Introduction
One of the significant developments in historical research during the last generation or so has been the increasing cooperation of textual historians with natural scientists and scientifically informed historians in creating a richer and more nuanced picture of the human past. We have seen important advances based on insights and data from climatology, seismology, genetics, medicine, animal husbandry, various aspects of scientific archeology, and surely other fields. Further developments ignite the imagination and all of us look forward to hearing about new research and wonder how it might impact our own fields.
At the same time, the need for careful and exact textual research in original languages – ancient and modern – has not been alleviated. One needs to maintain high standards of philological work, with texts put into proper historical and cultural contexts and the rules of evidence assiduously followed. Only then can a happy marriage between textual historians and those working in the gamut of natural sciences be achieved and sustained.
Recently, two new articles argued for the central role that the Mongols’ conquests in the eastern Islamic world during the thirteenth century played in the dissemination of plague causing the pandemic wave referred to as the ‘Black Death’. According to this recent theory, plague was transmitted from its long-term reservoir of the Tian Shan mountains, in present-day Kyrgyzstan, across Central Asia to Europe and the Middle East through Iran and Iraq (not through the northern steppe), and the disease’s westward transmission was facilitated by the Mongol Empire in the thirteenth century, not the fourteenth century.
This thesis was broached by Monica H. Green in her paper ‘The Four Black Deaths’, and then by Nahyan Fancy and Monica H. Green in ‘Plague and the Fall of Baghdad (1258)’.Footnote 1 The authors presented textual evidence from contemporaneous Persian and Arabic sources from Mongol-ruled Iran and Iraq, and from thirteenth to fifteenth century Arabic chronicles from the Mamluk Sultanate of Syria and Egypt, to support three main arguments: first, the Mongol conquests, and especially, the conquest of Baghdad by the forces of Hülegü Khan (grandson of Chinggis Khan, who ruled in the Middle East ca. 1255–1265) in 1258, also entailed the outbreak of plague, which occurred shortly after Mongol sieges were lifted; second, a ‘plague-like disease’, which spread in Syria and Egypt after 1258, was linked to this occurrence of plague during the Mongol conquest of Baghdad. Green and Fancy suggest that additional support for these two arguments is found in a discernable shift in descriptions of plague symptoms by Muslim physicians and religious authors in the second half of the thirteenth century, that is, after the fall of Baghdad to the Mongols.Footnote 2 Their third argument is that this connection between the outbreak in Baghdad and the epidemic in Syria and Egypt, which near-contemporaneous sources made, ‘fell out’ of Arabic histories from the mid-fourteenth century, when a new wave of plague swept the Middle East, Asia, and Europe.
However, a recent article published in the journal Nature, on the basis of new aDNA evidence retrieved from three specimens from the Christian (East Syriac) cemetery at Kara-Djigach in modern-day northern Kyrgyzstan, calls into question the hypothesis of the thirteenth-century spread of plague associated with the early Mongol campaigns. The victims of a lethal ‘pestilence’ that struck the area’s denizens in the years 1338–9 were buried in this cemetery in Kara-Djigach.Footnote 3 Green had suggested that the genetic diversification (i.e. the ‘Great Polytomy’) that produced four new branches of the bacterium Yersinia pestis, one of which caused the fourteenth century pandemic (the ‘Black Death’), can be dated to the thirteenth century, and thus, to the period of Mongol expansion. Yet, the authors of this recent article have decisively demonstrated that the genetic evidence points to a fourteenth-century emergence, which fits the 1338–9 outbreak near lake Issyk-Kul. Importantly, the authors establish the phylogenetic position of the Kara-Djigach genomes, falling on a node shortly preceding and giving birth to the ‘Great Polytomy’, which is the genetic diversification that created the four new branches of Yersinia pestis. They hypothesize that Central Asia trade networks, rather than military campaigns, served to disseminate the bacterium, which re-emerged in the outbreak that occurred in the Black Sea region in 1346, shortly after the 1338–9 plague outbreak near lake Issyk-Kul.Footnote 4
In our article here, we take a different approach to this debate by focusing entirely on the historical evidence used by Fancy and Green. We argue that overall, the textual historical evidence used by Fancy and Green does not support their conclusions. First, we examine the authors’ argument about plague outbreak after the Mongol siege of Baghdad. Their thesis relies almost entirely on three near-contemporaneous historical accounts of the siege. However, a re-examination of the sources has led us to conclude that the main text (in Persian) has been significantly misunderstood, and the two other accounts (in Syriac and Arabic) have been mis-contextualized, and thus not understood properly. They do not support the authors’ claim regarding the plague outbreak in Baghdad in 1258. Secondly, we re-examine the sixteen Arabic sources, almost all from Ayyubid and Mamluk Syria and Egypt, that Green and Fancy organized into ‘generations’, and then surveyed to support their arguments that plague spread in the area in the aftermath of the Mongol conquests, and, that ‘the role of epidemic disease in the Mongol attacks was commonly known among chroniclers in Syria and Egypt’.Footnote 5 We can already note that their division of these sources into ‘generations’ is problematic, and instead, we identify four original accounts (Urtexte, so to speak) that were copied, summarized, and sometimes also slightly altered by subsequent authors over many decades and even centuries.
We examine these Arabic accounts in detail and reach four main conclusions: first, while the Ayyubid and Mamluk sources describe, indeed, the spread of an epidemic disease (or perhaps diseases) in Syria and Egypt in 1258, the symptoms they describe are not specific to plague, and thus, cannot be used to support Fancy and Green’s argument. Secondly, while the Arabic/Persian term ṭāʿūn, which Islamic medical texts associate with plague (characterized by swellings (buboes) and fevers, i.e. Bubonic Plague)Footnote 6 is indeed employed by several Mamluk accounts for this epidemic in 1258, these references can all be traced back to one single, contemporaneous account (by Ibn Wāṣil), on which subsequent authors relied and from which they copied (generally shortening it in the process). This contradicts Fancy and Green’s conclusion that there is an ‘extensive body of evidence in Arabic historical chronicles attesting to the presence of a plague-like disease’.Footnote 7
Thirdly, we find that while some Ayyubid and Mamluk historians did report the outbreak of an epidemic in Baghdad after the Mongol conquest due to the deterioration of the conditions in the city, overall they did not connect this disease outbreak in Baghdad in 1258 with the outburst of another epidemic in Syria and Egypt. Instead, we find several authors linking the slaughter of Baghdad’s inhabitants by the Mongols to the subsequent outbreak in Syria and Egypt using a miasmatic explanation, i.e. that the disease was caused by the corruption of the air engendered by the massacres in Baghdad.
