In 1241, the Mongols advanced as far to the west as their forces were ever to reach. At Legnica (at one time Liegnitz) in western Poland, one of their columns destroyed a substantial army led by Duke Henry of Silesia, who was killed.Footnote ¹ Another column pursued King Bela IV of Hungary as far as the coast of Dalmatia. The main Mongol army almost annihilated the Hungarian army that had assembled to oppose them at the Battle of Mohi. That winter, they crossed the frozen Danube and despoiled western Hungary. Yet quite suddenly, early in 1242, they began to pull back. Soon they were in Bulgaria, and shortly afterwards in the southern lands of Rus'.Footnote ² Some stopped when they reached the Kipchak steppe, others made their way back to Mongolia. Never again would they pose such a grave threat to Europe.Footnote ³

The reason for the sudden withdrawal in 1242 has been the subject of considerable discussion and no little speculation. The first person to offer an explanation seems to have been John of Plano Carpini, who visited the court of the Great Khan Güyük in 1246, and afterwards related that: “their Emperor [Ögödei] was killed by poison [in 1241] and consequently they have rested from battle until the present time”. He also specifically says that the death of the Khaghan occurred “when their army was in Hungary and as a result the army in these parts retreated”.Footnote ⁴ For a long time this explanation was accepted by historians more or less without comment,Footnote ⁵ but from the early 1970s other explanations began to be advanced. Sinor put forward the theory that the Mongols' reliance on their horses was the fundamental reason: “the troops that left Hungary settled in the South Russian steppe where they were to remain for centuries and where there were grazing grounds vast enough to support their herds. In my view the Mongol evacuation of Hungary was motivated by Batu's logistical difficulties and his recognizance of the fact that the Hungarian pastures were insufficient to provide for his army's needs”.Footnote ⁶ It has also been argued that the Mongols had been greatly weakened by the resistance they faced from the peoples of the Rus' lands, and the Poles and Hungarians, so that they no longer had the military strength to pursue their campaign in Europe. There is some support for this in the contemporary account of Thomas of Spalato, who says that the Rus' resisted them strongly and held up their progress.Footnote ⁷

These and other arguments regarding the Mongol withdrawal from Hungary in 1242 have been admirably summarized by Rogers.Footnote ⁸ Briefly, they fall into four main categories, which Rogers calls the “political”, the “geographical”, the “military weakness” and the “gradual conquest” theories. The “political” theory is that supported by John of Plano Carpini's statements, that it was the death of the Great Khan Ögödei that caused the withdrawal. The “geographical” theory, which might perhaps better be characterized as “logistical”, is that of Sinor, that the Mongols could not find enough pasture for their horses in Europe. This theory has attracted considerable support, notably from D. Morgan.Footnote ⁹ The idea that the Mongols had suffered greatly at the hands of the Rus', Poles and/or Hungarians, so that they were too weak to continue their advance westwards, has found favour mainly with Russian and East European historians, sometimes for obviously nationalistic reasons.Footnote ¹⁰ The “gradual conquest” concept, which posits that the Mongols commonly conducted reconnaissance raids first, and only subsequently embarked on all-out conquest, in fact fits well with what happened in other areas. To give just one example, the Xi Xia state was first raided by the Mongols in 1205, and suffered further incursions in 1209 and 1217, but was not finally destroyed as a political entity until 1226–27.Footnote ¹¹

It is difficult to argue comprehensively and decisively against all these theories. It can be said, however, that Sinor's hypothesis is surely invalidated by a serious flaw in his calculations of how many horses the Hungarian pastures could support. He claims that: “a range area of 120 acres is needed to support one horse for one year”.Footnote ¹² This is wrong by a very large margin. In fact, even unimproved pasture will support one horse (and one large horse, not a small Mongolian horse) on five to ten acres (about two to four hectares).Footnote ¹³ The theory of the severe weakening of the Mongol army by the resistance it had faced in the Rus' lands, Poland and Hungary must also be questioned. Russian sources do not suggest any serious losses. The Mongols were not greatly held up in Rus': their campaign during the winter of 1237–38 resulted in the destruction of a dozen towns, most of which fell in a few days. They took Ryazan “on December 21, and they had advanced against it on the 16^th of the same month”. Only Torzhok managed to hold out for as long as two weeks.Footnote ¹⁴ In 1240, they broke the defences of Kiev in just nine days.Footnote ¹⁵ Moreover, when the Mongols invaded Poland and Hungary, they divided their forces into no fewer than five groups, advancing by different routes.Footnote ¹⁶ It seems unlikely that they would have split their strength to such an extent if they had been seriously weakened during their campaigns in Rus’. Nor did they commit all their strength to the campaign against Hungary: Ögödei recalled Güyük and Möngke from Rus' before the Mongol armies advanced against Kiev.Footnote ¹⁷

