1. See e.g. Ganji, M. H., “Climate,” Cambridge History of Iran (hereafter CHIr) ed. Fisher, W. B. (Cambridge: Cambridge University Press, 1968) 1: 212–49CrossRefGoogle Scholar; Ehlers, E., “Climate,” Encyclopædia Iranica (henceforth EIr) 5: 707–13Google Scholar; Balland, D., “Bārān,” EIr 3:745–53.Google Scholar

2. For an outline of the region's history, see Barthold, W., An Historical Geography of Iran, trans. Soucek, S., ed. Bosworth, C. E. (Princeton: Princeton University Press, 1994), 230–42Google Scholar; Spuler, B., “Gīlān,” The Encyclopædia of Islam New Edition (henceforth EI²) 3: 1111–12Google Scholar; and Minorsky, V. and Bosworth, C.E., “Māzandarān,” EI² 6: 935–40.Google Scholar Unfortunately the detailed history of Tabaristan promised under that heading did not materialize (q.v.).

3. See the snapshot of the situation in the early fourteenth century in the work of Kashani, Tārīkh-i Uljāytū, ed. Hambly, M. (Tehran, 1348/1969), 56–61Google Scholar; cf. Rabino, H. L., “Deux descriptions du Gilan au temps des Mongols,” Journal Asiatique 238 (1950):325–33Google Scholar; and Melville, C., “The Ilkhan Öljeitü's Conquest of Gilan (1307): Rumor and Reality,” in Amitai-Preiss, R. and Morgan, D. O., eds., The Mongol Empire and its Legacy (Leiden, 1999), 88.Google Scholar

4. In a recent series devoted to regional studies in Iran, the first to appear were three massive volumes on Gilan. See Ibrahim Islah ^ᶜArabani ed., Kitāb-i Gīlān, reviewed in Kelk 71-72 (1374/1995): 284-89. I have not been able to see this work.

5. For those discussed in the text, see Kasravi, A., “Tavārīkh-i Êṭabaristān va yād dāsht-hā-yī mā,” in Nawbahār-i haftagī (1301/1923) reprinted in Zoka, Y., ed., Essays from Kasravi (Tehran, 1334/1956), 14–17.Google Scholar See also Storey, C. A. and Bregel', Y. E., Persidskaia Literatura: Biobibliograficheskii Obzor (Moscow, 1972) (henceforth Storey/Bregel) 2:1074, 1077Google Scholar; Rosenthal, F., History of Muslim Historiography, 2nd. ed. (Leiden, 1968), 474Google Scholar, for a lost history of Mazandaran.

6. Therefore excluding principally Astarabad (Jurjan). For a lost history of Astarabad, see Sezgin, F., Geschichte des arabischen Schrifttums (Leiden, 1967) (GAS) 1: 352.Google Scholar Sahmi's Tārīkh-i Jurjān is primarily a biographical work; cf. Pourshariati, P., “Local Histories of Khurasan and the Pattern of Arab Settlement,” Studia Iranica, 27 (1998): 53–57.CrossRefGoogle Scholar For bibliographical details of the editions of the local histories discussed here, see below, pp. 88-89

7. Oxford Bodleian ms. Ouseley 214 contains brief chronological notes down to 842/1438-39.

8. Although the date of composition is 1044/1634 (Shaikh-^ᶜAli, 24), the narrative effectively ends with the Safavid subjugation of the province and the governorship of Farhad Khan, ca. 1007/1599.

9. Tārīkh-i Ṭabaristān: al-tadwīn fī aḥwāl jibāl Sharwīn, ed. Mihrabadi, M. (Tehran, 1373/1994), 301.Google Scholar

10. One such study, for example, might be the coverage of Mazyar's revolt; for a jaundiced view of Ibn Isfandiyar's account of this, see Rekaya, M., “Māzyār: résistance ou intégration d'une province Iranienne au monde musulman au milieu du ix siècle ap. J.C.,” Studia Iranica 2 (1973): 144–45.Google Scholar The difference between the “imperial” and “local” views might be more interesting than the “facts” of the matter.

11. Meisami, J. S. Persian Historiography to the Twelfth Century (Edinburgh, 1999), esp. 12, 281–98.Google Scholar As Meisami acknowledges, the ground had been prepared by Marilyn Waldman's study of Bayhaqi, Toward a Theory of Historical Narrative: A Case Study in Perso-Islamicate Historiography (Columbus, 1980)Google Scholar: see her introductory chapter, 3-25.