Finally, while the subtle shift to more accurate, yet still succinct descriptions of plague symptoms (as well as the association of plague with the appearance of the bubo specifically in the armpits, along with inflammation) in Arabic medical commentaries largely in the second half of the thirteenth century may be important,Footnote 8 we see no textual evidence for a connection between this shift and the 1258 epidemic in Syria and Egypt, and certainly no connection to the Mongol conquest of Baghdad. None of the thirteenth-century medical authors discussed by Fancy and Green explicitly state in any way that they had witnessed plague in Syria, Egypt, or Iraq. For example, in a recent article, Fancy introduced an important medical account of plague symptoms by the Damascene Ibn al-Nafīs (d. 1288, active, however, in Egypt); yet Ibn al-Nafīs’ description is based on an eyewitness account of a friend of his, and the latter witnessed and described a plague outbreak in Ethiopia (!), not in Syria or Egypt.Footnote 9 Moreover, Fancy argues that these more accurate accounts appear already in the 1240s, namely more than a decade before the Mongols besieged Baghdad.Footnote 10
Plague during the Siege of Baghdad in 1258?
The linchpin for Fancy and Green’s analysis of plague outbreak after the Mongol siege over Baghdad and thus of the role that the Mongols had in the ‘spillover event’ that spread the bacterium Yersinia pestis, which causes plague, from its reservoir in the Tian Shan mountainsFootnote 11 is a passage from a short Persian chronicle attributed to the savant and official Quṭb al-Dīn Shīrāzī, titled Akhbār-i Mughūlān (‘Mongol News’), an edition of which was published only recently, in 2010.Footnote 12 According to Green, this passage fits the pattern found in other contemporaneous textual accounts from China, according to which mysterious epidemics appeared shortly after Mongol sieges were lifted in China, as has been identified and discussed by Robert Hymes. He has further noted that Chinese medical writings record the appearance of a brand-new symptom—large purulent lumps or sores—in some descriptions of lethal epidemic outbreaks beginning in the Mongol incursions in the early to middle thirteenth century. This symptom fits plague buboes.Footnote 13 Following Hymes, Green (and then Fancy and Green) suggest that Akhbār-i Mughūlān indicates the same ‘epidemic profile’. They hypothesize that the Mongol armies brought with them plague, likely transmitted via the rats infesting the grain supplies they carried from further east, which struck the local population after the siege was lifted. Here is the translation of the text by Shīrāzī that Green uses:
When Hulegu [sic] reached Baghdad the rest of his army, who were already in the city, was standing on the ramparts. Because a great assemblage of people, namely all the people of the Sawād, had come to the city before the Mongol army arrived, there was a great dearth, want, and scarcity of provisions in Baghdad. Pestilence struck and many people died. The number of deaths reached the point that the Ministry’s priority was to prepare the corpses and bury them. Meanwhile the situation deteriorated so much that the people of Baghdad could no longer cope with ablutions and burial of the dead, so the bodies were thrown into the Tigris.… Even when the army arrived, they were unable to cope with the situation.Footnote 14
The original translation used by Green, however, is unclear as to when exactly the pestilence appeared in Baghdad. A close reading of the original Persian decisively shows that the author conveyed that the disease had struck the denizens of Baghdad before the Mongol forces arrived at the city’s walls. Here is our revised translation:
When Hülegü reached the gate of Baghdad, the rest of the army [the caliph’s army, according to prior passages] that had already been in the city [to defend the city from the oncoming Mongols] was standing on the ramparts. Before the Mongol army arrived at Baghdad, epidemic (wabāʾ) had struck [the city] and many people died [our emphasis]. This was due to the assemblage of many people [in the city], namely all the people of the Sawād,Footnote 15 who had come to the city [seeking refuge from the Mongols], and there was therefore great dearth, want, and high prices. The number of deaths reached such a point that [payment] would be made immediately from the Treasury to prepare the corpses [for burial] and they were to be buried [without delay]. Meanwhile the situation deteriorated so much that the people of Baghdad could no longer cope with ablutions and burial of the dead, and they would throw bodies into the Tigris…. Even when [Hülegü’s] army arrived, they [the residents of the Baghdad] were still not clear of it [i.e. the pestilence and the burial of the dead].Footnote 16
The Arabic/Persian term wabāʾ mentioned in Akhbār-i Mughūlān can refer to several contagious diseases, as it is a general term for deadly disease. It may refer to plague, although it usually appears more as an ‘umbrella term’ for epidemics. The term ṭāʿūn (below) specifically designates plague in Arabic and Persian sources.Footnote 17 A correct understanding of the text, considering what comes before and after it, clearly shows that an epidemic broke out in Baghdad before the Mongol armies arrived and began to besiege the city. One can suppose that in the crowded and disturbed conditions in the city caused by an inflow of refugees from the surrounding countryside, without adequate food, living in squalor, and lacking basic sanitation for many days, and compounded by uncertainty and dread, the city was ripe for a mass outbreak of a disease of some type, or many types (dysentery, typhus, typhoid fever, etc.). This was often the case in sieges, or in this case, the days and weeks before a siege. Thus writes A.D. Lee, when discussing parallel events in late antiquity:
[M]alnutrition left individuals more susceptible to disease, and in the often-crowded conditions of a siege could spread rapidly, with fatal consequences for many. Problems with the appropriate disposal of bodies in turn exacerbated insanitary conditions, as many of those at Amida in 359 and Rome in 409 found to their cost.Footnote 18
The same reservation must be made regarding the two other examples that Green provides for plague outbreak after Mongol sieges in the eastern Islamic world. She notes that a wabāʾ outbreak is reported in the Ismāʿīlī (‘Assassin’) fortress of Lanbasar (Lammasar) ‘at the end of the siege, in 1257’. She then proceeds to suggest that this outbreak ‘oddly’ parallels ‘the circumstances of the plague outbreaks in the sieges in China’.Footnote 19 Yet, neither of the two sources referenced – Shīrāzī and Rashīd al-Dīn – indicate that the epidemic appeared after the siege was lifted; rather, both sources confirm that the fortress lasted under siege for about a year until an epidemic broke out, killing most of the besieged men.Footnote 20 The outbreak is plainly mentioned by both authors as hastening the fall of the fortress to Mongol hands, and not as resulting from its fall. The same can be said about the earlier siege at Girdkūh in 1253. The outbreak of an epidemic during the Mongol campaign required the Ismāʿīlī ruler at Alamūt to send reinforcements to the castle to keep it from falling to the Mongol forces (which it managed to do for some time). Here, too, there is no evidence that the disease was brought by the besiegers.Footnote 21
Another key piece of evidence adduced by Green (followed by Fancy and Green) is from the Syriac Aramaic chronicle by Bar Hebraeus (d. 1286). We offer a slightly different version from Budge’s translation used by Green:
And then, there was a very severe famine and a pestilence (māwthānā) in the land of Babylonia (Senʿār), Assyria (Āthūr), Mesopotamia (Bēth Nahrīn), Syria (Sūrīya), and Anatolia (Bēth Rūmaye), so a young pigeon for a sick man was sold in Damascus for 12 Nasiri dirhams.Footnote 22
Fancy and Green see this as further proof of the spread of plague, which was earlier seen in Baghdad. We cannot agree, and state with a high degree of certainty that Bar Hebraeus is dealing here with a later event, some three or four years after the sack of Baghdad. The problem here is not the slight difference in translation, but the context. Looking at the entire paragraph, it is clear that the author is not at all referring to events right after the Mongol conquest of Baghdad, but has gone off on a tangent to discuss developments over several years in the city of Irbil (in northern Iraq of today) and the hill country to the north populated by Kurds (then, as today).
