¹ See, most recently, Lommel, H., ‘Zarathustras Priesterlohn‘, in Spies, O. (ed.), Studio, Indologica, Festschrift für W. Kirfel (Bonner Orient. Studien, NS, 3), Bonn, 1955, 187–95Google Scholar; Rudolph, K., ‘Zarathuštra—Priešter und Prophet’, Numen, VIII, 2, 1961, 81–116Google Scholar; Bailey, H. W., ‘Salca śśandrāmaia’, in Wiessner, G. (ed.), festschrift für W. Eilers, Wiesbaden, 1967, 137, 141Google Scholar.

² The importance of the Brāhmanas for a better understanding of Zoroastrianism was stressed by Molé, M., Culte, mythe et cosmologie dans i' Iran ancien, Paris, 1963Google Scholar; but doubts about the relevance of Zoroaster's earthly existence prevented him from considering this in the light of the prophet's own priesthood. Long before, J. Darmesteter also drew on the Brāhmanas for parallels for Zoroastrianism, maintaining that ‘ces théories qui paraissent subitement, en plein Brahmanisme, n'ont pas poussé tout à coup et leurs racines, serpentant sous les Védas, plongent dans le sol indo-iranien’(Ormazd et Ahriman, leurs origines et leur histoire, Paris, 1877, 302)Google Scholar.

³ See Haug, M., The Aitareya Brāhmanam of the Rigveda, Bombay, 1863, 17–18, 31–2Google Scholar; Henry, V., ‘Esquisse d'une Iiturgie indo-iranienne’, in Caland, W. and Henry, V., V Agnistoma, Paris, 1906–1907, II, Appendice in, 469–70Google Scholar.

⁴ On the functions of the hotar see Renou, L., Vocabulaire du rituel védique, Paris, 1954, 175–6Google Scholar; Henry, , op. cit., 475 fGoogle Scholar.

⁵ On traditional elements in Zoroaster's own hymns see further below, p. 26, n. 25.

⁶ The old type of priests' school was called by the Parsis a sār (< Gujarati niśāl). Each long day's work was divided into two, half being devoted to learning new texts, half to revising old. The priest recited a little, and the pupil after him; and then after repetitions the pupil withdrew to fix the words in his mind by repeating them over and over again to himself. Before the use of books as an aid, older pupils would help the younger ones at conning new material. The last sar in Navsari closed two decades ago, its last pupils being Ervad Edul Hiraji Kanga (now yōždaθragar in Navsari, whose father had learnt there before him) and Ervad Keresasp M. Kotwal (now ātaš-band at the AnjomanI Ātaš-Bahrām in Bombay). This sār was taught by Ervad Rustamji Kharshedji Kanga (d. A.D. 1929) and his son Ervad Sohrabji, men with phenomenal memories, rwho both became blind in adult life. [Information through Ervad Dr. Firoze M. Kotwal.] In Yazd also there were priestly ustāds who were especially entrusted with teaching. Dastur Khodadad S. Neryosangi recalls in particular in his own boyhood Dastur Mihragan, who had himself as a boy read Vendīdād with Dastur Shahriyar, the dastīr-mas when Manekji Limji Hatarīa came to Iran. Late tradition even knows the name of the teacher to whom the prophet himself was entrusted at the age of seven, see the Zarātusht nāma, ed. Rosenberg, F., Petersburg, St., 1904, 11. 351 ffGoogle Scholar.

⁷ In his valuable biographical notes on the life of Ervad Tehmuras D. Anklesaria (see the preface to Anklesaria's, The social code of the Parsees in Sasanian times, part II, Bombay, 1912, 29 ff.)Google Scholar, J. J. Modi states that Ervad Tehmuras completed his training at a sār in Surat by the age of 11, by which time he knew not only the liturgy but also the main rituals. He was then initiated nāvar and marātab at Broach. After this he became assistant to his maternal uncle, a priest, and learnt more ritual from him. Modi has interesting general remarks on teaching in the sār, ibid., 30–1. Nowadays priests' sons have religious instruction in the early morning before attending secular school; and accordingly do not master the texts and rituals until a later age, about 17 or 18.

⁸ See Staal, J. F., Nambudiri Veda recitation, '-Gravenhage, 1961, 40Google Scholar.

