THE WAR OF THE FACTORIES

At the outbreak of war the infantryman's weapon was a bolt-action magazine rifle, and each of the great armies had its own. The British Short Magazine Lee-Enfield (SMLE) fired a .303-inch cartridge from a detachable ten-round magazine loaded with two five-round chargers, and its smooth bolt made it the fastest of the standard rifles. Regular soldiers were drilled in the "mad minute", at least 15 aimed shots in 60 seconds at a target 300 yards away; the best marksmen managed 20 to 30, and in 1914 a musketry instructor, Sergeant Alfred Snoxall, was credited with 38 hits in a minute. The German Gewehr 98, a Mauser design firing a 7.92-millimeter round from a five-round internal magazine fed by a stripper clip, was robust and accurate. The French Lebel, a veteran of 1886, was the first military rifle to use smokeless powder, but its eight cartridges lay end to end in a tube under the barrel and had to be loaded one at a time; the army quietly adopted three-round Berthier rifles in its place in 1915 and 1916.

The machine gun

The weapon that really defined the war was the machine gun. Hiram Maxim's recoil-operated design of 1884 had been refined by every major army. The German MG 08, the British Vickers (adopted 1912) and the Russian Maxim all worked on the same principle: a belt of 250 cartridges fed through the gun, which fired at about 450 to 600 rounds a minute while a water jacket around the barrel cooled it. The price was weight. A Vickers with tripod, water and ammunition weighed about 40 kilograms, and the German MG 08 on its sled mount more than 60. A crew of four to six was needed to carry and serve it. Its worth lay in what it did once in place: a single gun could sweep an arc of ground with a stream of bullets that was commonly reckoned equal to the fire of dozens of riflemen.

Germany entered the war with far more machine guns than Britain, and ordered them in huge numbers; the British, who had issued two Vickers guns to each battalion, formed a dedicated Machine Gun Corps in October 1915. Of the roughly 170,000 men who served in it, about 12,500 died, which earned it the grim nickname of the "Suicide Club". The Vickers was also used as artillery. On 24 August 1916, to support an attack on High Wood on the Somme, ten Vickers guns of the 100th Company fired continuously for about 12 hours, laying a barrage of indirect fire over the German positions and expending close to a million rounds; the barrels were replaced repeatedly and the crew fetched water, even urine, to keep the guns cool.

Light and portable

An attacking infantryman needed a lighter weapon. The answer was the Lewis gun, designed by Colonel Isaac Newton Lewis of the US Army and manufactured in Britain, an air-cooled gun of about 28 pounds with a distinctive aluminum cooling shroud and a 47-round pan magazine, adopted by Britain in 1915. By 1918 a British battalion might have 30 or more. The French Chauchat (the CSRG), a light automatic rifle made cheaply in the hundreds of thousands, was a noted failure: its half-moon magazine was open at the sides, jammed in mud, and the American version made for the AEF was notoriously unreliable. The Germans answered with the MG 08/15, a water-cooled gun fitted with a bipod and a pistol grip, and in 1918 with the Bergmann MP 18, the world's first practical submachine gun.

The British Army was slow to appreciate the gun's effects: it began the war with only two Vickers per battalion, and a line often pinned on Douglas Haig, that the machine gun was "a much over-rated weapon", is probably a myth: it stems from the 1930 memoir of the machine-gun officer C. D. Baker-Carr, who named no one, and historians such as Gary Sheffield reject it. Events overtook the army's caution in any case. By 1916 both sides organized machine guns to fire not at visible targets but into fixed zones, sweeping gaps in the wire and the lanes between trenches in the dark. The machine gun, in the defender's hands, made it nearly impossible to advance in the open; in the attacker's hands, assisted by the Lewis gun, it became a means of keeping the defenders' heads down while the infantry rushed forward. Tactics of 1917 and 1918 depended on the second use.

If one weapon defined the war, it was the gun. Artillery fired the shells that did most of the killing, and produced the fragments that were responsible for most of the wounds. Medical records suggest that around 60 percent of battle deaths were due to artillery, and in some armies' statistics the share was higher. The guns ranged from the French 75 mm field gun and the British 18-pounder, which fired a shell some 4 to 5 miles to support infantry, through howitzers, which lobbed shells high into trenches, to railway-mounted and siege pieces that hurled shells weighing as much as a car.

