent German resistance meanwhile slowed the advance of the second armoured division – the Guards Armoured Division. And so it was that only 11th Armoured Division attacked Bourguébus Ridge on the 18th. Confronted by a strong and well-prepared adversary bolstered by the arrival of armoured reserves, they stood no realistic chance of breaking through. On the following day, 8 Corps’ armour and infantry reached the area in greater strength, and heavy fighting continued south of Caen for another 24 hours, but the projected breakout towards Falaise remained as elusive as ever.

One of the most decisive tactical actions during Operation Goodwood’s first day occurred at the village of Cagny, which lay towards the south of the armoured corridor. The village was targeted accurately by Bomber Command during the initial air bombardment.

The German defences in Cagny and the nearby villages comprised elements of 21st Panzer Division commanded by Major Hans von Luck. They included two battalions of panzer grenadiers, artillery and self-propelled guns. Von Luck’s account of the subsequent fighting at Cagny has been retold many times over the years but not always very faithfully. Returning from leave in Paris early on 18 July and still clad in dress uniform, he missed the Allied bombing but arrived at his command post at Frénouville to discover that all communication had been lost with his divisional headquarters and the more forward units. He therefore set out in a command tank for Cagny. He found the village in ruins and saw the tanks of 11th Armoured Division streaming past to the west.

Major Hans von Luck

Deciding that there was no hope for his forward units, von Luck was on the point of returning to Frénouville to organise a new defence line, when he saw the barrel of a single cannon pointing skywards over a wall near the church. Investigating further, he found a battery of four 88mm flak guns manned by Luftwaffe crews, which had somehow survived the bombing. The famous ‘88’ was a dual-purpose gun. Under the Luftwaffe, it functioned primarily in the anti-aircraft role, but it could also be used as an anti-tank gun, and ‘88’ batteries in Normandy were usually equipped with armour-piercing shells as well as anti-aircraft ammunition.

Von Luck found the Luftwaffe commanding officer - a captain - and ordered him to move his guns to the northern edge of the village to engage the British tanks. The man refused, stating that he was under Luftwaffe command and that the ground battle was none of his business. Von Luck’s response was to draw his pistol and ask the captain if he wished to be killed or decorated. He opted for decoration.

They both then made their way to an apple orchard on Cagny’s northern perimeter that offered clear views out across the arable fields, and von Luck gave orders for the battery to be moved to this new position. He told the Luftwaffe captain: ‘The corn over there is so high that you will be well protected and just have a field of fire across it.’ He also gave instructions to ignore the forward tanks and target those coming from the north instead. The guns were duly deployed, while von Luck returned to Frénouville.

An 88mm flak gun used in Operation Goodwood,
sometimes wrongly depicted as one of the Cagny guns, 

Back at his command post, von Luck discussed the situation with a number of other officers, including Major Alfred Becker, who informed him that two of his assault gun batteries would ‘be going into action at any minute on the right flank’. Becker also described how a Major Kurz had already set up a defensive front in this area.

Later, von Luck returned to Cagny, stopped by the church and ran to the position still occupied by the Luftwaffe battery. ‘In the extensive cornfields to the north of the village stood at least 40 British tanks, on fire or shot up. I saw how the tanks that had already crossed the main road were slowly rolling back. Becker’s assault guns had also joined in the battle. From the right flank, they shot up any tank that tried to bypass the village.’

The British perspective on these events has most commonly been associated with one particular element of 11th Armoured Division - Second Battalion, Fife and Forfar Yoemanry. The Fife and Forfars’ were the first to come under heavy fire from Cagny, with predictably dire consequences for the British tanks.

To the southwest, the battalion commander realised that the units behind him had come under attack and ordered his motorised infantry to take the village. However, the divisional commander, Major General ‘Pip’ Roberts, then learnt of the order and cancelled it, determined that there should be no loss of momentum and anxious to conserve his remaining infantry. Instead, he passed on the problem of Cagny to the Guards Armoured Division, telling them that it was ‘strongly held’. Heeding this intelligence, the Guards approached the village with great caution and took the whole afternoon to capture it. A considerable gap developed between 11th Armoured Division and the Guards, and, as we have seen, only 11th Armoured Division reached the Bourguébus Ridge on 18 July.

