Why Japanese Officers on Iwo Jima Couldn’t Explain How Marines Kept Finding Sealed Tunnel Entrances
100 to 300 men still underground. That was the estimate when Iwo Jima was declared secure and the army regiment sent to garrison it took 867 prisoners in the 100 days after. 216 were taken in the 36 days of the battle. The regiment was an orphan cut from its division in 1943. Half its veterans rotated home before it sailed.
15 of its men died finishing a battle the record says had already ended. On the 26th of March 1945, the Marines left an island they had been told was clear. Four days later a patrol opened a cave and found a vegetable garden, chickens, two cows, and American medical supplies. And on the 11th of April, 100 feet down, a Japanese major put surrender to his men as a vote and the count came back 69 to 3.
Major General Harry Schmidt commanded Fifth Amphibious Corps. His staff put the number of Japanese still alive under the island at 100 to 300. The figure carried now is around 3,000. Nobody was lying. There was no way to count men 60 feet down who had no intention of coming up. 21,000 had defended Iwo Jima.
216 had surrendered. The rest were assumed dead. >> >> So the word went into the file. Secure. 9 in the morning the 26th of March. Marines who had been on that rock for 36 days walked down to the beaches and went back aboard ship. And beneath their boots men were still holding ground and waiting for them to leave.
At dawn on the 21st of March, five days before any of that, an army regiment came ashore at Purple Beach on the eastern side of the island. Three battalions and roughly 3,000 men. On the movement orders it was task unit 11.1.2. On the roll it was the 147th Infantry Regiment and it was an orphan. In 1943 the army cut its infantry divisions from four regiments to three.
>> [snorts] >> The 147th was the piece left over when the 37th division, Ohio National Guard, was rebuilt. And it was detached permanently on the 31st of July that year. After that, it garrisoned Samoa, New Caledonia, and Emirau. Rotation home had taken roughly half its combat-experienced men. It reached Iwo Jima talk on the 14th of March.
A signal arrived the same day requesting the regiment proceed to Iwo Jima at the earliest practicable date. They were back at sea within hours. You could smell the island before you could see what was wrong with it. On the 30th of March, a patrol from the first battalion killed two sentries at a cave mouth and went inside.
They found a vegetable garden, chickens, two cows, American medical supplies, American weapons, American ammunition. All of it lifted off the surface at night while everyone above believed the fighting had stopped. That is not a hideout. That is a household. Colonel Robert F. Johnson stopped treating the job as garrison duty.

He put his regiment onto combat patrols, pre-sited ambushes, and what the men came to call cave cleanings. They started calling themselves the cavemen. Kuribayashi had built for this. A quarter of his garrison was put on digging. 11 miles of tunnel finished before the Marines landed. More than 1,500 chambers.
Entrances turned at 90°, so nothing could travel straight in. Every position given more than one way out. The engineering was aimed at two things, eyes and fire. Against both, it worked. What it could not beat was AIR. Iwo Jima is a volcano that has not finished. Near the mountain, the laborers cutting those galleries could work about 15 minutes at a stretch before the heat and the sulfur drove them back out.
Korean laborers, mostly, and not one word of theirs survives in their own voice. Living quarters had to be vented or the men inside them would not last the night. Every sealed position on that island was breathing. And on the surface, a technical sergeant from California was already walking up to entrances carrying a flashlight and nothing else.
If this is the kind of thing you want more of, a like and a subscribe help. Stay with it, though. The reason the estimate was wrong by a factor of 10 is worse than a scandal because nobody covered anything up. The fighting simply happened after the file had been closed. The regiment’s first return went in on the 31st of March.
387 Japanese killed, 17 taken prisoner, eight Americans dead, 53 wounded. 10 days ashore, 404 men accounted for. Schmidt’s ceiling had been 300 and the 147th had gone past it inside a fortnight, and it had barely started. James Maguire of B Company said the cave teams were equipped with guts, flashlights, and pistols.
That is the whole inventory. There was no doctrine for this. There was a man, a torch, and a hole in volcanic rock with something inside it that had already decided how it wanted to die. James Ahearn, a platoon leader in F Company, described how it looked from the other end. The openings were covered by marksmen who could take a man at distance.
If you got close enough to throw a grenade, there was often a second man inside whose whole job was to catch it and throw it back. Ahearn’s word for him was outfielder. Maguire was wounded twice in a single action by exactly that. He threw a phosphorus grenade. It came back. He threw a fragmentation grenade. It came back.
