WAR WITHOUT END.

COALITION WARFARE.

The First World War was a coalition war. Its intensity, its scale and its length were all the products of the alliances that sustained it. In 1918 one, that of the Central Powers, began to fall apart, but the other, that of the Entente, achieved a fusion, albeit flawed, which enabled it to wield greater military and economic power than any unit previously seen in the history of the world. In liberalism, however imperfectly expressed and however compromised by the business of waging war, it had a common ideological focus. 'What we demand in this war', Woodrow Wilson told the United States Congress on 8 January 1918, 'is nothing peculiar to ourselves.'1 Wilson went on to give shape to that ambition, spelling out his Fourteen Points which he believed could deliver a 'peace without victory'. His aim was to keep Russia in the war, and, in providing a counterpoint to Bolshevism, he was sufficiently anti-imperialist to alarm both Britain and France. Freedom of the seas (his second point) challenged British maritime supremacy, and his fifth called for the recognition of the rights of colonial populations. But the specifics were ambiguous and negotiable. In the event the allies' responses were muted; the Fourteen Points embraced their fundamental territorial ambitions within Europe, and included the German evacuation of Belgium, the return to France of Alsace-Lorraine, and the settlement of Italy's frontier with Austria-Hungary on the basis of nationality. Wilson's key audience was not governments but peoples. Only three days previously Lloyd George had delivered a speech on war aims to the Trades Union Congress in London, which was couched in similar terms and in which he had embraced open diplomacy and the idea of a 'democratic peace'. 'The days of the Treaty of Vienna are long past. We can no longer submit the future of European civilisation to the arbitrary decisions of a few negotiators striving to secure by chicanery or persuasion the interests of this or that dynasty or nation. The settlement of the new Europe must be based on such grounds of reason and justice as will give some promise of stability. Therefore it is we feel that government with the consent of the governed must be the basis for any territorial settlement in this war.'2 Clemenceau may have entertained reservations about some of the things said by his British and American counterparts, but the French left did not. Publicly, he endorsed Lloyd George. Moreover, Wilson's fourteenth and best known point, the formation of 'a general a.s.sociation of nations', had British and French authors too - including Jan Christian s.m.u.ts and Lord Robert Cecil; the latter had drafted Lloyd George's speech. 'The programme of the world's peace', Wilson said, 'is our programme ... the only possible programme.' 3 3 In the popular imagination he had captured the moral high ground for the Entente. In the popular imagination he had captured the moral high ground for the Entente.

America's partic.i.p.ation in the alliance, even if as a self-proclaimed a.s.sociate rather than as a full ally, also brought in other powers to what was evidently a going concern. The states of South America either entered the war or severed relations with Germany. In Asia, China declared war on the Central Powers on 14 August 1917 and offered 300,000 men for service in Europe. In the Balkans, Greece joined the Entente on 2 July 1917. But, after Bulgaria did so in September 1915, no other power in the world sided with Germany.

The combined populations of the four Central Powers totalled 144 million in 1914; those of the princ.i.p.al Entente powers of 1918 (including their colonies) 690 million.4 However, economic potential and military capability were not the same. Turkey, despite its backwardness, had twice defeated Britain in battle, and its military contribution to the war as a whole was greater than that of the United States. By the same token, the Entente's comparative strength rested on more than the sum of its resources. First, its princ.i.p.als were to a greater extent equals than were the Central Powers. That did not mean there were not great disparities of wealth between them: by 1918 America had the money, but France still had the biggest army and Britain the biggest navy. Second, the allies had in 1915 begun a process of organic growth which, however fractious, had reached a maturity and breadth by 1918 which ensured that joint military command was the coping-stone of the whole - not, as in the case of the Central Powers, its only foundation. Third, many of the fronts, particularly the princ.i.p.al one in the west, were genuinely shared responsibilities. Germany sh.o.r.ed up the western front single-handedly, but then on other fronts could expect its allies to act as virtual mercenaries, providing the cannon fodder for German commanders to use according to German priorities. There were Entente fronts, pre-eminently the Italian, which were largely sustained by the army of one nation, but that army sustained its legitimacy by acting in conformity with its own national objectives. However, economic potential and military capability were not the same. Turkey, despite its backwardness, had twice defeated Britain in battle, and its military contribution to the war as a whole was greater than that of the United States. By the same token, the Entente's comparative strength rested on more than the sum of its resources. First, its princ.i.p.als were to a greater extent equals than were the Central Powers. That did not mean there were not great disparities of wealth between them: by 1918 America had the money, but France still had the biggest army and Britain the biggest navy. Second, the allies had in 1915 begun a process of organic growth which, however fractious, had reached a maturity and breadth by 1918 which ensured that joint military command was the coping-stone of the whole - not, as in the case of the Central Powers, its only foundation. Third, many of the fronts, particularly the princ.i.p.al one in the west, were genuinely shared responsibilities. Germany sh.o.r.ed up the western front single-handedly, but then on other fronts could expect its allies to act as virtual mercenaries, providing the cannon fodder for German commanders to use according to German priorities. There were Entente fronts, pre-eminently the Italian, which were largely sustained by the army of one nation, but that army sustained its legitimacy by acting in conformity with its own national objectives.

Indeed that was the Entente's recurring problem with Italy: it timed its offensives to suit its own needs, not those of achieving unity of effect in s.p.a.ce and time. The collapse at Caporetto created the opportunity both to integrate Italy's efforts with those on other fronts and to complete the integration of the Entente by the establishment of a Supreme War Council. In July 1917 Robertson and Foch, as their countries' chiefs of staff, had met Generals Cadorna and Pershing, the commander of the American Expeditionary Force, to discuss the acceleration of the American military commitment to Europe in the light of Russia's collapse. They had broached the idea of amalgamating the Americans with more experienced British and French units. But Pershing's orders told him that 'the forces of the United States are a separate and distinct component of the combined forces, the ident.i.ty of which must be preserved'. This was not just a matter of national pride or public opinion, it was also one of policy: an independent army would enable America to retain a free hand at the peace negotiations. Dogmatic and rigid, a teetotaller, 'Black Jack' Pershing had lost his wife and three daughters in a fire in 1915; although he consoled himself with a mistress while in France, his men were banned access to brothels. He adhered to the principle of independence with similar vehemence. However, there was one major obstacle to its fulfilment: the lack of a sizeable body of proven American commanders and trained staff officers. The British experience had shown that it might be possible to improvise a ma.s.s army in comparatively short order, but that, as Haig's director of military intelligence put it, 'It will be a very difficult job for them [the Americans] to get a serviceable staff going even in a year's time'.5 The allies' military representatives had therefore concluded that some form of inter-allied organisation would be required to facilitate this process. The allies' military representatives had therefore concluded that some form of inter-allied organisation would be required to facilitate this process.

