The Somme

  While futile, penny-packet, but costly attacks of the type described were being conducted on the left flank, a larger operation to capture Guillemont was being planned for the right. Even then this operation was not to be that large. In all, just four battalions from 55 Division would attack the village direct while two battalions from 2 Division would advance on it from the north. There would be no flanking support from neighbouring divisions. To the north the entire XV Corps would be quiescent and to the south, for reasons that are incomprehensible, the French would advance but on the Hem Plateau near the River Somme, not in an area contiguous with the British.6

  In the event, Rawlinson had to postpone his attack until 8 August. As was becoming customary, the French went ahead on the 7th anyway, to very little effect. This now meant that on their right flank the British would not even have the support of the distant French on the Hem Plateau.

  The attack on the 8th was almost a total failure, although a small amount of ground south of Guillemont was gained by 55 Division. Explanations for the failure are not wanting. First there was the unimaginative nature of the plan – zero hour was at exactly the same time (4.20 a.m.) as all previous attacks on the village. Second, the heavy guns lifted from the German front line 15 minutes before zero, thus giving the defenders ample warning of the impending attack and sufficient time to bring their weapons to the parapet in complete safety. Third, the jumping barrage fired by the field artillery gave the troops no protection. Fourth, the British had still not grasped that the main defences of Guillemont were not the trench lines in front of the village, which were either lightly manned or evacuated completely before an attack, but the flanking fire from defenders in shell holes and occupants of the cellars in the village itself. Once again therefore, the troops swept through the village only to be cut down by fire from the flanks and rear.7

  Finally, as always, there were local factors. In some areas smoke and haze caused the troops to lose direction; in others the artillery had not cut any of the belts of wire in front of the German line.8

  The ordeal of the troops facing Guillemont was not yet over. Haig was convinced that some British troops were holding out in the ruins of the village and that insufficient effort was being made by Rawlinson or the divisional command to support them. He therefore demanded that the attack be renewed on the 9th.9

  So once again the hapless troops, supported by even less artillery than the inadequate amount that had failed them on the previous day, were committed to a repetition of the attack on the 8th. The result is best described in the War Diary of one of the participating units:

  3am Company officers warned of ... changes [to] scheme of attack. No time to go into detail, or for Coy Comndrs to explain to Coy officers and N.C.O's as attack was timed for 4.20 a.m. Owing to all N.C.O's and Btn and Coy runners having no idea of the country, and the sunken road being crowded with men from [a neighbouring battalion], great delay was caused in getting the Coys out in position.

  3.55 North Lancs just arriving and officer in command said he had only just heard they were going to attack.

  4.15 Our guns opened a heavy Barrage and the Germans dropped their Barrage [i.e. retaliated] within 3 minutes. Their back barrage line being on sunken road. Companies not all in position had to move thro barrage. Those in position moved forward behind assaulting waves but whole line was unable to make headway owing to machine gun fire and heavy barrage. Whole line fell back and manned our original front line, Coys hopelessly intermixed with men of other attacking Battalions.10

  Not surprisingly, soon after this fiasco, the battalion was withdrawn. It had suffered 138 casualties and had hardly taken part in anything that could be dignified as an attack.11

  It was clear to some close to the action why these attacks continually failed. As one observer later remarked:

  Their [that is the German] guns were located on the flanks away from the trench systems, and consequently in positions impossible to locate and destroy before an attack.12

  Rawlinson took a different view – he blamed the troops. In his diary on the 8th he remarked that ‘failure was mainly due to the want of go and inferior training of the infantry’.13 Next day he was still blaming them, noting: ‘I don't think the Infantry were for it’.14

  For once, Haig disagreed. He was becoming convinced that the responsibility for failure did not lie with the foot soldiers but with the plan. And he was not hesitant in blaming Rawlinson for slipshod planning.

