2. Earlier work on the advisory group has been published by Liu, F. F., A Military History of Modern China, 1924–1949 (Princeton, 1956)Google Scholar; Walsh, Billie K., “The German military mission in China, 1928–38,” Journal of Modern History 46 (09 1974) 502–13CrossRefGoogle Scholar; Liang, Hsi-Huey, The Sino-German Connection: Alexander von Falkenhausen between China and Germany, 1900–1941 (Assen/Amsterdam, 1978)Google Scholar. Two dissertations are Mehner, Karl, “Die Rolle deutscher Militärberater als Interessenvertreter des deutschen Imperialismus und Militarismus in China (1928–38),” inaugural dissertation, Karl-Marx Universität, Leipzig, 1961Google Scholar; and Seps, Jerry Bernard, “The German Military Advisers and Chiang Kai-shek, 1927–1938,” University of California, Berkeley, 1972Google Scholar. Fu Baozhen wrote an unfinished series of 10 articles, based on German and English sources, including the archives in Potsdam, entitled “Zai Hua Deguo junshi guwen shizhuan,” in Zhuanji wenxue, Vols. 24, 25, 26, 28, and 30 (1974–1977)Google Scholar. A rough official account is Deguo zhu Hua junshi guwentuan gongzuo jiyao, ed. Bureau of Military History, Ministry of National Defence, Republic of China (Taibei, 1969)Google Scholar, translated by the (U.S.) Office of Military History, A Summary of the Work of the German Military Advisory Group in China (Taibei, 1971)Google Scholar, hereafter Summary.

3. My calculations, from lists in Bundesarchiv Militärarchiv, Freiburg (hereafter DMAF), Deutsche Gesandtschaft in China (hereafter Msg) 160/9, and Falkenhausen Papers, N246/12. The totals at the end of each year were 1928: 4; 1929: 46; 1930: 47; 1931: 61; 1932: 65; 1933: 67; 1934: 59; 1935: 54. In March 1936 there remained 51, in May 1937, 42, and in April 1938, 33. The 11 out-of-service generals were Wetzell, Seeckt, Falkenhausen, Spemann, Streccius, Gudowius, Starke, Held, Lindemann, Link and Karlewski. At least five others would reach general rank in the 1939–45 European War: Nolle, Wilck, Voigt-Ruscheweyh, Newiger and Heinrichs.

4. Letter of Krummacher to his former commander, 20 April 1930, DMAF Msg 160/1: 151, “Chiang Kai-shek's speech to German advisers” (1 January 1930) in USWD 2657-I-357: 19.

5. Bülow (?), memo of 8 November 1933, Reichswehrministerium, in Politische Archiv, Auswärtiges Amt, Bonn (hereafter DAAB) “Militärischeangelegenheiten in China,” Politik 13, China, Vol. 7. See also note 9.

6. Trautmann despatch, 30 December 1933, DAAB Politik 13, China, Vol. 8; Wetzell, to Kai-shek, Chiang, 27 08 1933Google Scholar, DMAF Msg 160/3: 51–52.

7. Bauer was said to have been “ordered almost daily to Chiang Kai-shek.” Timann despatch, Hankou, 8 April 1929. DAAB Politik 13, China, Vol. 2; on Seeckt's and Falkenhausen's relations with Chiang, see Liang, , Sino-German Connection, pp. 91, 94, 102, 172–76Google Scholar. The seven who stayed are named in ibid. p. 134.

8. Walsh, pp. 504–505.

9. See in particular DMAF Msg 160/1, passim, on the war with Feng and Van (1930) and Falkenhausen's letters to Brinckmann of 18 June 1936, and 22 March 1937. DAAB Politik 8, No. 49, Vol. 1. Ambassador Trautmann (despatch of 30 December 1933) attributed Wetzell's unpopularity “in circles around Chen Yi” to his obtaining large military appropriations direct from the former finance minister, T. V. Soong, thus giving himself leverage over “the corrupt Chinese generals.” These, however, had been alienated too, for in seeking reforms Wetzell had “again and again ruthlessly stuck his finger into the open wounds of the corrupt Chinese army.”

10. In the early 1930s, some German weapons were made under licence by Swedish, Dutch and Swiss firms because of the prohibitions of the Versailles treaty.

11. See DMAF Msg 160/4, passim, for Wetzell's arrangements; and Ernst Bauer (Colonel Bauer's son) to Kriebel, 28 June 1934, confirming that Brinckmann had sought out experts for particular tasks “in contact with the Reichswehrministerium,” after which they were proposed to the Nanking authorities by Wetzell. Bauer Papers, 50: 231, Bundersarchiv, Koblenz.

