by Robin Prior
A further matter discussed at this conference was the difficulty of combining the cavalry with the artillery. After the German second line had fallen it would be necessary to move many of the guns forward to bombard the third line. However, if the cavalry were to have their maximum shock value, that would be precisely the time when they should begin their forward move from behind the British front. The two arms might therefore impede each other. And, even more alarmingly, the cavalry might mask the guns at the very time they were getting into position to support the infantry advance on the third line. The result would be that either the infantry would be denied artillery protection or many of the horsed soldiers would fall victim to their own guns. There was a further problem. If the cavalry did get through, all the available roads would be needed to supply them with food and ammunition, leaving most infantry divisions disastrously short of those vital commodities. It must be recorded that after a desultory discussion no solutions were provided to these problems.
It may seem incredible that such vital matters were left unresolved. But there might be a clue in Rawlinson's opening remarks at the conference. He noted that the first major objectives mentioned by the Commander-in-Chief lay around Bapaume. But he went on to say that ‘of course it will take us some time to arrive at that’. He then added:
We have got to break the line here first in order to put that proposed operation into effect. If we are successful, as we hope we may be ... it is hoped that there will be an opportunity of pushing the cavalry through.30
He concluded by saying: ‘I have just given you ... very broadly what the general trend of the operations will be in the event of our attaining a really decisive success. I will now turn to the operations of [the first day] itself.’
It seems reasonable to conclude that Rawlinson was not much alarmed at the difficulties that the cavalry operations would present to the infantry and artillery because he was not expecting them to eventuate. There are too many caveats in his opening statements (‘proposed operation’, ‘it is hoped that’, ‘it will take us some time to arrive at that’) to suggest any great faith in Haig's scheme. As had been the case for 1 July there is a strong suggestion that he was going through the motions of preparing for one kind of operation while actually preparing for another.
There were two artillery matters that remained to be considered. One was the perennial problem of cutting the wire and damaging the defences that could not be observed, the other was the weight of bombardment to be fired at the German defences.
The first problem particularly applied on the front of XIV Corps. There the Germans had constructed a defensive work called the Quadrilateral, which was supposed to have fallen into British hands by the time the offensive opened but had not. Neither this work nor the distant wire in the Combles Valley could be observed from any point behind the British front. The issue of bombarding unobservable defences was not new – it had arisen on many sections of the German front before 1 July. Rawlinson's solution was not new either. XIV Corps would have to rely on aerial observation to inform them if these defences were being destroyed. This had a chance of working in the case of the Quadrilateral – a large work not far distant from the British front line. Experience suggested that it had no chance of working in the case of a small objective such as wire. Time and again aerial photography had proved inconclusive as to whether belts of wire had been destroyed or merely rearranged by the bombardment. Yet in truth there was little more that could be done. The extension of the objective had left XIV Corps with a problem of observation which at this stage of the war was intractable.
The second artillery matter was conspicuous by its absence in any of the planning documents for 15 September. There was no discussion about the appropriate weight of bombardment needed to crush the enemy defences. Once more Rawlinson appears simply to have assembled as many guns and shells as possible and hoped that these would suffice.
This omission was not to prove as catastrophic as it had been on 1 July. It so happened that on 15 September the extension of the objectives by Haig to the German third line and beyond had little consequence for the artillery, because the third line was beyond the range of most British guns. Almost all of the 828,000 shells fired during the preliminary bombardment were therefore directed towards the first and second German trench systems.31 This gave a weight of bombardment of twice that of 1 July (though half that of 14 July).32 Nevertheless, the lack of attention to the bombardment was a matter of the greatest moment. On one section of the front (XIV Corps) the British did not have direct observation over the German trenches, a disadvantage not suffered on 14 July. Now, with a lesser ability to observe whether the shells were finding their target, they settled on a bombardment of half the intensity of that day. It must be presumed that by this period of the battle all commanders recognised the vital role of artillery in an attack. Yet they were prepared to send their men over the top without the sure knowledge that artillery calculations had been provided, and that their troops were adequately protected from both the distant enemy guns and machine gunners in the immediate defences which faced them.
A last-minute change was made to the 15 September plan by Haig. It involved an extension of the front of attack to include the capture of Martinpuich by the left of the Fourth Army and an attack on Courcelette by the 2 and 3 Canadian Divisions of the Reserve Army.33 Haig's reasoning was in line with his increasing sense of euphoria and his conviction that a large cavalry sweep was in the offing. He wanted to eliminate these two fortified villages quickly, thereby opening the way for the Reserve Army promptly to advance northwards on the left flank of the Fourth. If this were accomplished, the German forces facing Allenby's Third Army would find their communications cut and might surrender without even being attacked. As a consequence, a similar situation would confront enemy forces ranged against the First Army and a similar result eventuate.34 Rawlinson tried to resist this further commitment. He pleaded a shortage of troops. But Haig insisted and the Fourth Army Commander was forced to give way.35
In the days immediately preceding the battle Haig sent a series of exhortations to Rawlinson and to Gough not to neglect any opportunity to send the cavalry through.36 It seems clear that Haig had begun to sense something of Rawlinson's scepticism about cavalry action. Following a report by Davidson on a visit to Fourth Army headquarters on the 10th, Haig again impressed on Rawlinson the need for ‘bold action’ to ‘derive full value from the element of surprise, which, after all is fleeting! Moreover, the season for fighting is nearly over.’37 He repeated essentially the same message to Rawlinson on the 11th and on the 14th – clear evidence of how seriously he entertained the grander aspects of his project. Rawlinson duly passed on these instructions to his corps commanders, but without great conviction – for he invariably added the caveat that the cavalry must not impede infantry or artillery operations and he emphasised how difficult cavalry deployment on a large scale would be.
