² Journal of the Royal Geographical Society, XVI, 1846, 1–105Google Scholar.

³ London, 1887; second edition, abridged, 1894.

⁴ cf. Lorimer, D. L. R., ‘The popular verse of the Bakhtiāri of S. W. Persia—II: specimens of Bakhtiāri verse’, BSOAS, XVII, 1, 1955, 103Google Scholar and passim, where several other place-names mentioned in this article are also found.

⁵ Early adventures, 1894, 338; ‘A description of the province of Khuzistan’, 85.

⁶ Henning, W. B., ‘The monuments and inscriptions of Tang-i Sarvak’, Asia Major, NS, II, 2, 1952, 151–78Google Scholar.

⁷ Herzfeld, E., ‘Bericht über archäologische Beobachtung im südlichen Kurdistan und in Luristan’, Archäologische Mitteilungen aus Iran, I, 1929–1930, 65–75Google Scholar and plate VIII.

⁸ op. cit., 71: ‘Die vielen Felsbildwerke von Mālamīr, die altelamischen wie die subarsakiden, sind kaum bekannt; ihre genaue Aufnahme ist dringend nötig, konnte aber diesmal nicht unternommen werden. Einige von ihnen tragen Inschriften, bisher nur durch ganz unzureichenden Abzeichungen aus alter Zeit bekannt, die mir gerade genügten, um zu erkennen, dass die schrift mit der auf gewissen charakenischen Münzen übereinstimmt’. It has recently been confirmed that there is at least one inscription at Īzeh (Mālamīr) which corresponds to this description. Yet it is clear that Herzfeld's remarks refer also to Shīmbār, of which an account follows. As often, he refers to the Characenian script, when in fact the Elymaean is intended.

⁹ Debevoise, Neilson C., ‘Rock reliefs of ancient Iran’, JNES, I, 1, 1942, 76–105, and especially 102Google Scholar.

¹⁰ ‘A description of the province of Khuzistan’, 84.

¹¹ It may be intended to emphasize the dynastic connexions of the cult.

¹² Dittenberger, W., Orientis Graeci inscriptiones selectae, I, Leipzig, 1903, 589 (No. 383)Google Scholar, 1. 55, for ‘Artagnes-Heracles-Ares’. For the representation, see now Ghirshman, B., Iran, Parthes et Sassanides, Paris, 1962 66Google Scholar.

¹³ Especially to the point is the discussion by Seyrig, H., ‘Antiquités syriennes: Heracles-Nergal’, Syria, XXIV, 1944, 62–80, in particular 68 fCrossRefGoogle Scholar. For the identification of Heracles with Melqart, see Dussaud, René, ‘Melqart’, Syria, XXV, 1946–1948, 205–30, especially 214, 218CrossRefGoogle Scholar. If our site has indeed a connexion with the cult of Bēl (p. 276 below), it may be relevant that Heracles appears with apparently Syrian deities at the temple of Bēl at Palmyra (Seyrig, H., ‘Antiquités syriennes: 17. Bas-reliefs monumentaux du temple de Bēl à Palmyre’, Syria, XV, 1934, 172Google Scholar). His role in that context is not fully explained. In Asia Heracles became assimilated to many local cults, e.g. that of Śiva-Heracles in Bactria (Fischer, Kl., Archäologische Anzeiger, 1957, 417Google Scholar).

¹⁴ It resembles in size and shape that in the investiture scene of Ardashir I at Firūzābād, rather than the ‘incense-altars’ common in non-Iranian contexts (e.g. H. Seyrig, op. cit., pl. XVIII and XXIII; cf. Starcky, J., ‘Autour d'une dédicace palmyrénienne’, Syria, XXVI, 1949, 51Google Scholar).

¹⁵ As argued by ProfessorHenning, W. B. for the monuments at Tang-i Sarvak, loc. cit., 176Google Scholar.

