The Night of 7 July 1937: Shots at Wanping

On the evening of 7 July 1937 a Japanese infantry company went out on a night exercise on the open ground north of the walled town of Wanping, about a kilometer from the Marco Polo Bridge. At about 2240 shots were fired in the dark, and a private failed to answer the roll. He was back within about twenty minutes, but by dawn a Japanese battalion was deployed outside Wanping, the two sides had exchanged fire again, and at about 0530 on 8 July the Japanese attacked. This chapter follows that night hour by hour from both sides, and then asks the question that has never been settled: who fired first?

Why it mattered

The war between China and Japan that began at the Marco Polo Bridge lasted eight years and killed millions of people. It merged in December 1941 with the Pacific war that began at Pearl Harbor. Yet the incident that started it was small. The American ambassador in China, Nelson T. Johnson, estimated a week later that it had “probably involved not more than 200 Japanese soldiers” (Johnson to Secretary of State, 14 July 1937, FRUS 1937, vol. III). The first exchange of fire killed no one. The soldier whose disappearance led to the Japanese demand to search Wanping turned up unhurt within about twenty minutes.

That is why the night of 7–8 July has been studied so closely. The Mukden Incident of September 1931, told in Chapter 1, was planned by Japanese officers, who set off the charge on the railway themselves. The Tokyo tribunal after the war, and Chinese historians ever since, have seen the Marco Polo Bridge in the same light. Japanese veterans and many Japanese historians have argued the reverse: that Chinese soldiers fired first and the Japanese only answered. Western historians, working from both sides’ records, have mostly concluded that nobody can now say who fired the first shots, and that the more important question is why a skirmish that both sides’ local commanders tried to settle became a general war.

The answer to that larger question lies mostly in Tokyo, Nanjing (Nanking) and Tianjin (Tientsin) in the days after this chapter, told in Chapter 4. But it began here, in the dark, with tired men on both sides guessing at what the other side meant to do.

A tense summer on the Yongding

By the summer of 1937 the two armies around Beiping (Peking) lived uncomfortably close together. The Japanese China Garrison Army, headquartered at Tianjin and enlarged in 1936, kept the 1st Infantry Regiment of its infantry brigade in and around Beiping. Its 3rd Battalion, under Maj. Ichiki Kiyonao, was stationed at Fengtai, the railway junction southwest of the city where the Beiping–Hankou line met the lines to Tianjin. The regiment’s commander, Col. Mutaguchi Renya, was in Beiping. Japan claimed the right to station troops along the railway from Beiping to the sea under the Boxer Protocol of 1901, but the Fengtai garrison, established in 1936, had no clear basis even in that treaty. The American embassy later wrote that it knew of no legal basis for Japanese troops at Fengtai at all (Embassy despatch no. 1377, FRUS 1937, vol. III, doc. 484).

Around the Japanese posts lay the Chinese 29th Army of Song Zheyuan, the subject of Chapter 2. Song was chairman of the Hebei–Chahar Political Council, the semi-autonomous regime that Nanjing had accepted in December 1935 as a buffer against Japanese demands. In July 1937 he was away at his home at Laoling in Shandong, and the Chinese government later complained of his “long absence from his post” (FRUS 1937, vol. III, doc. 107). In his place, Qin Dechun, mayor of Beiping and deputy commander of the 29th Army, handled the army’s affairs in the city. The 37th Division of Feng Zhian held Beiping and the country to its west. Its 110th Brigade, under He Jifeng, held the Lugou Bridge sector, and the 219th Regiment of Col. Ji Xingwen had a battalion in Wanping under Jin Zhenzhong.

Wanping was a small walled town, built under the Ming as a guard post for the bridge, with a gate at each end of its long axis: the east gate faced Fengtai and Beiping, and the west gate opened onto the Lugou Bridge. North of the stone bridge the Beiping–Hankou railway crossed the Yongding on a steel bridge of its own, and the river’s broad, mostly dry bed was lined by embankments. On the embankment north of the railway stood a small temple to the Dragon King, Longwangmiao, where the Chinese kept an outpost. East of the town the ground was open: fields, sand and a low rise that the Japanese called Ichimonji-yama, Yiwenzi Hill in its Chinese reading, and the Chinese called Shagang.

The Japanese used that open ground for training. A note from the Chinese government of 15 November 1913 had designated four places, Lugouqiao among them, as drill grounds for the foreign legation guards, but only “temporarily” and on the condition that “ball cartridges are not to be used.” The American embassy noted that the Japanese company drilling there on 7 July carried ball cartridges, that is, live ammunition (Embassy despatch no. 1377, FRUS). Through June and early July the Japanese had exercised there almost every night, and the Chinese troops on the embankment and in Wanping had watched them with growing suspicion. Both armies were expecting trouble. Neither, as far as the records show, had planned it for that night.

1930 to 2240: the night exercise near Longwangmiao

The company that went out from Fengtai that evening was the 8th Company of Ichiki’s battalion, commanded by Capt. Shimizu Setsuro. According to the Japanese accounts, Shimizu changed his plan that day and, at about 1930, began a night exercise on the waste ground east of Longwangmiao, between the railway embankment and Wanping’s north wall. The exercise practiced an approach to an enemy position in the dark, followed by a dawn attack, a type of training the Japanese army stressed. A small detachment played the part of the enemy. More than 200 Chinese laborers had been at work on the river embankment during the day, and Chinese soldiers were in their positions on the embankment and at Longwangmiao. The two forces were well under a kilometer apart (Japanese official history, Senshi sōsho, as summarized in Japanese Wikipedia, “Rokōkyō jiken”; Iwatani, “From Marco Polo Bridge to Shanghai”).

