1 Introduction

In 1253 CE, the Great Mongol Khan, Möngke, appointed his brother, Hülegü, as the head of a massive army, consisting of a fifth of all imperial forces. In 1257 CE, however, after dealing with the Nizari Isma ‘ilis of northern Persia, Hülegü turned against the ‘Abbasid caliph and besieged Baghdad in February 1258 CE (Āshtiyānī 2010: 108).To intuitively visualize just how much wealth the Ilkhanid court inherited after the fall of Baghdad, the richest city of all time, Rashīd al-Dīn reports that endless supplies of gold were a consistent and readily available source of wealth for many years afterward and these treasures were redistributed several times among both the nobility and wider society during the subsequent reigns of Abaqa Khan (1234–1282 CE), Ahmed Tekuder (1246–1284 CE), Arghun Khan (1258–1291 CE), and Ghazan Khan (1271–1304 CE) (Rashīd al-Dīn 1994: 1288). After this event, the Ilkhanid capital of Tabrīz became one of the wealthiest cities on the planet as well as the center of architectural technology and the arts in the thirteenth century (Ibid: 1291).

When the Mongol ideology of world domination was halted after Hülegü’s withdrawal and General Kitbugha’s defeat by Sultan Qutuz at the battle of ‘Ayn Jalut in September 1260 CE (Stewart 2019: 321), Hülegü returned to Northwest Iran and ruled the Il-Khanate for the remainder of his life (Fig. 1). His empire covered the southwestern portion of the Mongol Empire and included the Iranian capital cities of Maraghe, Tabrīz, and Sultaniyya (Boyle 1968: 88). However, his territory was officially a subsidiary khanate of the Mongol empire, the capital of which was established at Dadu (modern-day Beijing) by Möngke’s successor Kublai Khan in the 1260s (Steinhardt 2018: 92).

Fig. 1
figure 1

Hülegü Khan with his Nestorian Christian wife Dokuz Khatun, from Jāmiʿ al-tawārīkh (granger.com)

Full size image

Rashīd al-Dīn provides a window back in time to when treasures and precious objects were transferred to Northwest Iran after Baghdad was burnt to the ground, thus marking the end of the Abbasid Caliphate and the Islamic Golden AgeFootnote 1: […Hülegü Khan conveyed all of the treasures to Azerbaijan (Northwest Iran) and ordered Malik Maj Al-Din Tabrizi to build an impregnable stronghold on a mountain called Tāle which is located on the shore of Lake Urmia and Salmas.Footnote 2 This mountainous stronghold was a place to shelter his war booty taken from Baghdad since the treasury in Tabrīz was not spacious enough for this purpose. All of the gold items were melted down and stored inside the fortress]. (Rashīd al-Dīn 1994:1022). Rashīd al-Dīn further suggests that the burial vault (Sanduq) of Hülegü was lowered into his grave amidst a huge qörugh (great taboo) on a hilltop on Shahi Island in Lake Urmia (ibid). According to Rashīd al-Dīn, this qörugh is located close to Dehkhareqn and the Chaqato River (ibid 1048). In his Shāhnāme-ye Changizi, Shāms al-Din Kāshani (fourteenth century) writes the following couplet concerning the credibility of Rashīd al-Dīn’s notes about the mysterious burial place of Hülegü:

  • نهادند اساسی به دریا کنار  به نزدیک سلماس و ارمن دیار

They established the foundation of Hülegü’s tomb nearby Lake Urmia close to Salmas and Urmia [city] (Shams al-Din 1998: 107).

The Persian historian and Ilkhan official, Wassāf (1299–1323 CE), narrates a similar report, adding the details that: after Hülegü’s destruction of Baghdad (1258 CE), all significant fortunes, works of art, etc. were carried off by the Mongol troop to the Ilkhanid capitals of Tabrīz and Maraghe (Wassāf 1966: 29). Wassāf reports that Hülegü’s intention was to invest part of this wealth in constructing a huge observatory under the directorship of Naṣīr al-Dīn al-Ṭūsī (a Persian scientist and astronomer). He wrote: […Hülegü passed away in 1265 CE when the construction process of the observatory was not completed yet. As was the Mongol burial custom, they buried him (Hülegü) in conformity after preparing an underground chamber (dakhme). Then, at the time of the interment, vast quantities of gold and jewels were poured into it and several beautiful young women dressed in sumptuous clothes were made his bed-fellows to protect the Khan from the horror of the darkness and narrowness of the grave] (Ibid: 30). The search for a burial site, the discovery of which might resolve many of the questions regarding Hülegü’s treasury and burial place, has thus far been unsuccessful. In this paper, the author shall reflect on this lacuna on the basis of information provided by chronicles and archaeological remains.Footnote 3

