^{page 472 note 1} Patmut'iwn Aluanic, ed. Šahnazarean, , Paris, 1860;Google Scholar ed. Emin, Moscow, 1860, reprinted Tiflis, 1912; translated by Patkanov, K. [Patkanean], Istoriya Agvan, St. Petersburg, 1861.Google Scholar

^{page 472 note 2} Patmut'iwn Aluanic, translated by Brosset, M., ‘ Histoire d'Aghovanie …’, Collection d'historiens arméniens, St. Petersburg, II, 1876, 191–220.Google Scholar

^{page 472 note 3} Patmut'iwn Aluanic, Valareapat, 1902 (vol. I), Tiflis,Google Scholar 1907 (vol. II).

^{page 472 note 4} patmut'iwn , Venice, 1865, 106–13, 237–43.Google Scholar The details which follow are taken from here; see also the summary by Mécérian, R.P.J. S.J., ‘Bulletin arménologique (premier eahier)’, Mélanges de l'Université Saint Joseph (Beyrouth), XXVII, 10, 1947–1948, 205–7,Google Scholar and Arutyunyan, B.M. [Yarut'iwnean, ] (ed.), Mxumapa Γowa , Erevan, 1954, pp. xiii–xviii.Google Scholar

^{page 473 note 1} J. Mécérian, op. cit., 207, assumes the existence of two Vaxtang, the one prince of , the other prince of Hat‘erk‘; but it seems clear from this passage in Kirakos, and in the colophons of Mxit‘ar Goš in the Datastanagirk‘, that these are one person.

^{page 473 note 2} 633 A; see Yovsep‘ean, G.Yišatakarank‘ , Anthelias, 1951, 507–8,Google Scholar for printed text of colophon; J. Meeerian, loc. cit., p. 212 [36], for a French translation of part of the colophon.

^{page 473 note 3} For a description of these and other MBS of the Datastanagirk‘, see J. Mécérian, loc. cit., 208–28.

^{page 474 note 1} The Armenian text is printed in Ališan, Ł, Hayapatum, patmut‘iwn Hayoc, Venice, 1901, II, pp. 276–8Google Scholar (extract 282), pp. 338–53 (extract 292).

^{page 474 note 2} This translation is made on notes taken in Bzommar in 1954; J. Mécérian, who summarizes the passage in French (loc. cit., 213), seems to have interpreted it slightly differently.

^{page 474 note 3} loc. cit. above.

^{page 474 note 4} loc. cit., 217.

^{page 474 note 5} Bazmavēp, 1942, 4–9; cf. Mécérian, loc. eit., 228–9.

^{page 474 note 6} Akōs, Beyrouth, 1944, fasc. 1–3; see Mécérian, 229, 233.

^{page 474 note 7} loc. cit., 216.

^{page 474 note 8} loc. cit., 214.

^{page 475 note 1} i.e. MS arménienNo. 177 of the Bibliothèque Nationale, Paris, dated A.D. 1231, MS NO. 1738 of San Lazzaro, Venice, dated A.D. 1238; see Mécérian, loc. cit., 218–25.

^{page 475 note 2} See Minorsky, V., Studies in Caucasian history, Cambridge, 1953, 25;Google Scholar this work is invaluable for the history of Ganjak and Ani under the Shaddādids and for the whole period with which our chronicle deals.

^{page 476 note 1} The author of the History of the Aluank‘ is more commonly known as Movsēs Kalankatuaci, although Mxit‘ar Goš's pupil Vanakan Vardapet also is said by Ališan, Ł (Hayapatum, ‘ … Hayoc , Venice, 1901, 175Google Scholar) to attribute the work to a Movsēs of the village of Dasxurēn. Ališan assumes that there were thus two Movsēs, while Manandean (Beiträge zur albanischen Geschichte, Leipzig, 1897, 22Google Scholar) dismisses the surname Kalankatuaci as an invention of later writers misled by a faulty interpretation of the text and prefers to call the author Moses of Uti. Akinean, N. (Handēs Amsorea, 1953, 1–3, p. 29)Google Scholar assumes that we have to deal with only one author, whose real name is Movsēs Dasxuranci. There is, however, no need to reject the name Kalankatuaci; by assuming, as Professor W. B. Henning first suggested to me, that this refers to the name of Movsēs’ monastery in Kalankatuk‘ while Dasxuranci refers to his native village, we can take Movses Kalankatuaci and Dasxuranci to be one and the same person. Mxit‘ar Goš himself is referred to as Ganjakeei from the place of his birth, Getkaci from the name of his monastery, and Goš (i.e. ‘beardless’) by virtue of a physical peculiarity (see Nor bafgirk‘ haykazean lezui, Venice, 1836–1837, I, 15).Google Scholar

^{page 476 note 2} See Mos. KaJ., I.6, ed. Ēmin, 7–9.

^{page 476 note 3} The greater part of this list has been taken from Mos. Kal., III.23; here, however, the latter has Movsēs in place of Karēn.

^{page 476 note 4} The MS has not as in Ališan, Hayapatum, patmut‘iwn Hayoc, II, 339.

^{page 476 note 5} This information is not in Mos. Kal. Incicean, Ł, Storagrut‘iwn hin Hayastaneayc, Venice, 1822,Google Scholar the principal work on Armenian geography, knows no or Łazarapat.