With hindsight, Roberts later blamed himself for passing an unduly pessimistic assessment of German strength at Cagny to the Guards. The Guards have also faced criticism for their alleged timidity. Yet the evidence suggests that neither of these critiques is entirely justified.

For many years, histories of Operation Goodwood tended to focus too much on the story of the Luftwaffe 88s at Cagny. Their precise location was something of a mystery and was the subject of much disagreement. Other aspects of von Luck’s account, notably his references to Becker's assault guns, received far less attention, although it was firmly established that at least one Pak 43 88mm anti-tank gun also participated in the village’s defence.

A Pak 43 88mm anti-tank gun and its wheels at Cagny;
this air photograph was taken in 1946.

Possibly the same gun or one very nearby, photographed
after Cagny was captured during Operation Goodwood.

Another view of the same position.

The same location today (Google Earth).

Location of the Pak 43s

More recently still, historians of Goodwood have studied air photographs taken on 18 July 1944 in the heat of battle. There has been particularly close scrutiny of imagery captured by a 16 Squadron Spitfire, which showed the north-west quarter of Cagny - extensively bomb-damaged - and the fields beyond, where several destroyed British tanks were clearly visible. However, there seemed to be no sign of von Luck’s 88mm flak guns. At best, only a few vehicles that were presumed to be German could be spotted.

One of the 16 Squadron photographs of Cagny, taken between
1130 and 1200 on 18 July 1944.

The most obvious problem involved in using this imagery to support historical investigation is the lack of resolution. Artillery weapons do not possess solid shape and mass. The only significant mass is located in the breech area, so artillery pieces often appear in wartime air photographs as little more than black spots. Extremely high resolution is necessary to disclose anything more. Second World War guidance to photo interpreters on the identification of artillery batteries focussed entirely on the tell-tale signs of gun emplacements and other prepared positions - not on the guns themselves. 

Also, mobile guns were frequently sited under trees, where they were invisible to aircraft. There were areas of tree cover west and north of Cagny that could have concealed tactical objects from the eye in the sky, as well as areas where such objects would have been very hard to distinguish because the ground had been severely scarred by bombing.

By contrast, tanks have considerably greater mass and usually show up more clearly from the air; a number are indeed immediately visible in the 16 Squadron photos. Yet many others are hard to spot. Their tendency to blend into the Normandy wheat fields was exacerbated by the fact that, during Goodwood, the vast majority were covered in dust, and tanks that had burned out typically became discoloured too. For all of these reasons, their presence in the air photos was often revealed only by tracks through the crops.

Thus, it is important that we should understand the limitations of the air photographs. And yet the 16 Squadron images do reveal the location of the Cagny defences with a fair degree of clarity for a reason unwittingly identified by the late Ian Daglish in an appendix of his book, Over the Battlefield: Operation Goodwood. There, he provided the following commentary, including references to a British Army battlefield tour:

Apart from the industrial estates of Caen, which nowadays spread across the western flank of the armoured corridor, the most conspicuous change from 1944 to be noted on a summer’s day is the appearance of the modern crops. Today’s much-modified strains are designed to direct more of their energy into the edible crop rather than the stem; in 1944, the ripening corn stood much taller. One point which the Battlefield Tour states entirely correctly is that “during the battle, the crops were shoulder high and it was hard to locate such field defences as were sited in the intervening ground.” Then, as in all previous summertime battles of European history, wheat grew to the height of a man’s shoulder or higher.