12. Spence, Jonathan, To Change China: Western Advisers in China, 1620–1960 (Boston, 1969), pp. 290–91.Google Scholar

13. The advisers' relationship to Nazism is complex, and beyond the scope of this article. While their views ranged from ardent support to aversion, they felt more in touch with Chinese than German politics. The kinds of personal animosities that appeared were between Bavarians and Prussians, or between those recruited by Bauer and Wetzell, not between Nazis and their opponents.

14. See note 2 above. On the divisive and centrifugal tendencies of militarism in the warlord period, see Ch'i, Hsi-sheng, Warlord Politics in China, 1916–1928 (Stanford, 1976Google Scholar), and Sutton, Donald S., Provincial Militarism and the Chinese Republic: The Yunnan Army, 1905–25 (Ann Arbor, 1980)Google Scholar. For a theoretical approach, see Sutton, Donald S., “Chinese Warlordism and organization theory: goal conflict and dual structure in the evolution of a militarist army,” Social Science HistoryGoogle Scholar, forthcoming. For the concept of residual warlordism, see Sheridan, James E., Chinese Warlord: The Career of Feng Yü-hsiang (Stanford, 1966)Google Scholar. For the politics of the Nanking decade (1928–38), see Tien, Hung-mao, Government and Politics in Kuomintang China (Stanford, 1972)Google Scholar, and Eastman, Lloyd E., The Abortive Revolution: China under Nationalist Rule (Cambridge, Mass., 1974).CrossRefGoogle Scholar

15. xuexiao, Lujun junguan, ed., Lujun junguan xuexiaoshi (Taibei?, 1964), 3244, 3283, 3302, 3333, 3344, 3370.Google Scholar

16. Military attaché reports on the Central Military Academy, dated 31 January 1933 (S. V. Constant); and 2 May 1935 (W. S. Drysdale), United States War Department (hereafter USWD) 2277-I-19: 13. At Whampoa, the Academy had not even had a firing range, and the course had lasted a mere eight months.

17. Constant, 31 January 1933, USWD 2277-I-19:13; and Zhongyang tujun junguanxuexiao shigao, bian 7 (political training). A critical view of the Whampoa clique in the 1930s and 1940s is presented in F. F. Liu, Military History.

18. Lujun junguan xuexiao xiaoshi, 3373.

19. Stilwell, J. W., “Military Schools and Colleges: The Central Military Academy,”Google Scholar USWD 2277-I-19:18. Wetzell called him “one of the people who value our work highly.” 5 May 1933, DMAF Msg 160/4: 125.

20. USWD 2277-I-24 (reporting Major Constant's visit on 16 December 1932); USWD 2277-I-19:17, 6 May 1936. Colonel Stilwell, as usual a sharp critic, called the War College “the most inadequately housed military school in or about Nanking,” on the basis of observations by Captains Crist and Tormey. He also commented on the presence of “elementary school subjects,” such as mathematics in the curriculum – these were needed since there existed field officers of artillery who did not even know what an “average” was. See also Lujun daxuexiao jiangguan ban yi ji disiqi tongxuelu; and Yaohuang, Wan, “Zhuchi lujun daxue shiqi de huiyi,” Part 1, Zhuanji wenxue, 23, 5 (No. 138, 11 1973), 80–85.Google Scholar

21. Lieut-Col. Drysdale, 2 May 1935, USWD 2277-I-19:26; Stilwell report, 17 June 1936, based on inspections by Capt. F. N. Roberts et at., USWD 2277-I-19:18; Major Constant, 27 February 1933, USWD 2657-I-357:23.

22. Constant report, USWD 2277-I-19:13; Wan Yaohuang, commandant of the War College from 1936 and after its migration to Cunyi in 1939, makes no mention of the Germans in surveying the College's history. Zhuanji wenxue, 23, 5: 82–85; 23, 6:108.

23. Zhongyang lujun junguan xuexiao, 181b, discussing the 8th Class.

24. At least 12 Germans taught at the military schools in 1933; 15 in 1934; 20 and five halftime in 1935; and 12 in May 1937. See USWD 2657-I-375:23, and DMAF N246/12, Msg 160/9: 179, 181.

25. Excluding seven advisers who had died in China, 68 advisers had already left by the end of 1935. Four had been dismissed before the expiration of their terms of contract, 13 left at their own request before completing their contracts, 14 because of illness, and six after their one-year contracts had expired. Thirty-one of the 68 stayed from two to seven years. See sources in note 3 above. Considerable dissatisfaction was expressed during the brief tenure of Colonel Kriebel, Max Bauer's successor in 1929–30. See Minister Borch's despatch, 28 November 1929, DAAB 1929 Politik 13, Vol. 2.

26. Among the full time War College instructors in 1934 were Generals Spemann (Artillery), Streccius (Air), Gudowius and Starke, and Colonel Lucht and Lt.-Col Guse; those at the War Academy at the same time were Colonel Bade and Lt.-Col. Nolte, and Captain von Hunolstcin. DMAF N246/12: 2, 3, 4.