^{page 476 note 6} An anonymous chronicle of the thirteenth century (Yakobean, V.A. (ed.), Manr žamanakagrut‘yunner …, I, Erevan, 1951, p. 23Google Scholar) calculates 266 years between St. Thaddaeus (Elisaeus’ predecessor) and St. Gregory. The usually accepted date of the conversion of Armenia is c. 300, N. Adontz's date of 288 (in ‘Les vestiges d'un ancien culte en Arménie ’, Mélanges Franz Cumont, Bruxelles, 1936, 513Google Scholar) being based upon ingenious but unconvincing arguments.

^{page 476 note 7} cf. Mos. Kal., 1.14 (after P‘awstos Biwzandaci, III.6, Movsēs Xorenaci, III.3). Mos. Kal., III.23, ed. Emin, 274, mentions Grigoris immediately after Elišay, thus following the order in which they appear in his Book I, whereas Mxit‘ar Goš places Šup‘xališoy end five others before Grigoris. Kirakos Ganjakeei, VIII, ed. Venice, 1865, p. 98, has noted this discrepancy: ‘ Then [after Elišē] Šup‘hališoy became bishop; concerning the chronological position of this man, however, we are in some doubt, for he who wrote the History of the Aluank’ puts this name in the days of Vačagan the Pious, and this is supported by the laws promulgated by King Vačagan and all the bishops of Albania, and he [Mos. Kal., 1.26] writes thus: “ I, Vačagan, king of Albania, and Šup‘hališoy, archbishop of Partaw ”; and after this there is no other mention of the name in the list of bishops, but we have set it down as we found it ’ [presumably in Mxit‘ar Goš]. The canons of King Vačagan in which Šup‘hališoy had a hand were drawn up some time after the accession of Balāsh (Arm. Valaršak) in Persia (484); see Mos. Kal., 1.17, ed. Ēmin, 35. He is further mentioned at Mos. Kal., 1.23, p. 61, as Vačagan's senior archbishop (awag episkoposapet) at the time of the discovery of Grigoris’ relics; Kirakos knows this (loc. cit.), but still follows the order given in Mxit‘ar Goš.

^{page 477 note 1} i.e., the 101st year of the saeculum rwvum, the 1101st year of Rome = A.D. 348; see Dulaurier, E., Recherches sur la chronologie arménienne technique et historique, Paris, 1859, 49.Google Scholar P‘awstos Biwzandaci (III.6), Movsēs Xorenaci (III.3), and Movsēs Kalankatuaci (I.14), relating the martyrdom of Grigoris, assign no date thereto. Xorenaci specifies, however, that it was King Trdat III who appointed him to the see of Albania and that his death at the hands of Sanatruk took place upon Trdat's death (c. 330). It is after Grigoris’ martyrdom that Sanatruk is said (M.X., III.9, P.B., III.7) to attack Armenia while Xosrov II Kotak (c. 330–40) is king and Grigoris’ father Vrt‘anēs is catholicos (c. 333–42). If we can trust these details, the date 348 is too late for Grigoris’ martyrdom, although it does coincide with the probable date of the death of his younger brother Yusik, Catholicos of Armenia (343–8; see M.X., III.ll, 14), at the hands of Tiran or Tigran VII.

^{page 477 note 2} In Mos. Kal., III.23, p. 275, Zak‘area is preceded by a certain Grigor, in place of whom Kirakos (loc. cit.), like Mxit‘ar Goš, has St. Grigoris.

^{page 477 note 3} See Mos. Kal., II.4, III.23. Movsēs gives 44 years as the term of Abas’ office, Kirakos, 14; Y. , Bazmavēp, 370–4, calculates his dates as 552–96, and Basmadjian [Pasmačean], K., ‘Chronologie de l'histoire d'Arménie’, Revue de l'Orient Chrétien, XIX, 1914, 366–9,Google Scholar quoted as an appendix by de Morgan, J., Histoire dupeuple arménien, Paris, 1919,367,Google Scholar as 552–94. A fresh attempt will be made below to determine the approximate dates of the patriarchs of Albania from the evidence of the present passage and corresponding passages in Movsēs Kalankatuaci and Kirakos, but in view of the many discrepancies and the fact that the length of their terms of office is given usually in round figures, the results are rarely entirely trustworthy.

^{page 477 note 4} The MS and Ališan’s text both have patrisak-, but this is a corruption of patruak ‘address’ (see M.K., III.23, ed. Šahnazarean, vol. II, p. 72; Ēmin's edition, p. 275, has a corrupt form paruak.

^{page 477 note 5} Mos. Kal., 34 years, Kirakos, 33 years; Acarean and Basmadjian, 596–630. These sources will henceforth be referred to as M.K., Kir., A., and B. respectively, and Mxit‘ar Goš as M.G. The dates in the Armenian era given by M.G. here and below seem to refer to the date of the death of the patriarch they follow and the date of the succession of the patriarch whose name they precede; here, however, the date of 75 A = 23.6.626–22.6.627 is too early for Viroy's death, for we know that he returned from captivity in Persia upon the death of Chosroes II in A.D. 628 (M.K., II. 14.)

^{page 478 note 1} M.K., Kir 15 years; A., 630–47, B., 630–45.

^{page 478 note 2} M.K. ed S., 25, ed. E., 27 years, Kir., 25; A., 647–72, B., 645–70.

