Regional Timeline · 廣東 / 廣西 抗戰
Homeland Defense — Wu Kun's South China War, 1938–1945
吳鯤 · 第九五八團 · 第一六〇師 — 廣東與廣西戰場
This page collects the key dates of the Second Sino-Japanese War as they were felt in Wu Kun's home region — the corridor from Guangzhou (廣州) and the Pearl River delta, west along the Leizhou Peninsula (雷州半島) and the Canton Bay (廣州灣) coast, and inland through Guangxi to Kunlun Pass (昆侖關), Nanning, and Guilin. The basic chronology is taken from the campaign essays in Mark Peattie, Edward Drea & Hans van de Ven, eds., The Battle for China — Essays on the Military History of the Sino-Japanese War, 1937–1945 (Stanford University Press, 2010); local Cantonese detail is corroborated from 《粵軍史實紀要》, 《廣東文史 資料》, and the 廣州灣歷史研究 series.
What the timeline makes plain is that Wu Kun was not fighting in an abstract national theatre. He was defending the coast where he actually lived: Leizhou, the Leizhou Peninsula, Guangzhouwan (his family's home), and Shaoguan — the wartime Cantonese capital. The war reached his door on 5 February 1938 when Japanese bombers first hit Guangzhou; it left it on 28 September 1945 when the surrender north of the 16th parallel was signed in Hanoi by his own chain of command.
The fight was hard for one reason above all others: China in 1937 was still a pre-industrial society. The Cantonese 粵軍 went into the war with local-pattern Hanyang and Mauser-derived rifles, almost no organic artillery above battalion mortar level, no native armour, no air force of its own, and a logistic tail that depended on coolie porterage and two single-track railways. Against the mechanised Japanese 5th and 18th Divisions, infantry like the 958th Regiment paid the price of that gap in blood — at Wanjialing, at Kunlun Pass, on the East River, and in the Guangxi corridor of Ichi-Gō.
Filter by period
Drag either thumb to scope the map's Japanese advance routes and the timeline below.
Map · 地圖
Guangdong & Guangxi — the regional theatre of the 958th Regiment / 160th Division. Click any marker for the date and significance.
Wu Kun's Theatre — Guangdong & Guangxi, 1938–1945
吳鯤的戰場 · 廣東與廣西 1938–1945
- 1938 Canton Operation — Bias Bay → Guangzhou (12–21 Oct 1938)
- Beijiang corridor — Guangzhou → Shaoguan (1939–40; seized 26 Jan 1945)
- South Guangxi Campaign — Qinzhou Bay → Nanning → Kunlun Pass (Nov–Dec 1939)
- Leizhou coast — Japanese landings & Guangzhouwan occupation (Feb 1942 – Feb 1943)
- Operation Ichi-Gō — Hunan → Guilin → Liuzhou → Nanning → Indochina (Aug 1944 – early 1945)
Solid lines mark a single named operation or landing; dashed lines mark a corridor used across multiple campaigns. Arrowheads show the direction of Japanese advance.
Red dots mark Japanese landings, occupations, and the leading edge of advance. Blue dots mark the Cantonese 粵軍 strongpoints in which Wu Kun's 958th Regiment / 160th Division served — Kunlun Pass (Guangxi, 1939), the Shaoguan / Northern Guangdong line (1939–45), the East River and Yunfu sectors (1940–43), and the 1944–45 Guangxi corridor. Yellow dots mark Wu Kun's family home country on the Leizhou coast (Gaozhou, Guangzhouwan, Xuwen).
Forces & weapons · 兵力與裝備
Order-of-battle and equipment for the three South China engagements that define Wu Kun's war — Canton (Oct 1938), South Guangxi / Kunlun Pass (Nov 1939 – Jan 1940), and the Guangxi phase of Operation Ichi-Gō (Aug 1944 – early 1945). Figures are taken from the campaign essays in The Battle for China (Peattie, Drea & van de Ven, 2010), Hattori's 《大東亞戰爭全史》, and the Japanese Bōeichō Senshi Sōsho 《戰史叢書》 volumes for the China theatre. Round numbers are deliberate — wartime returns conflict between Japanese, KMT, and post-1949 PRC sources.