Every entrance on that island was a decision about who went in first. It is worth being fair to the men who got the number wrong. Nobody had ever devised a way to count a garrison that lived underground, and the failure was on the record before the landing. The intelligence map compiled for the assault carries a line stating flatly that the thousands of caves on the island had not been plotted.
The 147th was not a bad regiment, either. It was a garrison unit by assignment. For 2 years, the army had handed it the islands after the shooting stopped, and that is what it was sent to do again. And the men below were operating inside a code that made surrender a disgrace rather than a choice. Ensign Satoru Omegari, who lived through it, said afterwards that the experience changed him at the root, that he came out of it with no faith left in people, and that there had been nothing glorious about any of it.
He had watched men fight each other over water. So, how were Americans finding sealed entrances at all? Not by looking. The photo interpreters had assembled on Guam in late January and built a joint enemy installation map out of every image taken since the 5th of that month, and the caves are not on it.
Senior intelligence officers did not work out that Kuribayashi was personally on the island until the 27th of February, the 9th day of the battle, or that his garrison was half again the 13,000 they had estimated. War dogs sit in the popular memory of Iwo Jima, and Marine dog platoons did land, but the ground was open and swept by artillery and small arms, and the conditions never warranted using them the way they had been used in the jungle.
What worked was food. The Americans began putting rations and water out at night, sometimes rigged, sometimes simply set down on ground already sighted for for Foraging parties came up in fours and fives because men who have been underground for 6 weeks will come up for water. Follow them back and you have the entrance. And the prisoners already taken went down again with the Americans who had taken them.
The entrances were not being found. >> >> They were being handed over. Technical Sergeant Terry Takeshi Doi came ashore with the assault waves on the 19th of February, the first morning alongside the Marines. He was an American soldier from California. He was also to a great many people back home a question.
Doi belonged to the Military Intelligence Service. Around 6,000 Japanese Americans went through its language school during the war and the work was classified while it was happening and stayed classified long afterwards. 50 of them worked at the Joint Intelligence Center in Honolulu in a converted furniture store downtown because the Navy would not allow Nisei onto the naval base at Pearl Harbor.
The men reading the enemies language were kept off the ground the enemy had bombed. Doi went into caves with a flashlight. On one occasion he walked up to a pillbox with his shirt off so that whatever was inside could see he was carrying nothing and talked. He was given the Silver Star. He was one of several.
Technical Sergeant James Yoshinobu landed with the Fourth Marine Division at 47 years of age fighting his Second World War having been in the first one in American uniform. He got a Silver Star as well. Sergeant Mike Masato Doi Guchi was hit by a mine and died of his wounds shortly after the war ended. Picture a man standing at an opening in volcanic rock in the dark, torch in one hand and nothing in the other, speaking Japanese into a hole he cannot see into while 3,000 miles behind him his family’s loyalty is a matter of public argument.
None of that was written down at the time under their names. It surfaced later in the army’s own official history of the linguists compiled by James McNaughton decades after the men had gone home. On the 11th of April, Captain James Kolb of A Company was setting an ambush on the east side of the island when two Japanese soldiers came up out of a hole near the spot he had chosen.
His men fired. One was killed. The other was hit and got back underground. Below him was a hospital of the second mixed brigade 100 ft down with 72 men in it. Kolb sent for his interpreter, Sergeant Ritsuo Tanaka. Tanaka spoke to the commander, Major Masaru Inouye, a senior medical officer, and put the position to him without decoration.
Come up or be burned and buried where you stand. Inouye did something almost no Japanese officer did in that war. He put it to his men as a vote. 69 to 3. The 69 were helped up to the surface. Corporal Kiutarou Kojima, one of the three, killed himself where he was. It was the largest group of prisoners taken at any single moment in the battle for Iwo Jima, and it was taken 16 days after the battle was declared over by an Army captain and a sergeant whose service file was secret.
In October 2004, the Journal of Military History published a piece by Robert Burrell, then teaching at the Naval Academy, called Breaking the Cycle of Iwo Jima Mythology. >> >> Burrell had spent years in the Army, Navy, Marine Corps, Air Force, and National Archives, and what he came back with was that the arithmetic everybody repeated was soft.
