Japanese troops march on Wuhan.
Battle of Wuhan.
In January 1938, Wuhan became the de facto capital of wartime China. All elements of China’s anti-Japanese “united front” were represented there, not only the Chinese Communist Party and the Nationalists of the Guomindang, but also third-party movements. Wuhan became a cultural center as well, with refugee intellectuals and military units from all over China mixing as equals on the streets. The defense of the city was likened by the Chinese press to the defense of Madrid by the Spanish Republican forces that was occuring at the same time.
In order to finish Chiang off and break the back of united-front patriotism, the Japanese high command ordered troops from the north and east to carry out a coordinated pincer movement in preparation for an attack on Wuhan that would be launched in the late spring. However, what happened next did not fit the Imperial Japanese Army’s game plan. The city did not fall until October 1938, after ten months and a series of costly battles. Both sides lost more men in 1938 than in any other year of the war. Moreover, from the Chinese standpoint it is clear in retrospect that these battles were critically important—win or lose—because they had the effect of prolonging the war. The fall of Wuhan brought mutual exhaustion and depletion of resources and set the scene for the next stage of the war, which was much slower in pace, consisting of smaller-scale engagements punctuated by Japanese bombing raids of varying intensity.
Why did it take the Japanese ten months to capture Wuhan? In other words, what happened militarily and how did the Chinese side prepare its defense? To answer these questions, we need to step back and take stock of Chinese military power at the outset of the war, with special attention to organization and preparedness.
China began the war with an estimated regular force of 1.7 to 2.2 million men. These troops can be divided into six categories based on their political loyalty to their commander-in-chief, Chiang Kai-shek. First, there were the troops controlled directly by Chiang. Second came troops who had been loyal to Chiang in the past but were less directly controlled by him. The third category were provincial troops over whom Chiang could exercise command in ordinary times. These were followed by provincial troops over which Chiang had little direct influence. The fifth category was the Communist forces: the Eighth Route Army headquartered in the caves of the northwest and the New Fourth Army taking shape in the hill country of the central Yangzi region. Finally, there were the Northeastern or Manchurian units that had been defeated and displaced by the Japanese in 1931. These were loyal to the “Young Marshal” Zhang Xueliang (one of Chiang’s kidnappers at Xi’an in 1936). The first two categories included roughly 900,000 men, with a million more in the independent provincial armies. About 300,000 men were split between the Communist and Manchurian forces.
As commander-in-chief, Chiang Kai-shek exercised effective command over only about half of the units that could be mobilized against the Japanese when war broke out in July 1937. This meant that military decisions were inevitably slow, procedurally complex, and almost always subject to negotiation. Coordination and communication at the division level proved especially difficult. All of this contrasted starkly with the clean, disciplined command structure on the Japanese side.
Geographically, most of the troops loyal to Chiang Kai-shek were located in the area between the Yangzi and Yellow rivers at the beginning of 1938. They had been heavily involved in the defense of Shanghai and Nanjing and had suffered tremendous losses, retreating to Wuhan at half strength under a decimated officer corps. They now numbered about 400,000 men, with command split between generals Chen Cheng and Hu Zongnan. The independent regional armies of the north—especially the divisions under General Han Fuju in Shandong province—had taken heavy casualties, with some units going over to the Japanese and later serving them as puppet troops. This meant that after the sixmonth Japanese onslaught, during which the coastal areas from Peiping-Tianjin to Shanghai and inland to Nanjing were lost, the relatively autonomous, sizable armies that remained were either from the southwest or the northwest. They were commanded by leading militarists such as Li Zongren and Bai Chongxi (Guangxi), Long Yun (Yunnan), and Yan Xishan (Shanxi and Suiyuan). About 700,000 of these troops, the bulk of whom came from Guangxi under generals Li and Bai, were committed to the defense of Wuhan. Of less importance were the Communist forces of the Eighth Route Army, who numbered about 100,000 and remained relatively unscathed and isolated in bases north and east of Xi’an. Altogether, there were probably about 1.3 million men under arms for the defense of Wuhan.
