Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Ancient fortifications, modern firepower, and warlord politics : subtitle a study on the Siege of Xi’an… Tsang, Kingsley 2002

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


831-ubc_2002-0585.pdf [ 4.77MB ]
JSON: 831-1.0090496.json
JSON-LD: 831-1.0090496-ld.json
RDF/XML (Pretty): 831-1.0090496-rdf.xml
RDF/JSON: 831-1.0090496-rdf.json
Turtle: 831-1.0090496-turtle.txt
N-Triples: 831-1.0090496-rdf-ntriples.txt
Original Record: 831-1.0090496-source.json
Full Text

Full Text

/Ancient F o r t i f i c a t i o n s , Modern Firepower, and Warlord P o l i t i c s A Study on the Siege of Xi'an and i t s H i s t o r i c a l S i g n i f i c a n c e by K i n g s l e y Tsang |fJP=!(Zeng Qingzhang) B.A., The U n i v e r s i t y of B r i t i s h Columbia, 2000 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF Master of A r t s . i n THE FACULTY OF GRADUATE STUDIES (Department of History) We accept t h i s t h e s i s as conforming to the re q u i r e d standard THE UNIVERSITY OF BRITISH COLUMBIA August 2 002 © K i n g s l e y Tsang, 2002 In presenting t h i s thesis i n p a r t i a l f u l f i l l m e n t of the requirements for an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y a v ailable for reference and study. I further agree that permission for extensive copying of t h i s thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. I t i s understood that copying or p u b l i c a t i o n of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of History The U n i v e r s i t y of B r i t i s h Columbia Vancouver, Canada Date August 28, 2002 Abstract T h e Warlord period (1916-28) is a much-neglected era in modern Chinese scholarship. Scholars tend to ignore it because the events were complicated and the warlords acted without an ideological commitment. They are seen as violent but unsophisticated thugs with minimum affects on the history of Chinese military. T h e Siege of Xi'an (April to November 1926) demonstrated the fallacy of this assumption and the uniqueness of the warlord military system. T h e warlords managed to fuse the Chinese and Western military experience in a hybrid warring style. This ad hoc system was utilized with great effectiveness under the circumstances of the time. O n e cannot transplant general military assumptions to this period since they fail to take into account the characteristics of the warlords. This study will ascertain the historical significance of the Siege of Xi'an during the Warlord period. T h e siege was the climax of the 1926 anti-Guominjun campaign between the Northern warlords Feng Yuxiang and the alliance of Zhang Zuolin and W u Peifu. This was the last major campaign of the Warlord period with the three main players engaged in ferocious battles all over North China. The Guominjun prevailed in the end because a small detachment of its soldiers managed to hold the strategic city of Xi'an in an eight months siege. It showed the hybrid nature of the warlord military system as well as relevant regional and local issues. This study is divided into three parts: Part One discusses the historical and geopolitical importance of Xi'an; Part Two briefly summarizes the anti-Guominjun campaign, the military dispositions of the warlords and their strength and weakness; Part Three details the eight months' siege. T h e siege and the circumstance that gave rise to it are reconstructed based on sources from Chinese and Western secondary analyses, newspapers, Shaanxi gazetteer, the biographies of Feng Yuxiang and Y a n Xishan. ii Table of Contents Abstract " Table of Contents iii List of Tables v List of Figures vi 1. Introduction 1 2. Summary of the Xi'an Siege 3 3. History of Xi'an 5 4. Background to the Anti-Guominjun Campaign 7 4.1. Realignment of Power Structure 8 4.2. Rise of Zhang Zuolin 8 4.3. Feng Yuxiang and the Expansion of Guominjun's Territories....9 4.4. Weaknesses of the Guominjun 11 4.5. Feng Yuxiang and Conspiracies 12 4.6. Anti-Guominjun Campaign 13 5. The Siege of Xi'an 15 5.1. The Defense of Xi'an 19 5.2. Tactics of the Siege 22 5.3. Feng Yuxiang's Rescue and the End of the Siege 23 5.4. Casualties and the Affects on Local Society 24 5.5. The Guominjun Motive in the Siege 25 5.6. The Allies' Motives in the Siege.. 26 6. The Nature of the Hybrid Military System 27 6.1. Regional Issues Concerning the Siege 28 7. Conclusion 29 8. Biblography 32 Appendix I Tables 38 Appendix II Maps and Figures 44 iv List of Tables Appendix I Table 1. Character List 38 Table 2. A Chronology of Events in China, 1911-1928 40 Table 3. A list of Guominjun's Territory and its Commanders in Late 1924 42 Table 4. The Guominjun and its Comanders, 1925-26 42 Table 5. Guominjun's Commanders and Units in Xi'an, April-November 1926 43 Table 6. Clarification on the Chinese Term "guan" 43 List of Figures Appendix II Figure 1. Map of Shaanxi, 1920s 44 Figure 2. Warlord China, 1925 45 Figure 3. The City of Xi'an, 1930 46 Figure 4. Modern Xi'an: City Wall and Moat 47 Figure 5. Towers on top of Xi'an City Wall 48 vi 1. Introduction [In a celebrated episode from the historical f ict ion Romance of the Three Kingdoms, the master strategist Kongming (Zhuge Liang) used an ingenuous stratagem to discourage his opponent f rom attacking his skeleton force held up in a small-fortif ied town. Directing a complex retreat after the Jieting military disaster, he soon learned of the imminent arrival of a massive force led by his archrival Sima Yi . Wi th only a handful of t roops and officers with him, the situation seemed hopeless.] There was panic among the cadre as Kongming cl imbed the wall to scout the horizon. Dust was bil lowing up to the skies. Two columns of northmen were bearing murderously down on Hsi. Kongming ordered all f lags and banners put away. The commanders of the watchtowers were caut ioned, on their lives, against any unauthor ized movements or audible conversat ion. Then Kongming had the main gates thrown open and twenty men stat ioned at each to sweep and damp down the roadway, They were to appear oblivious when the northerners arr ived. Kongming decked himself in crane feathers, wound a white wrap on his head and, fol lowed by two lads and carrying a zither, braced himself upon the city wal l . Incense burned as he struck up the instrument. W h e n Sima Yi 's forward units came with sight of the scene that Kongming had contr ived, they were afraid to advance. Incredulous, Sima Yi ordered a halt and raced to the front. There indeed was Kongming, seated on the upper wal l , appear ing palpably amused as he strummed his zi ther amid the incense. The lad to the left held his sword the lad to the right the yak-tail . In and around the gates, twenty-odd vi l lagers concentrated on their sweeping as if no one were near. 1 Romance of the Three Kingdoms In the fictional account, Sima Yi hastily withdrew his force, fear ing that Kongming had sprung a gigantic trap. In reality, no general would dare employ his celebrated stratagem of reverse psychology to defuse a siege. A siege is considered a very risky affair because it pitches two sides into a prolonged attrit ion campaign. To breach a fortif ied city with motivated defenders, the besiegers need to amass overwhelming manpower and resources. Still there can be no guarantee of victory as the besiegers may exhaust themselves trying to overcome the fortif ications. The introduction of gunpowder radically altered tradit ional siege craft. Once the symbol of feudal power and omnipotence, fortif ications such as city wal ls and castles became obsolete in the face of heavy caliber artillery. Yet the 1926 Siege of Xi 'an in 1926 showed that Western military 1 Kuan-chung, Lo. Three Kingdoms: China's Epic Drama Translated & ed. Moss Roberts. (New York: Pantheon Books, 1976), 288. 1 experience on siege warfare does not necessarily fit the circumstances of the Warlord period (1916-28 C E ) . T he Siege of Xi'an was part of an anti-Guominjun campaign between Feng Yuxiang and the alliance of Zhang Zuolin, W u Peifu and Y a n Xishan. Lasting from late 1925 to early 1927, the campaign raged across North China. T h e star of this period, Feng, was chased out of his enclave in Zhili (Hebei province) and his forces pushed to the brink of collapse. T h e war culminated in the dramatic Xi'an siege from April to November 1926. Isolated in the ancient fortress of Xi'an, a small Guominjun garrison headed by Li Yunlong and Y a n g Hucheng fought against a massive force led by allied general Liu Z h e n h u a . 2 W h e n the garrison faced imminent defeat in November, Feng Yuxiang brought the Guominjun back to life. Together with the National Revolutionary Army of the Guomindang (GMD), the Guominjun lifted the siege and broke the strongholds of the allies in North China. Subsequently, the G M D established a national government in Nanjing in 1928 and the Warlord period ended. Despite of the fact that this major campaign drained the resources of the Northern warlords and helped to expedite the G M D ' s Northern Expedition, it has received no coverage in either popular or academic circles. Perhaps this is due to the fact that it neither involved the dramatic political struggles between the Guomindang and the Chinese Communist Party ( C C P ) nor commanded the same pivotal military impacts as the Northern Expedition or the Xi'an Incident (December 1936). T h e Siege of Xi'an contradicted Western military experience on siege craft. T h e primitive fortifications in Xi'an should preclude the defenders any chance of success against an invading army armed with modern weapons. Though mighty city walls and elaborate networks of moats were known to crumble in the face of artillery, this siege showed the fallacy of this assumption. 3 T h e battle demonstrated the transitional nature of the Chinese military system. It was characterized by a hybrid warring style where contemporary Western experiences in military organization and weaponry were fused 2 In fact, the name Guominjun was disused by 1926. For the sake of clarity, I will continue to designate units associated with Feng as the Guominjun. 3 One of the most dramatic examples of the power of artillery over ancient fortification was Charles VIFs whirlwind campaign in Northern Italy in the spring of 1494. Even at this early stage, the forty artillery pieces Charles brought with him destroyed scores of medieval fortifications that used to take years to subjugate in a matter of months. John Keegan. A History of Warfare. (New York: Vintage Books, 1993), 321. 2 with the traditional Chinese wisdom in tactics and politics. T h e siege showed the immature status of the Chinese military evolution where the "old ways" of Chinese warfare retained their influence despite two decades of wholesale military reorganization along Western line. The siege also brought out salient regional issues. Regional feuding was primarily a result of traditional animosity and the "guest army" (kejun) phenomenon between Shaanxi and Henan. They played an important role in deciding the mindset of the actors involved in the siege. Sieges in Romance of the Three Kingdoms tend to conjure up romantic images where ingenuous stratagem and surreal heroism overshadowed the reality of suffering. In the context of Chinese Warlordism, the feats of the valiant generals were exaggerated to the extend that the mundane suffering of the people were completely ignored. Thus, it is very difficult to ascertain the social impact of the siege. These factors must be considered in analyzing the historical significance of this struggle. This study seeks to ascertain the historical significance of the Xi'an siege during the Warlord period. The siege represented the end point of a much-neglected era in modern Chinese scholarship. It demonstrats the adaptive nature of the warlord military system and the circumstances that brought it about. T o try to understand the result of the siege with a monocausal explanation will inevitably lead to a misinterpretation, for neither traditional Chinese politic nor Western military theory can adequately address the hybrid nature of the warlords as military and political entities. This study is divided into three parts: Part One discusses the historical and geopolitical importance of Xi'an; Part Two briefly summarizes the anti-Guominjun campaign, the warlords' military dispositions and their strength and weakness; Part Three details the eight months' siege. While numerous Chinese and Western sources has analyzed the campaign, very give details on the actual siege. Part Three is reconstructed from limited information derived from foreign news correspondence, secondary sources, the biography of Feng Yuxiang and Y a n Xishan. 2. Summary of the Xi'an Siege T h e Xi'an siege was the result of the anti-Guominjun campaign instigated by Zhang Zuolin and W u Peifu in late 1925. T h e Guominjun was in its zenith of both 3 military and political power at this time. Its army was as large as 40,000 and its territory stretched from G a n s u to Zhi l i . 4 Yet it was overwhelmed when the two most powerful Northern warlords joined force. With the provinces of Zhili and Henan fallen to the allies, scattered contingents of Guominjun fled towards their remaining strongholds in Shaanxi and Ningxia. A s the most heavily fortified position in Guominjun's remaining sphere of control, Xi'an was the last citadel of the beleaguered army. If it were to fall, the Guominjun would be cut in half and easily destroyed by the allies. Facing a massive allies assault force of 100,000 were less than about 10,000 Guominjun soldiers in Xi'an. Completely isolated and lacking in all essential supplies, the garrison managed the astounding feat of holding the city for eight months. Unlike other great sieges of the 20 t h century, where mobile armor battles played an important role, it was characterized by artillery bombardment and assaults along the elaborate fortifications encasing Xi'an. T h e surreal juxtaposition of modern firepower and ancient defense structures ended when Feng emerged from his retirement and lifted the siege with the reconstituted Guominjun in November 1926. 