We need not render the entire paragraph, but merely note that before dealing with the ‘severe famine and pestilence’, it talks about ‘rebellious Kurds’ in a place that Budge gives as Jûlmarg. In Arabic this is the city of Jūlamark (modern Çölemerik in Turkey, about 150 km north of Irbil). Mamluk accounts, indeed, report a rebellion there, but in 1261–2 and not earlier.Footnote 23 According to Bar Hebraeus, only then, at the earliest, did famine and pestilence break out in the regions listed by him. This fits very well with the parallel information from the Arabic sources that in late 1261, Syria suffered bad harvests (due, it is claimed by one author, to an outbreak of mice), and there was a shortage of food resulting in high prices.Footnote 24 In short, Bar Hebraeus provides no support for a raging epidemic of any type in Iraq, upper Mesopotamia, and Syria in 1258.Footnote 25 There may well have been a mass health emergency in these regions during this year (see below for Syria), but the Syriac version of Bar Hebraeus provides no evidence of this.
The third piece of evidence provided by the authors is taken from the anonymous Arabic chronicle of Baghdad in the seventh Hijri century (=1203–1300 CE), al-Ḥawādith al-jāmiʿa, which traditionally had been attributed to the Iraqi historian Ibn al-Fuwaṭī (d. 1323). In the description of the momentous events in 656 Hijri (=1258 CE), this is what we find relating to the massacre of the population and then the ensuing famine:
It was said that the number of dead in Baghdad was more than 800,000 souls, besides those children thrown into the mud, and those who died in the canals, wells and cellars, perishing from hunger and fear. Pestilence broke out among those who survived the massacre, from the stench of the dead, and from drinking water contaminated by corpses. People had much recourse to the smelling of onion due to [the reek] of the cadavers and the multitude of the flies, which filled the area, and alighted on foods, spoiling them.Footnote 26
This seems straightforward to us: given all the dead bodies—victims of the massacre and the subsequent famine—and contaminated water and food, is it any surprise that an epidemic (or epidemics) broke out? Does one need to resort to the plague to explain the mass outbreak of disease here?
In summary, we see no clear evidence in these three passages that there was an outbreak of plague in Baghdad or adjacent regions to the north around the time of the Mongol conquest of the city in early 1258. The first passage, by Shīrāzī, mentions an outbreak of wabāʾ in the city before the beginning of the Mongol siege after many people from the countryside of Iraq fled to the city. The second passage, in the Syriac version of the chronicle by the Gregorius Bar Hebraeus, does not apply to 1258 at all, but to events of three or four years later. Finally, we have a description in the anonymous Arabic Baghdadi chronicle of mass disease in Baghdad among the local survivors after the Mongol conquest. Certainly, in both the first and third passages, the outbreak of epidemics is readily explained by crowded conditions, poor sanitation, a plethora of unburied corpses, and the resulting pollution of water (and maybe food) sources, all compounded by fear and uncertainty.
Whatever affliction spread among the city of Baghdad either before the Mongol armies arrived or immediately following the city’s surrender to Hülegü, we might expect that it would have also affected the Mongol armies. According to Akhbār-i Mughūlān, after Hülegü’s army returned from Baghdad to Azerbaijan in the year 656 Hijri (1258), ‘the weather had become warm, and a great stench (ʿufūnat) [i.e. corrupted air] entered the people’s brains. [Subsequently] an epidemic (wabāʾ) struck, and most of the Mongol army was afflicted, and many died’. The source further reports that Hülegü himself became afflicted, but after twenty days recovered.Footnote 27 Did this epidemic have anything to do with the disease that afflicted the residents of Baghdad before (and likely during) Hülegü’s siege of the city? Possibly, although the affliction might have been related to the heat wave reported in the account. In either case, there is no evidence that we are dealing with plague. Rather, it is important to note that Akhbār-i Mughūlān reports a disease that afflicted the Mongol forces after Baghdad’s fall, making unlikely the possibility that the Mongols served as vectors that delivered the pandemic to Baghdad in the first place.
Another problem with Fancy and Green’s theory relates to the spread of the disease after the fall of Baghdad. Right after the above-cited description in al-Ḥawādith al-jāmiʿa, the anonymous Arabic chronicler of Baghdad continues:
The people of Ḥilla, Kūfa and Sīb [all in southern Iraq] brought to Baghdad food, from which the [local] population took sustenance. These [people from the nearby cities] took for the prices of [this food] precious books, inlayed copperware, and valuable furniture. A large number of these people were enriched in this way.Footnote 28
It seems, then, that visitors from other cities in Iraq, who immediately exploited the desperate situation of the surviving Baghdadis, remained untouched by the ravages of plague that supposedly affected the last mentioned. How so?
