Major-General Yamada's war journal can be found in Source Material Relating to the Battle of Nanking, Vol. 2. In it, he states that on December 12, his detachment was at Gaozizhen, about 12 kilometers north of Nanking. Gaozizhen was a remote village that had been "totally devastated" by Chinese troops.¹ The detachment received an emergency order "to take part in the assault on Nanking" at 1:00 p.m. on December 12.²

The order stated that the Yamada Detachment was "to capture the forts at Wulongshan and Mufushan, and to facilitate the advance of the Sasaki Detachment."³ Upon receiving these orders, the 65th Regiment and the 3rd Battalion, 19th Mountain Artillery Regiment departed from Gaozizhen. This was a night march, since it was 5:00 p.m. and already dark.

On December 13, Yamada and his men passed through Xiaqijie, which had been so thoroughly burned by Chinese troops that no usable buildings remained.⁴ At 1:00 p.m., the 1st Battalion, 65th Regiment succeeded in capturing the fort at Wulongshan, on the banks of the Yangtze River.

The next objective was to seize the fort on Mufushan, a mountain that overlooks the Yangtze, three kilometers north of Nanking. The Mufushan fort was the last Chinese defense line.

There was surprisingly little resistance from Chinese troops. Some Japanese soldiers were killed or wounded, but the Chinese did not put up much of a fight. Soon they began to surrender, waving white flags. Yamada described the situation on the morning of December 14 in his journal.

Fearing that the fort might be captured by another division, we departed for Mufushan at 0400 hours. We reached the fort at dawn. A host of soldiers surrendered, so many that we were at a loss as to their disposition.⁵

Every town and village in the area had been "burned by the enemy" - another instance of the scorched-earth strategy.⁶ Brigade Commander Yamada found a school at the base of Mufushan, where he confined the prisoners after disarming them. It was a very large group: Yamada reports that there were 14,777 in all.⁷ According to 65th Regimental Commander Morozumi Gyosaku, whose journal also appears in Source Material Relating to the Battle of Nanking, Vol. 2, there were upwards of 15,300 prisoners. In any case, the prisoners outnumbered the 65th Regiment 10 to one. Yamada truly faced a dilemma, writing, "There are so many that it will be just as difficult to kill them as keep them alive."⁸

According to Soldiers' Accounts of the Great Nanking Massacre: War Journals of the Yamada Detachment, 13th Division, once the prisoners were confined, Morozumi noticed that they were "dressed in motley outfits,"⁹ and wondered if they really were soldiers. Further inspection revealed that some of the prisoners were noncombatants, residents of Nanking who had fled the city, and that some of the soldiers were actually women, old men, or boys.¹⁰ Innocent children only 12 or 13 years old had been sent to the front line, and women had volunteered to fight with the men.

               13th Division
                     |
103rd Infantry Brigade (Yamada Detachment)
                     |
                     +---------------------------+
                     |                           |
         65th Infantry Regiment  19th Mountain Artillert Regiment
                     |
      +--------------+--------------+
      |              |              |
1st Battalion  2nd Battalion  3rd Battalion

According to the war journal of Miyamoto Shogo the Chinese soldiers were starving. While they were being escorted to the barracks, "some of them were eating whatever vegetation was to be found." Another war journal reports that "many of the Chinese soldiers had had neither food nor water for a week."¹¹

Hirabayashi Sadaharu's testimony in Suzuki Akira's The Illusion of a Great Nanking Massacre describes a similar case.

We interned a great number of prisoners. On the second day, there was a fire. I don't remember whether any of the prisoners escaped at that time. If they had wanted to escape, they could have done so easily, because the fence that enclosed them was made of bamboo. Our main problem was feeding them. It was all we could do to feed ourselves, and we simply couldn't prepare proper meals for 10,000 people. Besides, the Chinese were leaderless. When we gave them water, they would fight over it. Some of them even ate the grass in the garden.¹²

I dispatched Cavalry 2nd Lieutenant Honma to Nanking to receive instructions regarding the disposition of the prisoners and other matters.
We are told to kill all of them.
All units are desperately short of food.¹⁴

"We are told to kill all of them" gave rise to the belief that it was official military policy to execute prisoners. We must remember, though, that war journal entries were necessarily terse, and covered only the essentials. For that reason, they must be read with care. We shall proceed to explicate this entry.

