Somme: The Horrors They Don’t Tell You

The British Army suffered 57,470 casualties on the first day of July 1916, a single day of slaughter that remains the worst in its entire history, but the Battle of the Somme did not end there. It raged for four and a half more months, producing horrors that have been systematically minimized or forgotten. This is the full, unvarnished account of what happened when a generation of volunteers walked into machine guns and artillery, and what the world chose not to remember.

To understand the Somme, one must go back to the autumn of 1914. The German advance through Belgium and France had been stopped at the Marne in September. Both sides then attempted to outflank each other in a series of moves called the Race to the Sea.

Neither side won. The result was a continuous belt of trenches running 400 miles from the North Sea to the Swiss border, with no flank left to turn. From that point, any attack had to go straight through the enemy’s line head-on, turning the entire war on the Western Front into a single question: how do you walk into machine guns and artillery without your army being destroyed?

The army Britain sent to France in August 1914 was the British Expeditionary Force, six infantry divisions and five cavalry brigades of long-service professionals. They were the finest trained part of the British army, capable of putting 15 aimed shots a minute into a target 300 yards away. They were fed straight into the heaviest fighting of the year.

By the end of 1914, total British losses stood at roughly 95,000 men. The professional army of a world power was down to remnants. What Britain had left was nowhere near enough for the war it was now in.

The war secretary, Lord Kitchener, insisted from the start that the war would run for years and that Britain needed an army on the scale of the French and German ones. But Britain had no conscription. The only way to get men was to ask for them.

Kitchener put out his first appeal for 100,000 volunteers on August 11, 1914, and a second for another 100,000 on August 28. The response was overwhelming. Around half a million men came in, and by the end of October 1914, the total stood at almost 900,000.

About 40% of the volunteers were turned away on medical grounds, a reflection of a century of industrial poverty written into the bodies of the men who came to defend it.

The thing that made this recruiting boom work was a promise: if you enlisted together with your neighbors, relatives, or workmates, you would all serve together in the same unit. This is how the Pals battalions were born. Towns turned the whole thing into a competition over who could raise a battalion faster.

But there was a flaw built into this idea from day one. When a battalion of roughly a thousand men all comes from one place, that battalion can hold most of the fit young men that town has. If that one unit ends up in the wrong spot at the wrong time, the losses do not spread across the country.

The whole blow lands on a single town, arriving all at once in one round of telegrams.

What these volunteers thought they were signing up for had very little to do with what was actually happening in France. The common belief was that the whole thing would be over by Christmas. The real danger, as far as they could see, was missing it.

Meanwhile, the army they were joining was watching its professionals get destroyed at rates of tens of thousands of men a month. Among those volunteers were boys in numbers that are quite hard to take in. Over the course of the war, an estimated 250,000 recruits under the legal age got themselves into the British Army.

The rules said 18 to enlist and 19 to serve overseas, but back then, plenty of people had no birth certificate. Recruiting officers did not demand one, and the recruiting sergeants were paid for every man they signed.

The extreme case was a kid named Sydney Lewis, who managed to enlist in August 1915 at the age of 12. At 13, he was fighting on the Somme. Then there was Horus Isles, a 14-year-old who was handed a white feather in the street as a mark of cowardice.

He went and joined up. He was killed on the Somme at 16. Girls were sent out onto the streets to hand white feathers to any man who looked like he should be in the army.

The feathers landed on men too young, men too old, men already rejected as unfit, and even on soldiers home on leave from the actual fighting in France. The pressure did the job almost as well as conscription.

The decision that sent these men to the Somme was made in December 1915 at a conference at French headquarters in Chantilly. The Allied commanders agreed on the only strategic idea they had left: France, Britain, Russia, and Italy would all attack at roughly the same time in 1916, so the Germans and Austrians could not shuttle their reserves from one front to another. For the Western Front, the French commander-in-chief Joseph Joffre picked the river Somme simply because that is where the British and French armies happened to stand side by side.

That was the entire strategic reasoning. It also helped that it was a quiet sector where nothing big had happened for months, but that sector was quiet for a very good reason.

The Germans had held that ground since the line froze in 1914. For nearly two years, the only thing they had been doing there was digging. The country along the Somme is rolling chalk downland, and the German strong points sat on the high ground, looking straight down into the British lines.

Into that chalk, the Germans had cut shelters going down 20 and 30 feet, some fitted with electric light, proper beds, and stoves. Some of the dugout stairways even had rails running down them. The moment a bombardment lifted, the crews could haul their machine guns straight up to the surface.

Shells getting through to men that deep was practically impossible. The only real way to deal with the German infantry was to keep them sealed down there until your own men were standing right on top of the exits.

Then there was the wire in front of the German trenches, which had grown into two belts, each about 30 yards wide. The first position was three trench lines deep with villages turned into fortresses sighted so that each one covered the others. Behind that ran a line of strong points, and behind that sat an entire second position placed so far back that the British field guns could not touch it without packing up and moving forward first.

A third position was nearly finished by the time the battle opened. Even an attacker who did absolutely everything right and fought his way clean through the first position would arrive on the far side exhausted, beyond the reach of his own guns, staring at a second position just as strong as the one he just paid so dearly to take.

In February 1916, the Germans attacked the French fortress city of Verdun, and everything changed. Verdun swallowed French divisions month after month. The French share of the coming offensive kept shrinking until the Somme, planned as a mainly French battle, turned into a mainly British one with French support.

It also picked up a second purpose: to pull German strength away from Verdun and give the French some room to breathe. The main weight now sat on the British Fourth Army under General Henry Rawlinson. His army was made largely of those volunteers of 1914.

Britain was about to fight a battle on a scale its army had never attempted.

Through the spring of 1916, Rawlinson and the British commander-in-chief, Sir Douglas Haig, argued about what this attack was even supposed to be. Rawlinson had watched the failures of 1915 up close and wanted a limited, methodical battle: bite off about 2,000 yards of German line, stop while still under the protection of your own guns, destroy the German counterattacks as they come in, move the artillery forward, and then do the whole thing again. Haig called that too cautious.

He wanted the infantry through the entire German first position in one single bound on a front of some 25,000 yards, with the cavalry waiting behind to pour through the gap. What came out of that argument was a compromise that managed to collect the risks of both ideas and the strengths of neither.

The objectives were set deep, which only works if you have surprise and speed, but they would be prepared by a week-long bombardment, which threw away every bit of surprise. All of it was spread across defenses too deep for the guns they actually had. Then, because the commanders did not trust the training of their volunteer army, the infantry were ordered to advance in rigid waves on fixed timetables, with no man allowed to use his own judgment about anything.

That was the plan. But before the first day itself, the first horrors of the Somme did not happen in the trenches at all. They happened deep underground.

While the generals were arguing about objectives and timetables, another war had been running underneath the battlefield since the winter of 1914. In December of that year, the Germans exploded mines under British trenches. The British answer was to create the tunneling companies of the Royal Engineers.

They went after the men who actually knew how to work underground: the clay kickers, the specialists who dug London sewers, along with coal miners from all over Britain. What developed down there was one of the most nightmarish jobs of the entire war. Each side listened for the other with men pressing stethoscopes against the tunnel walls, trying to catch the sound of enemy picks somewhere in the chalk.

When you heard digging, you drove a counter tunnel toward the sound and fired an underground charge called a camouflet, built to collapse the enemy’s gallery with the diggers still inside it. The men the blast did not 𝓀𝒾𝓁𝓁 outright were left entombed down there. Even the blast was not the end of it, because these underground explosions left carbon monoxide hanging in the tunnels.

That gas does strange things to a man before it finishes him. Rescue parties kept finding that the poisoned men turned violent and hysterical, actually fighting the very men who had crawled in to drag them out. The tunnelers carried mice, canaries, and sometimes rabbits down into the workings because the small animals collapsed from carbon monoxide at concentrations a man cannot even feel yet.

For the opening of the battle, the tunnelers had a bigger job: to plant enormous mines directly under the German front line. The one called Lochnagar was started back in November 1915 and grew into a gallery 310 meters long, 16 meters below the surface, ending in two chambers packed with a combined 60,000 pounds of ammonal. The Hawthorne Ridge mine near Beaumont Hamel held 40,600 pounds at the end of a long gallery 57 feet under a German redoubt.

The Weissas mine held 40,000 pounds. Alongside these, there were smaller charges of 9,000, 15,000, and 25,000 pounds. The tunnelers also pushed shallow tunnels called Russian saps out under no man’s land, which would be broken open the moment the attack began so the infantry would have covered routes partway across.

Some of those saps were hiding a weapon that almost nobody on the British side had ever heard of: the Livens Large Gallery Flame Projector, a machine 56 feet long and weighing 2. 5 tons, which threw a jet of burning oil out to about 90 meters. It took a carrying party of 300 men to haul the thing forward in pieces and assemble it underground in a sap out under no man’s land, and then a crew of eight to actually work it.

Four of them were installed opposite the German line in the southern sector, but they would never all fire because German shell fire caved in two of the saps just days before they could be used.

On June 24, the bombardment that the entire plan rested on finally opened. It was laid out as five days of shelling that would cut the wire, flatten the trenches, and break into the dugouts while gas and trench raids wore down the garrison. On the third day, gas was discharged at six points.

The problem was what was actually flying at the Germans. British industry had been expanded in a huge rush after the Shell 𝒔𝒄𝒂𝓃𝒹𝒶𝓁, and in that rush, it sacrificed quality for quantity. Around two-thirds of the more than one and a half million shells fired that week were shrapnel.

A shrapnel shell throws a spray of lead balls forward, which makes it a weapon for killing men standing in the open. Against men sitting 30 feet down in chalk, it did exactly nothing.

High explosive, the thing that actually wrecks trenches, barely existed for the field guns. The heavy guns had been increased from 324 to 714, which sounds impressive until you spread them across a 25,000-yard front and defenses miles deep. On top of all of that, roughly a third of all the shells fired that week were duds that never went off at all.