A Re-examination of the Mamluk Sources for an Epidemic in Syria and Egypt in 1258
Beyond these three accounts, the Arabic sources written at the close of Ayyubid rule in Syria (1260 CE) and in the Mamluk Sultanate (in Egypt from 1250–1517; in Syria 1260–1516) offer some evidence for epidemics in Baghdad, as well as for Syria and Egypt in 1258, after the fall of Baghdad. In a table, and then in their discussion, Fancy and Green note sixteen sources conveying this material,Footnote 29 concluding that ‘the role of epidemic disease in the Mongol attacks was commonly known among chroniclers’.Footnote 30 They further argue that a peculiar historiographical ‘erasure’ took place, in which later, post-Black Death authors failed to identify the pandemic’s thirteenth-century episode.Footnote 31
To prove this ‘erasure’ of evidence, they put together an impressive survey of a large corpus of passages, usually in chronicles that are particularly verbose for the events of the annus horribilis of 1258.Footnote 32 Their argument here is based on a division of these sources by generations.Footnote 33 This division makes little sense to us, and we argue instead that a better approach is identifying four Urtexte from historians, apparently all contemporary to the events of that year, and then, observing how later authors copied, summarized, or subtly changed these earlier observations. We have thus concentrated on these original accounts, keeping our eyes open for novel information that might have been added by subsequent authors. To keep an open mind, we have not translated the two Arabic words wabāʾ and ṭāʿūn. Footnote 34
When looking at these sixteen texts (and several more that we have added), we can identify four separate accounts that are passed down from one historian to another over a period of almost two centuries, along with a couple of ‘dead ends’ that were not reproduced by later writers. Here the authors demonstrate, as noted above, a well-known phenomenon in the historiography of the Ayyubid and Mamluk eras: later writers cite passages (at times, also the names of the earlier authors are noted), sometimes almost word for word, but often summarized (occasionally in a rather terse way), and once in a while, new information is added, usually tacitly.Footnote 35 Actually, as we will see below, all four original accounts are repeated with few additions, and if anything, they generally get shorter over time. On the other hand, we do have one example where a later writer renders a much fuller text by an unnamed source, but we will suggest that this particular passage probably harks back to the time of the events themselves and is only partially rendered by historians of the next generation.
Briefly, the four original accounts are as follows:
-
1. Wabāʾ broke out in Baghdad following the fall of the city, and high numbers for the dead are provided. The original account here is the passage cited above from the anonymous (pseudo-Ibn al-Fuwaṭī) Baghdadi chronicle al-Ḥawādith al-jāmiʿa.
-
2. An epidemic (referred to as both ṭāʿūn and wabāʾ) erupted in Syria and Egypt in 1258. The original author of this account is Ibn Wāṣil. He himself came down with it in Cairo but recovered. There is a quick survey of mention of ṭāʿūn from a medical text and early Islamic history. The Ayyubid prince al-Nāṣir Dāwud got sick and died; according to the source, this was ṭāʿūn.
-
3. In the aftermath of widespread disease (maraḍ) and wabāʾ in Syria, large numbers died in Aleppo and Damascus, but in the former we get an exact daily figure. We learn of the suffering of the inhabitants, along with the high prices and shortages in both cities.
-
4. Mention en passant of ‘sickness’ in Egypt, as part of the biography of a poet and courtier there.
Account 1: We start with al-Ḥawādith al-jāmiʿa. Footnote 36 This passage has been cited above, but now we are examining its historiographical impact; we provide here a fuller rendition than above:
The dead bodies lay in mounds in the alleys and market. Rain fell upon them, and the horses trampled on them. Their shapes were disfigured, and they became an example (i.e., embedded in their memory) to whomever saw them. Then an amān (amnesty) was announced. Those who survived came out [from hiding], but their faces went white and they were shocked by the horror that they saw, that could not be described in words. They were like the dead that had emerged from the graves on the day of resurrection, [suffering] fear, hunger and cold…Footnote 37
It was said that the number of dead in Baghdad was more than 800,000 souls, besides those children thrown into the mud, and those who died in the canals, wells, and cellars, perishing from hunger and fear. Wabāʾ broke out among those who survived the massacre, from the stench of the dead, and from drinking water contaminated by corpses. People had much recourse to the smelling of onion due to [the reek] of the cadavers and the multitude of the flies, which filled the area, and alighted on foods, spoiling them. The people of Ḥilla, Kūfa and Sīb [all in southern Iraq] brought to Baghdad foods, from which the [local] population took sustenance. These [people from the nearby cities] took for the prices of [this food] precious books, inlayed copperware, and valuable furniture. A large number of these people were enriched in this way.Footnote 38
Later authors summarized this passage. Ibn Shākir al-Kutubī (d. 1363), author of ʿUyūn al-tawārīkh, gives a short rendition of both these passages:
It is said that the number of dead in Baghdad was more than 1,800,000 souls [N.B. 800,000 has become 1,800,000], besides those children thrown in the mud, and those who died in the canals, the wells and cellars, perishing of hunger and thirst. Wabāʾ broke out among those who survived the massacre, from the stench of the dead, drinking water contaminated by corpses, and the multitudes of flies, which filled the area, and alighted on foods, spoiling them.Footnote 39
A parallel, but somewhat different, text is from Ibn Kathīr’s (d. 1373) al-Bidāya wa’l-nihāya:
People disagreed about the amount of those who died in Baghdad among the Muslims in this event. It was said 300,000, [also] it was said 1,800,000, and [finally] it was said that the dead reached 2,000,000 … An amnesty (amān) was announced in Baghdad, and those who had been under the ground emerged from underground storerooms, subterranean water cannels and tombs, as if they were the dead when exhumed from their graves. One would disown the other: the father did not know his son, and one did not recognize his brother. A strong wabāʾ afflicted them, and they died, joining those who had preceded them in death.Footnote 40
This version is rendered in a slightly shorter form by al-ʿAynī (d. 1451) in his ʿIqd al-jumān. Footnote 41 Al-ʿAynī does not name his source, as he often does, but elsewhere in his account of the conquest of Baghdad, he cites Ibn Kathīr, so this later writer is most probably al-ʿAynī’s direct source here as well. Other authors give even shorter versions, concentrating on the numbers who died in Baghdad.Footnote 42 Al-Dhahabī (d. 1348) provides in Taʾrīkh al-islām the figure of 300,000 dead, not an unrealistic number.Footnote 43 We will return below to other passages by this last-mentioned author.
In summing up this particular Urtext and its derivative passages, we can first note again that there is nothing here indicating plague, but rather diseases caused by contaminated water and other conditions. We can also note for now that the graphic details presented by the anonymous author gradually disappear in most of the works of later historians (al-ʿAynī is an exception). What remains is the focus on the numbers of the dead, which are rendered in an exaggerated manner; al-Dhahabī (followed by Ibn Taghrī Birdī), however, returns us to reality by citing ‘just’ 300,000 dead. Finally, neither the anonymous writer nor any of the later authors who used his account, either directly or indirectly, linked this disease outbreak in Baghdad to the epidemic that struck Syria and Egypt later in 1258. Other authors, however, did imply a relationship between the epidemic that spread in Syria and Egypt and the Mongol assault on Baghdad. The first author to do so appears to have been Ibn Wāṣil (d. 1298),Footnote 44 to whom we now turn.