The French 75, adopted in 1897, was the most advanced field gun of its day. A hydro-pneumatic recoil mechanism held the barrel and carriage steady, so the crew could fire without re-aiming, at up to 15 rounds a minute; its shrapnel shell carried nearly 300 balls. French doctrine had come to depend on it, and the army entered the war short of heavy guns. The Germans, who had studied the fortress war of 1904 to 1905 in Manchuria, had far more heavy howitzers, and they used them at once. In August 1914 the Krupp 42-centimeter "Dicke Bertha" and the Austrian Škoda 30.5-centimeter siege howitzers pounded the forts of Liège and Namur into rubble; the magazine of Fort Loncin, hit on 15 August, blew up and killed most of its garrison.

The shell crisis

Every army ran out of shells. The British had planned for a small expeditionary force and were firing shells faster than their factories could make them by the autumn of 1914. In May 1915, after the failure of the attack at Aubers Ridge on 9 May, Field Marshal Sir John French blamed a lack of high explosive and briefed the military correspondent of The Times, Charles à Court Repington, whose report appeared on 14 May. Lord Northcliffe's Daily Mail attacked Kitchener a week later. The scandal, combined with a quarrel over the Dardanelles, brought down the Liberal government; the new coalition created a Ministry of Munitions under David Lloyd George in June 1915. The output of guns and shells rose dramatically, drawing in hundreds of thousands of workers, many of them women. But the workmanship was often poor, and a notable proportion of the shells fired on the Somme in 1916 failed to explode; farmers in Flanders and Picardy still plow up tons of unexploded shells every year.

Learning to shoot

The great artillery lesson of the war was that more shells were not enough: they had to land in the right place at the right time. The first problem was location. Gun positions were found by flash spotting at night, by observers with theodolites who triangulated muzzle flashes, and by sound ranging, in which a line of microphones recorded the arrival times of the report. The British system was developed by the young physicist William Lawrence Bragg, who had just shared the Nobel Prize, with William Sansome Tucker's low-frequency microphone, and by 1918 could place a hostile battery within a few dozen yards. Air observation, wireless and photography helped locate targets and spot the fall of shot.

The second problem was accuracy. A gun had traditionally to fire "ranging" shots at a target to find the distance, and this warned the enemy. Artillerymen learned instead to calculate ("predicted fire") using the exact survey position of each gun, the wear of its barrel, the temperature of the propellant, and the wind and air density. The British first used it on a large scale at Cambrai on 20 November 1917, when about 1,000 guns opened fire without warning. The third problem was timing: the creeping barrage, a curtain of shells that advanced just ahead of the infantry at about 100 yards every three or four minutes, had been tried in various forms from 1915; the French used it to great effect at Verdun in October 1916, and it became standard in the British Army in 1917.

Bruchmüller's hurricane

The most spectacular artillery planner was a retired German artillery officer, Georg Bruchmüller, nicknamed "Breakthrough Müller" (Durchbruchmüller). He developed a method of short, intense, scientifically timed bombardments that opened without ranging. A typical plan began with gas shells on the enemy artillery, then high explosive on command posts and communications, then a rolling barrage for the infantry. He refined it at Riga (September 1917) and Caporetto (October 1917). On 21 March 1918, against the British Fifth Army, his bombardment from about 6,500 guns and 3,500 mortars fell on a front of some 40 miles for five hours from 4:40 a.m., mixing high explosive with "Blue Cross" sneezing gas, which forced men to take off their masks, and "Green Cross" phosgene shells (Buntschiessen, "colored shooting").

No weapon of the war aroused such horror. On the evening of 22 April 1915, near Langemarck north of Ypres, a German unit opened the valves of 5,730 cylinders and released about 168 tons of chlorine gas. A yellow-green cloud rolled across no man's land on a light breeze and into the trenches of two French divisions, one Territorial and one Algerian. Men choked, fell, fled; a gap about four miles wide opened in the Allied line. Estimates of the dead range widely, from roughly 800 to 1,400 in recent studies to 5,000 or more in older accounts. The Germans, who had no large reserves ready to exploit the breach, failed to take Ypres, and Canadian and British troops plugged the gap in bitter fighting over the following days (see chapter IV).

