^{page823 note 1} Sprengling, M., “ A New Pahlavi Inscription,” American Journal of Semitic Languages and Literatures, vol. liii, No. 2 [01, 1939[, pp. 126–144 [═ Sprengling I].Google Scholar

^{page823 note 2} “ Zur Parsik-Inschrift an der ‘ Kaaba des Zoroaster’,” Zeitschrift der Deutschen Morgenländischen Gesellschaft, vol. 91, No. 3 [1939], pp. 652–672 [═ Sprengling II].Google Scholar

^{page823 note 3} I had at my disposal another, slightly better photograph which Professor Bailey had generously lent me (with Professor Sprengling's kind consent); my readings are throughout based upon this photograph which previously had been sent to Professor Bailey by Professor Sprengling.

^{page824 note 1} If ptk'ry were to be read, the meaning would be “ fight ”; this reading, however, is not very likely.

^{page825 note 1} The two words MN … kḥdṩ are to be translated “ together with ”. Sprengling's rendering “ vom Umkreis … aus zugleich ” (II, p. 656), is not quite correct. I still adhere to my proposal (Gütt. Gel. Am., 1935, pp. 15 sq.Google Scholar) to read kḥdẖ as ham(m)is (not hamē); the older translation of hamis as “ separated ” (again Herzfeld, Altpers. Inschr., p. 186) is wrong; cf. Sogd. “mrδ'n < *hamiθr- < *hammiʸr- ?

^{page825 note 2} Cf. Herzfeld, , Altpers. Inschr., p. 75.Google Scholar [Or “ citadel” ? Cf. Pahlavi Psalter plyw'ly ═ Syr. sāẖarīā, Andreas-Barr, p. 57.]

^{page825 note 3} Thus to be read.

^{page825 note 4} Ghirshman, R., Revue des Arts Asiatigites, t. x, num. iii, pp. 123–9.Google Scholar At the end of line 13, ZNH ptkly YHBWN is to be restored; krty … MN NPŠH BYT (line 10) probably means “ made at his own expense ”. As reward for his gift ‘ps'y receives “ (things) of gold and silver, slaves and slave-girls, gardens and estates ”. On ‘BDk wknysky see Herzfeld, , Altpers. Inschr., p. 87 n.Google Scholar; the spelling knysk- proves that the reading knyšk- in the Pahlavi Psalter was wrong (s and š can rarely be distinguished); Pahlavi (Frahang) also should be analysed as knysk; the last word of the inscription is Pahlavi ═ χvāstak (the ideogram possibly “ land ”; hardly Greek ἄxpa. as suggested by Herzfeld, ibid., p. 122 n.).

^{page826 note 1} Another proposal has been made by Christensen, see Ghirshman, ibid., p. 128.

^{page826 note 2} Hardly an Iranian; the name ends in -ā, not -āy; Āpasāy would be spelt ‘ps'dy.

^{page826 note 3} See Dussaud, R., Topographic historique de la Syrie antique et médiévale, p. 450Google Scholar: “ un des faubourgs d'Alep.” In the Pahlavi inscription Calecome is treated as a separate town.

^{page826 note 4} The date given by Noldeke, 251/52 (Tabari, p. 31, n. 3), holds good for the conquest of Nisibis and Carrhae only; that in 563 Sel. Šāapür should have been not only in Syria, but even in Cappadocia, is quite incredible.

^{page826 note 5} Irrespective of the traces on the stone (Sprengling: wzdly).

^{page827 note 1} See Dussaud, op. cit., pp. 199 seq. Note the old form found in Amarna tablets: Zi-in-za-ar (Dussaud, p. 200, n. 2). Assimilation of -nz- to -zz- is, of course, common in Aramaic.

^{page827 note 2} Pahlavi -nc- renders -nĉ-, -nž-, -nǧ-, and -nz-; in the third century A.D. the ordinary pronunciation of -nc- in Persian words was -nz-.

^{page827 note 3} The emplacement of Hemerium is not known, cf. Dussaud, op. cit., p. 500, n. 1. For Immirina, mentioned in Assyrian texts along with Bambyce, see Dussaud, p. 520 (═ Bethammaria ?).

^{page827 note 4} He hardly went down the Euphrates as far as Dura; this town, along with other places on the river (Circesium, Nicephorium, etc.), probably was occupied by an expeditionary corps at the same time that the King of Kings crossed the river to capture Antiochia.

^{page828 note 1} That does not necessarily imply that all the provinces enumerated in the inscription were provinciae in the strict sense of that term.