First, Yamada dispatched 2nd Lieutenant Honma to Nanking to receive instructions regarding the disposition of the prisoners. He did so because brigades (regiments) were required to inform division headquarters when they were faced with a large number of prisoners. For instance, the notice discussed in Chapter 4 states that, "When a great many prisoners of war are captured, they are not to be shot to death, but disarmed, assembled in one location, guarded, and their presence reported to division headquarters." That instruction emanated from the headquarters of the 13th Division, to which Major-General Yamada (Military Academy Class of 1905) was attached.

When Honma reached Nanking, Ogisu Ryuhei, commander of the 13th Division (Military Academy Class of 1904), instructed him to kill all the prisoners, as Yamada wrote in the second sentence of his entry.

There is a logical progression from the first to the second sentence. However, the third sentence, "All units are desperately short of food," does not seem to follow. There is no connection between killing prisoners and the food supply. In fact, the food shortage would have been a perfect excuse for allowing the prisoners to starve to death. Deprived of nourishment, they would be too weak to resist, making it easier for the Japanese to comply with an execution order.

However, without ammunition, it is impossible to perform executions. If Yamada had written: "We are told to kill all of them. All units are desperately short of ammunition," that would make better sense. But he wrote, "All units are desperately short of food." Did Brigade Commander Yamada actually obey the order to kill all the prisoners? Or is it true that, as Regimental Commander Morozumi wrote, he "stubbornly refused to carry out the order, and instead, instructed his men to intern the prisoners?"¹⁵ Did he defy the order, falling back on the military regulation that states, "Enemy soldiers who surrender are not to be taken prisoner, but removed from the battle zone after they have been disarmed?"¹⁶ Whatever the case, the three-sentence entry in his war journal does not tell us how he felt when he was ordered to kill all of the prisoners.

On the third day of internment, December 16, a fire broke out in one of the barracks. It could not have been the result of carelessness, for the prisoners were aware that they would have to sleep out in the cold if their barracks were destroyed. Regimental Commander Morozumi concluded that the prisoners had set the fire, hoping to escape during the ensuing confusion.¹⁷

The arson incident is mentioned by four of the 19 individuals who contributed accounts to the "War Journal of the Yamada Detachment, 13th Division." But the entries are inconsistent as to when the fire broke out. Three of the four soldiers wrote that it started at about noon on December 16. The fourth wrote, "In the evening, 20,000 prisoners started a fire." Morozumi wrote that it was dark when the fire started. All agree as to the date of the conflagration (December 16). The accounts also vary as to the extent of damage caused by the fire. Some of them stated that it destroyed half, and others, one-third of the barracks.

According to Regimental Commander Morozumi's account, half of the prisoners escaped during the commotion that resulted, while the fire was still raging. The Japanese fired their guns to prevent them from fleeing, but could not see their targets in the dark. Morozumi estimated that "if you moved away from the scene of the fire, you couldn't see anything, so at least 4,000 prisoners must have escaped."¹⁸ That leaves 4,000 prisoners remaining in the barracks.

If the Japanese had intended to execute the prisoners from the outset, they probably would have killed the remaining prisoners immediately, since the arson justified such action. However, they did not do so, which is proof that immediate execution was not official policy.

In his war journal, Private 1st Class Odera Takashi described a volatile situation: On December 17, when his unit had advanced about 18 kilometers to the Shangyuan Gate, "every once in a while, rifle bullets would fly very close to our heads."¹⁹ There was a real possibility that Chinese soldiers, once released, might engage in guerrilla warfare. Moreover, rather than surrendering to the Japanese, Chinese troops in Nanking were disguising themselves as civilians and waiting for a second chance to attack the Japanese. That behavior prompted the issuance of the division order instructing that all prisoners be killed.

