To Arms

by Hew Strachan

  It was argued that the German attack on Liège did not necessarily confirm the direction of the main German advance. The city’s railway communications gave it a logistical importance independent of its position as a gateway to the Belgian plain, and its fortresses would have presented a threat to the German right flank for an advance south of the Meuse. But in the latter case it would have been sufficient for the Germans to mask Liège rather than attack it. Moreover, from 8 August onwards intelligence of German concentrations between Thionville and Liège multiplied. On 15 August the Germans attacked Dinant, on the Meuse and midway between Givet and Namur. That evening Joffre gave way to Lanrezac’s pressure and ordered the 5th army to advance north and northwest, to Namur and the junction of the Sambre and Meuse rivers.

  Neither Joffre nor Lanrezac imagined that the 5th army’s advance would be in isolation: it would be extended to the left by the BEF and by the Belgians, and it would be sheltered by Namur. But French plans did not conform with the expectations of either of their allies. Lanrezac’s push to the north overthrew the planning assumption of the British war council of 12 August, that the BEF was to guard the left flank of a French offensive into Luxembourg. Safely shipped across in 240 requisitioned vessels, the BEF was about to be chivvied into a counter-offensive against the main German envelopment, while its supply columns were still struggling to keep up. At least the British did their best to meet French wishes; by contrast, the Belgians expected the French to conform to theirs. On 18 August, knowing that the German 1st army was north of Liège and the 2nd army facing it, the Belgians finally concluded that the French would not get to them in time, and began to pull their field army back to Antwerp. Only the 4th division remained forward, committed to the defence of Namur which, although now isolated, became a possible pivot for the Belgian, British, and French armies. It failed to fulfil such grandiose conceptions, falling on 23 August.

  The results of the pressures on Joffre’s headquarters from Lanrezac and from King Albert were a change in deployment, more than a change in strategy. Only the French 3rd and 4th armies now faced the Ardennes. Joffre began to create a new army, the army of Lorraine, found from reserve divisions and commanded by Maunoury, to form up behind them and to guard their southern flank. But his intention to advance into the Ardennes remained unswerving, confirmed rather than undermined by the knowledge that the Germans were moving across the face of the French armies to their right. The intelligence which he acquired after 15 August, in particular the capture of the order of battle of the 2nd German army on 17 August and the conclusion reached on 18 August that the Germans would advance north of the Meuse between Namur and Brussels, further buttressed this decision. The strength of the Germans on both their wings suggested—given the French underestimate of the total German strength in the west—that they would be weak in the centre. Moreover, a French thrust through the Ardennes into Neufchâteau and Arlon would outflank the German right wing on its own left and so frustrate the German manoeuvre. But Joffre’s picture of the German deployment was still only partial. He failed to realize that the thrust of the 3rd and 4th French armies would be directed into the hub of the German wheel. The German advance into the Ardennes was slow, as the 4th and 5th German armies had to mark time while the 1st and 2nd swung through Belgium. But, as the wheel took effect, so the Ardennes would fill. Thus, two key assumptions were valid only for a short period. First, the German weakness in the centre—Joffre imagined that six German corps would face his eight—would not last. Secondly, Joffre was wrong to conclude that the Germans would not fight in the wooded and broken ground of the Ardennes but would be found on the further side. On 21 August Joffre ordered the 3rd and 4th armies to attack through the Ardennes, and the 5th to do so on the Sambre. The latter would hold the Germans while the attack of the former took effect.