^{page829 note 1} ═ ‘[y] lw[: but ’ ֵ t] is also possible.

^{page829 note 2} Sprengling: ’sp ֵ y’y. Possibly k is the first letter; perhaps k’. tn’tey or k’.lk’wy ? k’[l]tk'wy ═ Carthago is not very likely.

^{page829 note 3} Cf. Schaeder, H. H., Ungar. Jbb., t. xv, pp. 574 sq.Google Scholar For this reason, sydy is not an adequate rendering of Arab, sayyidī (Herzfeld, Paikuli, s.v. hlw'nyk; this latter word probably means “ of Holwān ”, cf. Bailey, , B808., IX, p. 233Google Scholar, and hardly equals ‘arbāyā, as Herzfeld proposes; cf. also Altpers. Inschr., p. 114; in Pahlavi ‘arbāyā would be ar00DF;ay, arvāy or arßāayīk, arvāyīk, cf. hrōmāy beside hrōmāyīk and ZDMG., vol. 90, p. 8Google Scholar).

^{page830 note 1} It is presumed, as a matter of course, that Greek forms (not Latin) underlie the Pahlavi spellings.

^{page830 note 2} It does not seem superfluous to determine the meaning of Judaea. The following explanation is offered by Professor Sprengling: “ After this we are faced with a mystery near the beginning of line 13. Jkhvdia [yhwdy'y] can hardly be anything other than Jehudaia, hardly Judea in the time of Shahpuhr I, but almost certainly something, probably some place, hardly simply a group of people, which is Jewish. Exactly what it is and what is said about it, the writer cannot guess.’ (I, p. 36; on similar lines, II, p. 657.)

^{page831 note 1} In Syriac, e.g., Greek ot is usually rendered by w.

^{page831 note 2} štry MN, read by Sprengling in the second half of the gap, is not quite certain.

^{page831 note 3} Cf. the inscription found at Shapur (line 1: DD X IIII IIII), particularly the excellent photograph (Rev. Arts Asiat., t. x, n. iii, before p. 123).

^{page831 note 4} “ Fifty (odd) ” ia not completely excluded.

^{page832 note 1} Exceedingly dubious.

^{page832 note 2} Or at the beginning of the list; six names are needed for the large gap in Une 12.

^{page833 note 1} Of two provinces with the same name, if distinguished by superior and inferior only, a single name alone was mentioned (Moesia). Where the distinguishing words carried more importance, a complete list might have been given, as is suggested by the (assumed) presence of the three divisions of Syria. Of a twofold designation the first name alone was entered: Lycia (et Pamphylia), probably also Cyrenaiea (et Creta), Bithynia (et Pontus).

^{page834 note 1} Nöldeke, , Tabarī, p. 32, n. 3.Google Scholar

^{page834 note 2} Hardly ’Pn as in the Ḥāǧǧīābād inscription.

^{page834 note 3} The translation offered by Junker (Frah., p. 91), “ bond, pledge,” etc., is quite wrong.

^{page834 note 4} See Herzfeld, , Altpers. Inschr., p. 136.Google Scholar

^{page834 note 5} The final -i of Orhāi is lost (as, e.g., in Armen. Urha); Orhai would be spelt ’wlh’dy.

^{page835 note 1} Or: in, at, by, through, etc.

^{page835 note 2} See Gibbon-Bury, ⁶, vol. i, pp. 459 sq.Google Scholar According to Sarre-Herzfeld, , Iran. Felsreliefs, pp. 77 sqq.Google Scholar, Cyriades was still alive in 260 and assisted in the submission of Valerian, as depicted in the famous reliefs at Shapur and Naqsh-i-Rustam.

^{page835 note 3} Cf. G. Rawlinson, Seventh Great Oriental Monarchy, vol. 1, pp. 82 seq.Google Scholar

^{page835 note 4} The ending of the first name is hardly ]s'y; ]ly'y is perhaps more likely.

^{page836 note 1} Possibly derived from av. varәta- “ captured, captive ”, also “ booty ” (Pahlavi vartak, N, Pers. barda).

^{page836 note 2} Sprengling: hwdy.’hyn or hrd-.

^{page836 note 3} The first is Mopsuestia (see farther on).

^{page837 note 1} I do not question the fact that he entered Antiochia in the course of the whole campaign; he certainly did so on his return to Persia.

^{page837 note 2} The other two (1. Zeugma-Gindarus, 2. Caeciliana-Hierapolis-Aleppo) had been used by him in his first expedition.

