I finally got around to ordering a copy of Jack Chen's The Sinkiang Story. I was incorrect when I said in the original thread that Chen was only using secondary sources. He actually claims in the book to have conducted "scores" of interviews with Xinjiang residents. (Chen, p. 208.) The interviews were presumably conducted between 1957, when Chen says he first visited the province, and the mid-1970s, as the book was published in 1977. However, (a) he doesn't list the specific interview subjects in the bibliography (which is why I thought he was relying entirely on secondary sources from what I could see of the book on GoogleBooks' preview), and (b) he doesn't use either footnotes or endnotes, making it impossible in most situations to identify the specific source for his account. That said, here is his description of the Nationalists' use of armor at Wusu (which Chen refers to as "Hsiho") on or about September 4, 1945 (Chen gives September 6 as the date Wusu fell to the insurgent forces and the narrative appears to place the encounter with the tanks two days before that):
(Chen, pp. 233-234.)Hsiho was defended by some eight thousand Kuomintang troops armed with submachine guns in considerable numbers, light and heavy machine guns, some artillery and -- this was not known when the attack started -- a couple of light tanks.
The attack was mounted by the Suidun Infantry Regiment, the Sixth Independent Kazakh Cavalry Regiment, the Mongolian Cavalry Squadron and the motorized battalion. Though this force numbered three thousand men, none of these units was fully armed.
* * *
The first day the Suidun fighters advanced with spirit and according to all the lessons they had learned in their months of training. Their attack was well delivered and forced the Kuomintang back from their forward defense lines. The next day the elated fighters prepared to repeat their performance. They advanced straight down the toad and its flanks in good order, when suddenly the Kuomintang unmasked its tanks. This took the Suidun men completely by surprise. They had been taught how to deal with "tanks" simulated by carts covered with paper, but the real thing spitting fire caught them off balance. The Kuomintang counterattack threw some of the raw recruits into confusion and a disorderly retreat commenced. The tanks and enemy riflemen took full advantage of this near rout and mowed down men who had given up cover and were running in bewilderment. Hard-fought-for ground was lost in a matter of minutes. The Suidun veterans, however, kept their heads and brought up armor-piercing bullets and an anti-tank rifle. With his first shot their Uighur gunner Saud killed the driver of the first tank and put it out of action. The second tank, attempting to take evasive action, overturned in a ditch. It was captured intact. The situation was saved.
The description unfortunately doesn't do much to answer the question of what kind of tanks were used by the Nationalists, but the description of panic amongst the insurgents upon seeing the vehicles (and since Chen is clearly sympathetic to the insurgents in his writing, it seems unlikely that he would overstate their fearful reaction) does incline me toward thinking that these were actual tanks rather than armored cars. My admittedly unscientific reasoning is simply that I'd think tracked AFVs would be much more intimidating to unsophisticated troops than armored cars, especially given that the insurgent forces in this battle were already familiar with regular wheeled motor vehicles. I draw that latter conclusion from the fact that the "motorized battalion" referred to by Chen in the passage above is described by him elsewhere as consisting of captured Nationalist trucks. (Chen, p. 228.)
I'll also add that, not withstanding Chen's pretty strong partisan bent in his writing, the book probably has the best English language account of battles at the tactical level during the Yili Rebellion/Three Districts Revolution. (The other books on the subject, Under the Soviet Shadow: The Yining Incident: Ethnic Conflicts and International Rivalry in Xinjiang 1944-1949 by David D. Wang, The Ili Rebellion: The Moslem Challenge to Chinese Authority in Xinjiang, 1944-1949 by Linda Benson, and Clouds Over Tianshan: Essays on Social Disturbance in Xinjiang in the 1940s also by David D. Wang, focus more on the operational/strategic level when discussing the military aspects of the conflict.)
Conversely, Chen's account of the battles fought in Xinjiang during the early 1930s is worthless. His discussions of the 1931-32 and 1933-34 uprisings and Ma Zhongying's invasions during the same periods are cursory. For example, the entirety of the founding and destruction of the Turkish Islamic Republic of East Turkestan is covered in literally just two paragraphs. Chen also omits any reference to Soviet involvement in the 1933-1934 battles, which frankly has to be willful on his part because Chen includes Sven Hedin's The Flight of Big Horse, Aitchen K. Wu's Turkistan Tumult, and Peter Fleming's News From Tartary in his bibliography, all of whom discussed the Soviet intervention to one degree or another.