4

THE PRELIMINARY BOMBARDMENT:
21–24 SEPTEMBER 1915

The preliminary bombardment began between 7 and 8a.m. on 21 September 1915. Shells continued falling on the German lines in varying intensity until the morning of the attack, four days later. It was of vital importance. If the infantry were not to be ‘hung up’ in no-man’s-land, the thick belts of wire and the numerous dugouts and machine gun posts that protected the enemy trenches would have to be cleared away. The German artillery, lurking somewhere behind their defences, would also have to be engaged vigorously, or its efforts to hamper British troops and support counter-attacks could prove devastating. But these tasks were not all completed and the results of the bombardment were disappointing. Although considerable damage was inflicted on certain parts of the German line, enough of their defences were left intact, including most of their artillery, to pose a serious threat to the plans drafted by General Sir Douglas Haig.

Little of any real worth has ever been written about the preliminary bombardment at Loos, the accounts in the British Official History and Robin Prior and Trevor Wilson’s Command on the Western Front (1992) being the most useful.¹ This lack of attention is understandable, given the reliance placed upon gas and the well-known shortage of ammunition, but it is not altogether excusable and it underrates the importance of the bombardment. It was not an event separate from the infantry attack, but the first act of the fighting, which had a great effect on what would subsequently occur once the infantry left their trenches. Why did the preliminary bombardment fail to achieve its objectives? The shortages in artillery ammunition that dogged the BEF throughout 1915 have been well documented.² Less well explored has been how the British used the meagre resources they did have, and despite the technological limitations that the artillery was operating under, it is clear that, as had happened with the gas, the preliminary bombardment was spread not only over a wide length of front, but also to a considerable depth into the German position. This meant that although many more guns and a much greater quantity of ammunition were available at Loos than had been the case earlier in the year, its effect was not proportionally greater. The bombardment was dilute and hampered by poor staff work and bad weather.

By the first week of September the basic outline of what form the bombardment would take had been established. First Army issued its ‘General Principles for the Attack’ on 6 September.³ After discussing the role of the infantry and its reserves, the paper listed the aims and objectives of the preliminary bombardment. Spread over several days, it was to be a ‘deliberate and carefully observed’ shoot that would cut the German wire along the whole front of First Army and destroy not only the enemy’s observation posts, but also his strongpoints situated immediately behind the front line. The destruction of enemy batteries and the placement of barrages on German communication trenches in order to prevent, or at least hinder, the movement of reliefs and supplies, were further aims. Once the attack was launched, British guns were expected to support the infantry and to ‘gain superiority of fire over the hostile artillery’.

Although ‘General Principles for the Attack’ must, to some extent, be seen as a ‘wish-list’ of what, given favourable conditions, needed to have been achieved, it was remarkable for its ambition and optimism, and had little relevance to the poor state of British artillery during 1915. Why was this so? It seems that the type of bombardment chosen at Loos stemmed from a number of sources, including the lessons divined from British attacks earlier in the year, Haig’s attitude to artillery and the continuing influence of pre-war gunnery techniques. The former was probably the most pressing influence on the bombardment plans for the battle. During 1915 First Army had embarked upon three major offensive operations: the Battle of Neuve Chapelle (10–12 March), the Battle of Aubers Ridge (9 May) and the Battle of Festubert (15–27 May). They had all been preceded by artillery preparation of varying lengths and effectiveness. While short, intensive bombardments had been fired at both Neuve Chapelle and Aubers Ridge, Festubert had witnessed a much less intensive, more methodical shoot that had been spread over several days.

How effective were these bombardments and what conclusions did the British High Command draw from them? Prior and Wilson have examined Neuve Chapelle and Aubers Ridge in some detail, so it will sufficient to summarise their conclusions.⁴ The fierce thirty-five-minute bombardment prior to the attack at Neuve Chapelle had been the most effective. Even though the amount of ammunition and the number of guns used were minuscule when compared with those employed later in the war, First Army managed to achieve a density of fire upon the German front line that was not surpassed or even equalled until 1917. Although fire support was poorer on the flanks, the effects of the shelling in the central section of the German line were devastating, with trenches and wire being completely pulverised. But the subsequent break-in could not be converted into a break-out. Isolated enemy strongpoints held up repeated, albeit clumsy, British attempts to get forward. Because British infantry were not yet equipped with the array of equipment and fire-support weapons that would be placed at their disposal during 1917 and 1918, they found it virtually impossible to work their own way forward into the German defences without close artillery support. This secondary stage of the battle, however, seems to have obscured the success achieved during the initial shelling. Thus any future preliminary bombardment would now not only have to smash the enemy front line, but also eliminate those strongpoints deeper into the defences that had proved so stubborn at Neuve Chapelle.

That the next attack at Aubers Ridge was a ‘serious disappointment’ was clear to everyone concerned.⁵ Despite a bombardment five minutes longer than at Neuve Chapelle, it was weak in comparison. This had a devastating effect on the initial attack against the German front line. It seems that with a greater emphasis on the destruction of those strongpoints and defences further into the German position, combined with the deadening effects of wear on artillery barrels and limited stocks of ammunition, the intensity of the bombardment reached only a fifth of that deployed at Neuve Chapelle.⁶ The infantry suffered accordingly. Against strengthened German defences, only small groups were able to effect temporary lodgements into the enemy line. So bad was the congestion and confusion in the British front trenches that the attack was called off the following morning. Casualties numbered over 11,000. From the failure at Aubers a number of conclusions were drawn. According to Martin Samuels, Aubers Ridge marks ‘the turning point in British tactical development… in favour of a separation of firepower and assault powers’.⁷ Still not perceiving the importance of obtaining a sufficient weight of fire, both Haig and Lieutenant-General Sir Henry Rawlinson (GOC IV Corps) now rejected the short, intensive bombardments they had previously employed. Given adequate amounts of artillery and shells, they now believed that only a long and methodical bombardment could do enough damage to the German defences to make them practicable to infantry.⁸ Accordingly, First Army carried out the first lengthy bombardment (spread over three days) at Festubert between 13 and 15 May. With relatively modest objectives and a much more deliberate approach, the results of the battle were promising. Benefiting from good weather and innovative infantry tactics, First Army managed to get considerably further forward than they had at Aubers Ridge.