Canton Operation
12 – 21 Oct 1938 · Guangdong coast
- Japanese 21st Army (Furushō Motoo)
- ~70,000 men · 5th, 18th & 104th Divisions, plus Mixed Bde elements; supported by 5th Fleet carrier and land-based naval air.
- Chinese 12th Army Group (Yu Hanmou 余漢謀)
- ~80,000 on paper, under-strength — the best Cantonese divisions (64th & 66th Armies, incl. Wu Kun's 160th Division) had been sent north to Wuhan. Guangzhou fell in nine days.
South Guangxi / Kunlun Pass
Nov 1939 – Jan 1940 · Guangxi · Wu Kun present
- Japanese 21st Army southern wing
- ~30,000 — 5th Division (Imamura, the "Steel" division 鋼軍), Taiwan Mixed Brigade, supporting naval air from Hainan.
- Chinese counter-attack force (Bai Chongxi)
- ~150,000 across 14 divisions — Du Yuming's mechanised 5th Army in the main assault on Kunlun Pass, with the 35th Army Group (incl. Wu Kun's 958th Regiment / 160th Division) and 37th Army Group on the flanks. Costliest Nationalist counter-offensive of 1939.
Ichi-Gō — Guangxi phase
Aug 1944 – Jan 1945 · Hunan → Guangxi → Indochina
- Japanese 6th Area Army, southern wing
- ~80,000 in the Guilin–Liuzhou drive (11th & 23rd Armies elements); part of an Ichi-Gō total of ~500,000 along the full Pinghan–Yuehan–Xianggui axis — the largest Japanese ground offensive of the war.
- Chinese 4th War Zone (Zhang Fakui 張發奎)
- ~70,000 effectives — including the 64th Army (Wu Kun's parent army by this date) committed against the Guilin encirclement. Guilin fell 10 Nov 1944; Liuzhou the same day.
Typical infantry weapons, South China front
What a Japanese rifle company and a Cantonese 粵軍 rifle company actually carried into the same fight. The asymmetry is the story.
| Class | Japanese Imperial Army | Chinese 粵軍 / 958R (typical) |
|---|---|---|
| Service rifle | Type 38 Arisaka (6.5 mm); from 1940 the Type 99 (7.7 mm) | Hanyang 88 (local Gewehr 88 copy, 7.92 mm) and Chiang Kai-shek / Type 24 rifle (Mauser 98 copy) — German and Czech imports thinned out after 1938 |
| Light machine gun | Type 96 / Type 99 LMG | Czech ZB-26 (the 捷克式) — universally praised, but ammunition supply was chronic |
| Heavy machine gun | Type 92 HMG ("woodpecker"), 7.7 mm | Type 24 Maxim (Hanyang-built copy of MG 08), 7.92 mm |
| Grenade / mortar | Type 89 grenade discharger (the "knee mortar") at squad level; 81 mm and 90 mm mortars at company / battalion | Type 27 / Type 20 82 mm mortar at battalion; hand grenades local-pattern, often percussion-fused |
| Infantry gun / artillery | Type 41 75 mm regimental mountain gun; Type 38 75 mm field gun; Type 91 105 mm howitzer at division | Rare — Cantonese divisions usually had no organic gun above 82 mm mortar. A few German 75 mm Krupp pieces survived in the 5th Army; nothing comparable in the 160th Division |
| Armour | Type 95 Ha-Gō light tank, Type 89 medium tank in independent tank brigades | None in the 粵軍. The only mechanised force on the South China front was Du Yuming's 5th Army at Kunlun Pass — Soviet-supplied T-26 and a few BT-5 |
| Air power | IJA Ki-27, Ki-43 fighters and Ki-21 bombers; IJN A5M then A6M Zero, G3M/G4M bombers from Formosa & Hainan | No Cantonese air. ROCAF national strength was thin throughout; American Volunteer Group (Flying Tigers) from late 1941 operated mainly from Kunming, not over Guangdong |
| Logistic transport | Motorised regimental trains, rail along the Guangzhou–Hankou corridor once seized, coastal shipping | Coolie porterage, mules, two single-track railways (粵漢, 湘桂). Field telephones were scarce; runners carried orders at battalion level |
Casualties & combat experience
What the weapon gap above translated into on the ground. Figures hedged — Japanese (Senshi Sōsho 《戰史叢書》), Chinese (Hsu Long-hsuen 徐永昌 ed., 《中華民國重要史料初編 — 對日抗戰時期》, Taipei 1981) and PRC post-1949 figures rarely agree; round numbers are intentional.