  After the failure on the 9th he dispatched Kiggell and another staff officer, General Davidson to Fourth Army headquarters with a stern message:

  repeated failures to capture Guillemont have convinced the Commander-in-Chief that the method of attack adopted requires careful and full reconsideration.15

  Haig insisted that Rawlinson should do two things. He should discuss the problem of capturing Guillemont with officers – down to brigade level if necessary – who had already been involved in attacking it. And he should then, after careful consideration, himself draw up a plan for the next attack. Haig even gave Rawlinson a strong indication of the form the operation should take. It should consist of an attack ‘on a very wide front’, in conjunction with the French from the River Somme on the right to High Wood on the left, ‘so as to prevent the enemy from placing a barrage which might stop us if we attacked on a small front’.16

  After this démarche Rawlinson set about preparing to attack Guillemont in some force. He determined to attack immediately, which seemed contrary to Haig's statement that the plan should be given careful consideration but was in line with the general intention of the Commander-in-Chief as outlined in his memorandum of 2 August.

  As it happened, no operations took place for ten days. The cause was force of circumstances rather than Rawlinson imposing a delay in order to develop a sensible plan. On 10 August Congreve, the Corps Commander in the Guillemont area, became seriously ill.17 Haig ordered that he be replaced by General Cavan, the commander of XIV Corps, which was at the time operating on the front of the Reserve Army. Cavan too then became ill and his recuperation and the need to familiarise himself with the situation delayed the operation until 18 August.18

  In the interval, of course, operations did not entirely cease. On the left one battalion of 34 Division attacked the Switch Line on the 11th. Next day three battalions from 15 Division assaulted the same line. All failed. Also on the 12th one battalion from 55 Division tried unsuccessfully to gain ground south of Guillemont. Then the 16th and the 17th saw more attacks on the Switch Line and more attempts to advance closer to Guillemont. These last must have had the effect of alerting the Germans to the impending large-scale assault on the village but no one in authority seems to have considered this point.

  Though none of these operations had a scintilla of success it would be wrong to leave the impression that all attacks were poorly planned and executed. On 12 August, a batallion of Cameron Highlanders from 15 Division assaulted the Switch Line from a position called the Elbow. The artillery support was excellent, ‘a wall of flame’ (according to the battalion report), and the men advanced close under it and captured 300 yards of their objective.19

  More characteristic however, was the attempt of the East Surreys to capture a stongpoint south-west of Guillemont. One quarter of the battalion became casualties and precisely nothing was gained. The East Surreys’ War Diary lists the following reasons for failure:

  (1) the bombardment of the strongpoint by 9.2" howitzers did not take place.

  (2) the barrage plan was too complicated, the barrage being fired at an angle to the front line.

  (3) as a consequence of (2) the barrage commenced so distant from the British line that most of the men were not even aware that it had been fired.

  (4) the infantry plan was too rushed and resulted in the units commencing at different times.

  (5) machine guns to the south of the operation were left unattacked.20

  Rawlinson's large-scale attack was scheduled to commence on the 18th. But
even as the date was being decided, the plan for the vital right flank began to unravel. Fayolle, the commander of the French VI Army, decided unilaterally that his troops were too exhausted to launch a major offensive. He therefore reduced his participation in the battle to just one division. As a consequence Rawlinson considered that Guillemont was too tough a proposition to capture in one bound. He ordered XIV Corps merely to gain ground towards the village on the 18th and to effect its capture the next day.21

  So the British, deprived of significant French support on the right, would attack the Germans with five divisions along a front from Intermediate Trench on the left to just south of Guillemont on the right, a distance of 12,000 yards. This was a much more extensive front of attack than on 14 July and some comparisons between the two battles are instructive. On the previous occasion four British divisions attacked on a front of 6,000 yards – i.e. 1,500 yards per division. Now five British divisions were attacking on a 12,000-yard front, a ratio of 2,400 yards per division, a 60 per cent increase from that of 14 July. More significantly, artillery devoted to the wider front had not risen commensurately – in fact it had diminished. On 14 July the 6,000 yards of front had been subjected to a bombardment of 491,804 shells; for 18 August the figure was 396,912.22 In other words, while the width of front had more than doubled, the number of shells to be fired at it had decreased by 20 per cent.