The success, albeit limited, of the Festubert operation seems to have had a considerable impact on Haig and his staff. At a conference on 24 August, it was unanimously agreed that only a longer bombardment should be employed for the coming battle. Such a bombardment would, it was believed, ‘tend to destroy the enemy’s morale, his observations… it could be spread along an extensive front, and the gas would probably penetrate to his dugouts, and the enemy’s artillery would have no indication of the exact point of attack, and his fire would have to be distributed and not concentrated at the real point of attack.’⁹ On the other hand, it was agreed that if an ‘intensive bombardment took place on any particular part of the front, it would be a sure indication that we were going to attack there, and the German artillery would become active in this locality at once’. There were other reasons against the adoption of a short bombardment. Brigadier-General A.A. Montgomery (BGGS IV Corps) stated that there were not enough guns and ammunition for an intensive shoot.¹⁰ Given the length of frontage to be attacked, this was a point of vital importance, but it does not seem to have been the main reason why it was not adopted.¹¹

Gas was another factor in the decision to opt for a slower, more methodical bombardment. Barely two days after he had witnessed Lieutenant-Colonel C.H. Foulkes’s demonstration of chlorine gas at Helfaut, Haig was already incorporating it into his plans. He told the assembled officers at the conference on 24 August that he was aiming at a ‘lavish use of gas’. It was hoped this would deal with one of the key elements of the German defences: the deep concrete dugouts. Brigadier-General J.F.N. Birch (BGRA I Corps) was pessimistic about the feasibility of destroying them, about ten or twelve of which would prove problematic to any infantry advance. So well protected were they that ‘it would be scarcely worth while trying to destroy them’. Gas, however, could help to neutralise them, by either incapacitating the defenders, or forcing them to the surface where they would suffer from the accompanying shrapnel bombardment.

The abandonment of ‘hurricane’ bombardments and the adoption of the French concept of ‘L’artillerie conquiert, l’infanterie occupe’ was not without its problems, however. As John Bourne has written, the ‘pre-war tactic of fire and manoeuvre was replaced by the concept of fire then manoeuvre’.¹² A more methodical shoot would abandon any hope of surprise, and allow the enemy defenders the chance to repair any damage to parapets and wire at night. It would also prove a great strain on both the guns and their crews. But perhaps the most serious problem was Haig’s chronic optimism, which infected all aspects of the plans. It seems that he either did not recognise, or simply ignored, the grave limitations British artillery was suffering from in this period. The scale and effectiveness of the bombardment would depend on what implements Haig had at his disposal. But when compared with the ambition of ‘General Principles for the Attack’, they were woefully inadequate. A lack both of guns, particularly heavier varieties, and ammunition were the most pressing concerns. And while the creation of a Ministry of Munitions and the readjustment of British (and American) industry throughout 1915 would eventually allow bountiful supplies of guns and shells to be delivered to France, not until the summer of 1917 would the necessary equipment be in place, and the techniques that had been devised and refined during ‘many a doubtful battle’ between 1914–16 finally come of age.¹³

It has sometimes been assumed by historians that the British fought the Battle of Loos ‘with little increase in artillery’ than had been employed at Neuve Chapelle.¹⁴ This is incorrect. A much greater amount of field and heavy artillery was available for the Battle of Loos than had been the case earlier in the year, but it was still not enough (see Table 1). First Army managed to gather a total of 919 guns, with 238 going to I Corps and 225 to IV Corps.¹⁵

Table 1. Number of guns involved during the Battles of Neuve Chapelle (10–12 March 1915), Aubers Ridge and Festubert (9–18 May 1915) and Loos (21–27 September 1915)¹⁶

	Neuve Chapelle	Aubers Ridge and Festubert	Loos
Field guns	400	456	594
Field howitzers	54	78	132
Counter-battery	48	44	76
Heavy howitzers	33	47	69
TOTAL	535	623	871

This may have seemed impressive, but it masked serious limitations when compared with the length of the bombardment frontage. Whereas this had been a mere 1,450 yards at Neuve Chapelle, it had swelled to 11,200 yards for Loos (see Table 2).

Table 2. Length of bombardment frontage (yards) of the Battles of Neuve Chapelle (10–12 March 1915), Aubers Ridge and Festubert (9–18 May 1915) and Loos (21–27 September 1915)¹⁷

Neuve Chapelle	1,450
Aubers Ridge and Festubert	5,080
Loos	11,200

Was the ammunition situation able to help? First Army had access to a considerable amount of ammunition, far more than has generally been recognised, but again it was not enough. First Army received 650,000 shells for the field artillery (about 1,000 rounds per gun), with 100,000 shells for the howitzers and heavy guns (about 400 rounds per gun).¹⁸

Shortages of guns and ammunition were not the only concerns. Indeed, the bombardment was littered with technical problems. Because a large number of the guns First Army was employing had been under constant stress all year, mechanical breakdown was not uncommon. On 23 September, LI Brigade RFA (9^th Division) recorded that one gun had been put out of action because of a broken trigger, and the buffer springs had failed on another gun. No spares were available until the following day.¹⁹ Similarly, VI (London) Brigade RFA (47^th Division) reported one defective gun on 23 September.²⁰ According to the war diary of VIII (London) Brigade RFA, ‘a large number of misfires occurred’ during the bombardment owing to mechanical failure.²¹ The quality of the ammunition the BEF was receiving was also far from perfect; a result of the rapid, haphazard expansion in the British munitions industry that was currently underway. Historians have estimated that around 30 percent of the ammunition fired before the launching of the Somme offensive on i July 1916 did not explode. It is probable that a similar percentage of shells at Loos were ‘duds’.²² While Robert Graves (2/Royal Welsh Fusiliers, 19 Brigade) complained that American ammunition ‘contained a high percentage of duds; their driving bands were always coming off’,²³ Lieutenant-Colonel A.G. Prothero (CO 2/Welsh, 3 Brigade) estimated that around 100 unexploded shells littered no-man’s-land in his sector.²⁴ Even those shells that did explode were not without their problems. On 18 September Rawlinson asked First Army to replace certain stocks of 15-pounder ammunition. He complained that these shells were ‘absolutely unreliable and a serious source of danger to our own infantry’.²⁵ Unfortunately, owing to the chronic shortage of replacement ammunition, First Army refused two days later.²⁶

All along the front the enemy machine gun posts were recognised as a key cornerstone of the German defence. If these were put out of action an attack would stand a much better chance of getting through. Brigadier-General Hon. J.F.H.S.F. Trefusis (GOC 20 Brigade) knew as much and told Colonel H.H. Tudor (CO XIV Brigade RHA) about his problem.

Unless we could knock out the enemy machine-guns, our attack may fail. But it is not possible to knock them out, as they can be shifted and the best way to deal with them is to blind them and without smoke I don’t see how we can.²⁷

Tudor subsequently pestered Brigadier-General Birch for smoke shells. Birch sent an officer back to England to try and get some. Unfortunately, all Birch received was a number of smoke candles. Tudor was furious, lamenting that ‘it quite missed the point’ and only shells could effectively blind enemy machine gunners. He noted that ‘these candles could not do [this] unless the wind was ideal, blowing straight across the enemy and not too strong’.²⁸ This episode amply illustrates how seriously hampered the British were in trying to pierce the enemy defences. The equipment they needed (guns, smoke shells and high explosive ammunition) was simply not available.

