But the most ominous German advance was far to the north, where Von Buelow was profiting by the fall of Kovno, marching on Mitau and Riga, and threatening both to cut the railway between Vilna and Petrograd and confine the Russian retreat to congested and narrow lines of communication along which they could not escape. This northern advance was accompanied by a naval offensive in the Baltic, designed to seize Riga and turn the line of the Dvina on which the Russians hoped to stand in the last resort. Fortunately this part of the campaign broke down before matters had reached their worst on land. It looked like a naval operation planned, or at least attempted, by soldiers professionally incapable of grasping the elementary principles of naval or amphibious warfare. After an unsuccessful attack on the southern inlet to the Gulf of Riga on 10 August, the Germans during a thick fog on the 17th sought to land troops at Pernau in large flat-bottomed barges without having secured command of the sea; and the entire landing-force was captured or destroyed. Simultaneously the Russian Fleet engaged the Germans, who had eight destroyers and two cruisers sunk or put out of action; the only Russian vessel lost was an old gunboat. The Dvina lines were not to be turned by strategy like this, and Russia was henceforth free from naval interference until her sailors played her false.

Von Buelow was still, however, to be reckoned with, and he was the substantial danger. On 28 August he began his movement against the Dvina, which would, if successful, cut off all the Russian armies from direct communication with Petrograd. The blow was struck at Friedrichstadt, where the river is crossed by the only practicable road between Riga and Jacobstadt, but the design was to turn the whole front as far as Dvinsk; and Von Buelow held out to his troops the alluring prospect of winter quarters in Riga and a march on Petrograd in the spring. On 3 September the left bank was cleared for some miles, but all attempts to cross were frustrated. The out-march on the extreme German left had failed, and the critical point moved south towards Vilna. The danger here was serious enough, for the depletion of the Russian forces and length of their line had left a gap between Dvinsk and Vilna, and into this gap the Germans thrust a huge cavalry force which more nearly turned the Russian line than any other movement in the campaign.

The way was prepared by the great ten-days' battle of Meiszagola. The unexpectedly rapid fall of Kovno and Grodno had enabled the Germans to threaten the envelopment of Ewarts' army both on the south and the north, on the Niemen towards Mosty and Lida and farther north towards Vilna. The struggle for Vilna was decided at Meiszagola, a village about fifteen miles north-west of the old Lithuanian capital. It was captured on 12 September, and masses of German cavalry swept round from Vilkomir towards Sventsiany and crossed the Petrograd railway to outflank the retreating Russian troops. The evacuation of Vilna began on the 13th, and two days later the menace from the German cavalry became more apparent. Fresh divisions were apparently brought up from Courland with 140 guns; on the 16th they were at Vidzy and on the 17th at Vileika, nearly seventy miles due east of Vilna and in the rear of the Russians escaping thence. They were thus also close to Molodetchno on the railway along which Ewarts was falling back from Skidel, Mosty, and Lida; and control of that junction would have put two Russian armies at their mercy.

Just in time Ruszky was restored to the command of the northern group of Russian armies, and the victor of Rawa Ruska and Prasnysz was not doomed now to break his uniform record of success. The situation was not unlike that at Prasnysz, and it was relieved in a similar way by a Russian counter-offensive from Dvinsk against the flank of the German cavalry. Vidzy was recaptured on the 20th, and farther south the pressure slackened along the Vilna-Vileika railway; Smorgon was retaken by a brilliant bayonet charge on the 21st. The door had been kept from closing on Russian armies seeking to escape from the salient between Lida and Molodetchno, while the Germans were squeezed out of that which they had made to the north. They were driven out of Vileika, and gradually the lines were straightened and stabilized so as to run almost due south from Dvinsk by Postavy, Lake Narotch, and Smorgon. Other factors than Ruszky's brilliant strategy contributed to this dramatic defeat of the final German effort of the campaign to annihilate the Russian forces. The Germans had lost in men and impetus during their long advance. Superb though their organization was, lengthening lines of communication across a country ill-supplied with roads and railways, and the necessity of guarding against a hostile population told upon their armies in the fighting-line. The heaviest blow will spend itself in time against an elusive foe, and the longest arm will find the limit of its reach. The Germans had not planned a march on Moscow, but they had hoped to overrun the Russian armies and occupy the winter quarters of their choice. These were denied them on the Dvina, and they had not secured the coveted Riga-Rovno line.