What then was the final plan for Haig's war-winning operation? The infantry attack with accompanying tanks would be conducted by 11 divisions attacking on a 12,000-yard front distributed from left to right as follows:
Contingent on the success of these forces, the remaining British armies on the Western Front would come into play.
Four objectives were given to the infantry. The first was the German front line. The second (which concerned just III and XV Corps) comprised the immediate defences covering the village of Flers. The third involved all three corps from Fourth Army, and encompassed Flers and the ground to the east and west of it. The final infantry objective for the first phase was the capture of Gueudecourt, Lesboeufs, and Morval by XV and XIV Corps in order to establish a defensive flank for the cavalry sweep. The total advance to be achieved by the attacking force was 5,000 yards on the right and 4,000 on the left. The troops were expected to advance in four jumps spread over 4.5 hours. As the battle would commence at 6.20 a.m., the cavalry advance would therefore commence shortly after 10.50 am, its objective the high
ground around Rocquigny–Villers-au-Flos–Riencourt–Bapaume.
As for the tanks, they would generally operate in groups of three in line ahead. They would start at zero and aim to reach the first objective five minutes ahead of the infantry. Gaps 100 yards wide would be left in the creeping barrage along the route taken by the tanks. After consolidation of the first objectives the tanks followed by infantry would proceed to attack the second objective. Tank lanes would remain in the barrage up to this point, but then the barrage would extend along the whole front. The tanks were then to proceed to the third objective with the infantry still in the immediate rear. Here their participation would, for the moment, cease, and those still operational would be withdrawn to replenish fuel and ammunition. The fourth objective would be assaulted by the infantry alone, and only after the guns had been moved forward in order to bombard it.
21 Lumbering Tanks: The Battle of 15 September
I
The preliminary bombardment for the new attack opened on 12 September. The weather, although fine in the morning, deteriorated later in the day. Showers persisted through most of the 13th, greatly hampering the aircraft which were attempting to spot for the artillery. On the 14th the weather cleared and a full day's shelling was possible but it is certain that as a result of the variable weather the bombardment was adversely affected, particularly on the front of XIV Corps.1
According to plan all corps left lanes unbombarded of at least 100 yards in width for their tanks. An airman far above the battle recorded the scene:
When we climbed up to the lines, we found the whole front seemingly covered with a layer of dirty cotton-wool – the smoking shell-bursts. Across this were dark lanes, drawn as it might be by a child's stubby finger in the dirty snow. Here no shells were falling. Through these lanes lumbered the tanks.2
For this airman, the entry of the tank into battle represented a significant accretion of power. But for the troops on the ground the ‘dark lanes’ where ‘no shells were falling’ had in many instances an altogether different significance.
A good example of the negative impact is made evident by looking at the most rightward British division, the 56. This unit of XIV Corps had two tasks. First, it was to move forward on its right and establish a defensive flank facing Combles while French forces were carrying out a similar movement on the other side of the village. Converging attacks would then be arranged to capture it. Second, at the same time the 56 was to advance on its left to screen troops from the other two divisions of the corps as they advanced to capture the crucial villages of Morval, Lesboeufs, and Gueudecourt. The fall of these three villages would be the signal to unleash the cavalry, and was therefore vital if Haig's larger purpose was to come to pass.
Tanks in Action, 15 September
The 56 Division comprehensively failed in its tasks. Of its three tanks, two broke down before reaching the British line. Thus there were no ‘lumbering’ tanks in this area, just German machine-gunners free from bombardment and with unimpeded opportunities to destroy the attacking infantry. The troops in the vicinity of these lanes were mown down, some assaulting battalions losing 90 per cent of their officers and 75 per cent of their other ranks.3 The tank that did get through assisted a small advance around Leuze Wood.4 That was all. As a result no solid defensive flank faced Combles and there were no troops to screen the advance on the distant villages. As it happened this was not of consequence, because no British troops made an appearance on this flank. Something had obviously gone wrong with the attack of the remaining divisions of XIV Corps.
Those divisions were the 6 and the Guards. In 1914 they had constituted part of the original BEF. Although much leavened by volunteers, this was perhaps the reason they had been chosen by Haig to secure the cavalry breakthrough. What soon became obvious was that a notable history was no substitute for a decent fire plan.