The lack of heavy guns and quality ammunition was of great importance, but there were more intractable problems that needed to be solved. By September 1915 British artillery was undergoing a rapid and frequently painful process of growth and development, part of what has been termed a ‘revolution in military affairs’.²⁹ The type of war that emerged on the Western Front was simply not foreseen, and if a few bright thinkers had appreciated that the next war would involve barbed wire, entrenchments and the decline in traditional artillery techniques, they were firmly in a minority.³⁰ Because the front line now stretched for over 400 miles and to an ever-increasing depth, it was no longer safe to rely on old methods. Whereas in the pre-war Royal Artillery, under the dominance of the RHA and RFA, the speed and efficiency with which guns could be drawn into position and fired was a matter of professional pride, the techniques of accurate, long-range ‘scientific’ gunnery were little understood and only practised in the ‘ghetto’ of the RGA. But these were exactly the skills now demanded on the Western Front.

While traditional direct fire over open sights had occurred since the start of the war, by as early as the Battle of Le Cateau (26 August 1914), when II Corps lost thirty-eight guns to accurate enemy counter-battery fire, it was found necessary to pull artillery back and conceal it wherever possible.³¹ The resulting decline in direct fire had profound results. Now indirect fire, or the shooting at a target which could not be seen (often because it was on the other side of a terrain feature), had to be employed. This may not have been as glamorous, it required much less ‘dash’ than firing over open sights at visible targets did, but it depended upon greater technical knowledge, which took time to learn and experience to perfect. For a shell to be effective, it had to land on, or near to, its target. But by early 1915, as Prior and Wilson have aptly explained, British artillery ‘was not likely to hit, with any certainty or regularity, a target that could not be seen’.³² Why was this so? It seems that a combination of factors prevented the artillery from being the effective weapon it would later become. Merely finding a target presented great difficulties, let alone hitting it. Although FOOs were used, good observation posts in the flat districts of northern France and Belgium were rare, and while aerial reconnaissance was becoming ever more reliable, its technological difficulties, as well as survey problems, hampered these efforts.³³ There was also a lack of understanding of the effects of wind, air temperature and wear on gun barrels, and the calibration that was needed to combat this.

Technological limitations there certainly were. Most of the field guns available at Loos were 18-pounders, the workhouse of British artillery during the Great War.³⁴ Entering service in 1904, the quick-firing 18-pounder was primarily designed as an infantry support weapon. Although available in reasonable numbers and very reliable, its weaknesses echoed those of the French 75mm. It was issued mainly with shrapnel, which although lethal against infantry in the open, was decidedly limited in trench warfare. Wire-cutting with shrapnel fired from 18-pounders had first been tried at Neuve Chapelle and would again be employed at Loos. Although recent work has shown that shrapnel shells were preferred during wire-cutting operations, the results of firing were often not terribly clear.³⁵ And even when equipped with high explosive, which was only available in limited quantities by the time Loos was fought, the shell was too small to demolish sections of enemy trenches, especially the hardened concrete dugouts.³⁶ The 18-pounder’s low angle of fire also meant that its range was limited, with shells rarely landing on top of German trenches, where they would do the most damage.³⁷ Howitzers, on the other hand, were ideal for positional warfare. Because of their steeper angle of fire they could both engage targets at much greater distances than field artillery, and land on top of trenches. But howitzers and their high-explosive ammunition were in short supply, and only 96 of the vital 6" and 4.5” howitzers were available for the main attacks. Various other types of heavy artillery were used at Loos – including 15”, 9.2", 8” and 6” howitzers – but they were small in number (totalling 29 guns), with scarce ammunition and added little to the overall weight of the bombardment.³⁸

But why was there so much difference between what British guns could realistically achieve and what Haig wanted them to achieve? Although it is clear that at least some of these deficiencies were recognised at First Army HQ – indeed Haig had been grumbling about ammunition shortages all year – too little account was taken of the poor performance of British artillery. Although it would be inaccurate to say that Haig did not understand artillery, he was not a terribly great or original thinker about gunnery, and his views remained traditional. Haig believed artillery was primarily there to assist the advance of the infantry and lower the enemy’s morale.³⁹ This was made clear by Field Artillery Training 1914:

Artillery cannot ensure decisive success in battle by its own destructive action. It is the advance of the infantry that alone is capable of producing this result… To help the infantry to maintain its mobility and offensive power by all means at its disposal should be the underlying principle of all artillery tactics.⁴⁰

Haig, therefore, seems to have shoehorned his artillery into his wider ideas on the structured battle, in particular the importance of engaging the enemy on a wide front, wearing down his forces, and then delivering the decisive blow. First Army’s artillery was to do a similar thing. They were to be spread over a wide area, wear down the enemy defences and morale, before the final blow was struck by the release of the gas and the infantry assault.

The wide front of attack was of particular importance. There was simply no way, as was recognised, that First Army’s artillery could adequately cover the required length of front. Whereas the bombardment frontage had been a mere 1,450 yards at Neuve Chapelle, this had swelled to 11,200 yards for Loos (see Table 2). It does seem that Brigadier-General Birch’s later comment, referring to the Somme bombardment of the following year, that ‘poor Haig – as he was always inclined to do – spread his guns’,⁴¹ was not too unfair, because the same process of ‘spreading’ artillery occurred at Loos. Why was this so? Much of the answer seems to lie in the particular nature of the plan First Army adopted for the Battle of Loos, particularly the great reliance placed upon the use of poison gas. Because it was believed, especially by Haig, that gas would prove utterly devastating, the artillery bombardment lost much of its raison d’être, and its importance was lowered as a result.

Throughout September artillery was assembled for the battle. South of the canal the ground was flat and under direct enemy observation so that the heavy batteries could only be assembled between 3,000 and 4,000 yards from the front line. The field batteries were a little further ahead.⁴² The guns were hidden as far as possible in the clusters of mining buildings or any folds in the ground. In some sectors of the battlefield this was not too difficult. For example, North and South Maroc,Vermelles and Grenay all offered ‘good cover’ and excellent observation posts.⁴³ The batteries then had to construct gun-pits and dugouts, arrange their communications and register shots on the German lines. The battlefield was divided into an inner and an outer zone. Field artillery firing shrapnel would concentrate on wire-cutting within the inner zone, while heavier calibres fired high-explosive shells at various selected targets in the outer zone.