Meanwhile, south of the river, the French artillery had a much larger supply of high explosive for a much shorter stretch of front. The difference in what their shelling actually destroyed was visible. But the men in the British trenches were not told any of this.

They were told the wire would be gone, the Germans would be dead, and that they could cross no man’s land at a steady walk and simply take possession of what was left.

The orders were built on that exact belief. The men were not to run and not to stop to shoot. Many of them were carrying about 66 pounds of ammunition, tools, sandbags, wire cutters, rations, grenades, and gas helmets because they were expected to hold whatever they captured against German counterattacks.

A man loaded with 66 pounds could not have moved much faster anyway. The rigid waves existed because the command believed their half-trained volunteers could not handle anything more flexible. They wrote orders that left them no judgment of their own: walk forward on schedule and do not think.

Then the weather made everything slightly worse. Rain and mist ruined the aerial and ground observation that the guns depended on to aim. At 11:00 in the morning on June 28, the attack, which was set for the 29th, got postponed 48 hours to July 1.

The bombardment now had to stretch across two extra days, with ammunition economies ordered to make it last. The Germans got two more days of shelling to sit through, and the assault troops who had already moved forward got two more days of waiting in trenches soaked by the very same rain that had forced the delay. By June 30, the Germans did not even need to guess anymore.

A German listening post intercepted a British telephone message wishing the assaulting troops good luck.

The German command opposite read it right as proof that the attack was hours away. The Germans quietly pulled their men out of the threatened shelters and resighted their machine guns to sweep the approaches instead. The shelling had wrecked the trenches on the surface and killed men wherever it caught them in the open.

A direct hit from a heavy shell could blow in a dugout mouth or bury a shelter together with its garrison. Being sealed inside one of those is a way to go that is hard to describe. But the deep chalk shelters mostly held, and the garrisons inside them were alive.

They spent the week below ground, and every man down there understood the arithmetic of what was coming.

When the shelling lifted, it would be a race. The garrison had to reach the parapet before the attacking infantry reached the wire. The entire battle compressed into a sprint up a staircase.

The British knew things were not right. Night patrols sent out during the bombardment week kept coming back reporting that the wire was incompletely cut in many sectors. The night before the attack, parties even crawled out into no man’s land and blew gaps in the intact wire with explosive tubes called Bangalore torpedoes.

None of it changed anything at the top. The attack was not moved by a single hour.

There was one last argument over the preparation that never really got resolved: when to fire the mines. The commander in that sector, Hunter Weston, wanted the Hawthorne Ridge mine blown four hours before zero so his men could seize the crater and fortify it in darkness before the general attack went in. The army’s inspector of mines objected, and his reasoning was hard to argue with.

The British had never once beaten the Germans in the race to occupy a fresh mine crater. A mine fired four hours early would just hand the Germans a fortified hole, plus four hours of proof that the attack was coming. They settled on a compromise.

The Hawthorne mine would go up at 7:20 in the morning, 10 minutes before zero. Every other mine would go at 7:28, two minutes before the infantry.

By the evening of June 30, the assault battalions were filing into the forward trenches. Most of them were the men who had enlisted together in the autumn of 1914, standing shoulder-to-shoulder with their neighbors and workmates. Almost none of them had ever been in a battle.

Zero hour was set for 7:30 the next morning in full daylight. At 7:20 in the morning, July 1, 1916, under the German strong point on Hawthorne Ridge near Beaumont Hamel, those 40,600 pounds of ammonal went up. The official reasoning for the 10-minute gap was that the debris needed time to fall back to the ground before the infantry went over.

The result was the worst of both choices. The explosion killed about 30 Germans and trapped more in collapsed shelters, but it also told every German gunner on that part of the front that the attack was coming right now and handed them 10 free minutes to prepare.

The British had never yet won the race for a fresh crater. They did not win this one either. The Germans got there first.

Fire pits, wire, and defensive positions were built right into the new crater. The British party that rushed it managed to grab the near lip and hold it until about noon, and then a German counterattack threw them out. At 7:28, two minutes before zero, every other mine along the front went up.

Nineteen craters were blown that morning in total. They were supposed to help the attack, destroy strong points, and give the infantry cover. In practice, most of them became traps.

The Germans got machine guns onto the crater lips, and the men who climbed inside for shelter found out they could not climb back out under fire.

As for the Weissas mine, all 40,000 pounds of it, the Germans had quietly pulled their men out of those shelters. One of the biggest mines of the morning blew up ground the Germans had already left. Then there was the mine that went off late.

The officer at the plunger hesitated because he could see British troops already out of their trenches and moving. When he finally fired the 5,000-pound charge, the blast threw its debris outward across the advancing infantry instead of mainly upward, causing casualties in at least four British battalions. However, that same explosion destroyed German machine gun positions at the exact moment their crews were coming up to fire, and it landed on defenders who figured the danger of mines had passed.

British accounts of the day rate it, despite the men it killed on its own side, as the most effective mine of the morning.

At 7:30, the British barrage lifted off the German front line. The officers blew their whistles, and the infantry of 13 British divisions climbed out of their trenches. The next half hour had already been decided underground.

The German infantry had sat out the entire seven-day bombardment in those dugouts cut deep into the chalk, and the shelling had never reached them. The moment the barrage moved off, the race began: British infantry crossing no man’s land against German infantry running up the dugout steps with their machine guns. On most of the front, the Germans won it.

Their own accounts of the day record that the machine gun positions had survived the bombardment and opened fire immediately, and that the defenders were standing on their parapets while the attackers were still out in the open.

The British Army suffered 57,470 casualties that day, and 19,240 of those men were killed. It is still the worst single day in that army’s entire history. German losses are less certain because German units reported casualties in 10-day periods, but the estimates run from about 8,000, including some 2,200 taken prisoner, up to 12,000.

To understand what that number looked like on the ground, one must walk the front from north to south. At the far northern end at Gommecourt, the attack was designed to fail in a particular way. It was a diversion meant to pull German guns and attention away from the main assault.

For that to work, the Germans had to see it coming. Nobody even tried to hide it.

For five days, the British advertised themselves with smoke discharges, gas, wire cutting bombardments, and openly visible preparation. Four days before the attack, the corps commander told Haig plainly that the Germans knew they were coming. Two divisions were sent into the situation this had created.

The 56th London Division actually broke into the German lines and took the first two trenches. Then a German artillery barrage came down across no man’s land behind them and sealed them off. No ammunition and no reinforcements could get through.

They were pushed back trench by trench through the whole day. The last 75 men held out until about 9:30 that night, ran out of grenades and ammunition, and crossed back under fire.

The 46th North Midland Division, attacking through mud and a smoke screen that got its men lost, never got a foothold in the German line at all. The 56th lost about 4,300 men at Gommecourt, more than 1,300 of them killed or dead of wounds. The 46th lost 2,445 against about 1,255 Germans.

What happened afterwards tells you a lot about the army these men were serving in. The commander of the 46th, Edward Montague Stewart Wortley, was sacked on July 5. His division had suffered the fewest casualties of all 13 British divisions that attacked that day, and that exact fact was held against him as proof that his attack had not been pressed hard enough.

The first of several generals removed that summer was the one with the smallest casualty list. If you did not lose enough men, you clearly had not tried hard enough.

South of Gommecourt, in front of the village of Serre, the volunteer battalions of the northern towns attacked uphill against uncut wire. This is where that promise from 1914 that friends could serve together got paid for. The Accrington Pals sent about 720 men forward and lost 585 of them, 235 killed and 350 wounded, in about half an hour.

A battalion signaler watching from behind described the lines of men being mown down like meadow grass and wrote that he wept at the sight. Right beside them, the Sheffield City Battalion, a unit of students, clerks, teachers, and tradesmen, lost around 500. One of its survivors remembered Germans starting to come out of their trench with their hands up, then looking around at how few attackers were actually left, changing their minds, and going back to their guns.

The two Bradford Pals battalions, about 2,000 men between them, are reported to have lost 1,770 in the first hour. Most of these men fell in no man’s land before they ever reached the German wire. A few small parties made it into Serre itself, and they were killed or captured there.

The village did not fall. Below Serre at Beaumont Hamel, one of the most haunting pieces of film of the entire war was shot that morning. The official cameraman, Geoffrey Malins, had come up through a Royal Engineers tunnel at 6:30 to film the men of the First Lancashire Fusiliers, who were waiting in a sunken lane out in no man’s land.

They had been moved into the lane overnight because it lay closer to the German line than the front trench did.

Malins filmed them sitting against the bank, fixing bayonets, listening to the bombardment, men who knew full well what they were about to do. About an hour later, they attacked straight out of that lane, and the German machine gunners, who knew perfectly well that the lane was occupied, were waiting for them. The battalion lost 163 killed, 312 wounded, and 11 missing.

The footage went into the official film of the battle, and the faces in the sunken lane became some of the most watched images of the entire war. Most of them are the faces of men with minutes left to live. About two hours later, on the same ground, the Newfoundland Regiment was destroyed, and its destruction came down to a signaling error.

The first attack at Beaumont Hamel had already failed. A white flare was spotted and read as the British success signal. The problem was that a white flare was also the German signal telling their own artillery it was dropping short.

On the strength of that one flare, the divisional commander ordered the reserve brigade forward. The Newfoundland colonel questioned whether the German front line had actually been taken, and the answer he got was that the situation was not cleared up. The order stood.

The communication trenches were choked with the dead and wounded of the first attack. The regiment climbed out of a support trench called St. John’s Road behind the British front line and advanced in the open in full daylight down a bare slope, completely alone.

The battalion that was supposed to attack beside them had been delayed. The machine guns cut them down before most of them even reached the British front trench. The battalion’s own records say 721 men went forward and 619 of them became casualties in about 30 minutes.