The night was dark. By astronomical tables the moon was a day short of new, and the Japanese accounts describe pitch blackness. Shimizu’s men advanced in skirmish order toward the mock enemy and then, at about 2230, the exercise ended and the platoon leaders began to assemble their men for the march back. Chinese sources agree that the Japanese were exercising and add that, in the preceding days, Chinese officers had reported Japanese troops lingering near Wanping and digging. Wang Lengzhai, the magistrate of Wanping, had ordered the east gate closed for the night, and Jin Zhenzhong had sent his 9th Company to reinforce the east wall (Phoenix News, 3 September 2014, summarizing Chinese accounts).

2240: shots in the dark, and a missing private

At about 2240, the Japanese accounts say, the men playing the enemy fired blank rounds from a light machine gun, part of the exercise. Then live rounds came from behind the company, from the direction of the Chinese positions on the embankment: a few shots at first. Shimizu had the bugle sound assembly, and more shots followed, about a dozen, from near the railway embankment. Japanese soldiers later said they saw lights flashing on the wall of Wanping and on the embankment, as if signals were being passed (Japanese Wikipedia, “Rokōkyō jiken,” after the regimental and battalion reports; Iwatani). Iwatani, summarizing the battalion’s battle report, gives two shots at 2240 from the Chinese position near the riverbank.

The Chinese garrison in Wanping heard the firing too. Chinese accounts of what happened next differ from the Japanese ones. One version, repeated in Chinese anniversary histories, says that a handful of Japanese soldiers approached the Chinese post at the east end of the bridge and that the sentry fired back (Phoenix News, 2014). Other Chinese accounts say only that the shooting was heard and that the Chinese troops did not fire. The ambassador’s telegram written that morning recorded the Chinese guards’ claim that “the Chinese side did not respond” to Japanese firing later in the night (Johnson to Secretary of State, received 8 July 1937, FRUS 1931–41, vol. I).

When Shimizu took the roll, one man was missing: Private Shimura Kikujiro, a runner for the 1st Platoon leader. Shimizu sent two mounted messengers, Sgt. Iwatani Heiji and Pvt. Uchida Ichitaro, back to Fengtai to report to Ichiki that the company had been fired on and that a soldier was missing. The company searched for Shimura, and within about twenty minutes he turned up. The Japanese accounts say he had stepped away to relieve himself and lost his way in the dark. Peter Harmsen, writing in Shanghai 1937, gives a less flattering story, that he had slipped off to visit a brothel. Whatever the reason, his return was not reported to Ichiki before the messengers’ first report had already set the battalion in motion (Japanese Wikipedia, “Rokōkyō jiken”; Harmsen, Shanghai 1937; Phoenix News, 2014). Chinese writers have long charged that Mutaguchi knew of Shimura’s return and concealed it; the Japanese accounts say the news simply arrived late.

Shimizu pulled the company back to the hamlet of Wulidian, about 1.8 km east of the bridge on the road to Fengtai, arriving at about 0100 on 8 July (Japanese Wikipedia, after the battalion report). The missing soldier was found, and the shooting had hurt no one. On another night that might have been the end of it.

Midnight: Fengtai, Beiping and the demand to search Wanping

The messengers reached Fengtai at about midnight. Ichiki telephoned Mutaguchi in Beiping. The brigade commander, Maj. Gen. Kawabe Masakazu, was away from the city, and the commander of the China Garrison Army, Lt. Gen. Tashiro Kanichiro, lay seriously ill in Tianjin; he died there on 16 July. Mutaguchi was therefore the senior officer on the spot. According to the Japanese accounts, he ordered Ichiki to take the battalion out, occupy Yiwenzi Hill east of Wanping, and after daybreak negotiate with the commander of the Chinese garrison in the town (Japanese Wikipedia, “Rokōkyō jiken”; Iwatani). Mutaguchi’s order sent several hundred more Japanese soldiers into the dark toward a Chinese-held town.

At the same time the news went up the other Japanese chain, the political one. In Beiping, Col. Matsui Takuro headed the army’s special-service agency (tokumu kikan), the office that handled the army’s dealings with Song Zheyuan’s regime. Maj. Imai Takeo, the assistant military attaché at the Japanese embassy, was the army’s other main contact with the Chinese leaders in the city. Around midnight, according to the Chinese accounts, Matsui telephoned the Hebei–Chahar council’s foreign affairs office. He said that Japanese troops had been fired on and a soldier was missing, and demanded that Japanese troops be allowed to enter Wanping at once to search for him. Qin Dechun refused. Wanping and the bridge were Chinese territory, he said, and the Japanese had no right to search them; he ordered that no Japanese soldier be let in, and that Chinese troops fire only if fired upon. When Matsui called back and threatened to enter by force, Qin answered that his troops would resist (Qin Dechun’s account as summarized by Phoenix News, 2014; Chinese Wikipedia, “Qiqi shibian”).

The Japanese accounts differ on what was asked. Hata Ikuhiko and other Japanese historians have argued that the Chinese account is wrong on one point: by the time the agency approached the Chinese, its concern was to investigate who had fired, not to find the missing soldier, who had already been found (Hata, Rokōkyō jiken no kenkyū, 1996, as cited in later surveys). Qin’s own version, given at the time to the American embassy, was that the Japanese reported being fired on, that a soldier was then said to be missing, that the Japanese demanded to enter Wanping to look for the man who fired, and that they kept pressing to enter even after the soldier was back (Embassy despatch no. 1377, FRUS). Imai’s first statement to the Americans was that Chinese troops “had fired suddenly on Japanese soldiers at Lungwanmiao” (FRUS, same despatch). Both sides quickly settled on stories in which the other side had acted first.

