The picture as far as the state of the Fifth Army defences was concerned was a mixed one and was hardly an ideal situation for Gough and his corps commanders to inherit. Even if the new system of defence in depth – framed by GHQ Instructions dated 14 December 1917 – was going to be implemented, the Fifth Army would be hard pressed to complete it in time to repel a German Spring offensive. In truth the British were struggling with the entire concept of defence in 1918; having been largely on the offensive for most of the war, senior officers and the units they commanded were now faced with implementing a scheme of defence in depth without any real understanding of how it worked in practice. The result was a patchwork of defensive positions and an extremely ad hoc approach to defence. ‘They paid the Germans the compliment of copying the methods used against themselves in 1917’, wrote Corelli Barnett, ‘Unfortunately they copied the letter not the spirit.’ The rigid top-down command prevalent in March 1918 appeared to stifle the flexibility of deployment. In the German system of defence in depth, infantry were deployed in support of the machine gun with two thirds of the troops retained as a mobile reserve for counter-attack. In the British version the proportions were reversed, with the machine gun subordinate to a defence that was based on the infantry.

Defence in depth relied upon three separate zones of defence. First and foremost was the Forward Zone in which defensive outposts depended largely upon the tactical features of the ground they occupied. There was no continuous front line here; troops were deployed either in what Paul Maze describes as ‘lightly held outpost screens’, or in redoubts which were situated behind the first line of defence. The Forward Zone – sometimes referred to as the outpost or advanced zone – was designed primarily to delay an enemy advance and force him to deploy in strength and, as some divisional orders made clear, was not to be reinforced. Troops in the Forward Zone were quite simply expected to hold on – or were they? Evidence suggests that there was considerable inconsistency in the orders given to troops in the Forward Zone by divisional and corps commanders. Orders depended very much on an individual interpretation by corps and divisional commanders of how the zones of defence should be used; the inconsistencies in these directives underlining an overall lack of understanding and an inability to coordinate the three zone defence in depth system. Tim Travers cites the example of instructions given to men in 140 Brigade:

‘The Brigade Major of 140 infantry Brigade, in 47 Division, of Third Army, was told to fight for the front line and hold it at all costs. But on their immediate right 26 Brigade [9th Division, Fifth Army] was told that in the event of a serious attack, they were to fight for the battle zone and not for the forward zone.’³

This apparently contradictory approach between the Fifth and Third Armies perhaps originated from GHQ’s evident acceptance that in the worst case scenario the Fifth Army would have to fall back on the line of the Somme. Conflicting orders such as these would inevitably cause confusion – as indeed they did. But it went further than mere contradiction. According to Colonel Edmund Ironside, Gough had apparently been ordered to place at least 50 per cent of his machine guns behind his front line, an order he failed to implement, preferring, as Ironside later commented, not to do this, deriding ‘the idea of not fighting for the front line’.⁴ It was Gough’s over manning of his Forward Zone – which was again in complete contrast to Third Army directives – that contributed much to the disintegration of the Fifth Army on the first day of the German offensive.

Behind the Forward Zone was the Battle Zone which, like the Forward Zone, was organized in up to three parallel defensive systems with inter-connecting communication trenches. Built into the system were further strong points designed for all-round fire. But in effect this was never really achieved. Each strong point was designed to have interlocking arcs of fire with those on either flank and although this was completed in some sectors, much of the work was negated by the dense fog that covered the ground on 21 March. The Battle Zone was the area in which an attacking force was to be held – in effect a killing zone that could be reinforced quickly and where the main battle would be conducted and won – except there was one fatal flaw: there were very few reinforcements to be had.

The third zone, dubbed the Rear Zone, was designed to be another battle zone but, particularly in the Fifth Army sector, it had not actually been completed. The Royal Engineers’ (RE) historian described the rear zone as merely spitlocked, ‘or at best, excavated to a depth of one foot, but subsequent experience showed that troops retreating in the dark – or even in daylight – could not be directed to halt on a spit-locked line, if indeed they ever saw it’.