| Engagement | Japanese losses | Chinese losses | Outcome & combat lesson |
|---|---|---|---|
Canton Operation 12 – 21 Oct 1938 · Guangdong coast | ~1,900 killed & wounded (Senshi Sōsho 《支那事変陸軍作戦》 II, 「廣東攻略作戦」). Total air losses minimal. | ~10,000+ killed, wounded & captured across the 12th Army Group; figures contested (Hsu Long-hsuen). Coastal artillery and Humen forts surrendered largely intact. | Guangzhou fell in 9 days. The Cantonese rear was severed from the Wuhan front. Lesson: without organic air or coastal artillery answering Japanese naval gunfire and carrier strikes, a defending infantry army could not hold a contested landing — the asymmetry above translated directly into the speed of collapse. |
South Guangxi / Kunlun Pass Nov 1939 – Jan 1940 · Guangxi · Wu Kun present | ~5,000+ killed in the 5th Division & Taiwan Mixed Bde — incl. Maj-Gen. Nakamura Masao 中村正雄, 21st Bde commander (KIA 25 Dec 1939). Heaviest single-battle losses inflicted on the IJA in South China to that date. | ~14,000 killed, ~13,000 wounded across the 35th & 37th Army Groups and the 5th Army (Hsu Long-hsuen). Du Yuming's 200th Division alone took ~50% casualties. | Pass retaken 31 Dec 1939 but lost again Feb 1940. The one South China battle where Chinese armour (5th Army's T-26 / BT-5) achieved local parity — and the only one where the casualty ratio approached 1:3 rather than 1:5+. Wu Kun was decorated as battalion commander (营长) here. Lesson: mechanisation and concentration could close the gap, but only at a cost the Nationalist economy could not sustain. |
Ichi-Gō — Guangxi phase Aug 1944 – Jan 1945 · Hunan → Guangxi → Indochina | ~10–15,000 killed & wounded across the Guilin–Liuzhou drive (Senshi Sōsho 《一号作戦》). Losses lighter than Kunlun despite a far larger force — combat power had widened, not narrowed. | ~50,000+ casualties across the 4th War Zone; Guilin & Liuzhou both fell 10 Nov 1944. Most US 14th Air Force forward bases (Guilin, Liuzhou, Nanning) lost or evacuated — the rationale Stilwell had used to demand control of Nationalist forces collapsed with them. | Strategic catastrophe. Six years after Canton, the gap was wider, not closer: Cantonese divisions still had no organic armour, no air, and now had lost Lend-Lease priority to the Burma front. Lesson — sustained under-investment compounds. Wu Kun's 64th Army fought through this corridor and into the 1945 Nanning recovery. |
This is not a complete order-of-battle. It is the gap that Wu Kun's battalion fought across: rifles roughly comparable, machine guns roughly comparable, and from artillery upward almost everything owned by the other side. The casualty ratios above — roughly 1:5 at Canton, 1:3 at Kunlun Pass (the closest the 粵軍 ever got to parity), and back toward 1:4–5 by Ichi-Gō — are what that gap looked like in killed and wounded men.
Before the homeland: Nanjing breakout · 南京突圍 1937年12月
Before the Guangdong campaigns covered above, the same Guangdong formations — 66th Corps (第六十六軍, Ye Zhao 葉肇) and 83rd Corps (第八十三軍, Deng Longguang 鄧龍光), with the 160th Division inside the 83rd — were thrown into the defence of Nanjing in December 1937. How they came out of Nanjing is the reason there was a 160th Division left to defend Guangdong at all.
A three-month march, late to Shanghai
The Guangdong corps marched roughly three months north to reinforce the Shanghai front. By the time they arrived, Shanghai had already collapsed and the line was falling back on Nanjing. They were committed to a defence that was being improvised in real time, under a garrison commander — Tang Shengzhi 唐生智 — with no operational experience of a modern envelopment battle.