The two chains of command pulled in different directions. The special-service agency, the embassy and the Japanese diplomats in Beiping wanted the matter settled locally and quickly. The garrison headquarters at Tianjin ordered the troops to demand a formal apology and to be ready to use force. Beiping had been under a night curfew and martial law since 26 June (Iwatani, 2016), and the Chinese troops were already on edge. On the Chinese side, He Jifeng ordered the 219th Regiment to watch the Japanese closely and to counterattack if provoked, and Qin Dechun put Feng Zhian’s 37th Division on alert (English Wikipedia, “Marco Polo Bridge incident,” after Chinese and Japanese sources). Both sides agreed in principle to send a joint party to Wanping to look into what had happened.

0200 to 0500: Yiwenzi Hill and the joint investigation party

Ichiki left Fengtai with the main body of his battalion and, just after 0200, met Shimizu at the western edge of Wulidian, where he took command of the 8th Company as well. At about 0200 the Chinese authorities were told that Shimura had returned (Iwatani). By then, however, the Japanese purpose had changed. The question was no longer where a private had gone but who had fired on Japanese troops. At 0320 the battalion occupied Yiwenzi Hill, the low rise east of Wanping, and waited for daylight (Japanese Wikipedia, “Rokōkyō jiken”).

At about 0325, three more shots were heard from the direction of Longwangmiao. Ichiki took them as proof that the Chinese meant to fight (Japanese Wikipedia, after the battalion report). The Chinese side recorded the same hour differently. Ambassador Johnson’s first telegram on the incident, received in Washington at 0405 on 8 July (Washington time), reported that according to the Chinese guards at Wanping “the Japanese began firing on the city at about 3:30 a.m.” and that the Chinese had not answered (FRUS 1931–41, vol. I). Both sides heard shooting at about the same time; each said the other had fired.

Meanwhile the joint investigation party had set out from Beiping at about 0400. On the Japanese side it included Lt. Col. Morita Toru of Mutaguchi’s regiment, Maj. Akafuji Shoji, Maj. Sakurai Tokutaro, a military adviser to the 29th Army, and Capt. Teradaira Tadasuke of the special-service agency, with two interpreters and an escort. On the Chinese side went Wang Lengzhai, the magistrate of Wanping, who was also the Hebei–Chahar council’s commissioner for the district, and Lin Gengyu of the council’s foreign affairs commission. They reached Wanping at about 0500 (Japanese Wikipedia, “Rokōkyō jiken”).

Wang Lengzhai later wrote one of the main Chinese accounts of these hours. In the Chinese version, as the party came up to the town, the Japanese officers saw their own troops deployed east of Wanping and demanded that the Chinese troops withdraw from the east gate so that Japanese soldiers could enter. Wang refused and insisted that any talks be held inside the town and that the facts be established first (Chinese Wikipedia, “Qiqi shibian”; Phoenix News, 2014). Earlier, at about 0300, Sakurai had talked with Qin Dechun in Beiping. According to the daily log of the Japanese army agency in Beiping, the two agreed that neither side would move up more troops or spread the fighting, and Qin said that Feng Zhian’s troops had not been deployed outside the bridge area: any soldiers found outside it would not be Chinese troops, and the Chinese side would not object if the Japanese attacked them. Iwatani, who found the log, notes that this answer became the Japanese basis for attacking the Chinese positions outside the town (Iwatani, 2016). Ichiki, for his part, ordered his men not to attack the town itself but to advance on the Chinese positions along the embankment (Japanese Wikipedia, “Rokōkyō jiken”). Chinese accounts do not record Qin saying any such thing, and if he did, he seems to have meant it as a denial that his troops were out of place rather than as permission to attack them.

At about 0420, Ichiki telephoned Mutaguchi again to report the shots from Longwangmiao. Mutaguchi authorized him to fight. The regimental combat report gives the time of that authority as 0420; some accounts give 0400 (Japanese Wikipedia, after the regimental report; Iwatani gives about 0300 for Ichiki’s order to advance and prepare to fire). The Japanese battalion began to move forward toward the embankment in the half-light before dawn. Ichiki saw large numbers of Chinese soldiers on the embankment just after 0500 and ordered his mortars to fire. Morita, who had arrived at the front as acting regimental commander, stopped him: Morita had not yet approved combat (Japanese Wikipedia, “Rokōkyō jiken”).

0530: the attack on Longwangmiao and the railway bridge

At about 0530 on 8 July the Japanese attack began. The Japanese records do not agree with each other on exactly how. In one version, built on the regimental report, Chinese mortar and rifle fire struck the Japanese troops gathering near Longwangmiao and toward Changxindian at about 0400 to 0500, and the Japanese answered and took the temple. In another, given by the Japanese historian Yasui Sankichi from a Japanese veteran’s account, an attack order was given at about 0530; a Chinese officer on the embankment called on the Japanese to halt; the Japanese lieutenant answered that it was an exercise and kept advancing; and the Chinese opened fire at about 20 meters (Japanese Wikipedia, “Rokōkyō jiken”). Iwatani, summarizing the 3rd Battalion’s battle report, writes that the Chinese called on the Japanese to stop and opened fire when they did not, and that both sides went into action at 0530.