One staff officer serving with the 14th Division was dismayed that all the divisional defence preparations were continually under scrutiny from the air:

‘Day after day I saw German aeroplanes fly low over our positions as and when they pleased. I have personally seen German airmen wave their hands at our people and have seen them dropping bombs on our working parties digging trenches in the Battle Zone. You must therefore understand that every trench that was dug by us was at once photographed by the Germans and was, on 21 March, blown out by shell fire … in consequence every strong point and platoon post on 21st was drenched with gas to start with and then shelled to blazes.’⁵

When III Corps took over the southern sector of the line from the French on 29 January they found a well-prepared Forward Zone but nothing behind it. Their answer to constructing a Rear Zone was to make strong points at the rear of the Battle Zone which accounted for the close proximity of the 2/2 London’s strong points to their Forward Zone outposts opposite La Fère. In fact this was hardly a defence in depth at all; both the Forward and Battle Zones were confined by the line of the Oise-Sambre Canal on the right, and the St Quentin Canal on the left – a distance which, in places, was not much in excess of two miles. Compare that to the situation west of Gouzeaucourt in the 9th Division sector where the Forward Zone near Vaucelette Farm was nearly two miles in depth and the Battle Zone occupied a further three miles. To be frank, the Fifth Army Battle Zone – like the Rear Zone – was still incomplete on the opening day of the offensive, particularly in the sector south of St Quentin down to the Oise.

Understandably there was considerable uneasiness amongst the troops over this new defensive system. Many of the ‘old hands’ referred to the strong points as ‘birdcages’ and grumbled about the lack of a continuous trench line. Soldiers will always find something to grouse about but in this case their protests had foundation. As we know, the British Army on the Western Front had been used to fighting from two lines of parallel trenches since the end of 1914 and was totally unprepared for the fundamental change that was brought about by defence in depth. The worst of these uncertainties was perhaps with regard to the distance between redoubts in the Forward Zone. In many divisional areas there was no communication between battalion headquarters and the redoubts and, often or not, they were not sited to be mutually supporting. Where communication lines were in place, the opening German barrage on 21 March destroyed many of these vital links and isolated the redoubts from their forward posts, and, as it turned out, there was very little to prevent enemy infantry from breaking through between strong points and continuing the advance almost unhindered. But what was more disquieting was that some divisional and brigade commanders, while conscious that their troops were being asked to fight in unfamiliar circumstances, were also unsure themselves how to conduct a defence in depth.

North of St Quentin the three brigades of the 24th Division held the sector around Le Verguier. Lieutenant Claude Piesse found himself in command of a number of outposts in the Forward Zone and his description of the headquarters post at Shepherd’s Copse did not inspire confidence:

‘All the posts were badly equipped for defence, the HQ worst of all. The field of fire over the 180 degrees towards the enemy varied from 25 yards to nil over about 160 degrees, the remaining 20 percent, though much better, also enabled the enemy to infiltrate the entire position. On arrival at the post I examined the supplies – there was a limited amount of rifle ammunition, not a single grenade, a number of drums of water, which I had the curiosity to open and found them so tainted with petrol as to be undrinkable … it was under these conditions that I had written orders not to return and fight it out.’⁶

Although Piesse’s experience was by no means representative of the whole picture, it was one that was replicated in many brigade areas. The shortcomings in Piesse’s sector were without doubt exacerbated by the acute shortage of manpower, which, together with a lack of time in which to prepare defences, was beginning to raise the issue of defence to crisis point. In the III Corps sector for example, RE units in the shape of 4/Siege Company and 365/Forestry Company were in place by 1 February but 182 Tunnelling Company took another week to arrive. The entrenching units were delayed until 14 February and the final RE units only arrived on 7 March, a mere two weeks before the offensive began. ‘It was a pity,’ lamented Major General Noel ‘Curly’ Birch on 10 February, ‘that we ever had to take over the extra front.’

