СТАТЬИ

This study is an attempt to re-construct the history of the town of Hadjar, probably identical with antique Gerra / Gerrha, in Sasanid times. The history of ancient Gerra / Gerrha ended in the second half of the 2^nd century, when its inhabitants were driven away by the Arabic tribes of Tanūkh. Sasanids established their rule over Hadjar in the 1^st half of the 3^rd century and kept it till the late 5^th century. The advance of King Abkarib As‘ad of Himyar and Saba to the north in the mid-5^th century affected Hadjar and probably resulted in a clash with Sasanid forces, which was reflected in Arabic legends as his victory over Iran. However, Hadjar seems to have remained under the Sasanid power. Then, from the late 5^th through the end of the 6^th century Hadjar passed from Sasanids and their vassals, the Lakhmid kings of al-Ḥīra, to Banū Kinda and back. It was finally recovered by the Sasanids at the end of the 6^th century, following the migration of Banū Kinda from the Baḥrayn region. The Sasanids re-constructed and fortified Hadjar. By the beginning of the 7^th century Hadjar was an enough strong fortress to be selected as the place of the attack on Banū Tamīm. Like al-Ḥīra, Hadjar hosted then two governors, one Persian and one Arab. Economically, in the late 6^th through the early 7^th century Hadjar was an important centre of trade. In the 6^th century and later on, a Christian community headed by a bishop was present in Hadjar.

Скачать PDF Скачать JATS

In an article entitled Banū Kinda in Hadjar in the Second Half of the 6th Century and published in No. 1 of Vostok (Oriens) of 2019, I ventured to present a short history of the Banū Kinda rule in Hadjar, North-Eastern Arabia [Mishin, Banū Kinda, 2019]. That study only covered a part of the long history of Hadjar. To follow it, one would need a larger study.
A geographical localization of Hadjar has recently been suggested by ʻA. ʻA. al-Djanabī, who in 2004 published a book entitled ‘Hadjar, Its Three Fortresses (al-Mushaḳḳar–al-Ṣafā–al-Shabʻān) and Its River Muḥallim’. It undoubtedly goes to the author’s credit that he successfully combines written evidence with his own observations made on the spot. With no possibility to made an alternative investigation, I have to rely upon Mr. al-Djanabī’s conclusions. Those are that Hadjar was situated below the mountain of al-Shabʻān (present-day al-Ḳāra), to the north-west of it. The fortress of al-Mushaḳḳar is identified with a hill standing in the centre of the Ḳaryat al-Ḳāra village and bearing the name of Djabal Raʼs al-Ḳāra. The fortress of al-Ṣafā is a hill situated a little to the south-west of the Djabal Abī-l-Ḥaṣīṣ mountain [al-Djanabī, 2004, p. 238].¹

Another study which must be mentioned here belongs to Chr.-J. Robin, who suggests that Hadjar is to be identified with the town named Gerra or Gerrha of antique writers, situated in present-day oasis of a-Hufūf [Robin, Prioletta, 2013, p. 137–139; Robin, 2016, p. 226–227],² which in practice agrees with Mr. al-Djanabī’s conclusions above.³

The extant evidence coming from the times of Antiquity shows Gerrha as a town whose inhabitants were engaged in a long-distance trade along the Euphrates and across the Arabian Peninsula [Geographi ... 1882, p. 189; Polybius, 1925, p. 424–425 (Plb., XIII, 9); Strabo, 1930, p. 302–303 (Strab., 16, 3)]. The last, in terms of chronology, reference belongs to Claudius Ptolemy (wrote ca. 150 A.D.). He almost never goes beyond specifying the geographical position of a place and at this instance only states that the inhabitants of Gerra possessed, besides Gerra itself, two more towns [Klaudios Ptolemaios, 2017, p. 626–627 (Ptol., 6, 7, 16)].
At that point begins the history which is the subject matter hereof. The chronologically closest evidence is supplied by Muslim writers, Abū-l-Faradj al-Iṣfahānī (897/98–967) and Abū ʻUbayd al-Bakrī (ca. 1010–after 1090/1091). Their accounts show an obvious affinity and probably go back to one and the same source, which seems to be the account on settlement of Arabic tribes by famous mediaeval expert on antiquities Hishām al-Kalbī (ca. 737/38–819/20 or 821/22), found in the geography of al-Bakrī [al-Bakrī, 1983, p. 17]. The narration runs so that the tribe of Banū Taym Allāt Ibn Asad Ibn Wabara belonging to the tribal confederation of Ḳuḍāʻa, as well as some of Banū Rufayda Ibn Thawr Ibn Kalb Ibn Wabara (a branch of the Ḳuḍāʻā too) and of al-Ashʻariyyūn marched to the region of Baḥrayn⁴ and reached Hadjar. The latter was inhabited by ‘Nabataeans’ (nabaṭ). The Arabs prevailed over them by force and drove them out. Following that the Arabs created a tribal confederation of Tanūkh [Abū-l-Faradj al-Iṣfahānī, 1905, 11, p. 155; al-Bakrī, 1983, p. 21]. Later on they were joined by other Arabs, from the tribal confederation of al-Azd, who, according to al-Yaʻḳūbī (872/73–beginning of the 10^th century, but not earlier than 905), arrived from Oman [al-Ya‘ḳūbī, 1883, p. 233].