They were indeed left farther from it in the south than in the north. Their defeat east of Vilna enabled Ewarts to escape from the encirclement threatened by the advance from Kovno and Grodno; and although he had to leave Lida and was subsequently pushed behind the junction of Baranowitchi, thus surrendering to the Germans the control of the railway from Vilna to that point, it remained in Russian hands to Rovno. Mackensen was unable to advance from Pinsk, which he occupied on 16 September, to the railway at Luninetz, while Ivanov reacted successfully against the German attacks along the Kovel-Sarny line and recovered a good deal of the ground lost in the Volhynian triangle and eastern extremity of Galicia. Mackensen's army may have been weakened by calls from the north and from the south for a campaign which was already planned but not yet suspected; at any rate it was too weak to achieve its objectives, the capture of Sarny, Rovno, and Tarnopol, which would have completed the hold of the Germans on the Vilna-Kovno line and given them a base for a farther advance in the spring on Odessa and for the isolation of Rumania. On 7 September, as Mackensen's forces were moving on Rovno and the Sereth at Tarnopol and Trembowla, Ivanov counter-attacked from Rovno and Brussilov and Lechitzky on the Sereth. By the 9th the two latter had captured 17,000 prisoners and a considerable number of guns; and Ivanov followed up this success by retaking Lutsk and Dubno by the 23rd. Kovel was even threatened, but the pressure was not maintained. Sarny, Rovno, and Tarnopol were saved, but Lutsk and Dubno reverted to the Germans, and the line in the east was stabilized with the Volhynian triangle and the railway from Vilna to Rovno divided between the antagonists.

The success of Ruszky in the north and of Ivanov in the south in setting a term to the terrifying sweep of the German advance produced a temporary optimism in Russia comparable with that which followed the victory on the Marne; and in neither case did the Allies realize the extent of the advantage gained by the Germans or foresee the years that would pass before the loss could be recovered. The Grand Duke Nicholas was relieved of his command and sent to take over that in the Caucasus. He was succeeded by the Tsar himself, who was unlikely to interfere with the military measures of Alexeiev, his chief of staff; and the Duma seconded the Tsar's attempt to express the determination of the Russian peoples to withstand the Germans until victory was secured. Nevertheless, the profound effects of the Russian defeat could not be removed by any laudable efforts at keeping up appearances. It was a resounding disaster which condemned Europe to three more years of war, and Russia to a convulsion which would permanently alter the whole course of her history and position in the world. Miliukov raised in the Duma the question of responsible government, and if the debacle of 1915 was slower than Sedan in producing the downfall of the system to which it was due, it was not because the disaster was less, but because Russia was a less organized country than France, and her illiterate population reacted more slowly than the French.

At the moment the best face was put on affairs; and although one correspondent was allowed to report that the heart of the Russian people had grown cold to the Allies who had watched their misfortunes without raising a finger in the shape of a serious offensive to help, public opinion was fed on the comfort in which a facile optimism is so fertile. German casualties were multiplied at will, despondent diaries of individual German officers killed or captured were given unlimited publicity, and roseate pictures were painted of the colossal drain of man-power involved in winter trench-warfare in Russia and in holding vast tracts of hostile country. It was assumed that the Germans would suffer more than the Russians, although again and again whole Russian battalions in those trenches were wiped out by German artillery and machine guns to which the Russians had not the wherewithal to reply except with fresh masses of human flesh; and little was said of the millions of Russian prisoners and civilians who were put to far more effective use in making munitions and producing food for their enemies than they ever had been for Russia or themselves, and without whose labour Germany's man-power would have been exhausted one or two years before the end of the war. It was considered a triumph that the Germans had not reached Petrograd or Moscow, but it might have been well if they had. They had, however, no such ambitions. Just as the reconquest of Galicia had been mainly designed as providing the base for a flank attack upon Russia, so the conquest of Poland was to be used as providing protection for Germany against Russian interference with her plans in the Balkans. Sofia and Constantinople opened up more alluring prospects and a path that led farther than Moscow or Petrograd; and while public opinion in England and France was dreaming of a repetition of 1812, public opinion in Germany was feasting on visions of Cairo, Baghdad, and Teheran, and the possibility of evading the British blockade through outlets to the Indian Ocean.

