The Mongol Siege of Xiangyang and Fan-ch’eng and the Song military

The Mongol Siege of Xiangyang and Fan-ch’eng and the Song military

Chris Hanson

DeRe Militari (2004)

The Siege of Xiangyang and Fan-ch’eng was one of the longest sieges of the medieval world lasting almost 5 years, from 1268 until early 1273, and was the longest campaign in the war against the Song Dynasty (pronounced sung).

            Over the course of the campaign chroniclers record many heroic deeds were recorded, yet there is limited information about the siege itself.  Unlike the siege of Acre during the third crusade, between 1189 and 1191, where many separate events are recorded. Examples are the fraternization amongst Latin and Arab troops playing chess, and children wrestling where the parents of the losing child had to pay a “ransom”. Oriental sources quite often do not to cover individual battles/sieges in detail like the west.  However, oriental sources tend to give a broader perspective of the strategic war.  Finding detailed information about the siege of Xiangyang and Fan-ch’eng is difficult, however, there is some information in English sources regarding the siege of the two cities.

            After Khubilai stabilized his northern territories he established himself as the great Khan of the Mongols and the emperor of China.  Khubilai then turned his resources to the south against the Nan (Southern) Song.  Although, Khubilai took part in earlier invasions of Song territory he did not participate in the siege or other campaigns against the Nan Song.  Instead Khubilai appointed A-chu as the principal commander of the siege.

            A-chu was a gifted commander coming from a family of great generals. A-chu’s father was Urriyangkhadi, one of Khubilai’s greatest commanders, who in turn was the son of Subedei, the famous general under Chingis Khan.


            Emperor Duzong appointed Lu Wen-huan and Zhang Tianshun, both committed military leaders and loyal to the Song court, as the military commanders to defend the cities.  Xiangyang and Fan-ch’eng, known today as Hubei, were located on opposite sides of the Han River which led to the Yangtze River.  Its strategic location made it a stronghold which blocked the invasion of Song territory.  In Morris Rossabi’s bookKhubilai Khan: His Life and Times he states “Rashid al-Din’s”, (a Persian historian), “informant told him Xiangyang had a ‘strong castle, a stout wall and a deep moat.’”   The two cities were connected via a pontoon bridge allowing troops to move from one city to the other.  The garrison inside both cities were well armed, morale was high and they had plenty of supplies –supposedly enough grain to last 10 years.  However, they lacked less unessential supplies such as salt and clothing.  To support the two cities, the Song used their fleet of ships (river junks) to bring in supplies or to support troops sallying from the cities.  The Song had many forms of siege engines, including gun powder weapons, to counter Mongol siege equipment.

Khubilai realized the capture of both cities would enable the Mongols to use them as a base of operations during the war.  Liu Zheng, a former Song officer who defected to the Mongols, recommended several forts be built in the Sichuan area to cut off grain and food supplies to the two cities.

            In 1268 the Mongols, led by A-chu and Liu Zheng, surrounded Xiangyang with an unknown number of troops.  After viewing the defenses, A-chu realized he could not assault the city without taking heavy casualties.  Therefore, he decided to blockade the city and starve the people.  However, A-chu discovered without a fleet he could not stop supplies from coming up the Han River into the twin cities.  A-chu sent word to Khubilai explaining his problem that his troops consisted primarily of cavalry and were not entirely suitable for the siege.  He requested Shi Shu, a northern Han Chinese commander in Mongol service who commanded several thousand northern Chinese troops, be sent to support the siege.  After the arrival of Shi Shu and his troops, which consisted almost entirely of Chinese infantry, both cities were surrounded.  Again the number of troops is unknown.

            As the Mongols surrounded the twin cities, troops from Xiangyang tried to break out through the siege lines but failed.  There are no records of the number of soldiers who attempted to break through the Mongol lines, but Chinese sources state many Song troops were captured and beheaded.  From this point forward, the Song troops stayed within their walls and waited for supplies and reinforcements to arrive.