The arson incident served to galvanize the position taken by division headquarters. Prisoners who resisted were invariably shot. That is why the brigade commander abandoned the idea of defying orders. As the situation worsened, he acquiesced to the method of last resort - execution.

In his journal, Regimental Commander Morozumi wrote: "We have no choice but to comply with orders. Major-General Yamada came to my unit, fighting back tears, to convince us.²⁰ The brigade commander had gone to Morozumi to announce that he had agreed to shoot the prisoners to death. Armed with this information, we can offer the following interpretation of the third sentence in Yamada's journal entry.

We are told to kill all of them. However, I cannot bring myself to do that, and would like to find some way to release them outside the battle zone. But since all units are desperately short of food, the regiment cannot incarcerate the prisoners until it is safe to release them. [Italics supplied.]

The text in italics had simply been omitted from the original entry. Yamada's secret plan for the disposition of the prisoners was contrary to the order from division headquarters. There was never any agenda that prescribed "the systematic execution of prisoners of war," as Fujiwara Akira claims. Ultimately, however, Brigade Commander Yamada's plan to release the prisoners was thwarted.

By "I," 2nd Lieutenant Endo meant the 1st Battalion. The order he refers to was probably a brigade order that mirrored a division order. A description of the fire and the subsequent executions appears in the "War Journal of Miyamoto Shogo" as well.

Subsequent to the afternoon meal, a fire broke out, resulting in a terrific commotion. The fire spread to 30 percent of the buildings. At 1500 hours, the battalion led approximately 3,000 prisoners to the banks of the Yangtze River and shot them - the method of last resort. A sight like that could never be seen, and never will be seen, anywhere but in a war zone.²²

Second Lieutenant Miyamoto wrote about a "terrific commotion" after the fire broke out, and adds that the prisoners were executed - the "last resort." This was military action taken against rebellious prisoners, which is condoned by international law. Miyamoto's words - "A sight like that could never be seen, and never will be seen, anywhere but in a war zone" - eloquently communicates this fact.

It is important to note here that even though Nanking fell on December 13, the Chinese had not surrendered formally. Instead, they mounted an obstinate resistance against the occupying Japanese forces. We have already covered this territory, but to cite another example, on December 17, the day after the arson incident, Nakano Masao was standing guard at the Wulongshan fort. He wrote in his journal about deaths on the battlefield - events that occurred all too often.

While on sentry duty, two members of my platoon were wounded by hand grenades thrown by enemy stragglers. Every day we shoot and kill any number of these stragglers.²³

On December 16, Private 1st Class Nakano, a member of the 1st Battalion, which was under Morozumi's command, had been ordered to guard the Wulongshan fort. That day, a straggler threw a hand grenade at one of Nakano's comrades on sentry duty, injuring him. Skirmishes like these were frequent occurrences. For the Japanese, the battle was not over.

On the 17th, assemble all remaining prisoners on the south bank of the Yangtze River, north of Mufushan. Wait until dark, and then transport them to the north bank and release them. Procure boats and oarsmen from nearby villages.²⁴

Why did Morozumi ignore the order to kill the prisoners? According to his diary, he pondered it for a while, and then decided that "the unit chosen to carry out this order must have some discretionary authority. I will decide what to do on my own, and hope that all goes well."²⁵ He resolved to release the prisoners on an island in the middle of the Yangtze River. His plan was implemented on the night of December 17, the day of the ceremonial entry into Nanking.

The series of events that occurred when the boats were halfway across the river, events that could not possibly have been foreseen, is described in Morozumi's diary.