Examining the photos, a former colleague - an ex-RAF intelligence officer and an experienced imagery analyst - asked me the question: ‘What do you suppose these lines are?’ Sure enough, at high magnification, it was possible to see more or less straight lines running across the photograph, chiefly between Cagny and the advancing British tanks. He postulated that they might have been caused by artillery shells passing through fields of wheat, but it was Daglish who provided confirmation by establishing that the crops were so much taller than those grown today. Subsequent re-examination of von Luck’s memoirs pointed to the same conclusion. In his description of the action at Cagny, he described how ‘The 8.8cm cannons were firing one salvo after the other. One could see the shots flying through the corn like torpedoes.’

In the process, they flattened the crops and left tracks that showed up clearly from the air. In their appearance and orientation, the tracks were entirely different to 11th Armoured Divisions numerous tank tracks, which also featured prominently in the imagery. The shells left single tracks, whereas the tanks always left two. Moreover, the orientation of the shell tracks - from south and east to north and west - differed completely from that of the tanks, which were moving from the north towards the south and southwest. Further compelling evidence lay in the fact that, in areas of the photographs, several shell tracks suddenly disappeared at field boundaries. This reflected their passage from fields of high wheat into fields with lower-standing crops or wheat that had already been harvested; wheat sheaves were actually present in at least two areas, where several shell tracks abruptly stopped. Elsewhere, shell tracks ended at the carcasses of destroyed British tanks or narrowly missed them.

The main defended locations in Cagny suggested by the shell tracks.

Some of the British tanks destroyed or disabled northwest and west of Cagny
on the morning of 18 July 1944. There are undoubtedly more in this
photograph, but they are extremely difficult to identify with certainty.

Recorded in the photographs was the precise trajectory of the German artillery shells. This meant, of course that the lines could be traced back to their points of origin to show where the guns were.

There were evidently several variables that determined the propensity of shells to leave visible tracks. Apart from the height of the wheat and progress with the harvest, these would have included the height of the guns from the ground, their elevation and aim. Nevertheless, such tracks as can be clearly identified confirmed that the guns were not deployed in a single area of Cagny but were spread along woods and hedgerows north and northwest of the village, where a pronounced right-angle feature is created by the field boundaries.

Some guns were positioned east to west on Cagny’s northern edge, from where they could fire north and northwest. Some were lined up south to north on a field boundary and were positioned to fire west. A considerable number of shell tracks are visible in this area, and objects can be seen at track points of origin. These have too much mass to be artillery pieces. Moreover, while some points of origin are occupied, others are not. It therefore appears that there were tanks and/or self-propelled guns in this area, possibly including Becker’s assault guns, which were actually converted French Hotchkiss tanks. In summary, by the time this photograph was taken, the British tanks would have been confronted by a murderous crossfire.

German positions extending north from Cagny
revealed by shell tracks through the wheat.

The photographs suggest that the Germans were
using tanks or assault guns in this area; the object
arrowed here has the appearance of a tank.

This object has too much mass to be a static
artillery gun and is probably an assault gun.

One of Becker's assault guns in Normandy

There is but one location in the photographs that conforms with von Luck’s description of where the Luftwaffe 88s were positioned. This is the only point northwest of Cagny church where an orchard opens straight on to the wheat fields north of the village, and where there would have been unimpeded views across these fields. It is also served by a direct road from the church, along which the guns could easily have been transported. Most of all, it is the only location on the northwest edge of Cagny from which multiple shell tracks emanate. However, while several military objects are visible in the imagery, the guns are impossible to identify with certainty. Lack of resolution, tree cover and severe bomb scarring all reduce the utility of the photos in this area.

The objects arrowed here may be the Luftwaffe 88s; the shell tracks reveal that
a considerable volume of fire was directed north and northwest from this area.

The same objects can just be made out in this oblique air photograph taken by
430 Squadron; also captured in the image is a single clear shell track that
leads back to this area.

Enlarged, the objects are clearer; shadow is also visible, demonstrating that
they are three-dimensional.

Otherwise, apart from the Pak 43, just a single object in the imagery has the basic shape of an artillery gun. This can be seen in a field immediately north of Cagny. Its barrel is visible and is pointing in the direction of the British armoured corridor. The gun is also clearly the point of origin for at least one of the shell tracks through the wheat.