The Chinese accounts put it more simply: the Japanese, still insisting on entering Wanping, attacked when the Chinese refused (Chinese Wikipedia, “Qiqi shibian”). Wang Lengzhai was in the town, in the middle of the talks, when the firing began. Qin Dechun told the American embassy that “while negotiations were in progress, Japanese opened fire” (FRUS, despatch no. 1377).

The fighting was concentrated on the river embankment north of the town, around Longwangmiao and the east end of the railway bridge. The Chinese posts there were small, a platoon or so each. Chinese accounts say the two platoons defending Longwangmiao and the railway bridge were almost wiped out (Chinese Wikipedia). Ji Xingwen’s men in Wanping were under orders to hold the bridge; Chinese accounts quote the order that the Lugou Bridge would be their grave. The Japanese took Longwangmiao and the east end of the railway bridge. After about two hours, at about 0730, the fighting died down. At about 0930, at the Chinese request, the two sides agreed to a temporary cease-fire (Japanese Wikipedia, “Rokōkyō jiken”).

8 July: shells on Wanping and the Americans who drove out

In Beiping, the American military attaché, Col. Joseph W. Stilwell, had spent the evening of 7 July on a boat on the Beihai lake with Ambassador Johnson and others. A Marine officer in the party had heard that Japanese troops would go out that night for maneuvers near the bridge. At 0800 on 8 July Stilwell’s office learned of the skirmish. Stilwell sent his assistant, Maj. David D. Barrett, to drive out with the journalist John Goette. They found the scene quiet, with only an occasional rifle shot, and near the Japanese lines they saw the body of a dead Japanese soldier guarded by a platoon. Barbara Tuchman, telling the story from Stilwell’s papers, writes that the two men understood at once what the body would be used for (Tuchman, Stilwell and the American Experience in China; “Japan Strikes, 1937,” American Heritage). Johnson’s telegram that morning reported that “desultory firing” was still going on near the bridge as late as 0830 (FRUS 1931–41, vol. I).

The truce of the morning did not hold. Firing resumed in the afternoon, limited at first (Japanese Wikipedia gives about 1530). Chinese accounts describe a day of demands: Mutaguchi called on the Chinese troops to withdraw west of the Yongding; Mutaguchi came up to the front at Shagang in the early afternoon; at about 1600 he invited Wang Lengzhai, Ji Xingwen and Jin Zhenzhong to talks outside the walls, which Wang refused on the grounds that officers could not leave their posts; and at about 1700 the Japanese issued an ultimatum with a deadline of 2000 (Chinese Wikipedia; Phoenix News, 2014). About the same hour the brigade commander, Kawabe, now back from his absence, reached Fengtai and ordered his troops to gather near the Lugouqiao station for an attack the next morning. At about 1805 Japanese artillery opened fire on Wanping. Shells hit the district commissioner’s office in the town and destroyed it, and Jin Zhenzhong was wounded (Chinese Wikipedia; Phoenix News, 2014). In Tokyo, meanwhile, the Army General Staff at 1842 that evening ordered the China Garrison Army to avoid any further use of force, to keep the incident local (Iwatani, 2016).

That night the Chinese struck back. Men of He Jifeng’s brigade, many of them armed with the broadswords for which the 29th Army was famous, raided the Japanese positions on the embankment and, by Chinese accounts, retook the railway bridge and the Longwangmiao area (Chinese Wikipedia; Phoenix News, 2014). In Beiping, Qin Dechun met Japanese representatives through the night. Imai Takeo and Qin reached a verbal cease-fire at about 0200 on 9 July (Chinese Wikipedia). The terms that emerged over the following hours were that the Japanese would pull back toward Fengtai, the Chinese troops would withdraw west of the river, and Wanping would be held by a Chinese security force of a few hundred men. Even that went badly: Chinese accounts say Japanese guns fired on Wanping again early on 9 July, and that the security force was fired on as it came up (Phoenix News, 2014). Stilwell himself tried to drive to the Chinese garrison on the third day, and turned back when both sides opened heavy fire as he crossed the last five hundred yards (Tuchman).

The table below sets out the night as each side’s records give it. The times are approximate, and in several places the two sides do not even agree on what happened, let alone when.

Who fired first?

There are really two “first shots” in this story, and they need to be kept apart. The first are the shots of about 2240 on 7 July, fired at or near Shimizu’s company in the dark. The second are the shots of the early morning of 8 July, from the three shots at about 0325 to the opening of the battle at about 0530. Most of the argument concerns the first, because they set everything else in motion, but the second matter as much. By 0530 there was a Japanese battalion deployed outside a Chinese town, with its commander authorized to fight, and a Chinese garrison ordered to hold its ground. Whoever pulled the trigger at that point, a fight was very likely.

In their words

Presumably it will never be established whether a Japanese or a Chinese fired the shot which precipitated the fighting.

