¹ This follows Oliverio, 's text published in Riv. di. Fil. vi (1928) 222–32.Google Scholar Many of Oliverio's readings and reconstructions were seriously and sensibly challenged by Ferri, , Historia iii (1929) 389–96.Google Scholar

² Abh. Berl. Akad. Phil. Hist. Kl. 1925, no. 5, 19–24.

³ Riv. di. Fil. vi (1928) 250–4.

⁴ Abh. Berl. Akad. Phil. Hist. Kl. 1925, no. 5, 38–40.

⁵ See, e.g., Busolt-Swoboda, , Griech. Staatskunde ii 1265 n. 3Google Scholar; Bengtson, , Griech. Geschichte. (Munich, 1950) 90 n. 2Google Scholar; Chamoux, , Cyrène sous la monarchie des Battiades (Paris, 1952) 108 ff.Google Scholar

⁶ Thus Wilamowitz, who implies this when he says that the document shows that an account of her old history existed on Thera in the fifth century, and Chamoux.

⁷ Bury, , History of Greece ³ (1951) 862.Google Scholar

⁸ I should like to take this opportunity of thanking those who read an earlier discussion of this document and helped me with suggestions and criticisms, namely Sir Frank Adcock, the late Professor A. W. Gomme, Professor A. Momigliano, and Mr R. Meiggs. I am particularly indebted to Professor A. Andrewes for painstaking and encouraging criticism. Responsibility for all errors and failings remains my own.

⁹ Chamoux, 109 n. 3, does not consider Ferrabino's arguments convincing, but provides no detailed discussion of them.

¹⁰ The Cyrenaean account is dismissed by Chamoux, 95 ff., as a collection of fictional appendages, while Legrand, (Hérodote iv, Budé, 1945, 158 ff.)Google Scholar suggests with reason that Herodotos' informants at Cyrene were (1) favourable to national sentiment, (2) unfavourable to the Battiads, whose rule had ended comparatively recently. This observation is relevant to the question of the name Battos, to be discussed below.

¹¹ Macan, (Herodotus iv–vi vol. ii 265 ff.)Google Scholar attempted to show the scope of the discrepancy in the traditions, but, though he is partly right, I do not think he is completely so. He writes ‘the discrepant traditions concern only the actual foundation story of the Libyan colony, the person of the founder and the action of the first adventurers’; he then gives a perfectly adequate account of the differences and similarities in the two accounts of Battos and Delphi. But his statement (p. 266) ‘it is not perfectly clear at what point exactly the “Kyrenean” story is conceived by Herodotus as ceasing to contradict the “Theraean”’ is, I think, unfair to Herodotos, who, as I have tried to show, does make the matter clear. The reason for Macan's statement is his belief that the variation must go down at least to the abandonment of Platea. For the procedure whereby they left one of their number behind, the rest applying to Delphi (157.1), seems to him to be a variant account of the action described in the Theraean story (151.3), where Korobios is left behind by the reconnaissance party. At first sight these no doubt look like varying accounts of the same event. But if we take them as such difficulties arise. (1) Herodotos did not see them as such, for he does not include the event in the material on which the two traditions differed, the limits of which are, as has been shown, clear. (2) It would leave the Theraean account without an explanation of the move from Platea to Cyrene, whereas if we assume that the agreed account is resumed at 156.2 both accounts tell a complete story, though the Theraean would have two occasions on which a man was left behind. A possible explanation of this might be that it was a regular procedure in stories of colonial foundations. It is better, I feel, to assume two events than to reject Herodotos' exposition of the tradition. The differences are large enough, viz.: in 151.3 Korobios is left behind by a reconnaissance party, in 157.1 the man is left behind after two years' occupation by the colonists; the man Korobios is not a Theraean in 151.3, the unnamed individual of 157.1 is one of the colonists; in the first passage the spies return eagerly to Thera, in the second the colonists appeal dispiritedly to Delphi. One cannot deny, however, that in accounts with many mythical trimmings repetitions and confusions could occur.