            When Shi Shu arrived, he commanded his troops to start building 500 river junks, training the crews and he was able to capture some river junks from the Song.  Two fortresses, Po-ho-k’ou and Lu-men shan, were built downriver from the twin cities to deal with the Song navy coming up the river.  In August of 1269 the Mongols and their Chinese allies were to have their first naval battle.  A Song general named Hsia Kuei with 3000 junks attacked Lu-men shan.  Hsia Kuei was defeated and fifty of his junks were captured.  What happened to the crews and other junks is either not recorded or possibly not translated into a western language.

            Over the course of the siege it appears Khubilai became unhappy with the lack of results and sent one of his trusted military advisor’s, Shih T’ien-tse, to inspect, review and make suggestions to capture the two cities.  One of the suggestions Shih T’ien-tse made was to build a rampart between the two forts, P-ho-k’ou and Lu-men shan to stop Song river junks from supplying the two cities.  After his review Shih T’ien-tse authorized an additional 20,000 troops to reinforce A-chu’s army.  The Mongols even tried to dam the river or at least restrict the flow to render the pontoon bridge between the two cities unusable.  However, the garrisons of both cities still resisted the Mongols assaults and would not submit.  Even the Mongol changesin their siege tactics, Song officials were still able to find their way into the two cities.

            Throughout the siege, the Song imperial court remained adamant that it would not to surrender the two cities.  Some sources they indicate the Song emperor was kept ignorant of the siege and the threat of the Mongols.  Jia Sidao, the emperor’s chief advisor/councilor was in control of the imperial court and by some sources considered a treacherous minister full of corruption.  Other sources tend not to be as harsh and give him credit for changes within the court to reduce corruption.  However, he was not a military leader and was despised by many within the Song military.  Many commanders defected to the Mongols due to the decisions of Jia Sidao.  In one source, a concubine informed the emperor of the siege.  Later Jia Sidao had the concubine executed.  There is no doubt Jia Sidao was a ruthless and cunning individual who cherished his position within the imperial court.  The continued resistance of the two cities was important to Jia Sidao.  In an earlier campaigns against the Mongols, Jia Sidao claimed a victory after killing fewer than 200 of them, (yet in reality Khubilai was pulling back to deal with his succession).  If the two cities were to fall then the credibility of Jia Sidao would be undermined.  If they held, his position within the court would only strengthen.

            There seems to be little record about the Song Emperor Duzong’s views on the siege of the twin cities in any literature reviewed.  Emperor Duzong was the adopted son of the previous emperor, Lizong.  Emperor Lizong was one of the weakest emperors of the Song Dynasty, taken to the pleasures of his concubines and neglecting his duties as emperor.  Duzong seems to have followed the examples set by his step-father and left the duties of government to Jia Sidao.  Had the emperor put as much effort into the defense of China as his people did the Song army may have held out much longer or even may have stopped the Mongol invasions.

            There were many attempts made by the Song to break out of the cities.  One example occurred in March of 1270, when Lu Wen-huan decided to break out of the city with 10,000 troops consisting of infantry and cavalry.  The exact composition of the force is not given, but the Song took heavy casualties and had to retreat back into the city behind their walls.  The fact that 10,000 troops were involved is a testimony to the size of the garrison holding out inside the two cities.

            Over the course of the siege several attempts were made by the imperial court to relieve the siege, primarily by river.  In October of 1270 Fan Wen-hu tried to break through the blockade.  He failed, loosing 2000 men and thirty boats.  August of 1271 the Song experienced another failed attempt to relieve the siege, loosing 2000 soldiers.  In September of 1272 two famous commanders with 3000 men broke through the Mongols lines.  Chang Shun, one of the commanders, died and was considered a terrible loss.  Sometime later the other commander with the remainder of his original 3000 troops tried to break out.  This time the Mongols were ready.  To help them see at night the Mongols put straw on the banks of the river and set it on fire to light up the boats.  Many of the troops were captured and the commander was killed.

            Still, Lu Wen-huan would not surrender, and the Mongols could not storm the city without taking heavy casualties without a strong probability of taking the city.  Khubilai realized something had to be done and sent word to his nephew, Abakha, the Il-khan of Persia.  Abakha sent two renowned engineers, Isma’il of Hilla and Ala al-Din of Mosul, who built the first counterweight trebuchets in China called hui-hui pao, or “Muslim trebuchet”.

