To Arms

by Hew Strachan

  The initiative to recall Prittwitz and his chief of staff, Waldersee, did not originate with Moltke but arose from within the operations section of OHL. Max Bauer telephoned each of the 8th army’s corps commanders. All were agreed that the retreat to the Vistula was premature.85 The upshot was that Erich Ludendorff, freshly adorned with the laurels of Liege, and himself of course a former head of the operations section, was asked to report to Moltke at Coblenz at 6 p.m. on 22 August. Ludendorff was named chief of staff to the 8th army; the appointment of the new army commander, Paul von Hindenburg, was confirmed that night. The two men met for the first time on the platform at Hannover station at 4 a.m. on the morning of the 23rd. Hindenburg, square-headed and big, a veteran of Königgrätz as well as of East Prussian manoeuvres, had been retired in 1911 at the age of 64. Although sounded out by Moltke on 3 August as to his availability, he had spent the first three weeks of the war chafing in anticipation of the Kaiser’s call.

  Because the outline of the Tannenberg manoeuvre was already in place, it is easy to diminish the role played in the battle by Hindenburg and Ludendorff. The victory established their joint reputation, and yet they were not its authors. But they were its executors, and each made a vital personal contribution. Ludendorff had already issued his first orders on the evening of 22 August. Dismissing any intermediate contact via 8th army headquarters, he had dealt directly with the corps commanders. He confirmed the movement of I corps to XX corps’ right flank. He told XVII corps and I reserve corps to rest on the 23rd. Hoffmann complained that Ludendorff thereby lost a day and increased the already considerable demands which would be made on the foot-soldiers of those two corps. But that was with the wisdom of hindsight. What Hindenburg and Ludendorff inherited was defensive and reactive in conception; it made possible a more extensive manoeuvre, but it did not anticipate it. Furthermore, if any German formation had the potential for envelopment at this juncture it was I corps bearing down on Samsonov’s left, not XVII corps and I reserve corps, whose immediate task was to support XX corps in its frontal battle. By holding these two corps back for a day Ludendorff hoped to be clearer about Rennenkampf’s intentions. If Rennenkampf followed up, the two eastern corps should march south; if not, their movement could be to the south-west.86 On 24 August I reserve corps was instructed to advance in the direction of Bischofstein and XVII corps to Bartenstein. Whereas Hoffmann on the 20th had had to allow for a possible battle with Rennenkampf, Ludendorff on the 24th could be clear that his objective was Samsonov. Hindenburg’s task was less evident, but nonetheless vital: it was to steady the nerves of his chief of staff. Hindenburg’s personality has remained elusive; he had been favoured by Schlieffen as his successor as chief of the general staff, but his memoirs suggest a simple and somewhat unimaginative man. If these were his qualities, they were the ideal foil to Ludendorff. The latter, having adopted a plan, would then be seized with self-doubt. Even as he closed in on Samsonov, Ludendorff’s worries about Rennenkampf threatened to rob his actions of decisiveness. Hindenburg’s contribution, too easily underestimated—not least by the vainglorious Ludendorff—was to convey imperturbability.

  The speed of Samsonov’s advance, which had so alarmed Prittwitz, had been achieved at the expense of sufficient preparations for the march. And even then the 2nd army was not fast enough. The penalty of the pre-war neglect of supply was far heavier for Samsonov than for Rennenkampf. Samsonov had only one lateral railway line available to him, and that ran from Vilna through Grodno and Bialystok to Warsaw—in other words, well behind his front. The lines running at right angles, which crossed the frontier, lay outside the army’s initial frontage, to Mlawa on the left and Grajewo on the right, and were single-tracked. That directly behind the concentration area ended at Ostrolenka, 50 kilometres from the German border. The roads north of Ostrolenka were sandy and unmetalled. As a result units began their march still incomplete. The four corps of which the army was comprised came from four different districts, and therefore the army as a whole lacked homogeneity. They shared lines of communication, meaning that one corps tended to receive too much and another too little. The push to the west, encouraged by Stavka but opposed by Zhilinskii, became a necessity in order to be able to exploit the Mlawa line. However, the effect of the reorientation was to send the army first in one direction and then in another, and to confuse its deployment and communications yet further. By 22 August all units had marched for ten to twelve hours each day for the past week, and yet had only just crossed the frontier. Their daily distances were extended by the need to requisition. But the local hay and oats had not been cut, and much of the terrain was in any case a sandy desert. Samsonov’s army had still to make contact with the enemy; it was already close to collapse.