2 Literature

It was not until 1955 that Donald N. Wilber made an initial attempt to pinpoint Hülegü Khan’s royal tomb in Northwest Iran (Wilber 1969: 75). Based upon historical sources and chronicles, he was the first person to conclude that the summit of the great rock rising 1000 ft (c.300 m) above the shore of Jazīreh-ye-Shāhī could be the burial place of Hülegü, the founder of the Ilkhanates. However, Wilber’s excursion failed to provide detailed information regarding Hülegü’s enigmatic tomb and treasury, and he did not succeed in studying Hülegü’s fortress nor the site and archaeological remains to be found there. His publication on Jazīreh-ye-Shāhī encouraged other scholars to evaluate different theories and address this archaeological void, albeit without further physical investigations (Ahutja 1967; Boyle 1974). The account of Jazīreh-ye-Shāhī has left fragmentary evidence in archaeological and historical archives. The only systematic archaeological investigation on this island since the Islamic Revolution in Iran (1979) has attributed the first architectural phase of Hülegü’s so-called fortress to a Sassanid settlement (224–651 EC). Since then, no archaeologist has revisited the site.

In October 2006, the author tirelessly sought out any scholars able to assist him, in searching for evidence and obtaining all the information related to Hülegü Khan’s tomb. Four years later, in 2010, during a one-month investigation, the author interviewed several elderly people on the island, who stated that Kāboodan, located in the southern part of Lake Urmia, would be a promising site to visit surface finds. This survey uncovered 13th-century glazed potsherds similar to those found at other Mongol settlements in Northwest Iran, such as the summer capital of Ojan (Velayati et al. 2017). Moreover, atop the ground surface, the author noticed some lines of stone masonry in the lime mortar that indicated a relatively large former structure. The first course of the wall measured 10.21m × 1.12m, with only 30 cm of its height visible from the ground surface. However, this expedition did not conclude with any scientific publication.

3 Methodology

The methodology in this research adopts a multidisciplinary approach. The first section of this paper will examine contemporary historical narratives that supposedly pinpoint the location of Hülegü Khan’s tomb and treasury. Subsequently, it will focus on archaeological remains through a survey of the site to assess the credibility of written sources. The input to study in this research is the description of the architectural remains to produce basic data on their identity that can serve as a future point of reference for coming research on Mongol studies in the region and beyond. The study does not only provide a broad view of the historical setting that eventually motivated the construction of the fortifications in Jazīreh-ye-Shāhī, but also a catalog of very basic information, such as materials, construction techniques, architectural forms, and features. To this end, after the collection of relevant data from fieldwork and surveys the attempt was to correlate this information to the time of construction, initiators, and their origin as well as the historical setting.

4 Lake Urmia and the Legendary of Hülegü Khan

Lake Urmia (Daryācheh-ye Orūmīyeh) lies at the bottom of the large central depression of the Azerbaijan region in Northwestern Iran, at an elevation of 4,183 feet (1,275 m) above sea level. The basin is surrounded by mountains to the east (Kūh-ī Sahand) and north (Mishoodāghi Mount), plateaus to the south, and volcanic cones to the north (UNEP 2012).Footnote 4 The main rivers of the basin are located in its southern half, the most important of which are Ājīcahāy, Zarīnerud, Sīmīnerud (Tatavi), Mahabadrud, Bāranduzchāyi, Zolāchāy, and Nāzlichāy. Shahi Island (Jazīreh-ye-Shāhī), which translates literally to Royal Island, is the largest island in Lake Urmia (Fig. 2). This island is the only inhabited one in Lake Urmia, and is comprised of seven villages including Burāchālu, Qībchagh, Teymurlu, Aq Qonbad, Qāmichī, Sārāy, and Bāhrāmabad. All of these villages are located along the shores of the island. As a result of the declining water level, Jazīreh-ye-Shāhī is now connected to the mainland on its eastern side and thus forms a peninsula. In the southern portion of Lake Urmia, there is a cluster of about nine tiny islands encompassing Arīzu, Kāboodan, Chīyerlī, Dānālu, Gīzīl Konqur, Espīr, Kāmar Dāq, Azin, and Ashk. Among these islands, Kāboodan is the central island and differs greatly from its smaller counterparts in that freshwater is available there all year round. In Al-Tanbīh wa l-ishrāf, Alī Ibn Ḥusayn al-Masʿūdī (893- 957) mentions that Lake Urmia’s name comes from a village called Kāboodan, on the island at its center (al-Masʿūdī 1381: 70). In his Mu'jam ul-Buldān, Yāqūt al-Ḥamawī (1179–1229 CE) reports: [In Lake Urmia, there is an island which is called Kāboodan. This island includes four villages in which the life of the inhabitants is based on seafaring and agriculture. On the mountain of this island, there is a famous fortress whose inhabitants were rebels against the rulers of Azerbaijan for most of the time] (Yāqūt al-Ḥamawī 1995: 351). Ḥamd-Allāh Mūstūfī (d. 1344) and Ḥāfeẓ-e Abru (d. 1430) attest to Hülegü Khan’s motivation to rebuild and fortify this fortress to hide large amounts of gold ingots, and bags of outlandish coins after the ransacking of Baghdad (Mūstūfī 1913:56; Ḥāfeẓ-e Abru 1992: 104). In Taqwim al-Buldan (A Sketch of the Countries), Abū al-Fidā (1273–1331 CE), a Mamluk-era geographer, refers to “Bahirat al-Tāle” as follows: [Bahirat al-Tāle is the same Lake Urmia that is located between Maragheh and Salmas. In the middle of it, there is an island with an impregnable fortress called Qaleye Tāle. Hülegü Khan put his treasury in it, … and it is said that he is buried inside this fortress] (Abū al-Fidā 1996: 46). Shams al-Dīn Muḥammad ibn Abī Ṭālib Damīshqī (1256–1327 CE) in Nukhbat al-dahr fī ʿajāʾib al-barr wa-al-baḥr tells a similar story to that of Abū al-Fidā by emphasizing that the enigmatic fortress of Hülegü is located on top of a mountain called Tāle (Shams al-Dīn 1978: 121). In addition to Rashīd al-Dīn, Ḥamd-Allāh Mūstūfī, and Ḥāfeẓ-e Abru concretely report that Hülegü himself is buried somewhere inside the fortress and that according to a Mongol custom, the slaves involved in constructing the mortuary were executed to keep the location of his tomb secure (Mūstūfī 1913:56; Ḥāfeẓ-e Abru 1992: 108). The Mountain Tāle emerges in the post- Hülegü period as an ideal burial place for the Mongol elites. Rashīd al-Dīn narrates Hülegü’s grandson Abaqa Khan’s (1234–1282) desire to be buried in the same place as his father in Jazīreh-ye-Shāhī. He wrote: [On Sunday, the 16th of Muharram in the year 681, Abaqa Khan passed away and his coffin was taken to Shahu Tāle and buried near the great Ilkhan (Hülegü)] (Rashīd al-Dīn 1997: 1118). Regarding Ahmed Tekuder’s coronation, Rashīd al-Dīn further mentioned that: [He was elected on 6 May 1282 and enthroned on 21 June 1282 at Aladağ, east of Lake Van. After the feast and happiness, he [Ahmed Tekuder] ordered the treasurers to prepare part of the confiscated fortunes from Baghdad that were kept in Shahu Tāle. Then, he distributed it among the noblemen, princes, governments, and the poor] (Rashīd al-Dīn 1997: 1128). This research explores a question entirely ignored by archaeology to date; the probable location of the tomb and treasury of the great conqueror, Hülegü.