The problems of the 6 Division, although they were not to know it, had commenced long before zero hour. It will be recalled that this division faced a formidable German defensive work, the Quadrilateral. The task of destroying this work had been given to the heavy artillery, followed up by three tanks which would precede the infantry to subdue any defenders that had escaped the bombardment.5 As it happened, during three days of shelling the heavy guns hardly touched the Quadrilateral. Partly this can be accounted for by the weather, which made observed shooting difficult. It seems that in addition, however, the gunners had located the Quadrilateral in the wrong area – on the reverse slope of the ridge instead of its summit and forward slope. An aerial photograph taken just before the attack reveals that the main defences were intact.6
What of the three tanks? Two broke down well behind the British line and took no part in the battle. The third reached the Quadrilateral and poured fire into the German defences. Next bullets striking the outside of the hull caused slivers of red hot metal to sheer off and ricochet around the interior, wounding some of the crew. Then bullets actually began to penetrate the hull. The Bavarians opposite 6 Division were using armour-piercing ammunition brought up to deal with the steel shields used by British snipers. That decided matters. The tank commander withdrew. Remarkably, he made it back to the British lines. But tanks would play no further role that day on the front of 6 Division.
It was now time for the men of the division to leave their trenches. Not only did they find the Quadrilateral intact and with no tanks present but the creeping barrage by arrangement was also nowhere to be seen. The command apparently had assumed that the preliminary bombardment would have done its job and that in any case the tanks would have ‘flabbergasted’ what German defenders remained. That is: assumptions about a bombardment which could not be observed and a weapon untried in war inspired the command to deny its infantry its sure method of protection. The creeping barrage, it may be noted, had been entirely dispensed with. It commenced beyond the Quadrilateral, some 750 yards in front of the infantry.7 With enemy machine-gunners manning the parapet as the men deployed in no man's land, a bombardment this distant availed nothing and spelt disaster for the rightward attack.
In the leftward section of the divisional front, away from the Quadrilateral, it might be imagined that the attack stood a better chance. Here, the division was echeloned back some 500 yards from the Guards Division on the left. However, instead of trying to straighten the line to provide for a common start line for the infantry and artillery, the 6 Division's barrage was placed along an arbitrary line on the ground in order that it should conform with that fired for the Guards. So while the Guards' barrage coincided with an actual German trench, that in front of the left of 6 Division did not. In the indignant words of the divisional commander:
The stationary barrage was put on this [arbitrary] line, where there was not a single German, instead of on the trench which was full of Germans, as one well knew for we had already been repelled from it [on the 12th].8
The result of all this was a disaster for the division. On the right the battalions were swept away by concentrated machine-gun fire from the Quadrilateral. Further left a similar fate befell the battalions facing the unbombarded German trench line. Within minutes the attack had collapsed with 3,000 casualties.9 Throwing in a reserve brigade late in the day only added to the carnage.10 By evening the 6 Division had lost 4,000 men for precisely no gains.
The 6 Division was one of the units which was supposed to open the way for the cavalry. Within ten minutes it had been shattered. All that remained was the Guards. Clearly they could not redeem the entire plan but it seemed feasible that they could still capture Lesboeufs, which might provide some basis for cavalry action on the 16th. The tanks proved of no help. Ten had been allotted to them but five broke down at once and three more wandered rather ineffectually into areas occupied by other divisions, where late in the day they did some useful subsidiary work in ending pockets of German resistance.11
So in the Guards' area of attack, what the tank lanes managed to accomplish was the creation of clear fields of fire for G
erman machine gunners. Early on, this wreaked havoc on the Guards. Units soon became intermingled. Gaps appeared in some sections of the line and crowding in others; direction was lost; German trenches unknown to the British and therefore unbombarded added to the difficulty; flanking fire – especially from the right where the 6 Division had failed – caused further heavy casualties.12
Nevertheless, some attackers emerged unscathed through this hail of fire and managed to make some ground. How this happened is not at all clear. Some accounts suggest that the Bavarian troops in this area were in the process of being relieved and so were somewhat disorganised. Certainly, there could have been no ‘tank panic’ because none of the ‘armoured creepers’ were present at this time.13 Perhaps also, the ferocity of the Guards' attack unnerved the Bavarians. Many of the British accounts speak of the unknown German trench as if it were some kind of dastardly ploy that warranted the ‘slaughter’ of its inhabitants which was carried out by the Guards.14 Even after this trench had been captured more than one account notes that the men ‘bayoneted all the Germans they found’.15
After some hours of fierce fighting, the defenders opposite the 2 Grenadiers broke and ran back towards their second line. Passing through, a follow-up formation (3 Grenadiers) considered at one point that the road to Lesboeufs was ‘quite open’.16 This was also the impression of the Irish Guards who reported that ‘on reaching the top of the ridge the whole panorama of the landscape between Flers and Les Bouefs [was] disclosed in full view on the further slopes’.17
At this moment the Germans in the Lesboeufs area were certainly in disarray. But the position of the Guards was hardly secure. The leading units, looking towards the distant villages, were unsupported on either flank. Furthermore, a battery of German guns was coming into action against them. Shells soon started falling around the advanced troops.18 And because of the carnage caused in the early moments of the attack some of the battalions were much reduced. In the case of 3 Grenadiers just 100 men remained.19 Under these conditions it was thought prudent to consolidate the first objective and to dig in patrols slightly beyond to warn of German counter-attacks.20