² There is no good modern history of Mosul in this period. The best general introduction remains the article by Hönigmann, E. in the Encyclopaedia of Islam (first ed.), III, 609–11Google Scholar. For the historical geography of the area and the positions of the various tribal groups, see Canard, M., Histoire de la dynastie des Hamdanides, I, Paris, 1953, especially pp. 117–21Google Scholar. See also the article by Forand, P. G., ‘The Governors of Mosul, according to al-Azdi's Ta'rīkh al-Mawṣil’, Journal of the American Oriental Society, LXXXIX, 1, 1969, 88–105CrossRefGoogle Scholar.

³ For examples see al-Azdī, 197, 339–40.

⁴ ibid., 92, 371–2.

⁵ On the political significance of these terms see Shaban, M. A., Islamic history, 600–750, Cambridge, 1971, 120–21CrossRefGoogle Scholar. However, al-Azdī makes it clear that the Qays-Yaman conflict was by no means dead in the early 'Abbāsid period as Shaban implies.

⁶ Al-Azdī, 296–7, 332; Forand, , op. cit., 103Google Scholar.

⁷ On the titles of the governors see Forand, , op. cit., 102Google Scholar.

⁸ Al-Azdī, 156–78, al-Ṭabari, Annales, ed. de Goeje, M. J. et al. , Leiden, 1879–1901, III, 74Google Scholar.

⁹ Al-Azdī, 177–8.

¹⁰ ibid., 246, 251.

¹¹ ibid., 324; Forand, op. eit., 99.

¹² Al-Azdī, 335.

¹³ ibid., 332, 339–40.

¹⁴ ibid., 343–6.

¹⁵ ibid., 354.

¹⁶ ibid., 373–4.

¹⁷ Muḥammad b. Ḥumayd al-Ṭūsi was the son of Ḥumayd b. 'Abd al-Ḥamīd who had been al-Ma'mūn's senior commander in Iraq in the final years of the civil war, 201–4/816–19. Hārūn b. Abī Khālid was the brother of the wazir Aḥmad, see Forand, , op. cit., 101Google Scholar.

¹⁸ Al-Azdī, 429–31.

¹⁹ Ibn al-Athīr, Kāmil, ed. al-Najjār, 'Abd al-Wahhāb, Cairo, 1357/1938, v, 336, 339Google Scholar.

²⁰ Al-Azdī, 199.

²¹ ibid., 279, 282.

²² The first qāḍlī mentioned by Al-Azdī is Mu'ammar b. Muḥammad appointed by al-Manṣūr in 140/757, and Forand suggests that there was no such official in the province before this date, the qāḍī's functions being carried out by the governor. This contrasts with the position in Egypt where there had been qāḍīs from a very early date. See Al-Azdī, 173; Forand, , op. cit., 103Google Scholar.

²³ In 151/768, al-Ḥārith b. al-Jārūd, who wasqāḍī and chief tax official, was flogged on the orders of the caliph and died as a result, but this punishment was probably inflicted because of his failure as a tax-gatherer. See Al-Azdī, 214, 216.

²⁴ Al-Ashyab was appointed to Ṭabaristān after his dismissal from Mosul, but died before he could take up office (al-Azdī, 360—1). This is the only example from either Mosul or Egypt in this period of a man being appointed as qāḍī for two different provinces.

²⁵ See, for example, al-Azdī, 180, 199, 312.

²⁶ Al-Azdī, 181, 199, 312, 340.

²⁷ ibid., 217, 244, 365. In addition, it is probable that 'Abd Allāh b. Idrīs al-Hamdānī, appointed in 144, was one of the Hamdān of Mosul (al-Azdī, 181).

²⁸ Al-Azdī, 288, 302, 312. The exact form of the qāḍī's name is uncertain, see al-Azdī, 302, n. 1. It is clear, however, that he did not come from Mosul.

²⁹ For the distribution of the main 'Abbāsid armies see al-Ṭabarī, III, 304.

³⁰ Al-Azdī, 145–53.

³¹ For these numbers see al-Azdī, 177, 194–5, 267–9.

³² For Ḥarb's tenure of office see al-Azdī, 194–5, 201. Al-Azdī describes (p. 195) the function of the leader of the rawābiṭ, explaining that the post is separate from that of governor but responsible to him, cf. the somewhat ambiguous translation given by Forand, (op. cit., 93)Google Scholar. See also al-Ṭabarī, III, 328.

³³ Al-Azdī, 197.

³⁴ ibid., 203—4. Al-Ṣaqr is called al-Azdī and Mawṣilī.

³⁵ Al-Azdī, 267–9.

³⁶ It is significant that in the accounts of the two major disturbances of the late 170s, the rebellion of al-Aṭṭāf in the city and of the Khārijite al-Walīd b. Tarīf in al-Jazīrah, the rawābiṭ of Mosul was not mentioned at all. The decade was a period of increasingly oppressive taxation; in both Mosul and Egypt this caused widespread discontent. In Egypt the local troops became so disaffected that they were no longer willing to suppress rebellions and this may well have been true in Mosul as well.