^{page837 note 3} See Dussaud, op. cit., p. 479.

^{page837 note 4} Also spelt sumaisāṭ, see Minorsky, , Ḥudūd al-‘Alam, p. 393.Google Scholar

^{page837 note 5} Cf. Gibbon-Bury, ⁶, vol. i, p. 459.Google Scholar

^{page838 note 1} See Ramsay, W. M., The Historical Geography of Asia Minor, p. 385Google Scholar; Arab, . Maṣīṣsa, e.g. Ḥudūd al-‘Alam, § 38, No. 10 (p. 149, tr. Minorsky).Google Scholar

^{page840 note 1} Ramsay, op. cit., p. 384.

^{page840 note 2} Ramsay, op. cit., p. 66.

^{page840 note 3} See Ramsay's lucid exposition, op. cit., pp. 342 sq., 460.

^{page840 note 4} Traces of b can perhaps still be seen.

^{page840 note 5} Sprengling has …. sy only.

^{page840 note 6} It is admitted that this proposal is rather doubtful.

^{page840 note 7} Seo Ramsay, op. cit., pp. 280 sq.

^{page840 note 8} The only exception (Mallos) has been explained above.

^{page842 note 1} Uncertain.

^{page842 note 2} Sprengling: Karmanādi, suggesting Kirman (II, p. 658).

^{page842 note 3} Not quite certain, but probable.

^{page842 note 4} Sprengling considers also kspsty'y (i, p. 137).—Kaspatyros !

^{page842 note 5} Or a larger number, if the towns of the list constitute a part only of the places that were occupied.

^{page843 note 1} Nöldeke, , with his wonted partiality against everything Persian, surpasses all ancient accounts in his representation of this rather legendary story (Aufsätze zur Persischen Oeschichte, pp. 93 sq.).Google Scholar

^{page844 note 1} I have been unable to find out from Professor Sprengling's discussion of the passage what meaning he ascribes to ’d’ly; he transcribes pat adhar (I, p. 138).

^{page844 note 2} Different, of course, from hayār “ friend ”; this word occurs (unrecognized) in the last line of the inscription: hdb'r [Sprengling: hdbr(w)y(wj] with a noteworthy ligature (Pahlavi Psalter: hdyb'l- Pahlavi ). Possibly connected with “ minister ” in Daniel (e.g. iii, 24; vi, 8) ═ *hadabara- “ companions, friends (of the king) ”, with Av. haδa (Skt. sadha-).

^{page844 note 3} *ayār would result in New Persian *yār; perhaps one might think of NPers. yāra “ bracelet ”, also “ anklet ” and even “ necklace ” (ṭauq); Arab, loan-word yāraǧ (); the dictionaries record even a side-form ayāra (). But yāra is used for ornaments only, not for fetters.—The ordinary word for “ chain ” is NPers. zanǧīr from (Parthian) zēnāihr (BSOS., IX, p. 90Google Scholar). Under Varhrān I, Mani was fettered at hands, feet, and neck.—*ayāl < NPers. yāl “ neck ”, etc., is unlikely here.

^{page844 note 4} Probably K‘N ().

^{page844 note 5} In the first place, Sprengling reads šs|.

^{page844 note 6} Hardly YHWWN (Sprengling).

^{page844 note 7} Sprengling: kznẖ (cf. II, pp. 658 sq.).

^{page844 note 8} Sprengling: šm w. The reading ŠMk (═ n'mky ═ dipi) although not certain seems preferable.

^{page845 note 1} Shortly before the inscription discovered at Shapur which was inscribed in Sept./Oct. 264 (in the month Frawardln of the 24th year of Šāpūr's reign: see Christensen's interpretation in the above-mentioned article by Ghirshman, pp. 127 sq.; if Šāpūr acceded to the throne in the spring of 242, the first year of his reign is counted as from 22nd Sept., 241; his 24th year began on the 16th Sept., 264).

^{page845 note 2} Herzfeld (in Paikuli) proposed ZNH.

^{page845 note 3} Herzfeld (ibid.) considered LZNH.—Cf. Mir. Man., iii, b, 118 sqq.: ‘yd ‘ndr mrg, ’wr ’w mrg.

^{page845 note 4} The reading ’yw is beyond doubt. Cf. Sprengling, ii, p. 666. ’yw is optative particle, cf. Andreas-Barr, p. 120; klyty is passive pres. sg. 3, cf. Andreas-Barr, p. 133; pmr’ywmy = “ I order ”. Cf. ZII., vol. ix, pp. 205 sq., 232 sqq.Google Scholar

^{page845 note 5} Reading quite certain. Cf. Sprengling, II, p. 659, line 33.