12. Carr, E. H., What Is History?, ed. Davies, R. W. (Harmondsworth, 1990), 22Google Scholar: not often asked of Persian historiography.

13. Hampton, Timothy, Writing from History: The Rhetoric of Exemplarity in Renaissance Literature (Ithaca & London, 1990), 35CrossRefGoogle Scholar; quoted by Meisami, Persian Historiography, 287.

14. For the European context, see among many Guenée, B., Histoire et culture historique dans l'Occident médiéval (Paris, 1980), 332–54Google Scholar; Bagge, S., “Propaganda, Ideology and Political Power in Old Norse and European Historiography: A Comparative View,” in L'Historiographie médiévale en Europe , ed. Genet, J.P. (Paris: CNRS, 1991), 199–208Google Scholar; and Kennedy, E. D., “Romancing the Past: a medieval English Perspective,” in The Medieval Chronicle, ed. Cooper, E. (Amsterdam, 1999), 23–30.Google Scholar

15. Tārīkh-i Ṭabaristān, 1: 82.

16. See ^ᶜAta Malik Juvayni, Tārīkh-i Jahāngushā, ed. Qazwini, M. M. 3 vols. Gibb Memorial Series (London, 1912-37) 1:96–101Google Scholar; trans. Boyle, J. A., The History of the World Conqueror 2 vols. (Manchester and Cambridge, Mass., 1958) 1: 123–28Google Scholar for confirmation of the qualities of Khwarazm on the eve of the invasions; also, briefly, al-Hamawi, Yaqut, Mu^ᶜjam al-buldān, ed. Wustenfeld, F. (Leipzig, 1866) 4: 360–61.Google Scholar

17. Tārīkh-i Ṭabaristān, 1: 302, 2: 175; incidentally, this shows the remarkable speed at which calligraphers worked.

18. For a detailed discussion of the complicated manuscript tradition, see the editor's introduction, ii-xii. See also Melville, C., “Ebn Esfandīār,” EIr 7: 720–22.Google Scholar

19. M. Sotudeh, intro. to Tārīkh-i Gīlān, xiii, says the ms. was copied in 936/1530, which must simply be a misprint. The date falls within the reign of Khan Ahmad, who doubtless commissioned the copy. For his career, see ^ᶜA. Nava^˒i, “Sarguẕasht-i Khān Aḥmad Khān,” in Nāmah-hā-yi Khān Aḥmad Khān Gīlānī, ed. Nausad, Feridun (Tehran, 1373/1994)Google Scholar, unpaginated introduction (8-12). The final subjugation of Gilan by Shah ^ᶜAbbas is dated 1000/1592 by Iskandar Beg Munshi, ed. Iraj Afshar (Tehran, 1344/1956) 1: 448-51; trans. Savory, R. M., History of Shah Abbas the Great (Boulder,1978)2:621–25Google Scholar, though problems continued. The Safavid takeover of Rustamdar and Mazandaran followed in 1006/1597 and 1007/1599 (Munshi, 1: 534-37, 585-86; Savory, 2: 713-17, 772-23); cf. Shaykh-^ᶜAli Gilani, Tārīkh-i Māzandarān, 88, 100Google Scholar, and Edmund Bosworth, Clifford, The New Islamic Dynasties (Edinburgh, 1996), 201–2.Google Scholar The mss. of Ibn Isfandiyar were presumably commissioned by local rulers interested in publicizing their hereditary rights.

20. Tārīkh-i Ṭabaristān, 2: 131; see also below, n. 95.

21. The best account is still that of Bosworth, C. E., “The Political and Dynastic History of the Iranian World (A.D. 1000-1217),” CHIr 5: 1–202Google Scholar, esp. 180-92, although he relies only on Browne's summary of Ibn Isfandiyar. See also Madelung, W., “Āl-e Bāvand,” EIr 1: 749–52.Google Scholar

22. E.g. the former is called Khwārazmshāh-i sa^ᶜīd-i ^ᶜādil (2: 86, 113) and ṣāḥibqirān (2: 129), sulṭān-i jahān (2: 132), sulṭān-i a^ᶜẓam, shāhanshāh-i sa^ᶜīd (2: 157), mālik riqāb al-umam, ṣāḥib-qirān-i ^ᶜālam (2: 158); and the latter sulṭān-i a^ᶜẓam, Iskandar-i zamān (2: 168), khudāvand-i ^ᶜālam, sulṭān-i banī ādam, shāhanshāh-i kishvar-gushāy, Iskandar al-zamān (ibid.).