In the mediaeval Islamic literature the word ‘Nabataeans’ applied to Semites who were neither Arabs, nor Jews [Mishin, 2017, p. 47, note 36]. Such a description would fit well to the inhabitants of Gerra/Gerrha, who, as Strabo states, were Chaldaeans expelled from Babylon [Strabo, 1930, p. 302–303 (Strab., 16, 3)]. In the struggle against them nomad Arabs were in a better position from the start, since they could cut trade routes with their raids and thus deal a blow on Gerra/Gerrha’s commerce. The reference to the driving-out of the Nabateans means that the antique Gerra/Gerrha in fact ceased to exist.
It is possible to ascertain the chronological borders of that Arabic migration. The earliest point in time is around 150 A.D., where Claudius Ptolemy wrote his geographical treatise. It contains a mention of Thanouitai, doubtlessly identical with the Tanūkh yet located not in the North-East of the Arabian peninsula but in its southern part, near Katanitai, i.e., Banū Ḳaḥṭān [Klaudios Ptolemaios, 2017, p. 628–629 (Ptol., 6, 7, 23); Mishin, 2014, p. 273, note 396]. The latest point may be set up on the basis of statements by al-Ṭabarī (839–922/23) and Ḥamza al-Iṣfahānī (ca. 893/94–between 961/62 and 970/71) that the Tanūkh confederation was formed in the region of Baḥrayn in the time of ‘petty kings’ (mulūk al-ṭawāʼif) [Ḥamza al-Iṣfahānī, 1921/22, p. 63; al-Ṭabarī, 1881-1882, p. 747]. That term is used by Muslim writers to denote the Arsacid epoch, which they normally consider as the time of weak central power and actual dismemberment of the Iranian empire. Another 10^th century author, Muṭahhar al-Maḳdisī, who wrote around 966, states that Tanūkh chieftains controlled a number of regions in Southern Iraq during the rule of Ardashir I (225–240), the founder of the Sasanid dynasty, or shortly afterwards [Muṭahhar al-Maḳdisī, 1903, p. 196]. This means that the Tanūkh migration to the Baḥrayn region is to be put into the second half of the 2^nd century or, at the latest, to the beginning of the 3^rd century A.D.
Later on Tanūkh moved from the region of Baḥrayn, which was regarded to be poor in water, to Iraq, although a number of the Azd reportedly migrated to Oman [al-‘Awtabī al-Ṣuḥārī, 2006, p. 710–711]. As usually occurs in case of such migrations, some tribes must have remained where they lived.
The establishment of the Sasanid empire was in many respects the beginning of a new epoch in the history of the Middle East. Among other regions, Sasanids set foot in the Arabian peninsula. In the extant accounts of Ardashir I’s campaigns in Baḥrayn and Yemama it is stated that he killed the king of Baḥrayn, took his capital and granted security (amān), i.e., practically, received the vassalage of, three local rulers [al-Dīnawarī, 1960, p. 43; Nihāyat, 1996/97, p. 182–183]. He also built in the Baḥrayn region a fortress named Punyāt Ardashir, which is most probably to be identified with al-Zāra situated in the territory of present-day al-Ḳaṭīf [Mishin, 2020, p. 13–15]. Ardashir was able to seize Oman which appears in the inscription of his son and successor Shapur I (240–271) as a Sasanid possession [Mishin, 2014, p. 279].
This expansion is likely to have affected Hadjar. Late Muslim geographer Ḥamd Allāh Ḳazwīnī (1281/82 – after 1339/40) states in his description of the Baḥrayn region that its capital Hadjar was built by Ardashir I [Ḥamd Allāh Ḳazwīnī, 1915, p. 137]. Formally, this would provide enough grounds to conclude that Ardashir seized Hadjar, re-built it and made it his stronghold in Arabia. The usual critical approach to the sources suggests that, since this statement does not occur in earlier works, it may actually refer to Punyāt Ardashir wrongly identified with Hadjar as the regional capital of Baḥrayn. However, Ardashir is likely to have had reasons to build Hadjar as well. He needed a barrier between Punyāt Ardashir and the desert. Hadjar, howsoever in decline, could play that role and did not have to be built on an empty spot, for a number of its buildings were probably still standing.
Hadjar stays then out of sight for more than half a century. However, dramatic events were taking place in the Baḥrayn region. In the 3^rd century A.D. Banū Iyād moved there, followed later on by Arabic tribes belonging to the Banū Rabīʻa confederation. After a harsh struggle the Rabīʻite tribe of Banū ʻAbd al-Ḳays drove Banū Iyād out of Baḥrayn, to Southern Iraq. According to an account of the settlement of Arabic tribes, probably belonging to Hishām al-Kalbī, Banū ʻAbd al-Ḳays seized al-Khaṭṭ [al-Bakrī, 1983, p. 81; Kitāb, 1887/88, p. 8].
