The morning of September 15th, 1944, a Japanese artillery lieutenant stood inside a cave on the island of Pleu and looked out at the Pacific Ocean.

What he saw made him reach for his diary.

He wrote that the sight of the American fleet, hundreds of ships stretched to the horizon in every direction, made him so furious he could feel the blood pounding through his veins.

He wrote it the way a man writes when he cannot find any other words.

Then he closed the diary, picked up his weapon, and waited.

He had been waiting 5 months for this morning.

Not because he was afraid, because the colonel who commanded his island had spent those five months building something that had never existed before in the Pacific War.

500 caves, miles of tunnels, steel doors that could be slid shut faster than a naval shell could answer back.

A system designed from the ground up to do one specific thing, to make the United States Marines pay the highest possible price for every inch of coral on that island.

The Navy officer who directed the pre-invasion bombardment told his men afterward, “Everything over there is done.

You’ll walk right in.

” The lieutenant’s diary was found after the battle.

He did not survive to read what the Marines said about that prediction.

Kunio Nakagawa was not, by the standards of the Imperial Japanese Army, a remarkable man.

He was the third son of an elementary school principal from Kumamoto Prefecture on the island of Kyushu.

He had wanted to be a journalist.

He became a soldier instead because in Japan in the early 1900s, the army was one of the few paths available to a young man without money or family connections.

He graduated from the Imperial Japanese Army Academy in December of 1918, 30th in his class.

Not first, not last.

30th.

He spent the next two decades doing what capable, unremarkable officers do, showing up, doing the work, moving up the ranks one rung at a time.

He fought in China during the Second SinoJapanese War, where his superiors noticed he had what one later described as a supreme eye for the land, the ability to look at a piece of terrain and immediately understand how it could be used, how it could be held, and how it could kill the men who tried to take it.

In March of 1943, he was promoted to colonel and given command of the second infantry regiment of the 14th division.

In April of 1944, his regiment was ordered to a small island in the Palao chain called Pelu.

Before he left Japan, Nakagawa told his wife, according to accounts passed down through his unit, that he would not be coming back.

He was not being dramatic.

He was being precise.

What had happened at Terawa and Saipan, the systematic destruction of two Japanese garrisons by United States Marines had already filtered back through the army’s command structure by the time Nakagawa reached Pleu.

The Imperial Japanese Army had assembled a research team to study those defeats.

Nakagawa had read their findings.

What those reports described was something the Japanese military had not believed possible when the war began.

At Tarawa in November of 1943, the Japanese garrison, nearly [clears throat] 5,000 strong, had built one of the most heavily fortified positions in the Pacific.

14 coastal defense guns, concrete bunkers, interconnected pill boxes that had been months in construction.

The Japanese commander had reportedly told his superiors that a million men could not take Tarawa in a 100red years.

The Marines took it in 76 hours.

A Japanese anti-tank officer who survived fighting near Guadal Canal and later wrote about his experience described in detail the moment his unit realized that the Americans were not stopping, not slowing, not pulling back to regroup the way military doctrine said they should.

The Americans absorbed casualties at a rate that doctrine said would halt any advancing force and did not halt.

Whatever the training manuals had said about the limits of American will under fire, the men writing from those battlefields had concluded the manuals were wrong.

At Saipan the following summer, 32,000 Japanese troops mounted the largest bonsai charge of the entire Pacific War.

4,000 men, some of them walking wounded from field hospitals, threw themselves at the American line on the night of July 6th.

The Marines and Army soldiers held.

They fired until the barrels were hot and kept firing.

When 4,000 men were dead, the American line was still there.

The translators, who later worked through the diaries collected from those battles, described the tone shifting across the war.

In 42, Japanese soldiers wrote about the Americans with contempt.

in 43 with puzzlement.

By 44, those who had faced the Marines were writing with something that had no clean name in Japanese military vocabulary.

Not fear, exactly, more like the recognition that the thing they were facing did not operate according to any set of rules they had been trained to understand.

Nakagawa absorbed all of it.

He did not write a memo analyzing his conclusions.

He did not send a report to headquarters outlining a new theory.

He walked up Umour Brogal Mountain, the coral ridge that ran down the center of Pelio like a broken spine, and began counting caves.

There were 500 of them, most barely developed, some little more than gaps in the limestone left over from decades of phosphate mining.

The Palao Island chain had been a source of phosphate for Japanese industry before the war.

And the men who had worked those mines knew every shaft and chamber in the rock.

Nakagawa found those men, and he put them back to work.

The first order Nakagawa issued when he took command of Paleleu’s defenses was one that confused almost every officer under his command.

No one was to fire from the beach.

In every previous island battle of the Pacific War, the Japanese had met the Americans at the waterline.

The theory was sound.