At roll call the next morning, 68 men answered their names. There is a detail from that morning that is hard to shake off. The men of that division went over with triangles of polished tin fixed to their haversacks so artillery observers could follow the advance by the flash of the metal in the sun.

On a slope now covered with wounded men, the triangles kept doing their job. Every time a wounded man moved or tried to crawl back toward his own line, the tin flashed and handed a German sniper his exact position.

The badly hit lay out there through the day with thirst, blood loss, and pain, while the Germans put up flares through the night to keep watch over the ground. Clearing the wounded from in front of Beaumont Hamel took until July 6, done in short rushes under shelling and sniper fire. One Newfoundland soldier came in still alive on the fifth day.

The regiment’s casualty rate that morning, about 85%, was the highest of any unit on the entire front. Back home in Newfoundland, the casualty lists came out slowly and incomplete in the newspapers. Families waited for weeks to learn whether their men were among the 68.

At Thiepval, the fortified village on the ridge at the center of the front, the defenders of German Reserve Infantry Regiment 99 came up from cellars and dugouts. The bombardment had never broken, and their machine guns stopped the 32nd Division, Salford Pals included, out in no man’s land. The one gain of the day there was made by a battalion that simply refused to wait.

The forward companies of the 17th Highland Light Infantry crawled out at 7:23 and lay down 30 to 40 yards from the German trench. When the barrage lifted at 7:30, they were inside the Leipzig Redoubt before the garrison could get up the dugout steps from the quarry underneath it. They won the race by 𝒸𝒽𝑒𝒶𝓉𝒾𝓃𝑔 the start.

When they pushed on toward the next trench, the machine guns in a strong point behind it called the Wonderwork caught them in the open and drove them back to the redoubt. Parties of the Salford Pals and other battalions reinforced them there through the day, and that captured quarry was the only piece of German line the division still held when night fell. The day cost the 32nd Division 3,949 men.

Right next door, an entire division had the same idea, and it worked even better right up until it did not. The 36th Ulster Division, attacking next to Thiepval, did what almost nobody else managed that morning and was destroyed for it.

Its commander had made the same calculation as the Glasgow men: the gap between the barrage lifting and the infantry arriving was the entire problem. He pushed his leading waves out into no man’s land before zero hour to lie close under the shellfire. When it lifted, they rushed the German front line and took it.

Then they took the Schwaben Redoubt, the fortress on the crest behind it, and parties even pushed on toward the German second position. Then the plan around them collapsed into their flanks. Thiepval on their right had not fallen.

The ground on their left had not fallen. The captured redoubt turned into a wedge driven into a line that still held on both sides of it, under fire from everywhere, and cut off from reinforcement.

The redoubt changed hands in fighting through the day. At about 10 that night, a bombardment and an attack by two fresh German battalions drove the survivors out for good. The division lost more than 5,000 men, 2,069 of them killed.

By the end of the day, it held nothing but its own casualties. At La Boisselle, one brigade was asked to cross a mile of open ground before it even reached the battle. The four battalions of the Tyneside Irish, the reserve brigade of the 34th Division, started the day on the Tara and Usna ridges about a mile behind the British front line.

The plan had them advance overland from there instead of through the packed trenches. When the Lochnagar mine went up at 7:28, they came down that open hillside in long lines in full view.

The German machine guns reached them almost immediately with enfilading fire, coming from the side, running along the lines of men instead of across them. Men were falling a mile from no man’s land before the attack had even properly begun for them. Within minutes, the brigade had taken heavy casualties.

One account describes the machine guns cutting the infantry down the way a scythe cuts hay. Small parties kept going anyway, crossed no man’s land, and got as far as the German second line. Some of the survivors reached the shelter of the Lochnagar crater, where they held on among the wounded until evening.

The attack did succeed in exactly two places at the southern end of the British line: at Mametz and Montauban. The reasons were methodical, not heroic. The guns were simply denser there.

At Mametz, there was one heavy gun for every 58 yards of front. At Montauban, an artillery advantage of roughly six to one. The bombardment had actually done its job, cutting the wire and crushing the trenches.

The 7th Division at Mametz attacked partly out of those Russian saps, so the distance to cross was short, and four small mines went up under the German line there just before the assault. At Montauban, the 18th and 30th divisions walked in behind a barrage that lifted forward in time steps just ahead of them.

This was the first attack where British records used the word creep. They had been trained in mopping up, leaving parties behind to clear each captured dugout so the men in front would not get shot in the back. The German regiments opposite them, undermanned and with much of their artillery knocked out, were broken.

One of them was down to a few dozen organized survivors by the end of the day. Both villages fell at a cost of about 3,380 casualties to the 7th Division, heavy anywhere else, light on the Somme on July 1. Then there were the French, whose battle this was originally supposed to be, attacking astride the river to the south.

Their day looked like a completely different war.

North of the Somme, their 22nd Corps went over at 7:30 right beside the British and took its objectives. South of the river, the Colonial Corps and 35th Corps attacked at 9:30, two hours later on purpose. The German first position there was taken within about two hours at very small cost.

The French infantry did not advance in slow, even waves. They moved in small groups, rushing from obstacle to obstacle, behind far more heavy artillery than the British had, about 10 heavy batteries per kilometer of front, and with far more experienced men behind the rifles. Joffre had a standing instruction that infantry were only to be sent against positions the guns had already destroyed.

The colonial divisions raised from France’s colonies, including the Senegalese Tirailleurs, African infantry who fought in pretty much every major battle on the Western Front, took Dompierre and Fay, and the German first position south of the river. The Colonial Corps alone brought in about 5,000 prisoners. The French 6th Army’s losses on July 1 were about 1,590 men.

A few hundred yards away, the British had lost their 57,470 for objectives the French mostly took and the British mostly did not. Across its first two days, the entire Sixth Army lost about 3,300 men. None of that gap came down to courage or to the ground they crossed.

It came down to experience, artillery density, surprise where the French had it, and tactics built on the assumption that the bombardment might have failed instead of trusting that it had worked.

When the light faded on the evening of July 1, the day was not over. Between the lines, there were still thousands of wounded men lying out in no man’s land. What happened to them over the following days is a horror all of its own.

The sun went down on July 1, and out between the lines lay thousands of wounded men. There was no general truce that night and no organized rescue either. In a few places, small local truces were quietly observed while the stretcher parties worked.

But along most of the front, any movement in no man’s land drew fire. The wounded simply stayed out there. How long a man lasted came down to his wound, the weather, and pure luck.

Stretcher parties kept bringing men in for days, and individual survivors kept turning up long after that. The record probably belongs to Private A. Matthews, shot through the thigh on the 1st of July.

He lay in a disused trench for 14 days. He kept himself alive on his iron rations, on biscuits, and water passed to him by other wounded men stumbling past, and on rainwater he caught in his helmet after a shell splinter punched a hole through his water bottle. He was found completely by accident by an officer on patrol who nearly stepped on him.

He came into the dressing station a skeleton. That is one of the longest survivals in no man’s land recorded in the entire war.

The men whose job it was to bring the wounded in were few, and the arithmetic stacked against them was hopeless. A battalion of around a thousand men carried 16 regimental stretcher bearers, four per company, all of them unarmed. In peacetime, a lot of them had been the battalion’s bandsmen.

On firm ground, two men could carry a loaded stretcher. On shelled ground, it took four. In deep mud, it took six.

One bearer wrote about the exhaustion of six men hauling a 12-stone casualty a thousand yards across churned fields. A battalion that had just lost 500 men still had exactly 16 bearers to bring them all in. The shortage after July 1 was so obvious that battalions later in the war had their bearer numbers increased several times over.

Behind the bearers, the whole medical chain was drowning in the same way. The system ran from aid post to dressing station to casualty clearing station to ambulance train to base hospital, and it had been built for a fraction of what was now arriving. Somewhere between 16,000 and 20,000 wounded men came into it on the first day alone.

The stations filled up within hours. Men whose lives depended on being operated on quickly lay in rows on stretchers in the open instead, waiting for a table. Out of that overload came something called the Moribund Ward.

Men judged to be beyond saving were moved into a separate ward, their pain eased with morphia as far as supplies and time allowed, and left there to die.

The surgeons’ hours could go to the men who might actually live. It was rational and it saved lives in total. It also meant that some men who might just possibly have been saved were set aside on a judgment made in a few moments.

There was also something specific about the Somme that killed men who would have survived the exact same wound somewhere else. That something was the ground itself. This battlefield was farmland, plowed and manured for centuries.

Every shell burst drove that soil, absolutely loaded with bacteria, deep into torn flesh along with bits of cloth and metal. Gas gangrene, an infection that civilian surgeons working in clean conditions almost never saw, could take hold within hours, swelling and blackening the limb and killing the man through toxemia.

Tetanus came out of the same soil. The only real defense was speed and the knife, cutting away every scrap of dead and contaminated tissue. A man’s chances came down largely to how many hours passed before a surgeon got to him.

On those first days, it was usually too many. Some of the other numbers from this battle are just as stark. Early in the war, about 80% of men with a thigh bone shattered by a shell died, mostly on the journey back, until the army brought back a simple Victorian traction splint called the Thomas splint.

That one plain piece of kit turned the survival figures completely around. Then there were the faces. Thousands of men survived the Somme with their lower faces shot or torn away, alive only because the wound had missed the brain and the big blood vessels.

A surgeon named Harold Gillies treated some 2,000 facial cases from this one battle. His work led to the opening of the Queen’s Hospital at Sidcup in 1917, the first hospital dedicated to rebuilding faces. That is where he developed the tubed pedicle graft, where a flap of the patient’s own skin was rolled into a tube and then walked across the body toward the face in stages, still connected to its blood supply the whole way.

Rebuilding a single face took years and many operations. The artist Henry Tonks recorded the stages in portraits that are among the most unsparing medical documents of the war. These men survived the battle, and then their treatment went on for years after the battle itself was over.

