For a time he had the luck which lures the gambler on, and the scene of his second act was skilfully chosen. Before 21 March Haig had kept his line better manned north than south of Arras, and the reasons which made him anxious for the defence of his northern sector counselled Ludendorff to attack it when the defeat of the Fifth Army had compelled the British commander to divert ten divisions from the north and supply their place with the weary survivors of the battle of St. Quentin. He had little room to spare between his front and the sea, and a break-through, far less extensive than that which had been effected in March, would give the Germans the coast of the Straits of Dover, enable them to bombard the Kentish sh.o.r.e, hamper the port of London, and perhaps reach it with long-range guns like those with which they had occasionally bombarded Paris since 23 March. These annoyances would have been serious; but the British public paid itself a very bad compliment when it seemed to a.s.sume that the distant bombardment of London would have an effect upon the war disproportionate to that of Paris; and the notion that an impetus which carried the Germans to Calais would transport them across the Channel was merely another ill.u.s.tration of the comprehensive popular ignorance of the meaning of sea-power. Nieuport or Dunkirk might have taken the place of Bruges as a submarine base without greatly enhancing the success of that campaign; and Haig chose rightly when, having to weaken his northern front, he risked a sector north instead of south of La Ba.s.see and the Vimy Ridge. Defeat to the north of those points, even though it cost us the coast as far as Calais, would not entail retreat from the Artois hills between Arras and Gris Nez or threaten our liaison with the French which had been Ludendorff's first objective. The material comments on the value of his second thoughts were that the Germans might have had the Channel ports for the asking in 1914 but did not think them worth it, and that in April 1918 Ludendorff employed but nine divisions in his initial effort to break through. Probably his real ambition was merely to shorten his line and, in view of the possible resumption of his offensive in front of Amiens, to provide against a British counter-attack on the sensitive German position along the Belgian coast.

Antic.i.p.ating some such attack, Haig had deemed it wise to relieve the two Portuguese divisions which held part of the front between the Lys and La Ba.s.see of their arduous responsibility; but he could only replace them by weary British divisions, and the change had only been half effected when, on 9 April, Ludendorff s attack began after the usual bombardment with gas and high-explosive on the 8th. The Portuguese broke fairly soon, the British flanks on either side were turned, and the whole centre had gone in a few hours. By night the Germans had captured Fleurbaix, Laventie, Neuve Chapelle, Richebourg, and Lacouture, and were on the Lys from Bac St. Maur almost as far as Lestrem. But the key-position at Givenchy was splendidly held by the 55th Division, which set a permanent limit to the German success and prevented it from obtaining anything like the dimensions of the March offensive. It continued, however, to develop on the north. On the 10th Bois Grenier fell, Armentieres was evacuated, and the Germans poured across the Lys, taking Estaires, Steenwerck, and Ploegstreet and threatening the Messines ridge. That, too, followed on the 11th, while farther south the Germans secured Neuf Berquin and Lestrem. On the 12th they got into Merris and Merville and advanced to the La Ba.s.see ca.n.a.l, threatening to cross it and outflank Bethune on the north-west. Here, however, they were held up in front of Robecq, between the ca.n.a.l and the forest of Nieppe, and turned to exploit their advantage farther north. Their advance here was slower, but by the 16th they had mastered Wytschaete, Wulverghem, Neuve Eglise, Bailleul, and Meteren, and were facing the line of hills running from Mt. Kemmel to the Mt. des Cats.

British and French reinforcements were now arriving in considerable numbers, and Ludendorff would have been prudent to rest on his laurels. He had made a p.r.o.nounced bulge in our line, had diverted forces from other sectors of the defence, and compelled us to evacuate our dearly-purchased gains of the Flanders campaign in the preceding autumn. On the other hand, he had lengthened instead of shortening his own line, he had achieved no strategical object, and his troops were left in a salient which invited attack. Unless he could win the heights from Mt. Kemmel to Mt. des Cats, which commanded the country to the coast, he would be in a worse situation for defence than he was before. He was thus driven to prolong the effort, pour fresh divisions into the battle, and convert a diversion into a major operation. Doubtless popular visions of the Channel ports and the bombardment of London reinforced the sounder military reasons for persistence. There were three obvious lines of attack-on the Belgian front north of Ypres, on the Kemmel range, now held partly by French troops, and on Bethune. The first was defeated on the 17th by a brilliant Belgian resistance, and the third was repulsed on the 18th before Hinges and at Givenchy; but the second was longer delayed and more stubbornly pushed.

The effort began with an intense bombardment on the 25th, and a few hours later the Germans had captured the village and hill of Kemmel; our forces were driven back to a line running in front of d.i.c.kebusch lake, La Clytte, the Scherpenberg, and Locre. Mt. Kemmel had been regarded as the key to the position, and it looked as though the range would fall. But Kemmel was an isolated height, and the Germans were beaten in the valleys which separated it from the Scherpenberg. Their attacks reached a climax on the 29th, and after some partial success were everywhere defeated. Local fighting continued spasmodically till late in May, but it was clear that Ludendorff's second offensive had come to an end like his first. Its extension had also ruined the chance of successfully resuming the attack in front of Amiens. On 23 April the Germans attacked just south of the Somme and captured Villers-Bretonneux, but it was promptly retaken on the following day; and in the struggle along that line in May we advanced as well as improved our position. The Germans had fought their last offensive against the British front and had failed; and when after a four weeks' pause they resumed their attacks, they were directed against the French.

During the interval the British public had time to reflect upon the disaster and its effects. They were brought home by a new military service Bill extending the liability to all men under fifty-one and bringing Ireland within its scope. Panic had as much to do with these proposals as forethought. The raising of the military age was calculated to weaken our industrial more than to strengthen our military power; and the extension to Ireland handed that country over to Sinn Fein and necessitated the diversion thither of large British forces, which might otherwise have been sent to the front, without producing a single Irish conscript. The proposal was, indeed, so foolish that its authors made no attempt to carry it out. Wiser was the speedy dispatch to France of 300,000 superfluous troops who had been kept in England by nothing better than an ignorant fear of invasion. But it was the amazing rapidity with which the United States responded to Mr. Lloyd George's anxious appeals that saved the Government from the effects of its own blunders and reduced its military service Act to a measure for the infliction of gratuitous hardship. In April nearly 120,000 American troops landed in Europe, over 220,000 in May, and 275,000 in June. On 2 July President Wilson announced that over a million had sailed; that number was doubled before the summer ended, and in July General s.m.u.ts was antic.i.p.ating the possible presence in France of an American army as large as the British and French combined.