The Japanese dead, commonly given as 20,000 or 25,000, he put between 18,060 and 18,600. The strategic reasons for taking the island at all, he argued, had largely been assembled afterwards once the original justification failed to deliver. He was answered hard. Brian Hanley replied in the same journal the following year arguing that Burrell was judging 1945 with 2004 in his hand.
The argument is not settled. But hold it against the men underground. If the totals were never firm, then the 2,469 Japanese, the 147th killed or captured after the 26th of March, sat outside the battles accounting all together. They belonged to a fight that had officially finished.
The battle ended on the 26th of March. The fighting ended on the 30th of June. Lieutenant General Robert C. Richardson wrote the regiment a commendation. It praised their courage and their ingenuity, noted how few casualties they had sustained, and credited them with making a forward base finally safe.
A forward base, 101 days of tunnel war closed out in a garrison sentence. The 147th went to Okinawa on occupation duty, came home, and was deactivated on Christmas Day 1945. Doi’s file stayed shut along with every other linguist’s. >> >> Somewhere in those months a Japanese naval air group radio man named Saruji Akikusa got separated from his foraging party in the dark and walked into an American supply cache.
He was looking for water. He found a case of Coca-Cola which he had never seen before and drank until he nearly made himself sick. He said afterwards it was the most delicious thing he had ever tasted in his life. Two Japanese machine gunners, Yamakage Kufuku and Matsudo Rinsoki, came up out of the ground and surrendered on the 6th of January 1949 to American airmen of the 6,000th Support Wing, 3 years and 9 months after the island was declared clear.
The Military Intelligence Service was given a Presidential Unit Citation in 2000, 55 years late. In 2010, Congress authorized a Congressional Gold Medal for the Nisei soldiers of the Second World War, and on the 2nd of November, 2011, it was presented in Emancipation Hall at the Capitol.
Veterans arrived in wheelchairs. Some brought oxygen. The statute is precise about the form of it. A single gold medal awarded collectively to be handed to the Smithsonian and displayed there. So, it sits in a case. One medal for 6,000 men and not a name on it. Like the video and subscribe if you want more of these.
And if your family has a connection to the 147th Infantry Regiment on Iwo Jima or to one of the Nisei linguists of the Military Intelligence Service who talked men up out of those tunnels, I’d like to read about it in the comments.
Why Japanese Officers on Iwo Jima Couldn’t Explain How Marines Kept Finding Sealed Tunnel Entrances
100 to 300 men still underground. That was the estimate when Iwo Jima was declared secure and the army regiment sent to garrison it took 867 prisoners in the 100 days after. 216 were taken in the 36 days of the battle. The regiment was an orphan cut from its division in 1943. Half its veterans rotated home before it sailed.
15 of its men died finishing a battle the record says had already ended. On the 26th of March 1945, the Marines left an island they had been told was clear. Four days later a patrol opened a cave and found a vegetable garden, chickens, two cows, and American medical supplies. And on the 11th of April, 100 feet down, a Japanese major put surrender to his men as a vote and the count came back 69 to 3.
Major General Harry Schmidt commanded Fifth Amphibious Corps. His staff put the number of Japanese still alive under the island at 100 to 300. The figure carried now is around 3,000. Nobody was lying. There was no way to count men 60 feet down who had no intention of coming up. 21,000 had defended Iwo Jima.
216 had surrendered. The rest were assumed dead. >> >> So the word went into the file. Secure. 9 in the morning the 26th of March. Marines who had been on that rock for 36 days walked down to the beaches and went back aboard ship. And beneath their boots men were still holding ground and waiting for them to leave.
At dawn on the 21st of March, five days before any of that, an army regiment came ashore at Purple Beach on the eastern side of the island. Three battalions and roughly 3,000 men. On the movement orders it was task unit 11.1.2. On the roll it was the 147th Infantry Regiment and it was an orphan. In 1943 the army cut its infantry divisions from four regiments to three.
>> [snorts] >> The 147th was the piece left over when the 37th division, Ohio National Guard, was rebuilt. And it was detached permanently on the 31st of July that year. After that, it garrisoned Samoa, New Caledonia, and Emirau. Rotation home had taken roughly half its combat-experienced men. It reached Iwo Jima talk on the 14th of March.
A signal arrived the same day requesting the regiment proceed to Iwo Jima at the earliest practicable date. They were back at sea within hours. You could smell the island before you could see what was wrong with it. On the 30th of March, a patrol from the first battalion killed two sentries at a cave mouth and went inside.