23. For Amuli, see Madelung, W., “Awlīā^˒ Allāh Āmolī.,” EIr 3: 120–21Google Scholar; Sotudeh's edition is used here (see below, p. 88).

24. Amuli, 77-78. Cf. p. 79 for evidence of his visit to Baghdad, and Sotudeh, intro., xxiv-xxv.

25. We see here, as indeed in almost all surviving prefaces, a good example of what Sholeh Quinn calls the “key turn of events” that led to the composition of the history. See her “The Historiography of Safavid Prefaces,” in Safavid Persia, Pembroke Papers, no.4, ed. Melville, C. (London, 1996), 1–25.Google Scholar

26. Amuli, 121. For Gavbara, see Ibn Isfandiyar, 1:153-54. Compare the Bavandid genealogy presented by Ibn Isfandiyar, 2:126, effectively the same after Firuz son of Yazdagird the Sasanian, back to Adam via the “semitic” route through Jacob.

27. Mar^ᶜashi, Tārīkh-i Ṭabaristān va Rūyān va Māzandarān, ed. ^ᶜA. Shayan (Tehran, 1333/1955)Google Scholar, 35ff. and 229-30 (lengths of reigns). Subsequent citations of Mar^ᶜashi refer to this work and edition unless otherwise stated (see below, p. 89). See also W. Madelung, “The Minor Dynasties of Northern Iran,” CHIr 4: 218-19, idem, “Baduspanids,” EIr 3: 385.Google Scholar In contrast with Amuli's genealogy of the Baduspanids, Mar^ᶜashi, 102-3, traces the last stages from Nauzar b. Manuchihr back to Adam via the “Iranian” route to Kayumars, probably to emphasize its beginning and end with a ruler of that name.

28. E.g. Amuli, 156-67, 161-62, 168, 170, 172, 183, 198, 199-200, 202-3.

29. Ibid., 117-18; the same theme is taken up in the chapter on genealogy that follows, where the continual presence of these “Lords of the mountains” is asserted (120).

30. ^ᶜAbd al-Fattah Fumani, Tārīkh-i Gīlān, ed. Sotudeh, M. (Tehran, 1320/1941), 266.Google Scholar He does not say when he returned. Cf. Quinn, Sh., “Fūmanī,” EIr 10: 228–29.Google Scholar

31. See Bosworth, C. E., “Āl-e Afrāsīāb,” EIr 1: 742–44Google Scholar, and Calmard, J., “Mar^ᶜashīs,” EI² 6: 510–16.Google Scholar

32. Mar^ᶜashi, 386.

33. East of Daylam. Mar^ᶜashi's career is summarized by Kasravi, 42-44, and in more detail in M. Sotudeh's introduction to the Tārīkh-i Gīlān (Tehran, 1349/1970), xxxi–lv.Google Scholar

34. This set a precedent for Shams al-Din Lahiji's claim that he was merely drafting the material assembled by Khan Ahmad Khan, see Lahiji's, Tārīkh-i Khānī, ed. Sotudeh, M. (Tehran, 1353/1974), 6–7, 390-91.Google Scholar Tārīkh-i Khānī essentially continues Mar^ᶜashi*s Tārīkh-i Gīlān.

35. Referred to as still alive at this point, Mar^ᶜashi, 227 (also 229-30); his death in 881/1476 is mentioned by Madelung, “Baduspanids,” 389.

36. Mar^ᶜashi, 345-46.

37. Waldman, Toward a Theory of Historical Narrative, 6.Google Scholar

38. Amuli, 7-8, suggests some reluctance.

39. Tārīkh-i Ṭabaristān, 2: 33 (500 H.), 70 (521 H.), 105 (558 H.), 106 (558 H.), 159 (598 H.), 171 (602 H.).

40. Cf. Madelung, “Āmolī,” 121; idem, “Baduspanids,” 386. These dates are generally of the various rulers only, rather than of events.