^{page 478 note 3} M.K., Kir., 12 years; A., 672–84, B., 670–82; but M.K., II.38, 39, implies that his successor Eliazar was patriarch already in 681, perhaps even 680.

^{page 478 note 4} M.K., Kir., 6; A., 684–90 (wrong), B., 682–8. M.G.'s date must be that of Nersēs’ succession.

^{page 478 note 5} M.K., Kir., 17; B., 688–700, but 688–704 would be better; cf. M.K., III.3, 7, 8.

^{page 478 note 6} M.K., Kir., 1½; years; B., 700–2; better, 704–6; cf. M.K., III.7, 12.

^{page 478 note 7} M.K. ed. S., 35, ed. Ē., 37, Kir., 35; B., 702–37; better, 706–41; cf. M.K. III.13. Again M.G.'s date refers to the year of his death.

^{page 478 note 8} M.K., Kir., 4; B., 737–41; better, 742–6.

^{page 478 note 9} M.K., Kir., 17; B., 741–58, but this is definitely inaccurate. M.K., m.23, ed. Ē., p. 276, says ‘ in his fifth year the 200th year of the Armenian era was completed (or fuWiled: Icaw)’. This wording is strange, but taking it to mean that the end of 200 A = 21.5.752 and the beginning of 201 A = 22.5.752 occurred in the course of the patriarch’s fifth year, we can assume that his first year saw the beginning of 197 A = 23.5.748 and that his dates are 748–63.

^{page 478 note 10} M.K., Kir., 4; B. (whose dates seem inaccurate down to Movsēs, below, n. 19), 758–62. Apparently 763–7.

^{page 478 note 11} M.K., Kir., 9; B., 762–71. 767–76, M.G.'s date coinciding with the death of the patriarch.

^{page 478 note 12} M.K., Kir., 1½; B., 771–3. 776–8.

^{page 478 note 13} M.K., Kir., 2; B., 773–4. 778–80.

^{page 478 note 14} M.K., Kir., 2; B., 774–6. 780–2.

^{page 478 note 15} M.K., ½, Kir., 3; B. 776–76. 782.

^{page 478 note 16} M.K., Kir., 4; B., 777–81. Apparently 782–6, M.G.'s date again coinciding with that of the patriarch's death.

^{page 478 note 17} M.K. ed. S., Kir., 11½, M.K. ed. Ē., 11; B., 782–94. 786–97.

^{page 478 note 18} M.K., Kir., 25; B., 794–819. 797–822.

^{page 478 note 19} M.K., Kir., ½; B., 820–20. 822.

^{page 479 note 1} M.K., Kir., 28; B., 820–48. 822–50.

^{page 479 note 2} M.K., Kir., 25; B., 848–73. M.K. ed. Ē., p. 277 (ed. S., om.), informs us that the 300th year of the Armenian era was completed in his third year, which therefore saw the beginning of 301 A = 27.4.852, his first year being 849/50, which agrees with the calculated date of the death of his predecessor. If he held office for 25 years, then his dates were 850–75; if for 22 years, 850–72. One cannot reconcile M.G.'s date of 327 A = 878/9 as the date of his death with M.K.'s information or our own calculations hitherto, but since his successor is said by M.G. to have held office for 15 years until 893, the date may yet be correct.

^{page 479 note 3} M.K., Kir., 17; B., 873–88. Apparently 872–87 or 875–90 (if 15 years is correct), or 872–89 or 875–92 (if 17 years is correct); it is, however, probable that he held office until 893 (see below, n. 4).

^{page 479 note 4} M.K. (Yovhan), Kir. (Yunan), 8½ years; B., 888–96. Apparently between 887–96 and 892–901. We are informed by M.K. ed. E., 277, that he profited by the captivity of Georg II, Catholicos of Armenia, to have himself consecrated patriarch of Albania. Elsewhere (III.21, ed. Ē., 268–9) M.K. dates Gēorg's captivity as commencing in 342 A (inc. 17.4.893) and indicates that he was released in 894 (four years before the year in which 1 Nawasard fell on Easter Sunday, i.e. A.D. 898); cf. Asolik, III.3, tr. Macler, 104–9, and other refs. in R. Grousset, Histoire de l'Arménie, 403. Yovhan could not therefore have become patriarch before 893, and must have held office c. 893–902.

^{page 479 note 5} M.K., 21, Kir., 25; B., 896–917. 902–23.

^{page 479 note 6} M.K., 6, Kir., ‘5 years, the year 378 of the Armenian era ’; B. 917–23. M.G.'s figure of 35 is fantastic, and since Kir. seems to have borrowed his information directly from M.G., it is probably a mistake for 5. If M.K.'s figure of 6 years is taken to be correct, the year 378 fits perfectly: 923–9.

^{page 479 note 7} M.K. ed. S., Kir., 18, M.K. ed. Ē., 19; B., 923–41. M.G.'s figure of 25 is excessive. If 18 is correct, then 929–17; if 19, then 929–18, which agrees better with the dates of his successor.

^{page 479 note 8} M.K., Kir., 14; B., 941–58. M.K. ed. Ē., 278, says that 400 A was completed in his fourth year, therefore 401 A (inc. 2.4.952) began in his fourth year, and 398 A (inc. 3.4.949) in his first year. M.G.'s figure of 10 years seems too short; if one accepts the figure 14, then 948–62.

