page 6 note 1 St. Martin, Richter, Patkanian, Rawlinson, Mordtmann, Gutechmid, Justi, and some others also have in different degrees contributed to the progress made in this direction.

page 6 note 2 See his Excursus in the Geschichte der Perser und Araber zur Zeit der Sasaniden aus der arabischen Ohronik des Tabari, pp. 400–436, Leyden, 1879Google Scholar. This chapter contributed to a substantial advancement of the knowledge regarding the chronology.

page 7 note 1 This prince appears to have held the same post (the governorship of Kirmān) a very long time, for, as Dr. W. Henning informs me, he is mentioned in the inscription of his brother Shāpūlr Ka'be i Zardosht under the name and title of Ardashīr Kirmānshāh, about A. D. 262.

page 7 note 2 Reckoned from the conquest of Babylonia by Alexander the Great, i.e. 331 B. C.

page 8 note 1 The same correspondence occurs also in another passage, of doubtful authenticity, of the Acta relating to the martyrdom of Shāpūr Beit Niqtor, where again the two dates (31st year of King Shāpūr's reign and 117th of the Persian rule) are given up corresponding with each other, though in this case a third one, the 296th year from the crucifixion of Christ, is erroneously added (see Braun, O., Ausgewählte Akten persischer Martyrer, Munich, 1915, p. 8Google Scholar).

page 8 note 2 Die Chronik von Arbela, German by E., Sachau, Abhandl. d. kōnigl. preuss. Akad. d. Wiss., Berlin, 1915, No. 6Google Scholar.

page 8 note 3 The usage of the registration of the horoscopes of the reign of each of the Sasanian kings, i.e. those of the time of their accession to the throne, or of their coronation, as well as those of the first vernal equinox preceding (or succeeding ?) the accession, is confirmed by a table containing the horoscope, or rather the “ascendant”, of the reign of Ardashīr and those of most of his successors down to Yazdegird, the last Sasanian king, which I found in an Arabic book composed by the famous astronomer of the tenth century Ạḥmad ibn Muḥammad ibn 'Abd al-Jalīl as-Sajzī (Paris, fonds arabe, 2581 ff., 6a and 7b), as well as by the horoscope of the accession of Khosrau I, found in Kitāb al-masā'il of Qaṣrānī of the ninth century (see the text and the discussion of it in BSOS, ix, part 1, 1937, pp. 128–130Google Scholar).

page 9 note 1 I have discussed this question fully in ZDMG, vol. 91, part 3, pp. 675–7Google Scholar.

page 9 note 2 La Chronographie de Mar Élie Bar Šinaya ⃛ trad, par , L. J., Paris, 1910, p. 60Google Scholar.

page 9 note 3 Ibid., p. 61.

page 9 note 4 The record itself is not strictly historical, as the first historic and notable bishop (or archbishop) of Seleucia was Ṗapa (about the end of the third century). Šakhlūphā of Babylon is mentioned in the Chronicle of Arbela as the ninth bishop of Arbela, but the time of his succeeding the bishop Ḥairān is put in the reign of Ardashīr. Elias may have counted from the coronation of Shāpūr, in which case the calculation is correct.

page 9 note 5 Chronique de Michel le Syrien, éd. et trad. par Chabot, J. B., Paris, 1899, vol. 1 of the translation, p. 188Google Scholar. The same author also gives (ibid., p. 265) the interval between the 19th year of Shāpūr II and the first of Ardashīr as 98 years, which agrees with his first statement (A. D. 230–327).

page 10 note 1 See BSOS, x, part 1, pp. 125–6Google Scholar, where I have folly discussed that era.

page 10 note 2 Hoffmann, G., Autzüge aus syrischen Akten persischer Martyrer, Abh. für die Kunde des Morgenlandes, publ. by the Deutsche morgenländische Gesellschaft, vol. vii, Leipzig, 1880, p. 79Google Scholar.

page 10 note 3 The Sel. year 799 began on the 1st October, 487, and ended 30th September, A. D. 488.

page 10 note 4 323 B. C. (the date of the death of Alexander) + A. D. 209 = 532.

page 10 note 5 Zosimus has nearly the same story.

page 11 note 1 I have followed the English translation of Magee, D. of the Scriptores Histories Augustæ, the Loeb Classical Library, vol. 2, p. 381Google Scholar.

page 11 note 2 Revised ed. by Willy Liebenam, Bonn, 1909; see also Clinton, (H. F.), Fasti Romani, voL 1, pp. 256–260Google Scholar.

page 11 note 3 If the words of the Scriptores should be taken strictly and literally and mean really this sequence of events.

page 11 note 4 This is concluded from a Latin inscription written during the second consulship of Gordian and the 4th year of his reign, where the name of his wife occurs (see Acta Fratrum Arvalium, ed. Guil., Henzen, Berlin, 1874, pp. ccxxiv-v)Google Scholar.

page 11 note 5 E.g. Ibn Khaldun (ed. Cairo, A. H. 1284), vol. ii, p. 169.

page 11 note 6 This association or co-regency is a conjecture by some authors and not more than a possibility. It is, as will be shown later, by no means a certainty. Therefore we refer to it with the word “perhaps”.

page 12 note 1 Gord. C. 27.

page 12 note 2 Gutschmidyy, , Untersuehungen über die Geschichte des Kōnigsreichs Osroene, in Mémoires de l'Académie de St. Petersbourg, tome 35, 1887, p. 44Google Scholar. I am not competent to express any opinion as to the possibility of such emendation or as to the correctness of the sentence and its interpretation. However, some learned and competent scholars to whose judgment I referred this question pronounce against its possibility.

page 12 note 3 We have proof of Philip's succession to Timesitheus as prefect already in the Sel. year 554, which ended on 30th September, 243. (see Cambridge Ancient History, vol. xii, p. 131Google Scholar). Lehmann, (Kaiser Gordian III, Berlin, 1911) puts the death of the prefect at the end of 243Google Scholar, and Parker, (A History of the Roman World, London, 1935) puts it in the winter 243–4Google Scholar; but the reasons which they give for these datings do not seem to be convincing, and it is just possible that they have advanced the whole of the events about one year.

page 12 note 4 Loc. cit.

page 12 note 5 See Pauly's Realencyclopædie, article Resaina, where the battle is put in the winter of 242–3 (on the authority of Ammianus Marcellinus).