Fig. 2
figure 2

Lake Urmia basin before climate change; Landsat data from USGS, 1998. A: Jazīreh-ye-Shāhī, B: Arīzu, C: Kāboodan, D: Espīr, E: Ashk, F: Chīyerlī, G: Dānālu, H: Kāmar Dāq, and I: Azin

Full size image

5 Archaeological remains in Jazīreh-ye-Shāhī

Although there are many mountainous and coastal paths by which Lake Urmia may be reached, the most convenient and navigable route is via Ilkhchī, which can easily be reached from Tabrīz, the former Mongol Capital. As aforementioned, Jazīreh-ye-Shāhī is the biggest geological feature surfacing from Lake Urmia and has been cut off from neighboring localities by land salinization in recent decades. Historically, the distance between Jazīreh-ye-Shāhī and the surrounding high mountain ranges was covered by the green valley of the Zarīnerud basin, however, increasing soil salt concentration has made inter-village social life on this island impossible. From the names given by Ḥamd-Allāh Mūstūfī in Nuzhat al-gulub (written in 740/1340 CE), we know that the Mongolic name of Zarīnerud, designated to Lake Urmia, was Jaghātū (Mūstūfī 1983, 76).Footnote 5 According to Vladimir Minorsky, the name Jaghātū is most likely derived from the Mongolian word “jaqa”, meaning "border" or "bank", added to with the possessive suffix –tu (Minorsky 1957).Footnote 6 In his Ḥabīb al-siyar fī akhbār afrād al-Bashar, Ghiyāth al-Dīn Khwāndamīr (1475–1535 CE) highlighted the rivalry between Arpa Khan and Khwaja Ghiyath al-Din Muhammad, writing: […near Jaghātū, the two armies collided, they used swords and daggers, and Khwaja was defeated by Arpa Khan and escaped. To commemorate this triumph, Hossein Abād village, which is approximately 15 km away from Shāhīn Dej city, was chosen by Hülegü Khan as the summer capital, and Zarīnerud was named in honor of Genghis Khan's son Jagtāi (Jaghto)] (Ghiyāth al-Dīn 2001: 97).

In addition to Jaghātū, a number of other very interesting names are noticeable in local toponymy. The majority of villages in Jazīreh-ye-Shāhī and the Lake Urmia basin bear Mongol names and seem to reflect the period of Mongol domination of this huge area encompassing Northwest Iran. In Qībchagh, for example, which is a village situated to the north of the island, the toponymy bears witness to definite Mongol infiltrations (Hildinger 1997: 112).Footnote 7 Similarly, the village of Dash Kasan, some 28 km to the east stands as a reminder of the original Mongol localities. From archaeological research, we know that the same name, Dash Kasan, was given to imperial camping sites in the southeast of Sultaniyya (Moradi 2022).