^{page 479 note 9} M.K. lists before this one another Dawit‘ (VI, B., 961–8) holding office for seven years, apparently, therefore, 962–9. Dawit‘ VII (M.K., Kir., 6 years; B., 968–74) was, however, according to M.K., consecrated by Anania, Catholicos of Armenia, who died in 414 A (inc. 30.3.965) according to Asolik, II.8, or 415 A (inc. 30.3.966) according to Matt’ēos , 1.20. A.D. 967 is thus the latest possible date for Dawit‘ VII's succession and the death of Dawit‘ VI. Therefore, Dawit‘ VI from 962 to 967, Dawit‘ VII from 967 to 973.

^{page 479 note 10} M.K. (ed. S., Penelas, ed. Ē., Petros), 18 years, Kir., 16; B., 974–90. Taking M.K.'s figure of 18 as correct, 973–91.

^{page 479 note 11} M.K., 6 years; B., 990–6. 991–7. M.K.'s list of Albanian patriarchs in III.23 comes to an end with Movsēs. Kir. omits this patriarch and connects the abbey of P‘arisos with Petros.

^{page 480 note 1} Kir. also includes Markos and the following four patriarchs in his list without indication of the length of their rule. B. gives the following dates: Markos I, 996– ? Yovsep‘ III, 1038; Markos II, ?–1077; Step‘annos I, 1077–1103; Yovhannēs V, 1103–30.

^{page 480 note 2} Kir., ‘1½ years; he died in the days of his youth ’; B., 1130–2. If the see of Albania remained vacant for 8 years after the death of Step’annos until 1139 (see below, p. 481, n. 8), he must have died in 1131, having succeeded in 1129.

^{page 480 note 3} i.e. from 1131 to 11.6.1139 (see below, p. 481, n. 8).

^{page 480 note 4} In his account of these events, Kirakos, ed. Venice, 1865, 101, wrongly states the period during which Albania remained without a patriarch as 25 years, but nevertheless gives details not in M.G. He mentions Sahak as bishop of Karin, and mentions a Grigor Vardapet, son of T‘ok‘aker (see also ibid., 66), as one of the signatories of the letter sent to the Catholicos of Armenia; he gives the new patriarch’s original name as Gagik and says that he was renamed Grigores upon his election ‘after the name of the Catholicos of Armenia’, i.e. Grigor III (B., 1113–66). He says that 12 other bishops were consecrated at the same time, whereas M.G., whose account is valuable for the generally more detailed description it gives of the election, mentions only two.

^{page 480 note 5} i.e. Dawit‘, son of Alawik, of Ganjak, author of an important Book of canons, edited recently by A. Abrahamean, ‘Dawit‘ Alvkay ordu kanonner’, Sept. 1952–March 1953; cf. J. Mácárian, op. cit., 184–201, and Kirakos, 66.

^{page 480 note 6} Abas and Dawit‘, but not Step‘annos (mentioned below, p. 481, n. 2, as a bishop), are named as the sons of Kiwrikē (I, 1048–c. 1090; cf. Manandean, Y., K‘nnakan tesut‘yun hay žolovrdi patmut ‘yan, III, Erevan, 1952, p. 74),Google Scholar king of Loṙi, by Kirakos, p. 72: ‘And Kiwrikē Bagratuni, who was from the town of Loṙi, having opposed the Georgians all his life, kept his fatherland intact. But after his death his sons Dawit‘ and Abas were deceived by the Georgians and rose and went from the house of their fathers to the Persians, and received from them as a heritage Tawuš and Macnaberd and other places; then, after some days, the Persians took back Tawuš, and they dwelt in Macnaberd’. Cf. also Vardan, LXI, p. 106: ‘ This Kiwrikē was the son of Dawit‘, the son of Gurgen, the son of Ašot‘ Olormac [king of Armenia 953–77]; his [Kiwrikē's] father, who lies buried in Sanahin, built and 12 other fortresses. His [Dawif‘’s] grandsons Abas and Dawit‘, oppressed by the Georgians, go to the lords of an (Arrān) and take each a fortress and live in tribulation …’. Cf. Movsesian, L., ‘Histoire des rois Kurikian de Loṙi’, Revue de-s Ētudes Arméniennes, VII, 2,1927, 235;Google Scholar R. Grousset, Histoire de l'Arménie, 521.

^{page 481 note 1} In pr. Uti; see Hübschmann, H., ‘Die altarmenischen Ortsnamen’, Indogermanische Forschungen, XVI, 1904, 353;Google ScholarInčičean, Ł, Storagrut‘iwn hin Hayastaneayc, Venice, 1822, 349–51.Google Scholar Tawuš is called ‘the canton of the Sewordik‘ ’ by Vardan, LII, p. 92. Situated to the south of the Hasan-su according to Ł. Ališan (ref. apud Hübschmann, loc. cit.).

^{page 481 note 2} See above p. 480, n. 6.

^{page 481 note 3} See above p. 480, n. 5.

^{page 481 note 4} The fact that Kolt‘ and Šamk‘or formed one diocese confirms Hübschmann's supposition (‘Ortsnamen ’, 351) that Kolt‘ was the northernmost canton of pr. Arcax bordering on the canton of Šakašēn in Uti (‘Ortsnamen’, 352) in which lay Šamk‘or and Ganjak.