page 12 note 6 Even if the prefect died a few months later this does not interfere with the above conclusion, for the Latin author states that after the battle (or battles) with Shāpūr (possibly first around Antioch and then near Ra's al-'ain) the latter evacuated “the contested cities” (Ḥarrān and Nisibis) and returned to his country. Gordian did not pursue the Persians but quartered his army near the coast, spending some time for the administration of the recovered provinces and making plans for the march, probably in the next spring (243), along the Euphrates against the Persian capital. Thus Shāpūr was not disturbed for some time in Ctesiphon by the imminent resumption of the war.

page 13 note 1 He speaks of the succession as having occurred after Gordian had married and his rule and power were well established. This corresponds approximately to the period between the autumn of 241 and the summer of 242.

page 13 note 2 Kitāb al-Fihrist by an-Nadīm, p. 328.

page 13 note 3 In 240, 242, and 243 the sun was in about the 20th, 28th, and 17th degrees of the Fish respectively. In 242 the equinox occurred in Babylonia on the 21st March a few minutes after midday, whereas the Jewish Nīsīn began on the 19th March (a Saturday).

page 13 note 4 The oldest instance of the use of the Julian calendar by the Eastern Syrian Christians known to me is that which occurs in the old Syrian martyrology published by W. Wright from a manuscript dated A. D. 412. The first day of April was in none of the years 240–43 a Sunday.

page 13 note 5 Although there may have been such a calendar in local use in some parts of Mesopotamia or Syria—the calendar of the Ṣabians of Ḥarrān as described by an-Nadīm and Bīrūnī is one. example—we have no information of the use in Babylonia of any calendar other than the Babylonian (with lunar Seleucidan era) by the pagan and perhaps by the early Christian population and the Jewish calendar in most cases by the latter (e.g. in the early acts of martyrs). The “lunar year beginning with the vernal equinox” [sic] supposed by Nöldeke (op. cit., 409) to have been “used in the Manichæan books” is nothing but the lunar Seleucidan (Babylonian) year, which began almost invariably in the spring (but not with the equinox); and the lunar year used in the acts of martyrs (Moesinger, , Monumenta Syr, ii, 68Google Scholar) with regard to the dates of the martyrdom of the Christians under Yazdegird II (24, 25, and 26 Abh. and 26 Ēlūl) is simply a Jewish year and not the “Christian” lunar year based on the conventional calculation of Easter and constructed by assuming (though often contrary to reality) that this feast-day corresponded to 14 Nīsān, as Gutschmid supposed it to be (according to Nöldeke, ibid., p. 424). I am unable to understand how the 24th day of Abh of such a year in A. D. 446, when Easter was on 31st March, could correspond to 2nd August, although the 24th day of the Jewish Abh corresponded really to 2nd August and was a Friday (according to the tables of the Jewish calendar of Schram). I myself had also expressed the same opinion as Nöldeke in my article “Various Eras and Calendars in the Countries of Islam”, BSOS, x, 1, p. 126Google Scholar. But I do not wish to maintain it any longer. Professor Neugebauer has also expressed (in a letter to me dated 29th May, 1939) his extreme doubt as to the existence of such a lunar year, and he suggested that mistakes in the sources must have led to such a supposition.

page 14 note 1 See Neugebauer, , Hilfstafeln zur technischen Chronologie, p. 46Google Scholar. It fell in 241 on the 1st April (a Thursday) and in 242 on 20th April (a Wednesday).

page 14 note 2 Bīrūnī, (Chronology, 208)Google Scholar speaks of the great number of the followers of Mānī “when he appeared”.

page 14 note 3 Even if the mention of the Egyptian month Pharmuthi in Kephalaia (ed. G., Schmidt, p. 14)Google Scholar should prove to be an allusion to the date of the prophet's birth, which then would have occurred some time between 7th and 25th April, A. D. 216, he could not have yet completed 24 Babyl. years before or on the first Nīsān 551 (lunar Sel. era) unless one counted with the Julian years and months, which are not likely to have been the basis of the calculation. If Mānī was born in the latter part of the said Seleucidan year (527), which is quite probable, his birth would have then fallen in A. D. 217, but not in the Egyptian month Pharmuthi, which in the latter year began on the same day as the Sel. year 528 began. In that case he could have completed his 24th year only in A. D. 242. It is possible that the month Pharmuthi occurring in the above mentioned passage of Kephalaia relates rather to the year of the public proclamation of the new religion by Mānī on 9th April, 243, for this date, as a matter of fact, corresponds to the 14th day of Pharmuthi.

page 15 note 1 The spring of the year 240 is also untenable on other grounds, e.g. it is too early for Shāpūr's reign, which began after his father's reign of nearly fifteen years.

page 15 note 2 Schæder has suggested (Gnomon, ix, 1933, p. 349Google Scholar) that the Arabic word arba'īn in al-Fihrist is a miswriting for another Arabic word meaning “two”. But the Arabic expression for “two years” would be rather than .

page 15 note 3 Probably Central Asia and the south-eastern regions beyond the frontiers of Iran were meant.

page 15 note 4 I think that the whole story of the first public proclamation of the mission by Mānī in al-Fihrist must be considered as being composed of two separate pieces quoted from different reports, one inserted in the middle of the other. The one which is apparently the older and most authentic consists of the narrative in the first part of page 328 (Flügel's edition) up to line 20 continued in the last four and a half lines of the same page beginning with the words “”, and the second is a composition by the author himself based on various reports of different value (some erroneous) running from line 21 to the middle of line 27 of the game page.

page 15 note 5 Mention of both of these disciples is also to be found elsewhere.

page 16 note 1 As is known, the Babylonians used to call the year during which the accession of a king to the throne occurred “the year of the accession” (Antrittsjahr) and the year beginning with the following Nīsān, the year 1 of the reign.

page 16 note 2 Schæder expresses the opinion that this Bab. custom was adopted by the Achæmenians and continued to be in use later, and that the kings were crowned on the first feast of the spring after their accession (Gnomon, ix, p. 350Google Scholar).

page 16 note 3 Elie de Nisibe, sa Chronologie, par Lamy, Bruxelles, 1888, p. 17Google Scholar; in Delapdrte's transl., p. 61, where he states that the same occurs in the Edessan Chronicle.