- DocumentedGuangdong corps’ northward redeployment and late arrival to the Shanghai–Nanjing axis. Hsu Long-hsuen 《中華民國重要史料初編 · 對日抗戰時期》; Yang Tianshi on the Nanjing garrison.
- Family testimonyThe three-month figure and the felt experience of arriving too late.
Defend or retreat — the order kept changing
Inside Nanjing the command oscillated: hold the city to the last, then break out, then hold again. By 12 December 1937, when the order to withdraw finally came, the Japanese had already closed the perimeter and the Yangtze crossings were a killing zone. Tens of thousands of men were funnelled toward Xiaguan 下關 with no boats, no plan, and no covering fire. That is the chaos in which the Nanjing Massacre began.
- DocumentedTang Shengzhi’s vacillation between defence and withdrawal, and the collapse of the Xiaguan crossing on 12–13 Dec 1937. Yang Tianshi 楊天石; Hsu Long-hsuen; Iris Chang and the Rabe diary as corroborating eyewitness.
Deng Longguang’s dissent — and the route that saved the 160th
Deng Longguang disagreed with Tang Shengzhi’s withdrawal plan, which sent the bulk of the garrison west to the Yangtze. He recommended an alternative: the Guangdong corps would not queue for the river. The 66th and 83rd Corps instead broke out eastward through Taipingmen 太平門 / Zijinshan 紫金山 on the night of 12 December, cutting through the Japanese encirclement toward the northeast. They took heavy losses on the way out — but unlike the units funnelled to Xiaguan, most of the 160th Division (第一六〇師) and a substantial part of the two Guangdong corps came through intact. That is why, less than a year later, there was still a 160th Division to throw into the defence of Guangdong, and why Wu Kun’s 958th Regiment existed at all to fight the campaigns mapped above.
- DocumentedThe 66th and 83rd Corps’ northeast breakout through Taipingmen / Zijinshan on 12 Dec 1937, and their comparatively low losses relative to units that attempted the Yangtze crossing. Hsu Long-hsuen 《重要史料初編》; 《粵軍史話》; Yang Tianshi on the Nanjing breakout.
- Family testimonyThe specific attribution that Deng Longguang dissented from Tang Shengzhi’s route and recommended the alternative is descendant memory. The breakout itself is in the printed record; the dissent narrative is family voice.
- Not yet cross-checkedWhether Deng’s objection is recorded in Tang Shengzhi’s own post-war account or in the Guangdong corps’ war diaries. Flagged for follow-up.
The point is not to relitigate Nanjing. The point is that the unit that fought everything below — Canton, Kunlun Pass, Ichi-Gō — only existed because, in December 1937, its commander refused the route that was getting everyone else killed.
How the unit stayed alive · 士氣與同鄉
Weapons and casualty tables explain why units broke. They do not explain why this one didn’t. The 64th Corps (第六十四軍) under Deng Longguang 鄧龍光 was built largely from new recruits out of Leizhou 雷州 — same peninsula, same dialect, often the same villages. Wu Kun’s 958th Regiment fought inside that fabric. What follows is family memory and regimental practice, marked as such; treat it as testimony rather than archive.
Broth cooked eight hours before battle
Deng Longguang and Wu Kun were known among their men for feeding the regiment properly before an engagement — a heavy broth started eight hours ahead so the men went into combat fed, warm, and not on cold rations. In a war where Chinese infantry usually moved on coolie-carried rice and what they could carry on their back, this was unusual care, and the men remembered it.
- Family testimonyThe eight-hour broth and the specific attribution to Deng Longguang and Wu Kun. Descendant of Wu Kun.
- DocumentedChinese infantry ration shortages and reliance on coolie porterage — Peattie, Drea & van de Ven, The Battle for China (2010), logistics chapters.
- Not in archiveSenshi Sōsho and Hsu Long-hsuen 《中華民國重要史料初編》 do not record this regimental practice.
Simple lyrics, melodies from Leizhou
Songs were written to be easy to remember and carried tunes from home — Leizhou-hua 雷州話 cadences the recruits had grown up with. Singing on the march or in billets gave the men something normal to hold on to: a sound that did not belong to the war. The instinct is the same as soldiers in other armies who smuggled kittens or dogs into the line — one familiar, ordinary thing to keep a person anchored.