¹⁶ cf. Bartholomae, Altiranisches Wörterbuch, col. 953; *bājikara apud Cameron, George G., Persepolis treasury tablets, Chicago, 1948Google Scholar, index, s.v. But cf. Gershevitch, I., Asia Major, NS, II, 1, 1951, 139Google Scholar.

¹⁷ cf. W. B. Henning apud Driver, G. R., Aramaic documents of the fifth century B.C., abridged edition, Oxford, 1957, 81 and n. 1Google Scholar.

¹⁸ See notes by Schaeder, H. H. in Iranische Beiträge, I, Halle, 1930, 270Google Scholar.

¹⁹ Henning, W. B., ‘Mitteliranisch’ in Handbuch der Orientalistik, Abt. I, IV. Bd., Iranistik, 1. Abschnitt, Leiden, 1958, 41Google Scholar.

²⁰ Lidzbarski, M., Handbuch der nordsemitischen Epigraphik, I, Weimar, 1898, 289Google Scholar.

²¹ One might further compare for this word the uncertain Safaitic proper name b'š/bhš, on which see Ryckmans, G., Les noms propres sud-sémitiques, I, Louvain, 1934–1935, 48 fGoogle Scholar.

²² ‘šybt’ is probably related to the verb 'šp ‘to adjure’ which is commonly used, for example, in the magical texts (cf. Montgomery, J. A., Aramaic incantation texts from Nippur, Philadelphia, 1913, 303CrossRefGoogle Scholar), perhaps in contamination with šb', as this latter verb was also used in the same sense in the aph'el form. For the interchange of p/b there are examples in Babylonian Aramaic, cf. Levias, C., A grammar of the Aramaic idiom contained in the Babylonian Talmud, Cincinnati, 1900, 11Google Scholar; cf. also Nöldeke, Th., Mandäische Grammatik, Halle, 1875, 47 ffGoogle Scholar. This fact may explain why ‘šybt’ was deemed objectionable by the Jewish sages.

²³ The final -h of b'šybh could also conceivably be taken as representing the possessive pronoun of the third person singular masculine, the form of the noun being in this ease *'šyb.

²⁴ cf. Ledrain, E., Revue d'Assyriologie, II, 3, 1892, 93–5 and 32Google Scholar; Lidzbarski, M., Handbuch, I, 365Google Scholar; Ch. Clermont-Ganneau, , Recueil d'archéologie orientale, III, Paris, 1900, para. 22, 107–9Google Scholar. Lidzbarski, M., ‘Der Qaššīšā di daira und die Tracht der Palmyrener’, Ephemeris für semitische Epigraphik, Giessen, I, 1902, 87–9Google Scholar.

²⁵ cf. the clear occurrence of this usage in the dipinti of the Dura synagogue, e.g. Tile A, 1. 4–5, bqšyšwth dšmw'l khn' br yd'y 'rkwn, in C. C. Torrey apud Kraeling, C. H., The synagogue (The excavations at Dura-Europos. Final report, 8, Pt. 1), New Haven, 1956, 263Google Scholar.

²⁶ e.g. Brockelmann, C., Grundriss der vergleichenden Grammatik der semitischen Sprachen, I, Berlin, 1908, 121Google Scholar.

²⁷ ibid., 128; cf. also, for Neo-Syriac, , Th. Nöldeke, , Grammatik der neusyrischen Sprache, Leipzig, 1868, 39 fGoogle Scholar.

²⁸ cf. qṣyṣt'/gṣyṣt' in the incantation texts, on which see Rossell, W. H., A handbook of Aramaic magical texts, Skylands, N.J., 1953, 16Google Scholar. Also the verb šrg/šrq mentioned by Levias, C., A grammar of the Aramaic idiom contained in the Babylonian Talmud, Cincinnati, 1900, 10Google Scholar.

²⁹ cf. Th. Nöldeke, , Mandäische Grammatik, Hallc, 1875, 38 fGoogle Scholar.

³⁰ cf. Henning, W. B. in Gnomon, XXVI, 1954, 479Google Scholar.

