by Hew Strachan
However, Conrad’s mood was not so confident. On the evening of the 25th Brudermann’s 3rd army and Kövess’s group, securing the Dniester before the arrival of the 2nd army (commanded by Böhm-Ermolli) from Serbia, were being pulled north-east, not east, to conform with the movements of the 4th and 1st armies. However, the Russian 3rd army detached only a cavalry division to the north, and kept its main weight moving steadily—but slowly—west-wards towards Lemberg. Ivanov, the Front commander, and Dragomirov, the 3rd army’s chief of staff, were still convinced that the main Austrian concentration lay directly in the 3rd army’s path. On 26 August, Brudermann’s weakened 3rd Austro-Hungarian army, mistaking Russian deliberation for Russian weakness, attacked Ruzskii’s 3rd army at Zlocow and on the River Zlota Lipa. Brudermann’s task was to halt the Russians before they took Lemberg, but his attacks were poorly co-ordinated, and pushed in too quickly for effective artillery support. The Austrians’ losses were heavy, and the Russians—turning from defence to counter-attack—drove them back. Brudermann reckoned that he was threatened on both wings as well as to his front. Conrad’s policy of pushing to the northern flank had weakened his centre, and opened gaps between his armies. One hundred kilometres separated the 3rd army from the 4th, but if he brought the 4th army south-east he would open the distance between it and the 1st. Caught between the prospects of Auffenberg and the threat to Lemberg, Conrad twice recalled the three divisions released from the 3rd army to the 4th and twice gave them back again. Force-to-space ratios also restricted his options to the south: although tempted to bring the 2nd army up from the south to strike the Russian flank, he had to acknowledge that for the moment—with IV corps still extracting itself from Serbia—it lacked the strength to be effective.130
In the event the Austrians had time to take up defensive positions on the next river line, the Gnila Lipa. Ruszki’s pursuit remained cautious. He argued that his units were still dispersed as a result of their precipitate advance, their supply columns trailing in the rear. Ivanov’s deployment was complete by 30 August. He enjoyed a two-to-one superiority, with little more than two corps facing the four corps of Brusilov’s 8th army on the 3rd’s southern flank. Brusilov used his two central corps to pin the Austrians while his VII corps, on the right, crossed the river and outflanked Kövess’s group, driving it south and away from Brudermann’s 3rd. By the evening of the 30th the latter was routed, its troops falling back to Lemberg in disorder. Conrad, though on the verge of at least a local victory in the north, was being overtaken by defeat in the south. On 2 September the 3rd army gave up Lemberg without a fight and fell back 32 kilometres.
Stavka’s new priority was to recover the situation in the north, and above all to ensure that neither the 5th army nor the 4th suffered the same fate as Rennenkampf’s 2nd. On the 30th itself the immediate danger to Plehve was lifted. False reports that Russian troops were advancing to his aid caused the Austrian pincers to look to the safety of their own flanks and rears long enough to open a 32-kilometre gap through which the 5th army could escape. The 9th army in Poland was brought in against Dankl’s left, so supporting the 4th’s right. The gap between the 3rd and the 5th armies, which had opened to three or four days’ march, began to be closed by Stavka’s repeated insistence that Ivanov and Ruzskii put their preoccupation with Lemberg to one side, and push the 3rd army north of the city to Plehve’s aid. The task of occupying Lemberg and covering the Russian left was allocated to Brusilov’s 8th army.