According to local testimonies, it was almost impossible to access Shahi Island from the opposite shore twenty years ago. In fact, the island did probably not need special fortifications in this part because it was already well-protected by water. The remains of a huge fortress are visible on the top of a rocky hill in Jazīreh-ye-Shāhī, which can be accessed from Aq Qonbad village. The ascent from Aq Qonbad to the fortress takes about three hours. Dry riverbeds to the north indicate that the fortress’s northern approach was once surrounded by a ring of freshwater meres. Too shallow to be of any defensive use, these meres would have made the fortress look as though it were floating on an island, surrounded by water. Approaching the fortress from the foothill, one’s eyes are inevitably drawn to a massive fortification made of dry joint stone technique which stretches east–west (Fig. 3). Extremely thick masonry walls would arguably have been expensive and arduous to build around the fortress, therefore the dry joint construction technique was likely adopted to privilege time-saving and manpower as well as to conserve resources that could be better spent on other security factors. This wall is made up of irregular courses of untrimmed stone blocks and is called “qapi” [the gate] among villagers. Considering the width of the stone wall (ca. 2.80m), it is reasonable to conclude that it provided extra security for the fortress. The area inside the wall shows no remarkable masonry evidence. It looks to have been left empty and unoccupied, suggesting that the wall was to give an appearance of strength and imposition rather than having a practical utility. The whole path from qapi to the highest level of the fortress [qal-e ye Hülegü Khan] is marked by stone signs. Not tracking and following these indicators, put in place by compassionate climbers, may result in getting lost while hiking. This may not seem too dramatic at first, but when said lost hiker ends up wandering into the deep precipices all around the fortress, the situation can be hazardous.

Fig. 3
figure 3

Current remains of Qal-e ye Hülegü Khan in Jazīreh-ye-Shāhī, Lake Urmia. See from the south (Photo by Amin Moradi the author 2023)

Full size image

Unlike the entirely inaccessible northern and eastern sides comprised of rising perpendicular walls of solid rock, the slope on the eastern face is much more gradual. Here, most of the cliff face has been badly affected by erosion and is quite impassable without elaborate mountaineering equipment. A person wishing to enter Hülegü Khan’s fortress would instead first need to climb up a short flight of stone stairs constructed within a triangular fissure in the rock on the east side of the hill.

From the landing at the top, any visitor would proceed west, ascending a via Ferrata that is fixed to the rock and gives access to the upper part of the cliff. However, climbers should not be advised to secure themselves to it as a means to limit any falls. It is easy to envision that when enemies arrived at the site, soldiers and inhabitants would retreat inside the fortress, finding safety behind its walls and on top of the rock cliffs. After climbing the via Ferrata, one finds oneself standing in front of several rock-cut cells that have been uniquely carved into the sedimentary bedrock at various heights. These cells have an almost-uniform width of about 2.60 m but vary in length between roughly 3 to 6 m (Fig. 4). Remains of human skeletons and pottery shards are visible among the debris inside the cells. These cells have been ransacked by treasure hunters and repeatedly excavated by locals in the hope of finding the hoard of relics buried inside them, meaning that tons of soil have been shoveled in and out of them multiple times. Apparently, these antiquity hunters have handled enough soil and rock in this fortress so as to relocate an actual part of the mountain. In other words, if these lunatics that have tried to plunder these cells had hired out with road contractors, they could have earned lots of money instead of this terrible waste of energy. A vast part of the rock face has been flattened around these cells and the cleavage of picks is discernable on the resulting vertical surface. It is likely that the cliff was intentionally flattened with a view to accommodating more cells in the future. The other hypothesis would be that the cliff face was flattened so as to cut off access to the heart of the fortress and make it difficult for visitors and external threats to enter.

Fig. 4
figure 4

Rock-cut cells on the southern side of Qal-e ye Hülegü Khan in Jazīreh-ye-Shāhī (Photo by Amin Moradi 2023)

Full size image

There are two cylindrical cisterns at the level of the cells. While the smaller one measures ca. 13m in height and ca. 8.60m in diameter, the bigger one is almost double its size, measuring ca. 27.44m in diameter and 4.12m in height (Fig. 5). A hard layer of hydraulic plaster mixed with lime, ashes, and grit applied on the walls of both cisterns led the author to conclude that they were used as water storage containers at one time. Numerous rock-cut grooves around the cisterns were designed to convey the rainwater to said containers. As rains are usually heavy in this area, and can occasionally be torrential, the amount of water that these water tanks could catch could be several thousand cubic meters after each heavy rain. Even though history gives no account of who actually planned and supervised the construction of this site, cutting these channels was certainly not beyond the skills of engineers in the region. Several rock-cut hooks close to the cisterns appear to have been carved to tie animals’ leashes. On the eastern side of these cisterns, there is a room with an irregular plan (measuring 1.56 × 2.20m) with niches on all sides to hold oil lamps.