5 Evoking the most dramatic moment in Romance of the three Kingdoms, the "knights in shining armors" relieved the garrison when failure was imminent. Sources estimate at least 20,000 civilian and military casualties inside the cities. There were no reliable estimates on the casualties of the besiegers or the population. T h e siege presented its own set of unique characteristics. T he city of Xi'an, with its ancient fortifications, enabled the defenders to rely upon "defense in height" in addition to "defense in depth". T h e concept of "defense in depth" had showed its merit in modern sieges such as Leningrad and Stalingrad. Trading space for time, the defenders would heavily fortify the surrounding countryside of a strategically important city. T h e purpose is to drain the besiegers of resources and manpower long before they are in a position to attack the city. Though the technique has been used for a long time, the 4 Li Maosheng & Luo Chunpu & Y a n g j i anzhong • $ t # i 3 ' If YanXishan Quanchu IWiWi |i|_H¥» (B iography of Y a n X ishan) . (Bei j ing: Dangdai Zhongguo Chubanshe , 1997),322-324. 5 Soon after the Beijing coup in late 1925, Feng announced his retirement. Ostensibly to study Buddhism in Ningxia, Feng later traveled to the Soviet Union to "observe" the progress of the socialist regime. Scholars have variously interpreted his real intension as to negotiate foreign aids or remove himself as a target of the Wu-Zhang alliance. James Sheridan. Chinese Warlord: the Career of Feng Yu-hsiang, (California: Stanford University Press, 1966), 185. 4 advent of artillery gives the defenders tremendous advantages in holding territory and deterring the enemy from storming the city. In the siege of Xi'an, "defense in depth" was further supplemented by a supposedly obsolete technique—"defense in height".6 By 1926, the city was fortified by a series of elaborate fortification. Experience in European siege warfare after the "Gunpowder Revolution" (1300-1500 C E ) had demonstrated that high walls could not deter an enemy equipped with artillery. 7 Yet, the ancient city wall and moat proved to be a decisive factor during the siege. 3. History of Xi'an Xi'an is the capital of the province Shaanxi (Shenxi). Historically known as Chang'an (long tranquility), the city was once the capital of twelve dynasties (from 11 t h century B C E to 9 t h century C E ) . 8 It is also known as the terminus of the famed Silk Road that linked Han China to Imperial Rome. Situated on the south shore of the W e i River, the city is the central point of the W e i River valley. Better known as the "Area within the Pass" (guanzhong), this region is surrounded on all sides by mountain ranges . 9 Its land was highly productive due to the loess soil deposited by the Wei River and the mountains formed a natural barrier that protected the region from outside incursions. In the traditional period where military and political power stemmed from strong agriculture, whoever controlled this region was in a position to rule C h i n a . 1 0 Agriculture in this region demanded effective water control for the loess soil had a tendency to clog the rivers. A s the seat of numerous dynasties, the valley was highly coveted and was the site of 6 Though by no means a new technique, "defense in depth" was used to great effect in both mobile armor and siege warfare in the 20th century (i.e.: the Germans adapted this technique in the Western front by 1916-17. It was adapted in response to the increase sophistication of modern weaponry (primarily machine guns and quick fire bolt-action rifles). Instead of relying on one defensive line (it can be the city wall or a series of heavily fortified positions, such as the Maginot line), the defenders will draw the assailants into the batUefield litters with mutually supportive strong points. The goal is to exhaust the assailants before they reach the target. Geoffrey Parker ed. Cambridge Illustrated History ofWarfare. (Cambridge: Cambridge University Press, 1995), 290. 7 Ibid, 92-119. 8 The twelve dynasties were Western Zhou, Qin, Western Han, Xin, Western Jin, Former Zhao, Former Qin, Western Wei, Northern Zhou, Sui, and Tang. Shaanxi Shifan Daxue Dilixi ed. W&ffl&eJZ^i&W&M- Xi 'ansi Dilizhi S ^ f ) 5 % S ^ (Records of Xi'an's Geography). (Shaanxi: Shaanxi Renmin Chubanshe, 1988), 6. 9 Refer to Table 6. 1 0 The Han historian Sima Qian described this region as "protected on all four sides by barrier of mountains and rivers, and the land is rich and fertile. This is the place to make your capital and rule as a dictator." Qian, Sima. Records of the Grand Historian. Han Dynasty 1. Translated by Burton Watson. Revised ed. (New York: Columbia University Press, 1993), 33. 5 repeated battles. T h e destruction wrought by repeated disasters and war made the inhabitants recalcitrant, belligerent to outsiders and highly independent-minded. 1 1 In the absence of modern transportation, population inside the valley relied on the W e i River and the imperial highway through the treacherous mountains to carry goods and people. Hence the mountain passes and the city formed strategic checkpoints, controlling access to the central plain. Historically, the city symbolized the division between the Chinese conception of civilization and barbarity. T h e region borders diverse environments. T o the west are the desert regions of G a n s u and Ningxia. Until they were brought under the Manchu control, these regions were considered as de facto barbarian territory. It was also where the Silk R o a d linked "civilized" China proper to the rest of the world. T o the north is the Central Asian steppe. T o the east and south is the flood plain of China proper. Considered as the core of Chinese civilization, this is where Xi'an and its bureaucratic machinery governed for a millennium. T o the southwest is the province of Sichuan. Known as tiangu ziguo (heaven country), the province has abundant natural resource and is one of the richest parts of the country. By 1926, Shaanxi was part of the Guominjun's territory. T h e complex geography made the province the hotbed of bandit gangs and petty warlords . 1 2 Despite its best effort, the Guominjun only retained nominal control over part of Shaanxi. The absence of industries and the dire poverty of the population made the province a poor candidate for a warlord's base of operation. A s the capital of Shaanxi, Xi'an was the hub of the province's transportation network and economic activities. Since the Longhai railway had not yet been completed, Xi'an relied on the indirect Beijing-Suiyuan railway linking the Beijing Metropolitan region, Shanxi, Shaanxi, Ningxia, and G a n s u together. 1 3 T h e linkage between Xi'an and North China is contingent upon its link through Shanxi. This 1 1 Elizabeth J. Perry. Rebels and Revolutionaries in North China. 1845-1945. (California: Stanford university Press, 1980), 11. 1 2 The White Wolf (Bai Lang) bandit gang used Shaanxi as a staging group to raid surrounding province in 1910s. 1 3 Ma Liqian, Wang Kaiji, Lu Yizhi, ed. H J H ^ 3£pSPf % P—jits^, & Zhongguo Tielu Jianzhu Biannian Jianshi: 1991-1981 i^WW^M.WtWs^^'k-- 1881-1981)) (Brief History of China's Railway Construction: 1881-1981). (Beijing: Xinghua Shudian, 1983), 186. 6 dependence became almost fatal when the tide of the war turned against the Guominjun in 1926. 1 4 4. Background to the Anti-Guominjun Campaign T he war between Feng Yuxiang's Guominjin and the alliance of Zhang Zuolin and W u Peifu was a result of the 2 n d Zhili-Fengtian war (1924 C E ) . 1 5 Though the war eliminated W u as a power player in North China and led to the rapid expansion of Zhang's sphere of influence inside China proper, it was also inconclusive. Far from deciding the fate of the North, it brought about a new period of instability as Feng clashed with Zhang for control of the Beijing Metropolitan region and Zhili province. Despite of Wu's early success in holding off Fengtian assaults, a combination of treachery, bad luck, and warlord politic conspired against h i m . 1 6 T h e most pivotal factor contributing to Wu's defeat was Feng's sudden about face and occupation of Beijing. The Beijing coup (Beijing Zhengbian) in October 23 1924 derailed Wu's entire campaign, cut off his main force in Shanhaiguan, and invited Zhang Zuolin to exploit his weakened left f lank. 1 7 He was routed and had to beat a hasty retreat to Hankou. T h e once powerful Zhili Clique disintegrated into competing factions and lost much of its military muscle. After the disaster in Northeastern China in 1924, W u eventually managed to put together a new domain in Hubei. 1 4 Refer to Figure 2. 1 5 Known also as the "Old Marshall" or the "Manchuria warlord", Zhang headed the Fengtian clique. He took control of Manchuria in 1910s and gradually turned it into his personal kingdom. He maintained one of the largest armies during the warlord era and had the resources and industries of "Fortress Manchuria" at his disposal. Many scholars considered him as a puppet of Imperial Japan. Zhang was forced out of Beijing in May by the GMD's Northern Expedition He was assassinated by Japanese agent on June 4. Ronald Suleski. Civil Government in Warlord China. (New York: Peter Lang, 2002), 1-31. 1 6 Fengtian is the traditional term that describes the regions of northwestern Inner Mongolia, western and southeastern Jilian and a large portion of Liaoning. Zhongguo Lishi Diming Cidian. (Nianchang: Jiangxi jiaoyu Chubanshe, 1998), 463. 1 7 After the coup d'etat, Feng declared his intention to seek peace. He imprisoned Cao Kun and ostentatiously invited Sun Yatsen to the capital to discuss an end to the political disunity. According to Sheridan, the mutual tension between Feng and Wu grew out of a series of clashes and disagreement as early as 1922. Feng did not expect to gain anything from Wu's victory. Instead he accepted a huge bribe from Zhang (provided by the Japanese and amounted to approximately 1,500,000 yen), took over the capital and soundly handed a humiliating defeat to his old master. Sheridan, 136-147; Arthur Waldron.Frow War to Nationalism: China's Turning Point, 1924-25. (New York: Cambridge University Press, 1995), 182-184. 7 4.1. Realignment of Power Structure A s Lucian Pye points out in his study of warlord politic, the structure of warlord relations existed in an uneasy balance of temporary alliance and deal making . 1 8 The rapid expansion of power and territory by one warlord would inevitable lead to reactions from others, often violently in order to forestall the emergence of a hegemon. By the conclusion of the 2 n d Zhili-Fengtian war, Zhang was on his way of claiming this honor. 4.2. Rise of Zhang Zuolin Zhang's domain possessed three attributes which enabled him to push outward aggressively: security, resources, and proximity to the patron. Manchuria bordered the Soviet Far East and Korea, neither of which was in a position to invade Zhang's territory. Strategic security enabled him to concentrate his force towards China proper. Secondly, Manchuria had abundant resources. This provided Zhang with a huge resource base to develop. By the time Z h a n g consolidate his control over the provinces in 1910s, Manchuria already had an agricultural base and industries. It also had well-developed infrastructure (railway, ports). The combination of industries (particularly the arsenals) and transport networks allowed Z h a n g to wage war with a secure supply b a s e . 1 9 Thirdly, Zhang Zuolin had the patronage of Imperial Japan. With their burgeoning industrialized economy, the Japanese had eyed Manchuria longingly for its agricultural and industrial resources since the Meiji period. By the time Zhang rose to power, the Japanese had invested significant financial and political resources in M a n c h u r i a . 2 0 Above all, the Japanese sought stability in order to exploit the resource. T o protect its interest during the Warlord period, Japan provided aids to placate the "Old Marshall". Militarily, Zhang benefited greatly from his proximity to 1 8 Pye, Lucian W. Warlord Politics: Conflict and Coalition in the Modernization of Republican China. (New York: Praeger Publisher, 1971), 9-11. 1 9 This is especially important given the embryonic nature of Feng's defense industry. Soviet observers pointed out that Feng had very limited access to heavy industry and resources. His so-called arsenals were in fact improvised workshops. They could not manufacture gunpowder and hardly manage to keep up with basic maintenance of equipment. Julie Lien-ying How, ed. "Soviet Advisers with the Kuominchun, 1925-1926: A Document Study." Chinese Studies in History XIX, no. 1-2, (Fall-Winter 1985-1986), 125-127. 2 0 The Japanese invested both money and blood in Manchuria. It wrestled control of southern Manchuria from Czarist Russia after several bloody battles in the Russo-Japanese war (1904-5). It founded the Southern Manchuria Railway Company that alone worth 200 million yen. Suleski, 27. 8 J a p a n . 2 1 In so far as Japanese interest was concerned, Zhang had leverage- he managed to link his own political survival to the interests of Japan. In the name of maintaining stability over Manchuria, Zhang extracted a great deal of material and financial aids from the J a p a n e s e . 