Hit the landing force when it was most vulnerable, crammed into boats and wing through surf before it could organize and bring its firepower to bear.

It was the doctrine the Japanese army had built its island defense strategy around.

It had not worked at Guadal Canal.

It had not worked at Terawa.

It had not worked at Saipan.

Nakagawa understood why.

The American naval bombardment that preceded every landing was simply too accurate and too overwhelming for any position on the beach to survive it.

Whatever you put at the waterline, the guns would find it and destroy it.

The beach was a killing ground for defenders, not attackers, because the attackers controlled the guns offshore.

So, the beach would be left with mines and obstacles, enough to slow the first waves, enough to channel the landing craft into predictable paths.

But the men, the guns, the real defense, all of it would go underground.

The 214th Naval Construction Battalion had arrived on the island months earlier.

Most of its men had been miners and tunnel laborers in civilian life.

Men who understood how limestone fractured, where the natural weaknesses in the rock ran, how deep you could go before the cave floor became unstable.

They did not need to be told how to dig.

What Naka Gawa gave them was a reason and a design.

The existing phosphate mine shafts became the backbone of the system.

From those shafts, new tunnels branched outward and connected to the natural limestone caves that honeycombed the Umar Brogall.

Some chambers were enlarged to accommodate field hospitals with electric lighting powered by generators.

Others were fitted with water systems and rice storage.

The men who would eventually fight from those positions could move between firing points, resupply, treat their wounded, and rest, all without ever exposing themselves to American air or naval fire.

The cave entrances were modified with sliding steel doors.

The principle was simple and lethal.

A gun crew would open the door, fire, and close the door again before the naval gunners offshore could locate the muzzle flash and answer back.

The Americans would see the shell, mark the rough direction, and fire.

By the time their shell arrived, the door was shut and the crew was back inside the rock.

The loopholes, the firing slits cut into bunkers and cave walls, were constructed at angles.

A direct shot from outside could not penetrate them.

A marine with a flamethrower would have to position himself at precisely the right angle to use it effectively, and every foot of that approach was covered by fire from another position.

Many of the cave systems were built on two levels.

The lower level served as shelter during bombardment.

The upper level was the fighting position.

When American guns opened up, the defenders moved down and waited.

When the guns lifted, which they always did just before the infantry advanced, the men climbed back up.

By the time the Marines were moving across open ground, the guns were already back in position behind the rock.

There was one more factor Nakagawa had calculated that the Americans had not accounted for.

During the battle, surface temperature on the open coral would exceed 115 degrees Fahrenheit.

The Marines would be fighting in full gear, carrying weapons and ammunition across open ground with no shade.

The Japanese, underground in the cave system, would be fighting 15 to 20° cooler.

They could rest, eat, and treat their wounded without the heat eating through them.

Every minute a marine spent above ground was a minute the island itself was doing part of Nakagawa’s work.

Nakagawa’s written order to his troops was distributed to every unit on the island before the first American ship appeared on the horizon.

Each man was expected to kill 10 Marines before he died.

It was not a slogan.

It was arithmetic.

If each of the island’s 11,000 defenders fulfilled the order, the mathematics came out at something over a 100,000 American casualties, more than the entire landing force.

The American estimate for the operation based on aerial reconnaissance that could photograph surface features but could not see underground was 4 days.

Nakagawa made no changes to his preparations.

Major General William Rupertus had been fighting the Japanese since August of 1942.

He had been on Guadal Canal.

He had commanded Marines at Cape Gloucester.

He had watched the Pacific War evolve from the chaotic early days when neither side fully understood what kind of fight they were in through Tarawa and the grinding island campaigns that followed.

He was not an inexperienced man.

He was not a reckless man.

On the eve of the Pleio landing, he called his officers together and told them what to expect.

“We’re going to have some casualties,” he said.

“But let me assure you, this is going to be a fast one.

Rough, but fast.

We’ll be through in 3 days.

It may only take two.

” Six war correspondents came ashore with the Marines on Pelleu.

Only six.

The prediction of a short decisive battle had convinced most of the press corps to skip it.

There were larger stories elsewhere.

MacArthur was returning to the Philippines.

The allies were driving through France.

Pleio was going to be a footnote.

The intelligence failure that produced Rupertis’ prediction was not a matter of incompetence.

It was a systematic problem that had plagued American amphibious planning throughout the war.

Aerial photography, the primary tool for assessing an island’s defenses before a landing, can only see the surface.

The analysts in Pearl Harbor who studied the Paleo photographs were looking at the right island.

They were looking at the wrong part of it.

The Navy officer who commanded the pre-invasion bombardment, 3 days of sustained shelling that was among the most intensive in the Pacific War, reported that the island had been effectively neutralized.

What the bombardment had actually accomplished was this.

It had changed the island’s scenery.