The attack was organized by the chemist Fritz Haber. It was not the first use of chemical agents: the French had fired tear-gas grenades in August 1914, and the Germans had tried shells filled with xylyl bromide at Bolimów on the Eastern Front on 31 January 1915, though the cold prevented the liquid from vaporizing. Nor was it legal, quite: the Hague Declaration of 1899 had forbidden "projectiles" whose sole object was to spread asphyxiating gases, and the Germans argued that gas released from cylinders was not a projectile. Britain responded by retaliating at Loos on 25 September 1915; in places the wind blew the gas back onto its own trenches.

The arms race

Both sides at once developed defenses and better poisons. The first respirators were improvised pads of cotton waste soaked in urine or bicarbonate. The British "hypo helmet", a flannel bag soaked in sodium thiosulfate, was followed by the "PH helmet", and in 1916 by the Small Box Respirator, with a mouthpiece connected by a hose to a canister of chemicals, which was the best mask of the war. Germany and France developed masks of their own, the French M2 among them. Gongs, rattles, klaxons and empty shell cases warned of attacks; units drilled to put on the mask within seconds. Phosgene, deadlier than chlorine and harder to detect, was used by the Germans, mixed with chlorine, against the British at Wieltje on 19 December 1915. Diphosgene ("Green Cross") shells deluged the French guns at Verdun in June 1916. Mustard gas ("Yellow Cross", the French "Yperite"), a persistent liquid that blistered the skin and eyes and lingered in the ground for weeks, appeared at Ypres on the night of 12 July 1917. It seldom killed, but it caused huge numbers of casualties, and a British soldier who survived could spend weeks in hospital.

By 1917 most gas was delivered in shells, and in the British case by the Livens projector, a simple mortar-like tube dug into the ground and fired electrically in batches, and by the Stokes mortar. The prevailing winds on the Western Front blew from west to east, which favored the Allies when they used cylinders. Gas was an area weapon that never made a decisive breakthrough: once armies had good masks, it became a way to harass, to deny ground, and to force the enemy to fight encumbered. A defender in a mask could not see, shout or breathe easily, and gas often added to the misery rather than to the deaths.

By the end of 1914 the problem of the trenches was obvious to everyone who looked at it: a soldier on foot could not cross barbed wire under machine-gun fire. The answer, an armored vehicle on caterpillar tracks, was suggested independently by many. Lieutenant Colonel Ernest Swinton, a Royal Engineers officer serving as official war correspondent, proposed it in October 1914 after hearing of the American Holt tractors, and Maurice Hankey, the secretary of the Committee of Imperial Defence, took up the idea. Winston Churchill, as First Lord of the Admiralty, set up a Landships Committee in February 1915 under Eustace Tennyson d'Eyncourt, the Director of Naval Construction. Its chief engineers were William Tritton of the Lincoln firm of William Foster and Company and Lieutenant Walter Wilson.

The first prototype, "Little Willie", ran in September 1915. Its successor, built in a form with tracks running around the whole hull, was nicknamed "Mother" or "Big Willie" and was demonstrated to the Army in a trial at Hatfield Park on 2 February 1916, crossing a trench and a parapet to the delight of the officials who watched. The vehicles were disguised in shipment as water carriers, "tanks", and the name stuck.

Mark I and the Somme

The Mark I was a rhomboidal steel box weighing 28 tons in the "male" version (armed with two 6-pounder guns in side sponsons and machine guns) and a little less in the "female" (with machine guns only). A crew of eight, four to drive and four to fire, was crammed inside with a 105 horsepower engine, in the same compartment and without an effective silencer. Temperatures could climb above 120 degrees Fahrenheit; the fumes of carbon monoxide made men sick or faint; speed was about 3 miles an hour on a good surface. It was bulletproof but vulnerable to a shell or an armor-piercing bullet. It could cross a trench up to 10 feet wide.

Haig, impatient to use the new weapon on the Somme, sent 49 tanks forward on 15 September 1916 in the battle of Flers-Courcelette. Many broke down, bellied in shell craters or got lost, but a handful reached the German lines and helped to capture Flers, and the shock on the Germans was real. Critics, including Swinton, argued that sending so few before the machines were reliable squandered the surprise. It was an early example of an argument that runs through the history of every new weapon: use it as soon as it can help, or wait until it can decide the battle.

