Read history

The ninth day of the ninth month, 690

On the ninth day of the ninth lunar month — 16 October 690 — Wu Zhao proclaimed the dynastic change from Tang to Zhou, and changed the era name to Tianshou, "Heaven-granted."

She did not ascend a Tang throne that day. She replaced the dynasty with one of her own, conventionally called Wu Zhou to distinguish it from the ancient Zhou.

Three days to the title

The formal accession — the title "Holy and Divine Emperor," together with the demotion of her son Ruizong to imperial heir — followed on 19 October.

Popular timelines merge the two steps. Until the 19th, Ruizong formally retained imperial dignity.

The groundwork

The change had been argued for in advance: the official Fu Youyi led an earlier petition campaign urging that the state name be changed to Zhou.

A commentary on the Great Cloud Sutra, associated with Faming and Xue Huaiyi, presented Wu's rule as cosmically ordained and identified her with a prophesied Buddhist ruler.

The formal edict placing Buddhism above Daoism came later, in early 691.

March 749, at Jabalq

In March 749, Qahtaba ibn Shabib's Abbasid revolutionary army attacked Amir ibn Dubara's Umayyad "army of armies" at Jabalq, in the district of Isfahan.

Qahtaba reportedly raised a copy of the Quran on a lance and invited the Syrian troops to accept its authority. They answered with abuse, and he ordered the attack.

The Umayyad army was routed, Amir ibn Dubara was killed, and Qahtaba's men captured extensive spoils from its camp.

Numbers the sources cannot agree on

Al-Tabari's narrative reports 50,000 sent with Ibn Dubara to Isfahan, then gives 100,000 or even 150,000 at the battle itself, and puts Qahtaba's own force at 20,000.

These are claims within a tradition shaped under Abbasid rule, not measurements. After Jabalq, Qahtaba was reinforced with a reported 15,000 recruited from Khurasani villages.

Kufa taken without its commander

Qahtaba advanced through Nihavand and western Iran toward Iraq, aiming to link up with the Abbasid revolutionary organization in Kufa.

In August 749 he crossed the Euphrates and successfully surprised the Iraqi governor Yazid ibn Hubayra — then disappeared during the fighting and apparently drowned. Abbasid forces occupied Kufa two days later, under his son Hasan.

Abu al-Abbas was not publicly proclaimed caliph there until late October 749.

A window, not a date

At some point after 20 June 751 and before 1 March 752, Pepin ceased styling himself mayor of the palace and began appearing in legal records as king of the Franks. Those two charters are the firmest evidence for the transition.

November 751 is the conventional date for the Frankish assembly at Soissons, but it comes from a traditional reconstruction rather than from contemporary documentation.

The tonsured king

Later Frankish narratives place Pepin’s elevation at Soissons, where the Franks selected or raised him as king, and say Childeric III was tonsured and sent to a monastery — removing him from Frankish kingship without a recorded battle or execution.

The traditional reconstruction identifies Saint-Bertin as that monastery. Both details are less secure than the dynastic transfer itself.

The papal blessing may be retrospective

The Royal Frankish Annals say Fulrad and Burchard carried Pepin’s question about powerless kingship to Pope Zachary, who approved the principle that the possessor of royal power should bear the royal title. Contemporary Roman records are silent, and modern source criticism cautions that this may be retrospective legitimation.

Whether Pepin was anointed at all in 751 is now disputed. Conor O’Brien concludes that Boniface certainly did not perform such an anointing; the first securely evidenced anointing was Pope Stephen II’s ceremony in 754.

Almost all the narrative accounts of the transfer were composed after Pepin’s death, when his sons had an interest in presenting his accession as orderly and legitimate.

A forged edict at Fanyang

On the jiazi day of the eleventh month of Tianbao year 14 — usually rendered 16 December 755 — An Lushan mobilized his own troops together with Tongluo, Xi, Khitan and Shiwei forces and openly rebelled at Fanyang, near modern Beijing.