27. See Zhungyang lujun junguan xuexiao, pp. 188–90, for a list of over 80 works translated by Wu Guangjie and the Germans.

28. Baozhen, Fu, “Deguo guwen,” 25 (08 1974), 81–82.Google Scholar

29. Seeckt to Wetzell, N247/133. Falkenhausen memo, late 1936, Summary, pp. 23–24. Besides all of the German sources, Wang Qianan, then an officer interpreter, stresses the importance of field exercises and practical work, in his “Deguo guwen zai Nanjing shiqi gongzuo de huiyi,” Zhuanji wenxue, 27, 4. (No. 161, October 1975), 53–54. For a Communist criticism of Nationalist training as too impractical even from the vantage of 1939, see note 95. On the elusiveness of practicality in modern Chinese education, see remarks by Borthwick, Sally and Bernstein, Thomas P. in Rozman, Gilbert, ed., The Modernization of China (New York, 1981), pp. 402–404.Google Scholar

30. On the Artillery School, see Constant report on 16 December 1932 and Wang Qianan, “Deguo guwen.”

31. Wetzell, to (Chinese) Vice-Minister of War (Chen Yi), 12 1931Google Scholar; DMAF Msg 160/3:33.

32. Wetzell, to Vice-Minister of War, pp. 33, 35, 37, 39Google Scholar, and Busekist letter, 20 April 1930, DMAF Msg 160/1:149. Wetzell, to Brinckmann, , 4 12 1933Google Scholar. Msg 160/4: 54–55.

33. See note 15. Lujun junguan xuexiao diershiyi qi bubing diyi dier dadui biye tongxuelu (n.p., n.d.). Revised draft translation by Collier, Harry H. and Williamsen, Thomas M., The First and Second Infantry Battalions, 21st Class, Chinese Military Academy Student Directory (Taibei, 1970), 1: 4–12Google Scholar. In 1936, Stilwell reported total attendance at the Central Military Academy as 4,500 (3,000 ordinary cadets, presumably of the 10th, 11th and 12th classes, 200 anti-aircraft and 300 aviation cadets, and 1,000 student officers), 17 June 1936, USWD 2277-I-19:18.

34. Stilwell report, USWD 2277-I-29:18; Guodong, Peng ed., Jiang Jieshi, xiansheng jiayan leichao (Shanghai 1937), pp. 479–80, 456–62Google Scholar; Jieshi, Jiang, ed., Lushan junguan xunliian ji, (n.p., n.d.)Google Scholar; Jieshi, Jiang, ed., Emei junguan xunlian ji (n.p., 1935)Google Scholar; Whitson, William W., “Jiangxi jiaofei zhi jiantao,” Zhanshi huikan, 1 (04 1969), 35.Google Scholar

35. Constant, S. V., 31 01 1933Google Scholar, USWD 2277-I-19:13.

36. Seeckt insisted on a “Schul- oder Musterbrigade,” that would not be designated for military use. Memorandum to Chiang Kai-shek, June 1933, in Trautmann despatch, 26 August 1933, DAAB Politik 13, China, Vol. 7. On 12 May 1937, Hummel, Bauer, von Schmeling, von Boddien, Oehme and Bautz were still assigned to the Lehrbrigade; see DMAF Msg 160/9: 85, 87. In a letter to Brinckmann of 22 March 1937, still the main official contact in Berlin, Falkenhausen expressed his satisfaction with the Lehrbrigade as a “good means of propaganda for our work,” DAAB Politik 8, China (No. 49), Vol. 1.

37. On Bauer's preliminary work, see North China Daily News, 13 May 1929 cited in despatch, of 15 May 1929, DAAB Politik 13, China, Vol. 2; and letter from Fischer, Lieutenant-Colonel, 14 10 1929Google Scholar in DAAB Politik 13, China, Vol. 2. On the Henan campaign, see the account by an anonymous military adviser forwarded by the Ostasiatischer Verein, Hamburg-Bremen, to the Berlin Foreign Office on 29 November 1930. DAAB Politik 13, China, Vol. 4. See also Wang Qianan, “Deguo guwen”; on Feng, who died in June 1931, see Wetzell despatch to Brinckmann, 27 June 1931, DMAF Msg 160/1: 101. In reports from 1935, 1936, 1937 and January 1938, Captain Constantin Meyer, who helped to train the Lehrdivision in 1930, is given as the adviser to the 87th, a striking indication of German (and personal) commitment to this force.