While the Russians inclined to the north on 1 September, the Austrians moved to the south. Thus, at the very stage, after Tannenberg, when the Germans might have been free to push south-westwards into Poland, Conrad himself moved away from a joint envelopment in the centre of the eastern front to another independent Austro-Hungarian undertaking; although Conrad called on OHL to implement the Siedlitz operation four times between 23 August and 1 September, it was far from being AOK’s only and constant preoccupation. Auffenberg was told to abandon his pursuit of Plehve, and to turn on the 3rd Russian army. His immediate task was to save Brudermann’s army and, with it, to forestall the loss of Lemberg. But from the evening of 2 September, as the collapse of the 3rd Austro-Hungarian army became clear, a more ambitious conception gripped Conrad. He noted the slowness and deliberation of the Russian advance. He also, mistakenly, concluded that Plehve’s army was defeated. His cavalry was too broken to lift the veil over the Russian movements. What Conrad now aspired to was nothing less than the envelopment of the Russian 3rd and 8th armies. The entry of his 2nd army from Serbia, the continued resistance of Halicz on Brusilov’s left flank (it did not fall till 3 September), created the opportunity to attack the Russian rear from the south. Auffenberg was to take the rear of the 3rd Russian army from the north. As before, and as would happen again, Conrad’s imagination outstripped the dictates of common sense: the task of envelopment which he now set his troops took no account of their numbers or of the punishing schedule of battles and marches which they had already undergone. The 4th army could only engage its divisions seriatim. The 2nd army, its attacks uncoordinated and unsupported, made little impact on Brusilov; the 3rd—the body around which the movements of the 2nd and 4th armies pivoted—was already broken. The Russians in the south, although no longer enjoying a numerical superiority, were not advancing into a sack, as Conrad might hope, but into a paper bag with the bottom blown out.
Brusilov’s 8th army, on the line of lakes at Grodek, west of Lemberg, and the 3rd army to the north at Rava Ruska accomplished the tasks which they had been set. Attacking despite their numerical inferiority, they fixed the Austrian 2nd and 3rd armies from 6 September. Auffenberg’s 4th army, going south, met Ruzskii’s 3rd going north. The latter, aided by two corps of Plehve’s undefeated army, outflanked the former to the north-east. On 8 September Conrad ordered Auffenberg to hold on the front Rava Ruska-Janow, while attacking with the 2nd and 3rd armies at and to the south of Lemberg. But the 3rd army was now being pulled into the 4th’s battle on its left. Conrad hoped that the 2nd army could redeem the situation, relying on his greater strength in the south to roll the Russians up to the north. But the Russian 8th army held. On 9 September Dankl’s 1st army to the north, left isolated against stronger Russian forces, and with a large gap between its right and the main Austrian forces, was already falling back to the San. Plehve’s 5th army exploited the widening gap between Auffenberg and Dankl. As the Austrians fought to stabilize their position west of Lemberg, their weakness to the north exposed the 4th army’s rear. On 11 September Auffenberg, the full dimensions of the Russian manoeuvre revealed to him by wireless intercepts, pulled back to the south-east. Conrad had no choice but to order a retreat to the Dniester, and then to the San.
The defeat suffered by the Austro-Hungarian army in Galicia ought to have been shattering. Its total losses were about 350,000 men; some divisions were reduced to a third of their combatant strength. Particularly grievous in the long run were the disproportionately high casualties among junior officers and professional NCOs. Those that remained were exhausted: Auffenberg’s 4th army had been in the front line for twenty-one days and had been engaged in combat for eighteen of them. When, on 10 September, AOK drafted its by-now customary appeal to the Germans, its request was not that they manoeuvre on Siedlitz, but that they send troops straight from France to Galicia.131
The message was never sent. Pride intervened, but so too did the realization that the army was still intact. Its retreat was chaotic. After the surrender of Lemberg supply dumps were set on fire so that nothing but charred rubbish remained for the remants of the 3rd army. Its transport collapsed. One thousand locomotives and 15,000 wagons were abandoned.132 The roads, turned to mud by heavy rain, were crowded with refugees. Older reservists succumbed to their lack of fitness. But Conrad asserted himself. Brudermann, sacked on 5 September, was only the most distinguished of those purged. The 3rd army’s chief of staff also went, as did the corps commanders of the 4th army. Franz Joseph endeavoured to protect the re