The need for so colossal a force did not arise, but in April the position of his Government as well as the military situation agitated the Prime Minister and gave wildness to his words as well as to his actions. Apart from the casualties, we had lost 1000 guns, 4000 machine guns, 200,000 rifles, 70,000 tons of ammunition, and 250 million rounds of small ammunition, and 200 tanks. Circ.u.mstances wore a different complexion from the roseate hues of the early months of 1917, and Mr. Lloyd George could not escape the kind of blame he had heaped upon his predecessors. He sought to evade it in his speech at the rea.s.sembling of Parliament on the 9th by shifting the responsibility for the disaster partly on to M. Clemenceau as the princ.i.p.al author of the unfortunate extension of the British line, and partly on to the commander of the Fifth Army. The latter at least could not reply, and the unfairness of the attack provoked much ill-feeling in the army and elsewhere; it found expression in a letter from Major-General Maurice, lately Director of Military Operations, which was published on 7 May and challenged the accuracy of ministerial statements. His charges were so serious that the Government at once proposed a judicial inquiry. Mr. Asquith committed the tactical error of moving instead for a parliamentary committee. The Government naturally treated his motion as a vote of censure, and escaped all investigation on the ostensible plea that it preferred a different method from that proposed by Mr. Asquith. The House of Commons by 293 to 106 votes expressed its apparent satisfaction with that "ex parte statement from the Prime Minister himself" which "The Times"-then his strongest supporter in the Press-had the day before said could not dispose of a charge which "unless and until it is impartially investigated and disproved, will profoundly shake the public confidence in every statement made from the Treasury Bench." It was not, however, with the honour of ministers that the House was mainly concerned. Members were in that mood, which occurs at times in every nation's history, in which questions of morals seem irrelevant or unimportant; and what they wanted was not the truth but a plausible excuse for shirking inquiry and refusing to add a political to the military crisis. Conscious of their own responsibility for the Government, they were impatient of any discussion which might reveal unpleasant facts to their const.i.tuents or military information to the enemy.

It is difficult also not to trace a political motive, if not in the attacks on Zeebrugge and Ostend, at least in the contrast between the enormous publicity they received and the silence which shrouded the more normal but not less important or heroic work of the British Navy. The plans, indeed, had been prepared and sanctioned by Jellicoe before he left office some months earlier; but many plans have long to wait the ministerial word, and the naval operations of 23 April were as timely for political as for military reasons. The military objective was to block the submarine and destroyer exits from Zeebrugge and Ostend, both of which were connected by ca.n.a.ls with Bruges; and an operation of that kind against the elaborately fortified Belgian coast required favourable weather conditions as well as the highest courage. The plan at Ostend was simply to sink ships in the waterway; at Zeebrugge there were also to be diversions in the form of a landing on the protecting mole and the blowing up of the viaduct which connected it with the sh.o.r.e. Success was only possible if mist and smoke-clouds added to the concealment of night, and those conditions depended upon the wind. They seemed favourable on the night of 22-23 April, but a quarter of an hour before the Vindictive reached the mole, a south-west breeze dispersed the smoke-clouds and precipitated a torrent of sh.e.l.l-fire from the German batteries. In spite of it the landing party got on to the mole and systematically destroyed its works, while a submarine loaded with explosives was run under the viaduct and exploded. Meanwhile, the blocking ships were sunk at the mouth of the ca.n.a.l, and the survivors of their crews were picked up and got away in the Vindictive and her consorts. At Ostend the blocking ships had to sink outside the centre of the waterway; but the effort was repeated with better success by the Vindictive on the night of 9-10 May. Even Count Reventlow described these affairs as "d.a.m.ned plucky," but added that they were nothing more. The further attacks on the Belgian coast which were commonly expected did not come, and the operations had no appreciable effect upon the land campaign. But they hampered the German submarine campaign to some extent; and if they demonstrated once more that sea-power is limited to the sea, they also showed that on the sea German power had become a negligible quant.i.ty. That fact was, indeed, being proved in a more effective though less heroic fashion, by the safe transport of hundreds of thousands of American troops across the Atlantic; but possibly public opinion needed the more spectacular demonstration, and it certainly showed that the spirit of British seamen was unaffected by the tremors of politicians.

Politicians appeared, indeed, to be more nervous after the crisis had pa.s.sed than they were before it arose, although their alarms did not greatly affect the incurable sang-froid of the British public; and the way in which the middle-aged shouldered the unnecessary burdens imposed upon them by the improvidence of their Government, was as exemplary as the eagerness with which youth had volunteered early in the war. Their acceptance of the new obligations had its value in stimulating America to dispatch her hundreds of thousands of troops more fit for active service; and few, if any, of the elderly English recruits saw any fighting. Ludendorff's plans had already gone astray when he failed in March and April to break the liaison between the French and British armies; and his subsequent operations were ineffective attempts to prepare the ground for a final offensive which he was never allowed to begin. It would have been doomed to miscarry in any case, for his preliminaries exhausted the forces intended for the final effort, and the battles in Flanders had enhanced the failure of his original design. He took four weeks to prepare for a second subsidiary operation, and hoped to achieve a better success against the French than he had against the British. He had the advantage of taking them unawares, and on the eve of his offensive a French journal proclaimed that it would be another blow at the British front because the Germans knew that the French line was impregnable. Popular opinion in France had attributed German success at St. Quentin and in Flanders to British incompetence or cowardice, and British troops had even been hissed in the streets of Paris. The attack on the Chemin des Dames was to modify this opinion, although some tactless Frenchmen announced that reserves sent up to the British sector, which alone stood its ground, were going "au secours des Anglais."

Ludendorff's object was to widen his front towards Paris, for the lure of the capital had already diverted him from his original plan of breaking the liaison between the French and British armies in front of Amiens. That Paris was his objective in May, and not the diversion of troops from the critical junction with a view to resuming that attack, seems clear from the fact that his next blow in June was struck between Montdidier and Noyon. The Chemin des Dames would have been impregnable if properly held, but Ludendorff s information was not at fault, and the possession of the interior lines gave him the same advantage as in March of striking either right against the British or left against the French. He struck early on 27 May and achieved the most rapid advance of the war on the Western front. The line from Soissons to Reims was held by only eight divisions, four French and four British-one of these in reserve-and in a few hours the French had lost all their gains since October 1914 and were back again behind the Aisne. The British divisions, although they had been sent there to recruit after their hard work in March and April, made a better fight, and maintained themselves in their second positions all the day. But the French retreat had uncovered the British left flank, and in the evening they had to withdraw to the Aisne. By that time the French were nearer the Vesle than the Aisne, and on the 28th they were driven well south of the latter river. On the 29th the Germans broadened their front by taking Soissons, and on the 30th the apex of the salient they had made had reached the Marne between Chateau-Thierry and Dormans. For three days they had advanced at the rate of ten miles a day, capturing some 40,000 prisoners and 400 guns. From that date the pace slackened. The Germans did not attempt to cross the Marne, but endeavoured to widen their salient by pushing east behind Reims and west across the Soissons-Chateau-Thierry road. They had little success in the former direction, but in the latter they gradually pressed back the French to an irregular line which ran from Fontenoy on the Aisne southwards along the Savieres river across the Ourcq, and then turned eastwards down the Clignon and reached the Marne below Chateau-Thierry. American troops, who had on the 27th marked their advent into battle by capturing and holding Cantigny, a critical point on the Montdidier front, now took up an equally crucial position south-west of Chateau-Therry and drove the Germans back on 4-5 June, while on the 6th British troops recaptured Bligny south-west of Reims.