The name al-Khaṭṭ was used to denote the coastal zone between Basra and Oman, and island (possibly, identical with present-day Tārūt) near the Baḥrayn coast. Given the extension of the migration, the first of those meanings seems to be relevant. Arabs did not stop in al-Khaṭṭ, but began to attack Sasanid possessions on the southern coast of the Persian Gulf, from Abbadan to Busher. Those raids were facilitated by the fact that Sasanid king Shapur II (309/10–379/80) was minor, and the noblemen were divided between him and his elder brother Hormuzd in the struggle for power. Ultimately Shapur’s party got the upper hand. Shapur, upon reaching the age of 15-16 years, the Zoroastrian maturity, undertook a campaign in Arabia.
The accounts of that campaign are similar. With a selected army Shapur crossed to al-Khaṭṭ and then marched to Hadjar, then to the country of Banū ʻAbd al-Ḳays and reportedly reached Medina. As a result, North-Eastern Arabia was pacified for a long time, and a number of Arabic tribes became vassals of Shapur. To make his position in Arabia even stronger, Shapur settled some Arabic tribes in his possessions. According to one account found in the monumental history by al-Ṭabarī, Banū ʻAbd al-Ḳays and some of Banū Tamīm were settled in Hadjar, and Banū Taghlib, against whom Shapur had fought in the Euphrates region, near al-Khaṭṭ [Bal‘amī, 1959, p. 102; The Fársnáma, 1921, p. 68–69; al-Tha‘ālibī, 1900, p. 519–520; al-Ṭabarī, 1881-1882, p. 839]. The author of Fārs-nāmeh, who wrote in the first decade of the 12^th century, states that Banū ʻAbd al-Ḳays and Banū Tamīm were settled not in Hadjar itself, but in desert places of Hadjar, Yemama, and neighbouring regions [The Fársnáma, 1921, p. 69]. This, if correct, may suggest the presence of a Sasanid garrison in Hadjar, because some force must have been left in the fortress itself. But the general principles of methodology lead to prefer the earlier version by al-Ṭabarī, according to whom Arabic tribes were settled in Hadjar. They were probably governed by pro-Sasanid chieftains.
In another account no reference is made to Hadjar, but it is important that Shapur is said to have settled some Banū Taghlib and ʻAbd al-Ḳays in internal regions of the Sasanid empire [Bal‘amī, 1959, p. 107; al-Ṭabarī, 1881-1882, p. 845]. Should both accounts be correct, Shapur not only defeated Banū ʻAbd al-Ḳays, but also divided them, sending a part of them into exile. It would be natural to assume that the authorization to stay in the Baḥrayn region was only given to those whom Shapur did not expect to attack again, i.e., those who were now loyal.
Shapur’s campaign, thus, enabled the Sasanid empire to hold the region of Baḥrayn, including Hadjar, even more firmly than before. This state of affairs seems to have lasted fairly long. No Arabic attacks are recorded in the sources.
Afterwards Hadjar ceases to be mentioned in the sources for a while, but it appears again in connection with the events of mid-5^th century. That was the time when the rulers of the united state of Himyar and Saba (Yemen) began their advance to the north. King Abkarib Asʻad (ʾbkrb ʾsʿd of South Arabian inscriptions) left at Wādī Maʼsal⁵ an inscription known as Ry 509. The inscription has not been preserved in full, and its text is interpreted differently. Abkarib Asʻad, together with his son Ḥaṣṣān Yuhaʼmin (Ḥṣṣn Yhʾmn) is said to have stayed in the lands of Maʻadd (i.e., ʻAdnānite or Northern Arabs as different from Kaḥṭānite or Southern ones), have waged war and stayed in the lands of Maʻadd, or have undertaken a campaign and seized the lands of Maʻadd. The inscription goes back to the period of 433–455.⁶ Unfortunately for us, the text does not specify how far, if any further from al-Ma’sal, Abkarib Asʻad went. Yet this gap may be filled with information supplied by narrative sources. In Arabic stories preserved in Muslim sources Abkarib Asʻad, referred to as Tubbaʻ or Middle Tubba (Tubbaʻ al-awsaṭ), is depicted as a great conqueror king, who defeated Iran, subdued Transoxiana and successfully campaigned in China. All this is, undoubtedly, a legend. Yet one account belonging to al-ʻAwtabī al-Ṣuḥārī (fl. between the early 10^th and the early 13^th century) seems to provide a more realistic description of the king’s campaigns. In al-ʻAwtabī’s words, Abū Karib (Abkarib) Tubbaʻ campaigned against Maʻaddite Arabs, beleaguered Taif and then invaded Yemama. He ruthlessly destroyed his enemies. On some directions warfare was entrusted to separate detachments. One of those, commanded by Prince Ḥassān, probably identical with Ḥaṣṣān Yuhaʼmin, marched against Yemama. Another detachment reached al-Mushaḳḳar. The commander of the detachment, named ʻĀmir dhū Ḥiwāl, seized al-Mushaḳḳar and let his warriors do whatever they pleased with its inhabitants. Then, however, the narration progressively becomes legendary. It is asserted that another troop of Abū Karib’s army defeated the forces of the ‘king of the North’ (presumably, the Sasanid king) and took Ktesiphon [al-‘Awtabī al-Ṣuḥārī, 2006, p. 219–221].