³¹ It could conceivably be an Iranian loan-word, cf. YAv. stūra- ‘umfangreich, stark, derb’, Bartholomae, C., Altiran. Wörterbuch, 1609Google Scholar. For similar rendering of foreign st- by ṣṭ- in Mandaean cf. Nöldeke, , Mandäische Grammatik, Halle, 1875, 45Google Scholar.

³² Cooke, G. A., A text-book of North-Semitic inscriptions, Oxford 1903, 56Google Scholar. Cf. Jean, F. C. and Hoftijzer, J., Dictionnaire des inscriptions sémitiques de l'Ouest, I, Leiden, 1960, 24Google Scholar, with further references.

³³ Justi, Iranisches Namenbuch, s.v. ϐ⋯λουρος (Plutarch, , Artax., xxii, 5Google Scholar) connects this word with NP belūr; one may doubt, though, the appearance of a Greek loan-word in Persian of so early a date, and explain it rather as a Semitic proper name with Bēl-.

³⁴ Henning, W. B. in Asia Major, NS, II, 2, 1952, 172Google Scholar.

³⁵ cf. Benveniste, E., Journal Asiatique, CCXLVI, 1. 1958, 38Google Scholar; A. Dupont-Sommer, ibid., 19; Altheim, F. and Stiehl, R., East and West, IX, 3, 1958, 193Google Scholar.

³⁶ Bartholomae, , Altiranisches Wörterbuch, 1656 fGoogle Scholar.

³⁷ For this type of name cf. Caquot, A., ‘Sur l'onomastique religieuse de Palmyre’, Syria, XXXIX, 1962, 231 ff., esp. 249 fCrossRefGoogle Scholar.

³⁸ Alternatively, it could of course mean ‘the favour of Bēl’.

³⁹ This objection is raised by Benveniste, loc. cit.

⁴⁰ Though it is true that another Iranian loan-word, mzyšty', is also treated in the same way in that inscription. To take the ending -y' for the Old Iranian case-ending, as is done by Altheim, F. and Stiehl, R., East and West, IX, 3, 1958, 193Google Scholar, is surely impossible.

⁴¹ cf. Henning, W. B., Gnomon, XXVI, 1954, 478Google Scholar.

⁴² The personal names are treated separately below.

⁴³ cf. Henning, W. B., ‘The monumentṣ and inscriptions of Tang-i Sarvak’, Asia Major, NS, II, 2, 1952, 174Google Scholar.

⁴⁴ ibid., 167.

⁴⁵ cf. ibid., 174.

⁴⁶ ibid., 167.

⁴⁷ For further discussion of this phrase, see Appendix, below.

⁴⁸ cf. Lidzbarski, M., Handbuch, 382Google Scholar.

⁴⁹ Lidzbarski, M., Ephemeris, II, 293Google Scholar.

⁵⁰ Lidzbarski, M., Handbuch, 482Google Scholar; another occurrence in Lidzbarski, , Ephemeris, I, 77Google Scholar.

⁵¹ idem, Handbuch, 331.

⁵² idem, Ephemeris, II, 43; on which see also Ryckmans, , Les noms propres sud-sémitiques, Louvain, 1934–1935, I, 212 fGoogle Scholar.

⁵³ East and West, X, 4, 1959, p. 260, n. 123Google Scholar; the attempt to connect this with Wḥwḥ[štr ‘Oxyathres’ on coins of Persis seems therefore unfounded.

⁵⁴ See Justi, F., Iranisches Namenbuch, Marburg, 1895, 231 (s.v. Ōgī)Google Scholar.

⁵⁵ F. Altheim and R. Stiehl, loc. cit.; Justi, , op. cit., 341Google Scholar (s.v. Wahuka) did not connect with Ōgī.

⁵⁶ Yāqūt, Mu'jam al-buldān s.v. Ṣūl; Ṭabarī, Annales, index, s.v.; Justi, Iranisches Namenbuch, s.v. Sūl. Cf. Minorsky, in BSOAS, XII, 1, 1947, 29Google Scholar.