The French themselves defeated on the 5th a German attempt to cross the Oise at Lagache south of Noyon, which was intended to link up the German offensive on the Aisne with their next attack farther west. This was launched on the 9th between Montdidier and Noyon, and its purpose was to push southwards and envelop the French defences and forces in the forests of Compiegne and Villers-Cotterets which had stopped the German westward advance on Paris between the Aisne and the Marne. It was a dangerous threat, but this time Foch was prepared. The attack was, indeed, a matter of common antic.i.p.ation, and its adoption suggested that Ludendorff was getting to the end of his expedients. The Americans at Cantigny set a western, and the French success at Lagache an eastern limit to its front; and thus confined it advanced no more than six miles in four days. The French left stood firm and a brilliant counter-attack by Mangin on the German right flank between Rubescourt and St. Maur on the 11th determined its failure, although Foch was compelled to evacuate the salient which the German advance had created in the French line east of the Oise between Ribecourt and Mt. de Choisy. Hoping that this attack had diverted French forces from the defence of the forest of Villers-Cotterets, the Germans then renewed their push along the Aisne, but were promptly checked; and no better success attended their effort on the 18th to encircle Reims still farther east.

For the moment German trust in success had to repose upon the secondary efforts of her Austrian ally on the Piave, although no German troops could now be spared to give much substance to the expectation. That front had been quiescent since the winter, but a good deal had been done to strengthen it, and the Italians were doubtless well advised to stand behind their lines rather than risk an offensive until Austria was practically hors de combat. Austria herself had little stomach for the fight. Her domestic situation was deplorable; parliamentary government had been suspended; and nearly half the population of the Empire was in veiled or open revolt. Hundreds of thousands of Czecho-Slovaks and Jugo-Slavs had joined the enemy, and some were stiffening the Piave front. But German demands were inexorable, and it was hoped that German tactics would supply the place of German troops. There were two battles in the offensive which began on 15 June, one in the mountains, the object of which was to turn the whole Italian front on the Piave, and the other a frontal attack across that river between the Montello, the pivot of the mountain and river fronts, and the sea. The first was the more promising, but achieved the less success. That front was partly held by French and British troops, and the insignificant advance which the Austrians made on the 15th was stopped on the following day. The attack on the Piave was at first more fortunate; a good deal of the Montello was taken, a serious impression was made on the Italian right wing at San Dona di Piave, and fourteen new bridges and nearly 100,000 Austrian troops were thrown across the river. Fortune came to the rescue of the Italians, and torrents of rain flooded the Piave and broke ten of the Austrian bridges. On the 18th the counter-attack began, and by a brilliant dash of soldiers and sailors the Austrian left was turned on the 21st. On the 22nd a general retreat across the river was ordered; it was skilfully carried out, and the Austrians escaped with singularly slight losses considering their precarious position. Their offensive had been an utter failure, but Diaz did not think it prudent to follow up his success with an advance across the river.

The Austrian misadventure was a meagre morsel with which to fill the gap between the latest German offensives and the crowning mercy for which the German public had been led to look; and as the precious summer weeks flew by uneasiness must have filled any German minds that were capable of discerning the realities of the situation. But the wish is father to most men's thoughts, and unpleasant facts which were not concealed by the censor were sedulously ignored or explained away. "Foch's reserves" became a jesting synonym on German lips for something which did not exist, and it was the daily exercise of journalistic wisdom to show that American armies which could not swim or fly would be prevented by German submarines from crossing the Atlantic. Ludendorff was not so blind, and had he been a patriotic statesman instead of a Junker general he would have sought to make terms while he might do so with advantage. But it is the nemesis of militarism that it never can make a peace which is tolerable to its enemies, and Ludendorff had no choice but to persist with an offensive which had become a desperate gamble. His efforts since the end of May had profited him little; he had used up most of the divisions intended for a final resumption of his attack on the Franco-British liaison; and after more than a month's delay he could only launch his last bolt against an eccentric and subsidiary objective. Foiled in front of Amiens and Paris, he turned to Reims; but there was nothing in the previous history of the war on the Western front to suggest that, even were his last offensive as successful as his first, it would bring him within measurable distance of the victory he needed. The Marne might be crossed and the railway to Nancy and Verdun cut, as they had been in 1914, but the further advance for which he could hope from his attack on Reims would bring him no nearer to Paris, to breaking the Entente connexion, or to damming that fatal flow of American reinforcements which was providing Foch with as many reserves a month as Germany could recruit in a year.

The fateful attack began at 4 a.m. on 15 July after four hours of artillery preparation. Its object was to encircle the Montagne de Reims, the chief bastion of the line of communications between Paris and the eastern front on the Meuse, and to extend the German hold on the Marne from Dormans as far as Chalons. There were two converging attacks, one on the twenty-six miles of front which Gouraud held east of Reims between Prunay and Ma.s.siges, and the other on a twenty-two mile line south-west of Reims between Vrigny and Fossoy on the Marne above Chateau-Thierry. For each attack Ludendorff used fifteen divisions, with others in reserve. On both fronts he found Foch prepared to counter the tactics which had been so successful in the earlier stages of the offensive. The first line was lightly held, and the Germans were shaken by a skilfully arranged bombardment as they crossed the zone between it and the real French defences. Upon these in Champagne they made no impression whatever. Prunay, Prosnes, Auberive, and Tahure were yielded at first, but recovered by counter-attacks; the French lost no guns, and their casualties were insignificant. Gouraud more than anyone else had frustrated Ludendorff's last offensive. South-west of Reims the Germans were rather more successful. They pushed across the Marne to a depth of some three miles between Mezy and Dormans, and in three days advanced up it past Chatillon towards epernay as far as Rueil. Similar progress was made eastwards on the line between the Marne and Vrigny. But the gate-posts were firmly held at Fossoy with American a.s.sistance, and at Vrigny with that of the British and Italian divisions which under Berthelot did some of their best fighting in the war. By the evening of the 17th the Entente forces were successfully counter-attacking all along the line, and at dawn on the 18th Foch delivered the blow which converted the German advance into a retreat, and began a triumphal progress which did not stop until four months later the enemy sued for peace.