Small boats carrying two or three hundred prisoners had reached midstream, when Chinese troops guarding the bank toward which the boats were heading opened fire. They must have thought that the Japanese had launched an attack from across the river. The terrified Chinese oarsmen veered left and right to avoid the bullets, but the boats were swept away. Meanwhile, when the prisoners who had been assembled on the north bank heard gunfire, they jumped to the conclusion that the Japanese had taken their comrades onto the river only to shoot them. The silence was broken, and utter chaos resulted. About 2,000 of the prisoners began running for their lives, every which way. There was nothing we could do to stop them but fire our guns. But we could not see them in the dark. Most of them escaped inland, but some jumped into the Yangtze. The next morning, I saw the bodies of prisoners whom we had shot, but there were only a few. That was the end of it. The outcome was terribly disappointing, but I have written the truth. The stories that have been circulated about what happened are full of hyperbole and propaganda.²⁶

A map of the area shows an island called Baguazhou on the Yangtze opposite Mufushan, about 100 meters from the right bank of the river. Though small, it is larger than the city of Nanking. Since it was easily reached, many Chinese soldiers had escaped there from Nanking. Lieutenant Hashimoto Mochiyuki had already sighted them on the island from the battleship Hozu on December 13, according to The Battle of Nanking.²⁷

But Regimental Commander Morozumi had no way of knowing this. His 1st Battalion made preparations to release the prisoners, following his instructions. The boats, packed with prisoners, set out for the island. As they approached Baguazhou, Chinese troops on the island spotted them and assumed, unsurprisingly, that the Japanese had launched a night attack. They fired at the boats. The prisoners waiting to be released on the banks of the Yangtze heard the shots. They thought that their comrades were being executed. Desperate, they began to run.²⁸

According to Morozumi's account, this incident took place at "about midnight."²⁹ The Japanese fired their machine guns, but since it was dark, and since no one had anticipated such a situation, there were Japanese casualties as well.

The testimony of Sublieutenant Hirabayashi Sadaharu, the "walking encyclopedia" of the 65th Regiment (the Morozumi Regiment), appears in Suzuki Akira's The Illusion of a Great Nanking Massacre. Hirabayashi heard that "the prisoners were being sent back to the Division at Zhenjiang by boat."³⁰

However, Hata Ikuhiko has expressed doubts about a plan involving the release of prisoners in the dark of night, for two reasons. He wonders why Morozumi's unit didn't move them during the day, if it really intended to release them. Furthermore, he does not believe that prisoners could have revolted with their hands were tied behind their backs.

In this writer's opinion, there were two reasons for Morozumi's decision to carry out his plan at night. First, he had received division orders to kill all the prisoners and, therefore, wanted to keep his plan as secret as possible. The island on which he intended to release the prisoners was very near Nanking. A Japanese Navy battleship was moored on the Yangtze, and an Army unit was stationed on its banks, both war-ready. If Morozumi had released the prisoners in daylight, the Japanese would have spotted them easily. The prisoners might have been killed. That is why he chose to release them at night, when there was less risk of detection.

Second, if the Japanese had intended to execute the prisoners at night, they would not have removed the ropes that bound their hands, to avoid the possibility of rebellion. If their hands had been tied, the prisoners could not have run properly, much less rebel. Surrounded, most of them would have been killed by machine-gun fire. There would have been virtually no chance of their escaping.

To release the prisoners safely (and secretly), Morozumi deliberately chose the dark of night. Since the prisoners' bonds were probably removed just before they were released, they could have rebelled or escaped. Or, perhaps their hands were never tied since, given their numbers, it is unlikely that there would have been time for that between the issuance of the departure order and the actual departure. Whichever the case, it is clear that Morozumi intended to release the prisoners.

At that time, the Japanese executed enemy soldiers who violated international law on the banks of the Yangtze, in broad daylight. If Morozumi had intended to kill these prisoners, he could have done so during the day, to avoid overlooking any of them. After all, 4,000 of them had escaped on the previous day.

At nighttime, the Japanese could not have seen well enough to aim their rifles accurately. Nor could they have distinguished the prisoners from their comrades, whom they might have shot accidentally. It does not take an expert to realize the folly of nighttime executions, and Morozumi would never have jeopardized the safety of his own men to that extent.