page 16 note 4 Bīrūnī, (Chronology, 208)Google Scholar states that the cause of Mānī was increasingly flourishing in the time of Ardashīr and Shāpūr and Hormizd.

page 17 note 1 If, however, the last Persian year before that of the coronation (22nd September, 241–21st September, 242) is the one intended, which does not seem to be very likely, then Mānī may have left for India a few months earlier. His journey to that country in the lunar Sel. year 552, although possible, is less probable.

page 17 note 2 According to Ṭabarī (i, 833), after the death of A. and the accession of Shāpūr to the throne the latter appointed his son Hormizd to the governorship of Khorāsān. This may have happened in the early years of S., say 243 or a little later, though Herzfeld (Pāikulī) puts it in 252 and thinks that up to that date Fīrūz had always been there. This interpretation is not, however, in better accord with the literal meaning of the text.

page 17 note 3 The identity of a certain Fīrūz to whom and Rāshtīn = (another follower of Mānī in Marv: see Andreas-Henning, , MUteliranische Manichaica, iii, p. 13, n. 5Google Scholar) an epistle was addressed, according to al-Fihrist, 337, 10, with the well-known Sas. prince is, I think, by no means proved.

page 17 note 4 For the calculation of the Bema feast see excursus at the end of this article.

page 18 note 1 Ayātkār i Žāmāspik, publ. by Messina, Rome, 1939, pp. 64–5 —Ya'qūbī, history i, 179—Ṭabarī, i, 821 (from the death of Ardāvān to the death of Ardashār) according to the first version—Mas'ūdī in Murūj adh-dh ahab, ii, 159—Bal'amī, Persian Ṭabarī, Brit. Mus. MS., Add. 7622, fol, 127a, Zotenberg's transl., ii, 79—Tha'ālibī, Ghurar (ed. Zotenberg), p. 486—Anonymous (Sprenger, cod. 30), Rothstein, p. 24—Google Scholaral-Balkhī, Ibn, Fārs-nāma, GMS, new series, i, p. 61Google Scholar—Ibn Khaldūn(Cairo ed. A. H. 1284, vol. 2, p. 169)—Mir Khāwand, Rawḍat aṣ-ṣafā (after the death of Ardavān)—Ginza (the famous Mandæan book), 383, Lidzbarsky's transl., p. 411.

page 18 note 2 Fārs-nāma, p. 19 (however, the word “do” may be a miswriting or a misprint for dah).

page 18 note 3 Ibn Qutaiba (ed. Wüstenfeld), p. 322—Eutychius (Oxford ed.), p. 374, Beirut ed., p. 108—Ibn al-Faqīh, BGA, V., p. 198—Muqaddasī, al-Bad' wa't-tārīkh, iii, p. 156—Ḥamza of Iṣfahān, Annals (Berlin ed.), p. 13—Bīrūnī, Chronology, p. 123, and the author of Mujrnal at-tawārīkh, JA., 3rd series, tome xi, p. 523, both the latter authors following Ḥamza in this version.

page 18 note 4 Agathias, iv, 24—Ṭabarī, loc. cit. (second version)—Ḥamza (on the authority of Mōbad), op. cit., p. 21—Bīrūnī, in his first list, op. cit., p. 121, as well as in his third list (from Ḥamza from Mōbad) and in his fourth list on the authority of Zanjānī, p. 127—Mujmal, which again here as in most other cases follows Ḥamza, loc. cit.—Abū'1-Fidā' (ed. Istanbul), p. 50—Mustawfī, Ḥamdullāh, Tārikh i guzīda, GMS, xiv, 1, p. 105Google Scholar.

page 18 note 5 Mas'ūdī, , Kitāb at-tanbīh, BGA, viii, p. 99Google Scholar.

page 18 note 6 Mas'ūdī, Murūj (second version), loc. cit.—Elias of Nisibis, transl. Delaporte, p. 35—Syncellus, ed. Niebuhr, vol. i, p. 678—Michael the Syrian, ed. and French transl. by Chabot, vol. i, p. 265—Ibn al-Jawzī (Sibṭ), Mir'-āt az-zamān, Brit. Mus. MS. Or. 4215, f. 185b.

page 18 note 7 Ḥamza (on the authority of the less reliable Mūsā ibn 'Isā al-Kisravī), op. cit., p. 17.

page 18 note 8 Mīr Khāvand, op. cit. (for the whole of the reign of A., of which twelve years fell, according to this author, before the death of Ardavān).

page 18 note 9 Fārs-nāma, p. 61 (the whole period of A.'s reign since his rising in Fārs till the end of the reign).

page 18 note 10 Açoghik, p. 57, 114, according to Patkanian, Essai d'une histoire de la dynastie Sassanide d'après les renseignements fournis par les historiens arméniens, French transl. by Evariste Prud'homme, p. 45.

page 18 note 11 Tajārib al-umam (a Persian translation of Nihāyat al-arab of early eighth century A. H. The manuscript from which I took notes is in private possession and is dated A. H. 811).

page 18 note 12 Firdausī: “Chihil sāl⁰ bar sar do māh.”

page 18 note 13 Nihāyat al-arab, Brit. Mus. MS. Add. 18,305, f. 87a.

page 18 note 14 Bal'amī MS., loc. cit. (Zotenberg's ed., ii, 75) for the whole of A.'s reign, of which only fourteen years belonged to the period following the death of Ardavān.

page 18 note 15 Ḥamdullāh Mustawfī, loc. cit. (for the whole reign before and after the death of Ardavān).

page 18 note 16 Fārs-nāma, p. 19, where it is stated that out of this period thirty years had passed in fighting against the “tribal kings” or mulūk aṭ-ṭawā'if; Ḥamza has exactly the same (from Mōbad), op. eit., p. 21, also Ghazzālū, Naṣiḥat al-mulūk, Teheran edition, Persian version, p. 44.

page 19 note 17 Mkhithar of Airivank, p. 21, according to Patkanian, op. cit.

page 19 note 18 Samuel of Ani, p, 38, according to Patkanian, op. cit.

page 19 note 19 Sebeos, p. 14, according to Patkanian, op. cit.

page 19 note 20 Ayātkār i Žāmāspīk, ed. G., Messina, Rome, 1939, pp. 64–5Google Scholar.

page 19 note 21 Açoghik, p. 114, according to Patkanian, op. cit.