putations of his generals, demanding a careful inquiry into each case. His efforts, a vain attempt to reflect the monarch’s authority over his army, were fruitless: by the year’s end four of the original six Austro-Hungarian army commanders were dismissed. Conrad may have been covering his own back; he acknowledged to his staff that Franz Ferdinand, had he been alive, would have had him shot.133 However, his justification was that, in the succeeding days, both tactically and operationally the Austrians responded with good sense. They were reading Russian wireless traffic with a maximum delay of three days, and so could offset Russian numbers with Austrian anticipation.134 The Russian use of the defensive and of field fortifications, learnt in Manchuria and applied at the Zlota Lipa and on the Gnila Lipa, was emulated by the Habsburg army.135 From mid-September the Austrians too began to entrench. And, as at Mau beuge in the west, so at Przemysl in the east major fortifications were proved to have greater worth than the pre-war advocates of manoeuvre warfare had allowed. As the Russians advanced, so the line of the Carpathians, running south-east to north-west, shouldered their 8th and 3rd armies in the same direction. Due west of Lemberg, guarding the north-eastern approaches to the Carpathians and shielding Cracow to the west and Hungary to the south, was the railway centre and fortified city of Przemysl, on the River San. Conrad resisted the suggestion, made especially by Boroevic, Brudermann’s successor in command of the 3rd army, that it be abandoned. Its defences had not been fully modernized before the war, and its resistance had been planned to last a maximum of three months. But with its selection as AOK’s headquarters at the outset of the campaign, 27,000 workers had been brought in to create seven new defensive belts with twenty-four strong-points, to dig 50 kilometres of trenches, and to install 1,000 kilometres of barbed wire. Fields of fire for 1,000 guns were cleared by the felling of 1,000 hectares of woodland and the razing of twenty-one villages. From 21 September Przemysl was under Russian siege, and was to remain so, except for the period 3 October to 8 November, until 22 March 1915.136 Its defence sucked in the Russians and covered the Austrian recovery.
Brusilov, impatient with his neighbour, argued that if the 3rd army had advanced with unwonted rapidity it could have seized Przemysl. That it did not do so may have been a failure of command, but it was also a product of the very logistic constraints which now made it essential that Przemysl be stormed rather than screened. The Russians had deliberately not developed the railways and roads out of Poland into Galicia; the assumptions underpinning strategic communications, in Galicia as in East Prussia, proved to have been too defensive. Thus, the South-West Front now had more troops in Galicia than it could feed or move. By October the roads were at times waist-deep in mud.137 The problem was exacerbated by the change in railway gauge, that of Austria-Hungary being narrower than that of Russia, and by the destruction effected by the retreating Austrians. The Russian armies were being stopped by their own weight. They paused to consolidate their conquests. The dual monarchy could enjoy a brief respite.
POLAND
By mid-September 1914 the military prestige of Austria-Hungary had reached its nadir. The attempt to restore its authority in the Balkans by battle had failed; instead, Serbia had cocked a snook at its mighty neighbour, and Russia stood poised at the gates to Hungary. Conrad blamed his ally for his defeats. He claimed that the Germans had reneged on a written agreement that the two powers would mount the Siedlitz operation; in reality no such document existed. OHL responded by suggesting that Austria-Hungary make territorial concessions to Romania; an active ally to the south could transform the outlook on the Galician front. Conrad’s riposte was to attribute Romania’s neutrality to Britain’s entry to the war, itself the result of Germany’s provocation. Thus, a squabble over operational matters assumed a political complexion. Indirectly the dual monarchy threatened to conclude a separate peace.138
On 31 August OHL had already accepted that a thrust into Poland would follow the clearing of East Prussia. At this stage operations in Galicia still seemed evenly balanced. By mid-September, when the defeat of the Austro-Hungarians was patent, East Prussia was secure. Now it was Silesia, its Galician flank-guard in Russian hands, that was confronted with the prospect of invasion. As was the case further north, an attack seemed to be the best defence: by taking the initiative the Germans in the east could compensate for their lack of numbers and buy time for their armies in the west.139 Self-interest, as much as alliance obligations, directed Germany’s immediate attention to Poland and the aid of Austria-Hungary.