He displayed a forged imperial order claiming that Emperor Xuanzong had commanded him to march against the chancellor Yang Guozhong. His assembled officers did not openly object.

Marching by night

An left deputies to hold Fanyang, Pinglu and Datong while his field commanders led the rebel columns out during the night. The Old Book of Tang says the mixed cavalry and infantry force travelled at night, stopped for food near dawn, and covered 60 li a day.

The Zizhi Tongjian puts the army at 150,000 troops while An’s side claimed 200,000 — medieval chronicle reports, not audited headcounts.

A court caught unready

When the news reached the Tang court, officials improvised recruitment, opened the treasury to issue silk and cloth, and appointed generals including Feng Changqing and Gao Xianzhi to organize resistance. The Old Book of Tang says peace had lasted so long that some palace guards were marketplace merchants.

The Chinese date is secure as the jiazi day, eleventh month, Tianbao 14. Its usual modern rendering is 16 December 755, though some scholarship gives 20 December, because pre-1582 Julian and proleptic-Gregorian conventions differ.

Breaking ground on the City of Peace

What is conventionally commemorated on 30 July 762 was not a caliph moving house. It was the ceremonial foundation, or commissioning, of a city that did not yet exist.

Al-Mansur, the second Abbasid caliph, sited his new administrative centre near where the Sarat canal system met the Tigris. He gave it the official name Madinat al-Salam, the City of Peace. Astrologers including Nawbakht and Umar ibn al-Farrukhan chose an auspicious moment to begin the work.

Which day, exactly, remains contested

The 30 July date is widely repeated but not settled. Published reconstructions also give 31 July 762 — one specifying about 1:57 in the afternoon — and 1 Jumada 145 AH, equivalent to 2 August 762.

Part of the spread comes from converting early Hijri dates; part from disagreement among the narrative traditions themselves. The foundation, the caliph's own relocation, the completion of the Round City and the transfer of the administrative departments occurred in separate stages.

What actually existed by nightfall

Nothing like a completed capital existed by nightfall. Construction on the palatial settlement later known as the Round City ran from roughly 762 to 766 or 767, and the palace, mosque and administrative buildings all needed time before the caliph could live among them.

Al-Mansur transferred his residence around 763 or 764, after the initial building work. The name Baghdad, meanwhile, was older than his city and belonged to the area before he arrived; it outlived the official City of Peace.

A robe at Chenqiao

A report that Northern Han and Khitan forces were invading sent the Later Zhou palace army north out of Kaifeng under its commander, Zhao Kuangyin. At Chenqiao his officers acclaimed him emperor and, in the official narrative, placed a yellow imperial robe over him.

Most modern interpretations treat the coup as organized rather than spontaneous, and the self-exculpatory robe story cautiously. The traditional date of 4 February 960 is best kept for the formal start of Zhao's reign rather than for every stage of the affair, which unfolded over several days.

The condition he set before turning back

Zhao accepted his army's support on terms: the child emperor of Later Zhou, the empress dowager, the ministers, the population and the treasuries were not to be harmed or plundered.

He then led the troops back into Kaifeng. The Later Zhou court submitted, Zhao took the throne, and the capital changed dynasties with little fighting inside it.

What did not happen that day

Zhao adopted the dynastic name Song and the reign title Jianlong, and ruled from Kaifeng. What he did not do was reunify China: several southern states and Northern Han remained independent, and Northern Han survived until his brother conquered it in 979.

The Chinese Text Project's chronology puts the Chenqiao acclamation and accession sequence around 3 February, while its reign-era data begin Zhao's rule on 4 February 960.

Five books, finished on the road west

The year usually attached to Avicenna's Arabic Kitab al-Qanun fi al-tibb is 1025, though nothing about the Canon happened on a particular day. He began it in Gurgan (Jorjan) in northern Iran, continued it in Rayy, and completed it in Hamadan.

The finished encyclopedia was organized into five books: general principles, simple drugs, diseases of individual organs, diseases affecting more than one part, and compound remedies.