⁹ See ibid. These schools, like the modern ones for Zoroastrian priests' sons in Bombay, are endowed and require fees; whereas in the small sār, as among the groups of pupils attached to a teacher under old Brahman custom, there was no regular charge, the teaching being repaid by filial services and occasional gifts.

¹⁰ See Staal, , op. cit., 59–60Google Scholar.

¹¹ In an old Zoroastrian centre such as Navsari many yasnas are celebrated each day. Until recent decades there was also an endowed anjomanī yasna, solemnized daily on behalf of the community. [Information from Ervad Dr. Firoze M. Kotwal.]

¹² See Haug, , Essays on the sacred language, writings and religion of the Parsis, third ed., London, 1884, 279 if.Google Scholar, and Haug, (ed. and tr.), Aitareya Brāhmanam, I, 60–2Google Scholar; for a more general comparison, Henry, , op. cit., 467 ifGoogle Scholar. See also Pavri, Khurshedji Erachji, ‘Ancient ceremonies: additions and improvements made in them’, Dastur Hoshang memorial volume, Bombay, 1918, 165–92Google Scholar.

¹³ Haug was unable to press the parallels in this case, since he was not aware of the prescriptions about offerings to fire contained in the Nīrangistān; see the translation by Bulsara, S. J., Aērpatastān and Nīrangastān, Bombay, 1915Google Scholar, index, under ‘Animal for dedication’ and ‘Sacrifice of dedicated animal’, and further Boyce, , ‘Haoma, priest of the sacrifice’, W. B. Henning memorial volume, London, 1970Google Scholar.

¹⁴ See Renou, , op. cit., 102Google Scholar.

¹⁵ See Renou, , op. cit., 57–8Google Scholar. Kuśa is used for the barhis, on which, in connexion with the barsom, see in detail Thieme, P.,’Vorzarathustrisches bei den Zarathustriern und bei Zarathustra’, ZDMG, CVII, 1, 1957, 71–5Google Scholar.

¹⁶ The prescription for strewing barsom for the sacrifice is found in the Nīrangistān, ed. Sanjana, D. P., Bombay, 1894Google Scholar, fol. 85 v, 1.11 f.; cf. the Aitareya Brāhmana, II.2.11 (transl. A. Berriedale Keith (Harvard Oriental Series), xxv, 1920,143), and ibid., n.2.7 (Keith, 139), where the reason for this is said to be: ‘The victim has plants as its body; verily thus he makes the victim have its full body’. Cf. Y, XLVIII, 6Google Scholar; and the Old Persian practice of spreading ‘the softest herbage, trefoil (τρ⋯φυλλον) by choice’ for the animal-sacrifice, Herodotus, 1.132. It is noteworthy that trefoil is a fodder-crop.

¹⁷ See Haug, , Essays, 283–4Google Scholar.

¹⁸ These are now always of metal in the Zoroastrian rite, but earlier were evidently of stone (as was the mortar, not used in the Indian ceremony; but on this see Henry, , op. cit., 474f.)Google Scholar. Pottery was not acceptable for ritual use. Cf. p. 28, n. 37, below.

¹⁹ By Zoroastrian prescription this knife must be metal-hafted as well as metal-bladed; see Modi, J. J., The religious ceremonies and customs of the Parsees, second ed., Bombay, 1937, 271Google Scholar.

²⁰ On the dedication of the yajña in modern times see Dubois, J. A., Hindu manners, customs and ceremonies, transl. and ed. by Beauchamp, H. K., third ed., Oxford, 1906, 513Google Scholar.

²¹ See Lévi, Sylvain, La doctrine du sacrifice dans les Brāhmanas, Paris, 1898, 68Google Scholar (citing the Śatapatha Brāhmana, 1.1.4.24).

²² See Lévi, op. cit., passim.

²³ For the offering to fire see Y, XLIII, 9Google Scholar and xxix, 7 (on which see further below); for offering in general (myazda), Y, xxxiv, 3Google Scholar. The recurrent word yasna is probably to be understood as ‘sacrifice’, e.g., Y, XLV, 10Google Scholar (yasnāiš ārmatōiš ‘devout sacrifices’). On other traditional ritual terms see Hümbach, H., ‘Rituelle Termini technici in den awestischen Gathas’, Münchener Studien zur Sprachwissenschaft, VIII, 1956, 74Google Scholar ff.; Rudolph, , art. cit., 99–100Google Scholar.