38. Wang Qianan, “Deguo guwen,” and Krummachery's letter, cited above, of 20 April 1930, DMAF Msg 160/1: 149–61.

39. Lieut.-Col. Drysdale, W. S., “The 87th division, Chinese national army,” 24 01 1933Google Scholar in USWD 2009/230:1. Both divisions took in more recent graduates of the Military Academy, e.g. 40 each from the 8th class of 712, graduating in May 1933.

40. Seeckt memorandum to Kai-shek, Chiang, 06 1933Google Scholar; Liang, , Sino-German Connection, p. 90Google Scholar, citing the Starke papers for divisions “reorganized and trained by the Germans,” DMAF N218/1. By 1937, there were probably eight German-armed divisions (80,000 men), and about 220,000 more following German-style training. Lloyd Eastman, E., “The war years, 1937–1945,” China Under Nationalist Rule: Two Essays (Urbana, n.d.), p. 96.Google Scholar

41. Yuanliang, Sun, Yiwan guangnian de yishun (Taibei, 1972), p. 136Google Scholar. Dr. Chen Yung-fa kindly brought this work to my attention. In February 1933, a “confidential agent” reported to the U.S. attache the names of five advisers with the 87th, and four with the 88th. USWD 2657-I-357:23. In 1934, Seeckt took the advisers out of the fully-trained divisions, but after his departure Meyer was soon back with the 87th, and Vetter had joined the 88th.

42. T. V. Soong's Salt Division (so-called because it was funded by salt monies) had two German advisers in 1933 and at the end of 1935, one of whom, von Stein, held the same post in January 1938; the 36th division had been assigned Major Schimmelfennig by March 1936. DMAF Msg 160/9: 77, 79, 81, 167, 179, 187. USWD 2657-I-357:23. See also note 81.

43. For correspondence between Krummacher and Wetzell on the Lehrdivision in Henan, see DMAF Msg 160/1: 27–143; on Jiangxi, see shizhengju, Guofangbu, Jiaofei zhanshi (Taibei, 1962), 2: 297, 302–7, 320–21, 326–27Google Scholar; on the 1932 fighting at Shanghai, Wetzell expressed his satisfaction that the German-trained troops had stood their ground (“ihren Mann gestanden haben”) in his letter to Seeckt, 14 March 1933, DMAF N247/133:4. See also “The Sino-Japanese clash, Shanghai, January–March 1932,” special report prepared by the Far Eastern Section of the Military Intelligence Division, USWD General Staff, 3 June 1932, in United States Department of State (hereafter USDS) 793.94/5362 (pp. 13815–53). This report gives credit to the 5th Army's efficiency at defence and withdrawal, in the evident absence of the German advisers. The two divisions, known as the 5th Army, lost 1,566 dead, and the 19th Route Army 2,413. Reflecting the Japanese forces' superior arms, the Japanese lost only 591 dead and 1,173 wounded. Shō Kai-seki hiroku, Chinese version, Jiang Zongtong milu (Taibei, 1976), 8: 74–75Google Scholar. On the 88th division in Fujian and Guangdong, and representing the centre in Sichuan and Anhui in intervening years, see Yuanliang, Sun, Yiwan guangnian, pp. 136–39, 142–43, 150–60, 164–65.Google Scholar

44. Yuanliang, Sun, Yiwan quangnian, pp. 180–88Google Scholar. In 1937, as in 1934, Sun commanded the 88th; Wang Jingjiu the 87th. See Jiaofei zhanshi, 2: 264, table 19, and Qiyun, Zhang, ed., Kang Ri zhanshi (Taibei, 1966), pp. 50–52.Google Scholar

45. Rittmeister Lorenz's memo survives only in English, in USWD 2009–255/3. Song Xilian commanded the 36th.

46. See, for example, Lovat Eraser's critique cited in note 93.

47. Wetzell, to Krummacher, , 1 08 1930Google Scholar, DMAF Msg 260/1: 71. “General Fong believes like all Chinese division commanders that the division entrusted to him is only for his person. He does not consider that it is really supposed to be a Lehrdivision for the entire national army and that he has no right of possession over it.”

48. Yuanliang, Sun, Yiwan guangnian, pp. 212–20.Google Scholar

49. Some contemporary writers may have inferred German responsibility because they saw the Germans as part of a general struggle of Fascism against Communism, e.g. Smedley, Agnes, China's Red Army Marches (New York, 1934), xvGoogle Scholar; their responsibility was later asserted among others by Clubb, O. Edmund, Twentieth Century China (New York, 1964; 2nd ed. 1972), p. 200Google Scholar, and Tuchman, Barbara, Stilwell and the American Experience in China, 1911–45 (New York, 1971)Google Scholar. Whitson, William W., in his “Jiangxi jiaofei zhi jiantao,” p. 35Google Scholar, blames the Germans for encouraging tactics more appropriate to the First World War than to counterinsurgency; see also note 95.