Al-ʻAwtabī’s narration agrees with Ry 509 at least in that Abkarib Asʻad fought against Maʻaddite Arabs and stayed in Yemama for a while. Perhaps, his reference to the struggle with the ‘king of the North’, obviously legendary as far as the conquest of Ktesiphon is alleged, is based on some real clash. It would only be natural to suppose that Sasanid troops from Punyāt Ardashir and/or other places supported their Arab vassals and allies. If so, Arabic legends are likely to reflect a clash between an avant-garde of Abkarib Asʻad’s army and Sasanid forces. Those who fought for Abkarib Asʻad, in particular, his Arab subjects, apparently regarded themselves as winners, which, as the time passed by, developed into legends of the conquest of Iran.
Reverting to Hadjar, one may suppose that Abkarib Asʻad’s general fought against the Sasanid king’s Arab subjects, namely, Banū ʻAbd al-Ḳays and Banū Tamīm. The raid of the avant-garde is unlikely to have resulted in a long-term occupation of Hadjar. The recently discovered Sabaean inscription Maʼsal 3 contains a reference to another Himyarite campaign termed that of al-Khaṭṭ and undertaken in 474. The inscription has only been read in part, and its published fragments show that the Himyarite army marched to al-Maʼsal and then fought in Yemama [Prioletta, Arbach, 2016, p. 919, 922]. This would suggest that up to then kings of Himyar and Saba did not possess al-Khaṭṭ, which was still held by the Sasanids.
From the end of the 5^th century the Himyarite expansion was spearheaded by Arabs of the Banū Kinda tribal confederation. At about 513 or slightly later al-Ḥārith the War-Wager (al-Ḥarrāb), the king of the most powerful Kindite branch and former enemy of the Sasanids, switched his allegiance and became a vassal of Sasanid king Kawād I (488–498/99, 501–531). At that time Hadjar apparently belonged to Banū Kinda. Imruʼ al-Ḳays, the famous poet of pre-Islamic Arabia, states in a verse that his father Ḥudjr, who was son of al-Ḥārith and governed the tribe of Banū Asad Ibn Khuzayma on his behalf, commanded troups of Yemeni Arabs and dwelt in al-Mushaḳḳar [Dīwān, 2000, p. 688]. Imruʼ al-Ḳays himself is said to have lived in al-Mushaḳḳar or in some fortress in the Baḥrayn region [Abū-l-Faradj al-Iṣfahānī, 1905, 8, p. 61]. It is impossible to say whether Hadjar began to be governed by Banū Kinda before or after al-Ḥārith became vassal of the Sasanids. Yet it appears likely that King Kawād sanctioned the presence of Banū Kinda in Hadjar no matter whether they had occupied it before his contract with al-Ḥārith. Kawād reportedly expected al-Ḥārith to bring Arabia under the Sasanid power. Hadjar, in which Banū Kinda would have stayed at the beginning of the route to Central Arabia, but still under Sasanid control, would fit well to those plans. However, Kawād’s hopes did not come true. At a certain stage the Sasanids (it is not certain whether it was Kawād or his son Khusraw I Anōshag-rūwān (531–579)) lost their confidence in al-Ḥārith and, together with Lakhmid King al-Mundhir III (512/13–554) of al-Ḥīra, destroyed their Kindite allies in 520-30-es.⁷]