Account 2: In his chronicle Mufarrij al-kurūb, Footnote 45 Ibn Wāṣil (d. 1298) reports first how tāʿūn erupted in Syria and Egypt, in the aftermath of the Mongol conquest of Baghdad. This author himself came down with this disease in Cairo, but recovered. The author relates how the Ayyubid prince al-Nāṣir Dāwūd got sick and died near Damascus.Footnote 46
Ṭāʿūn spread throughout all of Syria and Egypt, and to other places. Corruption (fasād)Footnote 47 caused this, leading to the affliction of the wabāʾ and the changing of temperaments (amzija). I saw the most amazing thing in Egypt: in Bilbis high fever and coughing hit, so that nearly no one was spared it, but in Cairo no one was affected. Then, two days later, something like this happened in Cairo. At that time, I was staying in Giza; I rode to Cairo, and I found the matter (i.e., the sickness) there afflicting almost all the people of Cairo. I returned to Giza, and just settled in when I came down with it too, and it spread to the people of Giza. In most cases the coughing lasted more-or-less three days. Most people survived.Footnote 48 It passed from one town to the next, until it gradually reached the farther regions of Egypt.
Galen described it in a like matter: there was slaughter (malḥama) in Greece, and wabāʾ afflicted the land of Nubia after a while. ʿAbd Allāh b. al-Faḍl, one of al-Malik al-Nāṣir Dāwūd‘s entourage, reported:
When the wabāʾ and ṭāʿūn intensified in the aftermath of the Mongol conquest of Baghdad, we were dismayed (tasakhkhaṭnā) by it. Al-Malik al-Nāṣir said to us: ‘Don’t be dismayed.’ As for the ṭāʿūn, when it struck ʿAmwās during the caliphate of ʿUmar b. Khaṭṭāb (r. 634–44)—may God be satisfied by him—he said to some of the people: ‘This is the punishment of God (rijz); this is the disaster (ṭūfān) that was set upon the Children of Israel.’Footnote 49
This reached Muʿādh b. Jabal (d. 639),Footnote 50 may God be satisfied by him, who functioned as preacher (khāṭib) among the [Muslim] people, who said:
O people! Why do you call to your Prophet—may God’s prayer and peace be upon him—and for the mercy of God from the suffering [of the pestilence]. Think that the ṭāʿūn is the disaster (ṭūfān) that was set upon the Children of Israel.Footnote 51 The ṭāʿūn is the mercy of God, who forgives you, and the prayer of your Prophet upon you. O God, give to the family of Muʿadh their complete portion.
[ʿAbd Allāh b. al-Faḍl] said: ‘Al-Malik al-Nāṣir told us about the death of Muʿādh, his son and his family by ṭāʿūn.’ Then al-Malik al-Nāṣir prayed, saying: ‘O God, make us like them and make our lot what you gave them.’ He woke up the next morning or afterwards, afflicted by the ṭāʿūn (wa-aṣbaḥa … maṭʿūnan).Footnote 52 I was absent. When I heard of his illness, I came to him. He complained [of a pain] like a stab (ṭaʿn)Footnote 53 from a sword in his left side that prevented him from lying down. At dawn of Thursday, this eased up, and he lay down and awoke with a fever.
His son al-Malik al-Muẓaffar Shihāb al-Dīn Ghāzī, who now lives in Cairo, related that [al-Nāṣir Dāwūd] slept between the [first] two prayers, and then he woke up and said: ‘I saw my left side say to my right side, “My turn has come and I put up with it, and tonight your turn will come, and you will put up with it, as I did.”’ When evening came, he complained of a slight pain on his right side, and this began to get worse. We knew that this was ṭāʿūn. I was with him between the two prayers, and his strength was diminishing, when sleep overtook him. He woke up, and he was shaking. He motioned for me, and I drew near to him. He said, ‘I saw the Prophet—may God’s prayer and peace be upon him—and al-KhiḍrFootnote 54—peace upon him—came to me and sat with me; then they left.’ At the end of the day he said, ‘I have no hope, so prepare my funeral.’ I wept and all those present wept … He entrusted his family and children to me. Then at night, he became weaker… [Details of a further dream, and his funeral; death not explicitly mentioned].
As Fancy and Green note, Ibn Wāṣil provides in his account a miasmatic explanation based on Galen for the disease outbreak in Syria and Egypt.Footnote 55 According to this theory, corrupted air due to changes in the seasons or the temperatures, or from noxious vapours and unpleasant stenches from fires, stagnant waters, or the rotting of unburied corpses, can cause contagious diseases. The polluted air is inhaled by individuals corrupting their internal organs and spirit. Moreover, this corrupted air can spread quickly over a large distance with the winds. Footnote 56 As the reader can see, however, while Ibn Wāṣil mentions the miasmatic explanation for the epidemic in Syria and Egypt at the beginning of the account (due to corruption, fasād), he does not make the connection with the Mongol attack on Baghdad in this long passage explicit, just noting further along: ‘When the wabāʾ and ṭāʿūn intensified in the aftermath of the Mongol conquest of Baghdad, we were dismayed by it.’Footnote 57
A number of later writers provide fairly short versions of this long and detailed account, centering around the death of al-Malik al-Nāṣir Dāwūd.Footnote 58 Worth mentioning first is al-Dhababī (d. 1348), who cites by name Ibn Wāṣil:Footnote 59
The ṭāʿūn spread in Syria, after it had finished in Iraq. Al-Nāṣir [Dāwūd] was ‘stabbed’ (ṭuʿina, implying he was affected by ṭāʿūn). Ibn Wāṣil said that it spread in spite of the distance from Baghdad. Galen related that after the slaughter (malḥama) in the land of Greece, there was wabāʾ because of it in the land of Nubia, in spite of the distance.
Then he gives a shorter version of al-Nāṣir Dāwūd getting sick, telling his comrades not to complain about the wabāʾ (N.B., compare with his source, Ibn Wāṣil, who writes wabāʾ and ṭāʿūn here), the beginning and intensification of his sickness, and death. A brief version of this story is also related in the Mukhtaṣar by the Ayyubid prince and scholar Abū al-Fidāʾ (d. 1331),Footnote 60 who generally summarizes Ibn Wāṣil’s chronicle for the decades before he himself was a contemporary of events. He, in turn, is apparently followed by Ibn al-Wardī (d. 1348-9).Footnote 61 Abū Shāma (d. 1268), a contemporary resident of Damascus, notes briefly the death of al-Nāṣir Dāwūd, but does not mention his disease.Footnote 62
What do we learn from the above passage from Ibn Wāṣil and its derivatives? The original author reports that ṭāʿūn spread in Syria and Egypt, and he himself got it in Cairo. He recovered, however, as did most people. A prominent and well-respected Ayyubid prince in Syria, al-Nāṣir Dāwūd, who, however, was no longer enjoying any political power, came down with ṭāʿūn, eventually dying. We can note that Ibn Wāṣil and his ‘epigoni’ go back and forth between the terms ṭāʿūn and wabāʾ, when describing this disease, suggesting that perhaps there was no clinical distinction between them in their minds.