In his position as artillery adviser to GHQ , Birch’s concern lay not only in the lack of guns per mile of Fifth Army front but also with the deployment of artillery batteries in the new defence zones. Rather than deploy in more formal gun lines, artillery commanders were obliged to adopt a more mobile framework to give each of the battle zones sufficient coverage. Major Kenneth Cousland commanding 462/Battery, Royal Field Artillery (RFA) recalled filling his ‘days and nights preparing main battle stations that were to be held if the front lines were overrun, reconnoitring reserve positions and OPs, registering targets and working out barrages’. Battery positions such as Cousland’s that were within 3,000 yards of the front line were strongly wired in and prepared as defensive positions which the infantry could fall back upon if necessary. Great importance was given to Forward Observation Officers (FOO) having a good all-round view of the ground ahead so as to direct observed fire, particularly as the threat of German tanks was considered to be a very real one. So much emphasis was given to this, that around 10 per cent of the 18-pounder stock was installed in the forward areas for use as anti-tank weapons. In the event this proved to be a complete waste of resources and, as Birch acknowledged, such a deployment ‘created a greater set of communication difficulties’.

For the troops on the ground this ‘new fangled’ defence meant only one thing, each brigade – now reduced to three battalions – would deploy one battalion to the Forward Zone and a second the Battle Zone, leaving one in support with each battalion of the brigade taking their turn in each of these positions. It was a system of defence that Chaplain 3rd Class Father Henry Gill, who was attached to the 36th Division southwest of St Quentin, had serious misgivings about:

‘There was no system of communication trenches between the forward area and [the Battle Zone]. In fact there was no question or possibility of the battalion holding the front line being supported or relieved – or of getting away – except at night. I understand the forward battalion was expected to hold out for at least forty-eight hours before there would be any question of relief! It is not surprising that the people in the front trenches had anxious views on the whole position, and as the time of the attack was, of course, unknown, everyone in the division had an interest in the matter … it was quite clear to me that anyone who might happen be in the front line when the attack took place would never get back alive … it was a toss-up as to which battalion of each brigade would be sacrificed’.⁷

‘Sacrificed’ was a word used by Lieutenant Colonel Charles Howard-Bury, commanding the 9th Battalion King’s Royal Rifle Corps (9/KRRC) in the 14th (Light) Division. Charles Howard-Bury clearly felt quite troubled over his battalion’s limited prospects should it find itself in the forward zone on the day of the German attack; it would, he predicted, be wiped out without any hope of assistance from the remaining battalions of 42 Brigade. With the possibility of counter-attack apparently ruled out – which was contrary to the German view of defence in depth – Howard-Bury understood the dilemma facing those troops in the Forward Zone:

‘The troops of a Forward Zone or Outposts, or whatever you may choose to call those in front of the main line of defence, may accomplish their task in one of two ways. They may fall back fighting, in which case they are certain to mask the fire of the troops behind them, a very serious matter in these days of complicated artillery barrages; or they may stick it out to the last man. Now, no-one will deny that there are occasions when one part of a force must be sacrificed to save the rest, when troops must be expected to stick it out until the last man; but is it wise to make it a matter of routine?’⁸

Knowing his men and understanding what he called ‘human frailty’, he recognized instantly that if a soldier knows there is even a remote chance of relief he will generally stand fast and fight, but if there is no chance of relief the temptation to surrender before the last round has been expended is often an overwhelming one. A similar sentiment was echoed by Captain Charles Miller, a company commander with the 2nd Battalion Royal Inniskilling Fusiliers (2/Inniskillings):

‘The British army has certain shibboleths, one of which, and it has cost the lives of scores of thousands of soldiers, is that when you are attacked in overwhelming force you mustn’t run away. The French, who are much more logical than we, and who consider results and not prestige, invariably run away under such circumstances, and when the right moment comes run back again and deliver a counter-attack.’⁹