The Library of Congress counts roughly 800 individual drugs arranged in Book 2 and some 650 compound remedies in Book 5.

Why 1025 is a range rather than a year

Reputable accounts give the completion variously as 1025, circa 1025, or somewhere between 1020 and 1025. Other scholarly accounts put it more than a decade earlier — c. 1012, or approximately 1010–1015, with the first of the five books finished in 1015. No dated colophon settles the conflict.

The National Library of Medicine's catalogue describes the work as composed over an extended period while Avicenna moved from Gurgan to Rayy and then Hamadan.

'Publishing' is the wrong word for what happened

There was no publication event to date. No identifiable 1025 publication, edition or public release is documented, because the Canon originated in manuscript culture.

Its extensive copying, translation and institutional influence belong to the work's later reception history. Gerard of Cremona translated it into Latin at Toledo in the second half of the twelfth century, and European medical education took the work up after that.

What emerged in Hamadan was a completed five-book Arabic manuscript work.

What Zhu Yu actually wrote down

No shipboard episode is identifiable in the calendar year 1117. The date belongs to a book: Zhu Yu's Pingzhou Ketan, a Song-dynasty miscellany whose composition is usually assigned to the window 1111–1117.

In its maritime passage, seagoing-ship pilots follow coastal features, steer by the stars at night and the sun by day, and consult a south-pointing needle in overcast or dark weather.

Zhu Yu names no inventor and no particular voyage; the people described are unnamed pilots.

1117, 1119, and which date the record belongs to

1117 is the end of the stated composition window, not a dated first use. The three-volume work was published in 1119, which is why the record is sometimes dated to that year instead.

Either way the passage is evidence that the practice already existed. It is the earliest explicit surviving textual record of a magnetic compass used for maritime navigation, not an account of the compass being invented or used for the first time.

Shen Kuo's earlier account, dated about 1088, describes magnetizing a needle with lodestone and suspending it, but does not explicitly document use aboard ship.

What the passage does not say

The text does not describe how the needle was mounted. Calling Zhu Yu's device a pivoted, suspended or floating compass goes beyond his terse words.

It also does not describe the needle replacing anything. The pilots keep using stars and sun, and turn to the compass when the sky disappears.

A recent scholarly reassessment compares earliest maritime-compass dates of about 1100 in China, 1187 in Europe and 1242 in the Arabic world, and rejects a simple one-way chain of transmission between them.

9 January 1127: the Xuanhua and Tongjin gates

On the twenty-fifth day of the intercalary eleventh lunar month — converted to 9 January 1127 — Jin troops broke through Kaifeng's defenses at the Xuanhua and Tongjin gates and entered the city.

The fall came after more than twenty days of the second Jin siege. Two armies had converged on the capital: Wanyan Zongwang's eastern force and Wanyan Zonghan's western one, the latter arriving after taking Taiyuan.

Divine soldiers

Contemporary accounts blame the collapse partly on Guo Jing, whose promised supernatural soldiers persuaded Song leaders to withdraw or weaken the regular troops holding parts of the outer wall.

Those accounts were produced under the Southern Song and were politically invested in explaining the catastrophe; their claims should be cross-checked rather than read as neutral reporting.

What is not in doubt is that Jin troops began looting once through the wall, and that the occupation involved killing, sexual violence, coercive exactions, abduction and enslavement.

A week, then months

The breach did not make both Song emperors prisoners in a single stroke. Qinzong formally surrendered and became a captive on 16 January.

With the capital lost he relinquished his standing as an equal sovereign and declared himself a minister of the Jin ruler; the Jurchens took him captive anyway. On 20 March 1127 the Jin formally reduced both him and his father Huizong to commoners.

The conquerors did not carry off only gold: court books, maps, instruments, artisans, musicians and members of the imperial household went north as well.

Autumn on the Volga

Mongol armies under Batu's overall leadership opened the great western campaign in 1236 by invading Volga Bulgaria, in the Volga–Kama region of present-day Russia.