page 19 note 22 Mkhithar, p. 21, according to Patkanian, op. cit.

page 19 note 23 Elias of Nisibis, loc. cit.—Muqaddasī, op. cit., iii, p. 158—Tha'ālibā, op. cit., 498—Miskōye, GMS., vii, 1, p. 130—Tajārib al-umam (transl. of Nihāyat)—Ibn Khaldūn, op. cit., ii, p. 171.

page 19 note 24 Ṭabarī, i, 831—Ḥamza from Bahrām the Mōbad, op. cit., p. 21—Bīrūnī, in his third list (from Ḥamza)—Mujmal, also following Ḥamza—Anonymous, Sprenger, 30 (Rothstein, p. 24).

page 19 note 25 Ḥamza, op. cit., p. 13—Bīrūnī and Mujmal, both following Ḥamza.

page 19 note 26 Eutychius (Oxford ed.), p. 374—Ibn Qutaiba, op. cit., p. 320.

page 19 note 27 Firdausī: “sī sāl⁰ bar sar do māh.”

page 19 note 28 Bīrūnū, , Chronology, p. 121 (in the first list).Google Scholar

page 19 note 29 Dīnavarī, al-Akhbār aṭ-ṭiwāl, ed. Guirgass, p. 49—Bal'amī, op. cit., f. 128a (Zotenberg's transl., ii, p. 85)—Agathias, loc. cit.—Nihāyat al-arab, op. cit., f. 89a—Michael the Syrian, i, 265(French transl.)—Syncellus, loc. cit.—Bar Hebræus, Mukhtaṣar ad-duwal (ed. Pocock, p. 127)—Fārs-nāma, pp. 20 and 63—Sibṭ Ibn al-Jawzī, op. cit., f. 186—Mīr Khāwand, op. cit.

page 19 note 30 This is the result of the calculation of the reign of the Arab kings of Ḥīra, ‘Amr^u ibn 'Adiyy and his son Amru al-Qais al-Bad’ under Shāpūr (see Ṭabarī, i, 822 and 834, and Nöldeke's transl., p. 47, n. 3). The same calculation in Ḥamza gives thirty-one years two months, as this author gives a round number (twenty-three years) for the reign of Amru al-Qais under Shāpūr (instead of twenty-three years and one month).

page 19 note 31 Mas'ūdī, at-Tanbīh, p. 100—Abū'1-Fidā', op. cit., p. 50.

page 19 note 32 Bīrūnī, op. cit., p. 127 (in the fourth list according to Zanjānī).

page 19 note 33 Ṭabarī, i, 831 (second version).

page 19 note 34 Ḥamza (according to Kisravī), op. cit., p. 17.

page 19 note 35 Mas'ūdī, , in Murūj, ii, 163Google Scholar.

page 19 note 36 Sebeos, p. 16, according to Patkanian, op. cit.

page 19 note 37 Ginza, loc. cit.

page 19 note 38 Sebeos, p. 16, according to Patkanian, op. cit. (the reign of S. jointly with his father twenty-seven years and after the death of the latter forty-six years).

page 19 note 39 Sibṭ ibn al-Jawzī (second version).

page 20 note 1 Unless some of them or their original sources should have counted the whole career of A. as his “reign” and should have taken this or that important step of his life as the starting-point, e.g. his becoming the commander of the army in Dārābgird, etc.

page 21 note 1 There can be little doubt that A. as a founder of the dynasty had a long struggle with the members of the old royal house (the Parthians), as well as with many other princes and chiefs of the different provinces, and therefore he may have postponed his official coronation as emperor until the resistence of all the isolated pretenders and rivals had been overcome. We know that an Arsacide prince (by name Artavazd ?), probably a son of Ardavān, was still reigning in some parts of the country (possibly in the south-western mountains) during the Sel. year 539, that is at least until the autumn of the year A. D. 227, a fact borne out by his coins. Therefore it would be safe to say that the rule of the founder of the new dynasty was not yet well established in the whole country in his first years. Mr. J. Walker kindly showed me a coin of the said Parthian offshoot belonging to the British Museum collection, which was found in Khurramābād. a town lying much farther to the south than Ray, Nihāvand, or Adharbāijān.

page 22 note 1 It is therefore probable that A. was called by that title (Shāhānshāh) at his coronation in Ctesiphon after his final conquest of that city and on his ascending the throne of the Parthians, and that this happened after he had consolidated his hold on all the provinces of Īrān, as Firdausī put it at this time. He records this event, i.e. A.'s being called by the title of Shāhānshāh, after a series of successive events following his slaying of Ardavān, namely his staying one or two months in Ardavān's palace; his returning then from Ray to Pārs; his building a city called Ardashīr-Khurra (by which Gôr, the modern Fīrūzābād, was meant) and a citadel with fountains and a fire-temple there, which city was “called later (after his death) Shahrazūr’ (read Shahr i Gôr, as the text in Mohl's edition has it); his several expeditions against the Kurds, Haftwād and Mihrak, and his returning at each interval to Pārs; his sending another expeditionto Kirmān; his marching on Shahrazūr (the real town of this name in Assyria); and finally his moving to Ctesiphon, whereupon his coronation took place in Bagdād (sic), the king sitting on an ivory throne and holding in his hand a mace-shaped royal sceptre. Nihāyat al-arab (op. cit.) puts the time of A.'s appellation with the said title after his battle with a confederation of the kings of Mesopotamia, who met him under the supreme command of a certain Shād-Mihr, one of the descendants of Mihrak the son of Fādhān. This battle again took place, according to the same source, after A. had killed Ardavān, halted then one month in- Nihāvand, marched from there into Khorāsān, stayed a full year in Marv, going then to Ray and from there to Ādharbāijān and subduing the king of that province, attacking then Armenia and conquering that country, turning to Mawṣil, and marching from there over the Tigris to Irāq. Dīnavarī also gives a long, though different, list of towns and provinces into which A. successively marched and which he conquered. He makes A. in the meantime stay one year in Istakhr. Ṭabarī, after relating the account of the battle of Hormizdjān, puts the arrival of A. in Ctesiphon after his conquering Hamadān, Media, Ādharbāijān, Armenia, and Mawṣil. Thus the ceremonial coronation of A. in Ctesiphon as emperor of Iran, after consolidating his power over all Iran and establishing his rule firmly, is possible, and may well have taken place, as already stated, some months after the beginning of the Persian year (his first year), and probably in the spring of A. D. 227.We may add here that such late dates as A. D. 228 or even 229 for the coronation of A. or for the beginning of his reign might be also implied by the purport of a passage of Kephalaia, as well as by the corresponding dates given in the Acta of the Persian martyrs (see above, No. 10 of the data). In the first-named book (Schmidt's edition, pp. 14–16) Mānī tells of his first inspiration when “the living Paraclete descended on him and talked to him” as having occurred “in the year during which the king Ardashīr was about to receive the crown”. This statement (provided the text is perfect and the translation correct) would imply that the coronation of A. had taken place after March, 228, for the first revelation in the career of Mānī had, as we have seen (vide supra), taken place in the lunar Sel. year 539 (26th March, 228–12th April, 229). As to the Acts of martyrs, we have seen that the 261st year of “the Persian rule” was made to correspond to the 799th year of the Greeks. This would imply that the first year was either the Persian year beginning on 26th September, A. D. 227, or the following year (beginning 25th September, 228). On the other hand, as stated above, Michael the Syrian puts the beginning of the reign of A. even in A. D. 230–1. However, owing to the uncertainty in the text and translation of Kephalaia, which is not so clear as to justify a definite conclusion upon it on the one hand, and the doubt in the authenticity of the dates in the passage of Ada on the other, we can hardly consider these documents or the record of Michael of sufficient weight to set against the other indications pointing to the Persian year beginning in A. D. 226 as the first year of A.'s reign. In the case of the date given in Acta it is possible to suggest that it may be the result of the careless calculation by the authors of the Acta, who may have simply subtracted the number of Sel. year for the beginning of the reign of A., as they perhaps found it in Græco-Roman sources, namely, 538, from 799, the year of the martyrdom (instead of subtracting 537, the zero year of the reign). We find the same kind of wrong operation used to-day by not a few modern authors.