CHAPTER XIX

THE VICTORY OF THE ENTENTE

There were a few people in England who had some inkling on 18 July that it might prove a turning-point in history. Foch's simple piety had led him into what was almost an indiscretion; he had asked for the special prayers of the faithful, the request had spread to conventual schools in England, and by the 16th it was guessed by those who knew the fact that a special effort was in contemplation. But his great counter-attack owed its importance to what had gone before and what was to follow; and victory was due to more complex and comprehensive causes than the valour of the troops engaged upon the Marne or even the strategy of Foch. Greater efforts were made at other times on both sides than during the last fortnight of July 1918, and the destruction of the salient the Germans had made since 27 May was merely the last ounce which turned the balance of power and the scales of victory. There were many ounces in the total weight, and the pride of each belligerent points to the different contributions which it made. To the Americans their divisions at Chateau-Thierry seem the decisive factor, to the French it was Foch's genius. The British point to the fact that the greatest weight of German force was still in front of Amiens and not on the Marne, and an Italian prince has declared that it was Italy who won the war on 24 October; while Ludendorff has maintained that American troops counted for little, and that the crucial factor was the revolutionary propaganda which had begun to undermine the moral of German troops as early as 1916. None of these partial explanations contain more than an element of truth, and a more comprehensive view is suggested by the likeness of Germany to the "one-hoss shay" of Oliver Wendell Holmes' ballad, a vehicle so skilfully compacted of durable materials that each part lasted exactly as long as every other, and that the whole eventually crumbled into a heap of dust in a single moment. German resources were vastly inferior to those which were slowly mobilized against her, but she organized them with such skill that they resisted the wear and tear of the war for a period to which some observers could discern no end. The strength of materials is, however, limited, and no organization can make them last for ever. The German armies began to give on 18 July, and the decay went on increasing because she had not the means with which to make repairs. The wonder is not that the machine broke down, but that it bore so great a strain for so prolonged a time. The Germans could not command success because they defied the conscience of mankind, but from the military point of view they certainly deserved it.

In spite of Ludendorff's attempt, natural in a Junker, to debit revolution with his failure, it was American reinforcements which turned the scale. Few of them were as yet in the battle line, and there was no great disparity between the opposing forces on the front. But the mobilized strength of the Allies was growing to three times that of their enemies. Foch had an inexhaustible reservoir which enabled him to take risks which Ludendorff could not afford, and gave him a freedom of action which no Entente general had yet possessed. The extent of his command and his resources released him from the bonds of limited offensives. He could crush the German salient on the Marne without prejudicing the prospects of his plans at Amiens and Arras, in Champagne or at Verdun; and fear imposed on Ludendorff the dire alternative of weakening his powers of resistance to future attacks elsewhere, or starving his immediate defence. His plans for resuming the offensive at Amiens had already been ruined by the drain of his attacks on the Aisne and on the Marne; and his defensive prospects on the Amiens front were now to be jeopardized in the effort to avoid disaster in the salient he had rashly made along the Marne. For, except on the a.s.sumption that Foch was unable to attack on the western flank of that salient between Soissons and Chateau-Thierry, the German thrust deeper across the Marne was a wild adventure (see Map, p. 362).

Foch, however, had made his plan and his preparations. Concealed by the forests of Compiegne and Villers-Cotterets, he had a.s.sembled in the angle between the Oise and Marne reserves of which the Germans denied the existence. From the Aisne near Fontenoy southwards to the Ourcq Mangin commanded an army containing the pick of French colonial troops; and thence to the Marne Degoutte had another which included five American divisions. Before them ran the German flank weakly guarding the line of communications with the German front on the Marne. Led by a vast fleet of French "mosquito" Tanks something like the British "whippets," the French early on the 18th broke through the German defences on a front of twenty-seven miles and advanced from two to five miles towards the Soissons-Chateau-Thierry road. [Footnote: An error made in the British rechauffee of the French official news represented Mangin as having advanced eight miles on the 18th to the Crise on a stretch of five miles east of Buzancy. It was a mistake of nord-ouest for nord-est which was never corrected, and has got into most of the summaries and histories of the war, although it makes the subsequent French fighting in that area unintelligible. The history of the German evacuation of the salient would have been very different had the French got east of Buzancy on the 18th. As a matter of fact, it took them eleven days to secure the territory credited to them by this error on the 18th.] Mangin reached the Montagne de Paris within two miles of Soissons, and Berzy-le-Sec on the banks of the Crise, while south of the Ourcq Degoutte got to Neuilly St. Front and the Americans captured Courchamps, Torcy, and Belleau. Sixteen thousand prisoners and fifty guns were captured, but there was nothing like a German rout. They stubbornly defended their main line of communications for days until the bulk of their forces could get away; and they evacuated the salient slowly and in good order. There was, of course, no further hope for them south of the Marne, and by the 20th they had regained the northern bank without very serious loss; it was not till the 22nd that the Allies crossed the river in pursuit. On the 21st the Germans had abandoned Chateau-Thierry and the Soissons road as far as the Ourcq, but north of that river they held the road for a week, and Buzancy was not captured till the 29th. By the 23rd Berthelot was making progress on the other side of the salient, and the German centre had to relinquish the forest of Fere and Oulchy on the 25th. On the 31st the Americans drove in their centre at Seringes, and on 2 August the French forced their way into Soissons. By the 3rd the Germans had been driven across the Vesle and the salient had been flattened out.

Even the best of the critics in the French press had little idea of what was to follow. The Germans' latest offensive had been foiled, and they had lost the more adventurous part of their gains in May; but Foch's success was regarded as merely a promising detail, and men discussed the locality of Rupprecht's counter-attack. But the signs of the times did not point in that direction. On 4 July Americans and Australians fighting side by side had captured Hamel below the Somme. On the 19th the British had recaptured Meteren at the apex of the German salient across the Lys, and Merris fell on the 30th. On the 23rd the French between Amiens and Montdidier had advanced two miles on a four-mile front and captured Mailly-Raineval, Sauvillers, and Aubvillers in the Avre valley; and on 4 August the Germans withdrew from all their ground to the west of that river. Two days later they attacked and recovered some of the ground they had recently lost near Morlancourt. Both the withdrawal and the attack were signs of nervous antic.i.p.ation, but neither broke the force of the blow which Haig struck on 8 August on a twenty-mile front from Morlancourt to La Neuville on the Avre. The troops were mostly British under Rawlinson with a French army under Debeney cooperating on his right. Their success first opened the eyes of the public to the change in the situation on the front, and on Ludendorff's own testimony deprived him of his last vestige of hope. It was no weak flank that was attacked, but the sector of the front that was most strongly held by German armies. The drive was straight along the great road from Amiens to St. Quentin on which the Germans had made their westward thrust in March; and the first day saw them seven miles back at Framerville. To the south they lost Moreuil, Mezieres, Demuin, Cayeux, and Caix, and to the north Morcourt, Cerisy, and Chipilly, while 7000 prisoners and 100 guns had been taken by 3 p.m. On the 9th those totals had risen to 24,000 prisoners and over 200 guns, while the British continued their advance to Rosieres and Lihons, and the French to Arvillers and Beaufort. Nor was that all; for south of Debeney, Humbert interposed with another attack between Montdidier and the Oise. By the 11th the Germans had lost to the French most of their gains in the June offensive, and to the British further ground between Albert and the Somme.