¹² iv 153. This assumes that the passage represents an actual decree of Thera, as I show to be probable below.

¹³ Griech. Inschr. Rechtlichen Inhalts (Athens, 1951) 5–7.

¹⁴ There still remains the difficulty of exactly interpreting Neither Chamoux's translation (equal rights ‘pour chaque famille’ p. 107) nor Oliverio's (‘con equali diritti navigare da ciascuna casa’; p. 227) really represent the κατά. Wilhelm's paraphrase (p. 6) is as follows: ‘Gemäss der Zugehörigkeit eines jeden zu seinem οἶκος gewissermassen als Vertreter derselben.’ And to support this historically he adduces the parallel Arist. Pol. 1265b, 12 ff., where the Corinthian lawgiver Pheidon is said to have thought that the number of houses (families) should remain equal. The parallel seems admissible historically, but one wonders if κατὰ τὸν οἶκον can bear so much. Grammatically it must mean either ‘by family’ (LSJ s.v. κατά B) though the plural would be more usual (see examples Ibid.) or ‘for their home’ (LSJ B ‘direction towards an object’) though this seems redundant, or something like ‘according to house (family), etc.’ (LSJ iv), which is the way Wilhelm takes it. This last seems best, but it must be noted that the expressions with κατά in this sense are slightly different.

¹⁵ See, e.g., Legrand, 171 n. 1, who would amend to ἀδελφεῶν thus (incidentally) making Herodotos and the ὅρκιον in agreement.

¹⁶ E.g. Hesiod, , Works and Days 376.Google Scholar

¹⁷ Chamoux, 98, also considers this to be part of the Cyrenaean account. This suggests that he has not noted exactly where the two accounts diverged, in spite of his statement on p. 93.

¹⁸ I have discussed this matter with Miss K. Forbes, who is working on the dialect of Cyrene, and I gather that there are many difficulties in the dialect of the ὅρκιον. But this is not immediately relevant for my purposes here, as I hope to show.

¹⁹ See IG ii² 1.179, fr. c, lines 9 and 11, which can be taken to show this.

²⁰ C.R. Acad. USSR. 1929, 429 n. 2.

²¹ Diodoros xvii 49.3; Curt. Ruf. iv 7.9.

²² Mr John Cassels, who kindly examined the tone on my behalf, assures me of this.

²³ Hdt. i 51.3–4; cf. Volkmann, in Convivium, Festgabe für K. Ziegler (Stuttgart, 1954) 62Google Scholar, ‘Die Inschriften im Geschichtswerk des Herodot’.

²⁴ FGH ii B, 115 F.154.

²⁵ Konon, (FGH i 26)Google Scholar F. 1, no. 29. The tradition is fully discussed by Wilamowitz, , Hermes xxx (1895) 177 ff.Google Scholar See also Inschr. u. Magnesia p. 16.

²⁶ The ὅρκιον was inscribed at Cyrene c. 250 years after its presumed date; Inschr. u. Magnesia no. 20 some 800 years afterwards.

²⁷ Hermes xxx (1895) 191.

²⁸ Griech. Epigraphik 510.

²⁹ See FGH iii B, 491 and Commentary.

³⁰ It is used as a good example of a ‘Ktisissage’ by Schmid, P. B. in Studien zu griech. Ktisissagen (Diss. Freiburg i.d. Schweiz, 1947) 94 ff.Google Scholar

³¹ See Parke, and Wormell, , A History of the Delphic Oracle (Oxford, 1956) i 54.Google Scholar

³² Pp. 190 ff.

³³ See n. 30 above.

³⁴ Or possibly Cyrenaean, though it is slightly against probability and certainly against the principle of economy of explanation to imagine that the Cyrenaeans had a chronicle which the Theraeans exploited.

³⁵ See Jacoby, , Atthis (Oxford, 1949) 364 n. 62.Google Scholar Ion's and Hellanikos' prose ktiseis seem to be the earliest definite examples.

³⁶ Whether or not one accepts Jacoby's conclusions about early Greek historiography (see n. 38 below), he is clearly right to look at all other early writers in their relationship to Herodotos.