page 23 note 1 Moreover, this theory, viz. the reckoning of the reign from the coronation rather than from the possession of Ctesiphon, has the advantage of not conflicting with the statement of the author of the Chronicle of Arbela, and of being free from the difficulty involved in supposing a rather long interval between the defeat of Ardavān and the capture of Ctesiphon.

page 23 note 2 Read the Arabic word there and not , which is, no doubt, a misprint in the text.

page 23 note 3 Another version, which gives him twenty years, apparently counts roughly his reign from the death of his father, Vologæses IV, in A. D. 208 or 209, again assuming that the end was in 226–7.

page 24 note 1 This version is that of the not very reliable historian Mūsā ibn 'Isā al-Kisrawī (see Ḥamza, op. cit., p. 17). This author, though he added many imaginary kings to his list of Sasanian rulers, may have taken his version of the length of A.'s reign from one of the different copies of the Kh^watāi-nāmak which he had at his disposal.

page 25 note 1 Since the end can be considered as being the 21st September, 241 (the beginning of the first year of Shāpūr), or some time in the early part of 242 (the probable time of the retirement or the death of A.), or even the 9th April, 243 (the date of the coronation of Shāpūr), the beginning of the period may, therefore, have been respectively the end of March, 222, or some time in the second half of the same year, or the 27th September, 223 (the beginning of the “Sasanian era”).

page 26 note 1 Adding the six months and fourteen days which had elapsed of the year before the coronation (22nd September, 24–-9th April, 243) to the thirty years which constitute the round number of the reign of S. after his coronation, the result would be thirty years six months fourteen days, or only two days more than that given in version IX.

page 26 note 2 That A. retired some time before his death is mentioned by some authors and is quite probable. Mas'ūdī (Murūj, ii, 160) states that A. renounced worldly pleasure and kingship and withdrew to a fire-temple. Firdausū, after recording the accession of S. to the throne and his oration on that occasion, states that “afterwards the news spread that A. died …”. But his associating his son with himself in the rule some time before his retirement, though possible, is by no means supported by any reliable evidence. It is only a conjecture by certain modern authors based on coins bearing the effigies of A. and a young prince together, and perhaps on some reports of the historians who speak of the crowning of S. by A. before the death of the latter. Mention is also made of Gutschmid's inference from the Latin words ascribed to Gordian III by Capitolinus, namely, “rege Persarum” (Gordian XXVII), that the two kings (the father and son) were acting together before the first (Persian) year of the reign of S., which began in September, 241.The reports relating to the crowning of S. by his father during the lifetime of the latter do not all clearly point to the last year of A. On the contrary the statements of Ya'qūbī (i, 179), Ṭabarī (i, 820), Bal'amī, and Sibṭ ibn al-Jawzī on this matter, as well as the coins with the two effigies, point more probably to this event (the crowning) having taken place rather in the early years of A.'s reign, and perhaps just after he had established his power over the whole empire and had settled down in Ctesiphon. The above-mentioned authors in their narrative of these historical events and their sequence put it clearly on this point. The effigy of the young boy with a helmet by the side of that of A. shows a youth who can hardly represent S. towards the end of his father's reign, when he must have been about forty years of age. The difference between this effigy and that of S. on the coins of his early reign (with the simpler title) is great. The latter represents the king at least ten years older than the young man represented by the former. Only Mas'ūdī (loc. cit.) speaks of the crowning of S. before the retirement of A., when he states that after A. appointed S. as ruler and crowned him with his own crown he “lived still one year and according to some reports one month and according to others more than we mentioned”. Thus it is quite possible that A. had crowned S. once in his (A.'s) early years as heir to the throne, as Ṭabarī and some others put it, and again with his own crown in the last year immediately before his retirement, as Mas'ūdī has it, and that S. ascended the throne after the death of A. as Agathias, Moses Khorene (Hist. of Armenia, chap. 81), and Bal'amī report, and finally, that he had his solemn coronation as Shāhānshāh many months after the death of his father, which must have taken place a short time after the said retirement.The statement of Mānī (Kephctiaia) to the effect that A. died “in the same year” as that in which he (Mānī) travelled to India and that in which S. became king, may also refer to the accession of S. to the throne.

page 28 note 1 However, he again gives S. an additional twenty-seven years of reign, “jointly with his father,” and adding this to the forty-six years which he assigns to S. for his exclusive reign, he puts the sum at sixty-three years (for S.)!!! As according to Ya'qūbī (i, p. 179)A., after conquering the country, appointed S. as his susccessor and called him king, S.. must have held the royal title for a long time before his actual accession.