The Japanese account: Chinese troops fired on an exercise

The Japanese army’s version, given from the first morning, was that soldiers of the 29th Army fired without provocation on a Japanese company carrying out a lawful exercise, fired again before dawn, and fired first on the morning of 8 July, so that the Japanese acted in self-defense. The regimental and battalion reports, Ichiki’s judgment at 0325, and Imai Takeo’s first statement to the Americans all follow this line. Later the Japanese cabinet went further and claimed that the Chinese had planned the clash, pointing to cut telephone lines and closed gates (Japanese Wikipedia, “Rokōkyō jiken”). The most careful modern Japanese study, Hata Ikuhiko’s Rokōkyō jiken no kenkyū (A Study of the Marco Polo Bridge Incident, 1996), which Edward Drea and Tobe Ryoichi have called the single best source on the incident, concluded that the shots of 2240 most likely came from Chinese soldiers on the embankment. Hata treated it as an accidental shot, probably fired by a low-ranking Chinese soldier out of fear in the dark, rather than a planned act (Hata, 1996, as summarized in English Wikipedia, “Marco Polo Bridge incident”; Hata’s earlier English essay, “The Marco Polo Bridge Incident, 1937,” in James Morley, ed., The China Quagmire, 1983).

The Chinese account: a Japanese pretext

The Chinese government called the clash premeditated from the start. By 9 July the Foreign Ministry in Nanjing had protested orally to the Japanese embassy against an “attack by Japanese troops upon Chinese troops” (FRUS 1937, vol. III, doc. 107). In the Chinese account, the Japanese held a provocative exercise at night beside a Chinese town, used a missing soldier as an excuse to demand entry to Wanping, kept demanding entry after the soldier had returned, and attacked while the talks were going on. The search demand, on this view, was a pretext, just as the railway explosion at Mukden had been in 1931. This remains the account of the Museum of the War of Chinese People’s Resistance against Japanese Aggression, which stands inside Wanping today. After the war the Tokyo tribunal accepted much of it. Its judgment said that on the night of 7 July “Japanese garrison troops at Lukouchiao held an unusual manoeuvre,” that they demanded entry to Wanping “alleging that a Japanese soldier was missing,” and that fighting broke out “while the Japanese complaint was still under negotiation.” It found that the opening battle of the war “was planned, instigated and carried out upon the initiative of the Army itself” (IMTFE Judgment, ch. IV, pp. 183–185).

Theory: a Communist provocation

A third theory, popular with some Japanese writers since the war, holds that Chinese Communist agents fired the shots to provoke a war between Japan and the Nationalists, a war that would save the Communist movement. Its supporters point to Communist members and sympathizers inside the 29th Army, among them He Jifeng himself, who commanded the brigade at the bridge, was in contact with the Communists and later joined the party in secret; to the Communist Party’s call for national resistance issued on 8 July; and to the fact that the Communists gained most from the war. No document has ever been produced showing that Communist agents fired at Lugouqiao. Hata considered the theory possible but less likely than an accidental shot by a Chinese soldier (English Wikipedia, after Hata). Kimitada Miwa’s 1963 study in Monumenta Nipponica examined the Communists’ role in spreading the incident into a full-scale war, which is a different matter from firing its first shots.

Theory: nobody meant it

The fourth explanation is a mistake: that nervous Chinese sentries heard the blank fire of the Japanese exercise in the dark, took it for an attack, and fired, or that a stray or careless shot was fired by someone on either side. The Japanese historians Yasui Sankichi and the former staff officer Hattori Takushiro both treated the opening shots as accidental. Yasui added a pointed observation: the Japanese records themselves show an attack order and a Japanese advance on the Chinese positions on the morning of 8 July, so the Japanese were not simply responding to Chinese fire (Japanese Wikipedia, “Rokōkyō jiken”). Iwatani Nobu, whose 2016 paper drew on the Japanese battle reports, the Beiping agency’s log and Chiang Kai-shek’s diaries, wrote that neither high command wanted the incident to become a full-scale war but that both failed to contain it, and that mutual mistrust led each side to misread the other; his 2023 study, drawing on archives in several countries, likewise concluded that the incident was a minor clash that escalated because of deep distrust on both sides, and found no premeditated conspiracy by either the Japanese army or the Chinese Communists (Iwatani, “Revisiting the Outbreak of the Sino-Japanese War,” 2023, as summarized in English Wikipedia).

What historians conclude

James B. Crowley’s 1963 article in the Journal of Asian Studies, “A Reconsideration of the Marco Polo Bridge Incident,” was the first major Western study to use the Japanese army’s own records. Crowley found no evidence that the Japanese army had planned the clash. The army’s leadership in Tokyo, and above all the operations chief Ishiwara Kanji, wanted no war in China in 1937, because they were preparing for a possible war with the Soviet Union. Crowley placed the responsibility for the war not on the shots in the dark but on the decisions of the following days, in Tokyo and in China, by men who thought one more show of force would settle the North China question. He developed the argument in Japan’s Quest for Autonomy (1966). The American embassy reached a similar view in 1937. Johnson wrote on 14 July: “I do not believe that either the Japanese Government or the Japanese Army engineered the incident at Marco Polo Bridge” (FRUS 1937, vol. III, doc. 155). The embassy’s later despatch added that the Japanese army soon used the incident as a pretext to extend its influence in North China.

The leading recent histories in English take the same line. Rana Mitter, in Forgotten Ally (published in Britain as China’s War with Japan, 1937–1945), treats the origin of the shooting as unknown and turns at once to what mattered more: that this time neither Chiang Kai-shek nor the Japanese government would back down, as each had done after earlier incidents. Hans van de Ven, in China at War, likewise treats the clash as one that local commanders on both sides tried to contain, and finds the turning point in the political decisions that followed, Tokyo’s decision of 11 July to send reinforcements and Chiang’s decision to send central government divisions north and to make his stand at Lushan. James Huffman sums up the field: historians have widely rejected the idea that Japan planned the incident as a pretext in the manner of Mukden (Huffman, Modern Japan: An Encyclopedia), though Chinese official histories still hold that it did.

