¹ This is clear from Hude's text and apparatus in the OCT, but, for an egregious example, see especially Powell's, J. E.edition, Herodotus Book VIII (Cambridge, 1939).Google Scholar

² Ibid. The only justification offered was the suggestion that ‘the corruption arose from dittography of -eve'.The translation is Powell's, own; see his Herodotus (Oxford, 1949).Google Scholar

³ Grene, David, The History, Herodotus (Chicago, 1987).Google Scholar

⁴ Masaracchia, A., Erodoto. La Battaglia di Salamina. Libro VIII delle Storie (Milan, 1977).Google Scholar

⁵ Legrand, Ph. E., Hérodote. Histoires. Livre VIII Uranie (Paris, 1953).Google Scholar

⁶ Godley, A. D., Herodotus (Loeb, London and Cambridge, MA, 1925).Google Scholar

⁷ Rawlinson, G., History of Herodotus (4th edition London, 1880).Google Scholar

⁸ The Nine Books of the History of Herodotus (Oxford, 1827).

⁹ Goldhagen, J. E., Herodotus. Neun Bücher der Geschichte (Lemgo, 1756), but cited from Klassiker des Altertums, H. Conrad (ed.) (Munich and Leipzig, 1911), Vol. 2, p. 296.Google Scholar

¹⁰ Les Neuf Livres des Histoires de Herodote...traduiet...Pierre Saliat (Paris, 1556), p. CXCVII retro.

¹¹ Cited from the printed edition, Herodoti...Libri Novem...interprete Laurentio Valla (Coloniae, 1562), p. 232.

¹² I do not discuss those which are manifestly inaccurate, or too loose to reveal how the Greek was taken.

¹³ The History of Herodotus (London, 1709), Vol. II, p. 320.

¹⁴ Powell, J. E., A Lexicon to Herodotus (Cambridge, 1938) s.v. κ V.Google Scholar

¹⁵ Powell, Lexicon, s.v. ἒχω, B.2.a.

¹⁶ Renehan R. has argued well and attractively that οἶσπερ should be read instead of ὢσπερ here; see ‘Herodotean Cruces’, HSCP 89 (1985), 25–35, at 29–30. It is certainly true that the use of ὢσπερ here is not parallel to the two other passages put with it in Powell's Lexicon, s.v. 1. These are 1.78.1 and 5.53, in both of which ὢσπερ can simply be translated ‘as’, and has no implication of ‘although’. So I cite and translate this passage in accordance with Renehan's emendation.

¹⁷ One may also compare what Clearchus is reported to have replied to Cyrus just before the battle: (‘Clearchus... replied that he was seeing to it that they would be victorious’). Note also Xen.Anab. 5.8.26, where the better way to understand is ‘and he prevailed so as to be victorious’; cf. F. W. Sturz, Lexicon Xenophonteum (Leipzig, 1801–4; reprinted Hildesheim, 1964) s.v. περιγγνεσαι; though the erroneous interpretation,‘it turned out so that it was well’, is often found; e.g. ‘and in the end all was pleasant’ (C. L. Brownson, Loeb).

¹⁸ This text is regularly emended; see W. Mader, Die Psaumis-Oden Pindars (0.4 und O.5). Ein Kommentar (Commentationes Aenipontanae XXIX, Innsbruck, 1990), 88–89; but Farnell's cautious and sensible defence of the tradition is persuasive and attractive; see Farnell, L. R., The Works of Pindar: Critical Commentary (London, 1932) 39.Google Scholar

¹⁹ Mader does not think that this gnome refers to the old idea that no one is a prophet in his own country; see 89–90. His reasons are that κα should not be overvalued and that the connection of polis and victor runs through the whole poem, but these seem slight compared with the evidence, which Mader himself gives, that πολῖται usually have bad connotations in Pindar, and with the requirement that the gnome have some point and meaning. Mader's commentary is no help on εῒ δ’ ἒχoντες, because he accepts the emendation εῒ δ’ ἒχoντες, and comments on that.

²⁰ Abel, E., Scholia Recentia in Pindari Epinicia (Budapest and Berlin, 1891), Vol. I, pp.192–3Google Scholar

²¹ Lexicon, s.v.

²² Compare also the meaning of the noun πρoαγóρευσις in Arist. Poetics 1454 b5: ⋯λλ⋯

²³ History of Herodotus, Vol. II, p. 456.

²⁴ Selincourt, A.de, Herodotus. The Histories (Harmondsworth, 1954), p.199.Google Scholar

²⁵ Herodotus. A New and Literal Version (London, 1849).

²⁶ Herodotus (Oxford, 1949).

²⁷ Note, incidentally, that Pausanias felt no need to employ the meiosis of εῢ ἒχοντα κτλ. but plainly writes the proud word νκην.

²⁸ As was seen by How, W. W. and Wells, J., A Commentary on Herodotus (Oxford, 1912), ad loc.Google Scholar

²⁹ I owe this point to D r Karl Maurer.