- Family testimonyThat songs were written for the unit in Leizhou-hua and sung on the march. Specific lyrics not preserved.
- DocumentedVernacular marching songs and home-dialect ditties were standard morale practice across 粵軍 / Cantonese units; see also the Song page.
- Analogy (ours)The kitten / familiar-object comparison is an editorial framing, not a primary-source claim.
A Leizhou regiment fighting for itself
The 64th Corps was raised largely from Leizhou — new recruits who spoke the same dialect, ate the same food, and in many cases knew each other’s families. Unit cohesion in 1944–45, when better-equipped Chinese divisions were collapsing under Ichi-Gō, came less from training time (there wasn’t any) than from the fact that the man next to you was from your county. You were not fighting for Chongqing or for an abstraction. You were fighting so the people you grew up with would walk off the field too.
- DocumentedThe 64th Corps drew heavily from the Leizhou peninsula. Recruitment and order-of-battle in Hsu Long-hsuen 《中華民國重要史料初編 · 對日抗戰時期》 and 《粵軍史話》.
- DocumentedSame-locality recruitment as a cohesion mechanism is a standard finding in unit-cohesion scholarship (cf. Shils & Janowitz 1948 on primary groups).
- Family testimonyThe lived framing — “the man next to you was from your county” — is descendant memory, not from a printed source.
Colour key:Family testimony = first-person memory passed down within the family.Documented = corroborated in published scholarship or archive.Not in archive / Analogy (ours) = explicit gaps and editorial framings flagged for the reader.
Timeline · 1938–1945
Tagged by phase. Wu Kun present marks engagements his regiment fought. Homeland front marks events inside Guangdong / Guangxi. Global context marks the wider war that shaped the South China supply lines.
1937–38
3 events- 1937–38Homeland front5 Feb 1938· Guangzhou & Leizhou (雷州), Guangdong
Japanese naval bombing of Guangzhou begins
Japanese carrier and land-based air units begin a sustained bombing campaign against Guangzhou (廣州) and the Guangdong coast. On the Leizhou Peninsula the local administration orders the demolition of the old red-brick city wall on the grounds that its colour made the town an easy aiming-point for Japanese bombers — one of several civic acts in which Cantonese authorities physically dismantled centuries-old urban fabric in anticipation of the air war.
- 1937–38Homeland front12 Oct 1938· Bias Bay (大亞灣), Guangdong
Bias Bay landing — Japanese 21st Army
Japanese 21st Army (Lt-Gen. Furushō Motoo) lands three divisions (5th, 18th, 104th) at Bias Bay (大亞灣) east of Hong Kong. Yu Hanmou's 第十二集團軍, kept understrength after the 66th and 64th Armies had been sent north for the Wuhan campaign, is unable to hold the coastal road.
- 1937–38Wu Kun present21 Oct 1938· Guangzhou (廣州)
Fall of Guangzhou
Guangzhou (廣州) falls nine days after the Bias Bay landing — the home prefecture of the entire Cantonese 粵軍 is overrun while its best divisions, including Wu Kun's 66th Army (160th Division, 958th Regiment), are committed at Wanjialing 800 km to the north. The Canton Operation is the strategic pivot that pulls the 160th Division back south within the year and sets the conditions for every later phase of Wu Kun's war.
1939–40
6 events- 1939–40Homeland front24 Nov 1939· Qinzhou Bay (欽州灣) → Nanning
Japanese landing at Qinzhou Bay
The Japanese 5th Division (‘鋼軍’) and elements of the 21st Army land at Qinzhou Bay (欽州灣) and drive inland on Nanning to sever the Indochina–Guangxi supply corridor — China's last good overland route to a friendly port.
- 1939–40Wu Kun present19 Dec 1939· Kunlun Pass (昆侖關), Guangxi
KMT retake Kunlun Pass
Nationalist forces — Du Yuming's mechanised 第5軍 in the main assault, with the 35th Army Group (including Wu Kun's 958th Regiment / 160th Division) reinforcing the flanks — retake Kunlun Pass (昆侖關) in Guangxi after some of the bloodiest close-quarters fighting of the South China front. Wu Kun was a decorated battalion commander (营长) at this engagement.