How were the guns organised? As the British Official History proclaimed, Loos was a ‘landmark’ in the history of artillery development.⁴⁴ The most notable change was the creation of a formal corps artillery headquarters in early September. The idea of centralising all divisional artillery under corps control seems to have been an idea that had been milling around the BEF for some time, but as has been pointed out, the critical shortage in ammunition all year meant that there had been ‘little need’ for such a change until September.⁴⁵ Whereas BGRAs had previously only been advisers without any authority to actually command batteries, they were now able to direct all the field and heavy artillery (except those allotted to counter-battery fire) within each corps. This allowed a much greater level of co-operation to be achieved between batteries, and while there still remained some persistent problems that would not be ironed out until the thorough reorganisation of British artillery in the winter of 1916–17,⁴⁶ it was a considerable improvement.

The centralisation of corps artillery may have helped to standardise gunnery practices and clear up some lingering anomalies within corps, but organisation between corps could still differ widely. Whereas Brigadier-General Birch organised the guns of I Corps into a number of clearly defined sections, the system devised by Brigadier-General C.E.D. Budworth (formally appointed BGRAIV Corps, 10 October 1915) proved ‘a most untidy grouping, with an excessive use of groups of sub-groups’.⁴⁷ I Corps’ artillery was split into five parts. The artillery of each division contained three brigades of 18-pounders and another of 4.5” howitzers, while No. 5 Group HAR and a Siege Group, containing two brigades of RGA, completed the line-up. The organisation within IV Corps was much less straightforward, with 1^st Divisional artillery being split into different northern and a southern sub-sections and 47^th Division organised into the MacNaughten Group, the Massey Group, as well as a Divisional Reserve Group.

The organisation of artillery was not the only difference between I and IV Corps. Considerable differences existed in their respective ‘fireplans’. Birch wrote a simple step-by-step, unified plan, but Budworth’s ‘fireplan’, if it can be called such, was simply to break the bombardment down into five different tasks (ranging from wire-cutting to ‘special tasks’) and ask his group commanders to ‘discuss’ these with their divisional commanders. Why were Budworth’s plans so different from Birch’s, and why was he so reluctant to impose his will on his fellow artillery commanders? The answer seems to lie in the confusing situation that existed within IV Corps HQ during the late summer of 1915. According to Sanders Marble, Rawlinson lacked confidence in his BGRA (Brigadier-General A.H. Hussey) and had asked for his replacement on the same day that a formal corps artillery headquarters was created.⁴⁸ Approving of the new arrangements, but without the will to get rid of Hussey – perhaps the memories of the Davies affair were still sharp⁴⁹ – Rawlinson played a strange game of delegating all responsibility to Budworth (then BGRA 1^st Division), with the formation of the various groups and sub-groups, and avoiding giving Hussey anything to do. Budworth could therefore never actually impose his will on his fellow artillery commanders even if he had been possessed of such a desire.

The only artillery remaining within First Army were those heavy batteries allocated to counter-battery work, which had been retained under First Army control. This task was given to the HAR groups, but was hampered by a confusing chain of command.⁵⁰ While No. 5 Group HAR (Brigadier-General T.A. Tancred) was assigned to I Corps, and No. 1 Group HAR (Brigadier-General G. McK. Franks) supported IV Corps, they were not under corps control and were placed directly under Major-General H.F. Mercer at First Army headquarters. To compound matters further Franks was placed in charge of all counter-batteries within First Army.⁵¹ This system proved to be unhelpful and reflected the low priority given to counter-battery fire. Despite Sir John French’s promise that ‘particular attention’ would be paid to this task, relatively few guns were involved.⁵² Whereas at Neuve Chapelle just over 10 per cent of the total number of guns were assigned to counter-battery work, at Loos this had fallen to just under 7 percent. On the front of First Army, out of a total of over 900 guns, only sixty-one (split between 60-pounders and 4.7" guns) were set-aside for counter-battery work, with a mere thirty-four guns being available for I and IV Corps.⁵³ Whether this was the result of a lack of communication between Sir John and Haig is not known, although considering how stretched the artillery was in the crucial matters of wire-cutting and trench-bombardment, it is perhaps not surprising that it was relegated to a lower importance.

During the bombardment south of the canal a number of feint attacks were conducted. These were intended to confuse the enemy as to the time and date of the real attack, and compel him to man his front line and suffer casualties from doing so.⁵⁴ The feints all followed the same format. On the points of enemy trench that had been selected for demonstration, the divisional artillery would open an intense fire for five minutes. After this the guns would lift onto the second line and shell it for a further five minutes. During this lift the infantry in the front trenches would show their bayonets over the parapet, manhandle dummies into position and cheer loudly.⁵⁵ It was hoped that this would force the enemy garrison to man their trenches in order to repel the suspected attack. Infantry on adjacent fronts would then open an intense burst of rifle and machine gun fire towards the enemy positions. Finally, the enemy front line would be deluged with fierce shrapnel fire for three minutes. These feints were of some importance, but it was essential that the German defences were badly damaged before they would be practicable to infantry. But was this achieved?

Measuring the success of the preliminary bombardment is not easy. Although the available accounts, war diaries and eyewitness reports are all useful, they can be problematic. It is often impossible to give more than a rough idea of where the reporter was situated and which part of the front he was observing. Even if the location is accurately known, the observer could often only see only a small portion of the front being shelled, thus increasing the chance of misleading or inaccurate reports. It can also be assumed that the gunners of I and IV Corps only had a very rudimentary idea of what was going on at the time. The frontage of the main attack must also be considered. It totalled nearly six miles in length and was not uniform in terrain. While some parts of the German line were clearly visible from the British trenches, others were difficult to observe or covered with long grass. Attempts by the defenders to repair their trenches and wire entanglements further add to the difficulties of accurately assessing what progress was being made. Differences such as these could have a great impact on the local effect of the bombardment. These considerations make it extremely difficult to give a precise overview of the progress of the prehminary bombardment, but they do allow, however, general trends to be perceived.