Many senior officers harboured serious doubts about the distinct lack of reserves that were available for counter-attack purposes. Lieutenant Colonel Thomas Mudie, on the staff of the 9th Division, felt it was ‘suicidal’ to have no strong reserves, a factor he felt was responsible:

for the rapid disintegration of long stretches of the Fifth Army front and the loss of whole battalions, who sacrificed themselves heroically, but who have done much more good had they been in touch with one another and retired fighting.’¹⁰

Exactly how many of the surviving units in the Forward Zone actually did run away on the morning of 21 March will never be known, but ‘human frailty’ would have ensured there were plenty who preferred captivity to a battlefield death unless, as Claude Piesse remarked, they were under an officer who was willing to hold out until the bitter end himself – but, as Piesse himself acknowledged, many of them did not do so.

The manpower question continued to raise its head again and again as the need to dig and prepare the defensive line became the prime task. Despite Lloyd George’s assertion that the labour force was quite adequate for the task in hand, it was obvious to those who were actually in the front line that the labour force would have to be supplemented with troops. It was a source of much discontent amongst battalion commanders who, intent upon training their men and coping with front line duties, were constantly distracted by the need to supply digging parties. Writing to his wife on 5 March 1918, Lieutenant Colonel Rowland Fielding, in command of the 6th Battalion Connaught Rangers (6/Connaughts) who were positioned just south of Épehy, despaired over the amount of time his battalion had spent in front-line duty:

‘The battalion wants a rest. It had been up for forty-two days when, last night, it was relieved, and even now I doubt if the rest is in sight, since an order has just come in to go up to-morrow for the day to dig. I leave you to imagine the state of the men’s bodies and clothing, after so long a time in the line, almost without a wash’. ¹¹

No-one, it appeared, was immune from the constant demands of the working party. Lance Corporal William Sharpe, serving with the 2/8th Battalion Lancashire Fusiliers, was in charge of a wiring party in the 197 Brigade sector, north of St Quentin. Since 15 March he and his detail had been regularly wiring their positions:

‘Each night when it was dark, 12 nice lads and myself got gingerly over the parapet and took iron stakes and spools of barbed wire and constructed nervously, and as quietly as possible, barbed-wire fences, getting into all sorts of tangles, dropping down when the ‘zip-zip-zip’ of machine gun bullets came too close (i.e. when one ‘felt’ the whistle), and standing stock still like tree trunks when the firework display (Very lights) went up.’¹²

Charles Miller, recognizing that time was against him, attempted to keep up the spirits of his men by delivering a lecture on what he called ‘this new idea in defence’, but much to his annoyance the four days spent in so-called ‘reserve’ was generally taken up with digging the defence lines. He recounts an occasion when Brigadier General William Hessy, commanding 109 Brigade, decided in his wisdom to alter the positions of several strong points the battalion had been working on for several days:

‘The appalling and crass stupidity of it all! Putting up barbed wire entanglements at night time is a hard enough job in itself, but we reached the point where we had to uproot the entanglements we had previously erected on abandoned sites, cart them off and erect them on new sites, which is simply Herculean labour. And all this had to be done on the nights we were supposed to be resting.’¹³

But there was a far more serious outcome to a scenario that working parties on other divisional fronts would have instantly recognised. Miller evidently understood the need to find a balance between training his men to fight in the new defensive zone and that of preparing the defences. It was an unenviable task and one he felt was in great danger of tipping in favour of the enemy:

‘In the first place the men were tired to death; in the second place, since the position of the strong points was constantly being changed it was impossible to organize a regular drill by which every man knew his strongpoint and got there in the quickest possible time when ordered to do so; lastly, instead of being deeply dug, strongly revetted and wired it was quite obvious that when the moment came to use them, the strong points would hardly be strong enough to keep out a well-aimed snowball.’¹⁴

What really concerned Miller, along with Piesse, Charles Howard-Bury and others, was that when the offensive actually began, it was highly unlikely the Germans would be armed with snowballs. Had they been asked in hindsight, all three would probably have lent their support to the argument that one of the key contributory factors to the German success on 21 March was the failure of a defensive system that was incomplete and not fully understood and implemented by senior commanders.