- 1939–40Global context20 Jan 1940· Haiphong / Hanoi → Kunming
French close the Yunnan–Hanoi railway
Under Japanese diplomatic pressure, Vichy-aligned French authorities close the Yunnan–Hanoi (滇越) railway to Chinese war traffic — the first of the three Nationalist overland lifelines to be cut, leaving only the Nanning road and the Burma Road. The supply-line war that drove Kunlun Pass would now drive every Allied campaign in Burma.
- 1939–40Homeland front24 Dec 1939 – 10 Jan 1940· Conghua – Yingde – Shaoguan, Guangdong
First Battle of Northern Guangdong
Japanese 21st-Army formations from occupied Guangzhou push north up the Beijiang (北江) corridor toward Shaoguan (韶關), the relocated wartime Cantonese provincial capital and Yu Hanmou's 4th War Zone GHQ. The 12th Army Group holds, with elements of the 35th Army Group reinforcing as they become available from Guangxi.
- 1939–40Global context13 Aug 1940· Southern England
Battle of Britain — main phase opens
‘Adlertag.’ The Luftwaffe opens the main phase of the Battle of Britain. Britain's survival through autumn 1940 keeps the Burma Road open to British India and preserves the only remaining western Allied supply axis into China.
- 1939–40Global context18 Oct 1940· Lashio (Burma) → Kunming
Britain reopens the Burma Road
After a three-month closure under Japanese pressure (July–Oct 1940), Britain reopens the Burma Road from Lashio to Kunming. For the next sixteen months it is Nationalist China's only overland lifeline.
1941–43
2 events- 1941–43Homeland front16 Feb 1942· Leizhou Peninsula (雷州半島)
Japanese land on the Leizhou Peninsula
Japanese forces land on the Leizhou Peninsula (雷州半島) in the wake of the Pacific War's opening, extending the occupation southward along the South China coast where Wu Kun's family was living.
- 1941–43Wu Kun present21 Feb 1943· Guangzhouwan (廣州灣) — modern Zhanjiang
Japan occupies Guangzhouwan — Wu Kun's home
Japanese forces occupy Guangzhouwan (廣州灣 / Kouang Tchéou Wan), the French-leased treaty enclave on the Leizhou coast where Wu Kun's family lived. The occupation places his immediate family under direct Japanese administration while Wu Kun himself remains in the field with the 160th Division on the East River front.
1944
2 events- 1944Wu Kun present28 Sep 1944· Guilin (桂林), Guangxi
Japanese encircle Guilin
Japanese forces based on the Leizhou Peninsula advance north and complete the encirclement of Guilin (桂林) in Guangxi — the southern wing of the Hunan–Guangxi (桂柳) phase of Operation Ichi-Gō. Wu Kun's 64th Army is committed against the drive.
- 1944Homeland frontDec 1944· Guangxi–Guizhou border
Ichi-Gō grinds to a halt
The Guangxi–Guizhou phase of Ichi-Gō stalls at the Guizhou border. Japanese forces have driven a continuous corridor from north China to the Indochina frontier but cannot exploit further owing to severe logistic exhaustion — the high-water mark of the offensive.
1945
8 events- 1945Homeland front26 Jan 1945· Shaoguan (韶關), Guangdong
Japanese seize Shaoguan
Japanese forces seize Shaoguan (韶關) — the wartime Cantonese provincial capital and 第七戰區 GHQ that the 12th Army Group and the 35th Army Group had successfully defended in 1939–40. Its late-war fall marks the deepest Japanese penetration of inland Guangdong.
- 1945Wu Kun present27 May 1945· Nanning (南寧), Guangxi
Chinese forces retake Nanning
Chinese forces retake Nanning (南寧), Guangxi — the first major recovery of an Ichi-Gō-lost provincial capital, and the spring-summer counter-offensive into which Wu Kun's 64th Army is committed.
- 1945Global context28 Jun 1945· Manila, Philippines
MacArthur announces victory in the Philippines
General Douglas MacArthur declares the Philippines Campaign concluded — Luzon and the Southern Philippines are cleared, freeing US naval and air power for the projected invasion of Japan.
- 1945Global context10 Jul 1945· Japanese home islands
Final strategic bombing of Japan opens
US Twentieth Air Force and US Navy Fast Carrier Task Force open the final, intensified phase of strategic bombing against the Japanese home islands — area bombing of major cities continues until 14 August.