The first day of the bombardment seems to have opened promisingly, even on those sectors where observation was difficult. 2^nd Division occupied what was possibly the worst sector within First Army. As its front was longer than any other division involved in the main assault, there was much more wire to cut. One senior officer later complained of ‘an enormous front – few, too few, batteries to cover us’.⁵⁶ Its guns had to bombard Mine Trench, Brickstack Trench, the German Ducks Bill, Embankment Redoubt and the Tortoise Redoubt. One battery in 2^nd Division reported that the ‘wire is not very thick and wouldn’t stop anyone’.⁵⁷ Two lanes were cut in the wire around the Brickstacks and ‘a good deal of damage’ was done to some entanglements north of the La Bassée road. Similar success was achieved by the guns of 9^th (Scottish) Division.⁵⁸ Deployed on the right of 2^nd Division, with its front dominated by the imposing shape of the Hohenzollern Redoubt, 9^th Division had much to achieve. Its artillery was to cut ‘a great mass of black wire’ on the west face of the redoubt,⁵⁹ in front of the connecting trench (Little Willie), Fosse Trench and around Mad Point. Unfortunately, the wire around this latter stronghold (opposite 28 Brigade) was largely invisible from the British lines.⁶⁰

The ground conditions in the central sector of the British line also presented considerable difficulties. Between the Vermelles–Hulluch and Loos Roads the ground was open and devoid of cover. Observation was hampered by wide stretches of long grass.⁶¹ On the left of this sector was 7^th Division. Its artillery was split into two groups and one given to support 9^th Division. This latter group was to enfilade the north and south sides of the Hohenzollern Redoubt, which although not part of 7^th Division’s objectives, could only be observed from certain areas outside 9^th Division’s zone of operations. Accordingly, two batteries were sandwiched in between 9^th and 2^nd Division and were to cut the wire in front of Madagascar Trench, Railway Trench and the wire around the formidable Railway Redoubt. The remaining artillery of 7^th Division was to bombard Breslau Trench, Quarry Trench, the south-eastern edge of Big Willie and various communication trenches that led to the German second position around Cité St Elie. It seems that the results of the first day’s shelling in this area were hardly spectacular. Although some observers recorded that the wire was ‘cut to some extent’,⁶² others were much less sanguine. According to XXXV Brigade RFA (7^th Division) wire-cutting was ‘indifferent’, and ‘very little headway seems to have been made’.⁶³

Similar unfavourable ground conditions were experienced on the front of 1^st Division. Because the ground was so open, the field artillery had to be placed further to the rear. While favourable range was between 1,800 and 2,500 yards, 1^st Division’s artillery had to shoot at over 3,000 yards. There were, however, more serious deficiencies. Because it had not originally been intended that 1^st Division would take part in the main attack, much of its artillery had been distributed to the other divisions in IV Corps. When 1^st Division was belatedly ordered to take part, only three batteries of its own artillery were restored to it. As a result, the division had to make do with two batteries from 15^th Division, one 6" howitzer battery and some inexperienced batteries from 24^th Division.⁶⁴ With these meagre resources 1^st Division’s guns had to clear the wire from the main German frontline as well as two important communication trenches (known as Alley 3 and 4) leading to Hulluch.

The final two British divisions, deployed on the right of the line, seem to have had a good day on 21 September and benefited from the elevated ground conditions they occupied. The main attack had been entrusted to 15^th (Scottish) Division, but before any advance could be made, two heavily wired strongpoints, known as the Loos Road and Lens Road Redoubts, would have to be demolished. Although much of the German wire was invisible on the left of 15th Division, visibility did improve further south,⁶⁵ and ‘satisfactory’ progress was achieved against the Loos Road Redoubt.⁶⁶ With a view comparable ‘to that obtainable over the stage of a theatre from a corner seat in the dress circle’, 47^th Division formed the hinge of the entire attack.⁶⁷ With ‘distinctly favourable’ ground, its guns were able to enfilade parts of the German line that faced away from the British trenches, giving a distinct advantage not possessed by those divisions situated further north. Its field artillery had to cut the front line wire from the Lens Road Redoubt to the mining buildings of Puits 16 and its howitzers were to shell selected communication trenches. The imposing Double Crassier, a large slagheap bristling with machine guns situated in the centre of the German line, was also to be shelled with howitzers. The first day was reasonably successful. Major Hon. R.G.A. Hamilton (CVIII Brigade RFA, 24^th Division) recorded the ‘almost unlimited allowance of ammunition’ and the ‘tremendous rate’ of fire, which ‘must be fairly drowning him [the Germans] in shells’.⁶⁸ Wire-cutting was carried out with ‘very satisfactory’ results.⁶⁹ The Macnaghten Group (firing at the wire in the second German line) also believed that its shooting was accurate.⁷⁰

It was, however, a day of mixed fortunes. While the weather was clear and fine, an easterly wind interfered with the observation of falling shot and blew clouds of dust towards the British trenches. Nearly all the war diaries and operation reports written about 21 September (and the following three days) mention the difficult weather conditions the artillery had to cope with.⁷¹ 2^nd Division complained that ‘no bombardment results [were] visible from our observation stations’ and 7^th Division admitted that its heavy guns were ‘firing long and the damage to parapets was not extensive’.⁷² But these were not the only problems. The repeated firing of the 9.2" howitzers proved too much for their platforms and a number of them broke. The guns then had to be moved.⁷³ Major-General Mercer (MGRA First Army) also highlighted a number of gunnery accidents, some involving barrels that had become clogged and shells that had not been rammed home.⁷⁴ The infantry were not the only ones who were inexperienced.

The weather conditions stubbornly refused to improve throughout the second day and the familiar complaints about poor observation were repeated.⁷⁵ But progress was being made. One observer recorded how the German trenches were ‘simply one line of busting shells and smoke’.⁷⁶ On the front of 2^nd Division, a gap of 15–20 yards was reported on the Tortoise Redoubt, the Towpath was cleared and much wire was destroyed in front of the Brickstacks.⁷⁷ Indeed, XXXIV Brigade RFA (2^nd Division) believed that by the end of the day the wire ‘had nearly all been cut sufficiently to form no obstacle to an infantry assault’.⁷⁸ Further south, the wire around Mad Point was reported to be completely destroyed and about two thirds of the wire protecting Fosse Trench was ‘gone’. The west face of the Hohenzollern Redoubt was cleared and very little wire remained around Little Willie.⁷⁹ North of the Hohenzollern Redoubt, the 104^th and 12^th Batteries from 7^th Division had an encouraging day cutting the wire in front of Madagascar and Railway Trenches, but the Railway Redoubt, however, remained undisturbed. It was still protected by thick coils of ‘unusually strong’ wire.⁸⁰ On the more favourable terrain west of Loos, 47^th Division was doing well. German wire was significantly thinner than it had been and it was noted how effective 6" howitzer fire on machine gun emplacements was.⁸¹ The Fraser Group, bombarding the German line south of the Double Crassier, reported that the trenches covering it were ‘thoroughly smashed’.⁸²

On the more unfavourable fronts the bombardment was still struggling to make enough progress. After spending over four hours in the trenches watching the bombardment with Brigadier-General Trefusis, Colonel Tudor remarked that the wire ‘seems to require a lot of cutting yet’.⁸³ Trefusis was of the same opinion and spoke to his divisional commander (Major-General Sir T. Capper) about the possibility of getting more guns for the following day.⁸⁴ Trefusis’s request seems to have had some effect as both 104^th and 12^th Batteries, which had been bombarding Madagascar Trench in 9^th Division’s sector, were moved back to support 7^th Division that day.⁸⁵ Further south, on the front of 1^st Division, it was difficult to improve the rather meagre results of the previous day. Wire-cutting was ‘slow’, observation ‘difficult’ and the whole day extremely demoralising.⁸⁶ Problems were not confined to Grenay Ridge. Despite cutting large gaps in the German wire, 15^th Division’s artillery was experiencing trouble knocking out a number of chevaux de frise placed in front of the German trenches by its occupants. These were large and difficult to destroy.⁸⁷