  In the subordinate formations of the 3rd and 4th armies 21 August was a confusing and exhausting day. At its outset the 87th infantry brigade of II corps, on the right of 4th army, was charged with a defensive battle facing east; it was then told to turn north, albeit still with a defensive mission; by the evening its task was offensive. As a result it had three or four hours extra marching and was still not in its allocated start positions on the following morning. In common with the rest of 3rd and 4th armies, it did not halt until it was too late to establish its command links. In particular, II corps was not in contact with its right-hand neighbour IV corps, on the left of 3rd army. The reorientation of the 3rd and 4th French armies, from facing north-east to north, meant that their deployment was in echelon, each formation constituting a step in a staircase falling away to the south and east. The danger to each unit’s right flank was meant to be warded off by the support of the unit below it and to the east. But the lack of liaison meant that on the following day II corps attacked to the north, while IV corps was engaged by superior numbers to the east. The vulnerability of the French alignment was exacerbated by the southwesterly direction of the advance of the 4th and 5th German armies, which hit the French obliquely. Although on 21 August the 3rd French army had indications that it would encounter the German 5th, French reconnaissance was bad. The cavalry division attached to the 3rd army was poorly commanded and inadequately trained. The infantry divisions and brigades had been deprived of their integral cavalry in order to enable the formation of independent cavalry corps, and consequently each division had only three troops of reservists mounted on requisitioned horses. Aircraft had done good work so far, but on the 22nd they were grounded by early morning mist. The French began their advance in columns of march, with their 75 mms to the rear. As the guns came up they were deployed well forward for direct fire, and thus, when the mist lifted, were exposed to immediate German observation. The hilly terrain favoured the high-angle fire of howitzers, each German division having six 105 mms and each corps sixteen 150 mms. The 3rd army battle broke into a series of independent, ill co-ordinated engagements. V corps in the centre turned east to face the German line of advance and so opened a gap between it and its left-hand neighbour. On the 3rd’s left, the 4th army had a two-to-one superiority at the beginning of the day, although this was whittled away as the German redeployment took effect. More importantly, however, its centre and right were badly mauled before its left—with which it led and which could therefore have lapped around the German columns—was engaged. XVII corps’ right flank was exposed by the slowness of XII corps to advance in support, and suffered heavy losses at Bertrix. The 3rd colonial division on the left of 4th army’s II corps advanced furthest, but was not alerted to the cavalry’s reports that Germans were in the neighbourhood and so advanced in columns of four. As elsewhere on the front, it was not supported on its right. Thus it was caught in a devastating fire from its front and flank, and lost 228 officers and 10,275 other ranks killed, wounded, or captured in the course of the day. To the division’s south-east, at Virton, 3rd army’s IV corps was almost as battered. In the 130th Régiment d’ Infanterie the colonel was dead, all three of its battalion commanders killed or wounded, and only one of its twelve company commanders unscathed. Its neighbour, the 124th, had almost no officers, and had lost 770 men, or about a third of its strength. Nonetheless, both armies reckoned the day’s outcome had been indecisive. The German infantry had come on in dense formations, with their supports closed up, and where the 75 mms had been able to take effect had suffered horribly. The French planned to renew the attack the next day: the 4th army did so. But the losses, particularly of officers in battalions with a high proportion of reservists, were unsustainable without consolidation. The French retreat began on the 24th.73

  An integral part of Joffre’s plan for the defence of the northern frontiers was his expectation that the 5th army would attack on the Sambre at the same time as the 3rd and 4th armies to the 5th’s south advanced into the Ardennes. On 21 August, still underestimating by about half the German strength in Belgium, Joffre imagined that an attack by Lanrezac, with the BEF coming up from Maubeuge on his left, would li
nk with the Belgians in Namur and so create an effective barrier to the north. Lanrezac, however, was less certain. His army’s change of direction had been accompanied by a reshuffle of divisions which— although it reinforced it—disrupted its internal workings. His view of fortifications was that common among the pre–1914 students of mobile warfare: they were traps. Therefore, he was reluctant to hinge his operations on Namur or commit part of his army to support its defence. The Sambre itself was not a good defensive line: its valley was overlooked and filled with factories and pitheads which would inhibit manoeuvre and prevent the acquisition of clear fields of fire. Most compelling of all, his 2ème bureau was beginning to form an accurate picture of the German strength in Belgium. On the afternoon of 21 August Lanrezac finally recognized the scale of the threat in front of him and wisely decided to fall back south of the Sambre to fight a defensive battle. But he had dithered too long. That evening Bülow’s 2nd army secured crossings on the Sambre between Namur and Charleroi. The 5th army was now in contact with the enemy. Furthermore, von Hausen’s 3rd army was marching westwards, towards the Meuse south of Namur, and therefore threatening Lanrezac’s right flank. Whether he liked it or not, Lanrezac was committed to fighting a defensive battle in the Sambre valley—and not out of it—on the following day. His inability to decide earlier whether to attack or to defend meant that his troops were not entrenched and that he had not reinforced the Namur garrison.

