Editing Battle of Verdun (section)

==Aftermath==

===Analysis===
Falkenhayn wrote in his memoirs that he sent an appreciation of the strategic situation to the Kaiser in December 1915,

{{blockquote|The string in France has reached breaking point. A mass breakthrough—which in any case is beyond our means—is unnecessary. Within our reach there are objectives for the retention of which the French General Staff would be compelled to throw in every man they have. If they do so the forces of France will bleed to death.|Falkenhayn{{sfn|Falkenhayn|2004|pp=217–218}}}}

The German strategy in 1916 was to inflict mass casualties on the French, a goal achieved against the Russians from 1914 to 1915, to weaken the French Army to the point of collapse. The French had to be drawn into circumstances from which the Army could not escape for reasons of strategy and prestige. The Germans planned to use a large number of heavy and super-heavy guns to inflict a greater number of casualties than French artillery, which relied mostly upon the {{nowrap|75 mm field}} gun. In 2007, Robert Foley wrote that Falkenhayn intended a battle of [[attrition warfare|attrition]] from the beginning, contrary to the views of Wolfgang Foerster in 1937, Gerd Krumeich in 1996 and other historians but the loss of documents led to many interpretations of the strategy. In 1916, critics of Falkenhayn claimed that the battle demonstrated that he was indecisive and unfit for command, echoed by Foerster in 1937.{{sfn|Förster|1937|pp=304–330}} In 1994, Holger Afflerbach questioned the authenticity of the "Christmas Memorandum"; after studying the evidence that had survived in the {{lang|de|Kriegsgeschichtliche Forschungsanstalt des Heeres}} (Army Military History Research Institute) files, he concluded that the memorandum had been written after the war but that it was an accurate reflection of Falkenhayn's thinking at the end of 1915.{{sfn|Afflerbach|1994|pp=543–545}}

[[File:River Crossing NGM-v31-p338.jpg|thumb|{{centre|French [[Train (military)|train]] horses resting in a river on their way to Verdun}}]]
Krumeich wrote that the Christmas Memorandum was fabricated to justify a failed strategy and that attrition had been substituted for the capture of Verdun only after the attack failed.{{sfn|Krumeich|1996|pp=17–29}} Foley wrote that after the failure of the Ypres Offensive of 1914, Falkenhayn had returned to the pre-war strategic thinking of [[Helmuth von Moltke the Elder|Moltke the Elder]] and [[Hans Delbrück]] on {{lang|de|Ermattungsstrategie}} (attrition strategy), because the coalition fighting Germany was too powerful to be defeated. Falkenhayn wanted to divide the Allies by forcing at least one of the [[Triple Entente|Entente]] powers into a negotiated peace. Attrition lay behind the offensive in the east in 1915 but the Russians had refused to accept German [[peace feeler]]s, despite the huge defeats inflicted on them by the Austro-Germans.{{sfn|Foley|2007|pp=206–207}}

With insufficient forces to break through the Western Front and to overcome the reserves behind it, Falkenhayn tried to force the French to attack instead, by threatening a sensitive point close to the front line and chose Verdun as the place. Huge losses were to be inflicted on the French by German artillery on the dominating heights around the city. The 5th Army would begin a big offensive but with the objectives limited to seizing the Meuse Heights on the east bank, on which the German heavy artillery would dominate the battlefield. The French Army would "bleed itself white" in hopeless attempts to recapture the heights. The British would be forced to launch a hasty relief offensive and suffer an equally costly defeat. If the French refused to negotiate, a German offensive would mop up the remnants of the Franco–British armies, breaking the Entente "once and for all".{{sfn|Foley|2007|pp=206–207}}

In a revised instruction to the French Army in January 1916, the General Staff (GQG) wrote that equipment could not be fought by men. Firepower could conserve infantry but attrition prolonged the war and consumed troops that had been preserved in earlier battles. In 1915 and early 1916, German industry quintupled the output of heavy artillery and doubled the production of super-heavy artillery. French production had also recovered since 1914 and by February 1916 the army had {{nowrap|3,500 heavy}} guns. In May Joffre began to issue each division with two groups of {{nowrap|155 mm guns}} and each corps with four groups of long-range guns. Both sides at Verdun had the means to fire huge numbers of heavy shells to suppress the opposing defences before risking infantry in the open. At the end of May, the Germans had {{nowrap|1,730 heavy}} guns at Verdun and the {{nowrap|French 548,}} sufficient to contain the Germans but not enough for a counter-offensive.{{sfn|Jankowski|2014|pp=109–112}}

