The Somme

  To drive this point home Haig sent his Chief of Staff, General Kiggell, to Fourth Army headquarters with the following memorandum:

  With reference to your [plan] dated 28th August, 1916, the Commander-in-Chief considers that the situation is likely to be favourable for an operation planned on bolder lines. Accordingly he desires that the ‘tanks’ may be used boldly and success pressed in order to demoralize the enemy and, if possible, capture his guns.12

  He added, in a blow that ended one aspect of Rawlinson's plan, that information from the tank experts indicated that the tanks could not be used at night.13 And just in case Rawlinson had not got the point of Kiggell's message and the accompanying note, Haig penned an even longer set of instructions. He emphasised yet again that enemy morale was deteriorating, that they had suffered repeated defeats and that their resources were limited to very tired troops who had already suffered severely. Bold action now would perhaps yield ‘decisive results’, a point which must be stressed to all commanders. The slow methods of trench warfare were inappropriate. The cavalry available to Fourth Army would be increased to four or five divisions and their final objective extended. Now they were to aim first at establishing a flank guard along the line Morval–Le Transloy–Bapaume, then proceed north-west to roll up the entire enemy line.14 So what all this meant was that in a matter of a few hours Haig's thoughts had advanced beyond the enemy gun line to Bapaume and points north-west.

  What underlay Haig's optimism was his view that German morale was collapsing. The recent successes were being regarded not as fortunate incidents in a generally unrewarding period of fighting, but as proof that the German army had entered a decline possibly approaching terminal collapse which could be accelerated by a large offensive employing novel weaponry. What is to be made of this? Of course it is unlikely that the morale of German units facing the British on 15 September was universally good. Some of the Bavarian regiments in the line had been involved in the severe fighting around Ginchy, Delville Wood, and High Wood during the first two weeks of September. Three in particular had suffered heavy casualties.15 On 10 September one of these regiments described the mood in the battalions as ‘depressed’.16 Yet the majority of the Bavarian regiments facing the British were fresh and their morale was good.17 British Intelligence may once more have made the error of generalising about morale from a small sample of German prisoners.

  But these considerations did not weigh with Haig. He required a strong reason to believe – despite the scanty evidence – that German morale was all but spent. For if the German guns were to be captured and the cavalry unleashed, it was essential that enemy morale be on the point of collapse. Only then could the employment of cavalry be justified. So for Haig to believe that a major British victory was anywhere in sight, he was bound to look about for indications that the prerequisite circumstances for a great triumph were now present.

  II

  In the face of Haig's barrage of memoranda and messages. Rawlinson capitulated. He drew up a new plan. Perhaps Rawlinson was simply overwhelmed by the vigour, frequency, and imperious nature of Haig's memoranda. Perhaps also Rawlinson's position had been weakened by his lack of grip on the battle in August, a performance that had eventually earned him a stern lecture from Haig on his duties as an army commander.

  Rawlinson's response was to call an immediate conference of his corps commanders and senior staff officers. He told them the Commander-in-Chief wanted to go ‘all out’ and push the cavalry through and to this end he was increasing the number of infantry divisions for the attack from six to nine. Nevertheless, Rawlinson was clearly unhappy. He told the corps commanders that ways would have to be found to push the artillery forward to enable it to bombard the third German line (a point Haig had ignored in all his memoranda) and then support the forces consolidating its capture. And when the corps commanders indicated that the roads behind the British front were in such bad repair that this might not be possible, he merely said that the roads ‘must’ be improved. On the matter of tanks he expressed a ‘reasonable hope’ that they ‘may’ prove of some help and repeated Haig's assertion that the enemy, which would be subjected to the new weapon, was tired and demoralised. General Horne, commander of XV Corps, initiated a discussion on the still unresolved matter of how the tanks were to be integrated with the artillery. To appreciate this discussion it is necessary to say something about the capabilities and limitations of the Mark I tank, lest they be mistaken for anything resembling the modern equivalent.