41. For the distinction between annals, chronicles, and histories, see e.g. Gransden, A., Historical Writing in England, 2 vols. (London, 1974, reprinted 1982) 1: 29–35Google Scholar; Guenée, Histoire, 203-5. The debate still rages over the different genres, see e.g. Ward, “‘Chronicle’ and ‘History’: the Medieval Origins of Postmodern Historiographical Practice?” Parergon n.s.14/ii (1997): 112-22, and need not detain us here. For the Islamic context, see Rosenthal, A History of Muslim Historiography, 66-86; Stephen Humphreys, R., Islamic History: A Framework for Inquiry, rev. ed. London, 1995), 129–30;Google Scholar and recently, Malti-Douglas, F., “Texts and Tortures: the Reign of al-Mu^ᶜtaḍid and the Construction of Historical Meaning,” Arabica 46 (1999): 316–22.CrossRefGoogle Scholar

42. Lahiji, 391: the date is mid-Safar 922/20 March 1516 and the occasion is the reconciliation between the Ustandars of Nur and Kujur, brokered by Sultan Ahmad Khan—a positive note on which to end; cf. Madelung, “Baduspanids,” 389. Fumani also ends on an optimistic note, though nonetheless abruptly.

43. Ibn Isfandiyar, 2: 136, 153.

44. Such as Mafarrukhi's history of Isfahan, discussed by J. Paul in this volume.

45. The earliest of these late copies, British Library, ms. Ethé 568, fol. 64v, has a heading that corresponds almost exactly to the title of section two (1: 8), but it is identified as a guftār not as a new qism. The heading at this juncture in the earliest ms. is entirely different. Browne was evidently unclear where to mark the break, for he wrote Section II in the margin here and again on the next folio (65v) at the heading on the Al-i Vushmgir b. Ziyar, kings of Gilan. The latter heading is given much prominence in ms. Add. 7633, f. 81r, and also has a certain logic as the start of section two; cf. text, 1: 141, Browne, 91. It occurs in similar terms in all the manuscripts. According to Ethé's Bodleian catalogue, 160-1, headings in Ouseley 214 (which I have not inspected), are marked for section 2 (f. 94b), 3 (f. 123b) and 4 (f. 196a). The copy was made in Rabi^ᶜ I 1068/December 1657, a year after B.L. ms. Add. 7633. See also below, note 109.

46. It is not clear whether the original division into mujallads is coterminous with the division into qisms envisaged in the preface. The point is further confused by the division of the printed text into two volumes, vol. 1 containing section one and vol. 2 containing sections two and three according to ^ᶜAbbas Iqbal's scheme. Here I retain the term “section” for the principal divisions of the work.

47. See Madelung, W., “The Minor Dynasties of Northern Iran,” CHIr 4: 219Google Scholar; Khan, M. S., “A Manuscript of an Epitome of al-Ṣābī's Kitāb al-Tāğī” Arabica 12 (1965): 41CrossRefGoogle Scholar, and note 80 below.

48. This confirms the date as the 340s, when several outbreaks of epidemic illness (vabā^˒) are reported in Rayy, Khurasan, and the Jibal, e.g. al-Athir, Ibn, al-Kāmil fi'lta^˒rīkh, ed. Tornberg, C. J. (Beirut, 1965 reprint of 1867 edition) 8: 169, 170, 172.Google Scholar On the other hand, it could be that the invocation is taken verbatim by Ibn Isfandiyar from his source, for he does not mention the “plague” himself. It could therefore also be a later addition. In that case it is worth noting that when the Continuation of Ibn Isfandiyar was being written, during the time of last Bavandids, the Black Death was raging; cf. Mar^ᶜashi, 191. See also next note.

49. ^ᶜAbbas Iqbal considers this passage, and the “weak” verses that follow, to be scribal additions to the original text, but the basis for this—that they are not found in later mss., which do not mark any break in the narrative at this point—is hard to support in view of the obvious superiority of the earliest copy. It remains true, however, that this is an odd place for the division of the volumes. If the language of this passage is in a more elevated style than the preceding narrative, it is no more so than the preface, and shares its somewhat fatalistic outlook.

50. ^ᶜAbbas Iqbal, ix, assumes that the title intends the end of the dawlat in 750 H. and that it is thus necessarily a later addition, but there is no reason why it should not refer to the end of the second dawlat, ca. 606/1203. The late mss., e.g. Add. 7633, f. 184v, start qism-i chahārum at 2: 19 in the printed text.

51. The final passage, concerning a certain ^ᶜAlavi Musavi and his sons, however, seems a strange place to finish, and less well designed to shape a conclusion than the reconciliation of the brothers that precedes it. The two passages are reversed in the later manuscripts (see 2: 172-23; cf. Browne, 255-56).

52. The first mentions Ardashir, i.e. Rustam's father and Ibn Isfandiyar's benefactor, as its subject, though there may also be the suggestion of a lament in anticipation of the murder of Rustam. Later manuscripts, which omit the verses, move seamlessly into an account of his successors (see below).