[[File:Nieuport N.16 C.1 with Le Prieur rockets.jpg|thumb|{{centre|Nieuport 16 fighter in camouflage adopted during the Battle of Verdun}}]]
French infantry survived bombardment better because their positions were dispersed and tended to be on dominating ground, not always visible to the Germans. As soon as a German attack began, the French replied with machine-gun and rapid field-artillery fire. On 22 April, the Germans suffered {{nowrap|1,000 casualties}} and in mid-April, the French fired {{nowrap|26,000 field}} artillery shells against an attack to the south-east of Fort Douaumont. A few days after taking over at Verdun, Pétain ordered the air commander, Commandant [[Charles Tricornot de Rose]] to sweep away German fighter aircraft and to provide artillery observation. German air superiority was reversed by concentrating the French fighters in {{lang|fr|[[escadrille]]s}} rather than distributing them piecemeal across the front, unable to concentrate against large German formations. The fighter escadrilles drove away the German {{lang|de|[[Fokker Eindecker]]}}s and the two-seater reconnaissance and artillery-observation aircraft that they protected.{{sfn|Davilla|Soltan|1997|p=7}}

The fighting at Verdun was less costly to both sides than the war of movement in 1914, when the French suffered {{circa| 850,000 casualties}} and the Germans {{circa| 670,000}} from August to the end of 1914. The 5th Army had a lower rate of loss than armies on the Eastern Front in 1915 and the French had a lower average rate of loss at Verdun than the rate over three weeks during the Second Battle of Champagne (September–October 1915), which were not deliberately fought as battles of attrition. German loss rates increased relative to losses from {{nowrap|1:2.2 in}} early 1915 to close to {{nowrap|1:1 by}} the end of the battle, a trend which continued during the [[Nivelle Offensive]] in 1917. The penalty of attrition tactics was indecision, because limited-objective attacks under an umbrella of massed heavy artillery fire could succeed but led to battles of unlimited duration.{{sfn|Jankowski|2014|pp=114–120}} Pétain used a {{lang|ar|[[noria]]}} (rotation) system quickly to relieve French troops at Verdun, which involved most of the French Army in the battle but for shorter periods than the German troops in the 5th Army. The symbolic importance of Verdun proved a rallying point and the French did not collapse. Falkenhayn was forced to conduct the offensive for much longer and commit far more infantry than intended. By the end of April, most of the German strategic reserve was at Verdun, suffering similar casualties to the French army.{{sfn|Foley|2007|p=256}}

The Germans believed that they were inflicting losses at a rate of {{nowrap|5:2; German}} military intelligence thought that by 11 March the French had suffered {{nowrap|100,000 casualties}} and Falkenhayn was confident that German artillery could easily inflict another {{nowrap|100,000 losses.}} In May, Falkenhayn estimated that French casualties had increased to {{nowrap|525,000 men}} against {{nowrap|250,000 German}} and that the French strategic reserve was down to {{nowrap|300,000 men.}} Actual French losses were {{circa| 130,000}} by 1 May; {{nowrap|42 French}} divisions had been withdrawn and rested by the {{lang|ar|noria}} system, once infantry casualties reached {{nowrap|50 per cent.}} Of the {{nowrap|330 infantry}} battalions of the French metropolitan army, {{nowrap|259 (78 per cent)}} went to Verdun, against {{nowrap|48 German}} divisions, {{nowrap|25 per cent}} of the {{lang|de|Westheer}} (western army).{{sfn|Clayton|2003|pp=120–121}} Afflerbach wrote that {{nowrap|85 French}} divisions fought at Verdun and that from February to August, the ratio of German to French losses was {{nowrap|1:1.1,}} not the third of French losses assumed by Falkenhayn.{{sfn|Chickering|Förster|2006|pp=130, 126}} By 31 August, the 5th Army had suffered {{nowrap|281,000 casualties}} and the French {{nowrap|315,000.}}{{sfn|Foley|2007|p=256}}