On that day the German line ran in front of Bray, Chaulnes, Roye, and La.s.signy to Ribecourt on the Oise. They had brought up reinforcements to make a stand on that shortened front, and they stubbornly contested the French advance on the La.s.signy ma.s.sif. But its capture was completed by the 15th, and the number of prisoners had risen to 33,000 and of captured guns to over 600. The Germans were also being pushed out of their salient on the Lys, where Merville fell on the 19th; and Mangin was forcing his way forward in the angle of the Aisne and the Oise between Soissons and Noyon. But the next great blow was struck north of the Somme by Byng between Albert and Arras. The Germans sought to evade its force by a timely retreat across the Ancre, and there was no such rapid advance as marked the first day of Rawlinson's offensive south of the Somme. But it was less interrupted, and day by day some progress was made. Byng's attack on the 21st was along a ten-mile front north of the Ancre, and the first day gave him Beaucourt, Achiet-le-Pet.i.t, Bucquoy, Courcelles, and Moyenneville. On the 22nd he extended his attack from Albert to the Somme and advanced two miles to a line between Albert and Bray. On the 23rd his left was advanced another couple of miles to Boiry, Ervillers, Bihucourt, and Irles, while on his right the Australians captured Bray. The German centre at Thiepval was thus outflanked on both sides; it gave way on the 24th, and Byng pushed on to the outskirts of Bapaume. Bapaume held out for five days longer while Byng pushed his right forward along the Somme towards Peronne, and extended his left attack northwards beyond the Scarpe.

Byng's addition to the pressure the Germans had to bear from north of the Scarpe to south of the Oise imposed upon them a retreat as extensive as that of March and April 1917; but now they could not make it at their leisure. On the 27th they had to abandon the line south of the Somme on which they had stood since the 15th, when they recovered stability after Rawlinson's offensive. Roye was relinquished that day and Chaulnes and Nesle on the 28th. Noyon followed on the 29th, partly in sympathy with the northern withdrawal and partly owing to Humbert's pressure on the north-western bank of the Oise, but also because it had been outflanked to the south by Mangin's advance between the Oise and the Aisne. Beginning on the 17th with an attack on a ten-mile front between Tracy-le-Val and Vingre he had steadily pushed on until by the 23rd his left flank held the Oise as far as its junction with the Ailette and his front faced the latter ca.n.a.lized river as far as Guny. By the 29th he was across the Ailette and threatening to turn the whole German position south of the Somme at Chauny. Bapaume fell on the same day as Noyon, and it soon became clear that the Somme would not protect the Germans any more than it had done the British in March. For on the 31st the Australians stormed Mount St. Quentin the bulwark of Peronne, and Peronne itself fell into their hands on 1 September. Simultaneously Byng's army pressed forward from Bapaume to the Ca.n.a.l du Nord which runs north from Peronne.

But this after all was ground we had held for a year in 1917-18, and the Hindenburg lines might serve the Germans as well in 1918-19. More significant of the coming debacle was the success of Horne's First Army, which now intervened and extended the line of Byng's attack. Already Canadian and British troops, by the capture of Vis-en-Artois on the 27th, Boiry on the 28th, and Haucourt on the 30th, had seized ground which the Germans had held since 1914; and on 2 September in one of the outstanding actions of the campaign Canadian and British troops broke the Drocourt-Queant line on a front of six miles between etaing and Cagnicourt. On that day the British army fired 943,857 sh.e.l.ls. No single engagement caused greater depression in Germany, but the impression was somewhat fallacious; for behind this sector of the Hindenburg lines were waterways which were even worse obstacles to our tanks, and although the Canadians pressed on to L'ecluse, ecourt, and Rumancourt, they were hemmed in on their left by the Sensee and in front by the Ca.n.a.l du Nord, which protected Douai to the north and Cambrai to the east. The advance here was checked for some weeks, but it went steadily on along other fronts. The salient on the Lys was melting away: Bailleul fell on 30 August, Mount Kemmel on the 31st, and Ploegstreet wood on 4 September. Lens was evacuated on the 3rd. South-west of Cambrai the British were approaching their old lines, and east of the Somme the Germans were retreating to St. Quentin. On the 6th the French took Ham and Chauny, and on the 9th they were once more across the Crozat ca.n.a.l. Mangin was pushing his way towards the St. Gobain ma.s.sif, and French and American troops were driving the Germans back from the Vesle across the Aisne. It looked as though winter might come with the line of battle much where it was before the German offensive began in March.

But the latter half of September gave a novel aspect to affairs. A great deal, no doubt, was due to Foch and the unity of command; but that unity did not extend to the East nor account for the debacle of Bulgaria and Turkey. It was, however, partly responsible for the extension of our offensive in France beyond the limits of the year before and for the timing of the American attack in the Woevre. In the hour of his Allies' need President Wilson had consented to the brigading of American with French and British troops, and to the employment of American divisions as supports for French and British generals. But with the American Army growing equal in size to the French and the British and acquiring an independent skill in war, there could be no hesitation about an American command on an equal footing with the armies of Haig and Petain; and to the Americans under General Pershing had been allotted the right wing of the Allied front, the British forming the left and the French the centre. Some critics talked of Pershing's armies being used as the spear-head of an invasion of Germany through Lorraine; but this would have been an eccentric operation, and there were obvious reasons for restoring Lorraine, if possible, to France undevastated by war. North rather than east was the natural direction for an American advance, and in either case an indispensable preliminary was to eliminate that strange wedge at St. Mihiel which the Germans had held since September 1914. The task would also be a useful apprenticeship for an independent American command. The attack was made on both sides of the salient on 12 September, but the princ.i.p.al drive was from the south on a twelve miles' front between Bouconville and Regnieville. Part of the defending force was Austrian, but the whole salient collapsed under the blow; 15,000 prisoners and 200 guns were captured, and a new front was formed on a straight line from Fresnes to Pont-a-Mousson. The strategic purpose was to free the American flank and communications in view of a bigger offensive northwards, and on the 15th Austria and Germany began their overtures for peace, to which President Wilson at once returned an unsympathetic reply.