Rus chronicle traditions place the attack in autumn and report that the invaders took and burned the Bulgars' Great City, killed inhabitants and captured the land. No exact day is securely established; only the year and that autumn placement are supportable.

The opening conquest, not all of Rus

This was the opening phase of an expedition whose conquests across the steppe north of the Black and Caspian seas ran through 1240. The Rus principalities were not overrun in 1236: the northeastern campaign followed in 1237–1238, and Kyiv fell in 1240.

The imperial decision behind it was made in the mid-1230s under the great khan Ögedei, staffing a multi-prince expedition to complete the conquest of lands assigned to Jochi's house. Volga Bulgaria, which had resisted earlier Mongol pressure, lay directly across the westward route.

What the label gets wrong

Batu was not Jochi's eldest son. Encyclopaedia Iranica identifies Orda as his older brother.

The regular tribute system so often folded into the invasion year also belongs elsewhere: the comprehensive census operations associated with regularized tribute in Rus came in 1255–1259. Medieval Rus chronicles survive in later manuscripts, differ in wording and chronology, and describe killing in terms that cannot be converted into a reliable casualty count.

Nothing happened worldwide in 1250

1250 is a rounded label, not a switch. Continental-scale reconstructions by the PAGES 2k Consortium find no globally synchronous Medieval Warm Period or Little Ice Age, and Neukom and colleagues found no globally coherent warm and cold periods across the preindustrial Common Era.

The Intergovernmental Panel on Climate Change avoids leaning on either term precisely because their timing is poorly defined and varies by region.

What can be dated is regional: around the middle of the thirteenth century, expanding ice caps in Arctic Canada overran and preserved rooted tundra vegetation.

Onset dates that do not agree

A recent compilation of Arctic Canadian moss kill dates has an early ice-expansion peak near 1240. The 2012 Arctic study behind much of this line of evidence radiocarbon-dated roughly 150 rooted plant samples.

A separate study using Arctic Canadian and Icelandic evidence placed an abrupt regional onset of cold summers and ice growth between 1275 and 1300, followed by intensification in 1430–1455.

The regional order differs too: PAGES 2k found the transition arriving earlier in the Arctic, Europe and Asia than in North America and the Southern Hemisphere, and a 2019 global analysis found different regions reaching their coldest epoch in the fifteenth, seventeenth or nineteenth centuries.

The volcano arrives seven years late

Samalas erupted on Lombok in 1257, seven years after the listed date, and may have contributed to subsequent cooling. It cannot itself be an event that occurred in 1250.

No single global trigger has been settled. Long-term orbital forcing, internal ocean–atmosphere variability and clusters of volcanic eruptions are among the mechanisms studied for regional cooling.

Frost fairs, famines and glacier maxima are later and regional, and cannot be treated as immediate consequences of the year 1250.

The night of 24–25 July 1261

Alexios Strategopoulos approached Constantinople in July 1261 with a small Nicaean reconnaissance force, after local thelematarioi told him the Latin garrison and the Venetian fleet were away attacking Daphnousia.

On the night of 24–25 July his men hid near the Holy Well outside the Gate of the Spring. In George Akropolites's account, more than fifteen of them entered through a hole in the wall, overpowered a guard, and broke the gate's locks and bolts.

Dawn, and then fire

At dawn the infiltrators admitted Strategopoulos's force. Brief resistance collapsed, and the Nicaeans secured the city without a formal siege.

As the Venetian fleet came back down the Bosporus, Strategopoulos's men set fires in the Latin and Venetian districts — four places, by one account — diverting the Latins toward saving their families and property.

The Latin emperor Baldwin II escaped from the Blachernae palace by boat and reached the fleet.

Which entry you believe

The way in is told three ways. Akropolites describes a hole in the wall and fifteen or more men; Pachymeres adds ladders; the later Nikephoros Gregoras describes an old subterranean water passage and fifty men.

The figure of 800 describes the force originally dispatched, not necessarily the headcount at entry, since local thelematarioi joined or aided it.

Only one guard's death is described in the Akropolites account; claims that all the guards were killed go past the evidence.