²⁴ See Y, xxvin, 5, xxix, 7, xxxi, 6, XLIV, 14, 17. In Zoroastrianism, as in Brahmanism, ‘le rite ne se sépare point de la liturgie’, Lévi, , op. cit., p. 15Google Scholar. On mąθra see further Thieme, , art. cit., 68 ffGoogle Scholar.

²⁵ On the I E parallels see Schwyzer, E., Die Parenthese im engern und im weitern Sinne (APAW, Phil.-hist. Kl., 1939, 6), 10 f., 22, 25Google Scholar; Schaeder, H. H., ‘Ein indogermanischer Liedtypus in den Gathas’, ZDMG, xciv, 3, 1940, 399–408Google Scholar; and for further literature Rudolph, , art. cit., p. 98Google Scholar with m., pp. 100–1.

²⁶ Gray, L. H., ‘The double nature of the Iranian archangels’, AEW, VII, 1904, 345–72Google Scholar.

²⁷ See Haurvatāṯ et Ameretāṯ, essai sur la mythologie de I'Avesta, Paris, 1875Google Scholar; and Ormazd et Ahriman.

²⁸ See Die Amoša Spantas, ihr Wesen und ihre ursprüngliche Bedeutung (Sb. KA W in Wien, Phil.-hist. Kl., CLXXVI, 7), 1916Google Scholar.

²⁹ See Lommel, H., Die Religion Zarathustras nach dem Awesta dargestettt, Tübingen, 1930, 120Google Scholar ff.; and on Y, XLIV, 18Google Scholar, further Gershevitch, I., The Avestan hymn to Mithra, Cambridge, 1959, 201Google Scholar.

³⁰ See Geiger, , op. cit., 245Google Scholar; Gershevitch, , AHM, 12Google Scholar.

³¹ See Duchesne-Guillemin, J., La religion de l' lran ancien, Paris, 1962, 200Google Scholar. The interpretations which identify the seven Gāthic divinities with the Ādityas, or find in them representatives of the ‘trois fonctions’ (neither of which has won general acceptance) are dealt with trenchantly by Zaehner, R. C., The dawn and twilight of Zoroastrianism, London, 1961, 49–50Google Scholar. On the latter theory see also Gershevitch, , AHM, p. 48Google Scholar, n.

³² On Xšaθra and the sky see Zaehner, , The teachings of the Magi, London, 1956, 32–3Google Scholar. For fire ‘in all creation’ (andar harwisp dahišn) see Or. Bd., iii, 8, and cfGoogle Scholar. the more sophisticated Zādspram (ed. Anklesaria, B. T.), i, 25Google Scholar. The phrase xraoždištƏng asēno ‘hardest stones’, used of the sky in Y, xxx, 5, is paralleled in Yt., xiii, 2 by x^vaina- ayah- ‘*bright metal’. As Bailey has shown (Zoroastrian problems in the ninth-century books, Oxford, 1943, 125 f.)Google Scholar the term ‘metal’ included, in Olr. physics, crystal, which is used in several Pahl. passages for the substance of the sky. If these are all expressions of a unified tradition, then the Gāthic verse is in harmony with the doctrine which assigns the sky to the protection of Xšaθra. In speaking of the sky as of stone Zoroaster himself was evidently using inherited IE terminology (see, most recently, Biezais, H., ‘Der steinerne Himmel’, Ann. Acad. Beg. Scient. Uppsaliensis, iv, 1960, 5–28)Google Scholar; whereas the expression in Yt., xiii, appears to represent the particular Iranian determination of the tradition concerned. The link between Xšaθra, lord of metal, and the sky is confirmed in the Pahl. books (see Gr. Bd., iii, 16)Google Scholar, but was evidently allowed to recede generally from prominence with the development of other theories about the heavens. It is presumably for this reason that the association of the AmƏša Spantas with the gahāmbārs is not stressed by later theologians; but the fact that these festivals are holy days of obligation proves their fundamental religious importance.