One of the guns of Cagny and the orientation of the shell track.

Again, the dimensions of the gun are somewhat clearer when the photo is enlarged.

An associated and equally enduring part of the Cagny story concerns the fate of two German Tiger 1 tanks, part of 3 Company, 503rd Heavy Panzer Battalion, under Lieutenant Richard Freiherr von Rosen. The company had been positioned at Emiéville, northeast of Cagny, on the morning of 18 July and had suffered heavy losses during the bombing that preceded the British armoured offensive.

Nevertheless, von Rosen succeeded in extricating six tanks and began an advance west towards the defended farm of Le Prieuré at about 10 AM. Subsequently, to skirt an area of woodland and keep it between his force and the British armour to the west, he turned south towards Cagny. Then, following an engagement with two British Shermans, he lost two of his Tigers in rapid succession. Both suffered penetrations of their thick frontal armour, suggesting fire from powerful anti-tank weapons positioned ahead of them. Unnerved and unable to pinpoint the source of the incoming fire, von Rosen decided to withdraw.

For many years since, 3 Company’s fate has been the subject of much speculation. To some, at least, it appeared quite plausible that the two Tigers were the victims of friendly fire: that they were targeted by the four Luftwaffe 88s. Others have cast doubt on this proposition. The imagery cannot settle the argument once and for all. However, it does confirm that at least two German tanks advanced south towards Cagny from the direction of Emiéville on the morning of 18 July 1944. This is clear from the tracks they left through one of the wheat fields. They emerged from the hedgerow north of the village. One was destroyed soon afterwards, while the other joined a track heading towards Cagny. Its subsequent fate is unclear, but a shell track strongly suggests that it engaged a British tank that was also under fire from the south. The tanks had apparently been located by someone who examined the hard-copy prints at high magnification many years ago, as their location had been marked on one of the frames with a black pen.

The area where the tank tracks emerged, and the orientation of the shell track.

The tracks left by one of the tanks.

The second set of tank tracks, slightly further north; the shell track originates 
in this area and does not appear in the wheat field right of the farm access lane.

In summary, the scope and limitations of the air photographs must be kept firmly in mind when they are used as historical sources. They may contain a wealth of detail, but there are significant constraints in terms of both resolution and clarity. Beyond this, the air photos record only one specific situation at a particular time. They tell us far less about the previous chronology of events at the location concerned.

Nevertheless, the imagery captured over Cagny on the morning of 18 July 1944 broadly supports Hans von Luck’s story. Von Luck made no claims whatsoever about the Luftwaffe flak battery’s actions during his first visit to Cagny that day. He stated only that he ordered them to move to a location northwest of the village - an orchard that offered clear views out across the wheat fields and towards the British armoured corridor. His account of their subsequent role was based on his second visit to Cagny, by which time, on his own admission, other German units, including Becker’s, were active in the area.

In keeping with this narrative, the imagery confirms the presence of extensive German defences north and northeast of Cagny by the mid-morning period, which were inflicting significant losses on the British armour. Heavy fire was clearly emanating from the one area of the photograph that matches von Luck’s description of where the Luftwaffe 88s were positioned, and objects are visible at this location in both vertical and oblique imagery. The Germans also deployed anti-tank guns and self-propelled assault guns around Cagny.

Beyond this, the photographs record the movement of German tanks from the northeast towards Cagny and the apparent destruction of at least one of these tanks near the village. However, the unit involved and the reason for their movement south must remain the subject of conjecture, and we will probably never know for certain whether the tank was destroyed by Allied or German fire.

Blue-on-Blue: Bomber Command and Operation Tractable, 14 August 1944

Examined at the tactical level, the Normandy campaign sometimes appears as one long sequence of blue-on-blue or 'friendly fire' episodes. However, hardly any were actually caught on camera. One of the few that were involved RAF Bomber Command and First (Canadian) Army during Operation Tractable on 14 August 1944. This blog draws heavily on unpublished research by Group Captain Steve Lloyd RAF (Rtd), formerly of the Air Historical Branch.