Antic.i.p.ations as well as achievements counselled that diplomatic move, and Austria in particular had reason to fear developments on other fronts than the French. The Balkans had been quiescent during the summer, although the Greeks had on 30 May given an earnest of a better future by a victory at Skra di Legen, west of the Vardar, in which they captured 1500 Bulgarian and German prisoners, and on 18 June the fall of the pro-German Radoslavoff Ministry indicated that Ferdinand wished to present a less Teutonic appearance to the world. Italy, too, in pursuance of her a.s.sumed protectorate over Albania, thought in July that the time had come to a.s.sert herself, and with the a.s.sistance of some French troops began an advance towards Elbasan. The Austrians were taken by surprise, Berat was captured, and the country overrun as far as the s.e.m.e.ni and beyond the Devoli. The effort was not apparently serious; in August the Austrians returned to the attack, recaptured Berat, and drove the Italians back to their starting-point in a retreat boldly described in an Italian official p.r.o.nouncement as of no military importance. It helped to discourage Italy from taking an active part in the coming offensive against Bulgaria, but political motives were the princ.i.p.al reason for quiescence. Italy had a tenderness for Bulgaria arising out of her antipathy to Jugo-Slavs and Greeks, and while proclaiming that Austria must be totally destroyed, she exclaimed against the wickedness and folly of imposing on Bulgaria a second Peace of Brest-Litovsk (see Map, p. 151).

The success of the Balkan campaign did not, however, suffer much from the lack of Italian push. Franchet d'Esperey was commander-in-chief, and he was ably seconded by the Serbian Marshal Mis.h.i.tch. The Serbian Army was the spear-head of the attack, and it had with it an equally eager and effective force of Jugo-Slavs; with them cooperated the French on the west of the Vardar, while east of it were the Greeks and the British with the arduous and somewhat thankless task of facing the impregnable Demir Kapu defile and Belas.h.i.tza range. The offensive began on 15 September, and the main attack was on the Dobropolie ridge in the angle between the Tcherna and Vardar rivers. On the first day the Bulgarian line was broken on a front of sixteen miles, an advance was made of five, and 4000 prisoners and 30 guns were taken. On the morrow the front widened to twenty-two miles, and the advance increased to twelve; and within a week the Serbians had cleared the angle between the rivers and crossed the Tcherna on their left and the Vardar above Demir Kapu on their right. This cut the main Bulgarian communications with Prilep on the west and Doiran on the east, and compelled a general retreat along a hundred miles of front. On the 23rd the French occupied Prilep; on the 25th the Serbians captured Veles and Ishtip and pressed on towards Uskub, while their cavalry were at Kotchana almost on the Bulgarian frontier. The British, whose first attacks had been checked, had actually crossed the border at Kosturino on the road between Doiran and Strumnitza. Bulgaria had put her whole trust in the strength of her front, and with it she collapsed. An armistice was requested on the 25th, and Franchet d'Esperey's terms were accepted on the 30th. It was the most dramatic overthrow in the war, and within a fortnight the whole situation in the Balkans was transformed. The Serbians were bitterly disappointed at having to stay their avenging hands when almost at the gates of Sofia; but the elimination of Bulgaria made the recovery of their country a triumphal procession varied by the occasional defeat of Austrian rearguards. On 12 October they and their allies occupied Nish, and a week later they had reached the Danube. Nor was Serbia alone concerned. Austria had relied upon the Bulgarian buckler, and when it crumpled her entire hold not only on the Balkans but over her own Jugo-Slav subjects in Bosnia, Dalmatia, and Carinthia was relaxed. A general uprising of Jugo-Slavs in favour of union under the Serbian crown more than doubled the size of that kingdom which Austria had begun the war to crush.

Nor did this exhaust the effects of Bulgaria's capitulation. The terms of the armistice included the Allied occupation of Bulgarian railways, and this brought their military front up to the borders of Rumania on the north and of Turkey on the south. Presently Marghiloman's Ministry, which the Germans had imposed at Bukarest, fell, and Rumania prepared to resume her part in the war. Bulgaria, too, was willing to revive her quarrel with Turkey. The famous corridor had disappeared, and Turkey was an isolated unit. It was no wonder that the "Easterners" looked up again, and the Prime Minister's henchmen in the press began to tell stories about his single-handed and far-sighted championship of an Eastern campaign as the solution of the problem of the war. But the collapse of the Balkan front was ultimately due to the collapse of its German foundation. Berlin journalists talked of the German troops which would soon bring back Bulgaria to her senses and to the Teutonic fold. But they were mortgaged to the Western front, and instead of a German expedition to a.s.sist her under Mackensen, Turkey was faced with ruin at the hands of Allenby.

His blow had followed swiftly on that of Franchet d'Esperey, and four days after the Balkan campaign had opened, British forces began the battle which was to prove the most perfect operation of the war. Preparations had been in progress during the summer, and little had been done to modify the British line running a dozen or fifteen miles north of Jericho, Jerusalem, and Jaffa to the sea. A Turkish counter-attack on 13 July had even met with some initial success; but the Turks had been unable to maintain their strength, the Germans could not a.s.sist them, the Arabs were perpetually hara.s.sing them along the Hedjaz railway, and what reserves they had were sent on a wild goose chase for the recovery of Turkish dominion in Caucasia and Persia and along the sh.o.r.es of the Caspian. The pursuit was rendered attractive by Russian impotence and anarchy: Armenia was regained and subjected to a final and more extensive ma.s.sacre than ever; Northern Persia was overrun, and even the long and adventurous arms which the British Empire stretched out in August from Mesopotamia and India to the southern and eastern sh.o.r.es of the Caspian failed to save Baku from the combined efforts of Turkish troops and Bolshevik treachery on 14 September. But Allenby, the luckiest of British generals, brought down these airy Turkish castles with a single blow. He had been largely reinforced from India, which mobilized during the war nearly a million men and bore the chief burden of the Palestine and Mesopotamian campaigns; he had got a magnificent force of cavalry, and with it the terrain and open fighting wherein to exhibit a model of that traditional strategy from which the glory on European battlefields had departed for ever.