We are now coming to the events discussed in my article Banū Kinda in Hadjar in the Second Half of the 6^th Century. For the sake of continuity, it would be useful to briefly resume them. Since Yemen was conquered by Ethiopians in 525, its new rulers launched their own expansion in Arabia. In 553-554 Abraha, the Ethiopian ruler of Yemen, made a campaign to the north and, according to his inscription at Biʼr Murayghān termed as Murayghān 3, subdued Arabs of Hadjar (H.g.r.m.) and al-Khaṭṭ (Kh.ṭ.). That advance was backed by Banū Kinda who lived at al-Ghamr and were governed by rulers belonging to the Kindite royal family, although to a branch different from al-Ḥārith’s. Like their forefathers, those Banū Kinda served the rulers of Yemen. Kindite king Muʻāwiya Ibn al-Ḥārith (for the avoidance of doubt, not a son of al-Ḥārith the War-Wager’s) settled in al-Mushaḳḳar and re-built it. Soon thereafter Lakhmid King ʻAmr III (554–569) began a counter-offensive in Arabia, seized Hadjar and appointed his governor there⁸. Yet under his weaker successors Banū Kinda somehow recovered Hadjar. During the rule of Lakhmid King al-Nuʻmān III (579–601) Banū Kinda of Hadjar reportedly participated in some wars between Arabic tribes. However, they suffered considerable losses and left to the south of Arabia, being unable to hold Hadjar any longer [Mishin, Banū Kinda, 2019, p. 60–63].