  In fact the situation was even less favourable than these figures reveal. On 14 July all the shells had been directed against the German front system, owing to Rawlinson's erroneous belief that there were no formidable German defences behind the front, and almost all of this front lay within clear view of the British gunners. None of this applied on 18 August. As the British command was finally aware, German defensive arrangements around Guillemont had transmuted from a series of lines into a chequerboard arrangement situated in shell holes. This required many more shells to subdue. Moreover, from Delville Wood through High Wood to Intermediate Trench, the German positions were located on a reverse slope, also requiring a greater bombardment because of the need for aerial spotters continually to adjust the range of the guns. In every way, therefore, the British bombardment before the attack on the 18th was woefully inadequate. And Rawlinson, had he reflected for a moment on the figures he produced for his conference, should have been well aware of this.

  Command failures did not cease with the artillery arrangements. The divisions chosen to conduct the attack (1, 33, 14, 24, 3) were in some ways a decidedly peculiar choice. In the case of 1 and 3 Divisions, the fact that they had been in the line for so long made their choice questionable. In addition 3 Division faced the most unenviable task. It had been selected to capture Guillemont. Yet in its period at the front it had consumed half of its infantry strength (6,000 men), and had received as reinforcements just 3,000.23 What this amounted to is described by the commander of the division:

  All battalions were short of officers, many of them very short, whilst the reinforcements both of officers and men which had arrived had received but little training, and were quite inexperienced in war.24

  The broad-front attack went in at 2.45 p.m. on 18 August. The resulting gains varied from the non-existent or derisory on the left to extremely modest on the right.

  On the extreme left, the artillery had in the main missed Intermediate Trench, with the result that the left of 1 Division was driven back with heavy casualties. On the right of the divisional front near High Wood a more accurate bombardment and a creeping barrage enabled them to occupy a few hundred yards of the enemy front line.25

  On its right, the attack of 33 Division in High Wood was a complete fiasco. In this area the British had tried to innovate. They had brought forward two large flame-throwers. In addition pipes full of explosive were to be pushed under the German front line and detonated. And because the trench lines in the wood were so close together the forward troops were withdrawn and the wood subjected to a bombardment by the heavy artillery for 48 hours before the attack.26

  Nothing went according to plan. The heavy artillery comprehensively missed the German positions but succeeded in eliminating its own flame-throwers in the first few minutes of the bombardment. Then, the ‘push-pipes’ failed to explode. Moreover, the troops who had been withdrawn from the front line so as to facilitate the bombardment were moved back into it ahead of zero hour, which meant that no creeping barrage could be fired because once more they were too close to the German line to permit it. At zero these troops emerged from their trenches to find that not even the wire protecting the German front trench had been cut. Consequently they gained no ground and suffered heavy casualties. Outside the wood affairs went no better. Troops were shot down in numbers by fire from strongpoints in the north-eastern corner of the wood which had also escaped the attention of the British artillery.27

  The 14 Division immediately to the right also fell victim to the fire from the German strongpoint in the corner of the wood. Further right, however, beyond the range of the German guns in that strongpoint, British troops following closely behind an accurately fired creeping barrage succeeded in gaining ground on either side of Delville Wood. These gains made the wood vulnerable to converging attacks.28

  The most important gains of the day were made by 24 Division to the north and east of Guillemont in an attack which apparently caught the Germans unawares. It advanced the British line some 500 yards. In addition, all units speak of the accuracy of the creeping barrage. Nevertheless, the division sustained heavy casualties, probably from machine-guns in Ginchy and Guillemont which were beyond the range of the creeping barrage and untouched by the feeble preliminary bombardment.29