Cambrai and after

The tank's real debut came at Cambrai on 20 November 1917, when 476 tanks, 378 of them fighting vehicles, advanced without a preliminary bombardment, preceded by the predicted-fire barrage described above. Each carried a fascine, a huge bundle of brushwood, to be dropped into the wide German trenches of the Hindenburg Line. The Third Army broke through to a depth of about five miles in a day, with far fewer casualties than the months of fighting at Passchendaele had cost; church bells were rung in Britain. Ten days later, the Germans counterattacked with infiltration tactics and won back much of the ground. The lesson was that tanks could open a door, but without reserves, infantry and artillery, they could not hold it open.

The French developed their own heavy tanks, the Schneider and the Saint-Chamond, which proved disappointing, and then in 1917 the Renault FT, a small two-man machine designed by Louis Renault under the sponsorship of Colonel Jean-Baptiste Estienne. It was light, cheap, and had a fully rotating turret, a layout copied by almost every tank since. About 2,700 were delivered before the Armistice, and it first fought east of the Forest of Retz on 31 May 1918. The Germans, short of steel and unconvinced, built only about 20 of the cumbersome A7V; the first tank-against-tank action took place at Villers-Bretonneux on 24 April 1918, when a British Mark IV knocked out an A7V. Germany relied instead on antitank rifles, special armor-piercing bullets, and a field gun used as a direct-fire weapon.

On 8 August 1918 at Amiens, more than 500 British tanks, including Whippets, took part in the opening of the Hundred Days, though within a few days most had broken down or been knocked out. Historians now regard the tank as one useful tool among several. It was not the war-winning miracle that its advocates, such as J. F. C. Fuller, once claimed, but without it the infantry of 1918 would have taken heavier losses.

If a trench could not be taken from above, it could be blown up from below. Mining was an ancient art of siege warfare, and on the Western Front it was revived on a scale no one had foreseen. Germans began digging under French and British trenches in 1914 and made some small successes at Festubert in December. The British, short of expertise, turned to the mining engineer and Member of Parliament John Norton-Griffiths, nicknamed "Empire Jack", who in February 1915 recruited coal miners and the "clay-kickers" who dug London's sewers, a technique in which a man lay back on a wooden support and cut the clay with a spade held in his feet, almost silently. They were formed into Tunnelling Companies of the Royal Engineers. Eventually there were about 30, including Australian, Canadian and New Zealand companies, and tens of thousands of men served in them.

Underground war

The tunnelling war was a duel of ears. Miners worked in pitch darkness at depths from 20 feet to over 100, listening for enemy diggers with geophones, extraordinarily sensitive listening devices. A rival gallery might be detected by the sound of a pick; the answer was a "camouflet", a small charge that collapsed the enemy gallery without breaking the surface, or a bigger charge to destroy the enemy's chamber. Sometimes the two sides broke into each other's galleries and fought underground with knives, picks and revolvers. Men worked in foul air and the constant risk of being buried. The first major British mines were exploded on 17 April 1915 at Hill 60, a spoil heap from a railway cutting that dominated the Ypres salient.

The Somme mines

On 1 July 1916 the British fired a series of mines along the Somme front. The largest, at Lochnagar near La Boisselle, held 60,000 pounds of ammonal in two chambers and was fired at 7:28 a.m., two minutes before zero hour. It threw earth and chalk some 4,000 feet in the air and left a crater roughly 100 meters across and 20 meters deep, which survives today. At Hawthorn Ridge, near Beaumont-Hamel, a mine of about 40,000 pounds was fired at 7:20, ten minutes before the assault, and this early warning probably helped the Germans to man their parapet before the British arrived; it was filmed by a cameraman, Geoffrey Malins, and the footage is one of the best known of the war. Elsewhere in France, the French and Germans fought a ferocious mining war for the hill-top village of Vauquois in the Argonne from 1915 to 1918, exploding more than 500 mines which destroyed the village and left a moonscape that survives today.

Messines

The great mining feat was prepared over more than a year. The British Second Army under General Sir Herbert Plumer wanted to take Messines Ridge, which overlooked the southern side of the Ypres salient, before the main offensive of 1917. The Tunnelling Companies, aided by Canadian, Australian and New Zealand miners, drove galleries under the German lines, in some cases 100 feet down in blue clay to avoid the water-bearing sands above, and filled their chambers with nearly a million pounds of ammonal. Preceded by a bombardment of about 3.5 million shells, 19 of the mines exploded at 3:10 a.m. on 7 June 1917. The explosion was reportedly heard in London and was mistaken for an earthquake in Lille; among the biggest mines, the one at Spanbroekmolen held 91,000 pounds of ammonal and left a crater now called the "Pool of Peace". The mines are often said to have killed 10,000 Germans, but the historian Simon Jones traces that figure to a count of the missing and argues that probably no more than 500 died in the blasts themselves; their greater effect was shock and panic. The infantry then walked forward and took the ridge within hours. Plumer's chief of staff, Charles Harington, is said to have told reporters the night before that the army might not make history but would certainly change the geography (the wording varies, and the remark has not been traced to a primary source).