³⁷ See n. 6 above.

³⁸ E.g. Atthis 68, 289 n. 111.

³⁹ As Wilamowitz and Chamoux, see n. 6 above.

⁴⁰ This in particular will require further discussion in detail, but for the moment we are testing the document as it stands against the probable conditions of seventh-century Thera.

⁴¹ Clearly shown by the provision about return, 33 ff.

⁴² See BCH lxi (1937) 334 ff., with the correction in BCH lxii ( 1938) 194 f. Published in Buck, , Greek Dialects (Chicago, 1955) no. 116.Google Scholar

⁴³ Guarducci dated it to the beginning of the sixth or possibly end of the seventh century, Riv. di Fil. xvii (1939) 20.

⁴⁴ CQ xxxvii (1943) 14–18.

⁴⁵ Plut. Lycurgus 6. Although his own date seems extremely high Hammond showed the weakness of the argument for a low date (sixth century) in JHS lxx (1950) 42 ff. (see p. 42 n. 3 for bibliography). Andrewes, , The Greek Tyrants (London, 1956) 73Google Scholar, argues sanely for a date in the seventh century. It certainly cannot be later than Tyrtaios.

⁴⁶ Tod, 1; Wilamowitz, , Nordionische Steine (Abh. Berl. Akad. 1909) 64 ff.Google Scholar Arguments for a date between 570 and 550 are advanced by Jeffery, L. H., BSA li (1956) 156 ff.Google Scholar For the contents cf. Tod's commentary p. 2.

⁴⁷ IG xii 3.563 ff. and 762 ff.

⁴⁸ Well stressed by Klaffenbach, , Griech. Epigraphik (Göttingen, 1957) 50 f.Google Scholar

⁴⁹ That it is a title and not the bare record of an oath (like the frequent Attic ὅρκος) seems certain from the larger-sized letters and 18 f.

⁵⁰ E.g. Chamoux, 107; Oliverio, 226.

⁵¹ I exclude such phrases as Argos, , IG iv 557.2.Google Scholar

⁵² See ‘Tituli Calymnii’ by Segre, M., AS Atene, xxii–xxiii (1952) 1.A.1 f.Google Scholar, B. 11 ff.; 2.1 f.; Collitz-Bechtel, , Sammlung der griech. Dialektinschriften 1335.6, 5520.4Google Scholar; IG vii 4250.2 f., 4251.2 f. All these are, more or less confidently, assigned to the fourth century.

⁵³ Tod, 33.44 f.; but see the republication by Vollgraff, W.., Verhand. d.k. Nederl. Akad. van Wettenschappen (Letterkunde), Nieuwe Reeks li, 1948Google Scholar, no. 2. For the assembly see fr. vi 44 f. (p. 9).

⁵⁴ In addition to those in n. 52 I list the following examples: IG xiv 256.7, 612.1, 952.10; these, from the Western Greeks, are not quite parallel as the formula does not stand alone; it is, with slight variations, The same phrase is restored at line 6 of no. 13 of Asylieurkunden aus Kos, Herzog, and Klaffenbach, ., Abh. deutsch. Akad. (Berlin, 1952) no. 1Google Scholar, and the restoration is defended on p. 24 by reference to IG xiv 256.7. But the phrases are not exactly parallel, so there is room for doubt. Also, from Corcyra, we find in IG ix. i 685.1, 686.1, 687.1, 688.1. All these are honorary decrees of the third century or later. The earlier ones in n. 52 are also of this nature. In the case of the Western cities, Corcyra and Leros, we have no comparative evidence, but Calymna provides a large number of similar decrees with different sanction formulas. Nos. 3, 18, 25, 42, 43, 44, 47 have the majority have Here the formula clearly has no distinguishing significance. At Oropos we find, apart from the two mentioned in n. 52 above, two with (IG vii 4256.5 f.; 4257.5 f.). These are dated to the end of the fourth century by Dittenberger (notes to 4256). From a slightly later period onwardsthere is a long series of similar proxeny decrees, IG vii 236 ff., which use the more normal formula. Here we have only a change of fashion, for the two which use are a little earlier (see Dittenberger, notes to IG vii 4250).