2 2 W h e n combined with Manchuria intrinsic geographic and industrial advantage, Zhang was in a position to dominate China in late 1924. 4.3. Feng Yuxiang and the Expansion of Guominjun's Territories T h e biggest beneficiary of the war was the "Christian general" Feng Yuxiang. O n c e a subordinate of W u Peifu, Feng had by late 1924 carved out his domains and was a star among the Northern warlords. Unlike W u and Zhang, Feng's Christian heritage gave him an edge when dealing with foreigners (particularly Christian countries). 2 3 His demand for strict military discipline gave his forces a reputation that few warlord managed to achieve—an efficient armed force that looked after the welfare of the common people (laobaixing)24 In the aftermath of the Beijing coup, F e n g quickly expanded his army and territory. The National People's Army of China (Zhonghua Minguo Guominjun) was created on October 25 1924 and its numbers quickly swelled to around 400,000. Feng's sphere of control rapidly expanded from his base in the Beijing 2 1 In his study of the armament trade during the Warlord period, Anthony Chan points out that geography, regional infrastructure, and patron-client relationship all played a role in the warlords' ability to acquire weapons. Manchuria had developed industries. It was physically close to its major patron, and the ability to rapidly move the weapons from the ports to the army. This increased the combat effectiveness of Zhang's army and could potentially give him a deciding advantage over others, especially Feng Yuxiang. Anthony, Chan. Arming the Chinese: the Western Armaments Trade in Warlord China, 1920-1928. (Vancouver: University of British Columbia Press, 1982). 2 2 One of the most dramatic examples of Japanese aid was the huge bribe provide for Zhang during the 2nd Zhili-Fengtian war. Had Zhang Zuolin lost the war, Wu would control of Manchuria and possibly take over all of Japan's interests. This, coupled with Zhang's willingness to deal with them, gave the Japanese a powerful reason to get involved. Japan would again interfere on Zhang's behalf when it intercepted Feng's supply in Dagu in 1925. 2 3 Given his religious belief, Feng established very positive relationship with the foreign missionaries during his tenure as governors of various provinces and later as the commander of the Northwest Defense District. Foreign powers also judged him in more favourable light because of his Christian heritage and reputation as a "people's general". See Broomhall for a foreign missionary account and Stremski for the British reaction to Feng. The US Department of State also published a very short biographical sketch on Feng in 1948. It provides some insight on the Americans' perspective. Marshall Broomhall. General Feng: A Good Soldier of Christ Jesus. (London: China Inland Mission, 1923); Richard Stremski. "Britain and Warlordism in China: Relations with Feng Yu-Hsiang, 1921-1928". Journal of Oriental Studies (Hong Kong). 1973 11(1):91-106; US Department of State. Division of Biographic Information. "Biography of Feng Yu-Hsiang". (Washington, DC: Department of State, 1948). 2 4 According to missionary reports, Feng had made a positive impression to the local population while he was the military governor of Henan. He was well known for stressing the soldiers' role as the protectors of the people. NCH 143 (May22 1922), 661. Also see Sheridan, 122-123; "Document A, the Frist Kuominchun Army," Julie Lien-ying How, 107-108. 9 Metropolitan region to several provinces in North China . 2 1 5 While vast in scope, his territory was neither developed nor secure from rival incursions. Feng was in a perilous position in 1926. With his disciplined force and his control over Beijing, Feng stood head and shoulders above his peers in terms of combat power and political authority. T h e vast stretch of his domain put him in a position to strike and annex the rich regions of Manchuria, Jiangsu, and S i c h u a n . 2 6 But just as he could gain easy access to the rest of China, Feng's enemies could overwhelm his control over such a vast domain. His territory was fragmented into numerous spheres scattered in several provinces and regions, none of which Feng controlled completely. Since the transport infrastructure was underdeveloped, Feng relied on one railway link—the Beijing-Suiyuan l i n e - to maintain the flow of supplies and connecting the various spheres. Unfortunately, the strategic province of Shanxi bisected the vital link. W h e n its capricious overlord, Y a n Xishan, turned against the Guominjun in 1925, he crippled its ability to maintain an uninterrupted line of communication during crucial months in 1926. In spite of the fact that Feng had significantly expanded the Guominjun since the coup, the forces were largely uncoordinated and poorly equipped. Except for the 1 s t Guominjun, which Feng trained and commanded personally, his armies were composed of soldiers of very diverse backgrounds, many of whom were bandits or soldiers incorporated from the defeated warlords. 2 7 Indiscriminate recruitment enabled the Guominjun to rapidly expand its ranks, but at the sacrifice of military discipline, training and combat effectiveness. 2 8 Many of these soldiers never underwent the training stipulated by Feng and had little respect for the ideal of the Guominjun. In fact, their 2 5 Feng's sphere of control included Gansu, Inner Mongolia, Shaanxi, Henan, and the Beijing Metropolitan region. 2 6 The term Manchuria denotes the Northeastern provinces of Heilongjiang, Jilin, and Liaoning 2 7 At this time, the 2nd army was under Wu Jingyin and the 3rd army under Sun Yue. These two armies were never known for its tight discipline or military effectiveness. This got worst after the collapse of the Zhili-Shandong theatre. Most of the troops eventually ended up in Loyang. Along the way it picked up remnant of different forces (including bandits). The lack of discipline and disruption to local communities caused uproars from both civilians and the Red Spear society. Zhang Baifeng & Li Zongyi ed. |lf@i$ ±£§. Beiyang Junfa: 1912-1928 { 4t WWM » (Northern Warlords). Vol. 2 (Hubei: Wuhan Chubanshe), 1990. Also see chapters 2 & 3 of Billingsley's Bandits to under the origin of Henan and Shaanxi animosity during the Warlord period. While they were conducting "anti-banditry" activities, the Shaanxi warlords would purposely warned off local bandits so as to justify their continual existence in Henan. They would go as far as to attack the Henan Red Spear society and prevent them from terminating the bandits. Phil Billingsley. Bandits in Republican China. (California: Stanford University Press, 1988), 15-69. 2 8 Julie Lien-ying How, 184-198. 10 behaviors resembled more closely the enemies they sought to destroy.^ 9 With no time to impress upon them the espirit de corps the soldiers held little loyalty to the Guominjun. 4.4. Weaknesses of the Guominjun While vast in scope, Feng's territory was neither developed nor secure from rival incursions. In fact, his territory lacked all the elements that created "Fortress Manchuria": vulnerable to attack, dire poverty, and lack of outlet to foreign assistance. Feng's territory included some of China's most poverty-stricken regions. T h e four regions that made up the bulk of Guominjun's territory were scarcely populated and had no established financial or industrial base. Except for Shaanxi, Ningxia, Inner Mongolia and G a n s u were desert regions known for their pastoral life style. While Shaanxi benefited from the alluvial soil of the Yellow River, it also suffered from poverty induced by deforestation and the shifting of economic activities. Periodic flooding and the remoteness of the region discouraged foreign investment. Despite Feng's efforts to industrialize the regions, there were not enough infrastructures to support the Guominjun. Without a secure rear area or developed industrial base to replenish supplies and troops, the Guominjun lacked the resources with which to compete against the Manchurian warlord. Since there were no industries in which to extract revenue, the Guominjun relied heavily on taxation from the merchants and peasants. It forced local peasants to grow cash crops (such as opium) and disrupted local food production. 3 0 T h e 2 9 One foreign correspond wearily commented Sun Yue let lose his soldiers in Shaanxi. He "allowed the district to be bled white by his unpaid men. All the officers below the rank of colonel used to sally forth at night to collect their pay, as one man put it. Now they are commended to celebrate the 'virtues' of the man who openly connived at the highway robbery committee in his name. Strange times, when such men should control the destinies of the country." NCH, (Jan., 24, 1925), 137. 3 0 Regional food production was thrown into chaos when warlords imposed cash crops farming, such as opium and cotton, on the peasants in Shaanxi. With the already chaotic transport network and the limited rolling stock, food (if available) was unable to reach the people in need. Many regions experienced severe famine in 1924-25. Wang Jinxiang j E ^ l f - "Er, Sanshi Niandai Yapia Wenti" • H+^f^SJn fajgi" MinSU0 Dangan no.2 (1992): 71-76. Missionaries also reported that there was a man-made famine in Central China. Even though there were plenty of grain in Gansu and Suiyuan (Ningxia), no one could afford the exorbitant price to transport it. Warlords of the regions squeezed the population of revenue by setting up special tax and transport levies for intra-regional trade. The poor were suffering from warlord politics as they ate "leaves of the trees in lieu of food, and men [sold] their wives and children to save them from starvation". Missionaries pointed out that local cash crops and opium cultivation imposed by the warlords was a major cause of the famine. NCH 138 (Jan 15, 1921), 66 &141; 152 (July 26, 1924), 133. 11 transport network and the landlocked nature of Feng's territory nullified his effort to court foreign support. W h e n his foothold in Tianjin was taken by Fengtian troops in 1926, Feng lost his only seaport and access to the majority of the foreign powers. Without an outlet to the coast, he had no choice but to court support from the Soviet Union. But flirting with the "Reds" also created problems of its own. In addition to the difficulties of transporting the aid to Guominjun headquarters in Zhangjiakou, Feng also had to content with increasing Soviet interference in the military (on issues of political indoctrination and training) and foreign hostility. In spite of his Christian heritage, the close association with "Red" Russ ia aroused fear among the Western powers that the "Christian" general was turning "Red". 3 1 He faced the difficulties of appeasing his Soviet patron while minimizing their ideological encroachment and the danger of active foreign. 3 2 Furthermore, Feng's involvement in the G u o Songling rebellion destroyed what little chance he might have in peaceful consolidation. This opened the prospect of a two-front war against both W u Peifu and Zhang Zuolin. 4.5. Feng Yuxiang and Conspiracies T h e tension and competing interests would likely draw the two sides into open conflict, but Feng involvement in Guo's rebellion hastened the inevitable. 3 3 Apparently negotiated in secret between Feng and one of Zhang's crack generals, G u o Songling, the conspirators planned to take control of Manchuria and forced the "Old Marshall" to retire. Taking the designation 4 t h Guominjun, G u o launched a surprise attack against Zhang in November 1925. By December, G u o was in striking distance of Shenyang (Mukden), the capital of Zhang's warlord empire. After the shock subsided, Zhang rallied support among loyal troops and managed to stall Guo's assault outside of Shenyang. W h e n casualties began to mount, many of the rebel soldiers abandoned Guo's quest and returned to the Fengtian clique. G u o fled towards Zhili but was The British in particular was fearful. See Richard Stremski's "Britain and Warlordism in China" 3 2 In spite of Feng's best effort to insulate the Guominjun from the image of a "Red army", foreign powers did intervene on behalf of his enemies. For instance, Japan interception of weapons shipment in Dagu bounded to Feng's territory. 3 3 According, Zhang had indicated to foreign diplomats as early as March 1925 that he planned to attack the Guominjun very soon. In May, he sent troops to Beijing, overtly to protect the capital. But Feng pulled out the Guominjun from the capital and avoided a confrontation at the time. Sheridan, 179; Jian Youwen W^5L~!>C-Feng Yuxiangchuan OJ§3Eiji¥'PS)) (Biography of Feng Yuxiang). Vol. l&2(Taipei: Chuanji Wenxue Zazhishe, 1982). 12 captured and executed. Feng belatedly launched attacks against Fengtian positions in Zhili. He subsequently wrestled control of Tianjin from Fengtian's general Li J ing l in . 3 4 Feng then brought the war to Fengtian controlled S h a n d o n g . 3 5 After some initial success, Feng found himself facing a determined Fengtian resistance in early 1926. His action also provoked the cooperation of two former archrivals: W u Peifu and Zhang Zuolin. In spite of his intention of retiring from politic and avoiding a cataclysmic conflict between the Guominjun and the emerging alliance, W u and Zhang agreed that Feng's elimination overrode their long time mutual animosity. 