* * *

Apart from the infantryman, the remaining fighting troops of an infantry division in 1918 were made up from the artillery and ‘sappers’ of the RE Field Companies. The sapper was a fighting soldier whose qualities had been demonstrated during the retreat from Mons in 1914 which had begun unexpectedly after a single day’s fighting and, except at Le Cateau, heavy fighting was generally avoided. The four infantry divisions at Mons had two companies of RE per division and a supply of explosive material that was limited to what could be carried on the field company carts. Despite this, the field companies excelled in what were extremely trying circumstances. After the trench lines became established at the end of 1914, the RE settled down to the drudgery of trench warfare along with the rest of the army and for many of the young soldiers serving along the Third and Fifth Army fronts in March 1918 – and indeed a good proportion of the sappers themselves – the RE were perceived as a builder of dug-outs and the provider of trench-boards and revetting materials. The German offensive changed all that, and to use the words of Lieutenant Colonel Reginald Butterworth, the Chief Royal Engineer Officer (CRE) with the 16th Division, ‘the sapper came into his own again, laying down his shovel to pick up a rifle’.

Towards the end of 1917 when GHQ considered the notion of demolition schemes, discussion inevitably stumbled across the thorny question as to who was responsible for preparing the destruction of French controlled railway bridges, track and rolling stock. Anxious not to tread on the toes of the French, GHQ fudged the matter, accepting that when the time came the French would, if it became necessary, carry out their own demolitions. Sadly this did not always happen, and when the time did arrive it was left to the Fifth Army engineers to destroy over 250 bridges for which their ally had responsibility while others were left intact and fell into enemy hands. Typical of this ‘misunderstanding’ was the failure to completely demolish the railway bridge at Pithon, allowing German infantry to cross the Somme Canal at that point. It was, in the judgment of many RE officers, an example of where the ‘British should have retained the responsibility of dealing with all the bridges in their fighting area’.

Those RE officers who fought at Mons and during the subsequent retreat would all agree that the nature of demolitions carried out in 1914 were substantially different to those of 1918. For a start there was a demolition plan in existence in 1918 and bridges and supply dumps had been reconnoitred and numbered; there was even a list of demolition arrangements lodged with the CRE of each division. But the plan was incomplete, particularly as a number of bridges had been constructed after the plan had been completed! Writing with the benefit of hindsight in 1933, the Fourth Army Chief Engineer, Major General Reginald Buckland, highlighted what he saw as the basic shortcoming in the plan:

‘Defence works were pushed on as labour became available, and many additional bridges were built to improve communications, causing constant additions to the schemes for demolitions … [even when] an attack on a vast scale was felt to be imminent, there seems to have been no apprehension that the battle zone and rear zone might be overrun before reinforcements from outside of the army would arrive. Consequently we find complete schemes of demolitions prepared back to the line of the Somme and the Tortille, but behind those rivers little work of that nature was done before the fighting began.’¹⁵

Buckland also drew attention to the tactical necessity of keeping bridgeheads and crossing points under fire to prevent the enemy’s bridging equipment from being brought into use. The CRE to XVIII Corps, Lieutenant Colonel John Craster, had also raised this issue: ‘there was no bridge over the Somme, or the canal, the demolition of which need delay the enemy’s field artillery for more than two hours, unless the bridge was kept under rifle or machine gun fire at close range’. It appeared that no-one was listening at the time or perhaps no-one was prepared to put his head above the parapet and shout loud enough.

Bridges and their destruction were usually the responsibility of the corps concerned but to make matters more complex, some bridges were under the control of Army Headquarters. As the retreat unfolded the hierarchical command structure that the BEF insisted upon, broke down as positional warfare gave way to a rapidly moving warfare. In these circumstances some Corps commanders failed to carry out their responsibilities while others felt constrained by army orders which failed to materialize in time for demolition to be carried out.