In 1914 every army went to war with a doctrine built on the offensive: the French offensive à outrance, the German drive for envelopment, the British stress on the rifle and the "offensive spirit". The trench war made them all obsolete. From 1915 the armies tried a succession of formulas, each of which promised to break the deadlock, and each of which met the machine gun and the shell. This was the "learning curve", and its pace and reasons are among the best-known arguments in military history.

The old picture, and the new

For decades after the war, the dominant view, expressed in memoirs, in Alan Clark's book The Donkeys (1961) and in the play and film Oh! What a Lovely War, was that the generals had understood nothing, that "lions" had been "led by donkeys". (The famous remark said to have been exchanged by two German generals has never been traced to a German source, and Clark's account of it is generally taken to be invented.) Since the 1980s historians such as Paddy Griffith, Gary Sheffield and Peter Simkins have argued that the British Army went through a steep learning curve between 1915 and 1918 and emerged in the Hundred Days as one of the most effective armies of the war. Others, among them Robin Prior and Trevor Wilson, have been more critical of Haig and Rawlinson, whose planning on the Somme and at Passchendaele they judge costly and often unimaginative. Most modern writers reject the cartoon of stupid generals, but argue about whether the learning was fast or slow, and whether the lessons were learned at the top or on the front line.

From waves to platoons

The infantry's formation changed first. In 1915 and 1916, attacks were launched in lines or "waves" of men at fixed intervals. After heavy losses, tactical manuals shifted. French captain André Laffargue's pamphlet of May 1915 recommended small groups infiltrating gaps. The British issued SS 135 (December 1916, for divisions) and SS 143 (February 1917), "Instructions for the Training of Platoons for Offensive Action", which organized each platoon into four specialist sections: riflemen, bombers, rifle grenadiers and a Lewis gun team, to support each other by fire and movement. By 1917 platoons were attacking in loose, dispersed groups instead of lines.

Bite and hold

Rawlinson's phrase "bite and hold" described a deliberately limited advance of a mile or less, taking ground that the artillery could fully support and then holding it against the counterattack that would follow. Plumer perfected it at Messines and again at the Menin Road, Polygon Wood and Broodseinde in September and October 1917, the successful phase of the Third Battle of Ypres. Critics, including Haig, found it too modest; supporters argued that the Germans' defense in depth could be beaten only by repeated careful blows.

Combined arms

The synthesis emerged in 1918. At Hamel on 4 July 1918, the Australian Corps commander General John Monash used 60 tanks, aircraft that dropped ammunition by parachute, massed artillery and several brigades of infantry in a plan that had been rehearsed in detail; the objectives were taken in 93 minutes, almost exactly as scheduled. At Amiens on 8 August 1918 the same ideas were applied on a larger scale, with the Canadian and Australian Corps in the lead. Over 2,000 guns, more than 500 tanks and several hundred aircraft supported the attack. The Germans lost some 30,000 men on the first day, about half of them prisoners. Erich Ludendorff called it "the black day of the German Army". It did not end the war in a day, but it was the beginning of the end, and it showed what an army that had mastered every tool of the industrial war could do.

An army is a machine for moving information and supplies, and in 1914 neither could travel faster than a man or a horse, except where there was a railway or a telegraph. For most of the war that limitation shaped everything. A commander planning an attack could set a timetable, but once the infantry vanished into smoke and dust, he learned what had happened only by messengers, and he learned it hours late.

Getting the message through

The simplest solution was the runner, usually a young man chosen for speed, who ran the gauntlet of shellfire carrying a note. The runner was often killed. Telephones were a great advance: in the weeks before an attack, signal units laid cables from headquarters to the front, and in the later years of the war they were buried six feet deep to survive shelling. The enemy could still tap lines, especially by picking up the earth currents (the German "Moritz" stations), and the British developed the Fullerphone, a secure instrument that used direct current and could not be overheard. But a bombardment cut wires quickly, and once the attack began the telephone usually fell silent.