            The Mongol commanders employed the new siege weapons against Fan-ch’eng first.  Chroniclers report the siege engines shook heaven and earth when they fired and the projectiles destroyed everything they hit.  The projectiles weighed weigh 150 catties, (approximately 165 lbs. or 75 kg), and had a range greater than anything the Song had within the city so they couldn’t counter the attack.  The Mongols launched explosives and who knows what else into the city with their new siege trebuchets.  Fan-ch’eng could not hold out against such a barrage and fell several days after the weapons were employed.  When Lu Wen-huan received reports regarding the fall of one of the twin cities, he was in dismay but did not give up right away.  The hui-hui pao, or Muslim trebuchets, were moved to other side of the river and placed at the southeast corner of the Xiangyang.  Lu Wen-huan continued to hold out against the continued assaults of the Mongols.  Eventually he was offered the command of the two cities if he was to surrender and defect to the Mongols.  Khubilai Khan encouraged such defections and many Song commanders, seeing Khubilai as the one with the mandate of heaven to be the new emperor of the Yuan Dynasty, defected.  Lu Wen-huan surrendered, ending the long tiresome siege, and the two cities were given to him to govern under the Yuan Dynasty.

            The fall of Xiangyang and Fan-Ch’eng marked the turning point in the war against the Song Dynasty.  Morale collapsed within the Song imperial court, and the Song army was able to hold out for several more years before the Mongols conquered the rest of China.

Assessing the Song

Many modern books considered the Song dynasty militarily weak.  However, the Song military held out against the Mongols for over 40 years, longer than any other people the Mongols attacked.  If the Song military was so weak, how could they hold out so for so long?  In Peter Lorge’s dissertation on the establishment of the Northern Song he theorizes that the Song were just as capable as the Tang or Han dynasties and were not as weak as we are led to believe.  Rather the enemies of the Song were much more capable than the enemies of the Tang or Han dynasties.  Yet the Song military did fail to defend their country.  This failure can be traced back to the origins of the Northern Song Dynasty and their failure to take to the 16 prefectures, near modern day Beijing, from Khitan Liao.  The review of the Northern Song is beyond the scope of this article but can be read in detail in Peter Lorge’s dissertation mentioned above.

            Later the Jin, “Golden Dynasty”, pushed the Khitans west and took the northern part of China.  The Hsi Hsia, also called the Tanguts, invaded and took the northwest part of China further weakening the Song Dynasty.  Even though the northern part of China was lost, the imperial court along with many of the wealthy and much of the military were able to move south.  There is no one factor that can be used to state why the Nan Song could not hold out against the Mongols.  There are many factors, the first that must be reviewed, is the imperial court.

            Emperor Lizong, 1225-1264, neglected his duties as emperor as his dalliances with his harem of concubines.  Emperor Duzong, 1265-1274, who succeeded Emperor Lizong doesn’t seem to have put much effort into his imperial duties either.  Without clear leadership from the imperial court how could the Song military defend their country?  The leaders within the imperial court during the Song Dynasty distrusted their commanders.  These imperial advisors counseled the emperor to leave many high level positions within the military vacant.  Commanders and generals were moved from one military post to another to ensure they did not develop a bond with their troops.

            Another factor to consider in analyzing the Song’s defeat is the nature of their army.  The Tang and Han armies were conscripted soldiers trained to fight for the emperor. The later Song military differed substantially from the Tang and Han style armies consisting of hired professional soldiers.  Some sources describe the Song military as a mercenary force with many of the soldiers being hired from the dregs of society.  The Song inducted criminals into the army and branded or tattooed to discourage desertions.  Peasants who became landless due to natural disasters were conscripted into the military.  Many of the soldiers were ill led, ill trained, and ill equipped, however, some troops were highly motivated, well trained, and well equipped.


There are surviving descriptions documenting the formations and training of Song infantry.  Some of these descriptions depict spearmen and swordsmen standing (perhaps they are kneeling) while the archers and crossbowmen fired their shots.  Drums were used to give the commands for marching, firing, sitting/kneeling, and other commands.  It seems in later Song military manuals that archers and crossbowmen were not mixed with other infantry.  Whether this was adhered to by those in command is undetermined.  It would seem the generals would see the value of mixing missile troops with spearmen and swordsmen.