Fig. 5
figure 5

Panoramic view of the larger cistern (up) and sections of cisterns in in Qal-e ye Hülegü Khan (Photo and drawing by Amin Moradi 2023)

Full size image

Shards, both glazed and plain, can be found all around the site. Amongst the plain shards are parts of large vessels (Khumra) used for storing water or grain. Shards of different shades of cobalt blue are thought to be parts of smaller bowls and plates. These shards have a fairly uniform, fine-textured turquoise and blue composition, are mostly light in color and well oxidized in the firing, apparently thus not originating from local workshops (Fig. 6).

Fig. 6
figure 6

A selection of potshards from field surveys in Qal-e ye Hülegü Khan (Photo by Amin Moradi 2023)

Full size image

Three different construction material techniques were identified during the inventory of the fortress conducted in the summer of 2022. The oldest phase represents the stone masonry made up of natural volcanic blocks and lime-based mortar. Types of mortar used on, and textures and colors applied to structures within the fortress vary depending on the structural type and construction phase. The dominant type of mortar applied in the fortress can be identified by a gray tint, and is made up of coarse-grained sand and has a high lime content. The upper phase includes dry-laid stone masonry with weathered surfaces for fortification. Several small rocks were pieced together to fill the space between these stone blocks. Finally, a limited number of bricks and mud bricks, found scattered between the rubble, indicate the third variation for the material.

The geological outlook of the scenery of Lake Urmia presents other interesting features. The Kāzīm Dashī itself is an immense piece of a sedimentary bed, lifted up into a near-perpendicular position (Fig. 7). Local inhabitants have unverified stories about this rock mass which have been handed down from earlier times. According to a very long-standing oral tradition, the queen of this fortress was at one point kidnapped by an evil army and imprisoned in Kāzīm Dashī. The day before this event, the queen had thrown all the fortress’s gold plates and jewels into a very deep well in the grounds and pronounced a curse against anyone who should try to find it ere she returned. It is also said that these curses or bewitchments often damaged farmers' crops or caused outbreaks of disease. “Hossein Jan” the legendary hero of the region, defeats the army of evils, slays the chieftain of the demons, and returns the captured queen to the fortress.

Fig. 7
figure 7

General view of Kāzīm Dashī. See from the south (Digital collection of Cultural Heritage and Tourism Organization; Iran)

Full size image

During the First World War (28 July 1914 – 11 November 1918), Kāzīm Dashī was used as a shelter for the surrounding villages. Indeed, it is named after "Kāzīm Khan", the commander of the village guards, who himself was from Ghoshchi, the nearby village. Kāzīm Dashī was also known as Qiīkhlār. According to the elders of the village, in ancient times, forty people escaped from the cruel ruler of the region and took refuge in this place for a while. At the highest point of this rock mass, a well has been dug in the heart of the rock which was used to store snow and rainwater. Locals believe that this well was once used as a prison.

6 Discussion

As aforementioned, Hülegü’s fortress is known among locals as a fortification structure, and it is a convincing hypothesis that the remains can be linked to defensive architecture. The first key question concerning this fortress is: What kind of fortification was it? Chronicles give a clear answer that it was chosen as a secure place to house the Mongol royal treasures (Abū al-Fidā 1996: 46; Mūstūfī, 1913: 56; Ḥāfeẓ-e Abru 1992: 104; Wassāf 1966: 28), rather than being conceived as a governmental center. Hence, it was not expected to fulfill multiple functions within its walls, such as hosting: a magnificent great hall with ample space to hold the Mongol court and serve state banquets (to impress as well as intimidate its visitors); huge barns to store grain; stables and shelters for animals; lodging for workers; and all manner of workshops. Paradoxically, the idea of saving water in huge cisterns, which is a common feature of defensive fortresses in mountainous Northwest Iran, cannot actually be considered for practical purposes in a fortress controlled by the Mongols, the most powerful military army of the time. In other words, in Hülegü Khan’s fortress, the desire to ensure a safe place for tons of gold ingots surpassed the need for spacious, comfortable living arrangements inside the walls. However, it is likely that the fortress was associated with tax collection and might possibly have been the residence of an arrogant garrison, which made the citadel and its residents the frequent focus of attention for rebellions and independent citizens who tried to either ransack the treasury or staff them with their own men.