At that time the Christian community of Hadjar first appears in the sources. In a Syriac account of the synod of the Nestorian church of the Sasanid empire, hold in February 576⁹ it is stated that Bishop Isaac of Hadjar and Punyāt Ardashir did not attend the event but backed his fellows’ decisions in writing [Synodicon, 1902, p. 128, 387]. The presence of a bishop indicates a numerous Christian community which is likely to have come into being as a result of Nestorian preachers’ tireless efforts.

It is difficult to explain why Hadjar precedes Punyāt Ardashir in the bishop’s title. Given that Punyāt Ardashir was the capital of the region of al-Khaṭṭ, it would seem more natural for Hadjar to be mentioned second, not the first. Likewise, it is impossible to say, what prevented the bishop from attending. It may be conjectured that Hadjar was then under the Kindite power and the bishop transferred his residence there in order not to abandon his flock. Likewise, the bishop’s absence at the synod may be explained by the assumption that Hadjar was no longer under the Sasanid control. Both explanation, however, remain hypothetical, although they seem to fit to the historical context.
After the Banū Kinda left, Hadjar returned under the Sasanid power. It seems that the Sasanids learned the lessons of the past and took steps not to lose Hadjar again. Al-Ṭabarī states that one of the asāwira,¹⁰ named Basak Ibn Māhbūdh (Middle Persian Wasag Mehbūdān) built al-Mushaḳḳar at the order of Kisrā. In order to provide workers employed at the construction with feminine society, whores were brought from Sawād and Kuzestan. Those people and their descendants dwelt in Hadjar in later years as well [al-Ṭabarī, 1881-1882, p. 985–986]. In another treatise entitled ‘Book on How Pardon was Granted and Asked for’ (Kitāb al-ʻafw wa al-iʻtidhār) Muḥammad al-Baṣrī the Writer (al-Raḳḳām), who lived in the 10^th century, states that Kisrā ordered Hawdha Ibn ʻAlī al-Ḥanafī to build, in the place where Hadjar was, a town which was to be residence of the governor and his refuge in case of a revolt of the Arabs. The king sent workers, and they built the fortress of al-Mushaḳḳar [al-Raḳḳām, 1981, p. 432]. It is quite possible that Abū-l-Mundhir (probably, Hishām al-Kalbī) as quoted by Yāḳūt actually refers to those works while stating that al-Ḳāra is a small mountain which the Persians erected using bitumen and tar [Yāḳūt, 1977, 4, p. 295].

Kisrā is an Arabic version of Syriac K.s.rōn, Khusraw. Muslim writers could apply that name to any Sasanid king, but ceteris paribus a king named Khusraw is the first-choice identification. The mention of Hawdha, who appears in al-Ṭabarī’s account too, although not in connection with the building of al-Mushaḳḳar, practically rules out any alternative. Ibn Saʻd (ca. 784–845) states that Hawdha died soon after the end of the year in which Muḥammad established his power over Mecca [Ibn Sa‘d, 1990, 1, p. 201], i.e., 8 A.H. (1 May 629–19 April 630). So, Hawdha’s death occurred in the spring or early summer of 630, which shows that Kisrā in question is Khusraw II Abarwēz (591–629).
The works were probably undertaken immediately after the Banū Kinda’s departure, when there was no Sasanid governor, as Muḥammad al-Baṣrī’s account shows. They must have been accomplished in the last decade of the 6^th century. This is inferred from accounts on the massacre of Banū Tamīm, most dramatic episode in the history of Hadjar. Those accounts may be summarized as follows. Tribesmen from the Banū Tamīm confederation (the sources refer to Banū Yarbūʻ or Banū Saʻd) plundered a caravan which transported goods belonging to Khusraw or sent to him. Angry Khusraw ordered to chastise Banū Tamīm. Persians and Hawdha decided to employ a stratagem. They proclaimed that the king had ordered to hold a reception for Arabs in al-Mushaḳḳar, and invited them to attend. Arabs, chiefly of Banū Tamīm, arrived at the fortress front gate. At the entrance, whey were told to leave their arms and after the banquet to go out of the fortress through another gate. But that was a trap. Arab were let in through the front gate in small and then put to death [Abū-l-Faradj al-Iṣfahānī, 1905, 16, p. 75–76; Dīwān, 1979, p. 62–64; Ḥamza al-Iṣfahānī, 1971, p. 120; Ibn al-Athīr, 1987, p. 492–493; Ibn Rashīḳ, 1981, p. 217; The Mufaḍḍalīyāt, 1921, p. 708–709; al-Ṭabarī, 1881-1882, p. 984–987]. The date of the massacre is never specified and can only be conjectured on the basis of Abū al-Faradj al-Iṣfahānī’s statement that it took part before the the famous battle of Dhū Ḳār [Abū-l-Faradj al-Iṣfahānī, 1905, 20, p. 135]. In a recently published article I presented my reasons to believe that the battle took place in the summer of 602 [Mishin, 2019, p. 28–31]. By then Hadjar must have been re-built and arranged enough to become fit for such an important operation.
A number of important details is contained in the account of Ḥammād the Teller (al-Rāwiya, 694/95–between 772 and 774), quoted by Abū al-Faradj al-Iṣfahānī through the intermediary of Hishām al-Kalbī. It is stated that Hawdha left Hadjar, then was attacked by Banū Sa‘d and taken prisoner. Afterwards he was set free in exchange for a ransom which he had to fetch from Hadjar [Abū-l-Faradj al-Iṣfahānī, 1905, 16, p. 76]. In another account which Abū al-Faradj al-Iṣfahānī quotes after Abū ‘Ubayda (728/29–824/25) and Muḥammad Ibn Ḥabīb (d. 860), both regarded as realiable authorities on the Arabs’ pre-Islamic history, Sasanid horsemen stationed in Hadjar are mentioned [Abū-l-Faradj al-Iṣfahānī, 1905, 16, p. 75].
It is noteworthy that one sees in Hadjar both Persians and Arabs. It may be inferred that Khusraw II employed in Hadjar the same system as was used in al-Ḥīra where both Persian garrison and pro-Sasanid Arab ruler resided. That system seems to have applied till the end of the Sasanid rule in Hadjar. Al-Balādhurī (d. ca. 892) narrates that Muḥammad in 8 A.H. (1 May 629–19 April 630) wrote to Sībukht (Middle Persian Sībōkht), the marzbān (military governor) of Hadjar, and al-Mundhir Ibn Sāwī of the Tamīmite tribe of Banū Dārim, calling upon them to embrace Islam [al-Balādhurī, 1987, p. 106–107]. Similar information is supplied by Ibn Sa‘d, who, it should be said, distorts the marzbān’s name as Usaykhit Ibn ‘Abd Allāh, Hishām al-Kalbī, and Yāḳūt [Hishām al-Kalbī, 1986, p. 201; Ibn Sa‘d, 1990, 1, p. 211; Yāḳūt, 1977, 1, p. 172].¹¹