From July 1944 onwards, all the major set-piece Allied ground offensives in Normandy were preceded by heavy bombing operations against the German lines undertaken by the strategic bomber forces – Bomber Command and the US Eighth Air Force. Both forces were assigned at virtually no notice to a ground support task that differed fundamentally from their normal strategic role. The importance that Allied ground commanders attached to these operations may be measured from their repeated requests for heavy bomber support, which continued throughout July and into August, September and October.

Nevertheless, from a tactical perspective, this approach to offensive warfare raised a number of acute difficulties. Originally, Allied commanders intended the heavy bombers to bomb gaps in the German lines. However, while they certainly inflicted significant losses on German ground forces in Normandy, it was soon established that their primary effects were not so much physical as psychological – the shock and disorientation inflicted on the defenders. Yet these effects were temporary. They had to be exploited quickly by Allied troops on the ground, who had therefore to be deployed well forward when the bombers attacked. This inevitably increased the danger that they themselves might fall victim to any bombs dropped off-target, or to the tendency for bomb patterns to ‘creep back’ from their target areas.

A variation on the use of heavy bombers to bomb through the German lines involved using them to protect the flanks of advancing Allied forces. This could be achieved through the direct targeting of threatening German units or through extensive cratering to make ground impassable, or through a combination of both methods. Nevertheless, even then, quite close proximity between the bombed areas and Allied ground forces was an unavoidable feature of such operations.

Protecting the flanks: the results of Bomber Command 
bombing at Sannerville during Operation Goodwood, 
18 July 1944

Complicating matters further was an absence of formal command and control machinery for the use of heavy bombers in a ground support role. The formal command structures through which the Allied armies dealt with their tactical air forces were entirely absent, and planning was conducted on an ad hoc basis. It was not especially difficult for Allied air commanders to understand the essentials of British and American ground plans in Normandy, but their Army counterparts, for reasons that are entirely understandable, had virtually no grasp of the capabilities, limitations, tactics, techniques and procedures of the strategic bomber forces.

On 13 August 1944, with German forces resisting stubbornly on the road from Caen to Falaise, 21st Army Group decided to launch an attack from the north-east across the River Laison - Operation Tractable. An infantry and an armoured division from First Canadian Army were to advance from the area Soignelles-Estrees-La-Compagne. The object was to seize the high open ground southeast of the River Laison. After this had been accomplished, they were to continue the advance and capture the ridge dominating Falaise as a prelude to the capture of the town and the closure of the so-called ‘gap’ between Falaise and Argentan, cutting off German forces to the west.

Intelligence suggested that a powerful German armoured group was positioned in the area of Soumont-Le-Quentin, quite close to the Falaise road. First Canadian Army therefore requested that Bomber Command bomb this area after the ground assault had been launched, so as to prevent a German counter-attack on the Canadian right flank. The request was first submitted on 12 August in the expectation that the attack would commence the following day. However, Tractable was then postponed by 24 hours, and further discussions took place on the afternoon of the 13th on the employment of heavy bombers. Six aiming points were selected at Quesnay, Aisy, Soumont St Quentin, Bons Tassilly, Fontaine-Le-Pin and an area north of Hamel-Le-Marais. Heavy bombers were to start bombing the area two hours after the launch of the ground assault.

The Operation Tractable plan, involving a ground advance to the
east while heavy bombing protected First Canadian Army to
the west

This meant that the main target areas were on approximately the same latitude as the advancing Canadian forces, albeit further to the west. Moreover, to reach their aiming points, the bombers would have to fly over extensive rear logistical and assembly areas positioned to sustain the ground offensive.

Second Tactical Air Force was also to provide air support. Thirty minutes before H-Hour, medium bombers of 2 Group were to attack three defended localities in the thickly wooded valley of the Laison. At the same time, fighter-bombers of 83 and 84 Group would attack strong points and gun positions on the high ground between Olendon and Sassy, south-east of the Laison.