^{page 481 note 5} Šamiramajor (Semiramis Valley)—the form may be added to Hübschmann, ‘Ortsnamen’, 458–is possibly connected with the Berd Šamiramay (Semiramis Fortress) mentioned by Vardan, LVI, p. 100, in the canton of P‘arisos in Arcax (Hübschmann, ibid.).

^{page 481 note 6} gen. Kabroy. Not known; ? hardly Karbi in pr. Ayrarat (Mecobeci, T‘ovma, Patmut‘iwn Lank Tamuray …, Paris, 1860, 18–19:Google ScholarArarat—Karbi—Kotayk‘—Bǰni Yakobean, V.A., Manr žamanakaqrut‘yunner [Minor chronicles], Erevan, 1956, II, 534:Google ScholarT‘avrež—Naxčuan— Karbi— Bǰni——Tiflis; idem, I, 119: Karbi—T‘awrež—Sult‘ania). Inčičean, loc. cit., 528, mentions a river Karbi in Siwnik‘.

^{page 481 note 7} A monastery in the town of Miap‘or (Kirakos, 90), between Arcax-Gardman and L. Sevan (Hübschmann, ‘Ortsnamen’, 453; Inčičean, loc. cit., 528).

^{page 481 note 8} In A.D. 1139 Whit-Sunday fell on 11 June, i.e. 27 Trē 588 A.

^{page 481 note 9} 30 July 1139.

^{page 481 note 10} 17 September 1139.

^{page 481 note 11} Saturday, 30 September 1139. This earthquake is also described by Kirakos (101) who mentions that Grigor Vardapet was killed therein. Samuēl Aneci (ed. Venice, 1893, 132) gives 17 Areg, i.e. Friday, 29 September. Ibn al-Athīr, ed. Tornberg, XI.51, gives the year as 534 H (inc. 28 August 1139) and the number of killed as 230,000, including two sons of Karā Sonkur; ‘ al-dīn al-Isfahānī, ed. Houtsma (Recueil de textes relatifs à l'histoire des Seldjoucides. II, Leide, 1889, p. 190Google Scholar) gives the date (wrongly) as 533 H (inc. 8 September 1138) and the number of casualties as 300,000.

^{page 482 note 1} Job ix, 6.

^{page 482 note 2} Ps. ciii, 32 (A.V. Ps. civ, 32).

^{page 482 note 3} Said to be the same as K‘ustip‘ in pr. Arcax; see Inčičean, 309–10, Hübschmann, p. 350, n. 2, M.K., III.23, ed. ġ, 278, Vrac patmut‘ium, Venice, 1884, 32,Google Scholar Vardan, LVI, p. 100, Kirakos, 101. M. Barxutareanc, Patmut‘iwn Aluanic, I, mentions P‘arisos as the name of a fortress (p. 155), a monastery (p. 193), a town (p. 178), and a canton in Arcax (p. 182), and places the town on his map to the north-east of L. Sevan at 40° 28 N., 63° 25 E. (= 40° 28 N., 45° 46 E. on Lynch's map, just south of Karabulakh), therefore on a route between , Kambičan, and Iora as , loc. cit., indicates.

^{page 482 note 4} Late name of part of pr. Arcax, forming at this time a smaindependent Armenian principality; the earliest references to Xačēn are of the tenth century; see Hübschmann, 267, 349, Inčičean, 304.

^{page 482 note 5} Ganjak, in canton of Aršakašēn, at this time the capital of the Seljuk province of Arrān. It is said by Mos. Kal. III.20, ed. Ēmin, 265, to have been built by ‘Xazr Patgos’, i.e. b. Yazīd b. Mazyad, who died in 842/3; a later, though not a more correct date, is given in a (probably fourteenth century) list of dates: ‘In the year 295 A (A.D. 846/7) Ganjak was built’ (V. A. Yakobean, [Minor chronicles], I, 391). See V. V. Barthold, Encyclopaedia of Islam, article ‘ ‘, ed. 1936; Inčičean, 310–12, Hübschmann, 350, 416.

^{page 482 note 6} Demetrē I (1125–55/6), son of David IV the Builder (1089–1125). Kirakos, 102, adds that the Georgian king carried away the gates of the city to Georgia.

^{page 482 note 7} See below, p.483, n. 1.

^{page 483 note 1} On the atabek Karā-Sonkur, emir of and Arrān, see Barthold, loc. cit. According to al-Isfahānī, p. 190, he died in Ardabil in 535 H (1140/1). Professor Henning suggests that tiezerakal points to his rank of , misunderstood by the Armenian as .

^{page 483 note 2} This might possibly be a corruption of an Armenian form of Kutluk.

^{page 483 note 3} See below, p. 485, n. 1.

^{page 483 note 4} On the career of al-, see Ibn al-Athīr, XI.52, 77; al-Isfahānī, 161, 165, 175, 183, 191–3, 199–204, 213.

^{page 483 note 5} cf. Samuel Aneci, 132: ‘ In 592 A (inc. 14.2.1143) took the fortress of Xačēn by trickery, having gained the confidence of the princes by means of an oath, and he destroyed the land by famine and the sword and enslavement, whereby priests and holy gospels and monks were trampled underfoot by the infidels’.

^{page 484 note 1} The account of Samuēl Aneci, p. 133, differs: ‘In 594 A (1145/6) King Demetrē killed the great Iwanē. And in the same year amirapet came and besieged Tayuš [Tawuš] for 40 days and finally took it, taking away King Abas without harming him ’.