The bridges behind the Fifth Army front were either of British construction or built by the French after the German retirement to the Hindenburg Line in 1917. The larger British constructions were supported by steel girders but the vast majority were timber bridges of all patterns and sizes. Very few were more than a few feet above the water level which in March 1918 was very low. Lieutenant Colonel Arthur Walker, CRE to XIX Corps was concerned that bridge demolition alone would have a limited effect on an advancing enemy:

‘I am of the opinion that a great deal of the destruction herein arranged for can be easily made good by the enemy, owing to the width of the streams being small and the beds of the streams being merely a few feet below the bridge roadway. I think the enemy could be held up to an equal degree by blowing several large craters in roads as that through the marshland at St Christ.’¹⁶

It was a point that Fifth Army Headquarters did consider briefly before deciding that all available tunnellers were better employed in the construction of dugouts, a move that might be considered as short sighted. Causeways and roads across the Somme marshes, such as those at Brie and St Christ Briost, could have been easily cratered and the potential disruption to the enemy was one that was even obvious to the infantry. In his role as battalion Transport Officer, Lieutenant Claude Piesse had visited the St Christ Briost bridgehead before the offensive began:

‘No arrangements had been made to blow the earthen causeway [at St Christ Briost]. The wooden bridge would be replaced by the enemy in probably a couple of hours, but if the causeway had been properly blown, it would probably have been a two day delay.’¹⁷

Piesse had no way of knowing that his view was shared by Major General Gerard Heath, the Engineer in Chief at GHQ , who even tabled the importance of the destruction of roads and causeways on 28 February; sadly his advice, and that of Piesse, who forwarded his observations to Divisional Headquarters, were ignored.¹⁸

* * *

Much has been written of the German preparations for the March offensive of 1918 and the use of storm troops or sturmtruppen and the battlefield tactics employed by them. These cutting edge troops were certainly effective – as had been seen at Cambrai during the German counter-attack – but their tactics were not developed on the Eastern Front as is often supposed. The first sturm-bataillion was formed under Hauptmann wilhelm Rohr – Sturm Bataillon Nr. 5 (Rohr) – who famously led the battalion in taking the Hartmannsweilerkopf during December 1915. Typically each sturm-bataillon would consist of three infantry assault companies, a machine-gun company, a mortar company and an infantry artillery battery – armed with up to four 37mm guns – and a Flammenwerfer section. Ian Passingham and others agree that storm battalions were both a blessing and a curse, pointing to ‘their more elite status [which] led to the stripping out of the best infantry soldiers, NCOs and officers in many line infantry battalions for specialist training, so that the parent unit was weakened’.¹⁹

It also appears that the term storm troops was one loosely applied to high quality troops on both sides, German sources prior to 1918 occasionally referred to attacking British units as sturmtruppen although these references may have been aimed at Dominion troops who also utilized speed, aggression and fire-power as their modus operandi. One of the great myths of the March offensive often focuses on the superiority of the German sturmtruppen and the inadequacy of the British troops. Clearly the German front line attacking troops were of very high quality and extremely well trained but we should certainly not dismiss the number of exceptionally good British divisions and brigades who fought just as well and about whose exploits this book is primarily concerned.

German units were trained to advance at speed, taking full advantage of openings in the British defence lines. The artillery creeping barrage was designed to maintain pace with the infantry and the forward units were under orders to advance until they were exhausted. No daily objectives were set for these units, Ludendorff quite rightly assuming that if daily tactical objectives were set, troops would be inclined to dig-in once these had been achieved, slowing the impetus of the advance. It is practically impossible to put an accurate figure on the numbers of storm troops taking part in the 21 March 1918 attack but somewhere in the region of 6,000 is thought to be a reasonable estimate. Another of the myths of the March offensive was that every attack was spearheaded by storm troops, an impossibility given the limited numbers of these troops which adds weight to the consensus that storm troops were allocated to sectors where a particularly strong defence was anticipated. In other sectors German commanders did what their British counterparts would do in similar circumstances; allocated their best divisions to lead the attack.