  The Mark I was a very cumbersome weapon. It weighed 27 to 28 tons depending on whether its armament was cannon or machine-guns. It was 8 feet high and 32 feet long. It could cross rough ground, crush barbed wire, and engage enemy trench-defenders. But its maximum speed over average going on the Western Front was just 2 miles per hour, which made it slower than a walking infantryman. It was also decidedly prone to mechanical failure and very vulnerable to artillery fire.18 Moeover, the crews could not function for more than a few hours because they were breathing in ‘noxious fumes [from the engine] at very high temperatures’.19

  In the light of these considerations, Rawlinson's arrangements for the employment of tanks are open to question. He intended to employ the tanks against the most formidable objectives in the German line while denying them the artillery support they needed. He stated that

  the ‘tanks’ should go forward with the shrapnel barrage, that they should be concentrated in groups in trench junctions and strong points, and that the shrapnel barrage should be put on between the groups of ‘tanks’.20

  What this meant was that gaps would be introduced in the creeping barrage, so that it would fall in the areas where infantry were advancing but not in the areas where tanks would be operating.

  The ‘problem’ Rawlinson was attempting to solve here was that if a continuous shrapnel barrage was fired along the whole front it might hit the tanks. Why this was thought to be a problem is unclear – the armour of a tank could withstand shrapnel. Nevertheless, Rawlinson was decreeing that lanes be left in the barrage and – since the tanks would proceed in groups – these would have to be at least 100 yards wide. As the tanks were intended to be used specifically against enemy strongpoints this meant that the artillery would at this stage not be employed in this phase of the battle precisely against the targets where it was most required. This arrangement may have seemed to Rawlinson to constitute a solution to the problem of protecting his tanks from his own artillery. But it did so only by causing a much larger problem. At best, the tanks would only manage to deal with the strongpoints when they got within close range of them. The strongpoints, by contrast – which would be spared the creeping barrage – could exact their toll on the attacking infantry from the moment the latter left their trenches. Further, it was even questionable whether the tanks would go ahead of the infantry. Given the derisory speed of which armoured vehicles were capable at this time, the infantry – unless they were prepared to dawdle across no man's land and endure all the resultant losses – were likely to precede the tanks in their progress towards the strongpoints. But they would do so in the absence of a creeping barrage.

  So the appearance of the tank on 15 September did not provide a large body of British infantrymen with an additional form of protection and support. Ironically, it denied them the established sustenance of the creeping barrage and replaced it with a vulnerable substitute of doubtful efficacy.

  III

  The purpose of the conference between Rawlinson and his corps commanders was not only to decide on arrangements for the tank but to produce a revised plan for the Commander-in-Chief. After the conference Rawlinson informed Haig that he now understood that given the ‘undoubted’ deterioration in enemy morale his first plan had not been ‘sufficiently ambitious’. He now also realised the need to act boldly and to take every advantage of panic. He had therefore increased the infantry divisions from six to nine to give the attack the depth and weight needed to capture the third German line. As for the increased
force of cavalry, the gap for them would probably occur between Lesboeufs and Gueudecourt and they would proceed from there to Bapaume. The tank lanes had also undergone some modification. Because of the vagaries in the flight of shells, it was realised that the lanes could not be maintained beyond the German second line. Once that point had been reached the creeping barrage would be reimposed across the entire front. So for the advance on the third German position the infantry and tanks would both follow the curtain of shells.