In his war journal, Araumi Kiyoe, a member of the 1st Battalion Headquarters staff, wrote on December 17, "Some of the men in our battalion were killed or wounded."³¹ Another war journal describes the writer's sorrow when his comrade's belly was pierced by a stray bullet.

If Morozumi had been planning, secretly, to execute the prisoners, he would have done so during the day. That way, he could have avoided losing some of his men, and could ascertain that every single prisoner had been killed. The clues that help solve the mystery are: Morozumi did not allow the prisoners to starve to death, his operation took place at night, and some Japanese were killed during its course.

Suzuki Akira, who discovered the "Diary of Yamada Senji," views what happened as an accident,³² and he is correct. Even if Morozumi actually did plan to execute the prisoners, why would he be so foolish as to do so at night? Thus far, no one has provided a plausible response to this question.

When he heard what had transpired, Yamada, rather than reproving Morozumi, was relieved. According to Morozumi's diary, the brigade commander "had read my mind."³³ Although Morozumi's plan was partly unsuccessful, its outcome was close to what the brigade commander had originally intended. If Yamada's policy had been to execute all prisoners, he would have rebuked Morozumi severely for disobeying his orders. But he did not.

Thus, Yamada and Morozumi both made a strenuous effort to remove Chinese soldiers who had surrendered from the battle zone. But their plans ended in failure, betraying their good intentions. Morozumi believed that his efforts had been in vain.

As he wrote, "That was the end of it. The outcome was terribly disappointing, but I have written the truth."³⁴ But even if he miscalculated, he did everything he could to wrest the prisoners from the jaws of death.

Moreover, the rank-and-file had no way of acquiring an accurate grasp of the events that had taken place. Even if they had, battle fatigue prevented them from doing much more than scribbling a few lines in their diaries. That is why the "War Journals of Soldiers in the Yamada Detachment, 13th Division" describe only the final outcome, i.e., executions.

For instance, documents like "The War Journal of Meguro Fukuji" record only the executions.³⁵ Moreover, Corporal Meguro was a member of the 19th Mountain Artillery Regiment, not of Morozumi's unit (65th Infantry Regiment). His (and only his) diary contains a fatal misconception. He writes that on December 13 (actually December 14), 100,000 Chinese soldiers surrendered, and that on December 17 and 18 (actually only on the 17th) 13,000 were executed. He was mistaken about all the important facts.³⁶ His diary lacks accuracy and reliability, but he cannot be blamed for that. Journal entries were written after the fact, when the soldiers had a spare moment. It is not surprising that their memories sometimes failed them.

The first one is the "War Journal of Endo Takaaki." In it he writes that on or about December 16, "It seems that orders were received from the Division regarding the disposition of prisoners."³⁷ Thus, 2nd Lieutenant Endo knew that the brigade commander, in issuing an order instructing that all prisoners were to be killed, had made a decision that was very painful for him (see (3) above).

But Endo could not have known about (4) and (5), since he did not accompany Tayama's 1st Battalion. Therefore, he recorded only the tragic outcome of the regimental commander's carefully laid plan to release the prisoners. The entry in 2nd Lieutenant Endo's journal for December 17 follows.

Supplied nine soldiers for guard duty at the summit of Mufushan at 0700 hours. Selected soldiers to line the route of the entry into Nanking from "R" to represent "13D." Depart with 10 soldiers from the platoon at 0800 hours, and enter Nanking from Heping Gate ... Returned to barracks at 1730 hours, exhausted, as it was three miles from the barracks to the site of the ceremony. At night, provide five men for execution of more than 10,000 prisoners of war ... .³⁸ [Italics supplied.]

By "from 'R' to represent '13D'," Endo meant that he selected soldiers from the regiment (Morozumi's Regiment) to represent the 13th Division, and instructed them to participate in the victory ceremony.

The first part of this entry describes participation in the ceremonial entry. The second part refers to the fact that he dispatched soldiers for the executions. Viewing only this entry, one would get the impression that Morozumi's original intent was to execute the prisoners. Endo's entry for December 18 begins as follows.