Back home, the first news of the day was news of a victory. The reporting system could not have produced anything else. Five accredited correspondents covered the entire front, working under military censors and under the Defence of the Realm Act.

On top of all that, they censored themselves. The Reuters dispatch on the day itself said the day was going well for Great Britain and France. The headlines announced a great British offensive with casualties not heavy.

On July 3, the Manchester Guardian called the offensive very satisfactory and rich in promises. One of those correspondents, William Beach Thomas, admitted years later that he was deeply ashamed of what he had written.

The truth arrived anyway, just in a different form. It came as telegrams and casualty lists, and it came town by town because the army had been recruited town by town. In Accrington, a rumor spread that only seven men of the Pals had survived.

A crowd surrounded the mayor’s house demanding information. The rumor was wrong, but with 585 casualties out of about 720 men, it was near enough to the truth. Across the Pals towns of the north, the lists ran street after street.

The flaw built into the whole scheme back in 1914 was now sitting on the front doors of single neighborhoods.

The battle did not pause to catch its breath after the first day. Through the first fortnight of July, the fighting shifted into the belt of woods behind the German first position. Divisions were fed into them one after another.

Trones Wood changed hands over and over between July 8 and 14, attacked by the 30th Division until it was worn out, then by the 18th, while the Germans fed fresh regiments in from the other side. Shellfire brought the trees down across the path until the wood was one tangle of trunks, wire, and the dead of both sides. Units lost their direction and ended up fighting each other at a range of yards.

The 30th Division, already chewed up since July 1, lost another 2,300 men in there.

The final successful assault alone cost the 18th Division’s attacking brigade 1,527 men. Then there is Mametz Wood. The 38th Welsh Division, a New Army division raised in Wales, lost over 400 men in a first failed attack on July 7 and nearly 4,000 by the time it was relieved on July 12.

The wood was taken in fighting among shattered trees where survivors described severed limbs lodged in the branches and bodies strewn through the undergrowth. What happened to the 38th afterwards was a second injury stacked on top of the first. Its commander, Ivor Phillips, was removed in the middle of the operation.

Haig wrote that the division had not attacked with determination.

That judgment 𝓵𝓮𝓪𝓴𝓮𝓭 downward and hardened into barrack room talk across the whole army that the Welsh division had bolted. The division was never used on the Somme again, and its men carried that slur for decades. Never mind that the wood had actually been taken at the cost of a quarter of the division.

Never mind that the failures owed far more to artillery coordination and impossible ground than to the infantry. The division’s performance at Pilckem Ridge in 1917 began the correction, and historians finished it much later. But most of those men never lived to see their reputation repaired.

The French half of the battle was also being fought by men from far beyond France. One village is enough to show it. On July 4, the Foreign Legion’s marching regiment, about 2,000 foreign volunteers from many countries, attacked Belloy-en-Santerre across 800 meters of open ground.

The machine guns caught the leading waves. The regiment lost more than 800 men that day, and it still took the village along with about 750 prisoners. Among the dead was the American poet Alan Seeger, who joined the legion back in 1914 and who had earlier written a poem about keeping a rendezvous with death.

The Catalan poet Camil Campmany died in the same attack. The African infantry of the colonial divisions fought on through those same weeks south of the river.

Almost none of this ever made it into the English language memory of the Somme, which tends to begin and end with the British first day. The French themselves had spent that first fortnight advancing further than anyone else on the Somme, and then they were stopped south of the river. They took the entire Flaucourt plateau in the first 10 days, pushing up to 10 kilometers in places and taking around 12,000 prisoners and 85 guns for comparatively small losses.

On July 9 and 10, they took Barleux and the fortified estate of L’Ecluse on the south bank directly facing Péronne, the main German-held town of the sector. Then the front there just congealed.

One account has Joffre halting the exploitation as early as July 3 to keep the French effort tied to the pace of the British beside them. French records point instead to German reserves arriving and to fog and rain ruining the artillery observation that everything depended on. What the infantry knew was much simpler.

They had broken through further and cheaper than anyone. Péronne was sitting right there across the river, and the advance stopped without the breakthrough ever being used. Barleux and L’Ecluse then settled into attack and counterattack, with the Germans throwing six separate assaults at L’Ecluse on July 17 alone and Barleux changing hands twice over the following days until the ruins just sat in the front line.

On the German side, the first days of July produced an order that would shape the entire battle. The German command position, driven by Falkenhayn, was that not one foot of ground was to be given up voluntarily and that any position lost had to be retaken by immediate counterattack whatever it cost. The Second Army’s commander, Fritz von Below, enforced that on the Somme to the letter.

It meant German infantry spent the whole summer doing the very thing the British had done on July 1: climbing out of cover and crossing open ground against machine guns over and over again to retake trenches that were often not even worth holding. German losses climbed accordingly.

German divisions on the Somme lasted only two to three weeks in the line before they were burned out and had to be swapped for fresh ones. The battle turned into a mill that was grinding up both armies at the same time. German survivors of that summer left their own accounts, Ernst Jünger’s diary being the best known, and they describe ground churned over until literally not one blade of grass remained.

In the second week of July, the battle quietly achieved the thing it had been launched for in the first place. The German effort at Verdun had already been losing momentum, but the formal end came on the night of July 11 to 12 after one last push reached the top of Fort Souville and could get no further.

Falkenhayn ordered the entire Verdun front onto a strict defensive. The pressure on the Somme, together with the Russian Brusilov Offensive in the east, had drained away the reserves and the heavy guns that Verdun needed. The Somme offensive’s original strategic purpose, relieving Verdun, was accomplished within two weeks of the first day.

The battle then continued for another four and a half months. What those months looked like starts with a single morning that proved July 1 never had to happen the way it did. On July 14, two weeks to the day after the disaster, the British showed what they could do when they were allowed to fight differently.

Six brigades, 22,000 men, were assembled by night out in no man’s land, some of them within a few hundred yards of the German second position on Bazentin Ridge, guided through the dark by engineer tape laid out on compass bearings. Moving 22,000 men to within shouting distance of the enemy in the middle of the night sounds like a recipe for a massacre. The whole assembly cost the two most 𝓮𝔁𝓹𝓸𝓼𝓮𝓭 brigades exactly 12 casualties.

At 3:20 in the morning, instead of the usual half-hour bombardment the Germans had learned to expect, a hurricane bombardment of just five minutes came down on the trenches. The guns had already been quietly cutting up for days.

The infantry attacked out of the darkness right on its heels. The German wire and trenches were destroyed. The surprise was complete.

By mid-morning, Longueval and both Bazentin villages were taken with over 2,000 prisoners, a regimental headquarters staff among them. The attacking divisions lost 9,194 men, which is heavy by any standard in any war except the standard this battle had already set. One German regiment lost about 2,300 men that day.

This attack proved something that is uncomfortable to sit with. July 1 did not fail because the task was impossible. It failed because of how it was done.

That same morning also produced one of the emptiest moments of the entire battle. By about 10:00, officers walking forward from the captured ridge found that High Wood, the next feature ahead, was effectively undefended, with German troops visibly retreating past it. The divisional commander on the spot wanted to push his reserves straight in.

He was ordered to wait for the cavalry instead. The cavalry, which had been held far to the rear, took hours to come up over the shelled ground. In the early evening, the 7th Dragoon Guards and the 20th Deccan Horse charged across the fields below the wood in what is usually described as the only cavalry charge of the whole battle.

It was directed partly by an aircraft whose observer fired his machine gun into the German infantry hiding in the standing corn and dropped sketch maps down to the horsemen. Cavalry with lances being guided by an airplane. The charge killed or captured about 100 Germans, cost eight troopers killed, around 100 wounded, and 130 horses, and could not hold any of what it had ridden over.

By the time infantry entered High Wood that night, German reserves were already back inside it. The wood that had stood wide open at 10:00 in the morning then went on to consume British divisions for the next two months.

Right beside Longueval sits Delville Wood. On July 15, it was handed to the South African Brigade with the order to take it and hold it at all costs. That order was carried out almost literally to the last man.

The brigade, 121 officers and 3,032 men, took nearly all of the wood quickly, and then the German artillery closed over it. At the peak, shells were coming into that one square kilometer of trees at a recorded rate of 400 to 500 a minute, with some accounts claiming even more. Rain turned the ground into paste, and the men could not dig proper trenches because the ground was one solid mat of tree roots and buried bodies.

They lay in scrapes and shell holes among the falling trees for five days, beating off German attacks hand-to-hand in the pauses between bombardments. When the survivors were relieved on July 20, 29 officers and 751 men marched out. That is a loss of about 80%.

The fighting in and around that wood went on for weeks after they left it. The next horror did not happen on the Somme at all, but it happened because of the Somme, so it belongs here. Far to the north in French Flanders, an attack went in at Fromelles on July 19, meant to pin down German reserves that might otherwise be moved south.

The catch was that the reason for it had already evaporated, and the security around it was a joke. Civilians behind the lines were openly discussing the date. German observers sitting on Aubers Ridge with cameras and camouflage posts up in the trees had watched the entire assembly.

The infantry, the Australian 5th Division in its very first battle on the Western Front alongside the British 61st Division, attacked at 6:00 in the evening in full daylight against a concrete machine gun fortress called the Sugarloaf. The third German line their orders told them to consolidate turned out not to exist. It was a string of abandoned ditches that had been misread on air photographs as trenches.

In one night and the following morning, the Australian division lost 5,533 men, about 470 of them taken prisoner. The British division beside them took further heavy losses, and not a single yard of ground changed hands. A miscommunication even sent one Australian attack forward after the British half of it had already been cancelled.

An Australian corporal wrote that the dead lay as if the stock of a thousand butcher shops had been cut up and strewn about. The Australian brigade commander who met the survivors coming back wept and called the operation a tactical abortion. No man’s land was left full of wounded men who could not be brought in under the German fire.