It is quite possible that the Mongols were not seriously intending to advance much further into Europe in 1242. They may have planned no more than a reconnaissance in strength, or a raid to carry off slaves and booty. Determining now exactly what their intentions may have been then is virtually impossible. They themselves may not have had a clear idea of what they would do after overrunning Hungary. Their plans would no doubt have been flexible, depending on circumstances. Their campaign was probably not completed, however, for they had not captured King Bela IV of Hungary, as they clearly wished to do.Footnote ¹⁸ It may reasonably be assumed, therefore, that they would not have withdrawn eastwards from Hungary as early as they did, without some compelling reason.

It has been said that: “The Mongol withdrawal from East Central Europe is a unique event in the history of the region and in the histories of the Mongols, Rus’, and Inner Asia”.Footnote ¹⁹ In fact, however, there is an interesting parallel, in events immediately following the death of the Great Khan Möngke in 1259. Then, too, a major military campaign was suddenly interrupted. This certainly suggests that the death of a Great Khan could be sufficient reason for the Mongols to break off military activity. The full explanation may be more complex, however.

It is the case, as John of Plano Carpini observed (see above), that the Mongols did not engage in any major military campaigns for several years after 1242. It seems likely that there was very nearly a civil war in this period. Batu, of the Jochid ulus in the western steppes, was strongly opposed to the elevation of Güyük to the throne as Great Khan. In the event, the early death of Güyük averted hostilities.Footnote ²⁰ It was only after Möngke became Khaghan in 1251 that the Mongol conquests were actively pushed forward once more. The most important theatre of war for the Mongols at this time was in the Far East, where the main goal was the conquest of the Southern Song empire. This was a difficult task, however, as south China, with its rivers, lakes and mountains, was easily defensible. The Mongols were wary of a frontal assault from northern China, and had already opened another front in Sichuan.Footnote ²¹ This divided Song forces, but between Sichuan and eastern China there is a great expanse of mountainous terrain, which has always been notoriously difficult for an invading army to cross. Conquering Sichuan did not open the way to the conquest of the rest of the Song empire. In the early 1250s, a decision was made to open up another line of attack, by first subjugating the independent Kingdom of Dali, south-west of the Song empire, and then threatening south China from what is today northern Vietnam (variously called Annam or Jiaozhi in Chinese sources).Footnote ²² Such a stratagem had, in fact, been suggested to Chinggis Khan, by Guo Baoyu, some forty years earlier.Footnote ²³

This campaign has been described as a “failed second front”.Footnote ²⁴ This is nonsense, however: it was neither a “second front” nor a failure. As has been pointed out above, the Mongols already had a second front against the Song empire, in Sichuan. As they were able to use more than one line of attack east of Sichuan, it could be argued that they already had at least three fronts against Song.Footnote ²⁵ It should also be noted that the Kingdom of Dali had become an important source of horses for the Song empire.Footnote ²⁶ Once Dali had been conquered by the Mongols, Song must have lost access to these horses, causing a serious problem for its armies.Footnote ²⁷ Moreover, the campaigns against Dali, and later against Annam, allowed the Mongols to invade the Song empire, successfully, from the south. This has sometimes been overlooked.Footnote ²⁸ It has even been claimed that they were somehow forced to withdraw from Annam at this time (1258).Footnote ²⁹ This is based on an incorrect understanding of Chinese source material, which says that the Mongol forces, led by Uriyangkhadai, withdrew from the Annamese capital after only nine days, because of the unhealthy climate.Footnote ³⁰ Other records, however, state that the Annamese king submitted, and that the army then returned to Yachi (modern Kunming).Footnote ³¹ Shortly afterwards, in 1259, Uriyangkhadai's Yunnan army invaded the Song empire from Annam (Jiaozhi) and advanced towards modern Nanning, in Guangxi. This was part of a co-ordinated campaign against the Song empire, with Mongol armies attacking in Sichuan, under Möngke Khaghan, and in the east from Shandong and Henan southwards.Footnote ³²