^{page 484 note 2} cf. Kirakos, 72.

^{page 484 note 3} Mos. Kal. 11.50, ed. Ēmin, 223, mentions this Tanjik‘ in Arcax; cf. Hübschmann, 473, to which list one may also add Tanjijor in Tarōn (colophon of 1104 in G. Yovsep‘ean, Yišatalcaranner …, 290) and Tanjap‘arax in Siwnik‘ (colophon of 1393 in Xačikean, , [Colophons of fourteenth century Armenian MSS], Erevan, 1950, 606;Google Scholar see Inčičean, 300.

^{page 484 note 4} ‘ At the foot of the Mraw mountain near the r. Terter ’ (Ališan, , Hayapatum, patmut‘iwn Hayoc, II, p. 345, n. 1);Google Scholar see Barxutareanc, Patmut‘iwn Aluanic, 56.

^{page 484 note 5} Al-Isfahānā, 204, gives the date of death at as I541 H (9 October-7 November 1146).

^{page 484 note 6} P‘axradin is the amir Fakhr al-dîn ‘Abd al-Rahmān b ( ?) who is said by al-sfahānī (pp. 213–14) to attempt to take possession of Armenia and Arrān on the death of . He is found by the sultan Mas‘ūd to be plotting against him with Būzābeh (M.G. Bzap‘ay) and ‘Abbās, wālī of Raiy, and the sultan sends him to Arrān, commissioning Khāss-bek b. Balankirī (M.G. Xazbek) and Ildegiz (M.G. Eltkuz) to assassinate him; see Hamdu'llāh Mustawfī-1-Kazwīnī, ed. Browne, E.G., Gibb Memorial, XVI, I, p. 467.Google Scholar Our author's mention of his unpopularity in Arrān is an interesting detail. Al-Isfahānī also indicates that Fakhr al-dīn was preparing to attack Georgia when he was assassinated (p. 217) and that the assassination took place in 541 H (8 April 1147), which does fall in the Armenian year 596 (inc. 13 February 1147). The other indications of the date do not synchronize. Ališan, loc. cit., p. 345, n. 4, rightly takes the year of Rome to refer to the 367th year of the second 500–year period of the second millennium (cf. Dulaurier, Chronologie …, 49), i.e. the Varronian year 1867, but this corresponds to A.D. 1114. The 536th year of the Persians, i.e. 536 H, is A.D. 1141/2.

^{page 485 note 1} The rulers referred to are Manuel (not Michael) I Comnenus (1143–80), son of John II or Kalo-Ioannes, grandson of Alexius I; Demetre I (1125–55), son of David IV, grandson of Giorgi I I; Mas‘ūd (1134–52), son of Muhammad I, grandson of I. Mas‘ūd was the fourth Seljuk sultan of iraq, the sultan of sultans being his uncle (1118–56), but the latter's influence hardly exceeded the limits of Khorassan (Poole, Lane, The Mohammadan dynasties, tr. Barthold, , p. 125, n. 2).Google Scholar It was Mas‘ūd who held sway over Arrān, .and Armenia and who, shortly after his accession, appointed his nephew Dā'ūd governor over these lands with his capital at Tabriz (Kazwīnī, pp. 464–5).

^{page 485 note 2} Būzābeh, i.e. Buz-apa (Henning), lord of Fārs and Khūzistān (Ibn al-Athīr, XI.78), seized Isfahān and was defeated at Hamadan in 541 H (1146/7); cf. al-Isfahānī, 219–20, Kazwīnī, 467.3

^{page 485 note 3} represents the Turkish name written in Arab sources, e.g. Ibn al-Athīr and al-Isfahānī, as or Is this one perhaps the amīr for ٭ who was removed in 549 H (1154/5) from Ardabīl by Ildegiz and Nusrat al-dīn Arslān-ābeh (i.e. Arslan-apa), ‘the two lords of (al-Isfahānī, 242–3) ?

^{page 485 note 4} Rawādī, a cousin of Khāss-bek b. Balankirī, is mentioned briefly by al-Isfahānī, 232, as a ruler of Arrān deposed by Ildegiz (see below, p. 487, n. 5).

^{page 486 note 1} i.e. the daughter of Gurbuiay.

^{page 486 note 2} The Turkman Azadin (‘Izz al-dīn) is otherwise unknown. Is Herk‘an Berd a ‘ fortress of Herk‘ (Heret‘i, Hert‘) ’ ? The annotator of Kirakos, 90, suggests that the author's is Heret‘ in Albania and should be written Herk‘. Inčičean, 525, says he does not know whether Herg was in Armenia or Albania; but a colophon of 1417 reads: ‘ … a native of the gawaf Herg in the province of Afan, from the town of Bardawa (Partaw), from the village called Paris (nom. Parik‘ ?)’ (Xačikean, , [Colophons of fifteenth century Armenian MSS], I, Erevan, 1955, 205).Google Scholar

^{page 486 note 3} b. Mahmūd, son of Mahmūd II of Iraq (1118–31) and brother of Muhammad II of Iraq (1153–9), is mentioned in connexion with the revolt of Sulaimān , Ildegiz, and Arslānābeh against Muhammad II in 549 H (1154/5) and is said to have been at the time with the atabek Āyāz in (al-Isfahānī, 243–4; cf. Kazwīnī, 469).