- 1945Global context8 Aug 1945· Moscow
Soviet Union declares war on Japan
The Soviet Union declares war on Japan, abrogating the 1941 Soviet-Japanese Neutrality Pact.
- 1945Global context9 Aug 1945· Nagasaki, Japan / Manchuria
Nagasaki bomb; Soviet Manchurian offensive
A US B-29 drops a plutonium implosion bomb on Nagasaki. The same day, the Soviet Red Army launches the Manchurian Strategic Offensive across three fronts into Japanese-occupied Manchuria — within three weeks all Manchuria is in Soviet hands and the Kwantung Army's arsenals will pass, in part, to Lin Biao's forces.
- 1945Global context15 Aug 1945· Tokyo
Japan surrenders unconditionally
Emperor Hirohito's Imperial Rescript on the Termination of the War is broadcast (玉音放送). Japan accepts the Potsdam Declaration and surrenders unconditionally — V-J Day.
- 1945Global context2 Sep 1945· USS Missouri, Tokyo Bay
Formal surrender on USS Missouri
The Instrument of Surrender is signed on the deck of USS Missouri in Tokyo Bay. Foreign Minister Shigemitsu signs for Japan; General Hsu Yung-chang (徐永昌) signs for the Republic of China — the legal framework under which the regional surrenders at Nanking (9 Sept), Guangzhou, Suixi (21 Sept), and Hanoi (28 Sept), all involving the 64th Army's chain of command, were executed.
After surrender: why the Nationalists lost the mainland · 1945–1949
The standard PRC narrative explains 1949 as a moral verdict — corrupt Nationalists, virtuous Communists, inevitable victory. The operational record is narrower and more material. Two decisions between August 1945 and 1948 did most of the work.
1. The Soviets handed Manchuria’s arsenal to the CCP
When Japan surrendered in August 1945, the Kwantung Army (関東軍) in Manchuria surrendered not to the Nationalists but to the Soviet Red Army, which had entered the northeast a week earlier. Soviet forces occupied Manchuria until May 1946 and, during that window, transferred large quantities of Kwantung Army weapons — rifles, machine guns, artillery, and the Mukden / Shenyang arsenal infrastructure — to Chinese Communist forces moving north under Lin Biao. The Nationalists’ best divisions were still in Yunnan, Burma, and the south, and the US airlift could only move a fraction of them north in time. By the time they arrived, the weapons were already on the other side.
- DocumentedSoviet transfer of Kwantung Army stocks to CCP forces, 1945–46: Odd Arne Westad, Decisive Encounters: The Chinese Civil War, 1946–1950 (Stanford, 2003); Harold Tanner, The Battle for Manchuria and the Fate of China (Indiana, 2013); Steven Levine, Anvil of Victory (Columbia, 1987).
2. The US arms embargo cut one side only
In July 1946, General George Marshall imposed an arms embargo on the Nationalist government to pressure Chiang Kai-shek into a negotiated settlement. It was formally lifted in May 1947, but US deliveries remained slow and restricted through the 1948 China Aid Act and into 1949. There was no equivalent pressure on the Soviet pipeline into the CCP. The practical effect was asymmetric: the side that had already lost the Manchurian arsenal was also the side whose external resupply was throttled. By 1948 the PLA had grown from roughly one million to over four million — much of it raw peasant recruitment armed with captured Japanese and Soviet-handed-over weapons.
- DocumentedMarshall embargo (Jul 1946 – May 1947) and the 1948 China Aid Act: US State Department, United States Relations with China (the “China White Paper”, 1949); Daniel Kurtz-Phelan, The China Mission: George Marshall’s Unfinished War, 1945–1947 (Norton, 2018).
- DocumentedPLA growth from ~1.2M (1945) to ~4M (1948): Westad, Decisive Encounters; PLA official strength figures in 《中國人民解放軍戰史》.
- HedgedThe specific figure of “900,000 farmers armed” is a working magnitude; secondary sources vary on the exact count of newly armed peasant recruits between 1945 and 1947.