[[Image:French soldiers of the 87th Regiment shelter in their trenches at Hill 304 at Verdun.jpg|thumb|upright=0.8|{{centre|French trench at Côte 304, Verdun}}]]

In June 1916, the French had {{nowrap|2,708 guns}} at Verdun, including {{nowrap|1,138 field}} guns; from February to December, the French and German armies fired {{circa| 10,000,000 shells,}} weighing {{cvt|1350000|LT}}.{{sfn|Mason|2000|p=185}} By May, the German offensive had been defeated by French reinforcements, difficulties of terrain and the weather. The 5th Army infantry was stuck in tactically dangerous positions, overlooked by the French on both banks of the Meuse, instead of dug in on the Meuse Heights. French casualties were inflicted by constant infantry attacks which were far more costly in men than destroying counter-attacks with artillery. The stalemate was broken by the Brusilov Offensive and the Anglo-French relief offensive on the Somme, which Falkehayn had expected to begin the collapse of the Anglo-French armies.{{sfn|Foley|2007|pp=235–236}} Falkenhayn had begun to remove divisions from the Western Front in June for the strategic reserve but only twelve divisions could be spared. Four divisions were sent to the Somme, where three defensive positions had been built, based on the experience of the {{lang|de|Herbstschlacht}}. Before the battle on the Somme began, Falkenhayn thought that German preparations were better than ever and the British offensive would easily be defeated. The 6th Army, further north, had {{frac|17|1|2}} divisions and plenty of heavy artillery, ready to attack once the British had been defeated.{{sfn|Foley|2007|pp=249–250}}

The strength of the Anglo-French attack on the Somme surprised Falkenhayn and his staff, despite the British casualties on 1 July. Artillery losses to "overwhelming" Anglo-French counter-battery fire and the German tactic of instant counter-attacks, led to far more German infantry casualties than at the height of the fighting at Verdun, where the 5th Army suffered {{nowrap|25,989 casualties}} in the first ten days, against {{nowrap|40,187 2nd}} Army casualties on the Somme. The Russians attacked again, causing more casualties in June and July. Falkenhayn was called on to justify his strategy to the Kaiser on 8 July and again advocated the minimal reinforcement of the east in favour of the "decisive" battle in France; the Somme offensive was the "last throw of the dice" for the Entente. Falkenhayn had already given up the plan for a counter-offensive by the 6th Army and sent 18 divisions to the 2nd Army and to the Russian front from the reserve and from the 6th Army; only one division remaining uncommitted by the end of August. The 5th Army had been ordered to limit its attacks at Verdun in June but a final effort was made in July to capture Fort Souville. The attack failed and on 12 July Falkenhayn ordered a strict defensive policy, permitting only small local attacks to limit the number of troops the French could transfer to the Somme.{{sfn|Foley|2007|pp=251–254}}

Falkenhayn had underestimated the French, for whom victory at all costs was the only way to justify the sacrifices already made; the French army never came close to collapsing and causing a premature British relief offensive. The ability of the German army to inflict disproportionate losses had also been overestimated, in part because the 5th Army commanders had tried to capture Verdun and attacked regardless of loss. Even when reconciled to the attrition strategy, they continued with {{lang|de|Vernichtungsstrategie}} (strategy of annihilation) and the tactics of {{lang|de|Bewegungskrieg}} (manoeuvre warfare). Failure to reach the Meuse Heights left the 5th Army in poor tactical positions and reduced to inflicting casualties by infantry attacks and counter-attacks. The length of the offensive made Verdun a matter of prestige for the Germans as it was for the French and Falkenhayn became dependent on a British relief offensive being destroyed to end the stalemate. When it came, the collapse in Russia and the power of the Anglo-French attack on the Somme reduced the German armies to holding their positions as best they could.{{sfn|Foley|2007|pp=254–256}} On 29 August, Falkenhayn was sacked and replaced by Hindenburg and Ludendorff, who ended the German offensive at Verdun on 2 September.{{sfn|Foley|2007|p=258}}{{efn|Pétain praised what he saw as the success of the fortifications at Verdun in ''La Bataille de Verdun'' (1929) and in 1930, when building the [[Maginot Line]] ({{lang|fr|Ligne Maginot}}) along the border with Germany began. At Verdun, French field artillery in the open outnumbered turreted guns in the Verdun forts by at least {{nowrap|200:1.}} It was the mass of French field artillery (over {{nowrap|2,000 guns}} after May 1916) that inflicted about {{nowrap|70 per cent}} of German infantry casualties. In 1935, a number of mechanised and motorised units were deployed behind the Maginot Line and plans were laid to send detachments to fight a mobile defence in front of the fortifications.{{sfn|Wynne|1976|p=329}} At the [[Battle of Dien Bien Phu]] (1953–1954), General [[Christian de Castries]] said that the situation was "somewhat like Verdun". French forces at Dien Bien Phu were supplied by transport aircraft, using a landing strip in range of Viet Minh artillery; the French forces at Verdun were supplied by road and rail, beyond the reach of German artillery.{{sfn|Windrow|2004|p=499}}}}