Four days later, the Australians on the Somme itself went into Pozières. The fortified village sitting on the crest of the ridge the British had been climbing since July 1. The 1st Australian Division broke into it in a night attack in the early hours of July 23, and the rest of the village fell over the next two days.

Then came the part that made Pozières infamous. Once the village was taken, the Germans, obeying that standing order to retake everything, turned every gun that could bear onto the village and the two narrow approaches leading into it and just kept firing for seven weeks. The three Australian divisions that rotated through that place lost about 23,000 men, close to 6,800 of them dead.

Survivors of Gallipoli said the shelling was worse than anything they had known there. Australia lost more men in eight weeks at Pozières than in eight months at Gallipoli. The official historian Charles Bean later wrote that the windmill site above the village marks a ridge more densely sewn with Australian sacrifice than any other place on earth.

The particular horror of Pozières was burial alive. The shelling was so constant and so heavy that trenches simply closed over the men inside them. The living dug the buried out with their bare hands between salvos, got buried themselves, and were dug out in turn.

There is a set of letters from Lieutenant Alec Ross, a Melbourne journalist who came to Pozières in late July looking for news of his missing brother. He wrote home that the wounded and the killed had to be thrown to one side as the trench was rebuilt, that he was buried twice himself and thrown down repeatedly, that he lay buried among the dead and the dying, and that several of his friends were raving mad. He described himself as lousy, stinking, ragged, unshaven, and sleepless, with his tunic rotten with other men’s blood and partly splattered with a comrade’s brains.

His brother, Goldie, had already been killed on the night of July 28 to 29. Alec himself was killed on August 22 by shell concussion that left no visible wound on him. Neither brother’s body was ever found.

Beyond Pozières, the Australians were pushed on along the ridge toward Thiepval at an ordinary ruined farm called Mouquet Farm. That farm defeated attack after attack for a reason nobody on the attacking side fully grasped until far too late. The position they kept attacking was not really on the surface at all.

The Germans had linked the farm’s deep cellars into a whole system of dugouts and tunnels. Attackers could fight their way across the farm, report it captured, and then get shot from behind as the garrison came up out of the ground they had just walked over. It is the lesson of the first day all over again.

The Australians were thrown back three times in August and early September at a cost of thousands. The 4th Australian Division alone lost over 7,000 men across its two tours in that sector. The Canadians who relieved them captured parts of the surface twice and were driven off it again, while the cellars underneath stayed German the entire time.

The farm only fell on September 26 to the British 11th Division when pioneers finally attacked the underground exits directly and smoked out the last 56 defenders with smoke grenades. A few miles east, the same lesson was being learned at the same price at Guillemont, a village that broke one British attack after another all through August.

Its garrison sheltered in deep mine dugouts, in cellars, and in a quarry position sitting just 20 meters from the British line. Its machine gun crews would let the attacking waves pass right over them, then come up and fire into their backs. Whole battalions vanished into that village and did not come out.

It finally fell on September 3 to the 20th Light Division with the 47th Brigade of the 16th Irish Division attached, moving behind a strict creeping barrage rolling forward at 50 yards a minute. The Connaught Rangers overran the quarry at a rush. One German battalion defending the village was destroyed almost to the last man, losing around 90% of its strength.

The pattern was the exact one built into this battlefield from the start. Wherever the Germans had two years worth of digging underneath the position, the surface could be captured any number of times without the position actually falling. High Wood ran through the whole summer as that same story in miniature after standing empty on the morning of July 14.

It was attacked by the 7th, the 33rd, the 51st Highland, and the 1st Divisions in turn through July and August with saps, mines, flame weapons, and local assaults that each took a corner of the wood and then lost it again. By mid-September, the fighting for it had cost around 6,000 British casualties.

One later claim, which cannot really be verified, puts the total of British and German dead inside that one wood in 1916 at 8,000 or more. All through that summer, mines were breaking alongside the bodies. The army sorted the broken men into two categories that decided the rest of their lives.

A man diagnosed with shell shock W, W for wound, was ruled to have been damaged by enemy action. He counted as a battle casualty. He wore a wound stripe and he could receive a pension.

A man marked with shell shock S, S for sick, was ruled to be suffering from nervousness. He was not a casualty and he got no stripe and no pension entitlement.

The official dividing line between the two was supposed to be physical proximity to a shellburst. In practice, the decision often sat with commanding officers, and it carried a moral verdict with it. Wounded man or weak man, casualty or potential coward.

Between July and December 1916, about 16,000 cases classified as Shell Shock W were shipped back to Britain. That flood forced the army to set up forward treatment units deliberately labeled not yet diagnosed nervous to keep the men near the line and return them to duty. At the far end of that same scale, some men whose breakdowns were absolutely real were court-martialed for cowardice or desertion instead of being treated.

Some of them were shot.

Above everything, a whole separate war was running, and it was run for the guns. The Royal Flying Corps began the battle with 27 squadrons and around 421 aircraft. Its central work was not dog fighting.

It was artillery observation and contact patrol. Slow aircraft flying steady, predictable lines over the German positions while the observers corrected the shellfire by wireless or searched the churned ground below for the flares of their own infantry. Through 1916, the RFC was losing on average two air crew every single day, and none of them had parachutes.

A story that gets repeated everywhere is that parachutes were deliberately forbidden so pilots would not abandon their machines.

No document supporting that has ever been found in the War Office records. The researched explanation is more mundane. The only parachute available at the time was a heavy static line type that worked fine from a tethered balloon basket.

The kite balloon observers actually did use them, but it was impractical in the cramped cockpits of the day. A workable free-fall ripcord parachute was not issued until years after the war. To a man sitting in a burning aircraft, that distinction meant absolutely nothing.

In September, the balance over the battlefield tipped when the first Albatros fighters and the new German hunting squadrons arrived on the Somme.

Jasta 2 among them under Oswald Boelcke. By November, the Germans had taken back control of the air. In late August, this battle did something no battle had ever done before.

It arrived in British cinemas while it was still being fought. The official film, The Battle of the Somme, put together from the footage Malins and McDow had shot around the first day, opened in London on August 21, 1916 in 34 cinemas at once. In its first six weeks, roughly 20 million people saw it in some 2,000 cinemas.

That is close to half the population of the country. Audiences saw the bombardment, the Hawthorne Ridge mine going up, those men in the sunken lane, columns of prisoners, and British and German dead on a screen for the first time.

Newspapers reported people fleeing some screenings at the sight of the dead. One sequence in that film, the famous one of men climbing out of a trench and falling as they crossed the wire, was staged for the camera at a mortar school well behind the front in mid-July. Audiences took it as real.

It became the most reproduced piece of First World War footage there is, and it is still routinely broadcast as real combat film to this very day. The film had its critics, too, though on grounds that had nothing to do with authenticity. The Dean of Durham protested against an entertainment that wounded the heart.

One London cinema put up a notice saying it would not show the film because this is a place of amusement, not a chamber of horrors.

At the end of August, the battle claimed the German high command itself. On August 29, 1916, Falkenhayn was dismissed as chief of the general staff, undone by the failure of his Verdun offensive and the endless drain of the Somme. Hindenburg and Ludendorff took over the German war effort.

Within days, the new command accepted the very thing Falkenhayn’s no retreat doctrine had refused to even consider. At a conference in early September, the building of an entirely new fortified line far behind the Somme was approved. By the end of the month, construction had begun on the Siegfriedstellung, which the Allies would come to call the Hindenburg Line, running some 90 miles from near Arras to Laon, built by tens of thousands of laborers, mainly Russian prisoners of war.

That decision was an admission that the German army no longer believed it could hold the Somme indefinitely. A shorter line could be held with fewer divisions, and the army was quietly preparing to give up at a time of its own choosing the very ground the Allies were still bleeding to take. The ruins the summer had been spent on kept falling one by one.

After Guillemont on the 3rd, the 16th Irish Division took Ginchy on September 9 in an evening attack, capturing the village and 200 prisoners and losing 4,330 men between September 1 and 10 across Guillemont and Ginchy combined. Among the dead at Ginchy was Tom Kettle, a former Irish nationalist member of Parliament, an economist, and a poet who believed that Irish independence and fighting Germany in a British uniform served the same cause.

Days before his death, he had written in a poem that he would have died not for flag nor king nor emperor, but for a dream born in a herdsman’s shed. In the same week, the French Sixth Army, reinforced from Verdun, broke into the German positions north of the river and took Bouchavesnes on September 12. Then came September 15 at Flers and Courcelette, and the British finally committed the weapon they had been hiding.

49 Mark I tanks were assigned to the attack. 32 reached their starting points. 18 actually went into action.

A crew of eight was sealed into a steel box built around an unshielded engine in temperatures that could pass 50 degrees centigrade, breathing engine exhaust and gun fumes, in deafening noise from the engine and the six-pounder guns.

When machine gun fire hit the hull, hot flakes and droplets of metal called splash came off the inside of the armor and flew in through the joints and vision slits, which is why the crews wore chain mail masks over their faces. The machines could walk over wire and cross most trenches, but on shelled ground, they pitched like ships in a rough sea, and they broke down constantly. Where they worked, the effect was enormous.

Flers fell with tanks rolling through the village. Courcelette fell to the Canadians, and the advance that day was about 2,500 yards on a three-mile front, with German troops reported fleeing in terror from machines they had no answer to.

The tanks came with a cost that was built directly into the artillery plan. Because they were too slow to follow a normal creeping barrage, gaps were left in the barrage along the lanes the tanks were supposed to take, which meant the German positions inside those gaps went unshelled. Wherever a tank broke down or never arrived at all, the infantry assigned to its lane attacked untouched defenses.

Formations that lost their tank found themselves too far behind the barrage to ever catch it again. At High Wood, the four tanks sent into the wood itself ditched and blundered around among the stumps and trenches. One caught fire, and one fired on British infantry by mistake.