That leaves a narrow area of agreement. Almost no historian outside China now argues that the Japanese government or the army high command planned the shots of 7 July. Almost none argues that the Japanese army on the spot behaved innocently. Japanese troops exercised at night, with live ammunition, beside a Chinese town whose garrison they knew to be hostile; their commanders took a missing-soldier report, which they soon knew to be false, as the basis for a demand to enter that town; and by dawn they had a battalion deployed with authority to attack. As for who fired at 2240, the only defensible answer is the one the American embassy gave in 1937: it will probably never be known.

Timeline

Wanping, 7–11 July 1937

1. 7 Jul, about 1930Shimizu’s 8th Company begins its night exercise east of Longwangmiao.
2. 7 Jul, about 2240Blank fire in the exercise; live shots heard from the direction of the Chinese positions.
3. 7 Jul, about 2300Pvt. Shimura Kikujiro missing at roll call; found about 20 minutes later.
4. 8 Jul, about 0000Ichiki reports to Mutaguchi; Matsui demands entry to Wanping; Qin Dechun refuses.
5. 8 Jul, 0200–0320Ichiki takes command at Wulidian; the battalion occupies Yiwenzi Hill.
6. 8 Jul, about 0325Three shots from the Longwangmiao direction (Japanese); Japanese fire on the town (Chinese).
7. 8 Jul, about 0400–0500Joint investigation party with Wang Lengzhai reaches Wanping; Mutaguchi authorizes combat.
8. 8 Jul, about 0530Japanese attack on Longwangmiao and the railway bridge; two Chinese platoons nearly destroyed.
9. 8 Jul, about 0930Temporary cease-fire; firing resumes in the afternoon.
10. 8 Jul, about 1805Japanese artillery shells Wanping; Jin Zhenzhong wounded.
11. 8–9 Jul, nightChinese counterattack on the embankment; verbal cease-fire between Imai and Qin at about 0200 on 9 July.
12. 11 JulKonoe cabinet decides to send reinforcements to North China (evening); local cease-fire terms signed the same day.

People

Shimizu Setsuro, captain, commanded the 8th Company of the 3rd Battalion, 1st Infantry Regiment, and led the night exercise of 7 July. His report that his company had been fired on and a man was missing was the first link in the chain that led to war.

Shimura Kikujiro, private second class, a runner for the 1st Platoon leader, was the missing soldier. He returned within about twenty minutes. Japanese accounts say he had gone to relieve himself and lost his way; Harmsen says he had visited a brothel.

Ichiki Kiyonao (1892–1942), major, commanded the 3rd Battalion at Fengtai. He judged the shots of 0325 to be proof of Chinese hostility and led the attack of 8 July. Five years later, as a colonel, he led the Ichiki Detachment to its destruction at the Battle of the Tenaru on Guadalcanal in August 1942, where he died.

Mutaguchi Renya (1888–1966), colonel, commanded the 1st Infantry Regiment and was the senior Japanese officer in Beiping on the night of 7 July. He authorized Ichiki to fight. As a lieutenant general he later commanded the Fifteenth Army in the disastrous Imphal offensive of 1944, and he is remembered in Japan for both.

Matsui Takuro, colonel, headed the Japanese army’s special-service agency in Beiping and was the army’s main negotiator with the Hebei–Chahar authorities; he negotiated the local agreement of 11 July with the Chinese side. Imai Takeo, major, assistant military attaché at the Japanese embassy, gave the first Japanese account to the American embassy and reached a verbal cease-fire with Qin Dechun early on 9 July.

Qin Dechun (1893–1963), mayor of Beiping and deputy commander of the 29th Army, acted for Song Zheyuan in the city. He refused the Japanese demand to search Wanping and ordered that no Japanese soldier be admitted. He later went to Taiwan.

Wang Lengzhai, magistrate of Wanping and administrative commissioner for the district, went with the joint investigation party in the early hours of 8 July, refused to open the east gate, and was inside the town when the fighting began. His later account is one of the main Chinese sources for the night.

Ji Xingwen (1908–1958), colonel, commanded the 219th Regiment, whose 3rd Battalion held Wanping and the bridges. He became a national hero in China for the defense of the bridge. He was killed by Communist artillery fire on Kinmen in August 1958. Jin Zhenzhong commanded the 3rd Battalion in Wanping and was wounded when Japanese shells hit the town on the evening of 8 July.

Losses and uncertainty

For a battle so often described, the losses of 7–9 July are poorly recorded. The shots of 2240 hit no one. The fighting of 8 July did kill and wound men on both sides, but the figures depend on the source and on how many days they cover. Hata, using the Japanese records, gives about 510 Japanese soldiers in action, with 10 killed and 30 wounded, and Chinese losses reported as about 180, including about 60 dead (Hata, 1996, p. 218, as cited in the Japanese reference work Sekaishi no mado). Japanese Wikipedia, in a passage marked as needing a source, gives Japanese losses of more than ten killed and wounded and Chinese losses of more than 20 killed and more than 60 wounded for the morning of 8 July. Chinese accounts give no overall total but say that the two platoons at Longwangmiao and the railway bridge were almost wiped out. Barrett saw one dead Japanese soldier on the morning of 8 July.