Since executions not completed by 0100 hours, ordered to dispose of remaining prisoners of war. Proceeded to execution site. Cold wind blowing; blizzard commenced at about 0300 hours. Am chilled to my bones. It seemed as though daylight would never come. Completed at 0830 hours ... From 1400 hours until 1930 hours, mobilize 25 men to dispose of more than 10,000 corpses.³⁹

It would appear that 2nd Lieutenant Endo participated in the executions after they were underway. Since shots were fired from Baguazhou and the prisoners on the banks of the Yangtze began running for their lives at "about midnight," Endo must have been ordered to appear about an hour later. Therefore, he was not privy to the information contained in (4) and (5) above.

Therefore, he describes only the final outcome (see (6) above) - the executions. Prisoners were shot to death, so executions did take place, but Endo had no idea about the preceding events. That is why he wrote simply "executions."

Today some of our men participated in the parade into Nanking, but most of them had been entrusted with the disposition of the prisoners of war. I marched to Nanking, departing at 0800 hours ... By the time I got back it was evening. I immediately set out to take part in the disposition of the prisoners of war. There were more than 20,000 of them. There was a terrible mistake, and many of our men were killed or wounded.⁴⁰ [Italics supplied.]

Miyamoto had taken part in the ceremonial entry into Nanking, as had 2nd Lieutenant Endo. But Miyamoto participated in the "disposition" of the prisoners of war immediately after he returned from the ceremony, while Endo did not.

However, doubts remain. In the entry for December 16, Miyamoto had written, "The battalion has decided upon the method of last resort ... to shoot the prisoners to death."⁴¹ But on December 17, he wrote neither "shooting" or "execution," but "disposition." Was this because the executions had ended in failure? If the intention had been to shoot the prisoners from the outset, then the executions could not have ended in failure, nor would any Japanese have been killed.

Thus, it was the 1st Battalion, Morozumi's battalion, that shot some of the prisoners on December 16, after a fire had been set, and that removed the prisoners at night. Morozumi had instructed 1st Battalion Commander Tayama to release the prisoners. Therefore, Tayama was well aware of (4) and (5) above.

Miyamoto was not only a member of the 1st Battalion, but also an officer. Therefore, the battalion commander must have told him that the prisoners were being moved so that they could be released, not executed. It would have made more sense if he had written, "I immediately set out to take part in the shooting of the prisoners of war."

Why didn't 2nd Lieutenant Miyamoto write the truth? If he had, his entry would have read as follows: "I immediately set out to take part in the release of the prisoners of war. There were more than 20,000 of them. There was a terrible mistake, and many of our men were killed or wounded." But by writing that, he would have been criticizing the operation and his superiors.

Neither Regimental Commander Morozumi nor 1st Battalion Commander Tayama was responsible for the debacle. No one could have foreseen the events that took place. And by writing "release," "blunder," and "many of our men were killed or wounded," Miyamoto would have been intimating that the regimental commander and the 1st Battalion commander were to blame, which was not true. Therefore, to convey the idea that the prisoners were removed so that they could be "released," and that ultimately, they were "shot," he wrote "disposition."

If the original intention had been to remove the prisoners to execute them, 2nd Lieutenant Miyamoto would have written "shot" or "executed" in his entry for December 17, as he had on the previous day. Instead, Miyamoto recorded a lawful act of war (execution), not a massacre. However, since no one counted the actual number of prisoners who were removed or shot, the accounts do not match.

Our company supplied a great number of sentries. I was ordered to act as patrol officer, and was exhausted from having to be constantly vigilant. In the evening we provided meals to some of the prisoners. We didn't have enough rations for ourselves, so we had a difficult time feeding the prisoners.⁴³

This excerpt is from Miyamoto's journal entry for December 15. It tells us that, though "all units are desperately short of food," all possible was done to feed the Chinese prisoners, at the expense of the Japanese soldiers "who had been suffering from exhaustion for two days."⁴⁴ We must remember that the Japanese did not treat the prisoners harshly, even though the latter were not protected by international law.

Go to Chapter 8
Back to Index