53. The fact that the 1595 preface, 1: 1, 5, mentions the inclusion of the history of Husam al-Dawlah Ardashir's descendants (akhlāf), and an account of Rustam b. Ardashir's murder in the final book (mujallad-i ākhir), as well as the reference to a section of genealogies provided at the end of the book, 1: 152, might suggest it was actually completed.

54. Sotudeh, intro., xviii; the date is 1 Muharram 764/21 October 1362 (205). Madelung, “Āmolī,” 121, is skeptical about whether this was the date of completion of the work; there are also grounds for skepticism about whether it is the autograph (see below).

55. Amuli, 173; cf. Mar^ᶜashi, 68.

56. E.g. Amuli, 154, 167, 168, 173 and 175 (negative view of fraternal strife), 176.

57. The amir Mu^˒min and his son Qutlughshah. Although it is extremely tempting to try to equate one of these encounters with the campaign of Öljeitü's amir Qutlughshah, they must actually have occurred in the reign of Abu Sa^ᶜid. Cf. Melville, “Öljeitü's conquest.” The details of the family infighting are obscure and made confusing by the fact that the same names (Shams al-Muluk, ^ᶜAla^˒ al-Dawlah ^ᶜAli, etc.) were evidently held by various princes of different generations; see also below, p. 80.

58. Mar^ᶜashi, 191, is suitably noncommittal about the truth of the allegations. The Continuation of Ibn Isfandiyar, f. 215r, is less explicit. Cf. Madelung, “Baduspanids,” 387.

59. It is presumably to Abu Sa^ᶜid's death that Amuli, 204, inaccurately refers by the date 735 H., which is a non sequitur from what has gone before.

60. B. L. Ms. Add. 7633, f. 216r, actually reads 755 years, clearly an error.

61. Mar^ᶜashi, 2. The verse is also quoted by Amuli, 4: its use to justify the seizure of power by force was already invoked in Ghaznavid times, cf. Mottahedeh, Roy, Loyalty and Leadership in an Early Islamic Society (Princeton, 1980), 186–87.CrossRefGoogle Scholar

62. Incorporated into Rabino, H., Māzandarān and Astarābād (London, 1928), 133–49.Google Scholar For genealogy in Islamic history, see briefly Rosenthal, Muslim Historiography, 99-100. For a far more developed view, concerning thirteenth century France, see e.g. Gabrielle M. Spiegel, “Genealogy: Form and Function in Medieval Historiography,” reprinted in idem. The Past as Text: The Theory and Practice of Medieval Historiography (Baltimore and London, 1997), esp. 104–10.Google Scholar

63. Though twice mentioning the Tabari poem on spring, for example, he does not simply duplicate the report; see Mar^ᶜashi, 60, 190.

64. Amuli, 86. Cf. his similar views of the eclipse of the Sayyids by the Buyids, above, 16.

65. Guenée, Histoire, 27-29; Spiegel, “Political Utility,” esp. 89-94.

66. Cf. Lambton, Ann K. S., “Mirrors for Princes,” in La Persia nel medioevo. Accademia Nazionale dei Lincei, Quaderno n. 160 (Rome, 1971), 420–23Google Scholar; de Fouchécour, C.- H., Moralia: les notions morales dans la littérature person de 3e/9e au 7e/9e siècle (Paris, 1986)Google Scholar; Shaked, S. and Safa, Z., “Andarz i., ii.,” EIr 2: esp. 18–19.Google Scholar

67. Funduq's, Ibn Tārīkh-i Bayhaq, ed. Bahmanyar, A. (Tehran, 1317/1939), 7–17Google Scholar, is perhaps the best known prototype for such passages in Persian histories, cf. Meisami, Persian Historiography, 211-13, though Amuli's views clearly do not derive from this, except possibly in his conclusion that a knowledge of history is most useful for kings. His slightly earlier compatriot, author of Nafā^˒is al-funūn, 170, has a similar approach, cf. Rosenthal, Muslim Historiography, 39Google Scholar; Lambton, A. K. S., ‘Ta^˒rīḵẖ 2. In Persian,” EI² 10: 286–87.Google Scholar For the Arabic tradition, and particularly Miskawaihi's views on the use of history, see Khalidi, Tarif, Arabic Historical Thought in the Classical Period (Cambridge, 1994), esp. 170–76.CrossRefGoogle Scholar

68. Amuli, 26. Compare the briefer and less pointed account of Ibn Isfandiyar, 1: 150-52.

69. Cf. Mirkhwand, Rawżat al-ṣafā^˒ ed. ^ᶜA. Parviz (Tehran, 1338/1959) 1: 15.Google Scholar

70. Mar^ᶜashi's own treatment of this story (24—26) is similar but does not impart the same message for the servants of the ruler.