In 1259, Khubilai set out from Xingzhou (modern Xingtai in Hebei province) to lead an army into Song territory, passing through eastern Henan. He crossed the Huai River at the beginning of September. Later in the same month, as the army approached the Yangtze River, a messenger arrived from Sichuan bearing news of Möngke's death. Khubilai refused to break off the campaign, however. His army crossed the Yangtze and moved along its southern side to lay siege to Ezhou (modern Wuchang, part of the great conurbation of Wuhan). On or shortly after 17 November, a messenger arrived from Uriyangkhadai. He had led an army “from Jiaozhi [Annam] via Yong[zhou] [modern Nanning in Guangxi] and Gui[lin] [in northern Guangxi] to reach Tanzhou [modern Changsha in Hunan]. He had heard that [Khubilai] was at E[zhou], and sent a messenger to report.”Footnote ³³ Uriyangkhadai and his army had made a remarkable march, fighting all the way. “They had fought 13 battles, great and small, and killed more than 400 thousand Song soldiers, capturing three major and minor generals.” Subsequently they moved north, crossed the Yangtze River, and joined with Khubilai's main army (which by then was withdrawing northwards).Footnote ³⁴ Clearly, then, the campaigns against Dali and Annam had enabled the Mongols to inflict serious damage on the Song empire (although it must be suspected that the Song casualty figure is exaggerated considerably). Uriyangkhadai's army had successfully fought its way from the Annam border with the Song empire all the way to the Yangtze River, and had effected a junction with the main Mongol army. It may incidentally be noted that the ability of a Mongol army to campaign successfully in both Annam and China south of the Yangtze clearly indicates that problems associated with climate and geography were not insuperable. The army that Uriyangkhadai led in 1259 may well have included relatively few Mongols and many men from the former Kingdom of Dali,Footnote ³⁵ but this campaign nevertheless shows that the Mongols were able to find ways to fight successfully in areas far removed from their native steppe. The Mongols proved themselves to be remarkably adaptable during the 1200s.

It was at this point that Khubilai received news that moves were afoot to elevate his younger brother Arigh Böke to be the next Khaghan. When he received a message from “the empress” (probably his wife Chabi), urging him to return to Shangdu quickly, Khubilai withdrew from southern China to return to the north. He left the vicinity of Ezhou on 29 November 1259, and spent the remainder of the winter at Yanjing (modern Beijing). The following year, in early May, he was enthroned as Khaghan at Kaiping (Shangdu).Footnote ³⁶ Arigh Böke was not intimidated by this, however. Shortly afterwards, he arranged his own elevation to the position of Great Khan, at Karakorum in Mongolia.Footnote ³⁷ This began a conflict that was to last for more than three years. Arigh Böke was quickly forced onto the defensive. By 1264, most of his allies had either been crushed or had deserted him, and he had little option left but to surrender to his elder brother.Footnote ³⁸ More or less inevitably, Khubilai's preoccupation with dealing with Arigh Böke's challenge had meant abandoning the gains made in southern China in 1259–60. The small military force left to try to safeguard those gains was inadequate for the task, and Song was soon able to reoccupy the lands it had lost.Footnote ³⁹

It is also worth noting that another Mongol invasion force withdrew from conquered territory in 1260, with consequences that were, for the Mongols, something of a disaster. The army of the Ilkhan Hülegü had invaded Syria in September 1259. In January 1260, after a siege of a few days, Aleppo fell (its citadel held out until February). Subsequently, other Syrian cities, including Damascus, submitted.Footnote ⁴⁰ At this point, Hülegü learned of the death of Möngke, and withdrew the main body of his forces north-eastwards into Azerbaijan.Footnote ⁴¹ The precise reason for this is by no means clear. It has variously been suggested that it was because Hülegü himself wished to assert a claim to the Khaghanate, or that he wished to be in a better position to await developments in the civil war between Khubilai and Arigh Böke. He himself said, in a letter sent to King Louis IX of France in 1262, that there was insufficient fodder and pasture for the horses of his army in Syria. He moved quite slowly into Azerbaijan, and did not go further east, so it also seems possible that he was worried that Berke of the Jochid ulus might take advantage of the death of Möngke Khaghan to make a move to assert a claim over Azerbaijan.Footnote ⁴² Whatever the case, it is at least likely that the death of Möngke, and disputes within the Mongol ruling family, underlay Hülegü's withdrawal. An additional reason may well have been that a miscalculation was made regarding the strength of the army left to hold Syria, which was subsequently routed at ʿAyn Jalut by the Mamluks.Footnote ⁴³