The numbers of men present are uncertain too. Johnson estimated that the incident itself involved “not more than 200 Japanese soldiers,” and the embassy reported about 1,000 Japanese troops advancing to the Wanping area on 10 July (FRUS). Chinese accounts give Jin Zhenzhong’s reinforced battalion about 1,400 men, but only a few platoons were in the positions north of the town that bore the Japanese attack. English Wikipedia, following Chinese sources, says the immediate garrison of the bridge was about 100 men.

By the numbers

Men on the ground at Wanping, 7–10 July 1937

Approximate strengths, from the sources named below

Japanese, night exercise, 7 Jul≤200

Japanese in action, 8 Jul510

Japanese near Wanping, 10 Jul~1,000

Chinese battalion, Wanping~1,400

Times are the other great uncertainty. The Japanese and Chinese accounts often differ by thirty minutes to an hour for the same event, and in some cases they place an event in a different hour altogether. Mutaguchi’s authority to fight is given as 0400 or 0420. The Japanese date the start of the battle to 0530, while some Chinese accounts describe a first Japanese attack at about 0400, during the talks. The two sides were not using synchronized watches, and many of the times were written down afterward from memory, by men with every reason to show that the other side had moved first. This chapter gives the times in the Japanese regimental reports as summarized by Japanese historians, notes Chinese times where they differ, and treats all of them as approximate.

Questions readers ask

Was the missing soldier ever really missing?

Briefly. Pvt. Shimura Kikujiro missed the roll after the shots of 2240 and was found about twenty minutes later, but by then the first report had already set Ichiki’s battalion in motion. The Chinese were told of his return at about 0200, and the Japanese demand to enter Wanping continued, now framed as an investigation of who had fired.

Why were Japanese troops exercising next to a Chinese town at night?

Japan relied on the Boxer Protocol of 1901 and a Chinese note of 1913 naming Lugouqiao as a drill ground for the foreign guards. The American embassy pointed out that the note was temporary and barred live ammunition, which the Japanese company carried that night. The Chinese took the near-nightly exercises of that summer as provocation.

Did the Japanese plan the incident, as they did at Mukden?

The Tokyo tribunal found that they did, and Chinese historians still hold that view. Most historians outside China, starting with James Crowley in 1963 and including Hata Ikuhiko, Rana Mitter and Hans van de Ven, have found no evidence of a plan by the Japanese government or army high command, and note that Tokyo’s army leadership did not want a war in China that summer. They also agree that the Japanese army on the spot behaved provocatively and quickly used the incident to press its demands in North China.

Could Communist agents have fired the first shots?

It has been claimed, mainly by Japanese writers, but no document or witness has ever shown it. The Communists did have members and sympathizers in the 29th Army, and they gained a great deal from the war, but Hata, who examined the theory, considered an accidental shot by a Chinese soldier more likely.

How many people were killed?

No one was hit by the shots of 7 July. In the fighting of 8 July, Hata gives Japanese losses of 10 killed and 30 wounded, and Chinese losses reported at about 180, including about 60 dead. Other accounts give lower figures for the morning fight alone. The much larger battles for Beiping and Tianjin later in July are told in Chapter 5.

Why is it called the Marco Polo Bridge?

Because Marco Polo described a great stone bridge over this river in his account of his travels in the thirteenth century, and Western writers named it after him. Its Chinese name is Lugouqiao, the bridge over the Lugou, an old name for the Yongding River. In China the incident is usually called the Lugouqiao Incident or the Double Seven (7 July) Incident; in Japan, the Rokōkyō Incident. The bridge and the town are described in Chapter 2.

How it connects

The night at Wanping grew out of six years of Japanese pressure on North China, from Mukden in 1931 through the Tanggu Truce of 1933 and the agreements of 1935, told in Chapter 1: From Mukden to the Bridge. The garrisons that faced each other across the fields east of Wanping, and the bridge and town themselves, are described in Chapter 2: The Bridge, the Town and the Garrisons.

What happened next, the twenty days of talks, local agreements, reinforcements and ultimatums that turned the incident into a war, is the subject of Chapter 4: Twenty Days. For the United States, the shots at Wanping were the start of a chain of events that ran through the sinking of the Panay in December 1937 to the oil embargo of 1941 and the attack on Pearl Harbor, told in Chapter 8 and in the Pearl Harbor collection.

What happened afterward

The local cease-fire of 9 July did not end the crisis. On 10 July Japanese troops came back toward Wanping, about 1,000 of them by the American embassy’s count, and there was more firing. On 11 July, local commanders signed terms under which the 29th Army would apologize, punish those responsible, withdraw from the Wanping area and suppress anti-Japanese groups. The same day the Konoe Fumimaro cabinet in Tokyo decided to send reinforcements to North China from Manchuria, Korea and Japan itself (Tobe, 2019). Sources differ on which came first: one Japanese account puts the cabinet decision at about 1830 and the local signing at about 2000 (Bunshun Online, 2017), while Iwatani, from the Beiping army agency’s daily log, gives 1750 for the signing (Iwatani, 2016). On 17 July Chiang Kai-shek told the nation in his Lushan speech that China had reached the limit of what it could concede.

Within three weeks the war had spread across the North China plain. After clashes at Langfang and Guang’anmen on 25 and 26 July, the Japanese attacked on 28 July; Tong Linge and Zhao Dengyu were killed at Nanyuan, and the 29th Army left Beiping that night, as told in Chapter 5. In August the fighting spread to Shanghai (Chapter 6), and in December the capital, Nanjing, fell (Chapter 7).