page 30 note 1 The epagomenæ were at that time in the first half of the Persian year, hence one year (365 d.) + six Persian months (of 30 d. each) + 5 d. epagomenae + 14 d. of the seventh month = 564 d.

page 31 note 1 Only version VI, which gives the duration of the reign as fifteen years, may have counted from 26th September, 226, up to the beginning of Shāpūr's first (Persian Calendar) year (22nd September, 241). The first year of S. began almost certainly with the latter date, and this is another proof that the above-mentioned shorter numbers, i.e. fourteen years ten months or six months, do not begin with 26th September, 226, as otherwise those versions would not conform with the usual way of reckoning the reigns of the Sasanian kings. They used to count, as a rule, as stated above (p. 30), the year during which a king ceased to rule (through the death, abdication, or otherwise) as the first year of his successor.

page 31 note 2 Possibly October, 242, or at any rate after April of the same year.

page 32 note 1 Except for that of Yazdegird, the last Sasanian king, whose era continued in use for purposes of reckoning after his death up to the present day.

page 32 note 2 See No. 5 of the above data.

page 32 note 3 The existence of such an era is confirmed also by al-Quḍā'ū in his book Kitāb al-Inbā' bi anbā' al-anbiyā' wa- tawarūkh al-khulafā', composed in A.D. 1031 (Pocoek, Specimen Historise Arabum, Oxford, 1806, p. 178), in which he states that the Persians first used to compute their dates from the conquest of Iran by Alexander the Great and later from the accession of Ardashū to the throne.

page 32 note 4 77˚ 18' declination.

page 32 note 5 As already stated, according to the Babylonian reckoning the first year of a reign was the next calendar year following upon that of the accession of the king to the throne.

page 32 note 6 Būrūnī's words (“the ascendant of the year in which Ardashīr rose”) taken literally, however, may rather point to this latter interpretation.

page 32 note 7 ZDMG., 91, 3.

page 33 note 1 His death at the age of 100 is given by some authors as having occurred in 272 A.H., but Bīrūnī gives some observations made by him in 279 and 280 A.H. (Chronology, 340–1).

page 33 note 2 Because according to Ṭabarī, Pāb^hagh wrote to Ardavān immediately after this event asking for permission to put the crown of the king of Istakhr on the head of his (Pāb^hagh's)

page 34 note 1 In Pablavi “ēv-kh^watāyīh” or the monarchy in the literal meaning of the word.

page 34 note 2 This fire was lighted in the Persian year which began on 30th September, 208, and ended on 29th September, 209. There is mention of the fire-temple builty by A.in the books of Muslim authors.

page 36 note 1 The fact that according to Ṭabarī (i, 815), the king was in Baiḍā' when Pāb^hagh visited him to ask a favour for his son Ardashīr, as well as when later Pāb^hagh attacked and killed him, does not conflict with-Istakhr being the capital of the province, as the king may have spent part of his time in the adjacent towns.

page 36 note 2 See Gutschmid's Geschichte Irans und'seiner Nachbarlānder, pp. 156–9, as well as G. F. Hill, Catalogue of the Greek Coins of Arabia, Mesopotamia, and Persia, pp. 159–244.

page 36 note 3 The name of the royal house is most probably connected with Bāzrang a mountainous region in Arrajān about 100 miles north-west of Shīrāz, from where their ancestors may have come. This provincial dynasty must have been reigning in Pārs at least since the first half of the second century. For if Pāb^hagh was born about A.D. 160 or earlier, as we concluded, and if his mother was really “an offspring of a people who were kings in Pārs and were known as Bāzrangides” (Tabari, i, 814), she must have been born not later than A.D. 145, and her father, the maternal grandfather of Pāb^hagh and a Bāzrangide (if the reading of the word in Ṭabarī is correct) must have been nourishing at least before the middle of the second century.

page 36 note 4 Hill has six kings in the fourth or last series (of these local kings in Pārs), of whom the last is Artaxerxes, the son of Mīnōchihr. It is not certain that the king of Istakhr had any central authority, or that he was considered as suzerain over the kinglets of the other towns, though this is probable. Some modern authors speak of the anarchy reigning in Pārs before the rise of Sasanians, but I do not know on what authority.

page 37 note 1 In the inscriptions of Ka'ba i Zardusht. He may have the rank and title of frātadēra (the guardian of the fire) or fratakara or fraiarakā, which title is found on the coins of the older princes of Pārs.

page 37 note 2 His wife Rāmbihisht or Dēnakē was a princess of the Bāzrangian family. Nöldeke, however, considers this report to be legendary.

page 37 note 3 A fortress called Tīrkhudāi in the district of Khīr is mentioned by the Muslim geographers (e.g. in Fārs-nāmeh of Ibn al-Balkhī, etc.), which may be the same place as Tīrūdhē. The name may have changed through a misspelling and popular etymology into “Tir i Khudāi” (the arrow of God).

page 37 note 4 If this is true, Gōchihr, whose end we put about A.D. 220, must have been on the throne at least as early as 187.

page 38 note 1 The author of Nihāyat al-arab (f. 99a) states that A. was the youngest of the brothers. But Ṭabarī, who seems to have had better sources, says that A. had brothers other than Shāpūr, some of whom were older than he and some younger.

page 38 note 2 In the Kar-nāmak, in addition to the story of the last war, there is mention also of a previous war in which the army of Ardavān was defeated. Therefore the war must have begun in A.D. 223, if not earlier. Ed. Meyer in his article (Artabanus) in the Encyclopædia Britannica thinks that the war between A. and Ardavān continued for many years. The three battles mentioned by Dio also imply a considerable time.

page 38 note 3 Nöldeke thinks that all indications point to Babylonia or Susiana as the scene of this contest, but the author of Nihāyat puts it in the vicinity of Isphahān and the residence of Ardavān in Nihāvand, as Dīnavarī has it too. Ṭabarīstates that A. entered Hamadān after killing Ardavān. The Oriental sources call Ardavān king of Jibāl (Media), and Gutschmid is of the opinion that his capital was in North Persia (Geschichte Irans, p. 154).