The prime ministers of Britain and France, together with Robertson and Foch, met at Rapallo between 5 and 7 November 1917 to coordinate their response to Italy's plea for help. The Italians, according to their prime minister, Vittorio Orlando, 'were treated like servants'.6 The problem was Cadorna, who not only failed to attend but also submitted such inflated and inaccurate calculations of the German and Austro-Hungarian forces opposing him as to confirm the doubts about his competence. Foch and Robertson were being asked to commit eleven French and British divisions to a command structure in which they had no faith. Both Foch and Petain were adamant that the six French divisions should remain at the disposal of France, and should be handled in conjunction with those on the French front: in other words, that the line from the Somme to the Piave should be a unit. The problem was Cadorna, who not only failed to attend but also submitted such inflated and inaccurate calculations of the German and Austro-Hungarian forces opposing him as to confirm the doubts about his competence. Foch and Robertson were being asked to commit eleven French and British divisions to a command structure in which they had no faith. Both Foch and Petain were adamant that the six French divisions should remain at the disposal of France, and should be handled in conjunction with those on the French front: in other words, that the line from the Somme to the Piave should be a unit.

The French divisions in Italy were placed under Foch as chief of the general staff, not Petain as France's commander-in-chief: another step in the principle of dividing and ruling the military. Of the representatives at Rapallo, Foch was the only one whose motives in promoting the idea of a Supreme War Council smacked of altruism, and even he - having been tarred in 1916 with the brush of Joffre and of the Somme battle - had a career to resurrect. His prime minister, still Painleve for a few days more, saw the creation of a Supreme War Council as an interim step towards the promotion of Foch as allied commander-in-chief, but felt that he could not say so to the British: that was understandable after the Nivelle affair. Besides, Painleve and Petain thought the princ.i.p.al benefit in the creation of joint command would be that the British would take over more of the line from the battered French army. The French therefore proposed that Foch, being chief of the general staff, should also be their permanent military representative on the Supreme War Council. The corollary of such an arrangement would have been that Robertson would similarly exercise both functions for Britain. This was totally unacceptable to Lloyd George, who saw the Supreme War Council as an opportunity not only to subordinate Haig but also to break the axis between him and Robertson. He persuaded the French that Foch should not hold both appointments simultaneously and asked Henry Wilson, then languishing in a job at home, to be Britain's military representative on the Supreme War Council. Wilson, like his chum Foch, found his career reviving because of his usefulness to civilians as an alternative source of professional military advice.

The key issue facing the new body was the creation of a strategic reserve, not least in antic.i.p.ation of a German offensive in the west in 1918. Robertson argued that British reserves should be under his control as it was his job as chief of staff to know their capabilities, and Haig and Petain both said that they could not spare any divisions to create a reserve. Robertson resigned over the issue, and Wilson succeeded him, with Rawlinson moving into the Versailles post. Rawlinson backed Haig, and when the German offensive was unleashed between St Quentin and Arras on 21 March 1918 the Supreme War Council was no more than an advisory body, with no troops under its command and no ability to help General Gough's 5th Army.

It did not matter. Haig and Petain had agreed a scheme for mutual a.s.sistance which the French fulfilled to the letter. On 20 January Petain had undertaken to release up to twelve divisions immediately in the event of a German attack against the British, and up to half the French army if need be. By the end of 23 March, he had committed fourteen divisions to the battle, all of which were in action by the 28th. By then half the French army was on the move. The roads were jammed with troops and transport. Haig consistently underestimated his own requirements and the scale of French support. Before the blow fell he thought six French divisions would suffice, and on the 21st itself he asked for only three. He requested twenty divisions on the 25th, but by then he already had the support of twenty-one, with a further nineteen on the way. Petain created a whole new army group, under Fayolle's command, to straddle the front between the Somme and Oise rivers. Of its front of thirty-six miles, fully two-thirds was held by British troops, the elements of the 5th Army, which Haig told General Byng, commander of its northern neighbour, the 3rd Army, was finished.

Normally calm and imperturbable, Haig was exhibiting the nearest to panic of which he was capable. He confronted a crisis with unfamiliar tools: his staff had been emasculated over the winter, and Wilson, his new spokesman in London, was 'our only military black-leg', seemingly determined to curb the powers of the individual commanders-in-chief. When Wilson came to France and saw him on 25 March, Haig 'said that "unless the whole French army came up we were beaten" and "it would be better to make peace on any terms we could'".7 Haig was convinced that the greater danger lay to the north, towards Flanders. Petain disagreed: he correctly concluded that the thrust south of the Somme, spearheaded by Crown Prince Wilhelm's army group, const.i.tuted the major threat. Haig's desire to pull back to the north, using the British 3rd Army to guard his right flank, reflected the operational logic of what his enemy ought to have been doing, but not what he was actually doing. Ludendorff's offensive was being carried on the back of its own tactical opportunism, and Petain, who had paid more attention than the British to the evolution of German tactics, recognised that. Haig's subsequent claim that the French failed to support him and were pulling back to defend Paris was the exact opposite of the truth. Haig was convinced that the greater danger lay to the north, towards Flanders. Petain disagreed: he correctly concluded that the thrust south of the Somme, spearheaded by Crown Prince Wilhelm's army group, const.i.tuted the major threat. Haig's desire to pull back to the north, using the British 3rd Army to guard his right flank, reflected the operational logic of what his enemy ought to have been doing, but not what he was actually doing. Ludendorff's offensive was being carried on the back of its own tactical opportunism, and Petain, who had paid more attention than the British to the evolution of German tactics, recognised that. Haig's subsequent claim that the French failed to support him and were pulling back to defend Paris was the exact opposite of the truth.