A comparison of the events that followed the deaths of Möngke and Ögödei is instructive. In both cases, a major military campaign was abandoned when it appeared to be progressing well and could have led to further gains. In 1242, Hungary had been completely overrun, its king had fled, and there seemed to be very little to prevent the Mongols from extending their conquests in Europe. In 1260, Mongol armies had reached the Yangtze, from both north and south, and had begun the siege of a major city on the south bank of the river. The Song empire, even if it was not on the point of collapse, was certainly facing a grave crisis. It is recorded that consternation gripped the Song court when it was informed that a Mongol army had crossed the Yangtze.Footnote ⁴⁴ The Mongols were in a very advantageous position, and might well have gone on to complete the conquest of Song within the next two or three years. Yet, in both 1242 and 1260, the Mongols withdrew.

It was not, however, inevitable for Mongol campaigns to be interrupted on the death of the Great Khan. As has been seen above, when Khubilai first received news of the death of Möngke, he did not immediately withdraw, but pressed onward to besiege Ezhou. It was because he was later made aware of Arigh Böke's apparent intentions that he hurried back to north China. It seems highly likely that it was also because of disputes about the succession to Ögödei that the Hungarian campaign was curtailed. Batu had quarrelled with Güyük and was strongly opposed to him becoming Khaghan. The Secret History of the Mongols includes an account of a serious quarrel between Batu and Güyük at this time. Although it seems likely that this passage may be a later interpolation in the original text,Footnote ⁴⁵ it may none the less have had a foundation in fact. Rashid al-Din notes that Batu refused to attend the assembly that proclaimed Güyük Khaghan, in 1246, but says only that he “was offended … for some reason and held aloof”.Footnote ⁴⁶ Whatever the exact situation regarding Batu and Güyük, it is a fact that the proclamation of a new Khaghan was delayed for years, and that the delay clearly produced an unsatisfactory state of affairs. There was no very clear supreme authority in the Mongol empire and various Mongol princes took matters into their own hands, without reference to the regent, Ögödei's widow, Töregene.Footnote ⁴⁷ The refusal of Batu to attend a khuriltai was serious, as he was the senior Mongol prince after the death of Ögödei.Footnote ⁴⁸

It seems a reasonable assumption, therefore, that Batu's desire to block the elevation of Güyük to the Khaghanate was the fundamental reason for the Mongol withdrawal from Hungary. If there was no further attempt by the Mongols to push westwards into Europe after this withdrawal in 1242, it was very probably because of the lack of leadership during Töregene's regency, and the brief reign of Güyük, with a possible threat of civil war between him and Batu. This cannot explain, however, the continued lack of action in the west after Möngke became Khaghan, apparently with Batu's support, in 1251.Footnote ⁴⁹ In his disposition of appointments shortly after taking the throne, Möngke excluded the most westerly portions of the Mongol empire,Footnote ⁵⁰ most probably because he had accepted that these were entirely under Batu's control.Footnote ⁵¹ Möngke concentrated efforts at further conquests in the Far East and Persia. Meanwhile, the Jochid ulus was weakened by Batu's death, probably in 1255, and the deaths of two of his successors, his son, Sartakh, and his grandson (or possibly younger son) Ulaghchi, which followed shortly afterwards.Footnote ⁵² It may well be for these reasons that there was no further Mongol incursion into Central Europe during the 1250s. It is not true, however, that the Mongols never returned. In 1259, they again raided and plundered Poland, destroying Sandomierz, with its castle, and Krakow.Footnote ⁵³ By this time, however, the unity of the Mongol empire was breaking down, and the major period of conquest was already past.Footnote ⁵⁴ There could never again be a Mongol invasion of Europe involving forces of the entire Mongol empire. It has recently been suggested that the Mongol conquests were to a great extent facilitated by the use of Chinese gunpowder weapons, and that they began to falter by the late 1250s because the secret of gunpowder had been acquired by some of the Mongols’ enemies.Footnote ⁵⁵ If this is true, then the inability to resolve the succession crisis after the death of Ögödei, and the resulting failure to press forward with further conquests, may well have saved Europe, and perhaps other parts of Eurasia, from further large-scale Mongol invasions during the later 1240s.