They found a vegetable garden, chickens, two cows, American medical supplies, American weapons, American ammunition. All of it lifted off the surface at night while everyone above believed the fighting had stopped. That is not a hideout. That is a household. Colonel Robert F. Johnson stopped treating the job as garrison duty.
He put his regiment onto combat patrols, pre-sited ambushes, and what the men came to call cave cleanings. They started calling themselves the cavemen. Kuribayashi had built for this. A quarter of his garrison was put on digging. 11 miles of tunnel finished before the Marines landed. More than 1,500 chambers.
Entrances turned at 90°, so nothing could travel straight in. Every position given more than one way out. The engineering was aimed at two things, eyes and fire. Against both, it worked. What it could not beat was AIR. Iwo Jima is a volcano that has not finished. Near the mountain, the laborers cutting those galleries could work about 15 minutes at a stretch before the heat and the sulfur drove them back out.
Korean laborers, mostly, and not one word of theirs survives in their own voice. Living quarters had to be vented or the men inside them would not last the night. Every sealed position on that island was breathing. And on the surface, a technical sergeant from California was already walking up to entrances carrying a flashlight and nothing else.
If this is the kind of thing you want more of, a like and a subscribe help. Stay with it, though. The reason the estimate was wrong by a factor of 10 is worse than a scandal because nobody covered anything up. The fighting simply happened after the file had been closed. The regiment’s first return went in on the 31st of March.
387 Japanese killed, 17 taken prisoner, eight Americans dead, 53 wounded. 10 days ashore, 404 men accounted for. Schmidt’s ceiling had been 300 and the 147th had gone past it inside a fortnight, and it had barely started. James Maguire of B Company said the cave teams were equipped with guts, flashlights, and pistols.
That is the whole inventory. There was no doctrine for this. There was a man, a torch, and a hole in volcanic rock with something inside it that had already decided how it wanted to die. James Ahearn, a platoon leader in F Company, described how it looked from the other end. The openings were covered by marksmen who could take a man at distance.
If you got close enough to throw a grenade, there was often a second man inside whose whole job was to catch it and throw it back. Ahearn’s word for him was outfielder. Maguire was wounded twice in a single action by exactly that. He threw a phosphorus grenade. It came back. He threw a fragmentation grenade. It came back.
Every entrance on that island was a decision about who went in first. It is worth being fair to the men who got the number wrong. Nobody had ever devised a way to count a garrison that lived underground, and the failure was on the record before the landing. The intelligence map compiled for the assault carries a line stating flatly that the thousands of caves on the island had not been plotted.
The 147th was not a bad regiment, either. It was a garrison unit by assignment. For 2 years, the army had handed it the islands after the shooting stopped, and that is what it was sent to do again. And the men below were operating inside a code that made surrender a disgrace rather than a choice. Ensign Satoru Omegari, who lived through it, said afterwards that the experience changed him at the root, that he came out of it with no faith left in people, and that there had been nothing glorious about any of it.
He had watched men fight each other over water. So, how were Americans finding sealed entrances at all? Not by looking. The photo interpreters had assembled on Guam in late January and built a joint enemy installation map out of every image taken since the 5th of that month, and the caves are not on it.
Senior intelligence officers did not work out that Kuribayashi was personally on the island until the 27th of February, the 9th day of the battle, or that his garrison was half again the 13,000 they had estimated. War dogs sit in the popular memory of Iwo Jima, and Marine dog platoons did land, but the ground was open and swept by artillery and small arms, and the conditions never warranted using them the way they had been used in the jungle.
What worked was food. The Americans began putting rations and water out at night, sometimes rigged, sometimes simply set down on ground already sighted for for Foraging parties came up in fours and fives because men who have been underground for 6 weeks will come up for water. Follow them back and you have the entrance. And the prisoners already taken went down again with the Americans who had taken them.
The entrances were not being found. >> >> They were being handed over. Technical Sergeant Terry Takeshi Doi came ashore with the assault waves on the 19th of February, the first morning alongside the Marines. He was an American soldier from California. He was also to a great many people back home a question.
Doi belonged to the Military Intelligence Service. Around 6,000 Japanese Americans went through its language school during the war and the work was classified while it was happening and stayed classified long afterwards. 50 of them worked at the Joint Intelligence Center in Honolulu in a converted furniture store downtown because the Navy would not allow Nisei onto the naval base at Pearl Harbor.