The second major question then arises: Where was the treasure kept? To answer, Rashīd al-Dīn suggests that Hülegü’s treasury had an important role in finance management and budget planning during and after his reign (Rashīd al-Dīn 1994: 1118). Hence, it must have been a repository to which governments had relatively easy access to the spoils of. Although no dated historical evidence for different construction phases was found, the architectural history and development of the rampart confirm that the current fortification was built upon an older construction. It could be suggested that Hülegü ordered the reconstruction of the fortress (Mūstūfī 1913: 57; Ḥāfeẓ-e Abru 1992: 104) so as to use the huge cisterns by changing their functionality from water tanks to a treasury. The intentional attempts to block the water channel with rubble, thereby disrupting the rainwater harvesting system of these cisterns, is support of this idea. This fortress did not have any obvious practical military purpose at all, because it was the safest and most remote stronghold in the vicinity of the Mongol capital of Tabrīz.Footnote 8 Hence, there was no need for these cisterns to supply drinking water to the fortress for sustaining soldiers or inhabitants to guard the treasury and tombs. Considering the remarkable dimensions of these cisterns, they have the capacity to store hundreds of tons of gold ingots and this explains why the builders of the site had an obvious interest in choosing this location. These cisterns on the peaks of inaccessible cliffs and reliably isolated from the world were the ideal places for Hülegü to protect his treasure from uninvited guests. A vertical set of holes at the entrance to the cisterns allowed for them to be temporarily sealed by installing wooden sticks and leather sheets. Furthermore, a guard dog tied-up close to the cisterns using the rock-cut hooks could safeguard the treasury against unwanted or unexpected human entry while warning wardens. Rashīd al-Dīn reports that when Ghazan Khan sat on the throne (1295 CE), this treasury was empty because the treasurers were corrupt and its contents was gradually stolen from the fortress (Rashīd al-Dīn 1994: 1349).

To trace Hülegü and other Mongol elites’ tombs back in the history of this fortress, it would be helpful to refer to the historical records as well as Mongol burial customs. It is known that during the Ilkhanid era, no mounds in the form of a grave, raised platforms, or tombs were made which would echo Muslim traditions (Ibid: 1335). This explains the conspicuous absence of any such commemorative plaque or platform and also the untraceable burial places of Mongol Khans and chiefs. If we consider the deaths of the Mongol elites, we must first look to Genghis Khan’s death who died in August 1227 CE (Haenisch 1933: 503; Wright 1997: 425; Drobyshev 2006: 65; and Bold 2000: 95). According to Mongolian culture, kings’ corpses retained their divine power after death, hence their remains were buried in unmarked, highly protected and inaccessible places, particularly on mountains where they were thought to be resting closer to their final destination: Heaven. (Drobyshev 2006). According to legend, Genghis Khan lies buried somewhere beneath the steppe of northeastern Mongolia, entombed in a location so secretive that anyone who mistakenly encountered his funeral procession was executed on the spot (Weatherford 2005: 66). Marco Polo wrote: [It has been an invariable custom, that all the grand khans, and chiefs of the race of Genghis Khan, should be carried for interment to a certain lofty mountain named Altai, and in whatever place they may happen to die, although it should be at the distance of a hundred days' journey, they are nevertheless conveyed thither].

In a frequently recounted tale, Marco Polo documented that the 2000 slaves who attended Genghis Khan’s funeral were killed by the soldiers sent to guard them and that these soldiers were in turn killed by another group of soldiers who were tasked to dispatch anyone and anything that crossed their path, in order to conceal where he was buried (Polo 1960: 256). Finally, the legend states that when they reached their destination they themselves committed suicide (Levy 2007: 179). Historic accounts agree in stating that members of the Mongol Royal family, throughout the Empire, were buried in inaccessible places, for the most part, high up in the mountains (Polo 1960: 257; Carpini 1900; 112). Royal cemeteries were forbidden ground or taboo and were guarded by detachments of soldiers, though since the actual site of the grave was not indicated in any way, even members of the guard were ignorant of where the grave actually was (Rashīd al-Dīn 1994: 1349). These forbidden areas are everywhere designated by the same word, “qorugh”, which also occurs in the forms of “qorig” and “qhorug” (ibid).Footnote 9 In the Mongol period, the word qoruqh is used for anything sacred or taboo, private lands, most frequently Royal lands, the ruler’s place, and, in particular, his harem (Schéfer 1892: 27; Barthold 1970). From Rashīd al-Dīn’s accounts, we can certainly conclude that there were qorughs in Mongolia even before the time of Chingis Khan. He describes the sack of a part of the qorugh of the Kerait Wang Khan by his enemies (Rashīd al-Dīn 1994: 988). Hülegü’s qorugh, in Rashīd al-Dīn’s description, is reminiscent of the “yeke qorugh”, the “Great Forbidden Sanctuary” of Genghis Khan that Carpini reported in his travelogues. Carpini commented on Genghis Khan’s royal burial ground as a place where emperors, chiefs, and all nobles are interred and irrespective of where they died, were brought hereto if this could fittingly be done (Carpini 1900: 112). He also mentioned that a great deal of gold and silver was buried with them (ibid). Rashīd al-Dīn similarly speaks of the troops who, in his day, kept guard over Genghis Khan’s qorugh (Rashīd al-Dīn 1994: 1255). John Carpini distinguishes between the burial methods of elite and non-elite men. In the case of the former, after filling in the pit, it was covered over with grass so as to be undiscoverable (Carpini 1900: 122). Atâ-Malek Juwaynī (1226–1283 CE), the Persian historian, writes that when the tomb is covered up, horses are driven over it, in such a manner that not a trace of it remains (Juwaynī 1975: 229).Footnote 10 We are also explicitly informed of the immolation of young women in the case of Genghis Khan and his grandson Hülegü (Wassāf 1966: 28). It is believed that not only the Khans’ grave but also their encampments came under the classification qorugh, and these latter were maintained as they were during his lifetime (Barthold 1970).