In the fragments to which reference is made above, both Hishām al-Kalbī and Yāḳūt describe al-Mundhir as ‘lord of Hadjar’ (ṣāḥib Hadjar). Again, there are two governors, one Persian and one Arab. Yet the latter is not Hawdha but al-Mundhir. This may seem surprising, for the events in question took place during Hawdha’s lifetime. Explaining this shift is not easy. It is unlikely to be attributed to Khusraw II’s removal of Hawdha, for nothing of this kind is recorded in the extant sources, and Hawdha, so far as can be seen from his answer to Muḥammad, was a powerful ruler till the end of his life [Ibn Sa‘d, 1990, 1, p. 201]. Theoretically it is possible that Hawdha left Hadjar following the dethronement of Khusraw II in February 629. But the descriptions of Hadjar’s market discussed below show that Banū Dārim ruled it for a long time. The shift is, therefore, likely to have occurred earlier. It is most probably due to the fact that Banū Tamīm, to whom al-Mundhir belonged, lived in the Baḥrayn region, whereas Banū Ḥanīfa, the tribe of Hawdha, dwelt in Yemama. As a place of residence Hadjar was more fitting to al-Mundhir than to Hawdha, which seems the most plausible cause of the shift. The fact that Hawdha resided in Yemama at least for a while is illustrated by a verse of a contemporary of his, poet Maymūn al-A‘shā (the Weak-sighted), who states that once he visited him there [Gedichte, 1928, p. 66].
At that time Hadjar was centre of commerce and craft. A market was held each year, attended by merchants from Arabia and the Sasanid empire. According to a description of that market by Hishām al-Kalbī, which has reached us in treatises of Muḥammad Ibn Ḥabīb and al-Marzūḳī (d. 1030), the market was governed by the Banū ‘Abd Allāh Ibn Zayd family, to which, in particular, al-Mundhir Ibn Sāwī belonged.¹² Those men, like the Lakhmids of al-Ḥīra, were appointed by Sasanids and took a tithe from all merchants. Banū ‘Abd al-Ḳays and Banū Tamīm are described as protectors of the market by Ibn Ḥabīb and of al-Mushaḳḳar by al-Marzūḳī [al-Marzūḳī, 1996, p. 383; Muḥammad Ibn Ḥabīb, 1942, p. 264].¹³

The development of craft in Hadjar is illustrated by the fact that Hawdha, while sending Muḥammad’s envoy back, presented him with some clothes manufactured in Hadjar [Ibn Sa‘d, 1990, 1, p. 201].
Next in chronology was the Islamic conquest which was the beginning of a new epoch for the whole Middle East. The changes brought about by the conquest affected Hadjar considerably. Throughout the Sasanid period Hadjar was a border fortress of either Sasanids or their enemies. Following the conquest Hadjar was no longer at the border, and migrations of Arabic tribes were changing its population. That new situation well deserves a separate study.