What actually broke the defense of the wood in the end was not the tanks at all, but a hurricane bombardment of 750 mortar rounds in 15 minutes. That morning, the wood was also a stage for British weapons killing British soldiers. Drums of oil for the Livens flame projectors stored forward for the attack were buried by British shells falling short before they could ever be fired.

The pipe pushers, explosive-filled tubes driven under no man’s land to blast out instant trenches, fouled tree roots, and one charge turned back on itself and blew a crater in the very British front line it had been fired from. The 47th London Division took High Wood at last that day and paid over 4,500 men in four days for it.

Its commander was then sacked within a fortnight for what the corps commander called want and waste of men, a judgment that was later quietly walked back. As for the tanks, the wider lesson of September 15 was recorded both ways at once. The machine was mechanically nowhere near ready, and it was worth developing at full speed.

Haig ordered production increased on the strength of it. Courcelette that day also produced one of the battle’s frankest documents about what happened to men trying to surrender. During the Canadian attack, a group of Germans approached Captain Samuel Logan of the 18th Battalion, apparently to give themselves up, and one of them threw a grenade that killed him.

His company saw it happen. The brigade commander, Brigadier General Robert Rennie, afterwards wrote the whole matter into his report in a single sentence: this foul act of treachery was observed by the men of the captain’s company, with the result that none of the occupants of the trench were allowed to escape alive. Other Canadian accounts from the same day are just as plain.

One battalion colonel recorded that most of the enemy who offered to surrender were bayoneted by the advancing troops. A private of another battalion, asked about prisoners, said they were not taking any because the Germans had blown mines under them twice. The historian Tim Cook, who has studied this subject most closely, found ample evidence that the killing of surrendering men ran through the whole Canadian force from privates up to senior officers on many days besides this one.

In the last week of September, the objectives of the first morning finally fell. Three months late, on September 26, the 18th Division stormed Thiepval, the fortress village that had broken the 32nd Division back on July 1, with a single tank grinding through the rubble of the chateau to finish off its machine guns. Taking and clearing the village cost the leading brigade 840 of its 2,290 men.

That same day, the underground garrison of Mouquet Farm finally gave up. The fighting then rolled on up the ridge toward the Stuff and Schwaben Redoubts where the Ulster men had stood for one afternoon back in July, and it went on there into October.

That last week of September was also the end of the good campaigning weather on the Somme. What the weather did to this battlefield next is a horror story of its own. On October 2 at about 11 in the morning, rain began falling on the Somme, and it kept falling for the next two days.

There was one short break on October 4, and then the weather settled into rain, mist, sleet, and westerly gales that ran with barely an interruption through October and into November. The ground this rain was falling on had been shelled without pause since the end of June. There was nothing left in it to drain the water anywhere.

The topsoil had been churned into a layer of loose, broken earth pitted with shell holes.

The whole battlefield turned into a swamp. The supply routes crossing the ground fought over since July 1 became nearly impossible. That rain also set off a chain of events that decided everything for the next six weeks.

The British artillery could only shoot accurately at things its observers could actually see. Most of that observing was done from aircraft. Fog, rain, and high winds kept the planes on the ground, so the guns fired blind.

On October 2, eight German aircraft flew low over the British lines while every single British machine sat grounded. Without observation, the bombardments could not be corrected. The German wire was not properly cut, and the German batteries went unfound.

The attacks were ordered anyway. Those October attacks, grouped under the name of the Battle of Le Transloy, followed one after another straight into this swamp. On October 1, the New Zealand Division took its objectives, but by that point it was down to roughly 850 men, and it lost a good share of even that.

On October 6 and 7, the attacks were largely thrown back by machine guns firing from positions sited beyond the reach of the British barrages. On October 12, a bigger combined attack failed, partly because the fresh divisions fed into it had almost no time to study the ground or dig assembly trenches, and partly because the preparatory bombardment fired without air observation had not done its work.

On October 18, the attack went in at 20 to 4 in the morning, and the jumping off trenches were flooded at zero hour. The men climbed out already soaked, slipped and fell in the mud, and their rifles and Lewis guns clogged with it, which left them attacking with hand grenades and bayonets. They moved so slowly through the mud that they lost their own creeping barrage behind them.

The guns were wearing out from months of constant firing and were literally sinking into the mud where they stood, which threw their accuracy off even further. Bringing ammunition up to them got slower with every passing day. The German infantry holding the line through all this were in much the same state.

A German battalion commander in the sector reported cold food, constant artillery fire, a high number of casualties, and depressed morale. In some German units, a quarter to a third of the men were down with severe diarrhea. On both sides, the dead could no longer be removed from the trenches at all.

One place shows better than anywhere else what October cost. The Butte de Warlencourt is an ancient burial mound about 50 to 60 feet high near the village of Le Sars. The Germans had wired it and filled it with machine gun posts.

The 47th London Division attacked it on October 7. The 9th Scottish Division fought in front of it from October 12 to 24 and lost 3,255 men.

On November 5, the Durham Light Infantry battalions of the 151st Brigade attacked it after a rainstorm had turned the ground into mud so deep that moving a thousand yards took four hours. The infantry floundered in mud that was thigh deep in places, and several men drowned during the attack itself. Drowned during an attack.

The Durhams actually took the butte and got a machine gun up onto it. German counterattacks starting that same afternoon pushed them back off. By early afternoon the next day, they were right back where they had started, about a thousand men poorer.

On the other side of that mound, one Bavarian regiment holding the ground lost 1,177 men in 10 days and was down to about 350 men fit to fight.

Its own report said openly that being unable to bury the dead strewn around its trenches was wearing the survivors down. North of there, the Canadians spent those same weeks attacking one single German trench. Regina Trench was described as the longest German trench on the entire Western Front, and it swallowed the Canadian Corps for six weeks.

The 2nd and 3rd Canadian Divisions attacked it on October 1, got into parts of it, and were forced back out by a German counterattack at 2 the next morning. The second attempt on October 8 cost the most. The 1st and 3rd Canadian Divisions attacked at 10 to 5 in the morning in cold rain.

The bombardment, fired in high winds with poor visibility, had not cut the wire, and the Germans had spent the previous two days filling the few existing gaps with coiled concertina wire. The men who did find a way through took the front trench, ran out of grenades, and had to give it up again. The 1st Canadian Brigade lost 770 of the 1,100 men it sent forward.

The corps casualties for the day came to 1,364. On October 21, the 4th Canadian Division and three British divisions attacking beside it on a 5,000-yard front finally took the western part along with over 1,000 prisoners. The last section fell on the night of November 10 to 11 when the 46th and 47th battalions rushed it the moment the barrage lifted and caught the garrison by surprise at a cost of about 200 men.

By the end of the Somme, this battle would cost the Canadian Corps roughly a quarter of its entire strength. October also finally brought down the Schwaben Redoubt, a piece of ground whose history you already know. It is the fortress on the ridge above Thiepval, roughly 500 to 600 yards long and 200 deep, packed with machine gun emplacements and deep dugouts.

The one the Ulster men took in a single morning on July 1 and were thrown back out of before midnight. The British came back to it at the end of September when the 18th Division fought its way into the southern face on the 28th. For the next two weeks, the redoubt was fought over with grenades at close range.

The Germans counterattacked again and again to get it back, twice with flamethrower detachments on October 7 and 8 and again on the 15th. The last corner of it was taken on the night of October 14 by the Black Watch, the Cambridgeshires, and the King’s Royal Rifle Corps, who crossed the open ground and cleared it by 11 at night, taking about 150 prisoners. The 18th Division alone lost 1,990 men in the fighting for the redoubt.

German losses there were estimated at about 2,500. It had taken three and a half months to hold ground the Ulster had first reached in one morning. The French fought those same weeks in the same mud, and their part of it is the least told of all.

Their target on the northern flank was Sailly-Saillisel, a village on the Bapaume to Péronne road. The French 5th and 32nd Corps attacked toward it between October 3 and 8 and lost 5,987 and 8,244 men. The Germans fed in 36 fresh artillery batteries, and their pairing of machine guns with a mobile defense broke attempt after attempt while the mud crippled the resupply and the flying weather blinded the French guns just the way it blinded the British ones.

On October 15, the 152nd Infantry Regiment and the 68th Chasseurs got into the village. The fighting for it never really stopped. On November 11, the 150th and 155th regiments and the 9th Zouaves occupied it, only for a German division to counterattack the very next day and throw the Zouaves back out of the southern trenches.

Machine guns in St. Pierre Vaast Wood stopped any advance beyond it. South of the river, the French 10th Army took Ablaincourt and Pressoir by November 7 with over a thousand prisoners and paid for it in the other direction too.

On October 29, a German counterattack at L’Ecluse, that same fortified estate from July, supported by 20 aircraft, took 450 French prisoners. Among them was the son of a French general named Haire. That was six weeks of fighting.

The truth is that by November, the real enemy on the Somme was not wearing a uniform at all. If you asked any army on the Somme what the enemy was, the honest answer was not the men across the wire anymore. It was the mud.

Every account from every side describes it first. This was not ordinary wet ground. This was earth that had been shelled into powder for four months and then rained on for five weeks.

A deep paste with no bottom firm enough to stand on, spread across a landscape of flooded shell holes. Men lived in it up to their thighs for weeks at a time. A man who slipped off a wet duckboard track at night into a flooded shell hole could not always climb back up the greasy sides.

Men went under and they drowned. One veteran of that winter said it plainly: men simply drowned in the mud, and if nobody was beside them when it happened, there was no chance of rescue at all.

Getting somebody out took ropes and several men hauling. The 51st Highland Division issued ropes for that very purpose. It issued gum boots as well, and many of them were sucked clean off men’s feet and lost in the mud.