On 19 September his infantry drove the Turks from a sixteen-mile line between Rafat and the sea back a dozen miles to the railway junction at Tul Keram, while his cavalry burst through to the right towards the gap south-east of Mount Carmel and the plain of Esdraelon. It was a rare ride: on the morrow they were forty miles north and north-east at El Afuleh, Nazareth, and Beisan; and then wheeling south-east they cut off the retreat of nearly the whole of the Turkish forces. On the 22nd Allenby reported that 25,000 prisoners and 200 guns had been taken and counted, and that the Seventh and Eighth Turkish armies had virtually ceased to exist. The Fourth was pursued across the Jordan, and mostly mopped up between its pursuers and the Arabs to the east. On the 25th we were round the Lake of Galilee, and the number of prisoners had risen to 45,000 and of captured guns to nearly 300. There was nothing left to stop our advance, which was joined by some French battalions, while the Arabs kept pace on the other side of the Jordan. On the 28th we effected a junction with them at Deraa, and Damascus fell on the 30th. On 6 October cavalry, advancing between Mts. Lebanon and Hermon, seized Zahleh and Rayak between Damascus and Beyrut, which the French occupied on the 7th, while the British took Sidon. On the 9th we were at Baalbek, on the 13th at Tripolis, and on the 15th at Homs. On the 26th Aleppo fell, and on the 28th we reached Muslimieh, that junction on the Baghdad railway on which longing eyes had been cast as the nodal point in the conflict of German and other ambitions in the East.

Allenby played the leading part in Turkey's destruction, partly because Marshall's attention in Mesopotamia had been distracted towards the Caspian. But in October he resumed his interrupted march up the Tigris: on the 25th his troops captured Kirkuk and forced the pa.s.sage of the Lesser Zab; and on the 28th they took Kalat Shergat, and after a six days' battle forced the Turkish army on the Tigris to surrender. Turkey had taken a lot more beating than Bulgaria, but the end was the same. On 30 October an armistice was signed, which permitted the Allies to occupy the forts on the Dardanelles and Bosporus and make free use of the Straits. Marshall entered Mosul, and presently British ships commanded the Black Sea and British troops were holding a line across Caucasia to the Caspian and connecting with the chain of forces established between Krasnovodsk and India. An end was thus put to Germany's dreams of a Teutonic-Turco-Turanian avenue into the heart of Asia, but the search for an eastern front in Russia against the Central Empires was elusive. For the Bolsheviks, in spite of the murder of Count Mirbach the German amba.s.sador at Moscow on 6 July, grew ever more friendly to the Prussians, and the Entente had to go to Vladivostock for a basis of operations, and rely largely upon the romantic achievements of the Czecho-Slovak prisoners who had enlisted in the Russian armies and refused to lay down their arms at the Peace of Brest-Litovsk. At first the Bolsheviks promised them a pa.s.sage via Siberia to the Western front, but then, like Pharaoh hardened their hearts and refused to let the infant nation go. Thereupon the Czecho-Slovaks set up for themselves, seized the Siberian railway from the Bolsheviks, and after much hardship and fighting established contact with the motley Entente forces advancing from Vladivostock. With their a.s.sistance an anti-Bolshevik government, of which Admiral Koltchak afterwards made himself master, was set up in Siberia, while Entente forces, mostly British, were sent to Archangel and the Murmansk coast to prevent the Germans establishing their authority there as they had done in the Baltic provinces "liberated" by the Peace of Brest-Litovsk.

The Conquest Of Syria

But this Eastern front, which as late as August was regarded in high but civilian quarters as indispensable to the Allied success, failed to pierce the protection which the Bolsheviks gave to Germany or to penetrate farther west than the Urals; and Germany had after all to be beaten by professional strategists on the Western front. There was little fault to be found with their progress, and while Bulgaria, Turkey, and Austria were collapsing in the East, the Germans were being steadily driven towards disaster on a widening field of battle in the West. Simultaneously with Pershing's destruction of the St. Mihiel salient the British were thrusting the Germans back to the Hindenburg lines between Cambrai and St. Quentin, and Mangin was pushing forward towards the forest of St. Gobain. The Germans attempted to stand at epehy, but on 18-19 September they were driven back with the loss of 11,750 prisoners and 100 guns; and from the 27th to the 30th was fought the first phase of the battle for Cambrai and St. Quentin, in which the British First, Third, and Fourth armies took 26,500 prisoners and 340 guns apart from the gains of the French. The object was to complete the breach of the Hindenburg lines on the strength of which public opinion in Germany was stayed; and it was a critical operation. The lines themselves were reinforced by the Ca.n.a.l du Nord protecting Cambrai and the Scheldt-St. Quentin ca.n.a.l between Cambrai and St. Quentin.

The southern sector in front of the Fourth Army was the more strongly fortified, and an intense bombardment began on the night of 26-27 September which continued till the 29th. This tended to divert attention from the First and Third armies, which on the 27th forced the Ca.n.a.l du Nord south of Moeuvres and then spread fanwise along the eastern bank. By the end of the day they were more than half-way from the Ca.n.a.l du Nord to Cambrai, and on the 28th the advance was continued across the Scheldt ca.n.a.l at Marcoing and broadened from Palluel on the north to Gouzeaucourt on the south. On the 29th the Fourth Army began its attack on the ca.n.a.l to the north of St. Quentin. It was well supported by several American divisions, and the great episode of the day was the capture of Bellenglise by troops who crossed the ca.n.a.l equipped with life-belts, mats, and rafts. East of Bellenglise, Lehaucourt and Magny were also stormed, and north of it Nauroy and Bellicourt. Meanwhile the Third Army captured Masnieres and penetrated into the western outskirts of Cambrai while the Canadians threatened to outflank it on the north. On the 30th the Germans had to withdraw their centre at Villers Guislain and Gonnelieu, while the Fourth Army extended its gains southwards by the capture of Thorigny; and, thus menaced, the Germans had to abandon St. Quentin to the French on 1 October. On that day, too, New Zealanders and British troops took Crevecoeur and Rumilly south of Cambrai, and the Canadians Blecourt to the north of it. The Hindenburg line, apart from its tottering supports, had gone at the moment when Bulgaria was capitulating; and on the same 30 September Count Hertling and all his Secretaries of State resigned.

The British victory, while the critical movement on the Western front, was but one of the four operations which Foch had concerted with Haig in the middle of September. The other three were a Belgian attack at Ypres, an American advance on the Meuse, and a French offensive in close connexion with it in Champagne and the Argonne. The Belgian attack was an agreeable surprise, and nothing did more to illumine the change from 1917 than the contrast between its rapid success and the painful crawl of Gough's campaign. The cause was that which also accounted for the Germans' failure elsewhere; they had not the forces to sustain their vast and crumbling front, and they attempted to hold the line in Belgium with no more than five divisions. The attack began on 28 September on a twenty-three mile front, and in one day 50 per cent more ground was covered than had been gained in three months the year before. The whole of Houthulst forest, which then had hardly been touched, was taken at a stroke; and on the 29th Dixmude fell and the Belgians were across the Roulers-Menin road. As a consequence of this and of Haig's advance the Germans had to evacuate the rest of the Lys salient and draw back their front towards Lille and Douai. Armentieres was recovered on 3 October, La Ba.s.see and the Aubers ridge were abandoned without a struggle, and the Germans surrendered the remaining section of the Drocourt-Queant line, withdrawing to the Douai, Haute Deule, and Sensee ca.n.a.ls which protected Lille and Douai.