Early in 1938, as retreating troops regrouped around Wuhan and were joined by fresh divisions from Guangxi, war zones were reorganized and plans were laid for the defense of the central Yangzi valley. The Fifth War Zone (the area north of the Yangzi in Anhui, Hubei, and Henan provinces) was under the Guangxi general Li Zongren. Li commanded a mixture of his own Guangxi units under his longtime associate, the brilliant strategist Bai Chongxi, and troops more loyal to Chiang Kai-shek who were commanded by generals Zhang Zhizhong and Hu Zongnan. All together, about 280,000 men fought on the Chinese side in the Fifth War Zone.
A new Ninth War Zone to the south of the Yangzi (in Hunan, Jiangxi, and the southern part of Hubei, with Wuhan as its command center) was put under the most able of the Chiang loyalists, General Chen Cheng. He commanded some 380,000 men in seventy-eight divisions. Many of these troops, such as those under the Cantonese general Zhang Fakui, had strong provincial loyalties. They were expected to work closely with units that were more directly tied to Chiang Kai-shek and commanded by graduates of the Whampoa Military Academy (where Chiang had been the commandant). Added to these were displaced provincial troops of questionable loyalty who had formerly been under the militarists Feng Yuxiang and Zhang Xueliang, and the New Fourth Army of Communist-dominated guerrilla units being assembled under General Ye Ting.
At the beginning of the war, the Chinese had a four-to-one superiority over the Japanese in terms of men on the ground. But the advantage of numbers dissolved in the face of the massive superiority of the Japanese in equipment, mobility, and firepower. During the first six months of the war the Japanese swept south employing the latest in mobile warfare, including the free use of tanks, supporting air strikes, and sophisticated logistical support. The Chinese who faced them in north China and later in the defense of Wuhan and the central Yangzi were for the most part equipped more primitively with small arms, machine guns, and hand grenades. A Chinese air force existed, but it was defensive and not employable in a tactical sense. Chinese forces were most effective fighting at night, when their generals could take better advantage of their numerical superiority. Night fighting was a key factor in the biggest Chinese victory of the war: the outmaneuvering and defeat of several Japanese divisions by Li Zongren and Bai Chongxi at Taierzhuang (fifty miles northeast of Xuzhou) in March-April 1938.
With the exception of the Guangxi divisions, Chinese units were seriously handicapped by an unnecessarily complex command structure. Orders from Chiang Kai-shek had to pass down through six tiers of commanders before action was possible. Moreover, in the distribution of equipment Chiang favored the central army units over which he had direct control and which had loyal commanders belonging to the Whampoa clique. Needless to say, Chiang’s favoritism bred discord and insubordination at all levels of the Chinese field forces.
There were other problems as well. One was the mixture of approaches to the organization and positioning of forces. The Chinese forces had been reorganized using German models, with large armies grouped together as field armies (called war zones). At the same time, Russian influence was evident in the strategic positioning that Chinese units adopted, with division into “front” and “route” armies as well as rear area service units. Thus, there was a lack of coherence in the forces facing the Japanese. Troop placement and support procedures were not rational. On the one hand, Chiang and his generals consistently tried to avoid decisive confrontation with the Japanese so as to reduce the possibility of meeting with irreversible defeat (as had occurred at Shanghai). On the other hand, Chiang also eschewed a commitment to guerrilla warfare as a widespread tactic. The large, multilayered Nationalist-led units overemphasized holding onto communication lines such as railways, presumably for transport and logistical purposes. This had the effect of immobilizing or tying down their main fighting forces, around which the Japanese could easily maneuver.