Hindenburg’s intention was to cross the Narew and march in the direction of Siedlitz. In mid-September, therefore, it was the Germans, not the Austrians, that were lured by the prospect of a great envelopment battle in Poland. Freed from worries for the security of East Prussia by the victory of the Masurian lakes, Hindenburg spoke to OHL on 13 September about an advance in ‘a decisive direction’;140 not simply aid to Austria-Hungary, but a wider success became his objective. However, Conrad argued that it was now too late to implement the Siedlitz manoeuvre. His army was no longer advancing to the north-east but was being driven back on Cracow. His talk was not of attack but of further retreat. By the time a German thrust in Poland took effect against the Russian South-West Front the latter might already have completed its demolition of the Austro-Hungarian army. OHL was convinced by Conrad. It told Hindenburg that direct support for Conrad was politically essential, whatever Hindenburg imagined might be militarily desirable. What the Austrians wanted was German troops in Upper Silesia and around Cracow in order to meet the Russian advance, which Conrad (correctly at this stage) imagined would be directed westwards, rather than southwards towards Hungary. Conrad hoped in this way to give his army the opportunity to recuperate. When that was done, it might then resume the offensive.
Two totally distinct concepts for operations in Poland were thus being formed. Moltke, confronted with a strong-willed army commander, and preoccupied with the German recovery from the Marne, lacked the authority to impose unity. Hindenburg argued that the railway movements to effect the redeployment desired by the Austrians would take twenty days. Given that the Austrian need was for speed, the Germans should push from a position north of the Austrians, due east towards the Vistula. The proposal, whose origin reflected the need to compromise on the Siedlitz operation, counterbalanced by a refusal to meet Conrad’s demands, was easy to rationalize. If Ivanov continued to push Conrad westwards towards Cracow, the Germans would fall on his flank and rear. If he did not, the Germans would protect the Austrians on their unprotected flank, on the north or left bank of the Vistula. The Germans would then push north and east towards the line Warsaw-Ivangorod, and so fall on the South-West Front’s flank and rear. The proposal, in the hands of Hindenburg and Ludendorff, was not for the Germans to extend to meet the Austrians, but for the Austrians to reach out towards the Germans. Ludendorff told Conrad that the Austrians were in danger of being boxed in, wedged in a narrow space along the Visloka, the Vistula on their left and the Carpathians on their right. To avoid this, AOK should put Dankl’s 1st army across the Vistula to link up with the Germans. Thus, a limited manoeuvre designed to relieve the pressure on Austria-Hungary was becoming a grandiose scheme. By 22 September Hindenburg could tell the Austrians that ‘a greater success could be achieved against the Russians through envelopment’.141 Once again, a refusal by two independent commands to compromise was being dressed up as an envelopment operation. Not one plan but two were to be implemented. The Germans’ basic notion was still guided by the Siedlitz idea, and emphasized speed in the attack; the Austrians’ concern was to keep their army intact, and saw offence as a later stage. Neither command formed its plans with any accurate information as to Russian intentions. Conrad imagined that their main weight faced his armies. Ludendorff placed at least some forces further north on the middle Vistula, but still concentrated in southern Poland and not above Ivangorod.
The need to develop a German operation independently of the Austro-Hungarian was a product, not
only of personal pride but also of national prestige. If, instead of a German attack from the west and an Austrian from the south, a combined operation had been maintained in the Cracow area, the case for a joint command would have been irrefutable. The balance of forces was Austrian; moreover, although Hindenburg was the victor of Tannenberg and Conrad the loser of Lemberg, the former had been in retirement only a month previously. Therefore, such a command would have had to be held by Conrad. But the subordination of Hindenburg to Conrad was not only unacceptable to Germany’s de facto leadership of the alliance, it was also incompatible with Hindenburg’s and (particularly) Ludendorff’s self-image.