Standing in that water day and night destroyed men’s feet. Trench foot began as numbness. Then the feet swelled to two or three times their normal size and went completely dead.

So dead that, as one sergeant put it, you could stick a bayonet into them without the man feeling a thing. Then when the swelling went down, the pain arrived. In the bad cases, the tissue died and turned gangrenous, and the foot or the whole leg had to come off.

One soldier described watching men sob because their trench foot had gone gangrenous and they knew what was coming next. The army knew this enemy already because in the first winter of the war it had treated over 20,000 cases. By late 1916 there was a drill.

Men were ordered to carry three pairs of socks and change them at least twice a day and to grease their feet with whale oil. A single battalion in the line went through whale oil at a rate of around 10 gallons a day. Men worked in pairs and rubbed each other’s feet.

None of it could fully protect men who could not get dry for weeks on end. When the real cold came, frostbite joined trench foot.

The feeling came back into frostbitten feet only when the men were finally carried out, and some of them lost their feet entirely. Even the water they drank was its own small misery. It came up to the line in used petrol tins that were never properly cleaned out, so it smelled and tasted of petrol, and there was never enough of it.

Tea carried up communication trenches three-quarters of a mile long arrived cold, and in the deep winter, it arrived with ice on it. Water bottles froze solid. For washing and shaving, men used the warmest water they could find, which in practice meant shell hole water and saved tea dregs.

In the freeze, some men melted snow in cigarette tins and then watched it refreeze while they were still shaving.

The trenches they drank and slept in were shared with rats in swarms drawn in by the food and by the unburied dead. Rats gnawed through ration bags and ate men’s iron rations, and men were bitten in their sleep. One of them recorded his forehead bitten open.

Every single man was lousy. The lice carried a disease that nobody at the time understood. Trench fever came on with headache, dizziness, and a strange sharp pain in the shins, then a fever that faded and came back in waves for weeks.

It was not spread by the louse’s bite, but by the infected droppings of the louse being scratched into broken skin. There was no treatment.

Quinine, salvarsan, and everything else they tried against it did nothing. An average case kept a man away from the front for about three months. The British, French, and Belgian armies together had close to half a million cases across the war.

The actual cause was only officially pinned down at the end of 1917. Through all of this, the dead were part of the ground itself. Bodies from the summer fighting still lay in the open between the lines.

Neither side could clear its trenches of them during the October fighting. In places, the dead were built into the defenses. One soldier recorded that part of his parapet was made of sandbags and dead soldiers.

Sandbags and dead soldiers in the same sentence as the same building material.

The animals suffered on the same scale as the men. The British army bought over 460,000 horses and mules during the war, and it worked them hardest on ground like this, where every gun, every shell, and every ration had to be dragged forward. A horse needed about 10 times the food of a soldier, and all that fodder had to come up the same drowned roads as everything else.

Horses picketed in the open sank into the mud where they stood, and horses drowned in shell holes, just like the men did. Of all the horses and mules that died in this war, about three-quarters died of disease or exhaustion rather than enemy fire. The army was losing around 15% of its animals every single year.

Between the attacks, the line still had to be held, supplied, and relieved. That work killed men who appear in no account of any battle. Every ration, every water tin, and every box of ammunition came forward on men’s backs, mostly at night to avoid the German guns, across the same mud that took four hours per thousand yards.

Reliefs that should have taken an evening took whole nights, and men arrived at the front already finished. During one relief on the north bank of the Ancre that winter, a soldier died of exposure on the way in without a single shot being fired at him. Carrying parties and working parties were shelled on the tracks as a matter of routine.

Their dead were counted as the ordinary daily loss of holding the line.

The battle was not done. It had one more big attack left in it. The reasons behind that one were argued about at the time and are still being argued about today.

Haig wanted a success in hand before the Allied conference that opened at Chantilly on November 15, where the next year’s plans would be settled. He also argued that a blow would stop the Germans from moving divisions to Romania and would steady opinion in Russia. His own chief of staff later said that part of the purpose was to answer rumored hostility from Lloyd George back in London.

Some historians have called those justifications sophistry dressed around a battle that had run out of purpose.

Others have answered that the attack demonstrated who had actually won on the Somme. Either way, the Fifth Army under Gough was sent in the second week of November against the same ground north of the Ancre where the attack of July 1 had failed completely: Beaumont Hamel and Serre. The attack went in at a quarter to six on the morning of November 13 in the dark in a fog so thick that visibility was nil.

That fog hid the attackers in a way no bombardment of the entire summer had ever managed. A new mine of 30,000 pounds of explosive was blown in the old crater on Hawthorne Ridge, the exact spot where the famous mine of July 1 had gone up.

The 51st Highland Division went for Beaumont Hamel itself, and the Germans, blinded by the fog, simply could not see them coming across no man’s land. An early German account describes British troops appearing out of the mist in scattered groups, some of them smoking cigarettes, rifles still slung over their shoulders. The strong point called Y Ravine, a steep-sided gully riddled with tunnels and dugouts more than 20 feet down, held out all day with German parties surfacing behind the attackers out of tunnel mouths.

By evening, Beaumont Hamel had fallen. A village that had been an objective for the morning of July 1. The garrison that came up out of those dugouts had been living underground under constant shelling with their supplies almost completely cut off for the previous two days.

Even the German command judged the defending division exhausted. Down river, the 39th Division took St. Pierre Divion by 20 to 8 that same morning with well over a thousand prisoners, including an entire battalion headquarters.

The 63rd Royal Naval Division, a division of sailors fighting as infantry, took Beaucourt by the following morning at a cost of around 3,500 men. About 7,000 German prisoners were taken in the November fighting, more than in any comparable British effort of the whole summer. Ludendorff himself called the battle a particularly heavy blow.

Among the British dead of the follow-up fighting was the writer Saki, Lance Sergeant HH Munro of the 22nd Royal Fusiliers, killed by a sniper on November 14.

Serre failed again. The 3rd Division, attacking over the very ground where the Pals battalions had been destroyed on July 1, went into mud that reached the men’s waists. The official account calls the result a fiasco.

Their word, not mine. The two attacking brigades lost direction and cohesion in the swamp. By half past 6 in the morning, officers were trying to collect exhausted men scattered around shell holes.

At half 4 that afternoon, all operations against Serre were cancelled. Serre had now beaten a full-scale British attack at each end of the battle. It stayed German until the day the Germans chose to leave it.

Snow fell on the night of November 17. The last attack of the Battle of the Somme went in on the morning of the 18th through sleet that turned to rain, with visibility again at nil onto half-frozen mud that the sleet was busy thawing. South of the river Ancre, the 4th Canadian Division, attacking through that blinding sleet, took most of Desire Support Trench beyond Regina Trench along with 620 prisoners.

North of it, the 32nd Division attacked Munich and Frankfurt Trenches on Redan Ridge and were stopped almost at once by machine gun fire. Almost, because a few parties broke clean through the German line, and then they could not come back.

What happened to those men is one of the least known stories of the entire war. Somewhere between 130 and 165 men, the sources disagree, and recent research puts 145 names to it, most of them from the 16th Highland Light Infantry, the Glasgow Boys Brigade Battalion, a unit raised from a church youth organization, along with men of the 11th Border Regiment and the 2nd King’s Own Yorkshire Light Infantry, ended up holding a stretch of Frankfurt Trench 300 yards behind the German front line after the attack had collapsed behind them. Their orders said to hold for 48 hours until relief arrived.

No relief ever arrived, and every rescue attempt failed.

Around a third of them were badly wounded, and they were short of food, water, and ammunition from the very start. For the first five days, the Germans did not even realize they were there. They gave themselves away on November 23 by taking prisoner German soldiers who strayed too close.

Two German attacks on the 24th were then beaten off by men fighting hand-to-hand while sick and starving. On the seventh day, the Germans sent them an ultimatum: surrender or be wiped out. The garrison under Captain Alexander Welsh refused it.

The next day, November 25, German assault troops overran the trench. 15 men were still on their feet. The 141 taken prisoner included nearly all of the wounded, three of whom died in captivity.

A German general who inspected the survivors is said to have asked whether this was really what had held up a brigade for more than a week. With that, on November 18, the British stopped. That date became the official end of the Battle of the Somme.

But understand what that date actually marks because it is a British date and only a British date. The French Sixth Army fought on at Sailly-Saillisel through late November and was only withdrawn from the battle at the end of the month. French regiments were still moving into Somme sectors in the middle of December.

The end date everybody quotes marks the moment the British army stopped attacking, not the moment the fighting stopped.

What did four and a half months of this cost? There is a settled answer for only two of the three armies. The British and Empire forces lost 419,654 men.

Among them were 24,029 Canadians, around 23,000 Australians, 7,748 New Zealanders, and over 2,000 Newfoundlanders. 57,470 of the total fell on the first day alone. French casualties are given as somewhere between 154,000 and 202,000 depending on the source and the exact dates counted.

The German figure has never been settled and probably never will be. German official records support totals from the high 200,000s up to around 430,000. The British official historians then added roughly 27% on top of the German returns, arguing that Germany did not count its lightly wounded the way Britain did.

This eventually produced claims of 660,000 to 680,000 German casualties that later historians have largely rejected as inflated. Most later estimates run from around 400,000 to 500,000, with the old higher figures still being quoted here and there. The full range you will find in the books goes from 400,000 all the way to 680,000.

That entire gap comes down to a real difference in how the two sides counted their wounded. What is not disputed by anyone is that the year 1916, Verdun and the Somme together, cost the German army on the order of 1. 2 million men.

When the guns finally quieted down, the armies did not go anywhere. They spent the winter living on the very ground they had just spent half a year destroying. The official histories describe operations on both sides for the next two months as being mostly restricted to survival.

Survival in rain, snow, fog, mudfields, waterlogged trenches, and shell holes. That winter turned into one of the hardest anyone there could remember. Ambulance crews had to run their engines every 20 minutes to stop them from freezing solid.