  So far all of this was in line with Haig's wishes. But even now Rawlinson was not prepared to go further. There was no mention of what the cavalry might do after it had reached Bapaume. And he concluded his plan by emphasising to Haig the difficulty of constructing roads for the artillery and of supplying the large breakthrough force with food and ammunition.21

  Rawlinson had now devised a plan in which the tank, along some sections of the front, would be substituted for the creeping barrage. This decision seemed to indicate that he saw the armoured vehicles as possessing the equivalence of his most powerful weapon – the artillery. Yet at the same time he also seemed to be having doubts about the tank. He saw them perform again on 2 September and again ‘was not pleased with them’. He still considered the tank crews ‘green’ and ‘raw’ and felt that they did ‘not understand fighting the Bosche’. In addition he thought the CO (Colonel Brough) was ‘no sort of good’.22 He passed his complaints on to GHQ but clearly was not reassured. A few days later, in conversation with the Prime Minister (who was visiting the Western Front), he stated that he was placing no ‘great reliance’ on the tanks.23 In other words he was failing to notice that his plan, by expecting the tank to capture strongpoints in the absence of a creeping barrage, was placing every reliance on them.

  While Rawlinson's enthusiasm for tanks was waning, Haig's enthusiasm for the operation as a whole was increasing exponentially. On the 7th he told Rawlinson that he had informed the commanders of the Third and First Armies (Allenby and Monro) of the impending attack. He wanted them both to develop operations to seize points of tactical importance on their fronts (identified as Gommecourt and Monchy for the Third Army and Vimy Ridge for the First). These operations were to be launched by Allenby's forces on 20 September and Monro's on the 30th, so coinciding with the advance of the Fourth and Reserve Armies from the south.24 Now the Third Army's front ran from the north of Gough's forces around Gommecourt to Arras, and the First Army northwards from that point to Aubers Ridge – a distance of some 60 miles as the crow flies or 90 to 100 miles as the cavalry trots. In other words Haig's ambitions for the new attack were now virtually unlimited. He was envisaging nothing less than the collapse of the entire German army facing the BEF. The disintegration of the Germans opposite the northernmost British army, the Second around Ypres, would presumably have been a bagatelle if all else succeeded. Indeed, even the Second Army was given a role to play. On 4 September Haig had broached a plan to Rawlinson to distract the enemy's attention from his main attack by threatening to land a force from Plumer's Second Army on the Belgian coast behind German lines.25 Discussions with Admiral Bacon of the Dover Patrol and General Plumer followed.26 By the 13th a brigade of infantry from 4 Division was in place at Nieuport and asked to display ‘liveliness’ on the day of the great attack.27 Haig anticipated that if all went according to plan the Germans might abandon the Belgian coast, whereupon the threat of a landing would be converted into a reality.28

  In the event none of these expectations regarding the northern sector were fulfilled. Perhaps not surprisingly, the appearance of a few thousand men at Nieuport, however ‘lively’ their disposition, did not distract the Germans' attention from an impending attack by 100,000 troops on the Somme.

  It is important to reflect on the import of all this. Haig's conception was nothing less than the most grandiose vision for victory developed by any commander since Schlieffen. Ludendorff's subsequent conception for March 1918 can be compared to it, but even that lacks the audacity of Haig's operation. For to carry his plan through, Haig (unlike Ludendorff) had not assembled the largest number of guns seen on the Western Front. He had assembled the largest number of horses. Sharper eyes than ours have detected a ‘learning curve’ in the performance of the Commander-in-Chief during the Somme battle. Such an argument is scarcely applicable in this instance.

  While Haig was inhabiting his own world of collapsing enemy morale and decisive victories, Rawlinson and his corps commanders were grappling with less fanciful matters such as how to capture the front line of an army whose morale was not conspicuously disintegrating.

  A large difficulty presented itself: how to combine a new weapon, the tank, and an old weapon, the cavalry, with the main weapon that could ensure success, the artillery.