⁵⁵ Pp. 38 ff. (see n. 4 above).

⁵⁶ Though Professor Andrewes has drawn my attention to the at Sparta (see Wade-Gery, , Essays in Greek History, Oxford, 1958, 190Google Scholar), whose existence and name might suggest the possibility of an early assembly called at Dorian Thera.

⁵⁷ Cf. Meyer, , Forsch. Z. alten Geschichte i 295 f.Google Scholar

⁵⁸ Buck, 61 ff.

⁵⁹ See BCH lxi (1937) 342.

⁶⁰ The restoration is not certain; see Ferri, , Historia iii (1929) 392.Google Scholar

⁶¹ Cf. Ehrenberg, in RE xvi 1100Google Scholar; Taeger, , Hermes lxiv (1929) 453 ff.Google Scholar

⁶² See Hiller, in RE s.v. ‘Thera’ 2292 f.Google Scholar

⁶³ Parke and Wormell, 74 f.

⁶⁴ P. 155 (n. 10 above).

⁶⁵ Since the former name is so relevant to the argument Herodotos puts forward, the argumenta ex silentio seem admissible. Chamoux's (p. 97) ‘cet autre nom qu'Hérodote ne deigne pas préciser’ seems to gloss over the difficulty.

⁶⁶ Legrand, 155 n.; Chamoux, 95.

⁶⁷ It is a common enough Greek name (cf. Pape, , Griech. Eigennamen ³, 1911, s.v.).Google Scholar To his list add IG vii 1556; SEG i 352.1, ix 174, 175, 176, xii 237.12. The earliest examples, however, are the kings of Cyrene, and it is possible that all the others were derived from them. The earliest non-Cyrenaean bearer is the Corinthian leader in Thuc. iv 43.6.

⁶⁸ P. 156.

⁶⁹ See on this whole question Bertholdy, , Mélanges Boisacq i (Brussels, 1937) 47 ff.Google Scholar

⁷⁰ Naupaktos, Tod, 24; Brea, Tod, 44.

⁷¹ See Larfeld, , Griech. Epigraphik 343 f.Google Scholar His statement that the form of opening with the motivation expressed by an ἐπεί or ἐπειδή clause is found in early documents was supported by his first example only, and this is now differently restored and dated to the fourth century (see Hiller, , IG v 2 xix 65 ff.Google Scholar).

⁷² See Wilamowitz, 40.

⁷³ See the discussion by Hiller, , JdI xlvii (1932) 127 ffGoogle Scholar, where he defends the reading i.e. King Prokles, in preference to taking as a proper name. By connecting it with King Prokles of Epidauros and arguing from letter forms he comes to the date first half of the seventh century.

⁷⁴ Besides Hiller (p. 129) see Risch, , Mus. Helv. xi (1954) 34 n. 59.Google Scholar

⁷⁵ See Guarducci, , Riv. di Fil. xvi (1938) 264 ff.Google Scholar

⁷⁶ See Ehrenberg, , Aspects of the Ancient World (Oxford, 1946) 89Google Scholar, discussing the Harmodios scolium.

⁷⁷ The story of Aithiops, the Syracusan colonist mentioned by Archilochos (see Athen. 167d), merely shows that the colonists all had lots, as we should expect; it does not say that they were equal.

⁷⁸ E.g. the Corinthians, Thuc. loc. cit.

⁷⁹ See Busolt, , Griech. Staatskunde i 250 f., 327 f.Google Scholar

⁸⁰ AJP viii (1913) 133 ff.

⁸¹ A notable example is IG xii.2 15.

⁸² ÖJh vi (1903) 10 ff., especially 12 and 14.

⁸³ These are listed by Colin, , BCH xxiv (1900) 89.Google Scholar

⁸⁴ JHS lxx (1950) 44.

⁸⁵ See n. 23 above.