³³ See Y, xxxiv, 11Google Scholar; Li, 7; XLIII, 9. It is presumably because haoma is only one of the vegetable offerings that the deity of this plant is not invoked in the Gāthās. In the developed theology he and AmƏrƏtāt have much in common.

³⁴ See Modi, , op. cit., 292Google Scholar; Bartholomae, , Air. Wb., 1735–6Google Scholar.

³⁵ The facultative use of fine metal rods for the barsom is evidently of late date. Twigs are still usually used in Iran.

³⁶ This is given as th e doctrinal reason why the priest seats himself cross-legged, and why, if illness or age mak e this impossible for him, he may no longer celebrate the rituals.

³⁷ See above, p. 25, n. 18. That stone vessels were used at least into the Achaemenian period is shown by the mortars, pestles, and bowls of green chert found in the Persepolis treasury, see Schmidt, E. F., Persepolis, II, Chicago, 1957, 53 ff.Google Scholar, with pi. 23. A bronze mortar an d bronze pestles were also found in the treasury, see ibid., p. 9. Vessels both of stone and metal are of course mentioned in th e ‘Younger ‘Avesta.

³⁸ See šāyist nē-šāyist, xv, 5 (ed. Davar, M. B., 84Google Scholar; cited by Gray, , art. cit., 346–7)Google Scholar; Gr. Bd., iii, 12 (ed. Anklesaria, T. D., 33.15 f.Google Scholar; variant readings apud Anklesaria, B. T., Zand-ākāsīh, Bombay, 1956, 38Google Scholar; cited by Gershevitch, , AHM, 11)Google Scholar; Pahlavi, Sīrōze, I, 1 (ed. Dhabhar, B. N., Zand-i Khūrtak Avistālc, Bombay, 1927, 160.5 f.)Google Scholar.

³⁹ My own attention was first drawn to this traditional interpretation by behdīns of the village of Sharīfābād in Iran. To Irani priests the religious symbolism is naturally even more vividly present. In India, according to Ervad Dr. Firoze M. Kotwal, the symbolism is now understood only by the learned; but the traditional interpretation is preserved there also, in accord with Irani teaching.

⁴⁰ cf. Molé, , op. cit., 264, 265Google Scholar; Humbach, H., Die Gathas des Zarathustra, Heidelberg, 1959, 1, 73–4Google Scholar.

⁴¹ Darmesteter, , Haurvatāt et AmƏrƏtāṯ, 79Google Scholar.

⁴² ibid., 41.

⁴³ Darmesteter, , Ormazd et Ahriman, 14Google Scholar. See further ibid., 284, and cf. Molè, op. cit., 207; Kniper, F. B. J., ‘Remarks on The Avestan hymn to Mithra’, IIJ, V, 1, 1961, 41–2Google Scholar.

⁴⁴ Darmesteter, , op. cit., 17Google Scholar; see further Gershevitch, , AHM, p. 153, nGoogle Scholar.

⁴⁵ For a summary of various interpretations of Rta/Aša see Duchesne-Guillemin, , La religion, 194–6Google Scholar; and, in more detail, in ‘Heraclitus and Iran’, History of Religions, III, 1, 1963, 43–7Google Scholar.

⁴⁶ 6 cf. Y, XLIII, 16Google Scholar; XLVIII, 5. Even if one follows Bailey, Festschrift für W. Eilers, in separating Av. Ārmaiti from Vedic arámati, the concept postulated appear s to be similar.

⁴⁷ See Modi, , op. cit., 269–70Google Scholar. In the absence of an Indian parallel the antiquity of the varas cannot be proved; but since it is evidently the ritual vestige of a hair-strainer (cf. Vr., x, 2)Google Scholar, its ancientness is probable. Apart from the varas, cattle must be represented at the yasna by offerings of milk, and of meat, butter, or fat.

⁴⁸ Even in a village where there is only one priest, and the yasna cannot properly be celebrated, the ritual use of gōmēz makes the keeping of bull or cow obligatory.

⁴⁹ See Darmesteter, , Ormazd et Ahriman, 9, 248Google Scholar; and, recently, Schmid, W. P., ‘Die Kuh auf der Weide ’ IF, LXIV, 1, 1958, 12Google Scholar.