3 6 4.6. Anti-Guominjun Campaign T h e Guominjun faced the nightmare of defending several frontiers with scattered, under-equipped, and under-trained forces. Internally, it also had to deal with a fragmented command structure as a result of Feng's retirement. Though the Guominjun was usually associated with the charismatic Feng Yuxiang, in fact it was composed of three armies. Commanded by his subordinates W u Jingyin and Sun Yue, Feng had no direct authority over the operation of the 2 n d and 3 r d armies. T h e lack of a unified command became a serious obstacle to the operation of the Guominjun by 1926. At the onset of the war in early 1926 the Guominjun was scattered into six different spheres . 3 7 With Zhang Zhijiang in command in Zhangjiakou (Kalgan), the Guominjun set out defensive preparations around several key strategic cities: Duolun in Inner Mongolia, Zhangjiakou-Nankou in Zhili, Loyang in Henan, Xi'an in Shaanxi, and Lanzhou in G a n s u . 3 8 Despite the lack of supplies and coordination, the Guominjun managed to hold on to the fortress cities. T he W u - Z h a n g alliance launched a three-prong offensive against various sectors of Guominjun territories: Li Jinglin and Zhang Zongchang attacked Guominjun's 3 4 Incidentally, one of the original conspirators against Zhang was Li Jinglin. According to Sheridan, Li sudden about face was due to fear that Zhang might harm his mother. Li fled his domain in Zhili and joined Zhang Zongchuan in Shandong. Sheridan, 182 &184; Li Maosheng & Luo Chunpu & Yang jianzhong, 321. 3 5 The second Guominjun launched its Shandong offensive in mid-November 1925. 3 6 Sheridan 185; Wilbur, Martin C. & How, Julie Lien-ying, ed. Document on Communism, Nationalism, and Soviet Advisers in China: 1918-1927. (New York: Columbia University Press, 1956), 333. 3 7 They were the Beijing Metropolitan area, Shandong, Henan, Zhangjiakuo-Inner Mongolia (Rehe and Charhar), Shaanxi-Ningxia (Suiyuan), and Gansu. 3 8 Refer to Figure 2. 13 positions in Shandong and Southern Z h i l i 3 9 . Zhang Zuolin entered China proper and attacked Guominjun's strongholds in Zhangjiakou-Nankou. Moving from his base in Hankou along the Beijing-Hankou railway, W u Peifu rendezvoused with Fengtian troops in Zhili and launched a general offensive. 4 0 T o further disrupt the internal operation of the Guominjun, the allies sought support from the Shanxi warlord Y a n Xishan. He ' entered the war in March and launched a blitzkrieg along the Beijing-Hankou railway. Quickly occupying Shijiazhuang, Y a n proceeded to occupy key strategic cities along the length of the railway and opened the path for W u Peifu's invading force. 4 1 By April, Y a n had massed an assault force around Datong in Northern Shanxi. Responding to Zhang's call for support, he sought to launch a flanking attack against Zhangjiakou and clear Guominjun's positions along the way. The Guominjun strategy in the early 1926 was twofold: It maintained a defensive posture in both Inner Mongolia and Zhili while aggressively pushing into Northern Shanxi. Perceiving the danger of being permanently cut off, it sought to counter Yan's threat by occupying Shanxi and reopening the Beijing-Suiyuan railway. Difficulty of the terrain as well as internal dissention within the allies hampered their efforts to subjugate the 1 s t Guominjun in the Zhangjiakou-Nankou region 4 2 Nevertheless, the allied general offensive forced the Guominjun to pull out from Beijing by April 15. Five days later, Fengtian took control of Beij ing. 4 3 3 9 Wu Peifu made himself the commander of a fourteen provinces anti-banditry alliance on October 21 1925. Soon after he responded to Zhang's overture and agreed to a Wu-Zhang alliance against Feng Yuxiang. Before Lin joined Zhang Zuolin, he actually formed a tripartite alliance with Feng Yuxiang and Guo Songling in late December 1925. When Guo rebelled, Feng attacked Rene and Tianjin, Li decided to flee Zhili and joined stay with Zhang. Subsequently, he withdrew to Shandong and joined Zhang Zongchang. Li Maosheng & Luo Chunpu & Yang Jianzhong, 321. 4 0 Zhang Zuolin used the excuse of pursuing the remnants of Guo Songling's rebels and entered China proper., 4 1 Yan proceeded as far north as Baoding. According to his biography, Yan was at odd with the Guominjun over the control of key strategic cities along the two railway arteries. Yan suspected Feng's duplicity and felt the Guominjun garrisons around his domain constituted a direct threat to his power. Hence, he decided to heave Wu's call and preempted the Guominjun. Refer to Figure 2 for the location of Shijiazhuang. 4 2 Ibid, 326. Due to disagreement over the appointment of the Beijing leadership, Zhang Zuolin stalled the Zhangjiakou offensive during March and April and hoped to use the Guominjun to restrain Wu Peifu. The 1st Guominjun took this opportunity and shifted the bulk of its force to Ningxia in preparation of their Northern Shanxi offensive. Nankou was garrisoned by the 6th division under Liu Ruming and Zhang Wanqing with 16,000 men. In contrast, the combined strength of Yan and the Shandong-Hubei assault force numbered 500,000. Jian Youwen, 248. 4 3 Throughout the campaign, the allies used their small air force to harass the Guominjun. While their military effective was very limited, the bombing caused profound psychological shock to the population. New York Times (Apr. 13, 1926), 6. Also see Anthony Chan's article on the warlords' air force. Anthony Chan. "The Chinese Air Force During the Warlord Era, 1916-1928"^ 4rmy Quarterly and Defense journal (Great Britain). 1983 113(1): 81-89. 14 Under Song Zheyuan command, the Guominjun assault force was initially successful in pushing back the Shanxi army to Datong. But renewed efforts by the allied steadily eroded the defense of Zhangjiakou and forced the Guominjun to abandon its Shanxi operation. By August, Zhang Zhijiang ordered the evacuation of Guominjun from Zhili and Inner Mongolia to Wuyuan in Ningxia. Elements of the 2 n d Guominjun under Y u e Weijun retreated south to Loyang along the Beijing-Hankou railway. 4 4 Unfortunately, the undisciplined troops wrecked havoc along their path of retreat and led to popular uproar. Though he managed to hold out in Loyang, a combined force of the allied troops and the Henan Red Spear Society later captured Y u e and his troops . 4 5 5. The Siege of Xi'an With the Eastern fronts collapsing and the Guominjun fleeing in disarray, the Western spheres of the Guominjun's territory became the last citadel against the allies. Throughout the campaign, Shaanxi held the key in anchoring the scattered Guominjun's positions. Shaanxi was the strategic high ground since it provided the vital link to the Soviet Union as well as for the Guominjun throughout North China. In July 1925, the 1 s t mixed division under X u Yongchang gained partial control of Shaanxi and proceeded to systematically eradicate units associated with W u Pe i fu . 4 6 W h e n the Eastern theater was collapsing, the local garrison faced the impossible task of defending the province and safeguarding a path for the retreating units. T he opposing force centered on the former Shaanxi civil governor Liu Zhenhua. His assault force numbered around 100,000 and was lavishly equipped by both Zhang and W u . His force broke through Guominjun defense in Tongguan by mid-Apri l . 4 7 At this time, Liu was highly optimistic that he would finish off the Guominjun quickly. Xi'an was 4 4 Wu Jingyi died of illness on April 13th, 1925. Yue Weijun, his subordinate, took over the leadership. However, Yue did not possess nearly enough authority to lead the 3rd army. 4 5 The Red Spear Society was a local self-help organization originally created to resist pressure of government bureaucrats and local gentry for taxation. By this time, it had evolved into a quasi-militia resisting the warlords. 4 6 1st mixed division of the 2nd Guominjun was sent to re-conquer Shaanxi in July 1925. Under the leadership of Xu and Li Yunlong, the unit managed to impose Guominjun control over the southern part of the strategic cities of Shaanxi and took Xi'an by July 16th 1925. 4 7 In addition to the Zhenguojun and reinforcement from Wu Peifu's army, Liu also recruited scattered group of warlord units in Shaanxi not associated with the Guominjun. (Such as Ma Zhenwu & Gou Baojie). Mi Zanchen Yang Hucheng Jiangjunchuan i%^MMW-W^ (Biography of General Yang Hucheng). (Beijing: Zhongguo wenshi Chubanshe, 1986), 20. 15 defended merely by about 5000 under-equipped Guominjun soldiers. Furthermore, a number of gentry inside the city created a "Hope for Peace Society" (Heping Qichenhui) and were openly calling for the Guominjun to leave the city. With the Guominjun garrison vacillating between war or peace, Liu had reason to believe that his campaign would be short and decisive. Described by numerous Chinese sources as a "running dog" of Warlordism and a corrupt warlord who sucked the province dry, he was the head of a small warlord army that had been active in Shaanxi before the 1911 Revolut ion. 4 8 A holder of first-level degree [Shengyuan) before the Revolution, Liu became a bandit when the ancien regime collapsed in 1911. Based in the Mount S o n g region of Henan, his army, Zhenguojun, recruited among the destitute peasants, like-minded desperados and even remnants of the famed "White W o l f bandits. Liu was the civil governor of Shaanxi in the early 1920s and had acquired a bad reputation among the population. He was seen as an opportunist and was unscrupulous in his act ions . 4 9 T h e Zhenguojun blurred the distinction between outlaws and soldiers as it behaved no differently from a rampaging bandit g a n g . 5 0 Before the outbreak of the anti-Guominjun war, Liu was recruited by W u Peifu to head the Shaanxi campaign. T o strengthen the Zhenhuajun, Liu also recruited members of the Henan Red Spear Society, many of whom welcomed the chance to extract vigilante justice to past Shaanxi transgressions of the Guominjun. 5 1 O n e of their 4 8 The portrayal of various warlords in the literature is an interesting topic. Chinese have a long history of framing military figures in accordance with different historical and literary motifs. Even to the unschooled, these stereotypes instantly convey special messages and meaning. The "Jade Marshall" Wu Peifu, for example, fitted the motif of a capable, cultured scholar general. Following the examples of a long line of tragic heroes (both real or semi-fictionalized), he was betrayed and fell from grace. Liu, on the other hand, represented the worst type of villains populated in Chinese history and literature. Though educated and cultured in the Confucian order, he chose the outlaw profession and indulged in vices and corruption. In reality, both were selfish militarists worked for personal gains and benefits. 4 9 Before he was transferred out of the province, there were indications that Liu had amassed a huge personal fortune. It was alleged that he took with him $12,000,000 when he left Shaanxi in March 1925. His Henan soldiers were known to ransack the people. This contributed to the long running animosity between Henan and Shaanxi. NCH (March 30, 1925), 86. 5 0 The incompetence of the Zhenguojun could be seen in an earlier skirmish in Henan between one's of Liu subordinate and the Guominjun under Wu Jingyiu. Mi Zanchen, 20. 5 1 When Feng Yuxiang was made governor of Henan, he had brought with him detachment of Shaanxi units. Many of the Shaanxiness made it a sport to publicly humiliate and even execute the Henan residents for amazement. In addition, there was also the issue of taxation imposed by the "Christian" warlord. Many Henanese saw the siege as a chance to draw Shaanxi blood. Zhang Yungjia 3H|!f HC. Yu You fen Zhuan { ^f-tiUM ) (Biography of Yu Youren). (Taibei: Zhongwai Tongxunshe, 1958), 90-91; Sheridan, 207. 16 leaders went as far as to declare "he will not allow a person above 10 years old [within the city] to escape alive". 5 2 Supplied by its Soviet patron and from a variety of foreign arms merchants, the garrison in Xi'an had in its possession a great variety of European manufactured a r m s . 5 3 In additional to small arms of different types, the defenders possessed artilleries and trench mortars of different calibers. T h e opposing force's main advantage was in supplies. Despite the overwhelming number Liu Zhenhua managed to put together, a great many of his troops were militia and bandits in origin. His own private army was not known to be disciplined or militarily effective. Liu also had accessed to a limited number of airplanes from Wu's force as well as a wild assortment of small arms, artillery and cavalry . 5 4 In sharp contrast to Liu, the defenders of Xi'an were uniformly hailed by Chinese sources as saviors and heroes overcoming insurmountable odds from a ruthless enemy. 5 5 Given the fact that Yang Hucheng was involved in the defense, the historiography tends to heavily emphasize his integrity, loyalty to the people, and nationalism. 5 6 In some ways similar to Liu Zhenhua, Y a n g also had a special affinity with the province. Born in a small village in 1893, Y a n g fought for the "Han Revival Militia of Shaanxi" (Qinlong FuHanjun, part of the elder brother society) during the 1911 Revolution. By the summer of 1925, he was the commander of the 3 r d division of the 3 r d Guominjun and was tasked to exterminate the petty warlord W u Xingtian. W h e n he learned of Liu's victory in Tongguan, Y a n g abandoned his mission in Western Shaanxi (around Baoji) and fell back to Xi'an to reinforce the garrison. y2NCH, (Oct. 23, 1926), 155. 5 3 Anthony Chan points out that Warlord China became the "dumping ground" of European surplus munitions, "[reports] that the warlords possessed inferior and outdated arms did not consider which munitions were available to them. Because China needed arms, it became a dumping ground for the Western surplus.... Disposing of munitions that were anachronistic by Western standard, but modern in China therefore sometimes give rise to dishonesty and chicanery on the part of the sellers". Anthony Chan, 78; Julie Lien-ying How, 122-123, 186-187191-192. 