All eyes that could see were turned to the Dardanelles. There British troops were making the one serious counter-offensive to the German attack on Russia, and success would redeem the Russian failure and foil the hopes the Germans were building upon their victory. The immediate future of the Balkans, the Black Sea, and Asia Minor, and it might be the more distant future of Egypt and the East, hung upon the issue at Gallipoli. During July the reinforcements for which Sir Ian Hamilton had asked were gathering in Egypt and in Gallipoli; and on 6 August the new plan of attack was begun. There were to be four distinct items; a feint was to be made of landing north of Bulair, the attack on Krithia was to be renewed in order to hold the Turkish troops there and draw others in that direction, and a similar advance was planned for the Anzacs with a similar motive, but also to co-operate with the real and fresh offensive. This took the form of a landing at Suvla Bay, the extreme north-westerly point of the peninsula between Anzac and Bulair. The diversions were reasonably successful, as successful, indeed, as previous attacks had been in those localities when they were the principal efforts. The chief of them was a threefold advance north-east, east, and south-east from Anzac Cove on Sari Bair with its highest point at Koja-Chemen. Conspicuous gallantry was shown in the three days' fighting; and while, as earlier at Krithia, the summits defied the greatest valour, enough progress was made in these subsidiary attacks to justify the hope of general success if the principal effort at Suvla Bay went well (see Map, p. 107).

It began without any great mishap, and General Stopford's 9th Corps was successfully landed on the shores of Suvla Bay during the night of 6-7 August and deployed next morning in the plain without serious resistance. The surprise had been effected, but it would be useless unless the attack was pressed with energy and without delay. Yet torpor crept over the enterprise during that torrid afternoon; many of the troops were in action for the first time in their lives, and, understanding that water was obtainable from the lake close by, they had drained their water-bottles by eight o'clock in the morning. A thunderstorm mended matters a little, and Chocolate Hill was carried on the right. But all next day an inferior Turkish force, assisted by a planned or accidental conflagration of the scrub, managed by skilful use of a screen of sharpshooters to hold up our advance all along the line. Sir Ian Hamilton himself arrived that night and strove by persuasion to infuse some energy into the attack. But by the 9th it was already too late, for the Turks had had time to bring up reinforcements, and an attack on the Anafarta ridge on the 10th was repulsed. Five days later General Stopford relinquished the command of the 9th Corps, to which he had been somewhat reluctantly appointed by Lord Kitchener, and the 29th Division was brought up from Cape Helles to renew the attack on 21 August. It might have succeeded had it been originally employed in place of the inexperienced troops; but by this time there could be nothing but a frontal attack on a watchful foe, and it ended like the similar efforts in May and June. Some ground was gained, contact was established with the Anzacs, and a continuous line of six miles was secured from the north of Suvla Bay to the south of Anzac Cove. But before the Turks could be expelled from the peninsula and a passage cleared through the Dardanelles there would be a long and weary struggle, in which progress would be as slow and beset by as many obstacles as it was on the Western front. Russia was to obtain no relief that way; as a counter-offensive to the German campaign of 1915, the attack on the Dardanelles had failed; and the failure produced a deeper impression upon the Balkans than if the attempt had never been made. The way was clear for the next move of German diplomacy and war.