===Casualties===
[[File:Lot-3664-13 (32980411840).jpg|thumb|{{center|The remains of soldiers recovered in 1919.}}]]
In 2013, Paul Jankowski wrote that since the beginning of the war, French army units had produced numerical loss states ({{lang|fr|états numériques des pertes}}) every five days for the Bureau of Personnel at GQG. The Health Service ({{lang|fr|Service de Santé}}) at the Ministry of War received daily counts of wounded taken in by hospitals and other services but casualty data was dispersed among regimental depots, GQG, the Registry Office ({{lang|fr|État Civil}}), which recorded deaths, the {{lang|fr|Service de Santé}}, which counted injuries and illnesses and {{lang|fr|Renseignements aux Familles}} (Family Liaison), which communicated with next of kin. Regimental depots were ordered to keep {{lang|fr|fiches de position}} (position sheets) to record losses continuously and the {{lang|fr|Première Bureau}} of GQG began to compare the five-day {{lang|fr|états numériques des pertes}} with the records of hospital admissions. The new system was used to calculate losses back to August 1914, which took several months; the system had become established by February 1916. The {{lang|fr|états numériques des pertes}} were used to calculate casualty figures published in the {{lang|fr|Journal Officiel}}, the French Official History and other publications.{{sfn|Jankowski|2014|pp=257–258}}

The German armies compiled {{lang|de|Verlustlisten}} (loss lists) every ten days, which were published by the {{lang|de|Reichsarchiv}} in the {{lang|de|deutsches Jahrbuch}} of 1924–1925. German medical units kept detailed records of medical treatment at the front and in hospital and in 1923 the {{lang|de|Zentral Nachweiseamt}} (Central Information Office) published an amended edition of the lists produced during the war, incorporating medical service data not in the {{lang|de|Verlustlisten}}. Monthly figures of wounded and ill servicemen that received medical treatment were published in 1934 in the {{lang|de|Sanitätsbericht}} (Medical Report). Using such sources for comparison is difficult because the information recorded losses over time, rather than place. Losses calculated for a battle could be inconsistent, as in the ''Statistics of the Military Effort of the British Empire during the Great War 1914–1920'' (1922). In the early 1920s, [[Louis Marin (politician)|Louis Marin]] reported to the Chamber of Deputies but could not give figures per battle, except for some by using numerical reports from the armies, which were unreliable unless reconciled with the system established in 1916.{{sfn|Jankowski|2014|pp=258–259}}

Some French data excluded those lightly wounded but some did not. In April 1917, GQG required that the {{lang|fr|états numériques des pertes}} discriminate between lightly wounded, treated locally for 20 to 30 days and severely wounded evacuated to hospitals. Uncertainty over the criteria had not been resolved before the war ended. {{lang|de|Verlustlisten}} excluded lightly wounded and the {{lang|de|Zentral Nachweiseamt}} records included them. Churchill revised German statistics by adding {{nowrap|2 per cent}} for unrecorded wounded in ''The World Crisis'', written in the 1920s and [[James Edward Edmonds|James Edmonds]], the British official historian, {{nowrap|added 30 per cent.}} For the Battle of Verdun, the {{lang|de|Sanitätsbericht}} contained incomplete data for the Verdun area, did not define "wounded" and the 5th Army field reports exclude them. The Marin Report and {{lang|fr|Service de Santé}} covered different periods but included lightly wounded. Churchill used a {{lang|de|Reichsarchiv}} figure of {{nowrap|428,000 casualties}} and took a figure of {{nowrap|532,500 casualties}} from the Marin Report, for March to June and November to December 1916, for all the Western Front.{{sfn|Jankowski|2014|pp=259–260}}