  Failure was most comprehensive on the front of 3 Division, the unit directly in front of Guillemont. Some ground was gained by the troops well south of the village and therefore furthest away from the flanking machine-gunners. Further north, the battalions had not gone 20 yards when they were met with heavy machine-gun fire and an artillery barrage. The machine-gunners were located in cellars in the ruined village and in Wedge Wood and Falfemont Farm, beyond the range of the creeping barrage and left untouched by the bombardment. An attempt to bombard some of the German positions with trench mortars failed because the mortar crews consisted of raw recruits from Britain. At the end of the day the troops had gained no ground at all.30

  The attack on the 18th was therefore a failure except where ground was gained around Delville Wood and to the north and south of Guillemont. The anchors of the German defence, the Switch Line, High Wood, Delville Wood, and Guillemont, remained in enemy hands.

  There was no question of resuming the offensive on the 19th. The 3 Division was shattered, destroyed for the moment as a fighting unit. It would not reappear at the Somme until the very last days of the battle in November. Haig therefore arranged for the division to be relieved and for Guillemont to be attacked on the 21st by the successful 24 Division. After that, there would be another attempt at a joint operation with the French on the 24th.31

  These operations were largely unsuccessful. They were made in insufficient strength and with inadequate artillery support against a German defence that, although shaken by the experiences of the 18th, was still intact and thoroughly alert. Guillemont remained in German hands.

  For Haig this seemed to be the last straw. He was expecting at any moment the delivery of a new weapon in the form of the tank and he was determined to use it in mid-September in a large-scale operation to break the enemy line. But to give this operation every chance of success a reasonable start line had to be established, and this meant eliminating the awkward right angle by capturing such obstacles as High Wood, Ginchy, and, above all, Guillemont.32 Rawlinson, in his nibbling, poorly thought out attacks, showed no sign of being able to achieve Haig's desired start line. So on 24 August Haig penned him his strongest note yet. Regarding recent operations he told Rawlinson that ‘The only conclusion that can be drawn from the repeated attacks on Guillemont is that something is wanting in the methods employe
d.’33 He then went on to make crystal clear what this ‘something’ was. Rawlinson's attacks had been delivered on excessively narrow frontages, had employed insufficient forces, and had lacked the kind of oversight Haig had already suggested that Rawlinson employ. For the next endeavour, Haig wrote:

  The attack must be a general one, engaging the enemy simultaneously along the whole front to be captured, and a sufficient force must be employed, in proper proportion to the extent of front, to beat down all opposition.34

  He suggested to Rawlinson that two and a half divisions be used to attack Guillemont, a force sufficient to deliver a continuous attack along the whole front.

  But Haig's attempt to educate his army commander did not stop there. He reminded Rawlinson that his directive of 9 August had not been sufficiently taken to heart and proceeded to spell out in detail what was required.

  In actual execution of plans, when control by higher Commanders is impossible, subordinates on the spot must act on their own initiative, and they must be trained to [do] so. [However] in preparation for battle, close supervision by higher Commanders is not only possible, but is their duty, to such extent as they find necessary to ensure that everything is done that can be done to ensure success. [This did not constitute ‘interference’] but [is] a legitimate and necessary exercise of the functions of a Commander on whom ultimate responsibility for success or failure lies .... It appears to the Commander-in-Chief that some misconception exists in the Army as to the object and the limitations of the principle of the initiative of subordinates, and it is essential that this misconception should be corrected at once, where it does exist.35

  So far, in this boys’ own guide on how to command an army, Haig had been making good sense. But almost characteristically, he then put all this useful instruction at risk by seeking to deny Rawlinson sufficient time to implement the necessary changes. ‘Not a moment must be lost’, he told the Fourth Army commander, in carrying out the new ‘general’ attack.36 This was an invitation to disaster. Undue haste was simply incompatible with the changes in Fourth Army practice desired by Haig.