50. Wetzell, “The Campaign against the Communists, 1931,” DMAF Msg 160/2. Mehner, , “Militarismus,” p. 155Google Scholar, writes that 12 Germans were sent in to “take over anti- Communist fighting” during the Third Campaign (1 July–2 September 1931), but without success. For Seeckt's communications with Chiang in June 1933 and May 1934, see notes 40 and 64. In a communication to me Professor Liang has conscientiously withdrawn a statement in his book that credits Seeckt with advice on “a line of forts … to force out the Communist Red Army from South China.” This information came from the secondary literature, referred to in note 49, not German archives. Liang, , Sino-German Connection, p. 95.Google Scholar

51. Wetzell, , “Campaign,” DMAF Msg 160/2: 121Google Scholar; see also 99.

52. Wetzell to Brinckmann, DMAF Msg 160/4: 61, 70. A similar conclusion based on different evidence has been arrived at independently by Wei, William, “The role of the German advisers in the suppression of the central Soviet: myth and reality,”Google Scholarpassim.

53. Ambassador Trautmann believed Wetzell was sent north to get him out of the way. Tratumann, , 30 12 1933Google Scholar, DAAB Politik 13, China 8.

54. Seeckt, to Wetzell, , 06 1933Google Scholar, DMAF Msg 160/4: 17.

55. See Krummacher, to Wetzell, , 5 09 1932Google Scholar, reporting that in response to an urgent request he had asked the engineer expert Weber to stop at Nanchang on the way back from his holiday. DMAF Msg 160/3: 107.

56. Lushan xunlian ji, p. 613; for the text of the handbook, see pp. 578–643. For speeches by the commandant of the training group Chen Cheng, see Chen tuanzhang jiangyan lu(n.p., 1934?).

57. Xiong to Xiao n.d., 13–14; to Liu, , 28 04 1933Google Scholar; Cheng, Chen, 2 03 1933Google Scholar; to Liu duban, App. 9b, all in Gongzhe, Xiong, Jiaofei chuyi (n.p., 1935)Google Scholar. On jianbi qingye, see Kuhn, Philip, Rebellion and its Enemies in Late Imperial China: Militarization and the Social Structure, 1796–1864 (Cambridge, Mass., 1970)Google Scholar. For a full account of the “pacification” of Jiangxi, see the work of William Wei, forthcoming.

58. Wetzell, “Campaign,” DMAF Msg 160/2.

59. DMAF Msg 160/4: 59; Msg 160/2: 125.

60. Despatch of 8 December 1933, USDS 893.00/12617; and W. S. Drysdale, reports of 22 September and 6 November 1934, contrasting the 52nd with the 87th division. USWD 2657-Ixs-261/138, 139.

61. Cited by Zhen, Zeng, “Jiangxi wuci weijiao gongfei zhanshu shang zhi yanjiu,” Zhanshi huikan 2: 137 (03 1970)Google Scholar. Wei (p. 7) writes that Liu Weiyuan, Dai Yue, Qiu Qingquan and Liu Jimin all “probably played a role” in the development of the strategy.

62. DMAF Msg 160/2: 41, 71, 75, 85–89. This is a diary by adjutant Krummacher, “Sine-Japanese conflict: battles around Shanghai, from January 1932.” See also note 43 above.

63. But for his three months' work from March 1933, confided Wetzell to Brinckmann, “the Japanese would be in Peking and Tianjin today (13 October 1933).” Msg 160/4: 68; Summary, p. 30.

64. See the minutes of Seeckt's conversations with Chiang Kai-shek on 28 April, 2 May and 3 May 1934 in DMAF Msg 160/5 especially 207, 217, 235.

65. See the interview (Besprechung) on 2 May 1934, in Msg 160/5: 235, 237; and on 4 May, Msg 160/5:207.

66. DMAF N246/12: 36.

67. Falkenhausen, to Soong, T. V., 12 08 1935Google Scholar (apparently an English version of his proposal to Chiang), DMAF Msg 160/2: 3–21. See also Liang, , Sino-German Connection, p. 192–99Google Scholar. Seps, , “German advisers,” p. 461Google Scholar, points out that Falkenhausen followed Seeckt except on Japan policy. Unlike Seeckt, he thought that a “highly mobile small army equipped with the most modern weapons would be able to immobilize a larger attacking army,” even lacking artillery.