5 4 By the middle of 1926, Wu had to contend with the growing GMD threat from the south and could not spare any troops. 5 5 See the Biography of Yu Youren for an example. Chinese sources tend to give stock portray of the warlords. One was either a hero of the people (Feng Yuxiang or Yang Hucheng) or a puppet of foreign power (Zhang Zuolin). 5 6 Yang was best known for his direct involvement in the Xi'an Incident in 1936. Together with fellow conspirator Zhang Xueliang, they imprisoned Jiang Jieshi and facilitated the 2nd United Front. Tragically, he was burnt to death with his family aboard a ship in the post-war period. The tragic ending further perpetuates his legacy as a hero. 17 Y a n g brought with him the 3 r d division (with 5000 soldiers) and hastily entered Xi'an in April 18 t h. W h e n Liu learned that Y a n g Hucheng had withdrawn to the city, he immediately quickened the pace of the Zhenguojun. At this time, the Zhenguojun had penetrated Tongguan and occupied keys cities east of Xi'an. Liu ordered the forward elements in the vicinity of Northeastern Xi'an to launch a series of vicious assaults, aiming to breach the defense before Y a n g could catch his breath. Outlying defense were forced to pull back to the city and Yang's relief force had to join the fight before they had a chance to rest. W h e n the Guominjun finally repulsed the assaults, Liu resolved on a long attrition campaign to starve the garrison into submission. Breastworks were dug around the city and he harassed the Guominjun with incessant bombardment and weekly assault sorties along the wal l . 5 7 In the first few weeks of the siege, the city's West Gate remained open and supplies were brought in daily. Foreign missionaries stationed in Xi'an acted as peace envoys and sought (with limited success) to evacuate the civilian population. By June, Liu had isolated the defenders by occupying strategic towns around X i ' a n . 5 8 Vicious bombardment by "guns of very considerable caliber" rattled the population and assault along fixed defensive fortifications characterized the early phrase of the siege. W h e n the telegraph service was severed, the city was completely cut off. 5 9 At this time, the besiegers around the city numbered around 70,000. 6 0 By July, the momentum of the siege faltered and Liu concocted daring tactics to finish the Guominjun. Specially created "dare to die" commando units (gansiduF) would cross the moat at night and tried to scale the city walls with "cloud ladders". 6 1 Those who managed to cross the moat faced murderous defensive fire (machine gun, mortar, 5 7 More than 70 /;' of trenches and breastwork were dug. Jia Pingwa JPP [HI. Lao Xi 'an: Feidu Xieyang { ^ H ^ c : Jgtfl?4E§>> (Old Xi'an). (Nanjing Shi: Jiangsu Meishu Chubanshe, 1999), 96. 5 8 By May 15th, Liu's forces occupied the key town of San Qiaozhen between Xi'an and Xianyang. Xi'an was completely cut off. By the end of May, the besiegers controlled an area that stretched from Lantian south of Xi'an to Xianyang and Wugong to the west, Sanyuan to the north, and Tongguan and Lintong to the east. Mi Zanchen, 22. 5 9 According to eyewitness accounts, the besiegers would intermittently allowed peasants and foreigners to commute in and out of the city during the early days of the siege. As the casualties mounted and the siege dragged into the summer, the besiegers tightened their grip and not even foreigners were spared the hardship. NCH (Aug. 7, 1926), 247. 6 0 Mi Zanchen, 21-22. 6 1 Jian Youwen, 21. 18 artillery, and grenade). Rewards were promised to those who scaled the wall. 1" Other traditional siege tactics such as tunneling were also attempted near the East Gate. With the proximity of the two sides around the city, it was not uncommon for close-quarter, hand-to-hand battles to break out . 6 3 Liu also employed more subtle methods to destabilize the unity of the defenders. While the allied air force conducted bombing sorties against the garrison, propaganda leaflets were also dropped, urging the population to turn against the defenders. A price was put on Y a n g and Li's h e a d s . 6 4 T h e surrounding countryside of Xi'an was burnt and plundered by Liu's force . 6 5 5.1. The Defense of Xi'an T h e defenders of Xi'an faced several problems. Until May, there was no established command structure among the various units. Before Y a n g Hucheng's reinforcement arrive in April, the 5000 Xi'an garrison fell under the command of Li Yunlong and W e i D i n g . 6 6 W h e n Y a n g arrived, the three of them decided to cooperate in the face of imminent danger. The city was divided into three different defense zones: Y a n g was responsible for the north and eastern sectors (where Liu concentrated most of his attacks), Li for the south, and W e i for the west. T h e designation "Guominjun" was abandoned in favour of the more generic "Shaanxi army". This rallied the population to the defenders' cause against the Henan invaders. It also helped to reaffirm the common regional identity shared by the Guominjun and the population. At the outbreak of the siege, the population of Xi'an was around 500,000. 6 7 W h e n it became apparent that the city would be the next battleground between the allies and the Guominjun, people rushed to the safety of Xi'an. This added more burdens to the city's supply system as it 6 2 Ibid, 22. He put up rewards for: 1000 dollars for the first to scale the wall, 800 for the second, 500 for the third. Such tactics reminiscence the stratagems in Romance. 6 3 Some of the most ferocious battles were the assaults near Xiaoyanta in the northeast corner of the city and Liu's attempts to breakout in northwestern part of the city. Mi Zanchen, 23. 6 4 100,000 and 50,000 dollars respectively for Yang and Li's heads. 6 5 Chinese sources indicate that Liu had burnt tens of thousand of acres of cultivated land surrounding the city. Though it was not uncommon for the defenders to burn down the countryside and deny the enemy of shelter or supply. Jia Pingwa, 97. 6 6 2nd army 10th division was under Li Yunlong commanded; 4th division under Wei Ding. 6 7 According to the Xi'an record, there was no reliable census until 1938. Estimate of the city's population (permanent and transient) fluctuation from 20,000 to 650,000. Xi'anshi Difangzhi Bianzuan Waiyuanhui H^Tff i f t J S f i l t l t (Editorial Board for the Gazetteer of the City of Xian). Xian Shizhi Mf^rrJiS (Gazette of Xian) Vol.1 (Xi'an: Xi'an Chubanshe, 1996), 445. 19 relied on the surrounding counties for food supply. Hunger became more acute as the city's supplies dwindled. In desperation, the people ate anything that would fill their stomach, including bark and fertilizers. T h e military instituted rationing and commandeered the civilians' store of food. Foreign missionaries contributed to the defense with their supplies, medical expertise, and willingness to shelter the population. While the siege raged, the defenders also had to contend with the potential for internal dissent—gentry of the "Hope for Peace Society" continued to call for the Guominjun to leave the city. 6 8 Y a n g eventually reined in the movement by rounding up and executing their most prominent members . 6 9 T o reaffirm the Guominjun's commitment to defend the city, Y a n g and Li made the theatrical gesture of proclaiming their intentions to commit suicide in the bell tower should the besiegers breach the city's defense. By and large, the Guominjun managed to weather the assaults during the first few months of the siege. However, the reality of the siege snapped the early optimism of the defenders. Starvation, exhaustion from combat, lack of supplies and mounting casualties all took their tolls. Unable to communicate with the relief force, the passive defenders increasingly relied on the fortifications. 7 0 W h e n their early optimism evaporated, the fear among the people was almost palpable. With the prodigious usage of ammunition, the defenders ran the risk of running out of supplies. T h e situation became so desperate that the defenders utilized any and all means to forestall an imminent collapse, "when the ammunition is gone, we continue to use machetes, stones, our flesh and blood as weapons; when our food is gone, we ate dogs, horses, oil, and grass". 7 1 Two eyewitness accounts told of the suffering of the population: Corpses were everywhere in the city. Dogs ate many of the bodies as we could not clear them off the street fast enough. Some that have been starved for days simply lay beside the road. They too became the targets of dogs. Near the gate of Duanlu, I once saw a stricken old woman 6 8 Mi Zanchen, 24. 6 9 Ibid, 25. 7 0 Guominjun also recruited local Red Spear members to booster its ranks. Foreign accounts testified that they used the same bag of tricks as the Boxers once did in the face of overwhelming firepower—they claimed special rituals had rendered them invulnerable to firearms. NCH, (Aug. 28, 1926), 391. 7 1 Wu Changyi %^kM. Qiangu Gongchen Yang Hucheng { "6" JTJEfi (Yang Hucheng: Heroic General of a Thousand Generations). (Beijing: Zhongguo Wenshi Chubanshe, 1993), 24. 20 attacked by a pack of hungry dogs even when she wasn't dead yet! T h e old woman tried to fend them off but was too weak to even lift her arms. I helped to fend the pack away, but one of her arms and legs were already torn off. November 12 t h, blizzard and heavy snow. There were few if any people one can see outside. 2000 died in that day alone. Next day, I walked around and saw more dead bodies than I can count, many of them lying beside the houses. I also saw a great many families wearing funeral clothes and conducting funeral processions in the main road. Some wore the prescribed white cotton garment but most wore their ragtag clothes, carrying away their filial duty. Had it not been the siege, one would frown upon the inappropriateness of their clothing. Among the dead, some still had not swallow their last month full of food; some looked as if they just fell down and were trying to get up; some were tangled together as if they were trying to keep each other warm in an embrace. Many were curled up in fetal position laying among the grass field, looked as if they were homeless people trying to have a good night sleep. Apparently there were a great many looks when one died. O f the dead, many were men and very few women; many were the elders and few children; many were labours and few of other occupations. It seems that genders and occupations were also determinants of one's chance of dying..After seeing so much death, I am starting to feel as if I am already in hell with the hungry ghosts . 7 2 Because the Guominjun did not evacuate the civilians before the siege, the population became a crippling burden. Price of grain during the first few months of the siege remained moderately c h e a p . 7 3 Thanks to foreign missionaries' efforts, as many as 60,000 people (including foreigners residing in Xi'an) managed to flee the city. But conditions continued to deteriorate as the siege dragged onto the winter. T h e social classes were radically altered: T h e poor had long since disappeared and the roads were littered with well-dressed normally prosperous men and women, whose bodies were not only unclaimed for burial, but were not even robbed. Theft ceased to serve a purpose, because food was the only thing worth having and that could not be bought. W h e n relief came the deaths were averaging 700 a 7 2 JiaPingwa, 101. 7 3 2 catties of grain or 4 catties of sugar cost $1; 1 tin of kerosene oil cost $6. NCH, (Oct. 23, 1926), 155. One cattie equals to 500 grams or 1.1 pounds; one picul equals to 100 catties or 133.3 pounds. From American Heritage dictionary of the English Language, 21 day. Those who escaped or survived were mummified in aspect, fleshless and black-faced. 7 4 By the end of the siege, grain reached a price of $ 1 2 0 0 a picul. All animals down to cats and dogs had long been devoured. There were indications that cannibalism was openly pract iced. 7 5 The survivors often looked pale and sickly. Their condition is unsurprising given the severe shortage of food and essential supplies. 5.2. Tactics of the Siege T h e Xi'an siege revolved around the conventional definition of a siege: to prevent the assailants from breaching a setline of defensive fortifications. T h e fighting along the wall conjured up images of traditional siege: catapults hurling stones into the city with men-of-arms struggling to storm the wall with ladder; defenders raining arrows, rocks, and boiling water at them. 7 6 The tactics employed by both sides showed a curious mixture of new and old styles. Machine guns, small arms, and artillery duels characterized the incorporation of Western military technologies into the traditional art of siege warfare. Breastwork, city walls, starvation, and storming tactics showed the persistence of the "old ways" of traditional warring style. For Y a n g Hucheng and his fellow commanders, Xi'an was the last citadel of the Guominjun. Failure to hold on to this fortress would not only cut off the scattered Guominjun's chance of retreat back to safety in Ningxia, but would very well caused the collapse of the Guominjun . 7 7 Xi'an served as a magnet for attracting allied troops and resources, tying up resources that would otherwise have been used to attack the G M D ' s Northern Expedition. ,H NCH, (Oct. 23, 1926), 184. 