That is not to say that there was not deep despondency in Entente circles by 23 March. Paris was sh.e.l.led, and it was suggested that as in 1914 the French government should leave the city. The Australian corps commander, Lieutenant-General John Monash, an engineer in civilian life and a brigadier at Gallipoli, recalled from leave in the south of France, reached Doullens at about 3 pm on the 25th: 'Viewed from that particular locality it almost looked as if the whole British Army in this part of the world was in a state of rout'.8 On the following day Poincare, the president of France, arrived in the town to chair a meeting attended not only by the headquarters staffs of the British and French armies but also by Clemenceau and, for the British war cabinet, Lord Milner. Henry Wilson was also there. After seeing Haig the previous day, he had then met his friend Ferdinand Foch, the putative commander of the allied reserve if it had existed. Both of them were convinced advocates of a united command, both of them knew that Foch was the logical choice to exercise it, and they had little difficulty in persuading their political superiors at the meeting, Clemenceau and Milner, that that was the way forward. Haig's alleged role in the appointment of the allied generalissimo was little more than a face-saving exercise conducted for the benefit of his own dignity and posterity. On the following day Poincare, the president of France, arrived in the town to chair a meeting attended not only by the headquarters staffs of the British and French armies but also by Clemenceau and, for the British war cabinet, Lord Milner. Henry Wilson was also there. After seeing Haig the previous day, he had then met his friend Ferdinand Foch, the putative commander of the allied reserve if it had existed. Both of them were convinced advocates of a united command, both of them knew that Foch was the logical choice to exercise it, and they had little difficulty in persuading their political superiors at the meeting, Clemenceau and Milner, that that was the way forward. Haig's alleged role in the appointment of the allied generalissimo was little more than a face-saving exercise conducted for the benefit of his own dignity and posterity.

Some doubted Foch's mental faculties: his head had received a severe blow in a recent car accident. He was, in Clemenceau's words, 'not superabundant in nuances'.9 Furthermore, he had been educated by Jesuits, a definite disqualification in the mind of the anti-clerical prime minister. But will-power and faith were precisely the attributes that the situation demanded. On 26 March Foch was given the task of coordinating the actions of the British and French armies on the western front. At Beauvais, on 3 April, his powers were extended to embrace 'the strategic direction' of all the armies, including that of the United States. Foch conveyed the sort of rhetorical determination that politicians like to hear. 'I shall fight without ceasing', he was reported to have said to a group of officers. 'I shall fight in front of Amiens. I shall fight in Amiens. I shall fight behind Amiens. I shall fight all the time.' Furthermore, he had been educated by Jesuits, a definite disqualification in the mind of the anti-clerical prime minister. But will-power and faith were precisely the attributes that the situation demanded. On 26 March Foch was given the task of coordinating the actions of the British and French armies on the western front. At Beauvais, on 3 April, his powers were extended to embrace 'the strategic direction' of all the armies, including that of the United States. Foch conveyed the sort of rhetorical determination that politicians like to hear. 'I shall fight without ceasing', he was reported to have said to a group of officers. 'I shall fight in front of Amiens. I shall fight in Amiens. I shall fight behind Amiens. I shall fight all the time.'10 It could have been one of Clemenceau's own speeches. It meant nothing, but it conveyed an aura of command. It could have been one of Clemenceau's own speeches. It meant nothing, but it conveyed an aura of command.

Foch visited Gough the same day. He 'said in a loud, excited manner, "There must be no more retreat, the line must now be held at all costs," and then walked out of the room back to his car'.11 He offered no troops and no specifics. His role was to rally, not to plan. Haig grumbled that Foch 'spoke a lot of nonsense' and complained - with some reason - that he was 'unmethodical and takes a "short view" of the situation. For instance, he does not look ahead and make a forecast of what may be required in a week in a certain area and arrange accordingly'. He offered no troops and no specifics. His role was to rally, not to plan. Haig grumbled that Foch 'spoke a lot of nonsense' and complained - with some reason - that he was 'unmethodical and takes a "short view" of the situation. For instance, he does not look ahead and make a forecast of what may be required in a week in a certain area and arrange accordingly'.12 In late March and early April 1918, the only allied commander with that sort of vision was Petain, whom Haig, like Clemenceau, had convinced himself was defeatist. The French commander-in-chief was bent not simply on stopping the Germans and on closing the gap between the two allied armies, but also on preparing the counterattack that would strike north-east into the salient created by the German advance. The effect of his taking the 'long view' was ironically to widen the gap in understanding between himself on the one hand and Haig and Foch on the other. Petain's attacks were conceived within an overall context of mobile defence and of conserving lives. For Foch, there was only one sort of attack: 'Everything that will not be achieved rapidly will not be achieved at all.... Our offensive must therefore be mounted both with speed and force.' In late March and early April 1918, the only allied commander with that sort of vision was Petain, whom Haig, like Clemenceau, had convinced himself was defeatist. The French commander-in-chief was bent not simply on stopping the Germans and on closing the gap between the two allied armies, but also on preparing the counterattack that would strike north-east into the salient created by the German advance. The effect of his taking the 'long view' was ironically to widen the gap in understanding between himself on the one hand and Haig and Foch on the other. Petain's attacks were conceived within an overall context of mobile defence and of conserving lives. For Foch, there was only one sort of attack: 'Everything that will not be achieved rapidly will not be achieved at all.... Our offensive must therefore be mounted both with speed and force.'13 He spoke in the accents of 1914, as did Haig. Haig may have been a Presbyterian, but his inner certainty meant that more united than separated him from Foch, and in the second half of the year the circ.u.mstances of the war swung their way, bringing events on the ground into harmony with their notions of warfare. He spoke in the accents of 1914, as did Haig. Haig may have been a Presbyterian, but his inner certainty meant that more united than separated him from Foch, and in the second half of the year the circ.u.mstances of the war swung their way, bringing events on the ground into harmony with their notions of warfare.

The allied supreme command pose for the camera. Left to right Petain, Haig, Foch and Pershing

Petain's credibility was further dented on 27 May 1918. He had long forecasted a German attack on the Chemin des Dames, but he had not antic.i.p.ated its scale and his subordinates were reluctant to apply his defensive tactics: it was the most significant defeat suffered by the French on the western front since 1914. Five British divisions had been transferred to the sector to 'rest' after the spring offensives but Haig refused more, so showing the limits of reciprocity as well as of Foch's authority to bend the national commanders-in-chief to his will. By 1 June the Germans were held and Petain turned his thoughts once more to the counterattack, rather than a step-by-step withdrawal. Foch rejected the idea, but the French were ready when the Germans attacked again, in Champagne on 15 July. Petain's instructions on defence in depth were implemented as they had not been on 27 May. The French 'put up no resistance in front', a German officer, Rudolf Binding, wrote in his diary; 'they had neither infantry nor artillery in this forward battle-zone . . . Our guns bombarded empty trenches; our gas-sh.e.l.ls ga.s.sed empty artillery positions; only in little hidden folds in the ground, spa.r.s.ely distributed, lay machine gun posts, like lice in the seams and folds of a garment, to give the attacking force a warm reception. The barrage, which was to have preceded and protected it, went right on somewhere over the enemy's rear positions, while in front the first real line of resistance was not yet carried.' 14 14 The Germans had walked into a sack. This time Foch, even more than Petain, saw the opportunity for a successful counterattack. On 18 July, Crown Prince Wilhelm recalled, 'Without artillery preparation, simply following the sudden rolling barrage, supported by numerous deep-flying aircraft and with unprecedented ma.s.ses of tanks, the enemy infantry - including a number of American divisions - unleashed the storm against the 9th and 7th Armies at 5.40 in the morning.' The Germans had walked into a sack. This time Foch, even more than Petain, saw the opportunity for a successful counterattack. On 18 July, Crown Prince Wilhelm recalled, 'Without artillery preparation, simply following the sudden rolling barrage, supported by numerous deep-flying aircraft and with unprecedented ma.s.ses of tanks, the enemy infantry - including a number of American divisions - unleashed the storm against the 9th and 7th Armies at 5.40 in the morning.'15 Dubbed the second battle of the Marne, the blow drove the Germans back from Chateau-Thierry to Soissons on the River Aisne. Dubbed the second battle of the Marne, the blow drove the Germans back from Chateau-Thierry to Soissons on the River Aisne.