Meanwhile what had the RFC being doing? Attached to each corps artillery headquarters was a squadron of aircraft, which would spot targets for the artillery and help correct the fall of shot. While No. 3 Squadron was attached to I Corps, No. 2 Squadron was assigned to work with IV Corps, as well as No. 1 Group HAR for counter-battery work. The task of spotting targets and registering artillery fire was also given to No. 6 Kite Balloon section, attached to First Army. The balloon was deployed near Béthune, and as well as helping to co-ordinate the fire of I and IV Corps, it was to make observations on the strength and direction of the wind and transmit the results to General Haig’s headquarters.⁸⁸ During the first two days of the bombardment the aircraft of First Wing had been busy spotting for the artillery, and had been fortunate because of the relatively clear weather.⁸⁹ Their operations seem to have been reasonably successful. H.A. Jones recorded that ‘the squadrons were able to work to their full artillery programmes: our batteries were registered on the first and second line trenches, gun positions and dugouts’.⁹⁰ 21 September was ‘a good flying day’, with No. 2 Squadron engaging (and apparently silencing) four enemy batteries, while No. 3 Squadron did ‘much successful trench work’, including the taking of numerous photographs of the battlefield.⁹¹ So pleased was the Siege Group with the progress that No. 3 Squadron was congratulated on the ‘excellent work’ done. Air operations during the following two days ‘proceeded satisfactorily’ and on 23 September it was reported that the wire-cutting by heavy batteries attached to No. 2 Squadron was ‘apparently successful’.⁹²

Although the weather on 23 September was not bad enough to prevent artillery observation flights from taking off, the wind changed to the west and it started raining. Light and observation remained poor all day.⁹³ The volume of artillery fire did, however, increase substantially. Major Hamilton recorded the ‘simply appalling’ sound and according to Captain J.C. Dunn (2/Royal Welsh Fusiliers, 19 Brigade), the day’s gunfire was the ‘heaviest there has been’.⁹⁴ Enemy artillery, which had been noticeable by its absence, was also showing increasing signs of activity. Haig noted in his diary that First Army had received 170 casualties that day as compared with the usual number of 90 or 100 within the same twenty-four hour period.⁹⁵ Although XXII Brigade RFA (7^th Division) continued to make progress in its wire-cutting, its war diary records the ‘persistent shelling’ of its OP.⁹⁶ 2^nd Division’s artillery managed to clear another path south of the Embankment Redoubt.⁹⁷ Progress was, however, patchy. An officer in Robert Graves’s platoon complained about the resilience of the German wire, remarking that ‘our guns don’t seem to be cutting it’.⁹⁸ Indeed, according to XXXIV Brigade RFA, ‘much new wire had been thrown out during the night in front of the parapet, in loose coils’.⁹⁹ XXXIV Brigade RFA also complained about the ‘very poor attempts’ by the infantry to prevent the enemy rewiring its trenches.¹⁰⁰ But it was not all bad news. In front of the Hohenzollern Redoubt, 9^th Division continued to make definite progress. Reports indicated that the west face was clear and 250 yards of wire had been destroyed along Madagascar Trench.¹⁰¹ By the end of the day LII Brigade RFA (9^th Division) recorded that ‘practically the entire wire to be cut was cut’.¹⁰²

During the day a number of patrols reported that on 1^st Division’s front ‘favourable’ and ‘satisfactory’ progress had been made with wire-cutting.¹⁰³ The regimental history of 157 Infantry Regiment gloomily noted how its troops were ‘under heavy artillery fire’.¹⁰⁴ On the front of 15^th Division the enemy wire was by now ‘severely damaged’, and a number of chevaux de frise were destroyed by 6” howitzer fire.¹⁰⁵ By now the enemy wire was ‘sufficiently cut’ to give an attack a reasonable chance of success, although on the left of 46 Brigade, hostile fire prevented a patrol from examining the German wire. In the south of the battlefield on the front of 47^th Division, the important task of wire-cutting continued in earnest. Two large gaps had been blown in the German front line south of the Double Crassier, but north of it, things were not as satisfactory.¹⁰⁶ Reports during the night found that, contrary to earlier patrols, the wire around the Double Crassier remained impassable.¹⁰⁷

The second part of the aerial operation began on 23 September with a number of long-range bombing flights. The busy railway lines between Lille–Douai–Valenciennes were the targets for both Second and Third Wings.¹⁰⁸ Their efforts were greeted with some success and the German Official History noted that ‘considerable military damage’ was done.¹⁰⁹ Twenty-three planes of Third Wing bombed the line Douai–Valenciennes, destroying one goods train near Somain. The track near Wallers and the engine sheds north of Valenciennes were also hit. Eight planes from Second Wing and three from No. 12 Squadron (attached to GHQ) attacked the Lille–Valenciennes line, but the only damage recorded was to a signal box between Orchies and Templeuve. The attacks on the communications network did receive a boost, however, when British artillery fire destroyed the railway station at Pont à Vendin.

Rain greeted the final day of the bombardment. First Wing was able to continue spotting for the artillery, and although visibility improved during the afternoon, the low cloud and mist grounded aerial bombing flights. Despite the bad weather ‘a great deal of work was done’ with ‘numerous hostile batteries’ located and fire directed upon them.¹¹⁰ But more generally, 24 September seems to have been a mixed day. Some observers reported significant damage to parts of the German wire and trenches, while others remained much more guarded. 2^nd Division noted that lots of wire had been cleared from the first German trench, but what condition the second line was in was very much anyone’s guess.¹¹¹ There were also several worrying references to fresh wire that had been thrown out by the defenders.¹¹² On 9^th Division’s frontage, the results were again mixed. While the wire and parapets of Madagascar Trench were heavily damaged, contrary to previous information, some wire on the southern face of Mad Point was still intact. While the barbed wire left in front of Little Willie and the west face of the Hohenzollern Redoubt was deemed ‘satisfactory’, observers were unable to give a definite report on the state of Fosse Trench. The wire at the north-western end of Big Willie also remained uncut."¹¹³ L Brigade RFA (9^th Division), bombarding the cottages in Madagascar, recorded that 400 high-explosive shells were fired at this, with a good ‘visible effect’. It was, however, ‘impossible to tell what real damage had been done’.¹¹⁴

The results of the bombardment further south were equally divided. Brigadier-General Trefusis noted that although the German wire in front of his brigade was ‘nearly all cut’, there remained ‘some about halfway between our line and the Germans’ that needed to be cleared away.¹¹⁵ On the front of both 1^st and 15^th Divisions, the results of the day seem to have been satisfactory.¹¹⁶ IV Corps’‘General Progress Report’ on the evening of 24 September echoed this.