In relation to the Mongol funeral customs in Iran, our knowledge is limited to four folios from the Great Mongol Shahnama (Book of Kings), probably dating back to the 1330s and illustrated in Tabrīz.Footnote 11 These folios indicate some common features of the royal funerary in Ilkhanid Iran, such as the deploring of the dead by face-cutting and body-whipping, bringing the dead to the cemetery on a carriage or sled, sumptuous burials, a funeral repast, and the sacrifice of animals, especially horses, in honor of the dead. In some cases, the funeral ceremony was also accompanied by music, the beating of drums, ritual dancing, and song.Footnote 12 Whether Hülegü’s burial followed or deviated from these norms we do not know. There is no immediate iconography in the supposition that the abovementioned scenes are, in fact, part of a royal funeral procession. However, this illustration indicates the time when these traditions were alive and strong and it is probable that some commonalities could be found between this pictorial evidence and Hülegü’s funeral customs.

Considering the remarkable number of human skeletons inside the rock-cut cells, these spaces appear to have become a focus of burial ritual activity. Deposits of potsherds are thought to be the remains of offerings to those buried in these tombs. The category of burial gifts is well known from descriptions of Ilkhanid funerals where gifts from the mourners were carried in the funeral procession, with some of them being deposited on the pyre or in the grave. In Fārs-nāma-ye nāṣerī, Mīrzā Ḥasan Fasāʾī (1821–1898) on the death of Abīsh Khatun wrote: [Abīsh Khatun was the daughter-in-law of Hülegü Khan. She became ill and passed away in Tabrīz. She was buried in Tabrīz in the cemetery of the Mongol Sultans according to Mongol traditions. Although she was a Muslim, a believer, and a pure woman… gold and silver vessels filled with basil wine were placed in her grave (Mīrzā Ḥasan 2003: 277).

The exact parallel of these burial cells is visible in the neighboring Ilkhanid site of Rāsad Khāna.Footnote 13 In both cases, the alignment of these cells represents a north–south axis that connected with the Mongolian cult in which particular attention is given to the sunrise with the head facing east (Halbertsma 2005: 161). Additional evidence from the prehistoric Mongol graves in Alag Tolgoi and Khanan cemeteries in northern Mongolia, all of which represent a field of Mongolian burials of the imperial period (Nelson et al. 2011:216), indicates the fact that the north–south orientation was the standard norm of inhumation from the Early Iron Ages (ca. 800–400 BC) in Mongolia. Therefore, it is reasonable to conclude that people who had designed and cut these burial cells or the occupants followed the Mongolian tradition. The idea that the corpses of Hülegü and other elites were buried in these small chambers is not one that can be put forward. Historic resources emphasize that Mongol tombs had no mounds and were trodden over by horses so as to imitate ordinary ground (Polo 1960: 258). It should consequently be impossible to find these places afterward. Moreover, it is suggested that a royal Mongol tomb in Northwest Iran might be an underground chamber hewn out of the bedrock in which a long shaft, sunk into the side of a hill, gives access to the burial chamber (Moradi 2022). Additionally, to keep the element of secrecy, Hülegü’s funeral was the only Ilkhanid funeral to feature human sacrifice. Hence, the relatively small chambers that every visitor can see do not fit the funeral standards of the Mongol elites. These cells could belong to aristocrats, guards, warriors, servants, and persons who were not of noble birth. It is worth mentioning that the scattered remains of potsherds inside these cells stylistically indicate that they belong to the Ilkhanid period.

Recent findings concerning burials during the medieval period in western Mongolia, shortly before the emergence of Genghis Khan, confirm that both inhumation and cremation were practiced (Crubézy et al. 2006). These funeral alternatives can be studied in the contents of these cells, but fragments of potsherds found inside the chambers cannot take us further in identifying for whom any burial was intended. DNA analysis of the remains in these tombs could reveal any possible relationships between the tomb’s occupants and further tests could also reveal whether the bodies were mummified.

The word Rashīd al-Dīn and Wassāf use for Hülegü’s tomb, “dakhme” (Rashīd al-Dīn 1994: 1128; Wassāf 1966: 29), is commonly applied to a kind of rock-cut architecture which designates a space hewn in rocks, mountains, or hills in which the deceased is buried. Numerous dakhmes from different periods have been found in Western and Northwest Iran, which cannot be associated with the burial of a specific person (Huff 1989). The common feature of these burial spaces is that they have all been built on vertical surfaces after the removal of a huge amount of stone on the edge of the cliff.