THE TOOLS OF VICTORY.

There were now twenty-five American divisions in France. Pershing's insistence on independence seemed to have confirmed the Germans' expectation that the United States army would not make an effective contribution until 1919. In the event, although it undoubtedly delayed the Americans' impact, it was moderated in practice. Elements of eight divisions took part in the Marne battle, and they did so under temporary French command. The Americans' arrival was speeded by the decision to provide them with British and particularly French equipment, including the 75mm field gun and the Renault light tank. The shipping s.p.a.ce thus saved brought over men instead: 1.5 million US soldiers arrived in Europe in the last six months of the war, with the result that there were forty-two divisions in the field by the time of the armistice and twenty-nine of them had seen action. In the s.p.a.ce of eighteen months the army had grown from 100,000 men to 4 million, and had sent over 2 million overseas. By then the American Expeditionary Force was comparable in size with both the British imperial forces, which totalled 1.8 million in France, and those of France itself, which had fallen from a peak of 2.2 million in July 1916 to 1.7 million.

Equipped by their allies, from British helmets to French light tanks, the Americans go forward to the forest of Argonne, 26 September 1918

The effect of these numbers, and the prodigious effort that had produced them, was above all psychological. In April 1918, when the British army was fighting its desperate defensive action against the second of Ludendorff's offensives, Vera Brittain, serving as a nurse in etaples, saw a contingent of American soldiers march down the road. They looked like 'Tommies in heaven'. 'I pressed forward ... to watch the United States physically entering the War, so G.o.d-like, so magnificent, so splendidly unimpaired in comparison with the tired, nerve-racked men of the British Army'.16 That confidence and self-a.s.surance both helped and hindered the American Expeditionary Force in its adaptation to European warfare. It bred a courage not yet dimmed by age, loss and experience, a product of ignorance and naivety. But what Pershing could not accept was that in losing that vigour, which they, too, had possessed in 1914, the British and French armies had also learnt tactical wisdom. He believed that mobile warfare was the path to victory, that battle should be fought in the open, and that the key to success was aimed rifle-fire. He rejected the views of those who urged that machinery could subst.i.tute for manpower. The American division consisted of 28,000 men, twice the size of those of its allies, which were being restructured as smaller units with fewer men but greater firepower. It was short of lorries and guns, and it proved c.u.mbersome in manoeuvre and poor in its ability to coordinate infantry and artillery. That confidence and self-a.s.surance both helped and hindered the American Expeditionary Force in its adaptation to European warfare. It bred a courage not yet dimmed by age, loss and experience, a product of ignorance and naivety. But what Pershing could not accept was that in losing that vigour, which they, too, had possessed in 1914, the British and French armies had also learnt tactical wisdom. He believed that mobile warfare was the path to victory, that battle should be fought in the open, and that the key to success was aimed rifle-fire. He rejected the views of those who urged that machinery could subst.i.tute for manpower. The American division consisted of 28,000 men, twice the size of those of its allies, which were being restructured as smaller units with fewer men but greater firepower. It was short of lorries and guns, and it proved c.u.mbersome in manoeuvre and poor in its ability to coordinate infantry and artillery.

It also had few tanks. The French had attacked with 375 Renault light tanks on 18 July. These were the brainchild of Jean-Baptiste Estienne, a French artillery officer, whose service in an aviation unit before the war had awoken him to the possibilities of inter-arm tactical cooperation. He encountered his first caterpillar tractor, the key to the tank's cross-country capability, in service with the British army in 1915. On 28 December 1915 he wrote to Joffre urging the development of the tank as an armoured platform for a gun, so as to provide direct fire support for the infantry. Like those under development in Britain at the same time, the first French tanks were heavy machines which were also designed to cross trenches and crush wire. Excessive weight, the result of emphasising armour over mobility, often proved a false friend on the muddy battlefields of France. It also put great mechanical strain on the power-to-weight ratio of the tank. The British models of 1917 and 1918 weighed around 30 tons, and German tank production, based on British types, fell flat on its face because of its pursuit of excessive weight. The breakdown rate of tanks in the First World War was very high - the speed pf most ranged between 2 and 4 mph across country - and their range was restricted. The argument that the British had forfeited their surprise value by using them prematurely on the Somme in September 1916 is nonsense: here was an imperfect but evolving weapon which needed the benefit of combat experience. In essence the tank was only just moving into full series production when the war ended. However, in December 1916 Estienne proposed that tank development move in a radically different direction. He favoured lightness and manoeuvrability over protection, designing a two-man tank, with a weight of 5 tons. Unlike British tanks, which were rhomboidal in shape and carried their main armament on the sides, the Renault had a turret in which was mounted either a machine-gun or a 37mm gun. British tanks required special trains to transport them: the Renault did not, and could therefore be shifted from one sector to another with greater ease and concentrated for the attack in greater secrecy. Over 3,000 entered French service in 1918. The French manufactured about 800 heavy tanks, and the British about 5,000 of all types; about 20 of the monster German A7V were produced, and the handful of tanks the Germans deployed on the western front were mostly captured British models.

The tank was the most striking evidence of a number of points: that the Entente tackled the integration of science, technology and tactics with greater success than the Germans; that the link between tactical experience and factory production was a continuous loop, involving fresh blueprints and the rejigging of machine-tools and plant, as well as feeding munitions into the battle; that by 1918 the Entente, not the Central Powers, derived greater benefit from the trade-off between the ma.s.s army and ma.s.s production; and that the ultimate benefit was on the battlefield, in the reintegration of fire and movement.