The {{lang|fr|états numériques des pertes}} give French casualties as {{nowrap|348,000 to 378,000}} and in 1930, Hermann Wendt recorded French Second Army and German 5th Army casualties of {{nowrap|362,000 and 336,831}} respectively from {{nowrap|21 February to 20 December}}, not taking account of the inclusion or exclusion of lightly wounded. In 2006, McRandle and Quirk used the {{lang|de|Sanitätsbericht}} to increase the {{lang|de|Verlustlisten}} by {{circa| 11 per cent}}, which gave {{nowrap|373,882 casualties,}} compared to the French Official History record to 20 December 1916, of {{nowrap|373,231 French}} casualties. The {{lang|de|Sanitätsbericht}}, which explicitly excluded lightly wounded, compared German losses at Verdun in 1916, averaging {{nowrap|37.7 casualties}} per thousand men, with the 9th Army in Poland 1914 which had a casualty average of {{nowrap|48.1 per 1,000,}} the 11th Army in Galicia 1915 averaging {{nowrap|52.4 per 1,000 men}}, the 1st Army on the Somme 1916 average of {{nowrap|54.7 per 1,000}} and the 2nd Army average for the Somme 1916 of {{nowrap|39.1 per 1,000 men.}} Jankowski estimated an equivalent figure for the French Second Army of {{nowrap|40.9 men per 1,000}} ''including'' lightly wounded. With a {{circa| 11 per cent}} adjustment to the German figure of {{nowrap|37.7 per 1,000}} to include lightly wounded, following the views of McRandle and Quirk; the loss rate is similar to the estimate for French casualties.{{sfn|Jankowski|2014|p=261}}

[[File:18-9-27, inauguration de l'ossuaire de Douaumont, cercueils des soldats anonymes.jpg|thumb|{{center|[[Douaumont ossuary]] in 1927}}]]
In the second edition of ''The World Crisis'' (1938), Churchill wrote that the figure of {{nowrap|442,000 was}} for other ranks and the figure of "probably" {{nowrap|460,000 casualties}} included officers. Churchill gave a figure of {{nowrap|278,000 German casualties,}} {{nowrap|72,000 fatal}} and expressed dismay that French casualties had exceeded German by {{nowrap|about 3:2.}} Churchill wrote that an eighth needed to be deducted from his figures to account for casualties on other sectors, giving {{nowrap|403,000 French}} and {{nowrap|244,000 German}} casualties.{{sfn|Churchill|1938|pp=1003–1004}} In 1980, John Terraine calculated {{circa| 750,000 French and German}} casualties in {{nowrap|299 days;}} Dupuy and Dupuy (1993) {{nowrap|542,000 French}} casualties.{{sfnm|1a1=Terraine|1y=1992|1p=59|2a1=Dupuy|2a2=Dupuy|2y=1993|2p=1052}} In 2000, Hannes Heer and Klaus Naumann calculated {{nowrap|377,231 French}} and {{nowrap|337,000 German}} casualties, a monthly average of {{nowrap|70,000.}}{{sfn|Heer|Naumann|2000|p=26}} In 2000, Holger Afflerbach used calculations made by Hermann Wendt in 1931 to give German casualties at Verdun from 21 February to 31 August 1916 as {{nowrap|336,000 and}} French as {{nowrap|365,000 at}} Verdun from February to December 1916.{{sfn|Chickering|Förster|2006|p=114}} David Mason wrote in 2000 that there had been {{nowrap|378,000 French}} and {{nowrap|337,000 German}} casualties.{{sfn|Mason|2000|p=185}} In 2003, Anthony Clayton quoted {{nowrap|330,000 German}} casualties, of whom {{nowrap|143,000 were}} killed or missing; the French suffered {{nowrap|351,000 casualties,}} {{nowrap|56,000 killed,}} {{nowrap|100,000 missing}} or prisoners and {{nowrap|195,000 wounded.}}{{sfn|Clayton|2003|p=110}}