⁵⁷ But compare Sabaean Šptn, Ryckmans, Les noms propres…, 282.

⁵⁸ Annales, III, 1282.

⁵⁹ Contacts between Palmyra and Elymaia are attested in the Greek inscription published by Seyrig, H. in Syria, XXII, 1941, 256 ffGoogle Scholar.

⁶⁰ See further Appendix, below.

⁶¹ ‘A description of the province of Khuzistan’, 86: ‘The Lurs assured me that there was a long inscription near the Pul-i Nagīn, but I was unable to discover it. Yet after my return from the place, they still persisted in affirming that there was a large block covered with characters which had escaped my notice’. The truthfulness of Layard's Bakhtiāri informants is now fully established.

⁶² ‘Récherches archéologiques’ in Mission scientifique en Perse. Mémoires, IV, 298 and fig. 176.

⁶³ e.g. those discussed by de Mecquenem, R., ‘The early cultures of Susa’ in Pope, A. U. (ed.), A survey of Persian art, I, Oxford, 1938, 140 and fig. 10Google Scholar, which despite the stratigraphie indications can hardly be attributed to the fourth millennium B.C.

⁶⁴ I owe to Professor W. B. Henning a preliminary confirmation of my belief that several of the graffiti at this site are likely to pre-date the Sasanian period.

⁶⁵ The orthography agrees with the reading established by Professor Henning on coins of Elymais, , cf. Asia Major, NS, II, 2, 1952, 164Google Scholar.

⁶⁶ It is interesting that the introduction of the grape in Susis is attributed to the Macedonians in Strabo, XV, 731–2; though this need not be accepted as scientifically exact.

⁶⁷ Syncretic connexions between the Mazdayasnian religion and the cult of Bēl seem to be already attested in the Aramaic inscriptions from Arabissos (Harput) in Cappadocia, which need not long postdate the Achaemenian period, cf. Lidzbarski, M., ‘Aramäische Inschriften aus Kappadocien’, Ephemeris für semitische Epigraphik, I, 1900–1902, 59–73 and 319–26; especially pp. 67–9Google Scholar.

⁶⁸ Sachs, A. J. and Wiseman, D. J., ‘A Babylonian king list of the Hellenistic period’, Iraq, XVI, 2, 1954, 207Google Scholar.

⁶⁹ The narrative which follows is combined from Strabo, XVI, 744; Diodorus, XXVIII, 3, XXIX, 15; and Justin, XXXII, 2.

⁷⁰ In The Greeks in Bactria and India, W. W. Tarn used the evidence of Pliny, VI, 135 (…arcem Susorum et Dianae templum augustissimum…) to locate at Susa the temple of Nanaia attacked by Antiochus IV Epiphanes, in an episode often considered to be a doublet of that related about Antioehus III. This interpretation was queried by Henning, W. B., ‘The monuments and inscriptions of Tang-i Sarvak’, p. 177, n. 1Google Scholar, who rightly insists on Pliny's distinction between Susiana and Elymais. That the disaster of Antiochus III could itself have taken place at Susa is highly improbable; even by assuming this in the case of Antiochus IV, Tarn came into direct conflict with the evidence of 1 Maccabees vi, 1–4.

⁷¹ Henning, W. B., Asia Major, NS, II, 2, 1952, 171Google Scholar.

⁷² Cantineau, J., Inventaire des inscriptions de Palmyre, Fasc. 9, Beyrouth, 1933, 10Google Scholar.

⁷³ cf. in Schlumberger, D., La Palmyréne du nord-ouest, Paris, 1951, Recueil épigraphique, Nos. 2, 2 bis, 8Google Scholar.

⁷⁴ cf. above, p. 277.

⁷⁵ This is apparently a title of dignity, its position after the name being comparable to zy b' šybh in Shīmbār, see above. It seems to be composed of rabbȃn ‘lord, master’, with the possessive pronoun of the first person singular. A similar suggestion has been made by Altheim, F. and Stiehl, R., Supplementum aramaicum, Baden Baden, 1957, 91Google Scholar.