The Mongols did resume their offensive against the Song empire, however. Indeed, it was never fully suspended. Even during the regency of Töregene there were repeated attacks on Song.Footnote ⁵⁶ These all involved the Chinese general Zhang Rou, who had surrendered to the Mongols in 1218 and had played an important part in the Mongol conquest of the Jin empire in northern China.Footnote ⁵⁷ It must be possible that these continued actions against the Song empire were undertaken by Zhang Rou largely on his own initiative, so that the forces under his command did not forget their military skills. The Song empire could never be ignored, however. Song had never accepted the loss of northern China to the Jurchen Jin dynasty: for more than a century there had been either an uneasy peace or a state of war between Song and Jin. Song had co-operated with the Mongols in the final destruction of the Jin empire.Footnote ⁵⁸ If the Mongols had left Song alone, it is unlikely that Song would have accepted Mongol domination of northern China indefinitely. Moreover, the Song empire was a rich prize, one worth fighting for: once Arigh Böke had been dealt with and Khubilai had secured his position as Great Khan, at least in the Far Eastern portion of the Mongol empire, he again turned his attention to the conquest of Song.Footnote ⁵⁹ Hostilities began in earnest in Sichuan in 1265, but it was to be more than a decade before Song was finally reduced to submission.Footnote ⁶⁰

The similarities between the events of 1242 and 1260 suggest that the Mongol withdrawal from Hungary in 1242 was indeed the result of the death of Ögödei Khaghan. However, the death of the Khaghan did not in itself make the withdrawal inevitable: it was the lack of consensus regarding the succession that more or less paralysed Mongol military activity for most of the 1240s. It has been said that the fact that Batu stopped when he reached the Kipchak steppe and did not go to Mongolia indicates that the withdrawal was not a necessary result of Ögödei's death.Footnote ⁶¹ However, Batu's purpose seems to have been to frustrate the succession of Güyük as Khaghan: from his own ulus, he could hope to exert some influence on the succession, while at the same time delaying any final decision in Güyük's favour by his refusal to attend a khuriltai. Batu was the senior living Chinggisid prince. He probably believed that the lack of his approval, as expressed by his attendance at a khuriltai, would invalidate any attempt to make Güyük Khaghan. Indeed, his refusal to participate in a khuriltai did produce a long delay.Footnote ⁶² When Güyük was eventually proclaimed Khaghan, without Batu's approval, it appears that civil war would probably have resulted, if Güyük had not suddenly died. This is mirrored in the events of 1260, when Khubilai's elevation to the khaghanate resulted in a civil war with Arigh Böke and his supporters.

Finally, it may be noted that the difficulty of obtaining consensus regarding succession was a recurring problem for the Mongols. The original system, by which a supreme ruler was proclaimed at a khuriltai attended by all the Mongol princes (and, indeed, many of their wives and other family members), quickly revealed serious failings. If a general consensus could not be obtained at a khuriltai, with at least no important dissenting voices, then the succession had to be delayed, or, if it was pushed through despite disagreement, the result was often civil war. The only succession to the khaghanate that proceeded reasonably smoothly was that of Ögödei. Thereafter, disputes were normal, and civil war was common. Güyük's succession was delayed for years. When Möngke became Khaghan, there was apparently a plot against him, and there ensued a bloody purge of his opponents, principally of the family of Ögödei.Footnote ⁶³ Khubilai's succession was disputed not only by Arigh Böke, but also by Khaidu: the unity of the Great Mongol Empire broke down after 1260.Footnote ⁶⁴

Internecine feuding plagued the Mongol khanates throughout most of their history. Khubilai Khan was challenged not only by his younger brother, Arigh Böke, and by Ögödei's grandson, Khaidu,Footnote ⁶⁵ but also by Nayan, a descendant of a younger brother of Chinggis Khan.Footnote ⁶⁶ Khaidu remained a problem for the Mongol rulers of the Yuan dynasty until his death about half way through the reign of Khubilai's successor.Footnote ⁶⁷ The Chaghadai Khanate fell under Khaidu's influence and also joined his struggle.Footnote ⁶⁸ The Persian Ilkhanate became involved in a dispute with the khans of the Jochid ulus over lands in the Transcaucasus region and Azerbaijan: the two khanates began a long, intermittent war in 1262.Footnote ⁶⁹ As noted above, it is distinctly possible that the withdrawal of the main army of the Ilkhan Hülegü from Syria in 1260 may have been related to this issue. There were further disputes over succession, both in the Yuan empire,Footnote ⁷⁰ and in the Ilkhanate.Footnote ⁷¹ The Jochid ulus suffered similarly.Footnote ⁷² It must surely be the case that, if the Mongols had not dissipated so much of their time, energies, and resources quarrelling and fighting among themselves, then their conquests would have extended even further, and their empire might well have lasted longer.