Before moving on, we can also observe that to our understanding there is no symptom in Ibn Wāṣil’s report that leads us to consider that we are dealing here with plague. The disease that broke out in Egypt, which he himself came down with, was characterized by fever and coughing, with a low deathrate. Fever is certainly not unique to plague, nor is coughing. These suggest other diseases, influenza perhaps (see in next section for more on this). Beyond the use of the word ṭāʿūn (and the passive participle and verb derived from its root), there is little to tie this disease to plague. We, therefore, should consider the possibility that al-Nāṣir Dāwūd’s affliction was identified with ṭāʿūn, not based on symptoms, but rather to assign al-Nāṣir Dāwūd the same idea of divine merit (‘God’s mercy’) and martyrdom that Islamic traditions associated with those afflicted with plague in the seventh-century epidemic at ʿAmwās.Footnote 63
While Fancy and Green indeed observed that Ibn Wāṣil was the first chronicler to employ ṭāʿūn with regard to the disease in Syria and Egypt (as well as the first to use the miasma explanation),Footnote 64 they did not stress the fact that he was also the only source for all subsequent uses of the term ṭāʿūn for identifying the epidemic in Syria and Egypt.
Account 3: Ibn Wāṣil (d. 1298), al-Yūnīnī (d. 1326), and al-Dhahabī (d. 1348) provide additional important information regarding wabāʾ in Syria, in Damascus and Aleppo in 1258, in accounts that share many elements, but are not identical. They describe the deathrate in Aleppo, the high prices and shortages in both cities, and overall suffering. In addition, al-Dhahabī also unequivocally notes the miasmatic link between the Mongol slaughter of Baghdad’s denizens and the epidemic in Syria. These accounts have been surveyed and discussed by Fancy and Green, and they suggest a reliance of al-Yūnīnī and al-Dhahabī on Ibn Wāṣil.Footnote 65 On the contrary, we propose that these three historians share a hitherto unknown common source. In fact, we have uncovered the fullest version of this passage, which we now present, and suggest that it represents a contemporary account, as well as the common source for all three previously mentioned reports. Our source is an unpublished manuscript volume, found in the Vatican Library, from Taʾrīkh al-duwal wa’l-mulūk by the Egyptian historian Ibn al-Furāt (d. 1405). The translation is found below, while the Arabic text is in the Appendix:
One of the historians said that the sickness (maraḍ) and wabāʾ afflicted the people in Syria after the taking of Baghdad, and so that the people attained [access to] pharmacists only with great effort and difficulty. The pharmacists were enriched, medicines were used up, and physicians and bloodletters (or cuppers) were hardly found. Making a living was impossible, and there was no request for merchandise. The people acted towards the pharmacists as they would towards bakers during a famine. The situation lasted days. As for Aleppo, death there was greater: news arrived that every day 1200 funerals left the city. In Damascus, many people died on the street and in the hospitals, and both Qur’an readers and corpse washers were hardly found. Most of the people of the city were afflicted by fever and coughing [as if it was] one disease (maraḍ wāḥid). A number of notables died. As for those who did not have [medical] care, a large number [died among them]. At this time, a slice of green melonFootnote 66 was sold in Damascus for a dirham and a raṭl of tamar hindī Footnote 67 went for 60 dirhams. As for pullets (al-farārīj, i.e. young chickens), they were not to be found, but if one the size of a bird was found, they were selling (it) for one dirham, and this became three or four dirhams [for a bird]. As for Aleppo, the price of every pullet there reached 10 dirhams. A short [description] absolves [us] of a detailed one. God knows best.Footnote 68
Unfortunately, Ibn al-Furāt, who is usually scrupulous in citing his sources by name, leaves us in the dark here; yet this source was likely a contemporary or near contemporary of the events. It is clear, however, that he has brought us the fullest text for the of this particular chain of accounts, and thus we propose that it represents the common source for the versions by Ibn Wāṣil, al-Yūnīnī, and al-Dhahabī.Footnote 69 There is no a priori reason to reject the contemporary nature of Ibn al-Furāt’s account, and the fullness of its description adds to the consideration of being the earliest version of this particular account.
Ibn Wāṣil, whom we have seen was a contemporary of these events, has the following to say:
In this year, that is to say 656 [AH, i.e. 1258], the wabāʾ got worse in Syria, especially in Damascus, so that scarcely a washer of corpses could be found. The numbers who perished among the city’s population could not be counted. The price of pullets became dear, and they were not to be found. If one was found, it was quite small, and was sold for two or more dirhams.Footnote 70
Being a contemporary does not mean that everything Ibn Wāṣil wrote was in ‘real time’: the long chronicle Mufarrij al-kurūb was begun in 1272 and completed only in 1285.Footnote 71 This would have been plenty of time for another account of this outbreak of wabāʾ to have appeared and be summarized by Ibn Wāṣil. The fact that he placed this short passage far away from his long personal account (see Account 2 above), also leads one to think that it was not necessarily originally his own eyewitness version.