71. Mar^ᶜashi, Tārīkh-i Gīlān, 4-7.

72. Tārīkh-i Ṭabaristān, 1: 3.

73. Melville, “Ebn Esfandīār,” 21.Google Scholar

74. Meisami, , “Dynastic History and Ideals of Kingship in Bayhaqi's Tārīkh-i Mas^ᶜūdī,” Edebiyat 3 (1989): 59–61Google Scholar; idem, “Mas^ᶜūdī on Love and the Fall of the Barmakids,” Journal of the Royal Asiatic Society (1989): esp. 258-62; idem, Persian Historiography, 31-32.

75. Tārīkh-i Ṭabaristān, 1:190. The first reason is the relationship between Ja^ᶜfar and ^ᶜAbbasa, related on the authority of (^ᶜAli b. Muhammad) Nawfali; the second is the story of Ja^ᶜfar's release of a sayyid, the son of Yahya b. Zayd, taken from the Kitāb al-nawādir of al-Asma^ᶜi. Both stories occur in different form and on different authorities in many other sources, e.g. Tabari, Ta^˒rīkh al-rusul wa'l-mulūk, vol. 30Google Scholar, trans. Bosworth, C. E. as The History of al-Tabari: The ^ᶜAbbasid Caliphate in Equilibrium (New York, 1989), 205–6, 214-16Google Scholar, cf. Abu ^ᶜAli Bal^ᶜami, Tārīkhnāmah-i Ṭabarī, 3 vols., ed. Rawshan, M. (Tehran, 1366/1987) 2: 1195–96.Google Scholar

76. Ibn Isfandiyar draws no connection between the sayyid and his earlier revolt in Daylam and Tabaristan, which indeed he does not even refer to; cf. Tabari, trans. Bosworth, 115ff.; Bal^ᶜami, Tārīkhnāmah 2:1187–88.Google Scholar

77. Compare Ibn Isfandiyar, 1: 259; Amuli; 104; Mar^ᶜashi, 214.

78. E.g. Mar^ᶜashi, 52-54, 112, 155, 213-14.

79. Ibn Isfandiyar himself says (1: 125) that Yazdadi's compositions were too famous to need mentioning. This work is now lost; cf. Kasravi, 14—15.

80. See the index to volume 1 of Iqbal's edition for a list of titles. Khan, “An Epitome,” 41, draws attention to his heavy and underacknowledged use of Ibrahim b. Hilal al-Sabi's Kitāb al-Tājī; cf. Ibn Isfandiyar, 1: 139, 300. Madelung, W., “Abū Isḥāq al-Ṣābī,” Journal of Near Eastern Studies 26 (1967): esp. 23 n.33CrossRefGoogle Scholar, disagrees; it is more likely that he merely copied references to the Tājī from his source.

81. Tārīkh-i Ṭabāristān, 1:4 (Bāvandnāmah), 72, 83 (Kitāb-i Nīrūz va Mihrjān), 85. We may surmise that the Tārīkh-i Barāmika, as used by Yazdadi, was the Arabic original of the Persian version later made by Diya^˒ al-Din Barani; cf. Bouvat, Les Barmécides d'après les historiens arabes et persons (Paris, 1912), 9–10Google Scholar, and doubtless also the source of Yazdadi's version of the Barmakids’ fall (cf. above, n. 75).

82. E.g. concerning the association of Imam Hasan and Malik Ashtar with the province; on the latter's role as a rebel under ^ᶜUthman, he cites (46) the work of Ibn A^ᶜtham al-Kufi (d. ca. 926/314), i.e. his Kitāb al-futūḥ ed. ^ᶜAbd al-Majid (Hyderabad, 1389/1969) 2: 190–211.Google Scholar

83. Mar^ᶜashi, 9, 196 (Ruyani).

84. Amuli, 17; cf. Ibn Isfandiyar, 1: 60. Whereas Ibn Isfandiyar refers to the verse. of Firdawsi and the prose of Mu^˒ayyadi, Amuli refers to the verse of Firdawsi and Mu^˒ayyadi. The question remains open; cf. Storey, C. A. and de Blois, François, Persian Literature: A Bio-bibliographical Survey (London, 1994) vol. 5, part 1: 67–68.Google Scholar

85. Amuli, 19, throws in for good measure a eulogy of the cultured wazir.

86. See Tafażżoli, A., “Araš,” EIr 2: 266–67.Google Scholar

87. Cf. Rabino, Māzandarān and Astarābād, 32, 69.Google Scholar

88. E.g. Azhand, intro. to Mar^ᶜashi, Tārīkh-i Ṭabaristān, ed. Dorn, xix. See however, the more correct view already expressed by Kasravi, 44-45.