The exponential growth in the numbers of aircraft during the course of the war ill.u.s.trated similar arguments. Aerial combat at the start of the war was an affair of individuals, and generated its own heroes, the aces so loved by propaganda and the press. By 1917-18 it was a matter of ma.s.ses, and was therefore sustained as much by the capabilities of its industrial base as by the skills and courage of the pilots who flew the aircraft. In the last year of the war, Britain, France and the United States jointly produced an average of 11,200 machines and 14,500 aero engines per month; the German equivalents were each below 2,000. The corollary of this point was that the air forces were themselves being reshaped. At the start of the war, their role was reconnaissance; by the middle fighters were contesting control of the air above the battlefield; by the end bombers were targeting positions on the ground and interdicting lines of communication. The products of war industries themselves, heavy bombers were beginning to be used to target war production. In 1917 German Gotha heavy bombers raided Paris and London. Spurred to retaliate, the British established the world's first air force on 1 April 1918 and created an 'Independent Force' to target factories, railways and airfields. This strategy set missions which were beyond the capabilities of the existing aircraft, but its implications were already becoming clear: 'I would very much like it', Lord Weir, secretary of state for air, instructed the head of the independent air force, Hugh Trenchard, on 10 September 1918, 'if you could start up a really big fire in one of the German towns. If I were you, I would not be too exacting as regards accuracy in bombing railway stations in the middle of towns. The German is susceptible to bloodiness, and I would not mind a few accidents due to inaccuracy.'17 In practice, bombing against civilian targets produced negligible results in the First World War, but ground attack on the battlefield played a crucial role in checking the German spring offensives in 1918 and in the combined arms offensives of the late summer and autumn. 'The whistling of the falling bombs was like the noise of a thousand door-keys used to hiss a bad play', wrote Rudolf Binding as he fell back to the Aisne on 30 July. 'On explosion they burst into millions of splinters, which flew out horizontally and caused hundreds of casualties.' In practice, bombing against civilian targets produced negligible results in the First World War, but ground attack on the battlefield played a crucial role in checking the German spring offensives in 1918 and in the combined arms offensives of the late summer and autumn. 'The whistling of the falling bombs was like the noise of a thousand door-keys used to hiss a bad play', wrote Rudolf Binding as he fell back to the Aisne on 30 July. 'On explosion they burst into millions of splinters, which flew out horizontally and caused hundreds of casualties.'18 In the last year of the war the aeroplane as well as the tank embodied in one platform the ability both to manoeuvre and to deliver accurate fire on a target. In the last year of the war the aeroplane as well as the tank embodied in one platform the ability both to manoeuvre and to deliver accurate fire on a target.

Neither of them, however, was the true artisan of victory: that was the artillery. The biggest single intellectual shift in making war between 1914 and 1918 was that the combined-arms battle was planned around the capabilities of the guns rather than of the infantry. On 20 November 1917 the British had attacked at Cambrai with 378 tanks; they achieved complete surprise, penetrating up to 4,000 yards on a six-mile front. British tank advocates, including their most effective spokesman after the war, J. F. C. Fuller, who was involved in the planning of the battle, later used the victory as evidence of the decisive independent use of armour. It was not. Fuller himself had written in February 1917: 'It must be remembered that the Creeping Barrage will usually be more effective than the Tank and that the Tank is in no way intended to replace this Barrage but to supplement it when it breaks down or becomes ineffective'.19 By November the Royal Artillery had perfected the techniques of predicted fire. It used microphones to record from different points the low-frequency sound waves following the firing of a gun in order to take a cross-bearing and locate the position of an enemy battery. Unlike aerial observation or the visual spotting of gun flashes, both also methods of immense value in identifying targets, sound-ranging could be used in bad visibility. Consequently, artillery could register its guns in advance of an attack without the preliminary bombardment that had squandered surprise in the past, particularly when the tanks could themselves take on the tasks of crushing wire and destroying enemy machine-gun nests. By November the Royal Artillery had perfected the techniques of predicted fire. It used microphones to record from different points the low-frequency sound waves following the firing of a gun in order to take a cross-bearing and locate the position of an enemy battery. Unlike aerial observation or the visual spotting of gun flashes, both also methods of immense value in identifying targets, sound-ranging could be used in bad visibility. Consequently, artillery could register its guns in advance of an attack without the preliminary bombardment that had squandered surprise in the past, particularly when the tanks could themselves take on the tasks of crushing wire and destroying enemy machine-gun nests.

The success at Cambrai was the product of the short bombardment as much as of the ma.s.sed use of tanks. These were the principles carried forward into 1918. In the last year of the war bombardments were short, but their effects were greater than those of the long bombardments of 1916 and early 1917. First, second and third lines of defence could be isolated from each other by curtains of fire, which moved forward or back according to plan, matching the type of sh.e.l.l to the nature of the target. Confronted with mortal danger, and cut off from resupply or relief, the defenders had to respond in the most unnatural way of all: 'In absolute darkness we simply lay and trembled from sheer nervous tension'. Industrialised war enforced pa.s.sivity, as one of Pershing's officers, Hervey Allen, found out: 'There is a faraway moan that grows to a scream, then a roar like a train, followed by a ground-shaking smash and a diabolical red light.... Everybody simply shakes and crawls . . . A hunching of the shoulders and then another comes, and the thought - How long, how long? There is nothing to do. Whether you get through or not is just sheer chance and nothing more.'20 Accuracy of fire meant less wasted effort, but quant.i.ty still had a part to play. More guns were available, especially heavy ones. In 1918 France's holdings of field artillery, 5,000 guns, were comparable with those of 1914, but those of heavy artillery had risen from 300 to 5,700. Britain manufactured 3,226 guns in 1915 but 10,680 in 1918. Therefore as many sh.e.l.ls could be fired in a short period as could be fired in a longer period two years previously. Sh.e.l.l production was now always ahead of sh.e.l.l consumption. Petain focused on mobility to create concentrations of fire, but in the British case abundance meant that the guns did not necessarily have to be moved to create a local concentration, so attacks could be launched in rapid succession at different points of the line with minimum delay. At the end of the war, the French artillery const.i.tuted 37 per cent of the army, as opposed to 18 per cent in May 1915, and the Royal Artillery mustered half a million men, and const.i.tuted a quarter of the British army. Germany's artillery strength was greater than both Britain's and France's combined at the beginning of 1918, but by November was comparable with Britain's alone. Deliveries of new field guns fell from 3,000 per month in 1917 to 2,000 in February 1918 and 1,200 in September. Having had 7,130 guns at the front in February 1917, Germany had only 6,172 a year later. Lack of men and horses to provide gun crews was a more important reason for this decline than falling productivity, although this too had its impact: from July 1918 the monthly output of sh.e.l.ls was half that of 1917. In a war in which 70 per cent of all casualties were attributed to artillery it was a fatal weakness.