⁷⁶ i.e. at the Court (Henning). In inscription no. 1 this word is written btr‘’; in dialects where the laryngeals are no longer pronounced properly it often happens that the is omitted when it comes after a consonant and before a vowel, cf. the name šmwn quoted above. For this reason it is preferable to divide the words as is done here rather than zy btr 'br, where it would be more difficult to account for the loss of the in the construct state.

⁷⁷ loc. cit., 172 f.

⁷⁸ BSOAS, XVI, 1, 1954, 26 ff.Google Scholar; Anatolian Studies, III, 1953, 102Google Scholar. Professor Segal, who has shown kind interest in the present discussion, has made several helpful suggestions and has also agreed to let us use an article of his which was still unpublished when the present paper was in preparation (cf. p. 289, n. 83, below).

⁷⁹ It will be noticed that this translation varies in certain points from the one given by Professor Segal.

⁸⁰ gḥn, preceded and followed as it is by verbs in the imperfect, must be an active participle, although its spelling contrasts with that of n'syb. [Dr. S. Morag remarks on this that the inter-change of imperfect and participle, as well as that of the plene and the defective spellings of the participle, are particularly typical of the Babylonian dialect of Aramaic]

⁸¹ 'lyh ‘him’ represents a form similar to that which is the rule, e.g., in Judaeo-Aramaic.

⁸² It is clear from the context that 'bd must mean ‘worship’ in this inscription. This meaning of 'bd is attested in Syriac and in the derivative noun ‘bydt’ in Biblical Aramaic; cf. also its magical meaning in the incantation texts, Montgomery, J. A., op. cit., 297Google Scholar. [It has to be noted that Professor Henning, whose objections have given rise to the last three notes, has expressed reservations as to the interpretation proposed here.]

⁸³ For this title see esp. Segal, J. B. in Bacon, E. (ed.), Vanished civilisations, London, 1963, 217Google Scholar; cf. also Maricq, A. (ed. by Pirenne, J.) in Syria, XXXIX, 1962, 100–3 [reference kindly given by Professor Segal]Google Scholar.

⁸⁴ The peculiar aspects of the Sumatar religion, amply discussed in Professor Segal's publications, have not been mentioned here, as they may be irrelevant for the understanding of the Elymaean material.

⁸⁶ Iraq, II, 2, 1935, 209–10Google Scholar.

⁸⁶ Cylinder seals: a documentary essay on the art and religion of the ancient Near East, London, 1939, 159–60Google Scholar.

⁸⁷ Modi, J. J., The religious ceremonies and customs of the Parsees, Bombay, 1922, 321 ffGoogle Scholar. The shapes of the stools (x^vān) are described on pp. 270 ff. Cf. the eyewitness account of Drower, E. S., Water into wine: a study of ritual idiom in the Middle East, London, 1956, 199 ff.Google Scholar; on the stool cf. also p. 22 f.

⁸⁸ Barnett, , loc. cit., 210Google Scholar.

⁸⁹ ibid.

⁹⁰ loc. cit.

⁹¹ Drower, , op. cit., 212Google Scholar; Modi, , op. cit., 274Google Scholar.

⁹² One could also take up ProfessorHenning's, cue (loc. cit., 174)Google Scholar and regard š'ys' ‘the marble (pillar?)’, as the object of the following verbs: ‘He feeds, bows to, and worships the marble (pillar)’. This construction is perhaps preferable because it does not break up the phrase n'syb kwrsy'. [Dr. S. Morag proposes to regard š'ys' as derived from an Aramaic verb *šws/šys ‘to lead’, similar to Arabic . A related Hebrew form may occur in the Damascus Covenant word wysysw (i, 21), cf. Rabin, C., The Zadokite documents, second ed., Oxford, 1958, 5Google Scholar. š'ys' would then be an epithet ‘the ruler, the chief’ following n'syb kwrsy'.]

⁹³ See Henning, loc. cit., 159 and plate xvi.