⁵⁰ See Cameron, G. G., ‘Zoroaster the herdsman’, IIJ, x, 4, 1968, 261–81Google Scholar. The Manichaean missionaries adopted the Zoroastrian Vahman to represent their divinity the Nous, tutelary spirit of the Manichaean church (cf. Y, XLIX, 5), itself called, in Zoroastrian-sounding terminology, ‘the flock (or herd) of the just’ (ram ardāwīīft).

⁵¹ On GƏ uš Urvan, and the uniting with this being of animals at death, see Gr. Bd., xxvi, 26 (ed. T.D.A., 166.1–2) and iii, 14 (ed. T.D.A., 35.5–6). GƏuš Tašan is a perplexing divinity; but it seems unlikely (despite the carefully reasoned arguments of Gershevitch, , AHM, 55 f.)Google Scholar that he is to be identified with Spanta Mainyu. Zoroaster's followers appear to have had a good grasp on the interrelationship of the Gāthic divinities, including the subtleness of the link between Spanta Mainyu and Ahura Mazda; and GƏuš Tašan remains ritually a separate entity throughout Zoroastrian tradition. As Bartholomae says, Sīrōze, I, 14, appears decisive on this point. See also Lommel, , Die Religion, 178Google Scholar. In Irani Zoroastrianism certain sacrifices are still devoted to GƏuš Tašan and GƏuš; for the dedicatory mąθra see Unvala, M. R. (ed.), Dārāb Hormazyār's Rivāyat, Bombay, 1922, 1, 263.7–10Google Scholar.

⁵² On the bull-sacrifice in Iran see Boyce, , ‘Bībī Shahrbānū and the Lady of Pārs’, BSOA8, xxx, 1, 1967, 42–3Google Scholar. In India the soma-eeremony with animal-sacrifice is still solemnized in Kerala, the Godavari region in Andhra Pradesh, and the Tanjore District of Madras State (information which I owe to the kindness of my colleague Mr. J. E. B. Gray). The persistence of this observance in India, despite the doctrine of ahimsā, proves how strongly it is rooted in Brahman tradition. It s continuance at the heart of Iranian religion is therefore to be expected.

⁵³ See Dubois, , op. cit., 511, 512Google Scholar; Thieme, , art. cit., 78Google Scholar. On the old custom of killing a cow in honour of a guest see Pāraskara, Grihyasūtra, 1.3.26 (cited by Hopkins, B. W., Cambridge History of India, i, 232–3)Google Scholar.

⁵⁴ See Dubois, , op. cit., 110–11Google Scholar.

⁵⁵ The Zoroastrian evidence on this point has been brought together by the present writer in ‘Haoma, priest of the sacrifice’, in the Henning memorial volume; for the Indian evidence see, e.g., the Aitareya Brāhmana, n.l (transl. Keith, p. 142). The Irani Zoroastrians continued down to the present century to eat meat only when an animal was offered in sacrifice. To kill a creature without so consecrating it would have been to commit the sin of būdyōzadīh ‘destroying existence’.

⁵⁸ For the Zoroastrian practice see Nīrangistān, ed. Sanjana, , fol. 128 v, 1. 9 fGoogle Scholar.

⁵⁷ See, e.g., the Aitareya Brāhmana, n.7, transl. Keith, p. 139: ‘The victim as it was borne along saw death before it, and was not willing to go to the gods; the gods said to it, “Come; we shall make you go to the world of heaven”. It replied “Be it so; but let one of you go before me”. “Be it so” (they replied). Before it went Agni; it followed after Agni’.

⁵⁸ The only Gathic passage carrying any real suggestion to the contrary is Y, xxxn, 8, whose interpretation is difficult. It s connexion with Yima, who throughout Iranian tradition is held to have sinned, by disobedience, doubt, or pride, suggests that there is question here of wrong intent, of sacrifice without Vohu Manah, a dužyašti. On this see Bartholomae, , Air. Wb., 1866Google Scholar, and hi detail Molé, , op. cit., 221 ffGoogle Scholar.

⁵⁹ āzūiti is taken as ‘oblation of fat’ by Humbach, , Die Gathas, 1, 82, n, 17Google Scholar(see in more detail his ‘Milchprodukte im zarathustischen Ritual’, IF, LXIII, 1, 1957, 50–1)Google Scholar; and by Zaehner, , Dawn, 34 with p. 325, n. 8Google Scholar. See also Gershevitch, , JRAS, 1952, 178Google Scholar. There is no evidence for an offering of butter to the fire in Iranian ritual. The only attested oblation is of fat from a sacrificial animal. A purely figurative interpretation of Y, xxix, 7, is suggested by Cameron, , art. cit., 278Google Scholar.