68. This strategy was enunciated in Chiang's speech to political cadres in August 1935, but the ideas can be traced back to 1932. See Eastman, Lloyd E., “The war years, 1937–1945,” pp. 89, 195Google Scholar. A justification of “the determined policy of retreating into the interior at any cost,” citing intellectual supporters, is Wu, Hsiang-hsiang, “Total strategy used by China and some major engagements in the Sino-Japanese War, 1937–1945,” in Sih, Paul K. T., ed., Nationalist China During the Sino-Japanese War, 1937–1945 (New York, 1977), especially pp. 43–59Google Scholar. Falkenhausen explicitly praises Chiang's, “space-for-time” strategy in a letter from Germany on 17 May 1939, in Summary, pp. 18–20.Google Scholar

69. For the domestic situation, see Lary, Diana, Region and Nation: The Kwangsi Clique in Chinese Politics, 1925–1937 (London, 1974), pp. 197–99Google Scholar; and Falkenhausen to Brinckmann, 18 June 1936 and 22 March 1937, DAAB Politik 8, No. 49, Vol. 1. The rest of this paragraph is based on Liang, , Sino-German Connection, pp. 104–105, 127Google Scholar which uses material inaccessible to me.

70. Jiang Zongtong milu 11: 61–63.

71. Dorn, Frank, The Sino-Japanese War, 1937–1941: From Marco Polo Bridge to Pearl Harbor (New York, 1974), believed (pp. 78–79, 200)Google Scholar the stand at Shanghai was intended to provoke foreign intervention, and thought it was “disastrous” and “hopeless.” Chen Cheng's former associates make him in large part responsible for Chiang's decision, the underlying reason being to make time to prepare evacuation into the interior for protracted war. See Gillin, Donald G., “Problems of centralization in republican China: the case of Ch'en Ch'eng and the Kuomintang,” Journal of Asian Studies, 29, 4 (08 1970), 841–42.CrossRefGoogle Scholar

72. A contemporary journalist analyses the battle in Ji, Yang, Hu zhan shilu (Hong Kong, 1938)Google Scholar, Preface by Zhang Zhizhung. One of many admiring appraisals, by an informed American officer, is Carlson's, EvansTwin Stars of China (New York, 1940), pp. 31–33Google Scholar. On the defective leadership at the outset, see the Lorenz critique in the first part of this article. For criticisms that Chiang Kai-shek's instructions from Nanking obscured his commanders' boundaries of responsibility, failed to give battle plans, and destroyed their sense of initiative, see Dorn, pp. 72–73. For casualty figures, Eastman, , “War Years,” pp. 87–88Google Scholar, may be too high at 270,000, even including civilians. Carlson, Twin Stars, suggests well over 100,000 killed and wounded. The Junlingbu recorded 125, 130 deaths and 242, 232 wounded for officers and men in the whole of 1937. Editorial Committee for the Bloody Battle History of the Eight-year Resistance to Japan, ed., Ba'nian kang Ri xiezhan shi (Taibei, 1972), p. 512.Google Scholar

73. Jiang Zongtong milu, 11: 60.

74. Summary, p. 40.

75. Msg 160/8: 125–29.

76. Borg, Dorothy, The United States and the Far Eastern Crisis of 1933–1938 (Cambridge, Mass., 1964), pp. 456–65, 473–78, 644.CrossRefGoogle Scholar

77. (Falkenhausen) “Overview of Sino-Japanese war,” 6 April 1938, DMAF Msg 160/8: 17, 19. According to Zhang Fakuei, he and Zhang Zhizhung, the commanders of the right and central fronts, initially undertook to defend the lines for a period of three months, and proposed that the government set an upper limit to the number of divisions to be committed at Shanghai. Zhang Fakuei criticized Feng Yuxiang and others for overoptimism, and the “highest leadership” for procrastination in authorizing the final withdrawal only after orderly retreat was impossible. See Fakuei, Zhang, “Ba'isan Song Hu zhanyi huiyi,” Dacheng zazhi, 9 (08 1974), 2–5Google Scholar. In another memoir cited by Ziliang, Wang, Zhexi Kangzhan jilue (Taipei, 1966), pp. 24–25Google Scholar, Zhang criticized his superiors for ordering him at the end of October to hand over the right front command to General Liu Jianxu at a time when Liu's troops were still en route from Hangzhou.

78. I rely on American and British accounts for the retreat from Shanghai and Nanking. See in particular the summary in USWD 2657-H-439/649; Lovat Eraser, 2 January 1938, GBFO 371/22043: 78; and Consul Gage, GBFO 371/22111: 76b.

79. For the order of battle, see Kang Ri zhanshi, pp. 54–55. See USWD 2657-H-439/649 on the “futile” attempted defence of Nanking after giving up “virtually without a fight” what this author saw as “strong and heavily fortified natural positions between Chinkiang and Wuhu which had been prepared in advance.” See also Tang, Te-kong and Tsung-jen, Li, The Memoirs of Li Tsung-jen (Boulder, 1979), pp. 326–27Google Scholar; and Anon. (Falkenhausen?), “Nanking's Fateful Days, 12 December 1937–13 January 1938,” DMAF Msg 160/8:49–75, expressing contempt for the unsoldierly qualities of the Japanese Army.