75 Ibid. Foreign corresponds attested that "human flesh was sold openly in the street". Description of the condition in city often accompanied by historical stereotypes that does not fully convey the reality of the situation. In numerous sources, the authors attest that "people began to cannibalize and exchange their sons, ate bark, leather shoes, and bean to fend off hunger. Corpses resulted from starvation lied the street. The survivors were skinny and pale, looking as if they were mortally sick". Given the prevalence of writers and historians to use stock phrases such as "chaihai yizi" causally, it is difficult to judge their validity. Jian Youwen, 258. 7 6 Bruce Allen Watson. Sieges: a Comparative Study.(London. Praeger, 1993), 1. 7 7 Xi'an was famous for its heavily fortification. Giving it away to Liu uncontested would not only deal a fatal blow to Guominjun morale, but would also open a avenue of attack by the alliance towards Gansu and Ningxia. 2 2 5.3. Feng Yuxiang's Rescue and the End of the Siege A s a result of the evacuation of Zhangjiakou in August, the Guominjun in Eastern China withdrew westward towards Ningxia. O n September 16, the Guominjun began to reconstitute its force when Feng Yuxiang returned from the Soviet Union. Bringing with him not only support from the "Reds", but also a renewed sense of mission to join the G M D and rid China of Warlordism. Feng was made the commander of the National People's Alliance Army of China. O n May 17, Feng officially re-established the Guominjun and issued a proclamation signaling his formal alliance with the G M D . Focused more on indoctrinating the military and rallying the public to the new political ideals, Feng sought to stir up the people's passion as a prelude to his campaign. T h e newly reorganized Guominjun had about 250-260,000 troops in the Gansu-Ningxia region. 7 8 His new strategy was to "consolidate G a n s u and relieve Shaanxi, ally with Shanxi and take over Henan". Two relief columns were sent to Shaanxi. One column under Sun Liangcheng moved from Lanzhou to Xianyang. A s one of a new generation of capable generals, Sun managed to force his way into Shaanxi despite adverse weather conditions and lack of suppl ies . 7 9 Of the 10,000 under-equipped soldiers he brought with him, only 2,000 exhausted men were left when he arrived at Xianyang in October. His force was supposedly an elite unit but continuous fighting and force match had exhausted it. Lacking mutual support from friendly forces, he was pushed back by Liu. Fearing that a tactical withdraw would very well deal a fatal blow to the morale of the Xi'an defenders and affect the Northern Expedition, Sun was forced to stand his ground and took on Liu's counteroffensive. 8 0 Another column under Liu Ruming and Sun Lianzhong moved through Ningxia and rendezvoused with Sun in late October. However, it was not until in the middle of November that they gathered enough strength to launch a coordinated offensive 8 1 Jian Youwen, 253 7 9 The relief forces had their own horror stories to tell. Many troops perished while force match through blizzard and terrible weather. Those who survived often had to keep on marching while enduring terrible blisters on their feet. 80 Zhang Baifeng & Li Zongyi, 517. 8 1 The relief forces of Liu Yuming and Sun Lianzhong had around 8000 men. Sun moved from Baotou through Ningxia; Liu from Ningxia to Xianyuan. They were to meet each other in Xianyang. Then Liu was to attack from the left flank, Sun Lianzhong from the right. The embattled Sun Liangcheng was to attack from Xianyang from the centre. 23 Quarrels within the rank over distribution of supply delayed the offensive further. 0 2 By November 26, the combined force of the two relief columns launched an all-out attack. 8 3 Liu Zhenhua and the Zhenguojun collapsed under the offensive and were chased out of Shaanxi. The siege was officially lifted in November 27. 5.4. Casualties and the Affects on Local Society T h e city and the surrounding counties were devastated. According to one estimate, the daily casualties of the civilians during the last month of the siege ranged from 500 to 1,000. Overall between 15,000 to 20,000 died during the siege; unfortunately, this figure only represented soldiers and civilians within the city killed as a result of direct enemy f ire . 8 4 There is no estimate on casualties induced by starvation or revenge killing. But considering the amount of suffering and destruction incurred during the eight months period, total casualties would likely excess 20,000. Observers suggested that it would take 20 years for Xi'an "to recover anything like its former prosperity". 8 5 "Burn, kill, rape, and rob; there was nothing they did not do" (Fen sha yin lue; wu suo bu zhi) were the norm under the circumstances. 8 6 But long before their affects manifested among the people, starvation as well as the stress of the war would have taken their tolls. It is difficult to gauge the degree to which the war impacted local society, but one can still make several general assumptions: Inevitably, the physical and psychological stresses would seep into family life. Bickering and abuses among family members would tear families apart long before one had to sell his sons and wives. With so many persons dead or dying, the fundamental human relationships (parents-children, spouses, friends, neighbors) that formed a society would break up. Those who were old enough might live with basic survival strategy (every person unto himself). The elderly, the sick, and the orphaned would die due to lack of care. While the city can be rebuilt 8 2 Sun Liangcheng arrived at Xianyang on the November 23rd, but reluctant on the part of the Gansu unit (Feng Hongkui) to re-supply Sun led to a delay and invited counterattack from Liu Zhenhua . Sources indicate that the defense would fall within 3 days had Feng's relief columns not lift the siege in November 28th. 8 3 Jian Youwen, 257. 84 NCH, (Dec. 31, 1926), 617. 8 5 ibid. 86Edward A. McCord. "Burn, Kill, Rape, and Rob: Military Atrocities, Warlordism, and Anti-Warlordism in Republican China". Diana Lary & Stephen MacKinnon ed. Scars of War: the Impact of Warfare on Modern China. (Vancouver: University of British Columbia, 2002), 26. 24 and physical injuries cured, the psychological traumas and "scars" inflicted upon society would take much longer to h e a l . 8 7 These sufferings were further exacerbated when the local merchants and population were forced to fund the Guominjun's coffer ($20,000,000). 8 8 Local observer had sarcastically speculated about W u Peifu's sinister motives in maintaining the siege for so long considering the Guominjun was comparatively unarmed: Such an experience is more terrible than a massacre, because it lacks its thrill, and this was particularly thrilless because the investing army lacked the courage or energy for a serious assault, while the defenders would not attempt a businesslike sortie. T he troops of the 2 n d and 3 r d [Guominjun] monopolized the food supplies of the city. G e n [Liu Zhenhua's] attacking force looted the entire countryside most thoroughly to kill spare time, so the populace without the wall endured nearly as much hardship as the residents within, the armies engaged, as usual, suffered the least. 8 9 5.5. The Guominjun Motive in the Siege Compared to the combined forces of the W u - Z h a n g alliance and Y a n Xishan's Shanxi troops, the Guominjun was heavily outnumbered and widely scattered. It had very limited access to industrial resources and had no secure rear supply base. It relied upon two precarious railway links to maintain the integrity of its force and territory. By all accounts, the Guominjun was a failing military system. Yet Li Yunlong and Y a n g Hucheng made it a flexible and effective instrument during the siege. They demonstrated an unusual degree of valor and initiatives in holding out Xi'an instead of fleeing from Liu. Had it fallen to Liu Zhenhua, much of the non-aligned warlords within the provinces might join him. So instead of dealing with Liu's force from the east, Li and Y a n g would have to fight anyone that were hostile to the Guominjun from all sides. Heroism aside, there were practical considerations that made the defenders stayed in 8 7 Ibid, the idea of a "scar of war" is from Lary and MacKinnon's study of military atrocities on Warlord China. 8 8 Ibid, the saying "there is no free lunch in this world" rings truth in the Warlord era. Military campaign was an expansive affair and the warlords must recuperate their expense, particularly to one that had no stable revenue stream. Hence, the traumatized Xi'an population was squeezed of what little they had. 8 9 A O / , no. 162, 184. 25 Xi'an. Judging from the rapid pace of Liu Zhenhua's advance in April, it is reasonable to assume that the defenders would be routed had they pulled out. Beside, Xi'an was the heaviest fortified positions in western Guominjun territory. The defenders stood a better chance of surviving in Xi'an than retreating to the deep desert of Gansu . Fear of retribution was another factor in motivating the Guominjun to defend Xi'an with such gusto even when surrender was the norm of warlord battles. T h e y could not hope for a merciful treatment from the besiegers because the Henan Red Spear militia was out for blood. Had the city become undefended, the Henanese would slaughter the Shaanxi population. Beyond the summer of 1926, the Guominjun fear of elimination by the allies was replaced by preoccupation of defending regional strong points from regional enemies (Shaanxi and Gansu) as well as consolidating their scattered and demoralized force. 5.6. The Allies' Motives in the Siege For the allies, the cooperation between Zhang Zuolin, W u Peifu, and Y a n in early 1926 gradually broke down. Once the Guominjun pulled out of their bases in Eastern China, the ties that bound the Northern warlords together dissipated. Zhang and Y a n Xishan gained new territories and were too busy consolidating their gains. They allowed the demoralized Guominjun to flee to G a n s u rather than launching a "hot pursue" across the vast stretch of desert. A s for W u , the campaign increasingly became a secondary concern once the G M D launched the Northern Expedition. Like the Guominjun, W u faced the difficulties of defending his territory. Since the 2 n d Zhili-Fengtian war, his forces had diminished significantly. In supporting Liu's long siege, he further drained the limited resources he possessed. W u was caught off-balance once the Northern Expedition began. By late August, he faced intensive pressure south of Hankou and was desperate for supplies and re inforcement 9 0 Since neither of his allies offered any concrete assistance, W u had to redirect his limited resources to counter the National Revolutionary Army. In turn, this diminished Liu's ability to break the city's defense. 9 0 Liaoningsheng Danganguan. ed. Zhonghua Minguo Shiziliao Conggao. Vol.3 (Beijing: Zhonghua Shuju, 1985), 96-99. 26 6. The Nature of the Hybrid Military System The fighting during the siege demonstrated a hybrid military system within the warlord military structure. T h e wholesale importation of modern military technologies form Europe and North America since the first Sino-Japanese war (1894-95 C E ) had in several respects radically changed the traditional ways of warfare. New industries had to be created in manufacturing and maintaining the myriad of modern weapons. New military management technologies were introduced so that soldiers could be properly trained. Broader social, financial, and eventually political changes came about so that the dynasty could keep up with the financial burden of military modernization. Ironically, the military successfully transformed into a modern entity while the Qing dynasty collapsed in 1911. Modern military technologies also heralded changes in Chinese tactic and strategy. For the most part, artillery and warplanes rendered ancient fortifications obsolete; firearms forced soldiers and officers to adapt to new fighting techniques; logistics became much more important than in traditional warfare since generals must coordinate the movement of more hardware and supplies if an army was to be able to fight at all. Yet modern technologies had their limit. Despite the power of artillery and machine guns, Liu Zhenhua was forced to employ more traditional tactics (such as tunneling or scaling) in order to overcome the supposedly obsolete fortifications. Similarly, the defenders ended up relying oh the walls and moat to forestall a quick allied victory. While modern technologies had increased the power of an army exponentially, they also made warfare a complicate and costly affair. Without an interlocking series of industries to supply and service these modern weapons, they were no better than expensive t o y s - impressive to look at but completely useless. In an era where most warlords had little or no access to these industries, traditional technologies and weapons (sword, bows, stone, boiling water) were far more cost effective and reliable. T h e warring style constituted a separate component in this hybrid military system. The command structure as well as the personal quality of the warlords was often favorite topics among critics of this period. A rationally organized command structure based on an impersonal military hierarchy and Western military principles are often compared to the warlord system. No doubt the Western system is efficient with the support of a national government. But conditions that gave risk to the armies of nation-27 state did not exist in China in 1926. Foreign critics have consistently crit icized the hybrid nature of the warlord military system as comical and detr imental to their overall military power. But the "old" Chinese practices offered the defenders more flexibility. The personal command structure of the warlord armies gave commanders an unprecedented degree of autonomy in making tactical and strategic decisions. Given the isolated nature of the Xi 'an garr ison, a Western bureaucrat ic command system would be paralyzed for lack of central instruction. Though the hybrid training was ineffective by Western standards, it encouraged the soldiers to adopt alternative f ighting styles. A second point of content ion is the warlords' personal abil it ies. They are often seen as violent but unsophist icated thugs bickering among themselves. The "comic opera" analogy fai led to take into account the natural ability of the warlords to learn and adapt. They did not have to attend military academy or immense in the Western-centr ic espirit de corps in order to fight well . Often they drew from practical experience and a rich tradit ion of military literature to better their skills. Ambit ion, cunningness, personal charismas supplemented the lack of formal training. Likewise, crit icism of the warlords' ability to fight simply ignored the circumstance of the period. War lords such as Feng or W u had been praised for their skills as strategist or frontl ine commander. Yet these merits described only part of the picture. For they were frontl ine generals, strategists, polit icians, and quartermasters all woven into one. The fact that the Northern warlords managed to amass huge armies attested to their organizat ional as much as polit ical or military skill. 