  As alarming was the absence of effective communications. The army had two field radio stations. Most of their messages were in clear, in part because each corps had different ciphers, but principally in the interests of speed and accuracy; in this respect the Russians were little different from the other armies of 1914. There were twenty-five telephones and only 130 kilometres of wire for the entire force. While within Russia the army used the three government telegraph lines which ran as far as the frontier, to Myszyniec, Chorzele, and Mlawa. Each corps then extended the line as far as it could with its own stock of cable, but it had nothing spare with which to establish links between its own headquarters and those of its component divisions. A lateral line laid behind the frontier was then taken down to conform with the army’s advance. Thus, the army’s internal communications had collapsed by 20 August. Nor was its liaison with other bodies any better. Samsonov left his administrative staff in Ostrolenka and was accompanied by a field headquarters only. By the time he had established his command in East Prussia, at Neidenburg, messages between the two components of his staff were relayed through five stations. He had no direct contact with Rennenkampf, and little with Zhilinskii. Telegrams from the Front headquarters, whose main purpose was to hasten his progress, were forwarded from Warsaw by car. Therefore, the dispersion of Samsonov’s army and its separation from the 1st army were not offset by efficient communication. Loss of command and control was the inevitable consequence.87

  Samsonov’s advance was predicated on two totally false assumptions: first, that Rennenkampf was still pushing from east to west; and second, that the Germans were fleeing back to the Vistula and must be cut off before they got there. Pre-war thinking in Russia had praised the German practice of not having an independent advance guard to establish the enemy’s position and intentions, but of instead directing independent columns to converge on the battlefield. This practice led to German deployment straight from the line of march and to heavy losses in 1914. The Russians followed them, a 1911 manual dispensing with a vanguard for long-range security and leaving each column responsible for its own protection.88 Samsonov’s army, therefore, had no advance guard; each corps moved independently and was responsible for its own reconnaissance. But in this latter task the Russian cavalry, although three divisions strong, failed miserably: many mounted units stuck to the flanks, fearful of German envelopment, rather than pushed forward, and those that pushed forward lacked the means of communication with corps head-quarters.89 Once into East Prussia the Russians were on unfamiliar ground, referring to maps marked in the Latin alphabet, not in Cyrillic, and confronted with a population that (if it had not fled) was adamantly hostile. Captured German documents were rendered valueless for the lack of a German linguist. Aerial reconnaissance, of much value on other fronts in 1914, was largely denied Samsonov, since the bulk of the available Russian aeroplanes had been directed to Galicia. The reports of the 2nd army’s staff ‘were framed in the form of abstract statements and presentiments, but gave no assistance by means of clear and concrete appreciations’. Such specific intelligence as was gained thus lacked a context into which it could fit.90 Even as late as 27 August Samsonov could continue to underestimate the forces opposing him.91 Stavka’s preoccupation with Poland had tak
en full hold after Gumbinnen; Samsonov’s intelligence failures did nothing to correct the belief that the Germans were intent on the abandonment of East Prussia.

  On 23 August Samsonov asked Zhilinskii for permission to advance onto the line Allenstein-Osterode, in order to cut off the German retreat after Gumbinnen, rather than the line Rastenburg-Seeburg. Zhilinskii gave in to the prevailing optimism, setting Samsonov the compromise target of Sensburg–Allenstein by 26 August. To achieve this move to the north-west, Samsonov shed II corps on his right, now caught up in the Masurian lakes and transferred to Rennenkampf’s command. He was also instructed to leave I corps at Soldau, to guard his left. Thus, while Samsonov’s front widened to 96 kilometres, his strength fell to just over three corps.