During the battle of Wuhan in the summer and fall of 1938, the Chinese enjoyed a six-to-one numerical advantage, throwing 1.1 million men (or 120 divisions) against a Japanese force of 200,000 (10 divisions). But in fact the Japanese still had an enormous military advantage. According to one calculation by an American military observer, the fighting capacity (defined as numbers times firepower) of a Guomindang division ranged from one-third to one-twelfth that of its Japanese counterpart.2 Not only was Chinese weaponry inferior, but little replenishment of men and arms took place or was even possible. Leaving aside bravery and fighting spirit, this meant that at the front one hundred Chinese divisions were often no better than twelve Japanese divisions in terms of fighting effectiveness. The Chinese units with the least firepower were often exposed to the heaviest Japanese shelling. This was the situation at Shanghai in the fall of 1937, when the Twenty-Ninth Army commanded by Zhang Zhizhong fought valiantly before buckling under the Japanese bombardment.
The conduct of the war in and around Wuhan in 1938 also illustrated the major strengths and weaknesses of the Chinese forces. What made the victory at Taierzhuang possible during the spring of 1938 was the exploitation by General Bai Chongxi of a temporary balance between Chinese firepower and that of overextended Japanese units. The result was a short-lived Chinese victory. Casualties on both sides were heavy but also roughly equivalent. The Chinese failure to follow up this victory by pursuing the fleeing Japanese units as part of a larger strategy of counterattack reflected problems of command structure, especially Chiang Kai-shek’s jealousy and distrust of generals Bai and Li. The Japanese were thus given time to regroup and launch a full-scale attack on the strategically important city of Xuzhou in Jiangsu province, headquarters of the Fifth War Zone. Generals Li Zongren and Tang Enbo organized a spirited defense, producing one of the bloodiest and most destructive battles of the war. But the Japanese were now able to bring their full firepower to bear. They soon broke the Chinese defenses and closed in on Xuzhou from three sides. Losing two of their own men for every one Japanese soldier killed, the defenders were forced to retreat, abandoning the smoldering ruin of a city on May 18. It was just after the loss of Xuzhou, at the end of May, that Chiang Kai-shek made the famous decision to blow the Yellow River dikes at Huayuankou, just north of Zhengzhou in Henan, in a desperate surprise move designed to slow down the southward advance of Japanese mechanized units. With the Yellow River now flowing freely south through the Huai River bed to join the Yangzi just west of Shanghai, eleven cities and four thousand villages were inundated by flood waters, which left more than 2 million homeless.3 But the military objective was accomplished. For a few weeks the floods blocked the main route south to Wuhan and permitted Chinese troops to regroup in the Dabieshan area on the Anhui-Henan border. In the long run, however, the political cost of the move to Chiang’s government after the war far outweighed the short-term military benefit from the blowing of the dikes.
The Yellow River flood did not prevent major Japanese army and navy units from moving west from Nanjing into the central Yangzi valley, where they took Anqing with ease on June 12, 1938. Hoping to block the Japanese at Madang, halfway between Anqing and Jiujiang, the Chinese prepared an elaborate defensive position. But once again the Japanese outflanked Chinese forces by taking advantage of their failure to coordinate land and naval units in an effective way. The resulting rout and the fall of Madang by the end of June led to a brutal attack on defenseless Jiujiang shortly thereafter and the slaughter of its remaining civilian population. At this point, the lack of coordination between the war zones had become a major problem for which the Chinese were paying dearly. The defense of Madang had been under the Ninth War Zone commander Zhang Fakui. Throughout the battle, crack Guangxi units remained one hundred miles to the north in the Dabieshan area of the Fifth War Zone, the bailiwick of Guangxi general Bai Chongxi. As a result of the fall of Madang and Jiujiang, Bai’s forces became increasingly isolated and vulnerable in the mountains to the north along the Anhui-Henan border. Bai held out until the end of July, when he was finally outflanked by the Japanese and forced to retreat. By August 1938, Wuhan was nearly surrounded and the battle for its survival well underway. Bombing of the city intensified. Recognizing the superiority of Japanese firepower and the convergence on the city of Japanese armored divisions from the north and east, the Guomindang government began planning an orderly evacuation of skilled industrial workers and the shipping of whole factories upstream and into the interior. Wuhan held out longer than expected, but finally fell on October 25, 1938.