^{page846 note 1} Sprengling š … (cf. ii, p. 660).

^{page846 note 2} The numeral I has been regarded as “ a vertical dividing stroke ” (I, p. 130) which is “ like a long numeral I ” (ibid..) by Professor Sprengling.

^{page846 note 3} Sprengling reads ’yly; not quite certain.

^{page846 note 4} Thus Sprengling; cf. I, p. 141.

^{page846 note 5} Not cndy (cf. Sprengling, I, p. 141).

^{page846 note 6} Reading very dubious (Sprengling: wdys’d’y); see below.

^{page846 note 7} Not read by Sprengling.

^{page846 note 8} It is not intended to enter into a detailed discussion of Professor Sprengling's interpretation of the passage; his views differ widely from mine.—A similar (although not identical) explanation has been proposed by Professor Christensen, in a communication to the Congress of Orientalists held at Brussels in September, 1938, which I regret I was unable to attend. In the course of the discussion that followed Professor Christensen's paper, Professor H. W. Bailey kindly mentioned my interpretation which I had communicated to him by letter.

^{page847 note 1} That does not mean, of course, that the Sacred Fires were established by testament only.

^{page847 note 2} See Bartholomae, , WZKM., vol. xxvii, pp. 364, 369 sqq.Google Scholar

^{page847 note 3} See Unvala, J. M., Journ. K. S. Cama Inst., No. 11, 1928, p. 92Google Scholar; Benveniste, , Études d'Orientalisme … Raynwnde Linossier, t. i, pp. 155 sqq.Google Scholar; Andreas-Henning, , Mir. Man., ii, p. 317, n. 2Google Scholar; Schaeder, H. H., Iranica, pp. 19 sqq.Google Scholar

^{page847 note 4} Cf. Nöldeke, , Tabari, 43, n. 2.Google Scholar

^{page848 note 1} The second list, however, possibly represents the order of the sons according to age (cf. Herzfeld, , Kushano-Sasanian Coins, p. 34Google Scholar). Ōhrmizd-Ardasīr stands at the head of the first list, perhaps because his father designated him as his successor.

^{page848 note 2} Cf. Nöldeke, , Tabari, pp. 13, 19, n. 5Google Scholar; Guidi, , ZDMG., xliii, p. 410Google Scholar; Polotsky, , Manich. Homil., p. 44, n. b (annotation by Schaeder).Google Scholar

^{page848 note 3} Jamasp-Asana, Pahlavi Texts, p. 22, pu.: śahristān ī Ōhrmizd-Ardašir [uδ] šahristān ī Rām-Ōhrmizd Ohrmizd ī Nēv (the “ Brave ”) ī Šāpūrān kird. Cf. Marquart, , Ērānšahr, p. 144Google Scholar; Šahr. Erān, p. 95 sq.

^{page848 note 4} At an earlier date a brother of Sāpūr I, by name Mihršāh, was king of Mesene (Manich. fragment, M., 47).

^{page848 note 5} i.e., of those who were entitled to the succession.

^{page848 note 6} t. wrstn: there is a gap between t and w; h (or g ?) would fit in excellently. It is true, however, that in the parallel passage in Pers. I, 3, the reading twrst cannot be doubted. It seems necessary to assume two mistakes in one word. The identification of the name with Toχāristān is, therefore, rather doubtful; it would be valuable to have an early example of the spelling thwr- (as, e.g., in the Singanfu inscription, cf. B808., IX, pp. 545 sqq.Google Scholar). If we had to accept the spelling twrst(n) as correct, the identification with Taχwar, Toχār (first proposed by Herzfeld, , Paikuli, p. 246Google Scholar) would be excluded.–I regret that I have been unable to find a satisfactory reading for the last name.

^{page848 note 7} Pers. I, 2 sq. Andreas, who had at his disposal a good impression (beside photographs), read: šhpwhly sk’n MLK’ hndy [3] skstn Wtwrst wdgm’dnby. Professor ? ? ? Herzfeld's latest reading is: sakânšâh hinds sakastân u luχaristân dabirân dabir (Kushano-Sasanian Coins [Mem. Arch. Survey of India, No. 38], p. 36). The reading of the last two words (═ wdgm’dnby, or wdym’d’by) can hardly be accepted. It would be interesting to learn how Professor Herzfeld reads the traces which suggested Tuχaristân to him.