The French and Americans had a sterner task in the Argonne and on the Meuse, for here was the pivot of the Germans' whole position in the conquered territory. A possible retirement to the Meuse had been contemplated in 1917, and in September 1918 the Germans would have been glad to surrender everything west of it in return for safety on that line; hence their withdrawals and feeble resistance in Flanders. But the Meuse from Verdun to Mezieres was an indispensable flank for any German front in Belgium; it had now become more to the Germans than even that, for it was the only shield behind which their armies could escape disaster and get back to Germany at all. Whatever else might have to go, this flank must hold; if it gave, the Germans would have to capitulate or suffer the wholesale destruction of their forces. Hence the stubbornness of the defence the Americans encountered; the terrain gave it every advantage with which art could supplement nature; and a singular and serious breakdown of their commissariat added to the difficulties under which American troops fought with intrepid skill.

The attack was launched on 26 September. The American front ran for seventeen miles from Forges on the Meuse, eight miles north of Verdun, to the centre of the Argonne, whence the French extended it to Auberive on the Suippe. Pershing's First Army advanced an average depth of seven miles and captured Varennes, Montfaucon,-for long the Crown Prince's headquarters,-Nantillois, and Dannevoux. Gouraud's progress was less rapid but better sustained. His greatest advance was only three miles, but it extended along a wider front and developed during the following days, while the Americans were held up by defective organization. Somme-Py and Manre were taken on the 28th, while on Gouraud's left Berthelot began to move from Reims, and farther west Mangin pursued the Germans across the Aisne. Progress along the whole French front continued in October; Gouraud's right pressed on to a level with, and then in advance of, the American left towards Challerange and Grandpre; his centre advanced towards Machault, and on his left Berthelot took Loivre, Brimont, and forced the pa.s.sages of the Suippe at Bertricourt and Bazancourt, and of the Aisne at Berry-au-Bac. The Moronvillers ma.s.sif was thus outflanked, and by the middle of the month the Germans were evacuating the whole of their ground south of the Aisne. This retreat, coupled with the French advance east of St. Quentin, endangered the great apex of the German front in the St. Gobain forest, and by the 10th its abandonment was begun. On the 11th the Chemin des Dames was relinquished, on the 13th the French were in La Fere and Laon, and the Germans were retreating to the line of the Serre.

Nevertheless, the advance of the right wing of the Allied front had not quite come up to expectations. The prolonged maintenance of the German bastion in the Argonne and on the Meuse enabled their centre to withdraw more or less at its leisure and thus avoid the colossal Sedan with which it was threatened; and, the French centre having been cast for a part subsidiary to those of the two wings, the brunt of the fighting fell upon the British, whose advance was not so fatal as similar progress would have been on the other wing. They were greatly a.s.sisted by American divisions serving with the Third and Fourth armies, by the Belgians and French on their left, and by the French on their right; but the check to the American advance enabled the Germans-unfortunately for them, as it turned out-to transfer reinforcements from the Meuse to Cambrai and Valenciennes.

Cambrai did not therefore fall until another series of actions had been fought in the first nine days of October. The Scheldt ca.n.a.l to the north of it had proved a formidable obstacle, and Haig determined to press the attack from the south, where the Fourth Army had prepared the way on 29 September by destroying the Hindenburg line at Bellicourt and Bellenglise. On 3 October Rawlinson attacked again between Le Catelet and Sequehart and captured those villages, Gouy, Ramicourt, and the Beaurevoir-Fonsomme line. On the 4th and 5th further progress was made by the taking of Beaurevoir and Montbrehain, while north of Le Catelet the Germans were driven from their positions east of the ca.n.a.l, which were occupied by the Third Army. On the 8th the final phase in the battle for Cambrai began. The chief fighting was on the line secured on the 3rd. An American division captured Brancourt and Premont, and British divisions Serain, Villers-Outreaux, and Malincourt north-east of Le Catelet. New Zealanders south of Cambrai look Lesdain and Esnes, and three British divisions Serainvillers, Forenville, and Niergnies, penetrating the southern outskirts of Cambrai, while to the north of it Canadians captured Ramillies, crossed the ca.n.a.l at Point d'Aire and entered the city on that side. During the night the whole of it fell into our hands; the Germans were driven back in disorder to within two miles of Le Cateau; and Bohain was reached ten miles east of Bellicourt and a similar distance south-west of Le Cateau. By the 10th the advance had been carried to the line of the Selle river, on which the Germans made another stand, while farther south the French pushing east of St. Quentin, cleared the Oise-Sambre ca.n.a.l as far north as Bernot. On the 10th Le Cateau fell, and by the 13th the British had gained the west bank of the Selle as far north as Haspres.

A great wedge had thus been thrust into the German line, leaving p.r.o.nounced salients to the north of it round Lille and Douai, and to the south-east of it between the Oise and the Aisne. It was the policy of the Entente to eschew the destruction which fighting in cities involved, and it was particularly desirable to compel the Germans to retreat from Lille and its industrial neighbourhood by threats of encirclement rather than by frontal attack. To complete the process begun on the south, the advance in the north was now resumed; and on 14 October Belgian forces with a French army under Degoutte and the British Second Army under Plumer attacked the whole front in Flanders between Dixmude and the Lys at Comines. Their success was even more striking than it had been on 28 September; the Belgians and French carried Courtemarck, Roulers, and Iseghem, while the British pushed along the north bank of the Lys until on the 16th they held it as far as Harlebeke, farther east than Ostend and even than Bruges. On the 15th the Belgians captured Thourout and the British Menin, crossing the Lys at various points and taking Comines on the 16th. The effect of this advance was to precipitate a comprehensive German retreat both north and south. The coveted Belgian coast had at last to be abandoned: Ostend fell on the 17th, Zeebrugge and Bruges on the 19th, and by the 21st the Germans were twenty miles from the sea, striving to stand on the Lys ca.n.a.l in front of Ghent. To the south the withdrawal was no less complete: both Lille and Douai were entered on the 17th; Tourcoing and Roubaix soon followed; and by the 21st our Second and Fifth armies had advanced to the Scheldt on a front of twenty miles, forming nearly a straight line with the First, Third, and Fourth on the Selle.

There the battle had been renewed on the 17th, as soon as our advancing lines of communication had been sufficiently repaired to bear the s