80. Gage, p: 77a. Stennes, a police captain and a former member of the SA (Sturmabteilung), headed Chiang's bodyguard.

81. See list in DMAF Msg. 160/9: 77, 79, 81, dated 19 January 1938; and Gage report of 5 January 1938, GBFO 371/22111: 50–51. The following divisions now had advisers: 3rd (Schmeling, 6th (Wilck), 9th (Bruendel), 14th (Heinrichs), 185th (Streccius) and Salt (Stein).

82. In one of innumerable popular accounts, it is the only battle of 10 selected campaigns (all Nationalist) to be described as a “great victory.” Anon., Kang Ri shi da zhanyi (Hong Kong, 1971)Google Scholar. See also Ping, Wang, Kangzhan ba nian (Taibei, 1966)Google Scholar, and the special issue of Zhuanji wenxue (32: 10–34) commemorating the 40th anniversary of the battle with memoirs by He Yingqin, Sun Lianzhong and others. For confirmation of the seriousness of the fighting, see upbeat reports from MacKillop (Hankou) in GBFO 371/22043: 213–33, and Sir Archibald Clark-Kerr, GBFO 371/23044: 260–79, Shanghai, 20 May 1938; see also “Field operations in China, March 28-April 10, 1938,” by Captain E. E. Count, Jr., forwarded by Major H. L. T. Creswell, Military Attaché, Tokyo, USWD 2657-H-439/650.

83. For disappointment at the Chinese failure to follow through, see a comment on Sir Archibald's telegram, by Henderson, J. Thyne, dated 25 or 26 06, p. 279Google Scholar. Carlson (p. 152) reported that Falkenhausen complained, “I tell the Generalissimo to advance, to attack, to exploit his success … [b]ut nothing is done. Soon the Japanese will have 8 to 10 divisions before Hsüchow [Xuzhou]. Then it will be too late.” Chiang himself evidently saw the Xuzhou concentration as a bluff to draw the Japanese and gain time to dig in at Wuhan; see Jiang zongtong milu (1977), 11: 138. For Stilwell's analysis of why the “optimism created by Taierhchuang believed not fully warranted,” see USWD 2657-H-439/624 (received at War Department, 22 April 1938). The figures for the Xuzhou withdrawal are Falkenhausen's, a pencilled rejoinder to Japanese claims that the Chinese had lost 240,000 men, 96 cannon and 14 tanks, etc. There were no Chinese Tanks, or even armoured cars. DMAF Msg 160/8: 47.

84. Romanus, Charles and Sunderland, Riley, United States Army in World War II: China-Burma-India Theater, Vols. 2–3 (Washington, D.C. 1956, 1959).Google Scholar

85. Liang, , Sino-German Connection, p. 133Google Scholar. See also part of a Falkenhausen memo of 22 March 1938 including a sketched plan for the wide outflanking of the Japanese at Hanzhang and Yixian, ibid. after p. 152.

86. Gage, GBFO 371/22111: 76b. See also GBFO 371/27043, despatch of 28 April 1938, in which the German advisers are quoted to the effect that their advice is now being followed. At Tannenberg during August 1914 Field Marshal Hindenburg's vast envelopment killed 100,000 Russian soldiers and took 960,000 prisoner. The hardest-hit Japanese 5th and 10th divisions were recorded to have lost no more than 2,367 men, and 9,315 injured from March to May 1938: see Jiang zongtong milu, 11: 136. It is a depressing reminder of the state of Chinese army organization, even after German efforts, that no accurate Chinese casualty figures seem to have been kept, though the Chinese Army headquarters later reported 249, 213 killed and 485, 804 injured in the whole of 1938. Ba'nian kang Rixiezhan shi, p. 512. (The Nationalists in the Huai-Hai Campaign of 1948 were to suffer a fate worse than the Russians at Tannenberg, as a result of rapid Communist envelopments of would-be reinforcements for their Xuzhou garrison.)

87. Clark-Kerr, GBFO 371/23044:267, called the Chinese tactics “offensive … little attention to retention or loss of a city or barricade … main idea to draw enemy forces deep into Chinese trap.”

88. Phillips, Herbert, Consul General, Shanghai, despatch of 6 12 1937, GBFO 371/11643, pp. 43–44Google Scholar. Cf. Zhang Fakuei's analysis in note 77 above.

89. Phillips, loc cit.

90. Tsung-jen, Li, Memoirs, p. 357.Google Scholar

91. Tsung-jen, Li, Memoirs, pp. 344–54Google Scholar. A former staffer of Pang's claims that Li was mistaken, and that Pang and Zhang had old ties and were close friends, having fought the Japanese side by side in a previous engagement. But his very denial points to the personal relationship as a probable factor for Zhang being ordered to help Pang by the high command. See Li Fengming memoir in Zhuanji wenxue, 32 (June, 1978), 64 and Quo Xueyu quoting Li Zongren in Ibid. 31: (May, 1977), 75–78.