  His own uncertainty as to his operational intentions was mirrored by the secrecy of his 3eme bureau and their failure to communicate openly with either Joffre or the BEF. Consequently, both the latter continued to assume that Lanrezac’s overall objective remained offensive. Joffre therefore asked the BEF to advance northwards, in conformity with Lanrezac. The BEF’s own command decisions were caught in the thrall of those of the French. On 22 August both General Macdonogh, the BEF’s director of military intelligence, through aerial reconnaissance, and the British liaison officer with the 5th French army warned Sir John French that a massive German envelopment seemed to be taking place: rather than turning in behind each other as they reached the Sambre, the German units were pushing out to the west. Nonetheless, Sir John ordered the BEF to march on Soignies, a move which imperilled his command on both its left, to the German envelopment, and its right, as it opened the gap between the BEF and the 5th army now behind and to the east on the Sambre.

  Although French counter-attacks on the 22nd proved unsuccessful, the position of the 5th army was stable. Bülow was temporarily checked, and the gap between the French and the BEF was bridged by Sordet’s cavalry corps. Sir John French wisely refused Lanrezac’s request that he swing right to attack Bülow’s flank, since he would thus have exposed his own army to Kluck’s 1st German army, but he did agree to halt on the Mons-Condé canal. By fighting frontally, the BEF would help guard the French left from the German envelopment.

  The potentially dangerous position in which the BEF found itself—forward of its allies and directly in the path of the main weight of the German army—was rendered null by the Germans’ own ignorance of the BEF’s whereabouts. The German naval staff had expected the shipment of the BEF to begin on 16 August, a full week later than was the case.74 The German advance through Belgium began on 18 August. Kluck, commanding the 1st army on the German right wing, believed that the British had landed, not at Le Havre and Boulogne but at Ostend, Dunkirk, and Calais. His inclination was to protect his outer flank by pushing towards the Channel coast. However, the effect of doing this would have been to stretch the German armies too thinly and to have opened gaps in their line. Kluck was therefore subordinated to Bülow, commanding the 2nd and inner army, so that the movements of the two armies would conform to those of the latter. The effect was to bring the line of Kluck’s march through Belgium south, and not south-west, and to direct Kluck against the BEF’s front and not towards its flank. But Bülow’s alignment only increased Kluck’s worries about the possibility of the BEF lying outside his right wing. The German cavalry failed to clarify the fears and suspicions of either army commander. The 2nd cavalry corps was brought under Bülow’s command on 21 August and directed south towards Charleroi, but was then drawn northwest in the direction of the Scheldt and reallocated to Kluck. A false report that British troops were arriving at Tournai, 48 kilometres to the west of where they really were, caused Kluck to halt for two hours while the report was investigated. The effect was to redirect the march of II cavalry corps on to Tournai from the north. Exhausted by its peregrinations, the German cavalry played no effective part in the fighting of the 24th. At noon on the 23rd, 3 kilometres north of Mons, one battalion of Kluck’s infantry was told by cavalry scouts—who had presumably reconnoitred to the south-west and not to the south—that there was no enemy within 80 kilometres.75