THE ENTENTE'S VICTORIES 'The Boche holds firm,' wrote Charles Mangin, commander of the Soissons counterattack, to his wife on 28 July 1918.21 That was not how it seemed on the German side of the line. Opposite him, Crown Prince Wilhelm said the whole position had changed. His fellow army group commander, Crown Prince Rupprecht, concluded on 20 July that 'we stand at the turning point of the war: what I expected first for the autumn, the necessity to go over to the defensive, is already on us, and in addition all the gains which we made in the spring - such as they were - have been lost again'. But Ludendorff refused to confront realities, defying the advice of his senior commanders that they pull back or even negotiate, and ignoring the evidence of the collapsing morale of his troops. 'Poor provisions, heavy losses and the deepening influenza have deeply depressed the spirits of the men in the III Infantry Division', Rupprecht wrote on 3 August. Postal censors told him that letters home complained of the mounting numbers of Americans and of British aerial domination, and - even more importantly - called for peace in ways which linked front and rear; war, they said, was the product of capitalism, and 'at home, they must strike and strike hard, and cause a revolution, and then peace must come'. That was not how it seemed on the German side of the line. Opposite him, Crown Prince Wilhelm said the whole position had changed. His fellow army group commander, Crown Prince Rupprecht, concluded on 20 July that 'we stand at the turning point of the war: what I expected first for the autumn, the necessity to go over to the defensive, is already on us, and in addition all the gains which we made in the spring - such as they were - have been lost again'. But Ludendorff refused to confront realities, defying the advice of his senior commanders that they pull back or even negotiate, and ignoring the evidence of the collapsing morale of his troops. 'Poor provisions, heavy losses and the deepening influenza have deeply depressed the spirits of the men in the III Infantry Division', Rupprecht wrote on 3 August. Postal censors told him that letters home complained of the mounting numbers of Americans and of British aerial domination, and - even more importantly - called for peace in ways which linked front and rear; war, they said, was the product of capitalism, and 'at home, they must strike and strike hard, and cause a revolution, and then peace must come'.22 n.o.body on the allied side had yet realised that victory was possible this side of Christmas. In London the war cabinet was making preparations for 1919. In France Foch convened a conference of the national army commanders on 24 July. He appreciated that the opportunity to take the initiative had now arrived; morally, materially and numerically the allies were in the ascendant on the western front. But even he, ebullient spokesman of the offensive that he was, rejected a single decisive blow. Instead he envisaged a series of limited attacks. Their aims would be to free the princ.i.p.al railway lines radiating out from Paris, to regain the economic heartlands of France and above all to improve the armies' lines of communication for the next, more mobile, phase of operations. 'These movements should be executed with such rapidity as to inflict upon the enemy a succession of blows. This condition necessarily limits their extent . . . These actions must succeed each other at brief intervals, so as to embarra.s.s the enemy in the utilization of his reserves and not allow him sufficient time to fill up his units.'23 Here was a concept of operations which harnessed the Entente's superior resources to the constraints of trench war. Attacks would go no deeper into the enemy's positions than the reach of the artillery supporting them. By rapidly switching the axis of the advance, Foch would pull the enemy first in one direction and then in another. Most important of all, it was a framework to which Pershing, Petain and Haig were prepared to subscribe. The first recognised that the inexperience of his army required him to go gently; the other two needed to husband their forces after the battering of the first half of 1918. Foch was being realistic: he himself had only a skeleton staff, and his powers extended little further than any consensus he could forge. Individual blows, coordinated in time and s.p.a.ce, allowed each national army to do its own thing. Here was a concept of operations which harnessed the Entente's superior resources to the constraints of trench war. Attacks would go no deeper into the enemy's positions than the reach of the artillery supporting them. By rapidly switching the axis of the advance, Foch would pull the enemy first in one direction and then in another. Most important of all, it was a framework to which Pershing, Petain and Haig were prepared to subscribe. The first recognised that the inexperience of his army required him to go gently; the other two needed to husband their forces after the battering of the first half of 1918. Foch was being realistic: he himself had only a skeleton staff, and his powers extended little further than any consensus he could forge. Individual blows, coordinated in time and s.p.a.ce, allowed each national army to do its own thing.

Haig had drawn up a plan for a limited offensive at Amiens, a direct consequence of his precipitate withdrawal in March. The German line here lacked deep defensive positions. Orchestrated by Rawlinson, who had been recalled to the command of the 4th Army, and supported by the French to the south, the attack maximised firepower and method. Each gun was allocated twenty-five yards of trench, and thanks in part to allied aerial superiority (1,800 British and French aircraft were a.s.sembled) 95 per cent of German batteries were located in advance. On the Somme British battalions had numbered 1,000 men and were equipped with four Lewis light machine-guns and one or two light trench mortars. At Amiens they numbered 500 men but had thirty Lewis guns, eight mortars and - if in the first line - were accompanied by six tanks.24 In all over 400 tanks took part in the attack on 8 August; by the 9th only 145 were fit for action. Mechanical problems were their princ.i.p.al defect, but the Germans also learnt to overcome what Ludendorff called 'tank fright'. Stiffer anti-tank defences contributed to such high losses that never again in the war did more than 150 tanks go into action at any one time. In all over 400 tanks took part in the attack on 8 August; by the 9th only 145 were fit for action. Mechanical problems were their princ.i.p.al defect, but the Germans also learnt to overcome what Ludendorff called 'tank fright'. Stiffer anti-tank defences contributed to such high losses that never again in the war did more than 150 tanks go into action at any one time.

Ludendorff dubbed 8 August 'the black day of the German army'. Of 27,000 German casualties, fully 12,000 had surrendered, an unprecedentedly high proportion. But the German army did not cease fighting thereafter. Front-line resistance continued until the armistice; the problems of desertion and disobedience were more in its rear, on the lines of communication and at home. The significance of Amiens lay as much in its shock effect on Luden-dorff, who at last woke up to what others had been saying for weeks. But when the Kaiser summoned his princ.i.p.al political and military leaders to a council at Spa on 13 and 14 August, Ludendorff failed to present a realistic appraisal of the military situation. Instead he blamed the feeling at home. Over the next six weeks his mood showed alarming swings, as he oscillated between unfounded optimism and the search for scapegoats other than himself. A psychologist told him he needed to rest and to sing German folk songs on waking in the morning. At Spa, Wilhelm thought it sensible to seek an intermediary in order to open peace negotiations. But his foreign minister, Paul von Hintze, still made such an approach conditional on the next German victory, and could not resolve definitively to abandon Germany's claims to Belgium and Alsace-Lorraine. A week later, Hintze told the party leaders that the army believed that 'there was no reason to doubt ultimate victory. We shall be vanquished only when we doubt that we shall win'.25 Soldiers of the Devonshire Regiment, part of the British 62nd Division serving in the French sector, capture a German in the Bois de Reims during the battle of Tardenois in late July 1918.