The alternatives were crude. Pigeons were widely used; the British Army alone employed about 100,000, which carried messages from the line, from tanks, from ships and even from aircraft, and a high proportion of them got through. The American bird Cher Ami flew a message from the "Lost Battalion" in the Argonne in October 1918, though wounded. Dogs, signal lamps, flags, rockets of different colors and the "SOS" flare to call a barrage were all used. Wireless was bulky and unreliable on the ground but already valuable in the air: aircraft sent Morse messages to guns to correct their fire from the autumn of 1914, and by 1917 portable sets reached the forward battalions. The result was that the commander in the rear and the soldier at the front lived in different worlds of time. Many historians cite this "command and control gap" to explain why plans were so rigid and why the barrage timetable became sacred: it was the only way to coordinate an attack in the absence of real-time communication.

Feeding the guns

Supply was the other half of the equation. A British division in the line consumed scores of tons of food, forage, ammunition, wire, timber, water and mail every day, and far more in a battle. The system began at the Channel ports, where ships landed goods that were taken by standard-gauge railway to railheads some 10 or 15 miles behind the front, and onward by horse-drawn wagon, motor lorry, and then, closer to the line, by narrow-gauge "light railway" and finally by men carrying loads on their backs. The French Decauville 60-centimeter light railway was copied by the British and Germans on a large scale; by 1918, hundreds of miles of such track crossed the old battlefields, usually under shell fire. Roads of planks or rubble were laid across the mud.

The Battle of Verdun demonstrated the importance of logistics. The railways serving the fortress had been cut or were of low capacity, and the main supply route became a country road from Bar-le-Duc, which the writer Maurice Barrès named the Voie Sacrée, the "Sacred Way". At its peak, a truck was said to pass every 14 seconds; the road carried tens of thousands of men and thousands of tons of materiel a week, and organization (including an army of road menders who kept the surface repaired) kept it open. German failure to cut it was one reason Verdun held.

The industrial war produced casualties on a scale that no previous medical service had faced, and doctors improvised their way to techniques that reshaped surgery. The wounds were new: shell fragments and bullets drove soil, cloth and manure deep into tissue, and the infections that followed, gas gangrene and tetanus above all, killed men who might have survived a clean wound. The evacuation chain is described in chapter IX; here we look at what the surgeons did.

Triage and debridement

The first principle was triage, from the French trier, "to sort". At a casualty clearing station, an experienced medical officer inspected each arrival and assigned him to one of three groups: those who needed immediate surgery, those who could wait, and those who were beyond help and should be made comfortable. The second principle, adopted by most surgeons by 1916, was debridement: cutting out all dead and contaminated tissue and leaving the wound open for several days before closing it, which sharply reduced gas gangrene. Antitetanic serum, injected into every wounded man from late 1914, drove tetanus almost to extinction. The French surgeon Alexis Carrel and the British-born chemist Henry Dakin devised a further refinement, the Carrel-Dakin method, in which a dilute hypochlorite solution was dripped through rubber tubes into the wound every two hours; Dakin's solution, and the cruder British variant Eusol, were standard through the war.

The Thomas splint and the broken leg

A bullet or fragment through the thigh bone was among the most deadly wounds: the fracture was often compound, the bone fragments tore the large vessels, and the patient was jolted for miles in a wagon. In 1916 the orthopedic surgeon Robert Jones, drawing on the splint designed by his uncle Hugh Owen Thomas, saw to it that the Thomas splint, a ring and rods that held the limb in traction, was fitted to every such casualty at the front. Mortality for compound femur fractures, which had been around 80 percent early in the war, fell to well under 20 percent by 1917 and 1918 (one often-cited figure for 1918 is 8 percent).

Blood

Before 1914, a transfusion meant joining the arteries of donor and patient. The Belgian Albert Hustin used sodium citrate to stop blood clotting in 1914, a technique that Richard Lewisohn in New York perfected in 1915. In 1917 the American Captain Oswald Hope Robertson set up what is usually considered the first blood bank, collecting type O blood in citrate-glucose and storing it on ice at casualty clearing stations, where it saved wounded men in shock, in the battle of Cambrai in November 1917 and after. Blood groups had been discovered by Karl Landsteiner in 1901.