89. Khalidi, Arabic Historical Thought, esp. 39-44; Humphreys, Islamic History, esp. 71-87.

90. Sīrat al-Sulṭān Jalāl al-Dīn (Histoire du Sultan Djelal ed-Din Mankubirti, ed. O. Houdas, 1-2, trans. Houdas, 2-3; Persian trans. Khurandizi, 3.

91. Khalidi, Arabic Historical Thought, 200-1.

92. Ibid., 85, 113-15. See also Khaleghi-Motlagh, Dj. & Pellat, Ch., “Adab i., ii,” EIr 1: 431–44.Google Scholar

93. E.g. Ibn Isfandiyar, 1: 58, 61, 62, 106, 123, 125, 131, 161, 224; 2: 38, 62, 77, 84 (tower of skulls), 85, 90, 91; Amuli, 49, 77, 97, 114, 155-56.

94. Ibn Isfandiyar, 1 :156. Cf. Kasravi, 21.

95. Ibn Isfandiyar, 2: 134, 136. Cf. 2: 131 and 2: 91 for evidence of his other travels, for which his father rebuked him (for being absent so much).

96. Ibid., 2: 54, 72.

97. E.g. 169, 182; and 180 for his own information.

98. This seems paradoxical, but is probably due to Mar^ᶜashi's need to pay lip service to the refinements required when referring to current events. On irony, see briefly Meisami, Persian Historiography, 294.Google Scholar

99. Cf. Khalidi, 219. Ibn Isfandiyar is generally imprecise: the ulema in Rayy (1:5), or Khwarazm (1:7); “I heard” (1:55, 101, 108, 251).

100. Ibid., 1:140, 2:37 (father).

101. Mirkhwand, Rawżat al-ṣafā^˒, 1: 14–15.Google Scholar

102. Cf. Carr, What Is History?, 44.Google Scholar

103. Eg. Kasravi, 27-28 (written by Amuli); Iqbal himself, 1:ix (written by others based on Amuli); Yarshater, “Ibn-i Isfandiyār,” 810b; Madelung, “Āmolī,” 121.

104. Samadi, , “Awliya Allah Amuli,” in Nāmvārah-yi Dr. Maḥmūd Afshār, ed. Afshar, I. (Tehran, 1377/1998) 10: 5862.Google Scholar The third theoretical option, that the Continuation was written by an unknown author and extensively used by Amuli, is not a serious possibility.

105. Ibn Isfandiyar, 2: 5-17; cf. trans. Browne, 223-40 for a seamless summary of the later mss. For Jurbadhqani, see Meisami, Persian Historiography, 256-69. As his work was completed in 603/1206, this in itself does not disqualify Ibn Isfandiyar from incorporating this material into his own work (pace Iqbal's comment, 2: 5 n. 1).

106. So in Mar^ᶜashi, 121-44. For the corrected date of ca. 483/1090, see Bosworth, C. E., “On the Chronology of the Later Ziyārids of Gurgān and Ṭabaristān,” Der Islam 40 (1964): 33.CrossRefGoogle Scholar

107. Compare Mar^ᶜashi, 130-33, 136-44, with Ibn Isfandiyar, 2: 3-5, 7-18. The latter passage is based on Jurbadhqani, 228-29, 236-45, 347-57. The passage in Mar^ᶜashi, 133-6, rather than following “Ibn Isfandiyar,” 2: 5-7, is at least partly an extremely reduced version of Jurbadhqani, 59ff., esp. 80-82.

108. A Thursday. This copy was written by Muhammad Talib b. Maulana ^ᶜAbd Allah Amuli, for Mirza Muhammad Sadiq, who was wazir of the waqfs of Mazandaran, see Qazvini, Vahid, ‘^ᶜAbbāsnāmah, ed. Dehqan, I. (Arak, 1329/1951), 319.Google Scholar Thanks to Rudi Matthee for this reference.

109. Ethé 568 is India Office Persian ms. 1134; it was copied by Himmat b. Rustam Naq al-Mazandarani. British Library Or. 2778 is dated 1273/1857, and Cambridge Browne Coll. ms. 1.6.(10), 1268/1852. See also above, n. 45.