  Sir John French’s optimism about his strategic position—he was persuaded that at most only two German corps opposed him, when Macdonogh told him there were three—was thus somewhat vindicated by the nature of the battle the BEF was now to fight. Kluck’s initial attacks against the British were inevitably little more than part of his continued advance: they were not co-ordinated and were delivered in dense formations. The BEF had the advantage of prepared positions where it could develop the firepower of its infantry. But its corps commanders were uncertain whether French intended the action on which they were embarking to be the jumping-off point for a further advance or the preliminary to a general retreat. Tactically, the line of the Mons canal was no better for sustained defence than that of the Sambre: the canal itself did not constitute a major obstacle, and the buildings and pit-heads of the area obstructed fields of fire. On the right of the British position a bend in the canal created a salient which became exposed to German artillery as Kluck’s attack developed purpose and effort. The Germans broke into smaller groups and advanced by rushes, finding as they closed that the British were having to fire high because of the canal embankment and that once within 500 yards they were less exposed to its effects.76 The salient was abandoned in mid-afternoon. However, the BEF anticipated renewing the fight from its second line on the 24th.

  The salvation of the BEF was the sharp disillusionment of its commander. Sir John French’s reluctance to listen to Macdonogh’s accurate assessments of the German strength can only be explained by the dominance of Wilson and the consequent reliance on the conclusions of the French. Thus, it was the change in French intentions, not the efforts of his own intelligence services, that persuaded Sir John on the evening of the 23rd to order the BEF to retreat. Joffre was at last beginning to recognize the true nature of the German movement. The French commander therefore suggested that the BEF and the 5th army pull back to the line Maubeuge-Valenciennes. More immediately for the position of the BEF, Lanrezac had—independently, and in advance of Joffre’s orders—already ordered the 5th army to retreat. During the course of the afternoon of the 23rd the pressure of Hausen’s 3rd army on the right flank of the 5th army had begun to swing the balance against the latter. Without consulting the BEF, fighting—as Sir John saw it—on his behalf, Lanrezac began to pull back. From unrealistic optimism the British commander-in-chief switched to unjustified despondency. His erstwhile excessive faith in French assessments now contrasted with a deep distrust of Lanrezac.

  The two British corps spent the night of 23/4 August in positions 5 kilometres south of Mons, preparing to renew the battle the following day. Not until 1 a.m. were their chiefs of staff told that their orders now were to retreat, and even then no general plan for the withdrawal was issued. Sir John’s intentions fluctuated wildly. At one moment he spoke of falling back to Amiens and thus away from the 5th army; both it and the BEF would have been liable to defeat in detail. At another, he favoured placing the BEF within Maubeuge; most of its fortifications were out of date and trees obstructed its fields of fire. Ultimately, French did follow Joffre’s suggestions, but he still left the two corps commanders—Haig (I corps) and Smith-Dorrien (II corps) to make their own arran
gements. The 24th was well advanced before they could meet to do this. The opportunity to break contact with the enemy, if it had ever existed, was thus lost. II corps in particular, lying on the extreme left of the allied line, merited Kluck’s continuous attention as the Germans sought to envelop the BEF. On the night of 25/6 August the two corps were obliged to separate from each other by the forest of Mormal, II corps passing to the west and I corps to the east. In the course of the evening German advance guards clashed with a brigade of Haig’s corps at Landrecies. Haig, who, whatever his other failings was not a man prone to panic, seems—probably because he was also ill—to have done so on this occasion. French therefore concluded that the greater danger lay on his right than on his left, and that I corps, not II corps, needed nursing.

  In reality, Smith-Dorrien’s corps was scattered and exhausted, and had been fighting continuously for three days. By the time it was fully assembled around Le Cateau in the early hours of the 26th, daylight made it impossible for it to extricate itself. Smith-Dorrien decided he must turn and fight the Germans, the better to secure his retreat. Anxious to stiffen the infantry, the artillery of 5th division placed its guns well forward to give direct support. The effect was to ensure that the Germans’ counter-battery fire hit both elements simultaneously. Le Cateau cost II corps 7,812 casualties. The morale of some, regular soldiers though they were, snapped. On the following day the commanding officers of two exhausted battalions, now at St Quentin, agreed to surrender rather than carry on. The conclusion of Sir John French—that Smith-Dorrien should not have fought at Le Cateau and that the action had effectively broken II corps—seemed well founded. Smith-Dorrien did not agree. In the event his corps, though sorely tried, did not lose its cohesion.