^{page 486 note 4} cf. Kirakos, 146: ‘ in the monastery called Ganjasar, opposite Xoxana-berd ’; a colophon of A.D. 1232: ‘the monastery called Hawaptuk … in the province of Arcax, opposite the fortress called Xawxanberd ’ (Yovsep’ean, loo. cit., 882); Ališan, loc. cit., p. 347, n. 3, says it is ‘now called also T‘arxan-qale, on top of a hill near Ganjasar monastery, at the foot of which are the ruins of a village and large houses called Darpasner [“ Palaces ” ] ’.

^{page 486 note 5} This is the translation of the text as it stands (… Gurbulayin, glxawori t‘urk‘mani mioy, zor kočēr Tawlan i hnazandut‘iwn sultanin …). It makes better sense, however, if the text is amended to … t‘urk‘mani mioy, or kočēr zTawlan … ‘ a Turkman, who summoned Tawlan ’, since Gurbuiay is the loyalist, Tawlan the rebel.

^{page 486 note 6} The Tutak‘eank‘, men or descendants of Tutak‘, seem otherwise unknown.

^{page 486 note 7} cf. Incifiean, loc. cit., 314.

^{page 487 note 1} i.e. the present Gegamskii xrebet, west of L. Sevan; of. Inčičean, 264.

^{page 487 note 2} This Sewunč ( ‘joy’) and his master seem otherwise unknown.

^{page 487 note 3} Fortress in north-east of Areax, beside the Shamkhor chai; cf. Brosset, M., Histoire de la Siounie, I, 210; Kirakos,Google Scholar 88, 129; Inčičean, 536; Yovhannisik Careci (sixteenth century): … they scattered over the plains and mountains of Ganča, Bartay (Barda’a), (Čaraberd), Varanda, and Dizak, from the river called Dizak in Čarek‘ down to the Araxes ’ (Yakobean, [Minor chronicles], II, 242). See also Minorsky, V., Studies in Caucasian history, Cambridge, 1953, 24,Google Scholar 33.

^{page 487 note 4} al-dīn Ildegiz ; for his biography, see Encyclopaedia of Islam, article ‘Ildegiz ’. The events here described took place between 1147/8 and 1155 when Ildegiz was more than the mere ‘ lord of the town and canton of In 543 H (1148/9) he is referred to by Ibn al-Athīr (XI.87) as the ruler of , Arrān, and Kaysar, and in 549 H (1154/5) al-Isfahānī (243) calls him one of the two rulers of . There is no mention of his having killed who, according to al-Isfahānī (ibid.), was reconciled with, and then put under arrest by, his brother Muhammad II in 549 H.

^{page 487 note 5} cf. al-Isfahānī, 232, where Ildegiz is said to return to and Arrān and take it from Rawādī.

^{page 487 note 6} Ildegiz’s wife was the widow of I, sultan of Iraq (1132–3), and the mother of Arslān shāh (1161–77). It was not upon the death of Mas‘ūd II that Ildegiz made his stepson sultan of Iraq, but upon the deposition of Sulaimān shāh (1159–61). Samuēl Aneci, 133, shows greater confusion, and says that Ildegiz killed his son to become sultan in his stead. Muhammad II (1153–60) had Khāss-bek Arslān b. Balankirī assassinated in 548 H (A.D. 1153/4), but his agent was not Ildegiz (cf. al-Isfahānī, 230–1, Ibn al-Athīr, ‘ Histoire des atabecs de Mosul’, Recueil des historiens des croisades. Historiens orientaux, II, 1, p. 187Google Scholar). It is well known that Ildegiz became ‘the virtual ruler of the Seldjuk empire ’ (Enc. Islam). He is said, along with Saltuk and the Shah i Armen, by Vardan (LXXII, p. 124) to have been friendly towards Christians and to have brought prosperity to the country, though Manandean, Y. (K‘nnakan tesut‘yun hay žolovrdi patmut‘yan, Erevan, 1952, part III, p. 106Google Scholar) treats this statement with some reserve, saying that Xosrov of Ganjak was martyred at his command in 1167, and Yovsēp’ of Dwin in 1170.

^{page 488 note 1} Dawit‘ III reigned for less than a year between 1155/6 and 1156/7 and was succeeded by Giorgi III (1156/7–84). See Vardan (LXXIV, p. 126), who does not mention Vasak: ‘ And Dawit‘ died after one month; some say he was betrayed by Smbat and Iwanē Ōrbēleank‘ for having put T‘urkaš in their place, they having come to an agreement with Gēorg, brother of Dawit‘, whereby he would make them generals ’.

^{page 488 note 2} ‘Izz al-dīn Saltuk b. ‘Alī b. Abi '1-Kāsim of Erzerum was taken prisoner after an encounter with the Georgians in 603 A (inc. 11.2.1154) according to Samuēl Aneci, 134, in 548 H (inc. 29.3.1153) according to Ibn al-Athīr, XI.125–6, in 549 H (inc. 18.3.1154) according to al-Fāriki (in Amedroz, H.F., History of Damascus … by Ibn al-Kalānisī, Beyrouth, 1908, 328Google Scholar). Both Mxit‘ar and al-Fārikī name the Georgian king as Demetrē (1125–55/6), so that later datings at 1160 (e.g. Ibn al-Athīr, XI.185: 556 H; Mixayēl Asori, ed. Jerusalem, 1871, 441–2: 1471 Seleucid era) are wrong and probably confused with the date of Giorgi's capture of Ani. Apart from Mxit‘ar, no one mentions the role played by Vasak.

