² Nilsson, M. P., Geschichte der griechischen Religion ³, p. 81Google Scholar: ‘Es ist in der Tat erstaunlich, wie wenig die Griecher, in gegensatz zu vielen anderen Völkern an die wirksame Kraft des Bildes und auch des Names geglaubt haben u.s.w.’.

³ In his commentary on Aeschylus, , Agamemnon (OUP, 1950), line 160.Google Scholar

⁴ The Romans and their Gods, Chatto & Windus, 1969, pp. 24ff.Google Scholar

⁵ ‘De Graecorum precationibus quaestiones’, Neue Jahrbücher Suppl. 28, 1903, pp. 518f.Google Scholar

⁶ A seeming exception is Homer, Il. 10.462, where Odysseus invokes Athene simply as θε. This case is probably a literary feature. The poet has introduced the prayer by saying that Odysseus was praying to Athene. It may well be that, from the audience's point of view, that does the job of the invocation. Alternatively, it could be argued that the omission of the name is a result of the close relationship that Odysseus has with Athene. There is no other θε to whom he is likely to pray.

⁷ According to Augustine, De Civitate Dei IV.22, Varro drew up a list of the minor gods and their different functionsGoogle Scholar. Clearly this does show a legalistic interest in demarcation. I know of nothing similar in the Greek evidence. A passage in Pliny is particularly informative about the power of words. N.H. 28.3. tells us how the use of fixed formulae was seen as essential and we learn that in some ceremonies the prayer was dictated to the presiding magistrate from a script and that a piper played so that nothing but the prayer would be heard.

⁸ The seer does not explicitly say that Xenophon sacrificed to A when he should have sacrificed to B. What he actually says is that Xenophon should be sacrificing to Zeus Meilichios if he wants money. The reader has to remember for himself that up until this point in the narrative, Xenophon has been sacrificing to Zeus Basileus.

⁹ Od. 2.260 and 5.445 are prayers uttered by people who do not know which god is involved. They will be discussed later.Google Scholar

¹⁰ For the evidence of this, in the Lachish Letters, see Encyclopaedia Judaica, s.v. God, Names of.

¹¹ For the use of the name of God in Jewish theurgy, see Swartz, M. D., Mystical Prayer in Ancient Judaism (Tübingen, 1992), p. 18Google Scholar. Also important is Thiselton, A. C., ‘The Supposed Power of Words in the Biblical Writings’, J. Th. S. n.s. vol. xxv, pt.2 (10, 1974), 282–99.Google Scholar

¹² Although we have seen that for the Jews themselves the Divine Name was not originally magical, it is clear that ancient pagans thought it was. The words ‘Adonai’, ‘Sabaoth’, and ‘IAO’ frequently occur in magical texts.

¹³ See Renouf, P. L., Vorlesungen über Ursprung und Entwicklung der Religion, erläutert an der Religion der alten Ägypter, (autorisierte Übersetzung) (Leipzig, 1881)Google Scholar. On p. 181 we read, ‘an vielen Stellen gründet Osiris seine Ansprüche einfach auf seine Kenntnis der Namen der Götter’; p. 211Google Scholar (in a hymn) ‘Dein Name soll mein Schutz sein.’ For the power of the name in Vedic religion and Islam, see the references in Ausfeld, art. cit. p. 519.

¹⁴ On Aen. II.351. Cf. Macrobius S. III.9.4.

¹⁵ One thinks also of the secrecy surrounding the tomb of Oedipus in the Oedipus Coloneus.

¹⁶ That this is a prayer to Isis has been disputed by Griffiths, J. Gwyn, The Isis Liber of Apuleius, ad loc., and in Class. Phil. 63 (1968), 144ffGoogle Scholar. It may just be a prayer to the moon goddess. This makes no difference to our argument. If it were a prayer to Isis, one might suspect the influence of the well-known type of the Isis-aretalogy in the enumeration of the attributes of the goddess. As it is, though, it may simply be a highly wrought variation of the theme given in the other Latin examples. Griffiths also refers to Apul. De deo Socratis 15: cum ergo incertum est, quae cuique eorum sortitio evenerit, utrum Lar sit an Larva, nomine Manem deum nuncupant; scilicet ei honoris gratia dei vocabulum additum est.

¹⁷ Norden, E., Agnostos Theos (Leipzig/Berlin, 1923), pp. 143ff.Google Scholar

¹⁸ op. cit., pp. 143ff.