The men reading the enemies language were kept off the ground the enemy had bombed. Doi went into caves with a flashlight. On one occasion he walked up to a pillbox with his shirt off so that whatever was inside could see he was carrying nothing and talked. He was given the Silver Star. He was one of several.
Technical Sergeant James Yoshinobu landed with the Fourth Marine Division at 47 years of age fighting his Second World War having been in the first one in American uniform. He got a Silver Star as well. Sergeant Mike Masato Doi Guchi was hit by a mine and died of his wounds shortly after the war ended. Picture a man standing at an opening in volcanic rock in the dark, torch in one hand and nothing in the other, speaking Japanese into a hole he cannot see into while 3,000 miles behind him his family’s loyalty is a matter of public argument.
None of that was written down at the time under their names. It surfaced later in the army’s own official history of the linguists compiled by James McNaughton decades after the men had gone home. On the 11th of April, Captain James Kolb of A Company was setting an ambush on the east side of the island when two Japanese soldiers came up out of a hole near the spot he had chosen.
His men fired. One was killed. The other was hit and got back underground. Below him was a hospital of the second mixed brigade 100 ft down with 72 men in it. Kolb sent for his interpreter, Sergeant Ritsuo Tanaka. Tanaka spoke to the commander, Major Masaru Inouye, a senior medical officer, and put the position to him without decoration.
Come up or be burned and buried where you stand. Inouye did something almost no Japanese officer did in that war. He put it to his men as a vote. 69 to 3. The 69 were helped up to the surface. Corporal Kiutarou Kojima, one of the three, killed himself where he was. It was the largest group of prisoners taken at any single moment in the battle for Iwo Jima, and it was taken 16 days after the battle was declared over by an Army captain and a sergeant whose service file was secret.
In October 2004, the Journal of Military History published a piece by Robert Burrell, then teaching at the Naval Academy, called Breaking the Cycle of Iwo Jima Mythology. >> >> Burrell had spent years in the Army, Navy, Marine Corps, Air Force, and National Archives, and what he came back with was that the arithmetic everybody repeated was soft.
The Japanese dead, commonly given as 20,000 or 25,000, he put between 18,060 and 18,600. The strategic reasons for taking the island at all, he argued, had largely been assembled afterwards once the original justification failed to deliver. He was answered hard. Brian Hanley replied in the same journal the following year arguing that Burrell was judging 1945 with 2004 in his hand.
The argument is not settled. But hold it against the men underground. If the totals were never firm, then the 2,469 Japanese, the 147th killed or captured after the 26th of March, sat outside the battles accounting all together. They belonged to a fight that had officially finished.
The battle ended on the 26th of March. The fighting ended on the 30th of June. Lieutenant General Robert C. Richardson wrote the regiment a commendation. It praised their courage and their ingenuity, noted how few casualties they had sustained, and credited them with making a forward base finally safe.
A forward base, 101 days of tunnel war closed out in a garrison sentence. The 147th went to Okinawa on occupation duty, came home, and was deactivated on Christmas Day 1945. Doi’s file stayed shut along with every other linguist’s. >> >> Somewhere in those months a Japanese naval air group radio man named Saruji Akikusa got separated from his foraging party in the dark and walked into an American supply cache.
He was looking for water. He found a case of Coca-Cola which he had never seen before and drank until he nearly made himself sick. He said afterwards it was the most delicious thing he had ever tasted in his life. Two Japanese machine gunners, Yamakage Kufuku and Matsudo Rinsoki, came up out of the ground and surrendered on the 6th of January 1949 to American airmen of the 6,000th Support Wing, 3 years and 9 months after the island was declared clear.
The Military Intelligence Service was given a Presidential Unit Citation in 2000, 55 years late. In 2010, Congress authorized a Congressional Gold Medal for the Nisei soldiers of the Second World War, and on the 2nd of November, 2011, it was presented in Emancipation Hall at the Capitol.
Veterans arrived in wheelchairs. Some brought oxygen. The statute is precise about the form of it. A single gold medal awarded collectively to be handed to the Smithsonian and displayed there. So, it sits in a case. One medal for 6,000 men and not a name on it. Like the video and subscribe if you want more of these.
And if your family has a connection to the 147th Infantry Regiment on Iwo Jima or to one of the Nisei linguists of the Military Intelligence Service who talked men up out of those tunnels, I’d like to read about it in the comments.