⁶⁰ Transl. Keith, p. 136.

⁶¹ For a different rendering of hvō.urušaēibyō see Humbach, , Die Gathas, 1, 82Google Scholar. For details of the two offerings see Boyce, , ‘Ataš-zōhr and āb-zōhr’ JRAS, 1966, 100 ffGoogle Scholar.

⁶² Transl. Keith, p. 136.

⁶³ Kausītaki Brāhmay, a, x.6 (transl. Keith, p. 408 of the same volume as the Aitareya Brāhmava, see p. 25, n. 16, above).

⁶⁴ The references, for example, to the ‘cow in calf’ lead one away from the animal-offering to the symbolism of husbandry, since no animal with young may be slain. On the spiritual symbolism of Zoroaster's cattle-imagery in general see Cameron, art. cit.

⁶⁵ See Benveniste, E., ‘Deux noms divins dans l'Avesta’, RHR, cxxx, 1945, 13–14Google Scholar; ‘Éléments perses en araméen d'Égypte’, JA, coxlii, 3–4, 1954, 304Google Scholar.

⁶⁶ See, notably, , op. cit., 104Google Scholar. Easton, Earlier M. W. (‘Some divinities of the Gāthās‘, JAOS, xv, 1893, 197–8)Google Scholar wrote: ‘The peculiar difficulty… in the manner in which quality, personification, and person are considered to blend with each other, is common to all hermeneutics and depends less on the character of the special strophes than on the attitude assumed by the reader.… Considering the unmistakable evidence afforded by some of the passages [of the Gāthās], it is not easy to see what character could have been assigned… to the simple, unqualified, and monotonously repeated asha and vohumano other than that of… personality‘. Cf. Benveniste, , RHR, cxxx, 1945, p. 14, n. 3 (on Sraoša)Google Scholar; Humbach, , Die Gathas, I, 37–8Google Scholar.

⁶⁷ Only those Parsis who are under Western influence have come to accept the idea of the Gathic divinities as either abstractions, or aspects of God. Orthodox priests, in India and Iran, are strongly opposed to this, and hold that they are hōšdār beings, with their own independent existence.

⁶⁸ Y, iii, 20, iv, 23, LVII, 33; Vd.,.xviii, 14; Sīrōze, I, 17. Tanumθ is an epithet which Sraoša shares with only one other divinity, his great overlord Mithra (Yt., x, 25); and it appears to be one of the few attributes which Mithra borrows from him, instead of vice versa. On the translation of the term see Gershevitch, , AHM, 180–1Google Scholar.

⁶⁹ cf. Y, XLTV, 18; and see Lommel, , Festschrīft für W. Kirfel, 187–95Google Scholar; Humbach, , Die Gathas, I, 59–61Google Scholar. The Zoroastrian ašūdād, although no part of the ritual objects, is never lacking; and in the Brāhmanas the dakṣῑnā is considered essential to the sacrifice, see Lévi, , op. cit., 90–1Google Scholar. For Aši duly following Sraoša cf. Y, XLIII, 12.

⁷⁰ In the Brāhmanas there is insistence on offering the sacrifice with śraddhā ‘La confiance eat nécessaire à ce point que, sans elle, le sacrifice est stérile’, Lévi, , op. cit., 113Google Scholar; cf. Yt., x, 9. For verbal variations in the Gāthās on the theme of thought, word, and deed see Humbach, , Die Gathas, I, 55–6Google Scholar. Humbach observes, ibid., 73, ‘Die frommen Gedanken, Worte und Werke der Menschen beim Opfer sind der Ausdruck seiner frommen Haltung im täglichen Leben’. The connexion seems justly made, but with the priority reversed. Cf. Darmesteter, , Ormazd et Ahriman. 8 ffGoogle Scholar.

⁷¹ Lévi, , op. cit., 9Google Scholar. Cf. also ibid., 129 f., 152 f. On p. 152 Lévi speaks of ‘la superbe indifférence morale des dieux, ou plutôt de la théologie brahmanique’.