On 14 August, visibility was very good, but there was a slight wind blowing from the north. The Canadians moved forward from their start line at 1200 hours and made satisfactory progress. Bomber Command dispatched 811 aircraft, which began to arrive over the area at 1400 hours. The aim points were correctly marked and heavy concentrations of bombs fell on all six, a total of 3,723 tons being dropped by all the aircraft. Air photographs taken shortly afterwards confirmed the accuracy of the majority of the bombers, but the attack was called off at 1530 – half an hour early – after a number of aircraft dropped their bombs between four and six miles north of the target area. In total, 77 bombers were involved, 44 of which ironically belonged to 6 (Royal Canadian Air Force) Group. A total of 112 troops were killed while 142 were posted missing, and 376 were wounded; 265 vehicles, 30 artillery guns and two tanks were destroyed.

Bomber Command aircraft bombing Quesnay
during Operation Tractable

A field at Quesnay that still bears the scars

The best-known photograph of the blue-on-blue incident;
Allied troops on the Caen-Falaise road look on in horror 
as bombs fall behind them near the Haut Mesnil quarry

The Air Officer Commander-in-Chief, Bomber Command, Air Chief Marshal Sir Arthur Harris, subsequently ordered an enquiry and produced a report on the incident. In it, he addressed the key issues under three basic headings that might be described as strategic, operational and tactical. The fundamental strategic issue was that Bomber Command was a strategic bomber force and was trained as such – typically for night operations against urban and industrial targets in Germany that could hardly have been more different from the tactical support operations they found themselves conducting in Normandy. The Command was therefore trained primarily to operate in darkness, and it was not trained to fly in formation. In the dark, aircraft operated individually according to an overall plan, navigating by prescribed routes to the neighbourhood of their objective and bombing under conditions in which no details of the ground could generally be seen. Bombing was guided by pyrotechnic markers of varying types placed on or near the objective by Pathfinder aircraft. Subsequent corrections were made on the instructions of so-called Master Bombers.

Neither the pilot nor the navigator had much view of the ground. The air bomber alone could direct the aircraft  by means of map reading with reference to the ground beneath, but his training in this field was limited, mainly because the metier of the force was night bombing. Locating the aircraft’s position by map-reading from objects on the ground was normally impossible in darkness.

Air Chief Marshal Sir Arthur Harris, AOC-in C 
Bomber Command

As the task of supporting ground troops had only very recently been assigned to Bomber Command, it was not, of course, realistic to overhaul their entire training regime, and it would have made no sense to do so, in any case, given that they were still primarily intended to conduct the strategic night bombing role. As Harris put it,

This training and the organisation which produces it cannot be changed at a moment’s notice, or indeed at many months’ notice, to enable the force to operate with 100% efficiency in so entirely different a role as close direct support to troops on the ground in daylight … With these facts in view, when urgent demands began to arrive from the army for assistance by the heavy bomber force, it was invariably pointed out that we would do our best within the limits of our experience and training to meet their requirements, but that grave risk inevitably existed in these circumstances of some bombing going astray and taking effect upon our own troops. Nevertheless … the Army authorities concerned expressed ready acceptance of the risk in exchange for the approved decisive effects of such bombing in forwarding their military plans.

Under the operational heading, Harris pointed out that consideration of the Operation Tractable plan had only started on 12 August. This reflected the fact that the broader Allied plan for cutting off the German retreat from Normandy by closing the gap between Falaise and Argentan was only agreed on the 9th. Tractable was very much a last-minute venture devised in great haste. Bomber Command considered attacking the target areas by approaching them from the west. However, this would have required a routing directly above the German lines, during which time their aircraft would have been exposed to heavy anti-aircraft fire. Quite apart from the risks involved, only a few days before, during Operation Totalize, heavy bombers of the Eighth Air Force routing towards the battle area from this direction had been scattered by German flak, and some had mistakenly bombed Allied ground forces as a result. Fearing that something similar might happen in Tractable, Harris insisted that his aircraft should fly north to south instead.