The men of 7 July went on into the war. Ichiki was killed on Guadalcanal in August 1942, leading a night attack across the mouth of a creek against dug-in U.S. Marines. Mutaguchi rose to lieutenant general and led the Imphal offensive of 1944, one of the worst defeats in the history of the Japanese army; he lived until 1966 and spent much of his later life defending his record. Ji Xingwen fought on through the war, followed the Nationalist government to Taiwan, and was killed on Kinmen in 1958. Wang Lengzhai stayed on the mainland. In China the night of 7 July became a national day of remembrance, and the bridge one of the country’s most visited memorial sites.

Walking the ground today

The bridge and the town are now inside the city of Beijing, in Fengtai District, about 15 km southwest of Tiananmen Square. The Lugou Bridge itself is closed to traffic and preserved as a monument; you walk across it past the stone lions on its balustrades. Look north from the bridge and you will see the railway bridge that carries today’s Beijing–Guangzhou line, the successor of the Beiping–Hankou railway, roughly where the fighting of 8 July took place. The site of Longwangmiao lies on the embankment beyond it, and the open ground of the night exercise, north and east of the town, has long since been built over or turned into parks.

Wanping still has its walls and its two gates. Inside the town, on its main street, stands the Museum of the War of Chinese People’s Resistance against Japanese Aggression, opened in 1987, the national museum of the war. Its exhibits begin with this night and present the Chinese account of it. On the south wall of the town, marks left by the Japanese shelling of 8 July are picked out with a memorial plaque, and a row of inscribed stone drums along the wall tells the story of the war. Close to the town, the Sculpture Park of the War of Resistance holds large groups of monumental figures, and the tomb of Gen. Zhao Dengyu, killed at Nanyuan on 28 July, is not far away. Each year on 7 July a commemoration is held at the museum.

The museum generally requires a free advance reservation with a passport and closes on some days of the week, so check current arrangements before you go. From the town wall at dusk, with trains crossing the river to the north, it is not hard to imagine the dark ground below on the night of 7 July 1937, and how little either side could see of the other.

Sources and further reading

• U.S. Department of State, Foreign Relations of the United States: Ambassador Johnson’s telegram received 8 July 1937 (FRUS 1931–1941, vol. I, doc. 228); Peck’s telegram of 9 July 1937 (FRUS 1937, vol. III, doc. 107); Johnson’s telegram of 14 July 1937 (FRUS 1937, vol. III, doc. 155); Embassy despatch no. 1377 on the incident and its background (FRUS 1937, vol. III, doc. 484).
• International Military Tribunal for the Far East, Judgment, ch. IV, “The Military Domination of Japan and Preparation for War” (HyperWar), pp. 183–187.
• James B. Crowley, “A Reconsideration of the Marco Polo Bridge Incident,” Journal of Asian Studies 22, no. 3 (1963), and Japan’s Quest for Autonomy: National Security and Foreign Policy, 1930–1938 (Princeton, 1966).
• Hata Ikuhiko, Rokōkyō jiken no kenkyū (University of Tokyo Press, 1996); Hata, “The Marco Polo Bridge Incident, 1937,” in James W. Morley, ed., The China Quagmire (Columbia University Press, 1983).
• Iwatani Nobu (岩谷將), “From Marco Polo Bridge to Shanghai: Initial Phase of the Sino-Japanese War Revisited” (National Institute for Defense Studies, Japan, 2016); Nobu Iwatani, “Revisiting the Outbreak of the Sino-Japanese War” (2023).
• Tobe Ryoichi, “The Sino-Japanese War’s Expansion and the Japanese Army, July 1937–October 1938”, International Forum on War History, 2019.
• Kimitada I. Miwa, “The Chinese Communists’ Role in the Spread of the Marco Polo Bridge Incident into a Full-scale War”, Monumenta Nipponica 18 (1963).
• Rana Mitter, Forgotten Ally: China’s World War II, 1937–1945 (Houghton Mifflin Harcourt, 2013), published in Britain as China’s War with Japan, 1937–1945; Hans van de Ven, China at War: Triumph and Tragedy in the Emergence of the New China (Harvard, 2018); Mark Peattie, Edward Drea and Hans van de Ven, eds., The Battle for China (Stanford, 2011); Peter Harmsen, Shanghai 1937: Stalingrad on the Yangtze (Casemate, 2013), and “Deadly Clash at Marco Polo Bridge”; Jay Taylor, The Generalissimo (Harvard, 2009); Parks M. Coble, Facing Japan (Harvard, 1991).
• Barbara W. Tuchman, Stilwell and the American Experience in China, 1911–45 (Macmillan, 1971), excerpted as “Japan Strikes, 1937,” American Heritage.
• Reference articles used for the two sides’ detailed timetables: Japanese Wikipedia, “Rokōkyō jiken” (after the Japanese official history, Senshi sōsho: Shina jihen rikugun sakusen, vol. 1, 1975, and the regimental reports); Chinese Wikipedia, “Qiqi shibian”; English Wikipedia, “Marco Polo Bridge incident”; Phoenix News (ifeng.com), anniversary account of the incident (3 September 2014); Sekaishi no mado, “Rokōkyō jiken”; Bunshun Online, 7 July 2017.
• James L. Huffman, ed., Modern Japan: An Encyclopedia of History, Culture, and Nationalism (Routledge, 2013); Japan Center for Asian Historical Records, “Marco Polo Bridge Incident” timeline entry and documents.

Photographs: Japanese army and Chinese press photographs of 1937–1940, public domain; modern photographs by Ca.garcia.s, San Lie and Vmenkov, CC BY-SA, via Wikimedia Commons.