³⁰ See n. 13 above. ‘A New Edition, corrected’ was published as late as 1818 (Oxford), not to mention the reprint of the Third Edition (London, 1737) in 1976 (New York: AMS Press).

³¹ Herodotos erklä von H. Stein, 5th edition (Berlin, 1893: reprinted Dublin/Zurich, 1962); also in his Editio Maior (Berlin, 1884).

³² Many modern editors follow Stein's suggestion that ῥματα be excluded; Herodotos erklärt von Heinrich Stein, first edition(Berlin, 1862) and all subsequent; also in Editio Maior. When Stein proposed this, he thought that someone had inserted the word who did not understand Herodotus' usage of a plain genitive for a speaker. He cited 4.32, where he then read the genitives ‘Hσιóδυ and ‘Oμρν instead of the datives which are correct. In later editions he merely cited 1.109.1. . This passage in fact rather supports the pleonasm of 8.83.1. Legrand subtly defended the text on the grounds that the word ῥματα has the special meaning of affirmations accompanied by oaths (see the edition cited in n. 5 above, p. 78 n. 1), but the pleonasm hardly seems to need defence.Google Scholar

³³ Some editors prefer to write ὡς τε δ διϕαινε on the basis of the reading τε δ ἒϕαινε found in some manuscripts; see B. Hammer, De τε. particulae usu Herodoteo Thucydideo Xenophonteo (Leipzig, 1904), 8, who prefers that reading both here and at 7.217.1. The τε certainly looks forward to responsive κα here (against Hammer), as is shown by the similar expression at 7.44.1, . Compare also Xen. Anab. 3.2.1. But the responsive use of τε does not create asyndeton, since the particle can do double duty, serving both as a connective and as looking forward to κα, as Stein, note to 6.41.1, with many other examples, including 8.83.1, and Powell (n. 1 above) ad 8.64.1.

³⁴ The apparent anacoluthon, or hanging participle, here led Legrand to postulate a lacuna, but the transmitted text is correctly analysed by R. Kühner and B. Gerth, Ausführliche Grammatik der griechischen Sprache, Zweiter Teil: Satzlehre, Erster Band, 3rd edition (Leipzig, 1898; reprinted Hannover, 1966), 288–289, as an example of concealed partitive apposition, in which the subject of the finite verb is the part, and the whole is expressed by a participle standing in the same case (here nominative). More Herodotean instances of this usage were assembled by I. A. Heikel, De participiorum apud Herodotum usu (Helsingfors, 1884), 124. There is a striking example of the same construction, also connected with public speaking, by Xenophon at Hell. 2.3.54: (‘they entered with their attendants, led by Satyrus, the boldest and most shameless of them, and Critias spoke’). Kühner-Gerth explain that at 8.83.1 Herodotus underemphasizes the other speeches at Salamis as irrelevant.

³⁵ The rendering of μν and δ here might seem uncertain, because the words τ..ἒπεα..πντα appear to describe the whole speech, yet are contrasted with . The only analogous passage that I have found in Herodotus is at 7.5.3, where a speech is also briefly described with a μν and δ construction. Hude's text (OCT) runs as follows: . The textual variations are relatively minor and do not affect the use of the passage for this discussion. Rawlinson (see n. 7 above) translates ‘thus far it was of vengeance that he spoke, but sometimes he would vary the theme, and observe by the way that Europe was a beautiful region, rich in all kinds of cultivated trees, and the soil excellent; no one, save the king, was worthy to own such a land’; Powell (see n. 2 above) ‘thus far his argument was of revenge; but he made this addition thereunto, that Europe was an exceedingly fair country and brought forth all manner of garden trees, and was excellent in goodness, and that the king was the only mortal man that deserved to possess it’. The chief difference in these renderings stems from different views of the force of μν and δ and of the weight to be given to the iterative ποιεσκετο (Powell's ‘argument’ for λόγος seems tendentious and against the implication of παρενκην. Stein's idea that τιμωρς ‘helping’, i.e. to persuade Xerxes, seems weak and is against the consensus). This passage is not perfectly parallel to 8.83.1. We are given some of Mardonius' actual words immediately before, and the contrast is made clearer. But it is similar in that the first phrase seems to imply that the whole speech is described, except for the warning μν, and then that is explained by additional description. So, if we use the analogy of 7.5.3, we should follow Littlebury's ‘having first declar'....fram'd all his Discourse’; see the translation quoted above.

³⁶ Legrand (n. 5 above) 78 n. 3

³⁷ R. W. Macan, Herodotus. The Seventh, Eighth and Ninth Books etc. (London, 1908), Vol. I, Part II, 487–8.

³⁸ I am very grateful to my friends, Drs Don Lateiner, Karl Maurer, Martin Ostwald and Tim Ryder, who kindly read this paper in draft, and made helpful suggestions.