            The Song were lacking in cavalry due to their failure to secure adequate areas to raise horses from the Khitan Liao and the Tanguts.  Since the Song lacked in adequate cavalry, the song developed weapons, tactics and training for infantry to counter heavily armored enemy cavalry formations.  One of their innovations was a training program using two poles spaced approximately 2 feet apart, one short and one long, simulating the leg and head of a horse.  The infantry were trained to use their zhan ma dao or mei jian (see next paragraph for a description), and bend low, swinging at the lower post taking out the horse’s legs then turn swinging high aiming for the head of the horse.  The Song employed tie ji li or caltrops in front of these infantry to slow the impetus of the cavalry charge.

            The Song Dynasty invented many different weapons for the infantry to deal with heavily armored cavalry.  One such weapon was the zhan ma dao best described as a two handed “horse chopping sword.”  In the Official History of the Song Dynasty it states the sword was massed produced in the tens of thousands.  Another bladed weapon used in similar fashion was the mei jian or eye brow tip glaive similar to the Japanese nagamaki or the blade used by the elves in The Lord of the Rings.  Yue Fei, a famous Song general, used anti-cavalry troops trained with a large two handed axe or Ko.  In the mid 1100’s Yue Fei, used these highly trained anti-cavalry troops, defeated a Jin army and destroyed 70% of the Jin cavalry.  The armor worn by troops specializing in anti-cavalry was well made.  It consisted of shan wen jia or mountain pattern armor.  See photo’s 1 and 2.  Shan wen jia armor offered good protection against missile weapons and hand held weapons.  Whether these troops carried a bow or not is undetermined.  The Song considered the composite bow an important weapon and some sources imply that the infantry did carry it in addition to their sword or axe.

            Other weapons created in the Song dynasty were the da fu or great axe – halberd and the guan dao, basically a glaive on a long handle similar to the Japanese naginata.  How these troops were armored is unclear.  They may have worn the shan wen jia or mountain patter armor mixed with the more common lamellar armor.  It appears these troops supported the more highly trained anti-cavalry troops with the zhan ma dao, ko, and mei jian.

            Spears and swords were very common during the Song dynasty.  The length of the spears is not standardized in any of the texts reviewed.  In Peter Lorge’s dissertation it states the length of the longer spears were 5.5 meters to 5.7 meters long or 18.3 feet to 19.1 feet.  Extra long spears were called giang.  Shorter spears were used, and some were thrown.  A common sword used by infantry was the shoudao, a single handed single edged sword.  Most spearmen and swordsmen carried a shield.  There are very few depictions of shields but they appear to be rectangular shaped for the spearmen with some having a V cut shape at the top or an inverted V cut at the top, while others having a shallow V cut shape on the sides looking similar to an hour glass but with straight edges.  It seems those carrying the longer spears (pikes) did not carry shields.  A common shield for the swordsmen was smaller and round, making it easier to use in combat.  Shields were often painted with a terrifying face to frighten the enemy.  Such a display of shield art would make for a pretty army.  Some sources indicate swordsmen and spearmen were mixed together, swordsmen making up the first rank and spearmen supporting behind.  These troops made up the bulk of the army, but were not the quality of those specializing in anti-cavalry tactics and they did not have an extensive training program as the specialist troops.

            There are several types of crossbows used during the Song dynasty.  The most powerful was the “Shenbi nu” or the Divine crossbow, also translated as Divine Elbow crossbow, possibly an arbalest.  Records indicate the maximum range of the Shenbi nu was 460 meters with an effective range of 150 meters.  Other crossbows had a maximum range of 320 meters.  There were three methods of loading a crossbow; first was to place the butt against the hip and pull the string back with their arms, second was the use of the foot stirrup and the third was to sit down and pull with their back and arms.  Another famous crossbow was the “Cho Ko Nu” or repeating crossbow.  This weapon was not known for its long range or it’s accuracy but was well known for its rapid rate of fire.  There are several accounts of troops armed with the Shenbi nu or Cho Ko nu stopping a cavalry charge.  The Song considered the crossbow an important weapon and spent time training units to make up for the slow rate of fire.  The crossbowmen would line up in three ranks; the first rank firing, the second rank were those advancing to replace the first rank, and the third rank were those reloading.  Essentially creating the volley fire formation used by European troops armed with muskets in the 18th century.  Other larger siege crossbows such as the triple crossbow were used during siege operations.