Writing in 2005, [[Robert A. Doughty]] gave French casualties (21 February to 20 December 1916) as {{nowrap|377,231}} and casualties of {{nowrap|579,798 at}} Verdun and the Somme; {{nowrap|16 per cent}} of the casualties at Verdun were fatal, {{nowrap|56 per cent}} were wounded and {{nowrap|28 per cent}} missing, many of whom were eventually presumed dead. Doughty wrote that other historians had followed Winston Churchill (1927) who gave a figure of {{nowrap|442,000 casualties}} by mistakenly including all French losses on the Western Front.{{sfn|Doughty|2005|p=309}} R. G. Grant gave a figure of {{nowrap|355,000 German}} and {{nowrap|400,000 French}} casualties in 2005.{{sfn|Grant|2005|p=276}} In 2005, Robert Foley used the Wendt calculations of 1931 to give German casualties at Verdun from 21 February to 31 August 1916 of {{nowrap|281,000, against}} {{nowrap|315,000 French.}}{{sfn|Foley|2007|p=259}} (In 2014, William Philpott recorded {{nowrap|377,000 French}} casualties, of whom {{nowrap|162,000}} had been killed; German casualties were {{nowrap|337,000}} and noted a recent estimate of casualties at Verdun from 1914 to 1918 of {{nowrap|1,250,000}}).{{sfn|Philpott|2014|p=226}}

===Morale===
Fighting in such a small area devastated the land, resulting in miserable conditions for troops on both sides. Rain and the constant artillery bombardments turned the clayey soil into a wasteland of mud full of debris and human remains; shell craters filled with water and soldiers risked drowning in them. Forests were reduced to tangled piles of wood by artillery fire and eventually obliterated.{{sfn|Clayton|2003|pp=120–121}} The effect of the battle on many soldiers was profound and accounts of men breaking down with insanity and [[shell shock]] were common.<!--not unique to Verdun or the Great War--> Some French soldiers tried to desert to Spain and faced [[court-martial]] and execution if captured; on 20 March, French deserters disclosed details of French defences to the Germans, who were able to surround {{nowrap|2,000 men}} and force them to surrender.{{sfn|Clayton|2003|pp=120–121}}

A French lieutenant wrote,
{{quote|Humanity is mad. It must be mad to do what it is doing. What a massacre! What scenes of horror and carnage! I cannot find words to translate my impressions. Hell cannot be so terrible. Men are mad!| (Diary 23 May 1916){{sfn|Horne|2007|p=236}}}}

Discontent began to spread among French troops at Verdun; after the promotion of Pétain from the Second Army on 1 June and his replacement by Nivelle, five infantry regiments were affected by episodes of "collective indiscipline"; Lieutenants Henri Herduin and Pierre Millant were [[summary execution|summarily shot]] on 11 June and Nivelle published an Order of the Day forbidding surrender.{{sfn|Mason|2000|p=160}} In 1926, after an inquiry into the [[cause célèbre]], Herduin and Millant were exonerated and their military records expunged.{{sfn|Clayton|2003|p=122}}

===Subsequent operations===

====20–26 August 1917====

[[File:French attack at Verdun, August 1917.png|thumb|{{centre|French attack, August 1917}}]]
The French planned an attack on a {{cvt|9|km|order=flip}} front on both sides of the Meuse; XIII Corps and XVI Corps to attack on the left bank with two divisions each and two in reserve. Côte 304, Mort-Homme and Côte (hill) de l'Oie were to be captured in a {{cvt|3|km|order=flip}} advance. On the right (east) bank, XV Corps and XXXII Corps were to advance a similar distance and take Côte de Talou, hills 344, 326 and the Bois de Caurières. About {{cvt|34|km|order=flip}} of {{cvt|6|m|order=flip}} wide road was rebuilt and paved for the supply of ammunition, along with a branch of the {{cvt|60|cm|order=flip}} light railway. The French artillery prepared the attack with {{nowrap|1,280 field}} guns, {{nowrap|1.520 heavy}} guns and howitzers and {{nowrap|80 super-heavy}} guns and howitzers. The {{lang|fr|Aéronautique Militaire}} crowded 16 {{lang|fr|escadrilles de chasse}} into the area to escort reconnaissance aircraft and protect observation balloons. The 5th Army had spent a year improving their defences at Verdun, including the excavation of tunnels linking Mort-Homme with the rear, to deliver supplies and infantry with impunity. On the right bank, the Germans had developed four defensive positions, the last on the French front line of early 1916.{{sfn|Greenhalgh|2014|p=237}}