Al-Yūnīnī, who would have been 16 years old when the wabāʾ broke out, provides another version in his Dhayl mirʾāt al-zamān. However, rather than relating personal memories, al-Yūnīnī too probably summarized the long account that Ibn al-Furāt conveyed:
In this year, wabāʾ intensified in Syria, and an uncountable number of people of Damascus died. [The price] of pullets and other [foodstuffs] that serving the sick shot up. A raṭl of tamar hindī cost 60 dirhams and the slice of a green melon was a dirham.Footnote 72
From here we can go directly to his younger contemporary, al-Dhahabī, in his Taʾrīkh al-islām:
And during this year [AH 656] the wabāʾ got worse in Syria, and [so many] people died, so that it is said that in Aleppo every day 1200Footnote 73 funerals set out. As for Damascus, there was unlimited and indescribable sickness (al-maraḍ) and the pharmacists were enriched, medicines were used up, and physicians were hard to find. Pullets were sold in Damascus for three dirhams, and in Aleppo for 10 dirhams. The beginning of the wabāʾ was in Jumādā I (May-June 1258), due to the corruption (fasād) of the air in the slaughter (malḥama) at Baghdad. Footnote 74
These two texts, like that of Ibn Wāṣil, share data and formulations, but each is different. Again, at the risk of overstating the case, the reliance on a common source—evidently preserved by Ibn al-Furāt—seems very likely. There is here, however, an interesting and unique piece of information conveyed by al-Dhababī: ‘The beginning of the wabāʾ was in Jumādā I (May-June 1258), due to the corruption (fasād) of the air in the slaughter (malḥama) at Baghdad.’ This is certainly not in Ibn al-Furāt’s longer text, where it is only written that ‘the sickness (maraḍ) and wabāʾ afflicted the people in Syria after the taking of Baghdad.’ It is clear, however, that the terms ‘corruption’ (fasād) and ‘slaughter’ (malḥama) hark back to Ibn Wāṣil’s long account discussed in the previous section. Al-Dhahabī has deliberately, and seamlessly, combined elements from two accounts—the ‘Urtext 3’ preserved evidently by Ibn al-Furāt and ‘Urtext 2’ from Ibn Wāṣil’s chronicle. Writing at least half a century after the events he is describing, al-Dhahabī might have believed in the miasmatic connection between the slaughters in Baghdad and later pandemics in Egypt and Syria (which he refers to as both ṭāʿūn and wabāʾ). This, however, is far from compelling evidence that we are dealing with the same disease, let alone that it was transferred from Iraq to Syria, and then to Egypt.Footnote 75
One matter raised by Ibn al-Furāt, however,Footnote 76 needs to be addressed: the occurrence of ‘fever and coughing’ (bi’l-ḥummay wa’l-suʿāl) in Damascus might not be just some form of influenza or another ‘regular’ respiratory disease, but rather symptoms of pneumonic plague. About the latter, we can cite from Gage and Beard (insertions in square brackets are ours):
Pneumonic plague is the most rapidly developing and life-threatening form of plague. The incubation period for primary pneumonic plague is usually 2–5 days (range 1–6 days). Illness onset is most often sudden, with chills, fever, body pains, headache, weakness, dizziness and chest discomfort. Cough, sputum production, increasing chest pain, tachypnea [rapid breathing] and dyspnea [shortness of breath] typically predominate on the second day of illness; hemoptysis [spitting up blood], increasing respiratory distress, cardiopulmonary insufficiency and circulatory collapse can also occur. The sputum of primary plague pneumonia patients is typically watery or mucoid, frothy and blood tinged, and can be bloody.Footnote 77
We do not have detailed symptoms from any of the sources that can confirm or rule out the above in either Aleppo or Damascus. On the other hand, there was a high deathrate for at least a few days in both cities (with explicit numbers in Aleppo). One can only suggest that had there been a mass outbreak of pneumonic plague in these cities, the deathrate would have been even higher, and the symptom descriptions so much more vivid. The use by all sources of the term wabāʾ, even by the ‘ṭāʿūn informed’ al-Dhahabī, also inclines us not to see plague here, pneumonic or otherwise. We are still on safe ground when we claim that another type of respiratory diseases was wreaking havoc here.Footnote 78
Let us sum up this section as follows: ‘sickness and wabāʾ’ are mentioned in this Urtext and its derivative passages, not ṭāʿūn (except for al-Dhahabī, clearly influenced by Ibn Wāṣil). The symptoms were fever and coughing (and no more), again reminiscent of influenza, not plague. The number of dead was indeed high. Medicine, medical and funeral experts, and other supplies were scarce and thus expensive. This was a difficult situation, but it was not directly connected to the Mongol conquest of Baghdad earlier in the year, nor caused by plague.
Account 4: Another transmission thread about the epidemic in Egypt is provided in the biographical collection Wafāyāt al-aʿyān by Ibn Khallikān (d. 1282). In the biography of one Bahāʾ al-Dīn Zuhayr b. Muḥammad al-ʿAtaqī,Footnote 79 a poet and courtier who died during an epidemic in the year 1258 in Egypt, we learn that:
A great sickness (maraḍ) fell on Cairo and Fustat (al-Qāhira wa’l-miṣr), from which few were spared. It began on Thursday, 24 Shawwāl, the year 656 (AH, October 30, 1258). The above-mentioned Bahāʾ al-Dīn was one who suffered from it. He continued in this way for a few days, then dying before the evening prayer on Sunday, Dhū al-Qaʿda of this year … I did not pray for him, as I was myself preoccupied with the disease.Footnote 80
Some fourteenth century writers summarize this information: al-Dhahabī and al-Ṣafadī briefly note that after Bahāʾ al-Dīn returned to Cairo following a stay in Syria, ‘he was sick for a few days with wabāʾ and died.’Footnote 81 Clearly, maraḍ and wabāʾ were equated by contemporary and slightly later contemporaries.
To all the above we might add some independent pieces of testimony, none of which seem to have been picked up by other writers. Another contemporary was the legal scholar and historian Abū Shāma (d. 1268), a resident of Damascus, who wrote tersely:
During the time of Spring [656 AH/1258 CE], there was a lot of wabāʾ, and it was stranger than what was [usually] written in history. The disease (maraḍ) spread among the people, and there was much death.Footnote 82
Finally, the historical work by the famous Ibn Khaldūn (d.1406), Kitāb al-ʿibar, is unfortunately not very helpful here, writing only that ‘the Mongols had already conquered Baghdad, and then withdrew. In some of the villages of Damascus, [people] died.’Footnote 83
Conclusions
Rereading the sources surveyed by Green and Fancy, we have found no evidence in the various texts that would tie the various epidemics/pandemics in Iraq, Syria, and Egypt to plague, supposedly brought unintentionally by the advancing Mongol army that besieged Baghdad in early 1258. We have seen that the Persian-language text attributed to Quṭb al-Dīn Shīrāzī clearly stated that wabāʾ broke out in Baghdad before the Mongols conquered it. Another account, the anonymous (pseudo-Ibn al-Fuwaṭī) Arabic text from Baghdad describes the outbreak of wabāʾ after the taking of the city by the Mongols, but in circumstances that clearly imply that its origins were the terrible conditions of the survivors.