Originally, Bomber Command argued that they should strike the target areas starting with those furthest south and progressing to the more northerly locations. This was due to the forecast northerly wind direction. Harris was concerned that, if his aircraft bombed north to south, many of the more southerly aiming points and target markers would be obscured by smoke and dust during the later stages of the operation. However, First Canadian Army contended that the targets should be attacked from north to south to conform with their projected advance. Understanding their perspective, Harris agreed. The risk that the markers might be obscured was, in his words, ‘accepted at the particular insistence of the Army’. He felt the risk was justified given the potentially decisive role of the heavy bombers.

The most chilling of the surviving photos; the location
 is St Aignan, although the photo is labelled
'Falaise area'

At higher magnification, numerous Allied vehicles can
be seen; no such clarity would have been available to
the air bombers thousands of feet from the ground

Finally, at the tactical level, Harris’s investigation soon identified a familiar pattern. The bombing errors started when a Pathfinder aircraft and 13 bombers of 4 Group and 6 Group bombed an area around St Aignan. At the time, the Master Bombers, who were over the correct target area further south, instructed all crews to bomb on the yellow target indicators. Several of the crews who bombed St Aignan reported seeing yellow target markers in this area. Another 12 aircraft then released their bombs in the same vicinity after a Pathfinder incorrectly marked the target, having again seen yellow lights on the ground.

Harris then described how another 23 aircraft from 6 Group bombed the quarry at Haut Mesnil in error.

This was started by two aircraft of No 428 (RCAF) Squadron who bombed almost simultaneously. They had been briefed to expect to see smoke arising from aiming point 21, which should have been bombed by another force before their arrival. They appear to have mistaken the smoke rising from the erroneous bombing near St Aignan for the smoke from aiming point 21. The Master Bomber concerned, who was operating on and viewing the correct target, instructed arriving aircraft to ‘bomb yellow target indicators. You will find them when you have passed the first column of smoke.’ This appeared to fit the picture as these aircraft saw it and they also claimed to have seen yellow target indicators burning in the neighbourhood of their proposed target.

Next, an aircraft from 460 (RAAF) Squadron struck the same area, after the air bomber saw what he thought were red target indicators burning on the ground, ‘which he had previously seen cascading’, i.e., falling through the air. Another aircraft from 460 squadron then bombed what he thought were yellow target markers. This was enough to draw in a further 24 bombers from 1 Group, which all released their loads in the vicinity of the Haut Mesnil quarry.

An extraordinary photograph that captures the bombing
near the Haut Mesnil quarry

At very high magnification, it is possible to see the white Allied 
recognition stars on the vehicles; again it should be stressed that
the air bombers would have had no comparably detailed view 
of the ground

All the mistakes could have been avoided if the air crews involved had correctly estimated the interval of time between crossing the French coast and arriving over the target area, and they were under clear instructions to do so. However, this technique inevitably involved an element of approximation in itself. When their knowledge of the elapsed time was contradicted by a visual picture in which the location of the target areas was ostensibly confirmed by target markers, smoke and the sight of other aircraft bombing, and by the instructions of the Master Bombers, they chose to accept what appeared to be the evidence of their eyes and ears. At the same time, due to the direction of the wind, they were unable to see the target markers released further south over the correct areas. In Harris’s view, the aircrew ‘too light-heartedly abandoned’ their timed approaches from the coast. The perspective of the crews would doubtless have been very different, but several were afterwards disciplined, and revised instructions were issued for operations by Bomber Command in support of the Army.