The only discovered Mongol dakhma which has been plundered during the Qajar Dynasty in Iran (1789–1925) belongs to Arghun Khan (1258–1291 CE). The fascinating narration of Fasāʾī about this dakhma confirms beyond any doubt that this form of architecture is a rock-cut cell accommodating the Khan and his precious grave goods. Fasāʾī wrote: [Nearby the QidārFootnote 14 village there is a spring known as Arghun Bulaghi, and on the south side of that village and spring, there is a mountain. One day Karbali Fathali, a local shepherd, was herding his sheep on the slopes of that mountain, and he was lying down to relieve his fatigue. Suddenly, he saw a mouse that took out some white pearls from a hole and put them in the sun. The shepherd took those pearls and gave some of them to his fiancée, who was the daughter of Fathali Wali Nemet, and kept some for himself. When Fathali saw these pearls, he searched for the shepherd and went with him to the mouse hole. When they dug the rocky cliff, a dakhma was found, and they found some gold and jewels. That night, Fathali took them home and buried them. A young man called Majid who was from Garrus was informed about this incident. Since he had a book that recorded the events of the Mongol Sultans, especially those whose capital was in Khamsa province, they had written about Arghun Khan's dakhma at the foot of the same mountain. He (Majid), went to that place and carefully dug the plundered tomb and found other precious objects, and fled to the Shirvanat because of the fear of the rulers of that area. The shepherd man told the story to Nawab Abdullah Mirza, the Governor of Khamsa. After that, Nawab Muazi Aliyeh, with the help of several people, confiscated the treasure from Karbalai, Fathali. Among them was an inlaid gold vest that was made strangely and has tiny beads of pearl, emerald, and turquoise on it (Mīrzā Ḥasan 2003: 707).

At a glance, there appears to be no obvious physical indication of the location of a royal tomb inside the fortress. According to Mongol traditions, being buried close to the great Khans would have been a priority for any lay or noble person (Song 1976: 566). These individual graves could perhaps be attributed to the idea of improving the dignity of the deceased by having a tomb in the vicinity of Hülegü’s tomb. The original depth of the rock-cut cliff to the west of the burial chambers is unknown. Its bottom is now filled with limestone chips and antiquity hunters have never successfully reached the very base point of the wall. In search for a potential place to create a dakhme, the rocky cliff to the west of burial chambers harks back to the elevation of a cliff burial from the eastern Han (206 BC – 220 CE) and Jin (265–420) dynasties in East Asia (Steinhardt 2017: 126) (Fig. 8). The cliff burials in these periods are tombs positioned next to each other horizontally, and forming several layers vertically along the upper reaches of the hills and mountainsides (Luo 2000, 47). Regarding the size and layout of the tombs, they vary in scale, with some containing a single shallow burial chamber just a few meters in length, while others extend over tens of meters into the mountainside with multiple chambers to accommodate numerous coffins (Wu 2000, 79). The other important fact to be considered is that the rock-wall is facing south, an orientation which is so characteristic of Ilkhanid sites. From the architectural evidence we know that the entrance of imperial structures in Ilkhanid Iran has typically located in the south (Moradi 2022).Footnote 15 Hence, this possibility cannot be excluded that the entrance to an Ilkhanid dakhme might be located at the lower part of the rock-wall (Fig. 8). This hypothesis will be corroborated with further examination of the site and archaeological trial trenching in front of the rock-wall. It is also believed that in Mongol customs, a woman, upon dying a natural death, would be restored to her husband as a matter of course (William of Rubruck 1900: 70). It was supposedly in connection with this belief that Mongol concubines and widows were expected, or often compelled, to re-marry within their husband’s family (Boyle 1974).Footnote 16 Hence, it is quite possible that Hülegü’s wife, Doquz Khatun, is also buried in this fortress as, according to Rashīd al-Dīn’s account, she passed away only four months and eleven days after Hülegü himself (Rashīd al-Dīn 1052).

Fig. 8
figure 8

The southern side of the rock that has been hewn to create a vertical surface (Photo by Amin Moradi 2023)

Full size image

7 Conclusion

The most famous Mongol tomb, after that of Genghis Khan, lies in Lake Urmia on Shahi Island (Jazīreh-ye-Shāhī), Northwest Iran, and its mystery belongs to Hülegü. In fact, this is a true treasure story, par excellence, with a sufficient mystery to maintain its intrigue. Even though hundreds of treasure hunters have traveled to Jazīreh-ye-Shāhī to try to find the tomb and treasury of Hülegü, the site has never been found because there is no sign of a plundered royal Mongol tomb either inside or in the surroundings of the fortress. To be more precise, Hülegü’s successors left an empty treasury, but the tombs of Hülegü and other Mongol elites await discovery in Jazīreh-ye-Shāhī. It is hard to say how many tombs remain undiscovered but it is certainly more than just a few, in which immense sums of wealth remain buried. This discovery would be important because it could usher in another era for Mongol archaeology. This paper hopes to develop a consensus among archaeologists on such a topic.