6.1. Regional Issues Concerning the Siege W h y invest so much in a siege when other tactics were avai lable? It was not uncommon when one side would either abandon the f ield of battle or resolved to the "silver bullet" strategy. One reason why the Guominjun chose to f ight instead of surrendering was the involvement of the Henan Red Spear during the siege. The blood feud between the two provinces originated in the early years of the War lord period. Shaanxi t roops would station in Henan and vice versa. Ostensibly to exterminate marauding band of bandits, these troops were considered as de facto bandits by local populat ion. In addit ion to pil lage and random kill ing, it was not uncommon for these 28 "guest army" to cooperate with bandits and turned against anti-bandit organization such as the Red Spear . 9 1 "A rabbit does not eat the grass around its own burrow". 9 2 In the absence of a unifying national identity, the people could only identify with their regional identity. These "guest army" were ruthless because they saw their neighbors as vulnerable prey and not fellow C h i n e s e . 9 3 Hence, when the Henna besiegers attacked Xi'an, they saw the Guominjun as foreigners who had raped their land and the Shaanxi population as willing accomplices. It is unclear the degree to which regional issues influence the warlord struggles. W a s the siege the result of warlord politic or an expanded version of regional feud? Given the lack of relevant information, one can only draw several general observations. T h e longstanding hostility between the two provinces made the issue of regional relations an important determinant. Regional self-help organization such as the R e d Spear Society had an important role to play, but it remained to be seen their relationship to the warlords and the local gentry. T h e affluent gentry also had a role during the siege. But it is unknown how they would benefit from the change of order and their financial ties to the warlords. 7. Conclusion In Western historiography, the Warlord period is regarded as an anomaly among other pivotal political movements in the 1920s (the rise of the C C P and G M D , 1 s t United Front, establishment of the Nanjing national government). While the warlords did not have the same impact as the rise of the two modern Chinese political parties, it is problematic to cast the whole period as a mere sideshow. T h e warlords coexisted with the rise of the G M D and C C P and their actions in many cases directly influenced Chinese history in the 1920s. From a broad perspective, the Xi'an siege directly influenced the Northern Expedition and the re-emergence of a National government in Nanjing. It drew the attention of the Northern warlords and drained their resources in the critical early phrase of the Northern Expedition. T h e Northern warlords began to 9 1 Diana Lary. Region & Nation: the Kwangsi Clique in Chinese Politics, J925-37. (New York: Cambridge University Press, 1974), 43; Billingsley, 154-155. 9 2 Ibid, 181. 9 3 As Diana Lary points out in Region and Nation, the concept of "Chinese" probably did not register among the population at the time. Lary, 1974, 1-20. 29 recognize the potential threat of the G M D by late 1925. With his vacillation towards Sun Yatsen's ideology and close relation to Communist Russia, Feng was seen as a potential ally of the G M D . Hence, Zhang Zuolin and W u Peifu joined force to pre-empt the Guominjun. Ostensibly called the "anti-Red" war, the campaign was in fact an attempt to forestall the rise of a new political order. A s the high point of this campaign, the siege illustrated the fundamental weakness of the warlord alliance. Without the ideological underpinning, the allies easily lapsed back to everyone unto himself. 9 4 Militarily, the siege reflected the hybrid nature of the Chinese military system in the 1920s. T h e Chinese military was in a transitional period during the Warlord era. After the Sino-Japanese war in 1895, it had absorbed a broad range of Western military experience. Structurally, the Northern warlords imitated the modern Western military organization. T h e creation of the officer corps, arsenals, the opening of military academy produced a cadre of officers knowledgeable in the contemporary military art. Yet the impacts of this new system was limited vis-a-vis the "old" practice of tradition military experience. A s Jerome C h e n points out, China is far to complex to make a military only solution to reunification viable. T o the warlords who decided to resolve to violence alone, they would be foolhardy to ignore the well-established and battle-tested practices of the "old" ways. T h e professionalization and bureaucratization of the military certainty had their merits, but the fragmentary nature of the warlords precluded them from being fully utilized. T h e Western military institutions went hand in hand with the advent of the centralized nation-state system. But this system was significantly weakened when the central government was powerless and the nation's resources were pawned to the regional militarists. T h e Chinese institutions (such as personal based military hierarchy, indiscriminate recruitment in lieu of professional soldiers) persisted precisely because they were called for under the warlord system. What good was a professional army without a coherent ideology to attract and mobilize the 9 4 According to Diana Lary, regional militarism is a self-perpetuating movement. It could not fill the ideological gap left by the collapse of the old order, but it would not allow one to form. The only way to get rid of it was by militarism, albeit one with a "higher" ideological commitment. The GMD's National Revolutionary Army represented a new generation of "party army". Emerged in the 1920s, it was bounded by a sense of higher duty and ideological fulfillment. It is one stage of the Chinese military evolution where the fragmentary warlords gave way to the party army, with its higher ideology and national based character. Lary, 1974, 1-21. 30 population? There was little reason to invest heavily to convert their forces to a higher military standard as long as they facilitated warlords' personal advancement. In terms of the siege, the ancient fortifications offered the Guominjun enough protection to wear off the allied attack. Given the lackluster performance of both sides, the Guominjun could hide behind the walls and weather the initial assaults of Liu's army. O n c e the momentum dissipated, Liu would have to invest dearly in manpower and resources to breach the defense. Since he relied on his forces to hold on to power, understandably he chose to starve the defense rather than all out assaults. T h e "old ways" and fortification replaced Western military experience and they demonstrated their effectiveness under the warlord context. T he siege also represented a major traumatic experience to the civilians involved. T h e suffering incurred during the eight months extended beyond the city and it residents. Without the protection of the wall, those who lived in Xi'an's vicinity suffered just as much. Looting and pillaging by the invading army was offset by the purposeful destruction of properties by the defenders. T h e besiegers also shared the suffering. Consistently painted as the rampaging thugs, the besiegers were in such a berserk mood in part due to the past suffering inflicted by the Shaanxi "guest army". The regional feuding contributed to the dynamic of this siege and put the suffering in a broader context. This problematizes the portrayal of the two sides. No longer were the besiegers nameless villains perpetuating mindless violence acts, nor were the defenders simply heroes defending the helpless population. T h e siege showed the warlord army to be surprising resilient when its survival was threatened. T h e memory of the siege was manipulated or buried because it did not fit the traditional motif of historiography or the rhetorical framework of the political parties. In doing so, the suffering incurred was manipulated into misleading stereotypes. T h e memory of the siege should not be cast aside because it does not fit an arbitrary motif of heroic war memory or didactic value to history. 31 8. Bibliography Primary Sources: North China Herald New York Times Feng Yuxiang i i 3 E # Wode Shenghuo « a f i ^ f £ » (My life). Vol . 1-2 Heilongjiang: Heilongjiang Renmin Chubanshe , 1981. . Feng Yuxiang Zichuan (M^-^F^M) (Autobiography of Feng Yuxiang). Beijing: Junshi Kexue Chubanshe, 1989. Liaoningsheng Danganguan. ed. j g i r # ^ ^ | f l i . Zhonghua Minguo Shiziliao Conggao. {^W^M$.'MPrWM-) Vol .3 Beijing: Zhonghua Shuju, 1985. Zhongguo Dier Danganguan c^ffl^—M-fe^^it (Second Historical Archive of China) Feng Yuxizng Riji Olpsfi^ ElgB) (Diary of Feng Yuxiang) Vol . 2 Jiangsu Guji Chubanshe, 1992. English Sources: Billingsley, Phil. Bandits in Republican China. California: Stanford University Press, 1988 Black, Jeremy. A Military Revolution? Military Change and European Society, 1550-1800.London: Macmillan Education Ltd., 1991. ed. European Warfare: 1443-1815. New York: St. Martin's Press, Inc., 1999. Bonavia, David. China's Warlords. New York: Oxford University Press, 1995. Broomhall, Marshall. General Feng: A Good Soldier of Christ Jesus. London: China Inland Mission, 1923. Chan, Anthony. Arming the Chinese: the Western Armaments Trade in Warlord China, 1920-1928. Vancouver: University of British Columbia Press, 1982. . "The Chinese Air Force During the Warlord Era, 1916-1928". Army Quarterly and Defense journal (Great Britain). 1983 113(1): 81-89. Ch'en, Jerome. The Military-Gentry Coalition: China under the Warlords. Toronto: University of Toronto-York University, Joint Centre on Modern East Asia , 1979. Ch'i, Hsi-sheng. Warlord Politics in China 1916-1928. California: Stanford University Press, 1976. Clark, Alan. Barbarossa: The Russian-German Conflict, 1941-1945. London: Phoenix Giant Paperback, 1995. Department of State. Division of Biographic Information. "Biography of Feng Y u -Hsiang". Washington, D . C . : Department of State, 1948. McCord , Edward A. "Burn, Kill, Rape, and Rob: Military Atrocities, Warlordism, and Anti-Warlordism in Republican China". Diana Lary & Stephen MacKinnon ed. Scars of War: the Impact of Warfare on Modern China. (Vancouver: University of British Columbia, 2002), 26. Fairbank, John K., ed. The Cambridge History of China. V o l . 1 2 , Republican China 1912-1949 Part I. New York: Cambridge University Press, 1983. Gillin, Donald G . Warlord Yen His-shan in Shansi Province 1911-1949. New Jersey: Princeton University Press, 1967. How, Julie Lien-ying, ed. "Soviet Advisers with the Kuominchun, 1925-1926: A Document Study." Chinese Studies in History XIX, no. 1 -2, (Fall-Winter 1985-1986): 1-213. Kapp, Robert A. Szechwan and the Chinese Republic: Provincial Militarism and Central power 1911-1938. New Heaven: York University Press, 1973. Keegan, John. A History of Warfare. New York: Vintage Books, 1993. Kierman, Frank A. Jr. & Fairbank, John K. ed. Chinese Ways in Warfare. Cambridge: Harvard University Press, 1974. Kotenev, Anatol M . Chinese Soldier: Basic Principles, Spirit, Science of War, and Heroes of the Chinese Armies. Shanghai: Kelly and W a l s h Limited, 1937. Lary, Diana. Region and Nation: the Kwangsi Clique in Chinese Politics 1925-37. New York: Cambridge University Press, 1974. . Warlord Soldiers: Chinese Common Soldiers, 1911-1937. New York: Cambridge University Press, 1985. . "Warlord Studies." Modern China 6, no. 4, (October 1980): 439-70. 33 Lin, Alfred, H. Y. "Building and Funding a Warlord Regime: The Experience of Chen jitang in Guangdong, 1929-1936." Modern China 28, no. 2 (April 2002), 177-212. Lo, Kuan-chung. Three Kingdoms: China's Epic Drama Translated & ed. Moss Roberts. New York: Pantheon Books, 1976. McCord, Edward A. The Power of the Gun: the Emergence of Modern Chinese Warlordism. Berkeley: University of California Press, 1993. Mi, Zanchen. The Life of General Yang Hucheng. Translated by Wang Zhao. Hong Kong: Joint Publishing Co., 1981. Michael, Franz. "Military Organization and Power Structure of China during the Taiping Rebelion". Pacific Historical ReviewVol.18 (Nov. 1949): 469-483. Nathan, Andrew J. Peking Politics, 1918-1923: Factionalism and the Failure of Constitutionalism. Berkeley: University of California Press, 1976. Parker, Geoffrey. The Military Revolution: Military Innovation and the Rise of the West, 1500-1800. 