³⁷ Egypt in the early 'Abbāsid period has been almost completely neglected by modern scholarship. The only important work is the valuable account by Wiet, G. in L'Egypte arabe (Paris, 1937)Google Scholar, which is especially good on taxation and relations between Christian and Muslim communities, but suffers from the fact that it is entirely without notes. See also the article ‘Fusṭāt’ in El (second ed.), II, 957–9.

³⁸ Some idea of the composition of the Muslim population of Egypt in this period can be gained from the examination of the surviving tombstones as published in the Répertoire chronologique d'épigraphie arabe ed. Combe, E., Sauvaget, J. and Wiet, G., I, 1931Google Scholar, henceforward referred to as RCEA. Up to the year A.H. 243, the numbers are as follows:

These figures must be used with caution and accidents of survival have probably affected the proportions considerably but they correspond in some measure to the names we find in al-Kindī and probably give some guide to the composition of the population. It is noticeable that all these tribes, with the exceptions of Quraysh and Tamīm, are to be found in the Yamanī group. For their tribal affiliations see Caskel, W., Ǧamharat an-nasab, I, Leiden, 1961Google Scholar, table 176 for Qaḥṭān genealogy. For Ghāfiq, see table 220, for Tujīb, table 240, for Murād and Nakh', table 258. For the Ṣadifis, see al-Sam'ānī, , Kitāb al-anṣāb (ed. Margoliouth, D. S., London, 1912, 350)Google Scholar, who explains that they were a tribe (qabīlah) of Ḥimyar which settled in Egypt.

³⁹ Al-Kindī, 76–7.

⁴⁰ ibid., 101, gives the details of estates given to 'Abbāsid supporters in the province. 'Assāmah b. 'Amr had property in the Fayyum (p. 115).

⁴¹ See below, p. 34.

⁴² Al-Kindī, 128.

⁴³ ibid., 105. He took with him to Syria some leaders of the wujūh to maintain his connexions with the province. The celebrated qāḍī Ghawth b. Sulaymān al-Haḍramī, one of these wujūh, continued to benefit from his patronage long after Ṣāliḥ had left Egypt (al-Kindī, 362).

⁴⁴ Al-Kindī, 74–5; Wiet, , L'Egypte arabe, 57Google Scholar.

⁴⁵ Al-Kindī, 108–9. Previously the ṣāḥib al-kharāj had been appointed by the governor (ibid., 106), but Nawfal was responsible directly to the caliph.

⁴⁶ The problem with the sources is that al-Ṭabarī and al-Kindī, both meticulous historians, give completely different names for the governors in the years A.H. 152–62. For 152, al-Ṭabari (III, 370) gives Yazīd b. Manṣūr (Yazid b. Ḥātim is the correct version) as wālī and Muḥammad b. Sa'īd as 'āmil. From 153 to 157 he simply gives Muḥammad b. Sa'īd as 'alā Miṣr (III, 371, 373, 377, 379). Muḥammad b. Sa'īd is known from al-Kindī: he was appointed ṣāḥib al-kharāj by Nawfal b. al-Furāt (p. 110) and is also said to have been responsible to al-Manṣūr for the kharāj (p. 365). From 157 to 159, al-Ṭabarī says that one Maṭar, a mawlā of the caliph's, was responsible for Egypt (III, 380, 467), but he is unknown to al-Kindī. In 159 the position is further complicated: in that year, according to al-Ṭabarī, al-Mahdī appointed Abū Ḍamrah Muḥammad b. Sulaymān as 'āmil of Egypt. This poses several problems: it may be that he is referring to the famous Muḥammad b. Sulaymān, many times governor of Basra and a leading member of the 'Abbāsid family. There is no record, however, that he ever visited Egypt or that he was ever known as Abū Ḍamrah. We should therefore rely on the variant reading (III, 467, note a, confirmed on III, 492) which makes Abū Ḍamrah the mawlā of Muḥammad b. Sulaymān. That Muḥammad b. Sulaymān was responsible for the administration is confirmed in a ṭirāz inscription published by Wiet, (RCEA, I, 34)Google Scholar which records that it was made for Muḥammad b. Sulaymān by his 'āmil in Tinnis. For the same period, al-Kindī mentions a series of governors of local origin, notably 'Abd Allāh b. 'Abd al-Raḥmān al-Tujībī and Mūsā b. 'Ulayy al-Lakhmī, who are all mentioned as being 'alā-'l-ṣalāt, but not alā-'l-kharāj. Only with the appointment of 'lsā b. Luqmān in 161 is there again a governor who fulfils both roles (al-Kindī, 117–20). Abū Ḍamrah is mentioned once (p. 121) as being in charge of the kharāj, Muḥammad b. Sulaymān not at all. We can probably conclude from these confused notices that during this period, locally recruited governors were responsible for prayers and public order while the financial administration was in the hands of men answerable directly to the caliph. In 159/775, Muḥammad b. Sulaymān was responsible for the kharāj and until 161 or 162 he appointed subordinates to act as his deputies (al-Kindī, 120; al-Ṭabarī, III, 491, 492, 493). If this hypothesis is correct, it has interesting implications for the historiography of the period. It suggests that al-Ṭabarī was working from records which detailed those in charge of the financial administration, while the local tradition used by al-Kindī emphasizes the role of the wālī who supervised prayers and public order. This would seem to be confirmed in the differing accounts of the affair of 'Umar b. Mihrān (see below, note 51).