            Although the crossbow was considered the most important bow against cavalry the infantry still used the large composite bow.  The maximum range of the composite bow used by infantry was around 256 meters.  Accounts I have found describe mixed units with the archers behind the crossbowmen.  One account states the spearmen made up the first rank and would kneel allowing the crossbowmen to fire.  While the crossbowmen were reloading the archers, in the third rank, would fire.  It seems the archers killed more of the enemy than the crossbowmen did.  As I mentioned above Song military manuals indicated missile weapons in each unit were not mixed.  It could be they did not mix with those trained in anti-cavalry tactics but did mix with those with spears and swords.  The maximum range of the composite bow used by infantry was around 256 meters.

            Archers and crossbowmen as well as spearmen and swordsmen wore armor called buren jia or foot armor.  As described in the Wu Jing Zong Yao Song Military Manual, armor for spearmen consisted of 3782 plates weighing 32-35kgs taking the blacksmith 120 days to make.  Armor for the archers weighed 28-33kgs, and for crossbowmen it weighed 22-27kgs with 1825 iron plates.  Another form of armor used by archers was paper armour 1-3 inches thick called zhi jia.  When wet it became stronger and offered good protection against arrows and the cold but if it wasn’t maintained properly it became useless.

            Another form of armor was xouzi jia or locked chain armor and, liansui jia or linked chain armor.  Such armor was not abundant during the Song Dynasty but was used by those of higher rank.  The biography of Yue Fei mentions he used a form of chain mail, yet a statue of him does not show him wearing it, rather it depicts him wearing the shan wen kia or mountain pattern armor.  While armor was important for infantry during the Song dynasty but as to how many troops within the army wore it is unclear.  It is clear than many of the troops used it and plenty was available.


According to an ex-Song court official the greatest weakness to the Song military was its lack of cavalry.  Although the imperial court may claim it had many cavalrymen, there weren’t enough mounts for the men, maybe two horses per ten men, sometimes less.  To remedy this lack of cavalry the imperial court would make the army larger employing more infantry hoping to overwhelm the enemy.  In my estimation, the employment of additional infantry did not significantly benefit the effectiveness of the of the army.  However, it did was create a larger budget for the military straining the courts resources.

            There were some in the imperial court who realized they needed more cavalry and tried different ways of obtaining horses.  Breeding grounds were created in the lower Yellow River Valley in an attempt to procure more mounts, but the program failed.  Markets were established on the northern border trading tea for horses.  The price of the tea was inflated, but the quality of the horses did not improve.  Since the loss of its northern territories the Song had to rely on tribal cavalry, T’ung shih chun, deserting or fleeing the Jin or Mongols.  The T’ung shih chun would most likely be the lighter Mongol/Turkish type cavalry armed in the steppe nomadic fashion.  Imperial Song cavalry appeared to be heavily armored.

            I have only seen two depictions of Song cavalry.  One is in Osprey’s publication, titled Medieval Chinese Armies 1260 – 1520 by C J Peers, page 5.  Another is online  Both depictions show the men heavily armored.  The depiction online shows all the cavalry wearing barding, but the one in the Osprey depiction shows cloth over the horse’s neck and a metal plate over the rump.  In China A New History by John King Merle in between pages 104 and 105 picture #9 show’s a Song painting of a Uighur general paying homage to a Tang general.  Some of the cavalry are wearing the same barding as the cavalry shown on the web site mentioned above.  Although the event took place during the Tang dynasty, the painter used the armor of Song period in his painting.

            It appears the Song adopted the heavily armored cavalry most frequently used the Jin.  This would make sense; they would want to protect their limited number of mounts.  Even though barding was used and the men wore armor made up of either shan wen kia or mountain pattern armor or a lamellar suit, very few carried shields.  The shields that were used are described as round and painted with a terrifying face or painted with red lacquer.