  On the same day XV corps, in his centre, ran into the Germans’ XX corps on the line Orlau-Frankenau. A fierce frontal battle developed on 24 August, with Hindenburg and Ludendorff telling Scholtz to hold well forward in order to give I and XVII corps the space in which to deploy. If they had intended to envelop Samsonov at this stage they would have lured him into a sack by pulling XX corps back; instead, they opted to run the risk that Scholtz would himself be enveloped.92 The men of XX corps responded to the challenge: their peacetime headquarters lay only a few kilometres away at Allenstein, and at one stage their chief of staff, Emil Hell, was directing artillery fire onto his own house.93 Not until 25 August did XIII Russian corps, on XV corps’ right, begin to pressurize the Germans’ flank. Poor inter-corps communications and inadequate reconnaissance was costing the Russians the opportunity to accomplish their own envelopment. Scholtz now pulled his corps back, ready to align with François’s I corps, which was due to enter the line on his right on the morning of the 26th. However, Samsonov interpreted Scholtz’s move as fresh evidence of the supposed general German retreat, and secured Zhilinskii’s permission to advance on the line Allenstein-Osterode. The effect of yet another move to the west was to separate XIII Russian corps from VI corps, now on Samsonov’s extreme right and still directed northwards to Sensburg and Bischofsburg. During the course of the 25th Samsonov began to realize his mistake, and his doubts were fuelled by reports of the movements of the Germans’ I and VI corps.94 But Zhilinskii encouraged Samsonov to continue his advance, the Russians construing the threat as one from inferior German forces concentrating on the 2nd army’s left, where it had the support of I Russian corps at Soldau. Samsonov therefore reckoned on enveloping them with his right. The Germans were now threatened on both their flanks, and by 26 August the atmosphere at 8th army headquarters was strained. But what Samsonov did not appreciate, principally because of his misapprehension about Rennenkampf’s movements, was the perilous position in which his right now found itself. On 26 August Mackensen’s XVII corps and I reserve corps, by dint of a prodigious effort on the 24th and 25th, after their halt on the 23rd, came up to strike VI corps at Bischofsburg. Russian pilots had reported troops moving south-west, but had concluded that they were Rennenkampf’s. VI corps had failed to use its cavalry to reconnoitre and was preparing to swing to its left, onto Allenstein, to support Samsonov’s centre. It was totally surprised. Only the Germans’ exhaustion enabled its escape. Driven back in disorder, the Russians did not stop until they had reached Ortelsburg. Now Samsonov had no protection on his right and no reinforcements for the two corps, XIII and XV, in his centre.

  Ludendorff’s tetchiness on the 26th was fed by I corps’ slow approach to the battlefield. What he had in mind was a comparatively limited victory. The task of XX corps was to fight and fix the Russians, not to lead them on; the task of I corps was to extend XX corps’ front to the right and at right angles. He was becoming anxious about Rennenkampf and the threat to XVII corps’ flank; he was not privy to the intercepts of Samsonov’s orders, which made it clear that the Russians would not attack Scholtz’s corps on the 25th. He told François to attack XV Russian corps’ left flank at 10 a.m. on the 26th. François disobeyed. Despite the prodigious efforts of the railway departments, only twenty out of thirty batteries were at his disposal, and his corps was therefore not ready to attack the Russians in prepared positions; in any case, he preferred envelopment to a frontal attack. The effect of the delay was to allow Samsonov to continue to focus his attention on XX corps, and so draw in his two centre corps. François’s attack on the principal Russian position between XV corps and I corps, at Usdau, began on the morning of 27 August. The Russian trenches were continuous, without traverses or overhead protection, and thickly manned. I German corps’ artillery, now up, inflicted heavy casualties. Usdau fell by 11 a.m. François then turned south to push I Russian corps back to Soldau. The burden of the Russian attack therefore fell on XX corps. By the morning of the 28th François realized that the threat from Soldau was diminishing and that he could resume the attack on the Russian centre. Ludendorff ordered him to do so by a direct advance at Lahna. For the second time François disobeyed. He opted to direct his march on Neidenburg, so aiming to the east and achieving a fuller encirclement. By nightfall he was at Neidenburg, and his advance units were moving eastwards to Willenberg.

  If any German created the opportunity for the envelopment of Samsonov’s left between 26 and 28 August, it was François. But its principal author was not a German but Samsonov himself. The delay in I German corps’ attack until 27 August encouraged him to focus on his centre. On his right XIII corps was advancing on the railway junction of Allenstein, both to secure food and to converge with VI corps. Unable to establish radio contact with VI corps, and unaware that the latter was falling back to the south-east, XIII corps entered Allenstein on the afternoon of the 27th. It was now poised to break through to Rennenkampf (indeed, 8th army headquarters feared it might belong to Rennenkampf’s army), or to cut across the communications of the German corps to left or right. Hindenburg’s and Ludendorff’s priority, therefore, was to drive XIII corps out of Allenstein. I reserve corps was instructed to abandon its advance south, across the Russians’ line of retreat, and to turn west to Allenstein; XVII corps was told to conform. But Samsonov responded less to the movements of XIII corps and more to François’s attack. On the evening of 27 August he told XIII corps to move south from Allenstein to combine with XV corps in an attack on the flank and rear of Scholtz’s XX corps. The weary soldiers of XIII corps assumed that I reserve corps, now approaching it from the west, was VI Russian corps, which would therefore protect their right; at dawn on 28 August they set off on their 32-kilometre march to the south-west. It was these Russian movements which delivered the 2nd army into the jaws of a German envelopment.95