⁴⁷ Ibn Khurdādhbih, , Kitāb al-masālik wa'l-mamālik, ed. de Goeje, M., Leiden, 1889, 83–4Google Scholar, and al-Jahshiyārī, , Kitāb al-wuzarā', ed. al-Saqqā, M. et al. , Cairo, 1938, 287Google Scholar, also give figures for the reign of Hārūn well below the recorded Umayyad totals. For a discussion of these figures see Ashtor, E., A social and economic history of the Near East in the Middle Ages, London, 1976, 64–5Google Scholar.

⁴⁸ al-Kindī, 125.

⁴⁹ For his reputation see the Christian authors, especially the pseudo-Dionysius of Tell-Mahré, J. Chabot (ed. and tr.), Paris, 1895, 91 ff., and Michael the Syrian idem, Chronique, Paris, 1899–1900, II, 526. For his origins see al-Azdī, 226.

⁵⁰ His career in Egypt is described by al-Kindī, 124—8, and al-Ya'qūbi, , Ta'rīkh, ed. Houtsma, M., Leiden, 1883, II, 489Google Scholar.

⁵¹ The affair of 'Umar b. Mihrān provides another instance of the discrepancy in the sources discussed above (see note 46). He is not mentioned by al-Kindī or subsequent Egyptian sources. Al-Ṭabarī, III, 626–8, and al-Jahshiyāri, 217–20, give fairly full accounts. Once again there is independent confirmation of their version, in this case a papyrus published by Grohmann, H. (From the world of Arabic papyri, Cairo, 1952, 116)Google Scholar. For further discussion of these problems see Grohmann, , Corpus papyrorum raineri, Vienna, 1924, II, 119Google Scholar.

⁵² Al-Kindī, 139–40.

⁵³ ibid., 140.

⁵⁴ ibid., 110, 129, 132–3.

⁵⁵ ibid., 126.

⁵⁶ ibid., 103, 133.

⁵⁷ ibid., 132.

⁵⁸ ibid., 12.

⁵⁹ ibid., 15.

⁶⁰ ibid., 21.

⁶¹ ibid., 324.

⁶² ibid., 67.

⁶³ ibid., 329–30.

⁶⁴ ibid., 93.

⁶⁵ ibid., 98.

⁶⁶ ibid., 110. The author specifically states that he was appointed on the advice of the jund.

⁶⁷ ibid., 117–18.

⁶⁸ ibid., 117.

⁶⁹ For Hāshim see al-Kindī, 121, 139, 142, 148, 389.

⁷⁰ For Hubayrah see al-Kindī, 149, 151, 154.

⁷¹ ibid., 419.

⁷² ibid., 425.

⁷³ ibid., 371–3.

⁷⁴ For a discussion of this change see Tyan, E., L' Organisation judiciaire en pays d'Islam, Leiden, 1960, 120–4Google Scholar.

⁷⁵ cf. Tyan, , op. cit., 156, 170–4Google Scholar, who considers the reference to Ḥanafīs anachronistic.

⁷⁶ Al-Kindī, 369, 388 and 394–7.

⁷⁷ ibid., 148. Al-Sarrī is said to have arrived in 182/798, but he played no role in the politics of the province until after 193/808. It is not clear when, and in what circumstances the Khurāsānī soldiers who supported him came, but it must be assumed that many of them arrived at the same time as he did. Nothing is known of his previous career. He is described in some sources as a Balkhī and his father al-Ḥakam b. Yūsuf, a mawlā of Banü Ḍabba, had briefly been ṣāḥib al-shurṭah to al-Manṣūr in 158, which shows he was of some standing in the army. He had property in the Ḥarbiyah quarter in Baghdad. See al-Ṭabari, III, 384; al-Ya'qūbī, , Buldān (ed. de Goeje, M., Leiden, 1892), 248Google Scholar; al-Maqrīzī, , Kkiṭāṭ (Al-Shayyah, 1959), II, 86Google Scholar.

⁷⁸ The entire complex struggle is covered in exhaustive detail by al-Kindī, 147–84, and in rather more general terms by Wiet, in L'Egypte arabe, 66–71Google Scholar. The power of local jund was finally broken by the Khurāsānis in the battle at Fusṭāṭ described by al-Kindī, 159–60. This defeat sounded the death knell of the old order and it is significant that al-Sarrī moved the centre of government from Fusṭāṭ, dominated as it was by the wujūh, north to the area of al-'Askar (al-Maqrīzī, II, 76).

⁷⁹ Al-Kindī, 193–4; al-Maqrīzī, I, 168.