The same Japanese artillery lieutenant who had written about the blood pounding in his veins at the sight of the American fleet added a second entry that same morning.

Despite 3 days of shelling, exactly one man in his company had been wounded.

He and his men had spent those three days in the lower level of their cave system, waiting.

On the morning of September 15th, the first wave of Marines climbed into their landing craft and headed for the beach.

At 8:32, they touched shore.

5 minutes of silence followed.

The island was completely still.

The men looked at one another.

Then the steel doors opened.

By nightfall on the first day, 1100 Marines were dead or wounded.

60 tracked landing vehicles had been destroyed on the beach or in the water.

The heat, 115° on the open coral with no shade and no wind, had emptied every canteen within 2 hours of landing.

The water brought up to resupply.

The men had been stored in drums that had previously held oil.

It tasted like petroleum.

Men who drank it became sick.

Men who did not drink it passed out from heat exhaustion.

The Navy had said the island was done.

Rupertus had said 3 days, maybe two.

Inside the caves, the Japanese lieutenant opened his diary again.

He had been wrong about one thing that morning, he wrote.

He had expected the Americans to falter at the beach.

Every piece of training he had received, every report he had read about American soldiers in other theaters had suggested that an enemy hit that hard at the waterline would pause, would pull back, would need time to reorganize before pushing forward again.

The Marines did not pull back.

They walked over their own dead to keep moving.

The ones in the back did not stop when the ones in the front went down.

They moved to the side, found another angle, and kept going.

He had never seen anything like it, and he had been a soldier for 6 years.

Day one, the point.

On the northern end of the landing beach, a coral promontory called the point commanded the entire beach head with infiladating fire.

It was one of the strongest individual positions on the island.

A natural fortress reinforced with concrete and steel with firing lanes covering every approach.

Captain George Hunt led Company K of the First Marines against it on the first morning.

By the time they took the point, Hunt had less than a platoon’s worth of men still able to fight.

He held the position through the night against repeated Japanese counterattacks with no adjacent units in contact and no artillery support.

One tracked vehicle got through before dark to bring ammunition and evacuate the wounded.

In the morning, the position was still in Marine hands.

That was day one.

72 more days followed the first week.

On the sixth day, the first marine regiment, one of the most decorated units in the history of the core, was withdrawn from the line.

Colonel Lewis Puller had pushed his regiment into the Uma Brogal for 6 days.

They had taken ridgeel line after ridgeel line, cleared cave after cave, and lost 56% of their strength doing it.

What was left could no longer function as a fighting unit.

From inside the tunnel network, Nakagawa’s observers reported the same thing in every direction.

The Marines came back.

You drove them off a position and they came back the next morning.

You held a cave for 2 weeks against everything they threw at it.

Grenades, flamethrowers, demolition charges, bulldozers, pushing coral rubble into the entrances.

And when the cave finally fell, the men who took it moved immediately to the next one.

There was no celebrating.

There was no rest.

There was just the next position.

An anonymous diary recovered from a Pleio cave and later studied by historians described the Americans at this stage of the fighting in language that had no equivalent in the writer’s earlier entries.

He reached for a word that Japanese soldiers typically use to describe themselves in their own moments of deepest resolution.

Kakugo.

It translates roughly as settled determination, the acceptance of what is coming and the decision to continue anyway.

It was a word from the vocabulary of men who understand that there is no way out and who go forward regardless.

He used it to describe the Marines, not as a compliment, as an observation.

He was trying to name something he was watching happen outside his cave every day, and it was the only word he had that came close.

The second week, the canteen water still tasted like oil.

The coral cut through bootles within days.

Marines were fighting in makeshift wrappings of torn uniform cloth because their footwear had given out.

The temperature did not drop.

The Japanese snipers who emerged at night from tunnel exits behind American lines kept men from sleeping.

The sound of flamethrowers working on the ridge face was constant.

A Japanese soldier wrote that he could hear the Americans digging at night.

Not attacking, not shouting or celebrating, digging, extending their foxholes, running wire, positioning their guns for the morning.

He wrote that it reminded him of the way Japanese engineers prepared a siege, methodical, unhurried, as if there was no question about the outcome, and the only remaining work was the work itself.

He had been told before the war that Americans could not be trusted to fight like that, that they would grow bored, that they would lose focus, that the prospect of death sustained over weeks would erode their will to push forward.

He wrote that he no longer believed that the final weeks.

By late October, the First Marine Division had been withdrawn entirely and replaced by the Army’s 81st Infantry Division.

Nakagawa understood what this meant.

It was not that the Marines had broken.

It was that the Americans had an inexhaustible supply of men willing to do this kind of fighting.

When one unit was spent, another arrived.

The math he had given his men, 10 Marines for every Japanese soldier, was being fulfilled.