Reports from artillery and infantry observers show that practicable passages have been made at intervals all along the Fourth Corps Front. In some places the wire is reported [to be] very thick and though considerably thinned is not yet practicable to troops. In a few places in the front line and in many in the second line the wire is difficult or impossible to observe and the result is doubtful.¹¹⁷

By the evening of 24 September, after four trying and frustrating days, I Corps had almost to admit failure. Its war diary simply stated that ‘the wire was reported generally to be satisfactorily cut’.¹¹⁸ The final job was to cut a path through the British wire that lay in front of their trenches. This was cut diagonally on the evening of 24/25 September.

How effective was the preliminary bombardment? While there had been some success, particularly where the German lines lay under good observation, the cumulative effects of four days of shelling had clearly fallen short of the objectives listed in ‘General Principles for the Attack’. The German wire had proven extremely resilient to all but the most accurate shelling and still lay in large quantities along the front. Trench and strongpoint destruction had been equally haphazard and a considerable number of the deadly machine gun posts that dotted the German line were still intact and able to offer considerable resistance to any infantry attack. As befitted their limited resources, both the shelling of the second line and the attempts at counter-battery fire seem to have completely failed in gaining their objectives.

Cutting the barbed wire was perhaps the pressing and persistent problem British artillery faced in this period. But the references to uncut (or insufficiently cleared) wire that greeted the attacking battalions on 25 September are numerous and speak volumes about the failure of the preliminary bombardment. This was particularly noticeable on the front of 2^nd Division. All three brigade war diaries, as well as various other sources, mention uncut wire and unsuppressed enemy machine gun posts when they attacked on 25 September.¹¹⁹ In what was a disturbing comment on the failure of the bombardment in his sector, Lieutenant-Colonel H.C. Potter (CO 1/King’s, 6 Brigade) mistakenly thought ‘that no wire cutting had been done on the battalion front as reliance was placed entirely on the expected effect of our gas’.¹²⁰ 1^st Division was equally unsuccessful, testament to a lack of guns and an unfavourable sector of the front. Both attacking battalions of 1 Brigade complained about uncut wire,¹²¹ while 2 Brigade completely failed to clear away the enemy obstacles in its way. Faced with, as one officer put it, ‘one of the strongest and widest belts I saw during the war’,¹²² repeated attacks could not to reach the enemy line, which had been completely undamaged by the bombardment.¹²³

Poor staff work did not help. In particular, there does not seem to have been an effective transmission of information between those battalions in the front line and their supporting artillery batteries. For example, numerous patrols had warned Lieutenant-Colonel F.G.M. Rowley (CO 1/Middlesex, 19 Brigade) that the enemy wire in his sector was still intact and in places up to twenty-five yards thick, but nothing was done about it – with deadly consequences when the infantry tried to bypass it on 25 September.¹²⁴ Likewise, Lieutenant-Colonel A. C. Northey (CO 9/Scottish Rifles, 28 Brigade) complained that the machine gun posts at Strongpoint and Railway Redoubt had been practically undamaged by the bombardment. Although repeated efforts were made to communicate this to higher command, he did not ‘observe any alteration in the treatment of these points’.¹²⁵ This may well have reflected either a lack of ammunition or a belief that the gas would deal with these before the infantry went ‘over the top’, but Northey was not told why this was so.

For the bombardment to be effective, the infantry in the front line had to check the state of the German lines regularly and make sure that any repair work was kept to a minimum. But on some sectors this not carried out. Neither 28 nor 2 Brigade sent out patrols to check the state of the German wire before the attacks went in. This was an omission of crucial importance because of the difficulty of observing the state of the German wire in these sectors. If patrols had been sent out, it would have been obvious that a much heavier bombardment was needed.¹²⁶ It seems that in 28 Brigade, ‘the desire to save the men from being exposed to our own artillery fire’ contributed to the decision not to patrol,¹²⁷ but why 2 Brigade failed to do so is unclear. It is even more puzzling when considering that during August 1^st Division had been regularly sending patrols to the German wire, although a report had been filed on 3 August that ‘Lone Tree cannot be visited by day’.¹²⁸ There are perhaps similarities to what John Terraine observed about the preliminary bombardment for the Battle of the Somme the following year. It seems that the British tendency to ‘look on the bright side’ also occurred at Loos.¹²⁹ Captain J.C. Dunn even recorded how ‘I did not like to hurt the gunner’s feelings by saying how little sign there was of cut wire’.¹³⁰ Amazingly, this was not an isolated phenomenon.

Certain sectors of the front clearly had better artillery preparation than others, but results could still be mixed. The Hohenzollern Redoubt is a particular example. Because of its tactical importance it was heavily bombarded not only by 9^th Division’s field artillery and two extra batteries from 7^th Division, but also by much of the heavy artillery within I Corps. But it seems that only one side of the redoubt was properly shelled. The war diary of 26 Brigade, the unit that overran it on 25 September, noted that ‘Nearly all the German wire in the greater part of our front was efficiently cut by our artillery’.¹³¹ On the contrary, 28 Brigade, attacking from the northern side, was faced with ‘practically intact’ enemy wire, which completely stopped its attack.¹³² 7^th Division experienced a similar ‘mixed’ effect. While 22 Brigade reported that enemy wire was ‘very thick and had not been cut by our guns’,¹³³ the wire-cutting in 20 Brigade seems to have been ‘most effective’, with good gaps made.¹³⁴

The pre-battle problems with IV Corps’ artillery organisation do not seem to have been too debilitating. Indeed, the preliminary bombardment was most successful on the right of the British line. Both 15^th and 47^th Divisions managed to clear large amounts of enemy wire from the German front line. While both 44 and 46 Brigades reported that the German wire was ‘well cut’,¹³⁵ a report on 4 October confirmed that wire-cutting had been ‘excellent’ within 47^th Division.¹³⁶ Patrick MacGill, a stretcher-bearer with 141 Brigade, noted that the wire had been ‘cut to little pieces by our bombardment’.¹³⁷

The German trench [MacGill added] had suffered severely from our fire; parapets were blown in, and at places the trench was full to the level of the ground with sandbags and earth. Wreckage was strewn all over the place…¹³⁸