Faces, limbs and minds

The shell and the machine gun maimed faces and limbs on an unprecedented scale. The New Zealand-born surgeon Harold Gillies led the British effort to rebuild faces, first at Aldershot and, from 1917, at the new Queen's Hospital in Sidcup, Kent, where about 5,000 men were treated. Gillies refined the tubed pedicle flap, in which a strip of skin was raised, tubed and "walked" from one part of the body to another; the same technique was developed independently by the Russian Vladimir Filatov. The artist Henry Tonks, a surgeon turned Slade professor, drew pastel portraits of the patients before and after, and sculptors such as Francis Derwent Wood and the American Anna Coleman Ladd made painted metal masks for men whose faces could not be restored. Roughly 41,000 British servicemen lost limbs; prosthetic centers such as Queen Mary's Hospital at Roehampton fitted artificial legs and arms. In Germany, Ferdinand Sauerbruch designed an arm that could be controlled by the muscles of the stump. For the treatment of the mind, see chapter IX.

When the trenches closed to a few yards, an infantryman needed weapons that a rifle could not match: something to throw, something to burn, something to hit with. The war brought a startling revival of medieval weapons alongside the most advanced industrial technology.

Grenades

In 1914 nobody had the right hand grenade. British soldiers made jam-tin bombs, a tin of explosive and nails with a fuse lit by a cigarette; the Germans had only a modest stock of old-fashioned fortress grenades. The British adopted the Mills bomb in May 1915, the work of the Birmingham engineer William Mills: a serrated cast-iron body filled with explosive, with a spring-loaded lever held by a pin, and a time fuse of a few seconds. Tens of millions were made. The German army used the Stielhandgranate, a stick grenade with a wooden handle that gave leverage for throwing it farther and which the British called the "potato masher". Rifle grenades sent a bomb 200 yards. Grenades became the infantryman's main weapon in a trench fight, and each platoon had its "bombers" as specialists. The grenade also showed the dangers of the war: a man who dropped one in a narrow trench had no escape.

Trench mortars

A mortar lobs a shell on a very high arc into a trench, which makes it ideal for a front line only a few hundred yards from the enemy. The Germans, who had learned from the siege of Port Arthur, entered the war with several types of Minenwerfer, from the light 7.58 cm to the "heavy" 25 cm; their shells, sailing slowly through the air, were nicknamed "Minnies" by the British. The Allies had to improvise. The French used the crapouillot (the "little toad", a nickname for its squat shape), and the British, in the autumn of 1915, got the Stokes mortar, designed by the engineer Wilfred Stokes. It was a simple steel tube on a base plate: the bomb was dropped in, hit a striker pin at the bottom and flew out. It could fire many rounds a minute, was light enough to be carried by three men and was cheap. The Stokes design, improved by the French Brandt company, is the ancestor of almost every modern infantry mortar. The heavy British mortars, with their "toffee apples" (the 2-inch) and "flying pigs" (the 9.45-inch), became part of the soldiers' slang.

Flamethrowers

The German Flammenwerfer was first used against the French at Malancourt, near Verdun, on 26 February 1915, and against the British at Hooge in the Ypres salient on 30 July 1915. The portable version was a steel cylinder of oil and compressed gas carried on the back, with a hose and a nozzle, that projected a jet of burning liquid for 20 to 30 yards for less than a minute. The bigger version needed two men. The first British experience of "liquid fire" at Hooge, where a reserve battalion was driven out of its trench, was a shock, and the weapon seemed to symbolize the new frightfulness. Its crews were small units of pioneers, heavily targeted by every rifleman who saw them, who were reportedly not often taken alive. The Allies made their own flamethrowers, but it was the Germans who used them most, mainly in storm detachments to clear dugouts.

The return of the club

In a trench bay a few yards across, a rifle was often too long to swing and a bayonet too slow. Soldiers on both sides improvised knobkerries, clubs, spiked maces, bludgeons made from tent pegs, and sharpened shovels. Factory-made versions were soon available, and the trench knife became an official weapon of the American and French armies. The bayonet itself, despite the army's reverence for it, caused fewer than 1 percent of wounds. In its place came a machinery of close killing as brutal as any in history.

This is the general lesson of the chapter. The First World War did not replace old forms of violence with new ones. It brought the old, the new and the improvised into one space, and the soldier in the trench might be hit by a 15-inch shell, a machine-gun bullet, a mustard-gas droplet, a thrown bomb and a sharpened spade, all in the course of a single night.