110. Husam al-Dawlah Ardashir, d. 647 H. (f. 206r); Shams al-Muluk Muhammad, d. 663 (f. 208r; cf. Amuli, 166, 167); Taj al-Dawlah Yazdagird, d. 701. (f. 208r; cf. Amuli, 168) [but see below]; Nasir al-Dawlah, d. 714 (f. 208v); Rukn al-Dawlah Kay Khusraw, d. 728 (f. 209r); Sharaf al-Muluk, d. 734 (f. 209v). These dates are followed by Mar^ᶜashi, with a couple of differences; see n. 114 below.

111. Mar^ᶜashi, 60, gives only the first line, the meaning of which he explains.

112. So ms. Add. 7633; as indicated by Browne, 260, however, only the first lines of the poem are found in ms. Ethé 568.

113. Cf. Mar^ᶜashi, 190, writing over a century later, who does not simply copy this statement, but emphasizes it by remarking that he himself had seen them and talked with them; again in less detail, 82.

114. The Continuation here is followed by Mar^ᶜashi, 190, though he has ^ᶜAla^˒ al-Dawlah reign ten years, not four months, and Taj al-Dawlah succeed him in 675 H.

115. Cf. Madelung, “Āl-e Bāvand,” 753, for this episode and for the problem of establishing the correct date.

116. For example, Amuli, 167, 169, 179, 180, 200, 203.

117. Sotudeh, introduction to Amuli, xxvii-xxviii, appears to claim this identification for himself, but in fact he repeats Kasravi's argument, cf. n. 103. Mar^ᶜashi, 57, cites Amuli's eyewitnes report by name.

118. By contrast, Mar^ᶜashi, 72, again cites Amuli directly.

119. Again in Mar^ᶜashi, 192, who retains Amuli's 13 years of discord and strife.

120. Some of the confusion and repetition in the final section of Amuli's work may be due partly to the state of the manuscript, cf. the comment of Madelung, “Āmolī,” 121.

121. Amuli, 9, 197, also 5 (taqallub).

122. The order of the final two passages in Ibn Isfandiyar, 2: 172-73, has been reversed to make more natural transition, cf. above, n. 51.

123. E.g. Amuli, 177, 191 (repetitively, concerning Kujur). He seems more fatalistic about dynastic change in this latter part than in the earlier sections, where he is mainly reworking Ibn Isfandiyar, see e.g. pp. 153, 155, 175.

124. Amuli, 200; equivalent to the Continuation, f. 214v, 1.3.

125. Not only is there no other reference to Jalal al-Dawlah murdering his brother, but Mar^ᶜashi, 82, while following Amuli's ordering of the material, basically summarizes the text of the Continuation.

126. Close reading of the two texts reveals many such variant readings. One of the more significant is the apparent omission of a line from the text of the Continuation, f. 213r, losing the name of the mustawfi who reported the size of Mas^ᶜud's army, cf. Browne, 266. Amuli, 188, is followed by Mar^ᶜashi, 76. Concerning the murder of Nasir al-Dawlah Rustam in 606/1210, in contrast with the Continuation, f. 205r, Amuli, 153, appears to say that he had acted with nā-javānmardī first, but this could be due to the earlier omission of a bā. Mar^ᶜashi, 188, offers no help.

127. The relatively straightforward language of the work argues against the obvious candidate, Ruyani, used as a source by Mar^ᶜashi.

128. William Sayers, “The beginnings and early development of Old French historiography, 1100-1274,” (PhD. dissertation, University of California Berkeley, 1966), 284; quoted by Spiegel, Romancing the Past, 220. Spiegel's chapter suggests many possible lines of enquiry into the Persian historical literature of the same period.

129. Compare Amuli, 159, with Mar^ᶜashi, 85.

130. E.g. 1: 95, 105, 192, 227. See also translations by Amuli, 47, 49.

131. Mar^ᶜashi, 60-61 (cf. Amuli, 162), see n. 111 above. Kasravi, 28-35, discusses Ibn Isfandiyar's Tabari poetry.

132. Cf. Meisami, Persian Historiography, 296-98.

133. See Bosworth, C. E., “The Persian Contribution to Islamic Historiography in the Pre-Mongol Period,” in The Persian Presence in the Islamic World, ed. Hovannisian, R. G. and Sabbagh, G. (Cambridge, 1998), 233–35.Google Scholar