15^th Division had been equally successful. Its divisional history recorded how on ‘reaching the enemy trenches, the effect of the accurate and intense artillery fire was apparent. The front line was badly damaged – in fact, in some places was non-existent – and the communication trenches leading to Loos were filled with dead and dying men.’¹³⁹ When John Buchan visited the battlefield five days after the fighting had begun, he was amazed at what he found. The Loos Road Redoubt was now, he explained, ‘a monument to the power of our artillery. It is all ploughed up and mangled like a sand castle which a child has demolished in a fit of temper.’¹⁴⁰ But these successes remained the exception rather than the rule. As regards the long-range firing at the German second line, this seems to have been ineffective. According to 1^st Division, the machine gun posts and dugouts in the second line were ‘practically untouched’ and when 21 Brigade reached Cité Trench it found its wire obstacles were ‘undamaged’.¹⁴¹ This seems to have not been unusual. When the doomed attack of 24^th Division reached the German second line between Hulluch and Cité St Auguste on 26 September, numerous personal accounts and war diaries testify that its covering wire entanglements were completely undamaged and, as one observer put it,‘absolutely intact’.¹⁴²

How effective were the air operations conducted by the RFC? Its artillery spotting seems to have been generally successful, but considering the grave weaknesses inherent in the bombardment plan and within British artillery during this period, its effect was understandably limited. Teething problems with air-artillery co-operation were also in evidence. According to one observer:

The greatest difficulty was experienced at first in getting the gunners to conform to the best methods of ranging for air observations. They were a strongly individualistic corps and each battery commander had his own theories about working his guns and many at first intensely resented being interfered with.¹⁴³

There were also organisational difficulties. A memorandum issued on 25 October admitted that in future operations it would be necessary to clearly designate zones for each squadron to work in, ‘to prevent confusion caused by two machines trying to range on [the] same target’.¹⁴⁴ The effect of the aerial bombing campaign was also limited. Considering the fledgling nature of the air war, the damage that aerial bombing did to German communication networks and infrastructure was necessarily light, but it was considerably greater than had previously been achieved. Bombing flights went up to thirty-six miles behind the German lines and dropped over five and half tons of munitions. The main rail lines were apparently damaged in 16 different places, with 5 or 6 trains ‘practically wrecked’.¹⁴⁵ However, an RFC report concluded that although ‘considerable interference’ had been caused to the German rail network, ‘no great damage’ had been done.¹⁴⁶

The counter-battery operation was similarly ineffective. As Peter Chasseaud has explained, counter-battery results ‘before and during the battle were disappointing, and showed up the limitations of First Army artillery intelligence work, the artillery command structure and the gunnery itself’.¹⁴⁷ German artillery was undoubtedly quiet during the bombardment (Sixth Army only contained 325 field and 150 heavy guns), and this seems to have been a reflection of both poor German and British counter-battery work, as well as an understandable desire to avoid getting into an ‘artillery duel’.¹⁴⁸ Brigadier-General A.A. Montgomery believed that it ‘does not appear that the injury inflicted on the German artillery and personnel or material was very great’. He did, however, think that British batteries maintained ‘an undoubted superiority over the German artillery during the operations’.¹⁴⁹ But this was largely irrelevant. Lieutenant-Colonel W.R. Warren, serving as a brigade artillery commander in Indian Corps, noted the unrealistic optimism that many air observers and battery commanders had regarding artillery shoots.¹⁵⁰ He also voiced a familiar complaint that as German batteries tended to fall silent as soon as any shells neared their position, this was often taken ‘as evidence of destruction’, when in fact they were unscathed.¹⁵¹ Warren subsequently added that ‘all known batteries’ on his front came to life again on 25 September.

What of those ‘on the other side of the hill’? It is not known exactly how many German soldiers became casualties during the bombardment, but it was probably not a great number. An estimate by IV Corps reckoned that 117^th Division suffered only 120 casualties prior to 25 September.¹⁵² Because German defenders usually evacuated all but the most essential personnel from the first line when a bombardment was going on, they held their lines with only small numbers of men. The majority of its soldiers would therefore have been sheltering from the shells in deep dugouts or secure in the support lines. The regimental history of 157 Infantry Regiment confirmed that even after three days of bombardment the losses in the front-line battalion were ‘relatively small’ and the morale of the troops was ‘very good’.¹⁵³ So poor was the British bombardment that when Crown Prince Rupprecht (GOC Sixth Army) met his corps commanders on 22 September, he was unimpressed with British artillery preparation and refused to believe that an attack was imminent.¹⁵⁴

It is clear therefore that British artillery fire was simply not heavy enough. The Brigade Major of 26 Brigade complained that the bombardment ‘struck us all as quite childish and futile… no volume at all’.¹⁵⁵ This was a common concern and was echoed in Goodbye To All That (1929).¹⁵⁶ As Lieutenant Colonel Warren noted:

The whole programme appeared too deliberate, scattered and long drawn out for the amount of ammunition available. In many cases insufficient rounds were allotted to heavy guns for destructive tasks, particularly for counter-battery shoots.¹⁵⁷

This is confirmed by statistics taken from the papers of AA & QMG First Army. Table 3 shows that around five times more shells were fired during the preliminary bombardment at Loos that at Festubert, four months before.

Table 3. The number of rounds fired during the preliminary bombardments of the Battles of Festubert (13–15 May 1915) and Loos (21–24 September 1915)¹⁵⁸

	Festubert	Loos
Field guns	33,802	209,745
Field howitzers	5,102	32,554
Counter-battery	1,736	7,221
Heavy howitzers	3,307	16,958
TOTAL	43,947	266,478

But this is misleading. Table 4 puts these figures into sharp perspective. It shows the numbers of rounds fired per yard of front during the preliminary bombardments of both battles. Although at Loos there was a slightly higher amount of field gun, field howitzer and heavy howitzer shells fired, the differences between the two bombardments is hardly significant. The preliminary bombardment at Festubert had been a day shorter than at Loos and fired fewer shells, but the greater amount of front to be bombarded at Loos more than cancelled these factors out.

Table 4. The number of rounds expended per yard of front during the preliminary bombardments of the Batdes of Festubert (13–15 May 1915) and Loos (21–24 September 1915)¹⁵⁹

	Festubert	Loos
Field guns	11.2	13.1
Field howitzers	1.7	2
Counter-battery	0.57	.05
Heavy howitzers	1.1	1.8

From these statistics it is reasonable to assume that the results of the following infantry attacks would be similar. Festubert had been a success, with gains made and consolidated, but the attacks had been made at 11.30p.m. and 3.15a.m., at night and at dawn respectively. At Loos the British would be ‘jumping the bags’ hours later. This was the result of Haig’s ‘spreading’ of his guns. Enemy artillery strength may well have been spread over a wide area, but by doing so, Haig effectively neutralised his own artillery.