The Austrians did doubt it. Karl, accompanied by Count Burian, Czernin's successor as foreign minister, and Arz, attended the Spa conference and said that direct negotiations were needed as soon as possible. The Entente had been cautious about the dismemberment of Austria-Hungary, which if united would at least provide a counterweight to Germany in central Europe. In January 1918 Wilson's tenth point had allowed for the 'autonomous' development of the peoples of Austria-Hungary. But his thirteenth provided specifically for an independent Polish state, and, although he had been more cautious about Czechoslovakia, by the summer it, too, had received de facto recognition. With the coalescence of their war aims, the allies were able to coordinate their efforts in propaganda as in other spheres. This was an approach that appealed to the Italian high command. Throughout 1918 Armando Diaz resisted pressure from the Supreme War Council for an Italian offensive, and put his effort into the subversion of the Austro-Hungarian army instead. Ideas were less costly in lives than bullets, and so did not threaten the fragile morale of the Italian army after Caporetto. The Italians were not very successful: in June 1918 the Austro-Hungarian army attacked on the Piave, and the head of the British military mission reported that 'as a whole the enemy troops . . . are fighting with great determination'.26 The attack failed, but not primarily because of desertion or nationalist sentiment. The men were exhausted, without equipment and, more importantly, food. They had dubbed the attack on the Piave the 'bread offensive'. The pressures to which they were succ.u.mbing came from behind them rather than from in front; as in the German army, grievances at home were merging with grievances in the army. Allied propaganda aimed at the subject nationalities was blamed for strikes and mutinies generated by other causes. In August malaria added to the woes of the Austro-Hungarian army in Italy: two-thirds of its divisions were below half strength. When Hintze told Karl and his advisers on 3 September that Ludendorff was predicting eventual success on the western front, they decided on unilateral action. On 14 September Burian asked for talks on peace terms. The allies rejected him: in a belated bid to save the Dual Monarchy the Austro-Hungarians had at last resolved to abandon the alliance with Germany, but as in 1914 the Entente leaders saw the two as indivisible. The attack failed, but not primarily because of desertion or nationalist sentiment. The men were exhausted, without equipment and, more importantly, food. They had dubbed the attack on the Piave the 'bread offensive'. The pressures to which they were succ.u.mbing came from behind them rather than from in front; as in the German army, grievances at home were merging with grievances in the army. Allied propaganda aimed at the subject nationalities was blamed for strikes and mutinies generated by other causes. In August malaria added to the woes of the Austro-Hungarian army in Italy: two-thirds of its divisions were below half strength. When Hintze told Karl and his advisers on 3 September that Ludendorff was predicting eventual success on the western front, they decided on unilateral action. On 14 September Burian asked for talks on peace terms. The allies rejected him: in a belated bid to save the Dual Monarchy the Austro-Hungarians had at last resolved to abandon the alliance with Germany, but as in 1914 the Entente leaders saw the two as indivisible.

Haig had been surprised by the scale of his success at Amiens and had had to hold Foch back from pushing the offensive beyond its limits. But both commanders now realised, as Haig told a sceptical Winston Churchill on 21 August, that 'we ought to do our utmost to get a decision this autumn'.27 The American army was capable of independent operations, as it showed when it pinched out the St Mihiel salient south of Verdun between 12 and 16 September. It got the early success that Pershing had seen as essential to buoy the morale of his young army; it had done so in part because the Germans had taken the decision two days earlier not to hold the salient. But, rather than continue the advance, it was then redeployed, with considerable logistical difficulty, to the west of Verdun, to form the right flank of a joint attack with the French, going north through the Argonne forest. Foch hoped to get into the guts of the German rear areas from below. Beginning on 26 September and continuing through October, the French and American armies punched their way through deep German defences, the Americans' right flank following the line of the Meuse valley. The American army was capable of independent operations, as it showed when it pinched out the St Mihiel salient south of Verdun between 12 and 16 September. It got the early success that Pershing had seen as essential to buoy the morale of his young army; it had done so in part because the Germans had taken the decision two days earlier not to hold the salient. But, rather than continue the advance, it was then redeployed, with considerable logistical difficulty, to the west of Verdun, to form the right flank of a joint attack with the French, going north through the Argonne forest. Foch hoped to get into the guts of the German rear areas from below. Beginning on 26 September and continuing through October, the French and American armies punched their way through deep German defences, the Americans' right flank following the line of the Meuse valley.

The British attacked on the other side of the enormous salient that made up the German position in France. The 1st and 3rd Armies took over from Rawlinson's 4th in continuous fighting from late August into September. The battlefield was familiar from two years previously: 'Then we attacked en ma.s.se ma.s.se from the south-east, and fought for yards of ground thickly held by an enemy with no thought of retreat. Now the front in movement was wide and elastic, the fighting was open, and we were attacking positions from the flank.' from the south-east, and fought for yards of ground thickly held by an enemy with no thought of retreat. Now the front in movement was wide and elastic, the fighting was open, and we were attacking positions from the flank.'28 The Germans fell back to the Hindenburg line. Formidable obstacle though it was, it had been built at the end of 1916. It was an indication of the war's dynamism and the pace of its tactical innovation that its principles were now obsolescent. Made up of six defensive lines, it formed a zone 6,000 yards deep, with concrete emplacements and belts of barbed wire. But the southern part included the St Quentin ca.n.a.l, and did not therefore rest on a reverse slope out of artillery observation, and the whole was linear in design when more recent constructions had been made up of strongpoints, arranged in chequer-board fashion to create converging fields of fire. Unlike Amiens the Hindenburg line was strongly held, and unlike Cambrai it presented no opportunity for surprise. The answer was a 56-hour artillery bombardment, using 1,637 guns on a 10,000-yard front, twice the density of the Somme, and targeting the defences rather than the defenders. In the last twenty-four hours the British fired a record 945,052 sh.e.l.ls. The capture of the ca.n.a.l by the 46th Division was one of the great feats of arms of the war, helped by heavy early-morning fog and a creeping barrage which rained down 126 sh.e.l.ls for every 500 yards of German trench for eight hours. The Germans fell back to the Hindenburg line. Formidable obstacle though it was, it had been built at the end of 1916. It was an indication of the war's dynamism and the pace of its tactical innovation that its principles were now obsolescent. Made up of six defensive lines, it formed a zone 6,000 yards deep, with concrete emplacements and belts of barbed wire. But the southern part included the St Quentin ca.n.a.l, and did not therefore rest on a reverse slope out of artillery observation, and the whole was linear in design when more recent constructions had been made up of strongpoints, arranged in chequer-board fashion to create converging fields of fire. Unlike Amiens the Hindenburg line was strongly held, and unlike Cambrai it presented no opportunity for surprise. The answer was a 56-hour artillery bombardment, using 1,637 guns on a 10,000-yard front, twice the density of the Somme, and targeting the defences rather than the defenders. In the last twenty-four hours the British fired a record 945,052 sh.e.l.ls. The capture of the ca.n.a.l by the 46th Division was one of the great feats of arms of the war, helped by heavy early-morning fog and a creeping barrage which rained down 126 sh.e.l.ls for every 500 yards of German trench for eight hours.29 The Hindenburg line was breached on 29 September. The Hindenburg line was breached on 29 September.