2 n d ed. Cambridge: Cambridge University Press, 1996. ed. Cambridge Illustrated History of Warfare. Cambridge: Cambridge University Press, 1995. Perry, Elizabeth J. Rebels and Revolutionaries in North China. 1845-1945. California. Stanford university Press, 1980. Powell, Ralph, L. The Rise of Chinese Military Power 1895-1912.H&H Jersey: Princeton University Press, 1955. Pye, Lucian W. Warlord Politics: Conflict and Coalition in the Modernization of Republican China. New York: Praeger Publisher, 1971. Sima Qian. Records of the Grand Historian. Han Dynasty /.Translated by Burton Watson. Revised ed. New York: Columbia University Press, 1993. Seymour, William. Great Sieges of History. London: Brassey's, 1991. Sheridan, James, E. Chinese Warlord: the Career of Feng Yu-hsiang. California: Stanford University Press, 1966: Shyu, Nae-Lih. 1960. "Feng Yu-Hsiang and the Kuominchun, 1924-1928." MA. Thesis, University of Washington. Skinner, G. William. City in Late Imperial China. California: Stanford University Press, 1977. 3 4 Stremski, Richard. "Britain and Warlordism in China: Relations with Feng Yu-Hsiang, 1921-1928". Journal of Oriental Studies (Hong Kong). 1973 11(1):91-106. Suleski, Ronald. Civil Government in Warlord China: Tradition, Modernization and Manchuria. New York: Peter Lang, 2002. Sutton, Donald S. Provincial militarism and the Chinese Republic: the Yunnan Army 1905-25. Michigan: University of Michigan Press, 1980. Tracy, James D. City Walls: the Urban Enceinte in Global Perspective. Cambridge: Cambridge University Press, 2000. Ven, Hans van de. "The Military in the Republic". China Quarterly (1997): 352-374. Waldron, Arthur. From War to Nationalism: China's Turning Point, 1924-25. New York: Cambridge University Press, 1995. The Great Wall of China: from History to Myth. New York: Cambridge University Press, 1990. Watson, Bruce Allen. Sieges: a Comparative Study. London: Praeger, 1993. Wilbur, Martin C. & How, Julie Lien-ying, ed. Document on Communism, Nationalism, and Soviet Advisers in China: 1918-1927. New York: Columbia University Press, 1956.' Wilkinson, Endymion. Chinese History: A Manual. Revised & Enlarged Cambridge: Harvard University Press, 2000. Wou, Odoric Y. K. Militarism in Modern China: the Career of Wu P'ei-fu 1916-39. Dawson: Australian National University Press, 1978. Young, Ernest P. The Presidency of Yuan Shih-k'ai: Liberalism and Dictatorship in Early Republican China. Michigan: University of Michigan Press, 1977. "Warlords, 1925" (map). US Military Academy, West Point. Chinese Sources: Ding, Zhongjiang T ^ t L Beiyang Junfa Shihua (itW^M^M) (History of the Beiyang Warlords). Vol.4. Beijing: Zhongguo Youyi Chuban Gongsi, 1992 Feng, Hongda & Yu Huaxin Ml&M. & T N ^ L > Feng Yuxiang Jiangjun Hunsao Zhonghua VM^-K^-WW-^M^W} (General Feng Yuxiang's Spirit Swept China). Beijing: Wenshi Ziliao Chubanshe, 1981. 35 Gao, Xingya ^M^Feng Yuxiang Jiangjun (M^-ffiffi'W-} (General Feng Yuxiang). Beijing: Bejing Chubanshe, 1982. Jia, Pingwa MTO- Lao Xi'an: Feidu Xieyang Jliftf4G§> (Old Xi'an). Nanjing Shi: Jiangsu Meishu Chubanshe, 1999. Jian, Youwen fSf5£:&;. Feng Yuxiangchuan Ojf3£;#fl0} (Biography of Feng Yuxiang). Vol. 1&2 Taipei. Chuanji Wenxue Zazhishe, 1982. Kwan, Puiching ed. t l $ 5 f c $ f $ i . Qiannian guduXian ( r f 1 ^ " f i ' f | 5 M ^ » (Xi'an- China's Ancient Capital for a Thousand Years). Hong Kong: Commercial Press, Hong Kong Branch, 1987. Lai, Xinxia & Guo Jianlin & Jiao Jingyi ed. ^ f / f JC - $ ' M&'MM-Beiyang Junta Shigao HtWW-M^M) (Historical Document of the Beiyang Warlords). Wuhan: Hubei Renmin Chubanshe, 1983. Lan, Mengbo WJSM- Xian <M§5:> • Hong Kong: Zhenzhong Shuju, 1957. Li, Maosheng & Luo Chunpu & Yang jianzhong ^h%$& ' ' H i t 1 ^- Yan Xishan Quanchu (Hflf^[_L|^'ff» (Biography of Yan Xishan). Beijing: Dangdai Zhongguo Chubanshe, 1997. Liaoningsheng Danganguan. ed. Zhonghua Minguo Shiziliao Conggao. (^IS-KKifefl l&WM) Vol.3 (Beijing: Zhonghua Shuju, 1985) Ma, Liqian, Wang Kaiji, Lu Yizhi, ed. JSJirP % H I I T S ^ * ^.Zhongguo Tielu Jianzhu Biannian Jianshi: 1991-1981 ( ^ S i i ^ ^ W ^ : 1881 -1981» (Brief History of China's Railway Construction: 1881 -1981). Beijing: Xinghua Shudian, 1983. Mi, Zanchen ^WlK- Yang Hucheng Jiangjunchuan {WofjtM$&'W-\%} (Biography of General Yang Hucheng). Beijing: Zhongguo wenshi Chubanshe, 1986. Peng, Jingzhong ^ f ? ^ - Zhongguo Fangzhi Jianshi < 4^119 ^ Sffiife) (A Brief History of the Gazetteer of China). Chengdu : Sichuan Daxue Chubanshe, 1990. Shaanxi Shifan Daxue Dilixi ed. ^^^WK^^M^M- Xi'ansi Dilizhi { g ^ c r f j i t e Mfc) (Records of Xi'an's Geography). Shaanxi: Shaanxi Renmin Chubanshe, 1988. Tao, Juyin f^WM- Beiyang Junta Tongzhi Shiqi Shihua (4fc#¥HaMtoB#^^IS) (History during the Beiyang Warlords Period) Vol. 1-3 Beijing: Sanlian Shudian, 1983. 36 Tseng, Hua-pi <t|l|g. Minchu Shiqide Yan Xishan <S#7Bf I T O B i l U O (Yen Hsi-shan in the Early Repiblican China: 1912-1927 AD). Taipei: Taiwan University, 1981. Wang , Jinxiang 3 E ^ # . "Er, Sanshi Niandai Yapia Wenti" " ~ - =+^ft$t> i f « 1 I I " Minguo Dangan (The Opium Problem in the 1920s and 30s) (BM^-fe) no.2 (1992): 71-76. Wang , W a n g BEH. XinXi'an ( | j f lS$» (New Xi'an). Beijing: Zhonghua Shuju, 1924. W e n , Gongji 3c£t~\M.- Zuijin Sanshinian Zhongguo Junshishi (SI^H-f-^^IS^ lljfe) (Chinese Military History in the Most Recent Thirty Years). Vol . 1&2 Taibei: Wenxing Shudian, 1962. W u , Changyi i^SR. Qiangu Gongchen Yang Hucheng {:f-~&xjj&M&ti&) (Yang Hucheng: Heroic General of a Thousand Generations). Beijing: Zhongguo Wenshi Chubanshe, 1993. Xi'anshi Difangzhi Bianzuan W a i y u a n h u i ^ ^ l r j i f e ^ ' ^ | ^ ^ # M # (Editorial Board for the Gazattee of the City of Xian) Xian Shizhi ( H ^ r f J ^ ) ) (Gazette of Xian) Vol.1 Xi'an: Xi'an Chubanshe, 1996. Zhang Baifeng & Li Zongyi ed. ±$S Beiyang Junfa: 1912-1928 # ^ ^ » (Northern Warlords) Vol . 2 Hubei: W u h a n Chubanshe, 1990. Zhang, Yungjia WS^. Yu Youren Zhuan (T^frfifii} (Biography of Y u Youren) T a i b e i : Zhongwai Tongxunshe, 1958. Zhongguo Tiedao Chubanshe ed. ^MWt^tfci^iLWi- Zhongguo Tielu Jioatong Diyuji (^'MWMr^C.M^MM) (Railway Transportation Atlas of China). Henan: Zhongguo Tiedao Chubanshe, 1989. Ciyuan (WtM) • Beijing: Commercial Press, 1998. Cihai {^M) • Taiwan: Zhonghua shuju, 1967. Zhongguo Lishi Diming Cidian ((^IMM^i#&if?^l>> • Nianchang: Jiangxi jiaoyu Chubanshe, 1998. 37 Appendix I Table 1 Character List This list provides Chinese characters for the Chinese terms introduced in this essay (including individuals' names, location, terms). B a i L a n g Etf&CfilP) Baotou l3jg| Beijing Zhengbian ihtitik^ chaihai yizi ^ ^ . ^ ^ Chang'an j l ^ r Dagu Datong j^M Duolun Feng Yuxiang $f 3sJ£ Feng Hongkui Fengtian guanzhong ||cfp gansidui WffiM G o u B a o j i e mGkfa G u o Songling Guominjun fflj^jil Hankou yj|n Heping Qichenghui frjzp$J$ft Kejun Kongming Jim laobaixing ^ H ^ f e Lantian §tEH Lanzhou |U ' |H Li Jinglin Li Yunlong ^ « f | Liu Ruming %[J^BJ Liu Zhenhua H!jfi#? Longhai WM Li Mingzhong^P.IM Liu Ruming fflft&BJ! Liu Yufen L u Z h o n g l i n J H H K Mount Song Lintong M a Z h e n w u Nankou i ^ p Ningxia S=£J| Qinlong FuHanjun S an Qiaozhen ELM Sanyuan Shanxi [iiiS Shaanxi Shanhaiguan LUj Table 1 MB Shenyuan Shengyuan Shijiazhuang Sun Liangcheng Sun Lianzhong Sun Y u e Song Zheyuan Tongguan MM W u Peifu J ^ f l ^ W u J i n g y i n mWU W u g o n g ^Xjj Wuyuan TLW> Xianyang Xiaoyanta /hW& X u Yongchang f ^ z k l l Y a n Xishan W\Wi\h Y a n g Hucheng #?JTI$< Y u e Weijun fiffiSg. Zhang Zongchang H T H H Zhang Wanqing MMJS. Zhang Zhijiang Zhang Zuolin Zhangjiakou • \2K Zhenguojun f t l l j p Zhonghua Minguo Guominjun Zhuge Liang 39 Table 2 A Chronology of Events in China, 1911-1928 1911 October Establishment of the Republic of China Feng Yuxiang became military governor of Shaanxi. 1 s t Zhili- Fengtian war. Feng led troops to support W u Peifu through Henan. 2 n d Zhili-Fengtian war. Jiangsu-Zhejiang war between Zhang Zuolin and Sun Chuangfen. Zhang was forced to withdawfrom Shanghai due to vulnerable supply line. T h e brief war rekindled his desire to occupy the strategic Zhili province. Beijing Zhengbian, Feng Yuxiang betrayed W u Peifu and occupied Zhili controlled Beijing. W u was defeated and fled to Hankou. 1924 September-October October 23 1925 Spring July November December 1926 March April May Minor clashes between Guominjun and Fengtian troops in the Beijing area. Feng and Zhang met to discuss a compromise. Both sides pulled troops out of the Metropolitan area. Feng took control of the Beijing-Hankou railway. Guominjun took Xi'an G u o Songling rebellion. Zhang Zuolin reached an informal alliance with W u Peifu to destroyed Feng Yuxiang and the Gouminjun. Guominjun belatedly launched attacks against Fengtian troops in Tianjin, Southern Liaoning, and Shandong. Feng announced his retirement from political and military affairs. Briefly stayed in Ningxia before traveling to the Soviet Union. T h e W u - Z h a n g alliance launched massive assaults against the Guominjun in Eastern China. Y a n Xishan joined the allies and attacked the Guominjun along the Beijing-Hankou railway. Guominjun evacuated from Tianjun and Beijing to Nankou. Liu Zhenhua attacked Guominjun in Shaanxi. Siege of Xi'an officially began on the 15 t h. T h e alliance launched assaults against Nankou. Y a n Xishan attacked Guominjun from Datong. Liu occupied key towns around Xi'an. The city was completely isolated. Northern Expedition began. Table 2 1926 June August September November 27 28' 1927 January 26 t h March 1928 June •th th Zhang and W u met in Beijing on the 26 t h and decided to continue the anti-Guominjun campaign. In reality, both were preoccupied with their own affairs and the campaign stalled. Due to lack of reinforcement, the Guominjun's positions in Inner Mongolia crumbled. Zhangjiakou lost the protection on its flank. Guominjun soldiers from Eastern China were ordered to withdraw westward. Y a n Xishan gained control of Suiyuan (Inner Mongolia) from the Guominjun. Feng returned from the Soviet Union on the 17 t h. Declared his intention to join the G M D and rid China of Warlordism. Guominjun reconstituted. Guominjun pushed Liu out of Shaanxi. T h e siege of Xi'an officially ended. Feng arrived at Xi'an G M D took Nanjing and Shanghai. G M D took Beijing. Many scholars considered this as the end point of the Chinese Warlord period. < Table 3 A List of Guominjun's Territory and its Commanders in Late 1925 Guominjun's territory in Northwest China was divided into five sections: Beijing Metropolitan Region Northern Hebei & part of Southern Mopngolia Ningxia Southwestern Liaoning & part of Inner Mongolia G a n s u Lu Zhonglin Zhang Zhijiang Li Mingzhong Sung Zheyuan Liu Yufen Source: Minchu Shiqide Yan Xishan Table 4 The Guominjun and its Commanders, 1925-26 The Guominjun was officially created after the Beijing coup on October 25, 1925. However, the generals listed below had been with Feng and his armies since at least 1924 9 5 . Feng Yuxiang (When he retired, Zhang Zhijiang became the Commander in Chief of all Guominjun) W u Jingyin (Yue Weijun took over after W u died in April 1925) S u n Y u e G u o Songlin 1 Guominjun 2 n d Guominjun 3 r d Guominjun 4 t h Guominjun 9 5 The exception was Guo Songlin and his 4TH Guominjun. He took up the designation after his launched surprise attacks against Zhang Zuolin. 42 Table 5 Guominjun's Commanders and Units in Xi'an, April-November 1926 Li Yunlong W e i Ding Y a n g Hucheng Total number of Guominjun soldiers in Xi'an at the ontset of the siege was around 10,000. Table 6 Clarification on the Chinese Term "guan" T h e Chinese term guan means a pass. It can be part of a geographic feature or purposely built to regulate the flow of traffic in a particular locale. In traditional period, the term "guanzhong" (IS 40 or "guannef (ISF*3) are both used to describe the W e i valley. T h e word "pass" denotes the numerous mountains passes in Shaanxi (such as Tongguan). In modem period, the term "guannei" is used to describe China proper. T h e "pass" refers to Shanhaiguan. Source: Ciyuan, 1998, 3253; Cihai, 1967, 107-108 2 n d Guominjun 10 t n Division 2 n d Guominjun 4 t h Division 3 r d Guominjun 3 r d Division 43 Appendix II Figure 1 Map of Shaanxi, 1920s SHAANXI PROVINCE 0. 25 50 75 100 km. ,J\ _ . „ V {*J«fcBIAN i HENQ5HAN YANCHUAN . Ksu. Sketch nup drawn for Oris book b* run Zkiung Sketch Map of Shaanxi Province SOV^ -UL '• L"&?. <£ \<\^ I-WWJU, Figure 2 Warlord China, 1925 Source: "Warlords, 1925". US Military Academy, West Point. 45 Figure 3 Th e City of Xi'an, 1930 1. Revolution Park 16. Barrack 2. Guannan Huiguan 17. Farmer's Market 3. Orphanage 18. Ministry of Construction 4. Red Cross 19. Ministry of Education 5. Guangren Hospital 20. County Office, Ministry 6. Telephone Exchange of Civic Affairs 7. Gongshan University 21. Ministry of Finance 8. Telegraph Office 22. Provincial Party Office 9. Lianhu Park (GMD) 10. Workshop 23. City's Party Office 11. Post Office (GMD) 12. Educational School 24. City Hall 13. Office of the Provincial 25. Police Station Government 26. Army Survey Office 14. Local Court 27. High Court 15. Xi'an Park 28. Ministry of Agriculture 29. and Mining Ministry of Industry Figure 4 Modern Xi'an: City Wall and Moat Source: Xi'ansi Dilizhi 47 Figure 5 Towers on top of Xi'an City Wall Source: Qiannian Gudu Xian There are a total of 98 towers on top of the city wall. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items