¹⁹ There is something superficially similar at the beginning of the fragmentary h.Hom.l, to Dionysus. The poet relates several versions of the god's birth, only to reject them and give his own. Unlike the Hymn to Apollo, though, the primary object of this catalogue does not seem to have been praise. Rival versions are rehearsed like this in order to impress upon us that the poet's own version is even more prestigious than these traditional and well known ones.

²⁰ This might refer to present whereabouts but it could equally be a reference to the practice of mentioning the god's place of birth in one's invocation. One can, for example, invoke Athene as Tριτογενεα (Ar. Lys. 347). It is difficult to know which sense is intended here. Both may be meant. One may favour the idea that birthplace is meant if one considers that this is the usual force of the word πθεν in the common phrase τς πθεν εἰ;

²¹ Callimachus is in a similar quandary at Hy. i.4ff. This, though, is rather like h. Hom. 1ff., in that the poet is enumerating different legends about the god primarily in order to suggest that they are false.Google Scholar

²² For further reference to the importance of τ πατρῳον, cf. ∑ AD on Il. 2.371. The Scholiasts tells us that the invocation of the Zeus-Athene-Apollo triad has led some scholars to assume that the poet was an Athenian since it was customary to invoke this triad in Athens. Plato, Euthyd. 302b–d shows how it is your religious affiliations in the phratry that determine under what epithets you worship your gods.

²³ A full bibliography on this point can be found in Richardson, N. J., The Homeric Hymn to Demeter (OUP, 1974), p. 152.Google Scholar

²⁴ Paus. I.i.4; V.xiv.8; Philostratus, Vit. Apoll. VI.iii.5; Tertullian, Ad Nat. ii.9; Lucian, Phil. 9, Zeus Trag. 6.; Diog. Laert. I.x.110.

²⁵ Van den Horst, P. W., ‘The Unknown God’, in Van den Broek, R., Baarda, T. and Mansfeld, J. (edd.), Knowledge of God in the Graeco-Roman World (Leiden, 1991), pp. 19ff.Google Scholar

²⁶ It is very common to find the formula ξορκζω σε followed by catalogues of exotic, foreign names or Ephesia Grammata. The Sethianorum Tabellae (Audollent, , Defixionum Tabellae (Paris, 1904), Nos. 140–87Google Scholar) provide ample attestation of this. See also, for a fuller list, L. Cesano, ‘Defixio’, in Ruggiero, E., Dizionario epigrafico di antichità romane, II, 2, fasc. 49–50, (Rome, 1908–1909), pp. 1575ff.Google Scholar

²⁷ See Dieterich, A., Eine Milhrasliturgie (Darmstadt, 1966), pp. 110ff.Google Scholar

²⁸ In Faraone, C. A. and Obbink, D. (edd.), Magika Hiera (OUP, 1991), p. 192.Google Scholar

²⁹ In Armstrong, A. H. (ed.), Classical Mediterranean Spirituality (RKP, 1986), Ch. 20.Google Scholar

³⁰ I am not, of course, suggesting that Greek religion had any kind of canonically defined orthodoxy. It is surely beyond dispute, however, that there is a world of difference between the public prayers and sacrifices in the traditional religious calendar of a Greek polis and the lone activities of the practitioner of magic. This issue is well discussed by Fritz, Graf, ‘Prayer in Magic and Religious Ritual’, in Faraone, C. A. and Obbink, D. (edd.), Magika Hiera (OUP, 1991), pp. 188ff., esp. pp. 196f.Google Scholar

³¹ On this question, see How, and Wells, , A Commentary on Herodotus (OUP, 1912), pp. 157ffGoogle Scholar. and especially the ref. to Sourdille, C., Hérodote et la Religion d'Égypte (1910), pp. 2–26.Google Scholar

³² On the secrecy surrounding the Eleusinian Mysteries, see Richardson, N. J., The Homeric Hymn to Demeter (OUP, 1974), pp. 304ff.Google Scholar

³³ In the ordinary narrative of the poem, speakers can also leave us in some doubt as to which god was responsible for a particular event (Od. 3.166, 173f.) is counterbalanced by certainty in other cases (Od. 3.119, 160). We might reasonably suppose that the prayer referred to by Nestor in Od. 3.173 was of a similar type to Od. 2.260 and 5.445, i.e. expressed uncertainty about which god to address. Such uncertainty over names is not a conspicuous feature in the evidence as a whole.