page 39 note 1 Ṭabarī, after relating the account of the battle of Hormizdjan, puts the arrival of A. in Ctesiphon after his conquering Hamadān, Media, Aarbaijān, Armenia, and Mawsil. Dīnawarī and the author of Nihāyat, as well as Firdausī, who follows the Kārnāmak in most of the detail of his narrative about A., give long though different lists of towns and provinces to which A. successively marched and which he conquered. As is already stated, the former makes him stay in the meantime one year in Istar, and Nihāyat attributes to him one year's sojourn in Marv. The whole story told by these authors implies the lapse of soms considerable time, perhaps not less than three years.

page 40 note 1 The author of the Chronicle of Arbela (op. cit., p. 60) states that in earlier times before Aṛdashīr strove to remove the Parthians from the throne the Persians had tried their strength many times in war but were repulsed. However, in consequence of many wars and continued strife, the Parthians were paralysed and were defeated (in A.D. 224).

page 40 note 2 He must have reigned in Western Babylonia at least till the closing months of 221, as we have coins of him from 519 Sel. era up to 533 (A.D. 207–8 to 221–2). See W. Wroth, Catalogue of the Coins of Parthia, introd., P. 1xii.

page 41 note 1 He began the collection of Avesta texts and the composition of the holy canon of the Zoroastrians. The whole of his career was one of religious zeal and orthodoxy which was the most important factor in his success and popularity. The Parthians either did not recognize Zoroastrianism as their official religion or at any rate were not very zealous in that respect. Ya'qūbī (ed. Houtsma, i, p. 179) expressly states that A. was the first of the Magian (i.e. Zoroastrian) kings.

page 41 note 2 According to the Manichaean Homily Shāpūr came to the city of Bīshāpur (?) in Pars and fell ill and died. (SeePolotsky, H. J., Manichāische Homilien, 42Google Scholar.)

page 41 note 3 One year and ten days according to the older sources.

page 41 note 4 26th April, if the numbers for his reign and that of his predecessor are strictly counted.

page 41 note 5 Strictly, 27th July. Should the reign of Hormizd be counted one year and six months, or one year and ten months as given in some other sources, the dates must be put some months later respectively.

page 41 note 6 Contrary to what I have formerly held, I believe now that this date is more in accordancewith the data at our disposal, as is explained in the excursus below.

page 42 note 1 Bīrūnī mentions also two other minor fasts of the Sabians, namely, three days at the beginning of the month of Shabāt and one on the 14th Ēlūl, but none of theseare mentioned by an-Nadīm. Ibn al-Jawzī mentions also the same fasts with the same dates and same number of days as given by an-Nadīm (see Talbīs Iblīs, ed. Cairo, 1928, p. 75). Barhebraeus has also the same, but with a mistake regarding the date of the first fast, which he puts on the 9th of Kānūn I (instead of nine days before the end of the month).

page 43 note 1 In the intercalary years when there were two Adārs their fast-month must have begun in the first of the two and not in the intercalary month, for otherwise the beginning of the fast (the 8th day) might then fall in the solar month of Aries. In this case, i.e. in the intercalary year with two Adārs, which occurs more than thirty times in a century, the end of fasting or their fast-breaking feast must have fallen inthe second Adār and not in Nīsān as usual. In the first six centuries of the Christian era perhaps only in six different years the 8th day of the Bab. Adār I fell a little after the vernal equinox.

page 43 note 2 The only difference was perhaps that the former (the lunar month) began apparently with the new crescent, whereas the latter (the Chinese month) begins with the new moon.

page 43 note 3 The text reads as follows: “… afterwards when the new moon took place and the sun entered Aquarius and eight days elapsed of the month …” (p. 333).

page 44 note 1 These exceptional years were: A.D. 430, 449, 468, 487, 506, 525, and 593.

page 44 note 2 Ed. Bang, Le Muséon, 36, 1923. The conclusion, from the text of the passage, ofthe Bema's being at the end of the great fast, does not seem to be indisputably clear.

page 44 note 3 These were the years 375, 378, and 381. The 8th day of Shabāt fell on 27th, 24th, and 20th February respectively.

page 45 note 1 SeeMueller, F. W. K., Mahrnamag, ii, 311Google Scholar and 373; Henning, W., Mitteliranische Manichaica,iii, Glossar, 53Google Scholar (where two references are given) and Ein manichseisches Bet- und Beichtbuch, by the same author, Glossar, p. 123 (ten references); von Le Coq, Tūrkische Manichaica aus Chotscho, Nos. 8 (twice), 9, 21, and 22.

page 45 note 2 Ed. cit. (see page 44, footnote 2), xiv, A and B, pp. 162–i.

page 45 note 3 Called in Greek Bema (platform-step), in Persian gāh and in Turkish chāidān.

page 45 note 4 See Allberry, “Das Manichaische Bema-Fest” in Zeitschrift fiir die Neutestamentliche Wissenschaft, 37, 1938, 6.Google Scholar

page 45 note 5 BekMbuch, p. 9.

page 45 note 6 See No. 16 of the Uigur manuscripts publ. by Eadloff in Gruenwedel's Berichte ueber archseologische Arbeiten in Idikutschari, Abh. d.K. bayer. Akad. d. Wiss., philos.-philol. Kl., 1906, pp. 181–195; Radloff, Uigurische Spraehdenlcmaeler, Leningrad, 1928, MSS. 16, 88, 112, 117, and 125; Rachmati,' Tuerkische Turfan-Texte, Abh. d. preuss. Akad. d. Wiss., phil.-hist. Kl., 1936, xii (where the name of this month occurs six times); the Uigur vocabulary publ. by Klaproth in his Abhandlung ueber die Sprache und Schrift der Viguren, Paris, 1920, 19; and Zīj i Īlāni of Naşīr ad-Dīn Ṭūsī, MS. Cambridge 0. 2. (7) f. 15a, as well as the Zīj of Ulugh-Beg.Google Scholar

page 46 note 1 Die alttilrkischen Inschriften und die arabischen Qvellen, p. 4, Radloff 2te. Folge, 1899.

page 46 note 2 20 u'l-qa'da must be read 20 u'l-hijja and 2 Ilcindī must have been in the original copy 10 Ikindī (in Persian dahum instead of duvum, a very easy miswriting in the Arabic script).

page 46 note 3 SeeRadloff, W., Die alttürkischen Inschriften der Mongolei, Petersburg, 1895,Google Scholar as well as the second series of the same with the annexes by F. Hirth and W. Barthold, 1899.