⁷² Y, xxvn, 11; cf. Y, xxxm, 14. Rudolph, , art. cit., 102Google Scholar, has cited the Gāthic verse as proof that Zoroaster abandoned the blood-sacrifice for an ethical offering of the self; but the two are not incompatible, as Judaism and Islam both testify. The double offering remains valid through out the history of Zoroastrianism. The Dēnhard, e.g., both contains prescriptions for tangible sacrifices, and enjoins: ‘Everyone should make offering of himself and deliver himself to the gods’, harw kas tan xwéš ustofrid be kunišn ud ō yazdᾱn abspᾱrišn (Dk., ed. Madan, D. M., 478. 22–3Google Scholar; cf. ibid., 522.6–8).

⁷³ See Gershevitch, , AHM, 286Google Scholar.

⁷⁴ See ‘On Mithra's part in Zoroastrianism’ BSOAS, xxxn, 1, 1969, 26–7Google Scholar.

⁷⁵ Gershevitch, , AHM, 193Google Scholar.

⁷⁶ See Bailey, , Zoroastrian problems, 66 f.Google Scholar; Gershevitch, , ABM, 217–18Google Scholar.

⁷⁷ Dhabhar, B. N. (ed.), Pahlavī Rivᾱyat accompanying the Dᾱdistᾱn ī dēnīg, Bombay, 1913, lvi, 3Google Scholar; for other passages see Dhabhar, , The Persian Bivayals of Hormazyar Framarz, Bombay, 1932, 166–7Google Scholar with notes. Sraoša is nevertheless also indwelling, being the opponent in the mind of the demon Wrath.

⁷⁸ Y, xxx, 9; xxxi, 4.

⁷⁹ In Y, xxxi, 4, the Ahuras are invoked together with Aša, and in connexion with Aši and Ārmaiti. Zaehner, , Dawn, 39–40Google Scholar, identifies the Gātbic ahuras with the asuras, but supposes that, although Zoroaster invoked these divinities, ‘he found them inconsistent with his own religion and therefore studiously avoided mentioning them by name’. It is wholly incredible, however, that a prophet of Zoroaster's burning faith should invoke other deities together with Ahura Mazda without holding them in reverence.

⁸⁰ DB IV.61. Moulton, J. H., Early Zoroastrianism, London, 1913, p. 352, n. 2Google Scholar, accepting this comparison, commented: ‘Provided that we limit the Ahuras to Mazdāh and the Six, with the other Gathic abstractions of the same class, we do not compromise Zarathushtra's unmistakable monotheism’. But (a) the Gāthic ‘abstractions’ cannot all be classified together; (b) there is no evidence for such a limitation of the term ahura by the prophet; (c) so far is Zoroaster's monotheism from being ‘unmistakable’ that it needs special pleading to establish it in any strict sense from the Gāthās.

⁸¹ See BS0A8, XXXII, 1, 1969, 18Google Scholar.

⁸² Such a particular dedication is ritually necessarȳ, see above, p. 25, n. 21. It is presumably the occasional element in Y, LI, 22, which made it impossible for the yeƏhē hᾱtᾳm prayer, designed for general use, to be modelled on it with absolute fidelity. Henning himself had no explanation to offer for the significance of the initial yehyᾱ but he was not a scholar who allowed attempts at interpretation to distort grammar. The yeƏhtē hᾱtąm he understood to mean: ‘We shall worship those among the existing gods whom Ahura Mazda knows to be best for worship‘. (Verbal communication made repeatedly during expositions of the yašts.) Cf. Nyberg, H. S., Die Rdigionen des alten Iran, trs. Schaeder, H. H., Leipzig, 1938, 270Google Scholar.