Apart from the veteran ‘Easterner’ Oscar von Hutier, whose reputation followed him to the Western Front, Ludendorff brought in Oberst Georg Bruchmüller who is sometimes credited – quite incorrectly – with masterminding the whole of the German artillery plan for 21 March. There was in fact no single artillery commander and while important to von Hutier, Bruchmüller was not the key figure he is often made out to be. However, he did introduce a new approach to artillery offensives and his experience did – with Ludendorff ’s backing – influence the Seventeenth and Second Army barrages, but in effect he was only directly responsible for the Eighteenth Army artillery.

For much of the war artillery bombardments prior to offensive action had achieved mixed results, any element of surprise was immediately lost by the protracted and intense bombardment on enemy lines which only served to loudly announce the forthcoming infantry attack. Serving on the Eastern Front on the staff of the 86th Infantry Division, Bruchmüller introduced the short ‘hurricane’ bombardment that was designed to take out all the enemy defences from the front line emplacements to the gun lines at the rear. In short he advocated a saturation barrage using a mixture of high explosive, gas and smoke which would neutralize the enemy’s defences, prevent or inhibit counter offensive action and disrupt the effectiveness of command and control centres. Firing ahead of the infantry in what was a totally predicted timetable, the infantry were able to advance with the guns. Such a bombardment preceded the infantry assault at the Battle of Lake Naroch in March 1917 and again at Riga six months later. Both battles were a complete vindication of his methods and earned him the nickname of Durch-Bruchmüller.²⁰ Bruchmüller’s artillery preparations and methodology may well have influenced the other two artillery commanders as by 20 March the German front from Arras to La Fère boasted 6,473 guns and 3,532 mortars in support of Ludendorff ’s three German armies that were spearheading the attack.

Incredibly the Germans managed to conceal their offensive intentions from the French and British with remarkable skill. For months the withdrawal of divisions from the front line for training and reorganisation had taken place without attracting attention, attack divisions had been assembled in the rear of the line, well beyond the reach of observing aircraft, while the more obvious dumps of ammunition and gun positions had been well camouflaged. Finally towards the middle of March the huge concentration of men and materials began moving up to the front, marching by night and concealing their movements by day.

By way of contrast to the meticulous preparation that went into the infantry and artillery attack, the more mundane but essential aspects of transport appears to have been neglected by Ludendorff. Randal Gray’s observation that ‘Ludendorff had no arm of exploitation beyond his men’s feet, the hooves of insufficient and underfed horses and motor transport often on iron tyres (owing to the rubber shortage)’, is a crucial one and points to the shortages of basic raw materials. Horses were in short supply as were motor vehicles, and those that were in service were more often than not fitted with iron ‘tyres’. These would prove to be practically useless in crossing the battlefield as would the nine tanks prove to be ineffectual against the greater numbers possessed by the allied forces. Yet the March offensive did see Germany’s first A7V Sturmpanzerwagen in action, five of which were deployed north of St Quentin – although two had broken down before the attack. They would be in action again near Villers-Bretonneux when three A7Vs met three British Mark IVs in what has been declared as the first tank versus tank battle in history.

On the night of 20 March Charles Miller received word that the expected German attack was imminent and there was to be no question of retreat: ‘I had time to make a visit to one of the points occupied by my platoons. I hated the look of it: a couple of dug-outs, hastily made and inadequately wired, and I wished I had disobeyed orders and put them in one of the previously discarded strong-points’. It was only a matter of hours before Kaiserschlacht would be unleashed and Miller along with thousands like him would soon be tested to the limit of their endurance by a revitalized and powerful enemy intent on ending the war once and for all.