Sheepskin coats and layers of extra clothing were issued against it. In mid-January, the ground froze hard for about five weeks, which at least made walking possible again.

A thaw beginning on February 16 turned everything back into mud, worse than before. Even in that state, the fighting never fully stopped. Small British operations along the Ancre from January into February pushed the Germans back about five miles on a four-mile front and took over 5,000 prisoners.

That same winter, a new group of men was brought onto the old battlefield, and their story went untold for a very long time. In 1916, Britain asked South Africa for laborers. The South African Native Labour Contingent was raised from September onward.

About 25,000 black South Africans enlisted, and around 21,000 of them went to France, where they unloaded ships in the channel ports, quarried stone, cut timber, and built and repaired the roads and railways that kept the entire front supplied.

They were forbidden to carry weapons. Off shift, they were locked into closed compounds ringed by six-foot fences topped with barbed wire, with armed guards on the exits, forbidden to enter a shop or a bar or to even speak with the local population. Observers compared the conditions to prisoner of war camps.

They were the only labor contingent in France housed that way. The South African government insisted on that system to preserve the racial order back home. In January 1918, it chose to disband the contingent entirely rather than relax the compounds.

The contingent’s worst single day did not even happen on land.

On February 21, 1917, the troopship Mendi, carrying a fresh draft to France, was rammed in thick fog off the Isle of Wight by a larger ship, the Darro, and sank quickly, drowning over 600 men, 607 by the most careful count. A story spread that the men met death performing a barefoot drill together on the tilting deck. That is the version remembered in South Africa to this day, though it has never been confirmed.

The men of the contingent were never awarded any war medal. Compare that. The white South African Brigade destroyed at Delville Wood, the one that went into the wood over 3,000 strong and came out with barely a quarter of its men, got a national memorial on the spot, inaugurated in 1926 on land France sold to South Africa for exactly one franc.

The black South Africans who worked the same battlefield got no medal, no memorial for decades, and a contribution that was deliberately forgotten. Through that same autumn and winter, the army was also shooting its own men. The instrument was the Field General Court Martial, a battlefield trial in front of a panel of at least three officers.

The trials were short, commonly under an hour, and the courts got through several cases in a day. The accused had no right to a lawyer. He could be assigned an officer as a so-called prisoner’s friend to speak for him, but about one executed man in 10 had nobody at all.

There was no appeal.

A death sentence traveled up the chain of command, brigade, division, corps, army, all the way to the commander-in-chief, who confirmed or commuted it. The process took two to three weeks. The sentence was carried out at dawn by a firing squad of 12 men, usually drawn from the condemned man’s own unit, his own friends.

Private Harry Farr of the West Yorkshire Regiment had served since August 1914. His case shows how those trials handled a man who had already broken. In May 1915, he collapsed with shell shock and spent five months in hospital at Boulogne.

In April 1916, he broke down again and spent two weeks at a dressing station. In July 1916, he was back at a medical station once more.

On September 17, 1916, with his battalion moving up for another Somme attack, he reported sick, was found to have nothing physically wrong with him, was ordered forward, and refused, saying he could not stand it. He was charged with cowardice. His court martial on October 2, 1916, lasted 20 minutes.

He defended himself without a prisoner’s friend, and he was not able to call the medical officer who knew his history. He was sentenced to death. At 6:00 in the morning on October 18, 1916, at Carnoy on the Somme, Harry Farr was shot by 12 men of his own regiment.

He refused the blindfold and looked at the firing squad. The chaplain who attended him wrote that a finer soldier never lived.

His wife Gertrude received a telegram telling her that her husband had been sentenced for cowardice and shot at dawn. His pension was stopped, and she and her daughter were left homeless. The wave of trials ran on through the winter.

The next case happened just off the Somme in the sector near Arras that the men had been moved to after the battle, but it belongs to the same story. On the night of November 26, 1916, a German raid hit a trench held by the 19th Durham Light Infantry. Lance Sergeant Joseph Stones, a small man who had been commended more than once for bravery and had fought through the Somme, was beside his officer when the officer was mortally wounded and was ordered to go for help.

He was found 700 yards back without his rifle, described by the NCO who stopped him as being in a pitiable state of terror. He said he had jammed the rifle crossways in the trench to slow down the Germans behind him. He was charged with shamefully casting away his arms and tried on Christmas Eve 1916.

This time the accused even had a defending officer who was a qualified solicitor, and his own company commander testified that Stones was the last man he would have thought capable of any cowardly action. It made no difference. Two lance corporals from the same night, Peter Goggins and John Macdonald, were condemned for leaving their posts during the same raid.

On January 18, 1917, at Rollencourt, the three men were brought in a motor ambulance, manacled, tied to three posts, and shot simultaneously by three firing squads of 12. The chaplain said he had never met three braver men. Across the whole war, the British army passed 3,118 death sentences and carried out 361 of them, around one in eight.

About three-quarters of the executed men were deserters. The courts were at their busiest in the months right after each great offensive. The French army passed around 2,500 death sentences and shot around 650 men.

The German army, whose courts martial sat with a legally trained judge and four officers and demanded proof of intent, passed about 150 death sentences, carried out 48, and shot exactly 18 men for desertion in the entire war against 272 British deserters shot.

In 2006, Britain pardoned 306 of its executed men by act of parliament. Harry Farr was among them. The Germans, meanwhile, drew their own conclusion from the Somme, and it changed how they fought for the rest of the war.

Holding to Falkenhayn’s order that the front line be defended to the death and any lost ground retaken at once whatever it cost through four and a half months of Allied artillery had helped cost Germany casualties that its own senior officers openly called out of all proportion to the ground involved. On December 1, 1916, the high command issued a completely new defensive doctrine.

The packed front trench was abandoned. The forward line would now be held thinly. The defense would stretch back in depth through zones of shell hole positions and machine gun nests.

Ground could be given up temporarily instead of being died for. Specially held back divisions would deliver the counterblow once an attack had worn itself out in the mud. From February 1917, the army was running week-long courses to teach its division commanders the new system.

The deeper decision had already been taken while the Somme was still burning. The Hindenburg Line was being built. One German commander summarized the reasoning behind it in a single sentence: that enemy superiority was now so great that Germany could not prevail in a second battle of the Somme.

What followed in the new year was not an Allied attack. It was a German departure. The Kaiser signed the withdrawal order on February 4, 1917.

Local pullbacks began first. On February 24, British patrols pushing through the mud found the German trenches in front of Serre empty. The next day, the 21st Manchesters walked into Serre, the village that had destroyed the Pals battalions on July 1 and beaten the 3rd Division in November, for the loss of one man killed.

The main operation, codenamed Alberich, ran from March 16 to 20, 1917, and pulled the German armies back up to 25 miles onto the new line, shortening their front by about 25 miles and freeing 13 to 14 divisions.

They destroyed everything as they went on orders. Across some 580 square miles, roads and railways were dug up, bridges were dropped, orchards and shade trees were felled, wells were wrecked and fouled, though German records state that outright poisoning of wells was forbidden. About 200 towns and villages were systematically demolished with delayed action mines and booby traps left behind in the ruins, in the dugouts, and on the roads.

More than 100,000 French civilians, some accounts say 125,000, everyone judged capable of work, were deported into occupied territory as labor. The children, the mothers, and the old were left behind in a few designated towns with rations for a few days.

Even Crown Prince Rupprecht of Bavaria, the army group commander who had to carry the order out, opposed the destruction as a stain on Germany’s name and on his troops’ discipline, and considered resigning over it, staying only because his resignation might have suggested a split between Bavaria and the rest of Germany. In a few weeks, Germany voluntarily handed over more French territory than the Allies had managed to take from it by force since September 1914. Incomparably more than the Somme offensive had gained in four and a half months of attacking.

The withdrawal opened up the old battlefield.

In the spring of 1917, the ground in front of Serre could be walked on in daylight for the first time since the 1st of July. The burial parties went onto it in May. Working parties sent to clear the old no man’s land and the slopes where the attacks of July and November had died.

What they found was the dead of the whole battle lying where they had fallen after eight months of summer heat, autumn rain, and winter frost in the open. The clearers gathered them into new cemeteries along the Serre road. The numbers from those cemeteries tell you how that work went.

Serre Road Cemetery Number One holds 2,434 men, and 1,728 of them, about seven in every 10, could not be identified. They lie under headstones without a name.

Serre Road Number Two, begun by the same parties in the same month, kept growing all the way until 1934 as searchers went on combing the battlefield for bodies year after year. Some of the graves made in 1917 were themselves later destroyed by shellfire. Those men are remembered by special memorial stones instead.

The men of whom nothing findable remained at all became names on the memorial to the missing later built at Thiepval. It carries 72,311 names, and every single one of them is a man of the Somme with no known grave. At Delville Wood, where the South Africans fought, the cemetery made afterwards holds more than 5,500 men, and over 3,500 of those are unidentified.

The German dead lying on the same ground were collected separately and buried in a completely different way. After the war, the French military authorities on whose soil all these graves stood took charge of the German burials. In 1920, they began concentrating German remains from 79 communes of the district into a single large cemetery at Fricourt.

It holds about 17,000 German soldiers, roughly 10,000 of them dead of the 1916 battle. Around 5,000 lie in individual or shared double graves under plain crosses, wooden at first, replaced with metal ones decades later. 14 of them are Jewish soldiers with headstones instead of crosses.

The other 11,970 lie together in four mass graves, the Kameradengräber, with plaques carrying the names of those who could still be named.

There was no individual headstone for every man there and no garden cemetery of the kind Britain was already planning for its own dead. This is how the Battle of the Somme actually ends. The army that held the ridges of the Somme for two years lies in shared graves under communal grass.

The army that spent four and a half months and over 400,000 men failing to take those ridges buried its dead one by one, most of them without a name, on ground the enemy finally gave up without a fight.