Looking at the many late Ayyubid and Mamluk historians, we have identified four original and independent accounts that we have referred to as Urtexte. By identifying these original versions and then following how later authors incorporated and modified them, we have observed some important points: 1) A group of historians reported the outbreak of an epidemic in Baghdad after the Mongol conquest due to the deterioration of the conditions in the city. They seem to be based on, or derived from Urtext 1, the anonymous [pseudo-Ibn al-Fuwaṭī] chronicle al-Ḥawādith al-jāmiʿa, but they did not connect this disease with the outbreak of another epidemic in Syria and Egypt in the following years. 2) The only accounts to use the term ṭāʿūn are all based on Urtext 2, the long passage by Ibn Wāṣil. We suggested above that Ibn Wāṣil may have identified al-Nāṣir Dāwūd’s affliction with ṭāʿūn (plague) not based on his symptoms, but in attempt to credit al-Nāṣir Dāwūd with the death of a martyr, reminiscent of those who had died of plague in seventh-century ʿAmwās. 3) Two accounts probably based on Urtext 3 (contained in Ibn al-Furāt’s chronicle) suggest that it was the slaughter of Baghdad’s inhabitants that led to the subsequent outbreak in Syria and Egypt. One is the passage by Ibn Wāṣil, explaining that the corruption of airs had caused the epidemic in Egypt and Syria, and then further along alludes, or rather ‘hypothesizes’, that Mongol conquests related to this by referencing a tradition by the Greek physician Galen. This insinuation of a miasmatic connection was more clearly and decisively articulated by a later author, al-Dhahabī. However, let us remember that the unidentified historian cited by Ibn al-Furāt only writes that ‘One of the historians said that the sickness and wabāʾ afflicted the people in Syria after the taking of Baghdad’; no miasmatic connection here. Thus, the majority of Mamluk authors did not connect reports on a disease outbreak in Baghdad with the epidemic in Syria and Egypt. In short, there is simply no reason to conclude that ‘the role of epidemic disease in the Mongol attacks was commonly known among chroniclers’ as Fancy and Green argue.Footnote 84
We have found the use—interchangeable at times—of three terms used to describe the various epidemics in Syria and Egypt in 1258: wabāʾ, maraḍ, and ṭāʿūn. While the last mentioned was also applied to plague, there is nothing concrete in the various passages that we have examined here (mostly also presented in Fancy and Green’s paper) to indicate the epidemics in in Syria and Egypt were plague. Secondly, when symptoms are described in these passages, we encounter fever and coughing – indicating perhaps some type of respiratory-centered ailment, such as influenza; there is no indication in the Arabic sources of the presence of extreme symptoms of either bubonic or pneumonic plague.
In short, when looking closely at the numerous passages, as we have done, and then integrating these with a critical reading of the sources describing epidemic disease in Baghdad during and after the conquest (in section 1 of this article), we have come to the conclusion that there is no cogent historical argument for the presence of plague in Iraq, and then in Egypt and Syria in the year 1258, the year that Baghdad was conquered by the Mongols. The connection between the epidemic in Baghdad and that in Syria and Egypt is questionable, and certainly not as obvious as presented by Fancy and Green.Footnote 85
Fancy and Green have attributed the appearance of detailed and relatively accurate descriptions of plague symptoms in Arabic Ḥadīth and medical commentaries from the second half of the thirteenth century to this supposed thirteenth-century plague event in Iraq, then Syria and Egypt.Footnote 86 This is, of course, an important and interesting matter, but at this point it appears to be far-fetched to connect this shift specifically to a supposed 1258 epidemic in Syria and Egypt, and certainly not to the Mongol conquest of Baghdad. In fact, that the plague symptoms were well known in post-1258 Egypt and Syria, yet none of the medical authors connect them to the 1258 events, only strengthens our argument that there is no relationship between the two phenomena.
In conclusion, our close reading of the texts brought by Fancy and Green (along with other passages) convincingly showed that there is no evidence that connects the epidemics in Baghdad, Syria and Egypt in 1258 to the Black Death. Nor does the recent paleogenetic evidence support a thirteenth-century outbreak of the plague in Central Asia or elsewhere. The Mongol conquest of Baghdad and the Black Death are both medieval catastrophes that accumulated mythical dimensions, and linking the two is very tempting. The textual evidence, however, does not support this connection.
Acknowledgements
The authors wish to thank Philip Slavin for his valuable comments on an earlier version of this article, as well as the anonymous reviewers. The authors are also grateful to Dr. John Yaphe, Dr. Iyas Nasser, Dr. Yosef Witztum and Mr. Raz Amitai-Preiss for their assistance on specific matters (see above in the notes).
Funding
No funding.
Competing interest
No competing interests.
Appendix
From Ibn al-Furāt, Taʾrīkh duwal al-mulūk, MS. Biblioteca Apostolica Vaticana Ar. 726, fol. 203a-b.
As is to be expected from manuscripts from the period, elements of the spoken language frequently have entered the text, a register that is referred to as ‘Middle Arabic’. We have left these forms as they appear, but to note the deviation from the standard Literary Arabic (what is called by many Classical Arabic), we have added in square brackets كذا = sic. We again take the opportunity to thank our colleague from the Department of Arabic Language and Literature at the Hebrew University, Dr Iyas Nasser, for checking our reading and translation, and helping make sense of one particularly recalcitrant sentence.
وقال بعض أهل التأريخ لحق الناس في الشام مرض ووبا (كذا) عقيب أخذ بغداد حتى كان الإنسان لا يصل إلى العطارين إلا بالجهد والمشقة واستغنى العطارين (كذا) ونفدت الأشربة والعقاقير وعزّ الأطبا (كذا) وأهل الحجامة وتعطلت المعايش (كذا) وكسدت البضايع (كذا) وكان الناس على العطارين في دمشق كما يكونوا (كذا) في القحط على الخبازين ودام ذلك الحال أياما وأما أهل حلب فكان الموت فيهم أكثر وورد الخبرFootnote 87 أنه خرج منها في يوم واحد ألف ومائتين (كذا) جنازة قال ومات في دمشق على الطريق خلق كثير وفي المرستانات وعزّ القُرّا (كذا = القُرّاء) والمغسلون وبقي أكثر أهل البلد بالحُمَّى والسُّعال // مَرض واحد وتوفي من الأعيان جماعة وأما من لا يُؤبَهFootnote 88 له فخلق كثير وأبيع البطيخ الأخضر في دمشق في هذه المدة كل حزّة يعني شقفة بدرهم وبلغ كل رطل تمر هندي بستين درهم (كذا) وأما الفراريج (كذا) فكانت لا توجد وإن وُجدت يكون الفروج في قدر العصفور بدرهم ثم صار إلى ثلاثة وأربعة دراهم وأما في حلب فبلغ ثمن كلفروج عشرة دراهم وبالجملة يغني عن التفصيل والله أعلم