Strategic surprise was impossible; the Germans had {{nowrap|380 artillery}} batteries in the area and frequently bombarded French positions with the new [[Sulfur mustard|mustard gas]] and made several spoiling attacks to disrupt French preparations. The French counter-attacked but [[Émile Fayolle|Fayolle]] eventually limited ripostes to important ground only, the rest to be retaken during the main attack. A preliminary bombardment began on 11 August and the destructive bombardment began two days later but poor weather led to the infantry attack being put back to 20 August. The assembly of the 25th, 16th, {{lang|fr|Division Marocaine}} and 31st divisions was obstructed by German gas bombardments but their attack captured all but Hill 304, which fell on 24 August. On the right bank, XV Corps had to cross the {{cvt|3|km|order=flip}}-wide Côte de Talou in the middle of no man's land. The French infantry reached their objectives except for a trench between hills 344, 326 and Samogneux, which was taken on 23 August. XXXII Corps reached its objectives in a costly advance but the troops found themselves too close to German trenches and under observed fire from German guns on high ground between Bezonvaux and Ornes. The French took {{nowrap|11,000 prisoners}} for {{nowrap|14,000 casualties}} of whom {{nowrap|4,470 were}} killed or missing.{{sfn|Greenhalgh|2014|pp=237–238}}

====7–8 September====
Guillaumat was ordered to plan an operation to capture several trenches and a more ambitious offensive on the east bank to take the last ground from which German artillery observers could see Verdun. Pétain questioned Guillaumat and Fayolle, who criticised the selection of objectives on the right bank and argued that the French must go on or go back. The Germans counter-attacked from higher ground several times in September; holding the ground captured in August proved more costly than taking it. Fayolle advocated a limited advance to make German counter-attacks harder, improve conditions in the front line and deceive the Germans about French intentions.{{sfn|Doughty|2005|pp=382–383}}

A XV Corps attack on 7 September failed and on 8 September XXXII Corps gained a costly success. The attack continued and the trenches necessary for a secure defensive position were taken but not the last German observation point. More attacks were met by massed artillery fire and counter-attacks and the French ended the operation.{{sfn|Doughty|2005|pp=382–383}} On 25 November after a five-hour [[hurricane bombardment]], the 128th and 37th divisions, supported by {{nowrap|18-field artillery,}} {{nowrap|24 heavy}} and {{nowrap|9 trench}} artillery groups conducted a raid on a {{cvt|4|km|order=flip}} front in appalling weather. A line of pillboxes were demolished and the infantry returned to their positions.{{sfn|Greenhalgh|2014|pp=238–239}}

====Meuse–Argonne Offensive====
{{main|Meuse-Argonne Offensive}}

[[File:Meuse-Argonne Offensive - Map.jpg|thumb|{{centre|Meuse–Argonne Offensive, 26 September – 11 November 1918}}]]
The French [[Fourth Army (France)|Fourth Army]] and the American [[First United States Army|First Army]] attacked on a front from [[Moronvilliers]] to the Meuse on 26 September 1918 at {{nowrap|5:30 a.m.,}} after a three-hour bombardment. American troops quickly captured Malancourt, Bethincourt and Forges on the left bank of the Meuse and by midday the Americans had reached [[Gercourt-et-Drillancourt|Gercourt]], [[Cuisy, Meuse|Cuisy]], the southern part of [[Montfaucon-d'Argonne|Montfaucon]] and [[Cheppy]]. German troops were able to repulse American attacks on Montfaucon ridge, until it was outflanked to the south and Montfaucon was surrounded. German counter-attacks from {{nowrap|27 to 28 September}} slowed the American advance but Ivoiry and Epinon-Tille were captured, then Montfaucon ridge with {{nowrap|8,000 prisoners}} and {{nowrap|100 guns.}} On the right bank of the Meuse, a combined Franco-American force under American command, took Brabant, Haumont, Bois d'Haumont and Bois des Caures and then crossed the front line of February 1916. By November, {{circa| 20,000 prisoners,}} {{circa| 150 guns,}} {{circa| 1,000 trench mortars}} and several thousand machine-guns had been captured. A German retreat began and continued until the Armistice.{{sfn|Michelin|1919|pp=24–25}}