The casualty numbers were real.

And it was not enough.

The water inside the caves was running out.

The food was running out.

The ammunition was running out.

Tunnel sections were being sealed one by one.

A charge placed at the entrance.

A bulldozer pushing rubble until the opening was gone.

the men inside it gone with it.

One of the last diary entries recovered from the Umar Brogall was written by a man who gave no name and no unit.

He wrote about his family in Osaka.

He wrote about the smell of rain.

He wrote that he had stopped being afraid, which he said was either a sign of courage or a sign that something in him had simply stopped working and that he was no longer sure which one it was.

He did not write about surrender.

None of the diaries did.

Not one.

The last organized Japanese resistance on Pleu officially ended on November 27th.

But not everyone accepted that.

A band of 34 soldiers, 26 from Nakagawa’s own second infantry regiment and eight naval personnel, had survived the battle in the caves and continued fighting under Lieutenant A.

Yamaguchi.

They did not believe Japan had surrendered.

Leaflets dropped from aircraft, announcements over loudspeakers.

They dismissed all of it as enemy tricks.

In March of 1947, more than 2 years after the battle ended, Yamaguchi’s group attacked a Marine patrol on the island with hand grenades.

Only 150 Marines were stationed on Pelleu at that point with 35 dependents.

Reinforcements were called in.

American authorities flew in a Japanese admiral specifically to convince the holdouts that the war was over.

On April 21st, 1947, Yamaguchi’s group emerged from the jungle in two groups.

Yamaguchi saluted, bowed, and handed over his sword and his units battle flags.

They were the last men still keeping faith with the orders Nakagawa had given them.

The last formal surrender of the Second World War.

On the evening of November 24th, 1944, Colonel Kuno Nakagawa sent his last message.

It did not go by radio.

The island’s radio equipment had been destroyed weeks earlier.

It went by underwater telephone cable, a line that ran from Pleu to Japanese headquarters on the island of Kor, 25 mi to the southwest.

He reported to his superiors, “Our sword is broken and we have run out of spears.

” Then the operator at the Pleio end of the line began sending the call sign for the Coror switchboard.

Sakura, Sakura, Sakura, sack.

The line went dead.

At Corore, General Inu heard the voice stop and assumed the Americans had destroyed the command post with artillery.

He did not know what had actually happened.

What had happened was this.

Nakagawa burned his regimental colors.

He destroyed his command documents.

He performed the ritual suicide of a Japanese officer whose position has been lost.

His senior adviser, Major General Murai, died beside him.

Tokyo received the report and promoted Nakagawa two full grades from colonel to lieutenant general postumously.

It was one of the highest honors the Imperial Japanese Army could bestow on a man who was already dead.

When American soldiers searched the final cave complex in the days that followed, they found the ash of the burned flag.

They found evidence of what had happened.

They found no body.

Nakagawa’s remains have never been located.

He lies somewhere in the coral of Paleo.

After 80 years, no one knows exactly where.

The battle was declared secure on November 27th, 73 days after it was supposed to last three.

Of the approximately 10,900 Japanese defenders on the island, fewer than 300 survived as prisoners.

Most of those were Korean laborers, not combat troops.

The Marines and soldiers who took the island lost 9,800 killed and wounded, the highest casualty rate measured against the size of the force engaged of any American amphibious operation in the entire Pacific War.

The National Museum of the Marine Corps in Quantico, Virginia calls Pelu the bitterest battle of the war for the Marines.

Not the biggest, not the most famous, the bitterest.

The Japanese lieutenant who wrote about the blood pounding in his veins on the morning of September 15th did not survive the battle.

His diary was carried off the island and translated at Marine Corps headquarters.

The men he described in those last pages, the ones who kept walking, the ones who came back to the same ridge line day after day, most of them went home to the United States and did not talk about Pelleu much.

It was not the famous battle.

It was not Euoima.

It did not have the photograph.

What it had was 73 days of men doing the hardest kind of work against an enemy who had spent 5 months studying exactly what they were capable of and building the most complete answer anyone in that war ever built.

In April of 2015, Emperor Akihito and Empress Micho of Japan traveled to Paleo, the first time a Japanese emperor had set foot on the island.

They laid white chrysanthemums at the Japanese memorial.

Then they walked to the American memorial and did the same.

They stood on the same coral ground that Nakagawa had walked in 1944.

The ground above the tunnels that still run beneath the island, still sealed, still there.

His body is somewhere in that coral in the dark below bloody nose ridge where no one has found it in 80 years.

The Marines he was waiting for.

Their names are on the wall at Quantico.

If your father or grandfather served at Pleu, the first Marines, the fifth, the 7th, or the Army’s 81st infantry, leave their name and unit in the comments.

Those names belong to this story.

They deserve to be written down.