⁸³ On the general preliminary dedication to Ahura Mazda see BSOAS, XXXII, 1, 1969, 30, 32Google Scholar. When a prophet preaches monotheism in the strict sense, there are to be expected reiterated declarations of the existence of only one God, and denials of that of any other. One has only to compare the Gāthās in this respect with the utterances of the Hebrew prophets, or Muhammad, to be struck by the total absence from them of any such declarations. The problem appears, however, to be largely one of terminology, since it is not easy to find a word to define the unique teachings of Zoroaster. The Parsis, conscious of the supremacy of Ōhrmazd over all the good creation, divine and earthly, and battered by the verbal assaults of Muslim and Christian, have indeed themselves resorted to a declaration of monotheism as the best way of describing their own beliefs; but this has never led the orthodox to modify the actual doctrines of their faith, or to abandon the veneration of lesser divine beings, as practised evidently by their prophet. It is juddīns who have attempted to reconcile an apparent contradiction by postulating strict monotheism for Zoroaster, followed by syncretism to account for the beliefs of his followers. This has led, however, to another contradiction. It is incontestable that Zoroaster affirmed the existence of false gods, the daēvas, with power under Angra Mainyu to delude mankind; and yet it is supposed (against the evidence) that he denied that of their traditional opponents, the righteous ahuras. Thus by the monotheistic theory Ahura Mazda in his majesty is alone against a host of enemies—a strange interpretation of the doctrines of Zoroaster, with their balanced opposition of the forces of good and evil.

⁸⁴ Y, XLI, 3; see Wesendonk, O. G. von, Die religionsgescMchtliche Bedeviung des Yasna Hapta hᾱti, Bonn and Köln, 1931, 45Google Scholar. For the general meaning of the term Aməša Spanta in Yasna Hapta hᾱiti, see Y, xxxix, 3 and XLH, 6. Nyberg, , op. cit., 280Google Scholar, has justly commented that from these passages it appears ‘dass Aməša Spənta hier noch eine allgemeine Bezeichnung für Götter und göttliche Mächte ohne Begrenzung auf eine bestimmte Zahl ist’. Cf. Zaehner, , Dawn, 63Google Scholar.

⁸⁵ Even here the expression has been interpreted as limited to the seven Gāthic divinities, see von Wesendonk, , op. cit., p. 49Google Scholar with n. 2, Gershevitch, , AHM, 163 f.Google Scholar; but the arguments, although ingenious, appear a trifle forced.

⁸⁶ See, e.g., Y, XLII, 1.

⁸⁷ There are passages in the yasna and yašts where the meaning may be taken either as particular (the six with Ahura Mazda) or general. The general meaning is especially striking in some mythological passages, e.g., Yt., x, 51 (the Aməša Spəntas built Mithra's abode above high Harā); Yt., x, 89 (Haoma is ‘priest of the Aməša Spəntas’ i.e. of the gods); Yt., vii, 3 (the Aməša Spəntas distribute the moon's radiance over the earth). In Y, i, 2, Ātar is called ‘the most active of the Aməša Spəntas’ and in the often-cited Vr., VIII, 1, the Aməša Spəntas are numbered by hundreds and thousands. In the Pahlavi books, and in living Zoroastrianism, the term is regularly applied to the 33 divine beings associated with the calendar, and to other yazatas as well. Geiger, , Die Aməša Spəntas, 87 ff.Google Scholar, while acknowledging the general use of the term, interprets the facts in the opposite way. He suggests, that is, that the term was first applied to the seven, and then occasionally extended to other yazatas, as (p. 88) ‘eine leicht erklärliche, durch Üngenauigkeit verursachte Übertragung des Namens auf ausserhalb stehende Gestalten’. The evidence of Yasna Hapta hᾱiti suggests, however, that this ‘Ungenauigkeit’ is original to the term, and that its limitation was a theological development within Zoroastrianism.

⁸⁸ cf. Gr. Bd., xxvi, 125 (ed. , T.D.A., 180.14–15)Google Scholar: ‘Ōhrmazd and the six Amahraspands are the source of all which one calls Amahraspand’ (Ōhrmazd ud ᾱn 6 amahraspand bun ῑ hamᾱg ῑ amahraspand gōwēd).

⁸⁹ See, e.g., Gr. Bd., xxvi.

⁹⁰ The same categories of ritual objects are used at this lesser service, although haoma and varas are lacking. Even at the ᾱfrῑnagᾱn, Amərətāt is represented by flowers and fruit, and Vohu Manah by the offering of milk.

⁹¹ All six are further honoured by their special feast-days, recurring annually. There is no ground whatsoever for postulating a neglect of the six Aməša Spəntas in the practice of the later religion (pace Zaehner, , Dawn, 96)Google Scholar.