page 46 note 4 SeeGinzel, F. K., Handbuch der mathematischen und technischen Chronologie, Leipzig, 1906, 1., appendix, 501–2.Google Scholar

page 46 note 5 There is no reason to believe that at any time of which we have any Turkish document theTurkish calendar differed from that of China. The divergence of sixty-three days which was formerly believed to have existed between the Chinese and Turkish dates relating to the death of Bilga Kāghān, was satisfactorily explained by Pelliot in T'oung Pao (1929, p. 230, n.). Even in Bīrūnī's list of the old Turkish months the numbers given to the months (though with some confusion in the order) may have corresponded to the same numbers of the Chinese months, and, as Barthold has suggested (op. cit.), the big month (ulugh āi) and the little month (kichig āi) may have corresponded with the 11th and 12th Chinese months. From the names of the months used in the Uigurian inscriptions of Mongolia in the runic script published by Ramstedt (Zwei uigurische Runeninschriften in der Nord-Mongolei, Journal de la société finno-ougrienne, 30, 1913) such as ikinti āi, tōrtunj āi, sakizinj āi, and particularly from the way of counting the days of the first decade of the month by adding the word yanghiqa to the number of the day, it appears that even the calendar used by these early Uigurs before their conversion to Manichseism was probably not different from that used by the later Uigurs of Eastern Turkestan except perhaps in the name of some of the months (Ramstedt ascribed the inscriptions of the tomb in Sin-usu to Moyunchur orrather Bayanchor.who reigned in the Silinga valley in A.D. 746–759). Ramstedt, however, puts the beginning of the year of these Uigurs in November, perhaps basing his theory roughly on Bīrūnī's above-mentioned list and taking birinj āi of that list as corresponding to the third Chinese month, which theory is, as stated above, by no means certain.

page 47 note 1 The Chinese months begin with the new moon, and therefore the 8th day of the lunar month beginning with the day following the first visibility of the crescent would correspond with the 10th (or 9th) day of the Chinese month. But as the Arabic expression “and eight days elapsed from the month” used by an-Nadīm in this connection (the beginning of the fast) may be interpreted as meaning that the first day of the great fast was the 9th of the lunar month as well as the 8th, it could correspond also with the 11th day of the Chinese month.

page 47 note 2 In exceptional cases it may correspond to the last day of February.

page 47 note 3 See Chavannes and Pelliot, op. cit., pp. 112 and 310, n. According to the same scholars the Chinese word meant literally retirement in the Buddhist religious sense for the Buddhist monks. This maythus have become the name of the same period of retirement and observance - of religious duties.

page 47 note 4 In the S.W. and N.W. Iranian languages the equivalent word used in the same sense is andarz.

page 47 note 6 Similar to the Arabic word imsāk.

page 48 note 1 In Eastern Turkish (in Āzarbāijān) the word orūj means fast and Orūjliq the fast-month of Ramaḍān.

page 48 note 2 Perhaps this was a fasting habit practised in that particular Buddhist community. The Buddhist fasting was developed apparently in Tibet and adjacent countries. (See the article “Fasting” by Oscar Hardman in Encyclopaedia Bntannica, 14th ed., vol. 9, p. 107.)

page 48 note 3 If, however, the use of the word Chaqshapat as the month-name in earlier times should be proved, it'is also quite possible to suppose the reverse of the above-mentioned process; namely that the word was used first by the Buddhist Uigurs before their conversion to Manichseism for the Buddhist ten commands, and that later it was adopted for the. Manichaean commands of abstinence of the same number.

page 48 note 4 In the Turkish vastvaneft the word bachag and vusanti are both used side by side in the sense of fasting. Was the latter used for a complete abstention from all foods, like ṣawm of the Muslims, and the former rather for abstention from certain foods only, like the Christian fast, or the longer fast of the pseudo-Sabians, who abstained only from the eating of meat ?

page 48 note 5 The author of this Zīlj (astronomical tables), Naṣīr ad-Dīn Ṭūsī, was well acquainted with the astronomy and calendars of the Far East, the Eastern Turks and Mongols.

page 48 note 6 The word Bashī or bagsī used by the Mongols is of Chinese origin (modern Chinese po-ṣī ―po-shih, Karlgren, Analytical Diet., No. 50 and 877 = Middle Chinese pâk-ḍż'i; Japanese hakushi) means etymologically “man of wide learning”. I am indebted to Professor H. W. Bailey for the above explanation and references.

page 48 note 7 According to Henning (BSOS., viii, p. 588) the word bachag is borrowed from the Sogdian.

page 48 note 8 A Manichsean Psalm Book, part ii, ed. and transl. by C. R. C. Allberry, Stuttgart, 1938, p. 8.

page 48 note 9 The occurrence of the word chashapat mā(chaqshapat-month) in the Sogdian text mentioned above also confirms its being used in the literal sense and not as the name of a calendar month, since in the latter case probably the corresponding Sogdian month-name at the time would have been used.

page 49 note 1 This is, howeyer, doubtful, and it is more probable that the epagomense were at the end of Amerdād and before Shahrēvar, as it is more likely that the adoption of the Zoroastrian calendar, based on the Egyptian model by the state, and the introduction of the 120-yearly intercalation occurred about 441 B.C. rather than 487 (see my Old Iranian Calendar, prize publication of the Royal Asiatic Society). In this case the 4th of Shahrēvar in 276 would be Saturday, the 19th February.

page 50 note 1 In the years A.D. 233 and 322 Mihr and Nīsān corresponded strictly with each other.

page 50 note 2 The epagomenæ must have been before Shahrēvar, and most probably were not moved to the end of that month until A.D. 279 (see Old Iranian Calendar by present writer, prize publication of the Royal Asiatic Society, 1938).

page 51 note 1 Neugebauer himself points out that owing to the dependence of the reckoning of the first day of the lunar month upon the observation of the new crescent and clear sky, a difference of one day between the Julian dates corresponding with the Babylonian (lunar Sel.) dates obtained by the use of his tables and the real dates is always possible. Since the new moon occurred in Babylon only on the 2nd February, 276, at about 2 h. 15 m. p.m., the visibility of the new crescent in the evening of the following day (the 3rd February) was quite possible, and this would make the first day of Shabat fall on 4th February.