^{page 489 note 1} Mxit‘ar Goš is said by Kirakos (XI, p. 107) to have been welcomed at Erzerum by a prince K‘urd who had been expelled by the Georgian king; see above, p. 473. He and the K’urd of the present passage are surely identical (as Ališan, loc. cit., p. 351, n. 1).

^{page 489 note 2} Giorgi took Ani in 1161 and 1174, but the two phases mentioned by Mxit‘ar probably refer to (a) the first capture of Ani in 1161, and (b) the revolt of the priests and the arrival and defeat of the Shah i Armen. Vardan (LXXIV, p. 126) indicates a lapse of 50 days between the two events; cf. also Samuēl Aneci, 137. Al-Fāriki (loc. cit., 361) mentions the role of the priests of Ani in revolting against the Shaddadid Fadlūn b. (Vardan: P‘at’lun), but does not mention the subsequent revolt of the bishop BarseJ, etc., against Giorgi; it is possible that it was they who appealed to the Shah i Armen to drive off the Georgian king. The priests referred to in al-Fārikī are those mentioned here by Mxit‘ar, hardly the Shanshe Dadian of the Georgian Chronicle (as V. Minorsky, op. cit., p. 89, n. 1; see this work for the eyewitness account of the capture of Ani by the kādī Burhān al-dīn); the bishop Barsel is mentioned by Vardan (LXVI, p. 130) as bishop of Ani in 623 A (inc. 6.2.1174). Ibn al-Athīr (XI.184) gives the date of the capture of Ani as Sha‘bān 556 H (26.7–23.8.1161), Vardan and Samuēl Aneci (loc. cit.) as 610 A (inc. 9.2.1161). As to the particular form of punishment meted out by Giorgi, Ibn al-Athīr (XI. 188–9) says it was inflicted by the Georgian king upon the Muslim women after his capture of Dwin in 557 H (inc. 21.12.1161).

^{page 489 note 3} i.e. Sadun (Vardan, p. 126), Sa‘dun (al-Fārikī, loc. cit., 361). He was suspected of plotting rebellion because of his fortification of Ani and divested of his office; he went over to Ildegiz, but was captured by the erist‘av of Šak‘i and put to death, 4,000 ‘Persians ’ dying with him. Sargis Zak‘area was appointed governor of Tiflis in his place (Vardan, ibid.).

^{page 489 note 4} i.e. Fadlūn, more likely a Kurd; see above, n. 2, and Minorsky, Studies …, 2, 5, 88.

^{page 489 note 5} The Georgian general was Iwanē Ōrbēli (Brosset, Histoire de la Oeorgie, I, 1, p. 387), called Yvān by Burhān al-dīn (Minorsky, op. cit., 89). The enemy raids refer to those of Arslān whom Ildegiz, after his defeat at Gag (below, p. 490, n. 1), urged to besiege Ani; the attacks lasted for four years, until Giorgi surrendered the town (Vardan, LXXIV, p. 127). Al-Fārikī (loc. cit., 364) says the Georgians suffered a heavy defeat in 559 (inc. 30.11.1163) and that ‘ towards the end of the year Ildegiz gave Ani to the amīr brother of the amirs and Fadlūn, sons of , who had been its rulers ’; see Minorsky, loc. cit., 96.

^{page 490 note 1} On the fortress of Gag (Kāk), see Vardan, pp. 119, 127, 145, Inčičean, 513, Minorsky, loc. cit., 99–100. Ildegiz's attack on Gag took place after Giorgi's attack on Dwin (see Vardan, p. 127, al-Kalānisī, 361). Samuēl Aneci, who mentions the serpents, gives the date as 612 A (inc. 9.2.1163). There is an account in Ibn al-Athīr, XI.189, of the treachery of a certain Georgian who helped Ildegiz to victory during an encounter in Safar 558 (9.1.–6.2.1163), but no name is given. Vardan, LXXVI, pp. 129–30, relates how, after Giorgi's capture of Ani in 1174, his spasalar Iwanē, threatened by the sultan, intended to surrender the town, but was prevented from so doing by the inhabitants; possibly Mxit‘ar has confused the two events. Xalagučean may be a corruption of ٭Mxargrjean = Georgian Mxargrdzdi; on the Georgian Longimani, see Minorsky, loc. cit., 102–3.

^{page 490 note 2} Both in the canton Aragacot in pr. Ayrarat; see Inčičean, 508, 441–42, Samuēl Aneci, 138.

^{page 490 note 3} Ibn al-Athīr, dates the attack on Dwin at 557 H (16.7–13.8.1162), adding that the Georgians took many prisoners, stripped the women naked, and burned the principal mosque and others. Al-Fārikī (loc. cit., 361) gives the same month and year and says that they demolished the minaret which Kurtī b. al-Ahdab had built with the skulls of Georgians killed during a previous encounter, i.e. at Garni in 579 A (inc. 17.2.1130); see Vardan, LXXI, p. 123; Minorsky, Studies …, 85–6. Ališan, loc. cit., 353, reads the last phrase as ew ztunn … Ormzdi …, but the manuscript is here illegible, with the naked eye, at least, and a ‘house of Orcnizd’ is certainly not what one expects.