For Wu Kun’s 64th Corps, this is the missing third act. They had spent eight years fighting the Imperial Japanese Army in the south on Hanyang rifles and Czech machine guns. After August 1945 most of those troops were still in Guangdong and Guangxi when the centre of gravity moved north — to Manchurian cities where the weapons they had just helped take from Japan were being handed to the army that would defeat them. The collapse of 1949 was not, in the first instance, a moral verdict. It was a logistics and timing problem with an embargo attached.
This section is included because the standard PRC narrative erases the role of Soviet weapons transfers and the US embargo — both of which materially shaped the outcome — and credits the result entirely to political virtue. The point is not to absolve the Nationalists of their many failures of governance after 1945. It is to restore the two facts that get edited out.
Lesson · 教訓
One pattern shows up across this page and across centuries of warfare on Chinese soil: you do not need the best technology in the world. You need technology that is better than your opponent’s, in the specific theatre where you are fighting them.
Mongols vs. Song
The Mongols did not have the most advanced civilisation of their age — the Song did, by almost any metric: metallurgy, gunpowder, hydraulics, printing, finance. The Mongols had the better operational stack: composite recurve bows, remount-based cavalry logistics, siege engineers absorbed from conquered Khwarazmian and Jin specialists, and a command system that learned faster than the enemy could reorganise. The gap that mattered was not absolute; it was differential, and it was at the point of contact.
Japan vs. China
Japan was not the world’s leading military power — the United States, Britain, and Germany all out-built it. But across the South China battles mapped above, Japan had what China did not: standardised Arisaka rifles in volume, Type 92 heavy machine guns, Type 41 75mm infantry guns, Type 95 Ha-Gō tanks, Ki-27 / Ki-43 air cover, motorised rail logistics. China fought with Hanyang 88s, scattered Czech ZB-26s, almost no armour and almost no air. The weapons tables higher up on this page are exactly that gap. The men compensated with ground, time, and lives — see the casualty ratios.
AI and the same rule
The rule generalises beyond warfare. In any contested deployment of new technology — AI today included — you do not need the frontier model. You need a deployed stack that is meaningfully better than the one your opponent is actually using in the field. Differential, not absolute, and at the point of contact. The Song lost because the Mongols out-deployed them with worse base technology. The Republic of China lost ground because Japan out-deployed it with technology that was not world-leading. The lesson for any contemporary competition — commercial, strategic, or otherwise — is the same: ship better than the other side ships, in the theatre that decides the outcome.
Don’t let the press mood decide for you
A corollary to the rule above: do not let the prevailing negative emotional tone in the press about AI override a sober assessment of what the tool actually does. The system was built to solve problems. Much of it was designed and trained by scientists. It runs probability calculations over simulated possibilities and selects among them — which is, structurally, a version of what humans do when we reason under uncertainty. Treating it as taboo because the discourse is anxious is its own failure mode: it cedes the differential advantage to whoever is willing to use the tool seriously while you hesitate. The historical pattern on this page is consistent — the side that under-deployed available technology lost ground it did not need to lose.
This is an editorial opinion, not a historical claim. It is offered as the author’s reading of the lesson, not as a finding of any cited source.
This is an editorial reflection, not a historical claim from primary sources. The Mongol–Song and Sino-Japanese comparisons are conventional military-history readings (cf. Morris Rossabi on the Yuan; Peattie / Drea / van de Ven on 1937–45). The application to present-day technology deployment is the author’s own.
Sources for this page
- Mark Peattie, Edward J. Drea & Hans van de Ven (eds.), The Battle for China — Essays on the Military History of the Sino-Japanese War, 1937–1945 (Stanford University Press, 2010).
- Hattori Takushirō, 《大東亞戰爭全史》 (Tokyo, 1953).
- 《粵軍史實紀要》 and 《廣東文史資料》 (Guangdong provincial historical materials series).
- 廣州灣歷史研究資訊 / La Correspondance des Études de Kouang Tchéou Wan — for the Guangzhouwan occupation (Feb 1943) and the Suixi surrender (21 Sept 1945).
- Barbara Tuchman, Stilwell and the American Experience in China, 1911–1945 (Macmillan, 1971) — for the Burma Road and the Stilwell–Chiang axis context.
For the granular phase-by-phase narrative of Wu Kun's own service — with quotes, casualty tables, and family attestation — see the main timeline. For editorial method and confidence marking see methodology.