For three days the entire western front had been under coordinated allied attack. To the north of the Hindenburg line the Canadians in the 1st Army, supported by the 3rd Army on their right, had crossed the Ca.n.a.l du Nord, and above them two more British armies and - for the first time since 1914 - the Belgians were pushing into Flanders. Paul von Hintze, briefed on the realities of the military situation by Ludendorff's subordinates, realised that defeat was likely to precipitate a revolution. To avert it, he proposed a 'revolution from above'. Germany's government should be reformed on more democratic lines as a preliminary to any peace negotiations. This might achieve two objectives: it would preserve the monarchy and it might channel the opprobrium for defeat on to the left and not the right.

These careful calculations needed time to be put into effect. They were wrong-footed by Ludendorff. On the night of 28 September his nerve cracked: he fell to the floor and according to some accounts foamed at the mouth. At a crown council convened at Spa on the following morning, he demanded an immediate armistice on the basis of Wilson's Fourteen Points. He had suddenly changed his a.s.sessment of the military situation, and as a result had at last usurped the political process. Georg von Hertling, who reached the meeting too late to play an active part, resigned the chancellorship. His successor, Max von Baden, arrived in Berlin to take up office at 4 a.m. on 1 October, and promptly made it clear that his whole policy depended on holding out for some while longer. He protested that Ludendorff was inverting the logical sequence: 'a request for an armistice makes any peace initiative impossible'.30 Germany was in danger of forfeiting its long-term powers of negotiation because the army insisted it needed a short-term breathing s.p.a.ce. Germany was in danger of forfeiting its long-term powers of negotiation because the army insisted it needed a short-term breathing s.p.a.ce.

The advance east of Arras: Canadians pa.s.s through a burning street in Cambrai, 9 October 1918 Dominion forces spearheaded many of the British attacks of the last hundred days of the war.

None the less, Ludendorff's erratic behaviour was not in any immediate sense the product of the situation on the western front. 'Deep but partial break-ins; on the whole the front still holds', one of his staff officers wrote in his diary on 28 September. 'I believe that we are still parrying the a.s.sault this time.'31 The crisis was triggered by the news that Bulgaria had sought and been granted an armistice. In some respects the First World War ended where it began, in the Balkans. But the impact of Bulgaria's decision makes another more substantial point: that in the First World War no front stood in total isolation from another. The crisis was triggered by the news that Bulgaria had sought and been granted an armistice. In some respects the First World War ended where it began, in the Balkans. But the impact of Bulgaria's decision makes another more substantial point: that in the First World War no front stood in total isolation from another.

The Salonika front had been locked in stalemate for much of 1916 and 1917. For the French and British troops there it may have been a side-show, a theatre of war which exposed them to boredom, extremes of weather, and disease - above all malaria: British non-battle casualties in Macedonia exceeded battle casualties by twenty to one. But for their allies this was where the Third Balkan War now had its focus. The Serb army, evacuated from Albania via Corfu, represented the nation in exile and carried its hopes for a greater Serbia. To their left, at Valona on the Adriatic coast, were the Italians, harbouring ambitions of conquest in Dalmatia and so set to rival Serbia. To their right were the Greeks, who had failed to support Serbia in 1914 but who had been dragooned into joining the Entente in June 1917. King Constantine's pragmatic neutrality in 1915 had been undermined by his prime minister, Eleutherios Venizelos. Venizelos argued that joining the Entente would open the door to a greater Greece, and in 1916 secured the Entente's recognition for a provisional government in Thessalonica. In 1917, with a disregard for neutrality which accorded ill with their defence of Belgium, the British blockaded Greece and the French landed at Piraeus; Constantine abdicated.

The mountainous front that faced this multi-national force was naturally strong, but very hard to supply. Moreover, the occupants, the Bulgarians, had lost their reasons for fighting. Their war aims had been gained with the conquests of Serbia in 1915 and Romania in 1916. They were now fighting to meet German objectives, but the Germans showed them little generosity. They were not included in the deal cut at Brest-Litovsk. They hoped that the Treaty of Bucharest would give them all Dobrudja and its grain. It did not. The north was put under joint German-Austrian-Bulgarian administration. While the Bulgarians starved, the German troops in Bulgaria bought up supplies for transport to Germany. The withdrawal of German divisions for the western front prompted warnings that the Central Powers' position in the Balkans was now dangerously exposed. Hindenburg and Ludendorff were powerless to respond: they had not got the men and they refused to compromise on Dobrudja. On 17 September they simply signalled that 'it must be accepted that the decision of the whole war' rested on the western front.32 At Megiddo. Allenby enjoyed at least a two-to-one superiority. The transport of the British 5th Cavalry Division crosses the River Auja.

By then the Serbs were ten miles into the Bulgarian positions and about to prove them wrong. Entente forces under the French general, Louis Franchet d'Esperey, 'Desperate Frankie' to the British, had attacked on 15 September. By 22 September the campaign was mobile enough for the continuing value of horses to become evident in another theatre, as well as Palestine. A brigade of 3,000 French cavalry, Spahis from North Africa, covered sixty miles over terrain which rose to 5,000 feet above sea level, entering Skopje, inside Serbia, on 29 September, the day that Bulgaria agreed terms. Its comm