            Silk was also used for cavalry and infantry protection.  Silk offered protection against missile weapons by imbedding into the flesh with the arrowhead.  The edges of the silk were tugged allowing the arrowhead to come out without tearing the flesh which helped reduce infections.

            Weapons for the cavalry appear as varied as the infantry with different types of pole arms such as the da fu, guan dao, and some depictions show the langya bang or wolf’s tooth stick which looked like an oblong mace on the end of a shaft.  Another weapon depicted is the Sanjian Liangrend dao, 3 pointed, 2 edged glaive, and the Erland dao, a glaive like weapon portrayed to be used by a daoist god.  The lance or shuo was used and seems to be of similar length to that of western or middle-eastern lances and it was used couched like their western counterparts.  Most, if not all, cavalry carried the composite bow.

            There is much debate amongst historians on the effectiveness of the Song cavalry.  The depictions of the Song cavalry are quite spectacular, but fancy looking troops do not win battles.  In most battle there was probably not enough cavalry to make a difference. In the dissertation by Peter Lorge there is a description of cavalry training during the Northern Song Dynasty. Whether such training was used or another form of training was conducted during the Nan Song is unknown or possibly not translated into a western language.  Documentation exists indicating that the cavalry troops were regularly rotated to the northern borders for duty, making the soldiers hardier and more experienced.


The Song Dynasty was well ahead of the rest of world in terms of technology and culture.  The Song were the first to develop the use of gunpowder for warfare and created the first printing press with movable block type.  Siege weapons were created for various uses and the Song developed the well known traction trebuchet.  Yet, these technological advantages did not save the Song Dynasty.  Several types of gunpowder weapons were used and I have chosen to briefly cover a couple of them.

            The first one is the li hou-ch’iang or Firelance, also spelled t’u hou ch’iang which translates to “flame-spouting lance”.  It was placed at the end of a spear or lance and shot out a flame as well as a small projectile.  The fire lance is considered by some modern historians as the precursor to the gun.  The weapon was used by both cavalry and infantry.  Archers used a fire weapon attached to the end of an arrow called zhun pao heng pen qian or hand rockets.

            Another well documented gunpowder weapon was the zhen tian lei or thunder crash bomb which was launched from a catapult or traction trebuchet.  Accounts state when it exploded it could be heard 100 li away, approximately 30 miles.  There are accounts of it being used against enemy troops besieging Xiangyang and Fan-ch’eng, yet how much damage this bomb did or how many people it killed is unclear.  Was that because it really wasn’t that effective?  Did it explode at the incorrect time, such as in flight or when igniting it.  I find it hard to believe a weapon that could be heard so far away didn’t cause some type of damage worth noting.  If anything, it would be good for frightening horses or against enemy siege weapons.

            There are many depictions of Chinese ships loaded with bags of gunpowder and used to destroy other ships or bridges.  Other types of gunpowder weapons are recorded, some thrown by infantry or attached to animals, but again the effectiveness of such weapons is not clear.  The Song used gunpowder weapons of virtually every sort and it is doubtless that some were effective, while some were not.

            The Song Dynasty is the least studied dynasty by the west.  Since it was not a “conquering” state many see it as weak and not worthy of study.  I contend the Song were not militarily weak, but rather their northern neighbors were much stronger, better organized, and had gifted leaders.  In the long run the Song military and the imperial court failed to stop the invasion of the Hsi Hsia, Jin, Khitan Liao, and later the Mongols.  After the collapse of the Song imperial court Khubilai united China under a new dynasty called the Yuan meaning “origin”.  Additional information regarding Song dynasty warfare can be found in Peter Lorge’s new book Chinese Warfare 900-1795.


China: A New History, by John King Merle

Dissertation on the establishment of the Northern Song Dynasty, by Professor Peter Lorge

Imperial China: The Art of the Horse in Chinese History, exhibition catalog

Khubilai Khan: His Life and Times, by Morris Rossabi

Medieval Chinese Armies 1260-1520, by Chris J. Peers

This entry was posted in Articles, Uncategorized and tagged , , , . Bookmark the permalink.

Comments are closed.