92. Noting that the “co-ordinating command which had proved so successful at Taierh-chuang was non-existent in the other war areas,” Liu, F. F., Military History, p. 201Google Scholar, cites one case of failure attributable to personal militarism. In May and June the Doihara division escaped destruction near Lanfeng (Henan), and General Gui Yungqing, former Lehrbrigade commander, along with the replenished 87th and 88th divisions, effected “heroic resistance” without the support “for some reason” of the nearby forces of Hu Zongnan. This cost “irreparable damage in trained soldiers and materiel.” (A very different picture is presented in the offical Kang Ri zhanshi, p. 72, map after p. 78.) See also Cohen, Warren I., “Who fought the Japanese in Hunan: some views of China's war effort,” Journal of Asian Studies, 27, 1 (11, 1967), 111–15CrossRefGoogle Scholar, showing that the fierce Fighting in Hunan in late 1943 and 1944 was undertaken by General Xue Yue without support from Chungqing, and Sheridan, James E., China in Disintegration (New York, 1975), p. 203Google Scholar, on the importance of personal relations in the cohesion of wartime armies.

93. Lovat Fraser, “Hsiichow and after,” in Clark-Kerr despatch, 31 May 1938, GBFO 371/22044, p. 30. He continued with a critique of high Chinese military leadership that pointed to a mixture of military backwardness with militaristic behaviour. “To make matters worse the Chinese have yet to produce leaders worthy of the name. My experience has been that when leaders find themselves in a tight place, they bolt and leave the wretched troops to fend for themselves. The Chinese soldier knows this and fights accordingly – in other words he is always glancing backwards towards his avenue of escape. Not very good for morale! Chinese tactics in the field are of the most hazy description. Co-operation as we envisage the word hardly exists. In consequence divisions are pushed out with incomplete orders and told to hold a certain locality. For some reason best known to themselves, Chinese troops like to have the protection of a walled city. Their liaison with other divisions is of a most nebulous variety. Add to this the language difficulty; the difficulty of communicating orders; personal animosities; lack of good equipment, particularly artillery and anti-aircraft guns; lack of good communications and finally, the inequality between the two forces both in the air and on land, and it's then possible to appreciate the difficulties under which the Chinese have been fighting.” Ibid. pp. 30–31.

94. After Taierzhuang, Falkenhausen's optimism was grounded on the new small-scale warfare (Kleinkrieg), see Anon. (Falkenhausen?), 11 April 1938, DMAF Msg 160/8. “Japan,” the writer believed, “will press forward in China, but it is impossible that Japan will really conquer a single province or overthrow the Chinese Wehrmacht.” He justified his earlier belief that the Japanese were unprepared for war by arguing that they had seen how rapidly the Chinese were strengthening themselves, and decided on a preventive strike. 1 February 1938, Msg 160/8. Seps, “German Advisers,” p. 473, believes that Falkenhausen's encouraging reports to Berlin helped to prolong German aid. In the first half of 1938, in an exchange agreement whereby China returned raw materials, Germany supplied over 70% of the bombs and shells imported by sea, 30% of rifle and machine gun ammunition (113·15 million rounds), 25% of machine guns, and was the main source of supply for explosives, anti-tank guns, howitzers, and handguns. (Information from the Government of Hong Kong.) GBFO 371/11075: 257–62.

95. Xiang Ying, summary of remarks to “an American pressman,” in USWD 2009/259/1, forwarded by Stilwell, , 29 03 1939Google Scholar. Another irony here is that Xiang and 9,000 men of his New Fourth Army were wiped out in perhaps the most successful positional encirclement ever carried out by the Nationalists, in January 1941. The view that German advisers were responsible for the Nationalists' fixed-position and trench-warfare mentality is found in many secondary works, e.g. Bianco, Lucien, Origins of the Chinese Revolution, 1915–1949 (Stanford, 1971; French ed., 1967), p. 18Google Scholar

96. Liang, , Sino-German Connection, p. 133Google Scholar; Jiang Zongtong milu 11:60, diary excerpt, dated 7 November (1937).

97. Yifei, Liu, ed., Kang Ri yingxiong texie (Hankou, 1938Google Scholar; reprinted Washington, D.C., 1969), pp. 62–67. Chiang acknowledged such criticism in a speech at Hankou, on 6 07 1938.Google ScholarChinese Ministry of Information, comp., The Collected Wartime Messages of Generalissimo Chiang Kai-shek, 1937–1945. (New York, 1946), 1: 71.Google Scholar

98. [Falkenhausen], “Judgment of the Japanese forces on the basis of the battles since July 1937,” DMAF Msg 160/8.