  The first issue was to ensure that the tanks were on the start line at zero (6.20 a.m.) so that they could move forward as the preliminary bombardment reached its crescendo. It was decided to accomplish this by bringing the tanks forward under the cover of darkness. This would also ensure concealment. The main difficulty lay in the area of III Corps where the front line ran through High Wood, a treacherous enough place for the transit of men let alone for tanks at night. Should the tanks in this area begin their forward move earlier? Surprisingly the corps commander, General Pulteney thought not. He stated: ‘The tanks will go quickly through High Wood because they will have cover all the way.’29

  It is difficult to know what Pulteney had in mind here. There was no ‘cover’ in High Wood. By September there was only a scattering of trees standing and these certainly had no leaves attached. What High Wood did consist of in September 1916 was a tangle of tree stumps, fallen branches, trenches, dug-outs, and belts of rusting wire. It was doubtful whether tanks could even traverse this landscape let alone ‘go quickly’ through it. If the opportunity presented by tanks on this part of the front was not to be cast aside, it was desperately necessary that Pulteney's views be corrected. Colonel Ellis, a tank expert at GHQ, attempted to do this, but he dissipated his whole message by telling Pulteney that tanks could not get through Trones Wood, a position that had fallen two months previously. Yet more remarkably, he did not mention High Wood at all. The end result of this dialogue of the deaf was that Rawlinson decided it to be a ‘matter of consideration’ and left the decision to Pulteney. The results would be quite disastrous.

  The discussion then turned to the matter of the formations to be used by the groups of tanks. Rawlinson wanted them deployed in groups of four in diamond formation. Horne, in an isolated burst of insight, pointed out that this would mask the inward guns of the two centre tanks. Rawlinson made the bizarre reply that it would not matter if they shot into each other, providing these tanks were only equipped with machine-guns! In this instance common sense eventually prevailed – most tanks would be used in groups of three in line ahead, thus giving them the ability to use all their guns against the enemy.

  The next item covered was how the creeping barrage might affect the tanks. Rawlinson, it will be recalled, had produced an uneasy compromise whereby a creeping barrage was to be fired but not in the areas where the tanks were bunched. At this conference General Horne (XV Corps), who was an artilleryman, put his finger on the weakness of this plan. It will, he told Rawlinson, ‘leave enormous unprotected places in the line’. Horne was making sense here, but he then went on to say that it would be better to have standing barrages on each of the enemy trench lines towards which the tanks and the infantry could work. This of course was the type of plan which had failed so disastrously in the north on 1 July. The barrages then missed the German trench lines and thus gave the German machine-gunners complete freedom to mow down the attacking infantry. Anyway by September it was clear that many German machine-gunners were located in shell holes well away from the main trench defences. So Horne was substituting for a scheme which gave only partial protection to the infantry one which, unless the guns fired with pinpoint accuracy, would give them no protection at all. Rawlinson recognised
this. He told Horne that he ‘was very reluctant to give up the creeping barrage. I would prefer to maintain it and have a hole for the “tanks”.’ Horne replied that he was aware that a creeping barrage was useful to clear out men ‘lying about in shell holes’ but apart from that he ‘could never follow what is the value of a creeping barrage’.

  This exchange is extraordinary. Horne's characterisation of German machine-gunners situated outside the trench lines, who would present a deadly threat to his advancing infantry, as ‘men lying about in shell holes’ appears to show little appreciation of the reality of battle. Even more puzzling is his statement that he could not make sense of the creeping barrage, the surest method of artillery protection yet devised and one which, in the previous two months, had clearly established its effectiveness. Indeed the first group to fire a creeping barrage was Horne's own XV Corps artillery on 1 July. It had achieved decidedly favourable results on an occasion where instances of success were conspicuously absent

  Rawlinson, for his part, was not noticing that the gaps that he was instituting in the barrage for the sake of the tanks were opposite strongpoints from which would emanate the most devastating German defensive fire. In the end, and in one sense not surprisingly given the incoherence of Horne's arguments, Rawlinson's plan remained. But what is remarkable about this is that no one questioned the wisdom of withholding the creeping barrage from areas in front of the tanks. For one thing the hulls even of the Mark I tank could well withstand a shrapnel barrage. For another, as long as the infantry were required, for good reason, to place their un-armoured bodies at hazard by clinging close to the advancing curtain of shrapnel it made no sense to deprive them of this protection. The arrival of the tank in this instance was causing confusion where none had existed before.