Clearly, the appearance of target markers in the wrong locations played a key role in the Tractable blue-on-blue incident. After it was firmly established that the indicators could not have been dropped by Bomber Command Pathfinders, Harris sent an officer – Group Captain SC Elworthy (later to be Chief of Air Staff and, subsequently, Chief of Defence Staff) – to France to investigate. He duly reported to Headquarters, 84 Group (which was co-located with Headquarters First Canadian Army) on 20 August. There, he saw the Senior Air Staff Officer (SASO) and the Group Captain Operations and explained the sequence of events as they were so far understood. In particular, he raised the question of the yellow lights, which had allegedly been seen on the ground around St Aignan and the Haut Mesnil quarry. He was ‘immediately told that yellow smoke, flares and celanese strips were the standard signals for troops of all allied armies in France to indicate their positions to our own aircraft’. He was shown a copy of the First Canadian Army Operations Standing Orders confirming that this was so. Elworthy’s subsequent record deserves to be quoted verbatim.

I pointed out that crews claimed to have seen yellow lights looking like target markers before any bombs fell in the areas occupied by our own troops, and asked, therefore, whether these yellow flares and smoke would only be fired by troops after they had actually been attacked by friendly aircraft. I was informed that, owing to there having been numerous minor incidents involving the attack of friendly troops by our own tactical aircraft, it was quite common for these signals to be displayed if the troops imagined they were going to be attacked ... I could get no direct evidence that yellow flares were alight on the ground prior to the first bomb falling behind our lines, but it would appear most probable that they were. The manner in which these yellow flares, smoke and strips are laid in such circumstances is in the form of a cluster, and thus they resemble the pattern of target indicators on the ground.

No information had previously been given to Bomber Command that this system of ground-to-air recognition was in use. Even more remarkably, Harris’s SASO, who had hurriedly arranged the operation with First Canadian Army in France on the 13th, had specifically asked if any confusing pyrotechnics might be employed by the ground troops, and had been assured that they would not.

Bombs falling near St Aignan on 14 August 1944; the
Army's yellow recognition flares are clearly visible.
Curiously, this photo was incorrectly labelled as
Bons Tassilly

Again, at very high magnification, it is possible to see
Allied vehicles in the photograph

The appearance of the yellow flares was thus explained. However, at least one of the crews reported seeing red target markers too, both in the air and on the ground. Where could these have come from?

Shortly after Operation Tractable, a sensational story appeared in the press, written by war correspondents in Normandy. It described how a wholesale disaster had had only narrowly been averted during the operation following the intervention of several of the Army’s Auster observation aircraft, which flew across the areas being bombed in error, firing red Verey lights. The story suggested that this courageous action prevented the entire bomber force from committing the same mistake. Harris perfectly summed up the reality:

My comment on this is that in the first place the rest of the bombing was under way, firmly controlled by the Master Bombers and achieving excellent results on the correct aiming points. In the second place, red Verey lights fired into smoke or seen through smoke burning on the ground are likely to and did, in fact, give a misleading imitation of target indicators. However well intentioned, therefore, these Auster aircraft succeeded only in making confusion worse confounded.

Two very obvious points emerge from the story of the Operation Tractable blue-on-blue incident. First, when operations are planned at the eleventh hour, there is an increased risk of planning blunders or mistaken actions at the tactical level, which, among other things, may jeopardise clear combat identification. Second, particular care is needed when joint operations are mounted that involve ground and air forces unfamiliar with one another and unused to collaboration.

The precise impact of the Operation Tractable blue-on-blue incident is difficult to gauge. The Allies had hoped that Falaise would be captured on 14 August, but Canadian troops did not enter the town until the 16th. This was largely because of German resistance north of the town and was not a direct result of the fact that St Aignan and Haut Mesnil were bombed in error by Bomber Command. Nevertheless, it is most unlikely that the blue-on-blue exerted no influence at all on the progress of the Canadian offensive and, to that extent, it must be seen as one contributory factor in the Allies’ failure to close the Falaise-Argentan gap before many thousands of German troops escaped through it.

First (Canadian) Army's advance towards Falaise, showing the
intended bombing areas for Operation Totalize and Tractable

The area of the Operation Tractable blue-on-blue