  On the morning of the 28th Samsonov learned of the defeat of both his flanking corps, I corps on the left and VI corps on the right. The precariousness of his position at last became clear. But even now his situation could be construed as serious rather than hopeless. The Germans, although enjoying a superiority of 155 batteries to 132, had only been strong enough to envelop a part of Samsonov’s army, XIII corps and XV corps, leaving both I corps and VI corps outside the net. François had had to spread twenty-five batteries over a distance of 50 kilometres in order to form a thin cordon to the south. The Germans had communication problems of their own, and only I corps was in regular contact with Ludendorff. Below’s I reserve corps was slow to start its march on the morning of the 28th, and it was 10.30 a.m. before it realized that Allenstein was only lightly held. Below switched his route back to the south, but Mackensen was not told of this until 12 noon. Although XVII corps had reported that VI Russian corps was in full retreat to the south at 1.20 a.m., its movements remained subordinate to those of I reserve corps until 2.35 p.m. By the time Mackensen received definitive instructions to push south as fast as he could, the day was hot; moreover, the roads before him were poor. Even now Hindenburg and Ludendorff continued to think defensively. Fearful of an intervention by Rennenkampf, they were inclined to leave both their left-flanking corps in the vicinity of Allenstein. Furthermore, the fact that XIII corps had evaded the clutches of I re
serve corps increased the possibility of it helping XV corps to escape east. Thus, their orders of 10 p.m. on 28 August told I reserve corps to attack south and east on the next day, while XVII corps was to be ready to face either Samsonov or Rennenkampf. But the order never reached Mackensen. At 6.30 a.m. on the 28th Ludendorff learnt that XVII corps was approaching Ortelsburg. It was now too far south to protect 8th army against Rennenkampf, and so Ludendorff authorized it to proceed. Closing the gap with François was a protracted process.96

  On the morning of 28 August, Samsonov went forward from his army headquarters at Neidenburg to join XV corps: he was therefore unable to direct the operations of I corps and VI corps. I corps fell back to Mlawa, thus drawing away yet further from the remnants of the 2nd army. XV corps tried to break out to the south, but found German artillery and machine-guns posted at the exits to the forest. Turning east, on the night of the 29th, it blundered in the dark into XIII corps, in the process of falling back to the south-east. The Russians’ problems of command, communications, and supply, multiplied by the forests in which they found themselves fighting, broke XIII corps and XV corps into isolated groups of exhausted and hungry men. Samsonov went off into the woods, and killed himself. Russian attempts to fight their way out on 30 August were ill co-ordinated and demoralized. By the 31st the Germans could claim the capture of 92,000 prisoners and nearly 400 guns.

  Rennenkampf, meanwhile, had barely moved, advancing with excessive deliberation, and doing so in the wrong direction. His losses had been heavy. For three days after Gumbinnen his cavalry, still preoccupied with their flanks, failed to push forward to establish the direction of the German retreat. Confident that the 8th army was falling back to the Vistula, Rennenkampf could argue that an energetic pursuit would only push it out of the clutches of the 2nd army, toiling northwards from the Narew. His concern was not with events to the south but with his northern flank, where Königsberg’s fortifications presented a continued threat. Zhilinskii did nothing to suggest that greater urgency was required; his aim was still to hold the 1st army back, while the 2nd army pushed the Germans north. Nor was this particularly surprising, when even Samsonov himself continued to regard the situation as favourable until the morning of 28 August. Thereafter, it was too late for Rennenkampf to save the 2nd army. However, it was not too late to exploit the vulnerability of the German 8th army, exhausted by battle and concentrated to the south. But the actions of the 1st army and the responses of the North-West Front remained dilatory. On the 30th Zhilinskii told Rennenkampf that Samsonov had been completely defeated. Zhilinskii’s preference for security, only reluctantly suppressed in the previous few days, resurrected itself. The position of the 1st army was now seen as one of acute danger, as the Germans were free to concentrate against it. Therefore, rather than seize the opportunity to achieve a victory before the 8th army could redeploy, Zhilinskii approved Rennenkampf’s adoption of a strong defensive position along the rivers Alle and Omet and down to the Masurian lakes. When, on 4 September, he did consider the possibilities of an offensive he was not only too late, he was also the victim of fantasy. The activities of German reserve units round Mlawa led him to believe that the Germans were ignoring Rennenkampf and were already turning against Poland. He therefore instructed the 1st army to attack into East Prussia on 14 September. He somehow assumed that the shattered remnants of the 2nd army would be able to play a supportive role.