¹ For the purposes of a study of the history of the prescript which I hope to publish elsewhere.

² Apart from abbreviations in normal use I shall employ the following: BM = Bradeen, D. W. and McGregor, M. F., Studies in Fifth-Century Attic Epigraphy (University of Oklahoma Press, 1973)Google Scholar; D = ATL II; ML = Meiggs, R. and Lewis, D. M., A Selection of Greek Historical Inscriptions to the End of the Fifth Century B.C. (Oxford, 1969)Google Scholar. In referring to inscriptions in the Corpus I omit the letters I.G.: thus i²39 = Inscriptions Graecae vol. I² n. 39.

³ See Denniston, J. D., The Greek Particles ² (Oxford, 1954) 190Google Scholar ff.

⁴ All quoted by Denniston, loc. cit.

⁵ See Dover, K. J., ‘Der Stil des Aristophanes’, in Aristophanes und die alte Komödie (Darmstadt, 1975) 138–9Google Scholar.

⁶ i²39 = D 17 = ML n. 52. I quote from ML.

⁷ i²10 = D 10.

⁸ Note that Meiggs, and Lewis, do not accept the extremely doubtful ούδέ (connective) + ποτε (vv. 25–6)Google Scholar of ATL.

⁹ Cf. e.g., ML n. 73, vv. 46 and 54 ff.

¹⁰ i²57 = D 3/6.

^10a The restoration originally stems from Kirchhoff, , Abh. Berl. Akad. (1861) 561Google Scholar n. 2.

^10b See Denniston, op. cit. 194–5.

¹¹ Cf. e.g., ML n. 40 v. 10: ii²558 vv. 19–20:

¹² i²22 = D II.

¹³ Op. cit. 63–5.

¹⁴ Op. cit. 58.

¹⁵ I opt for 3 letters in 2 spaces rather than 2 in 1 since H and O are both wide letters.

¹⁶ Op. cit. 31–2.

¹⁷ i² 50 (+i² 102) = D 18.

¹⁸ ATL II pl. XI; Meritt, B. D., Athenian Financial Documents of the Fifth Century (1932) 53CrossRefGoogle Scholar fig. 9.

¹⁹ See Meritt, op. cit. 54. The Corpus had assumed a line of 44 letters.

²⁰ As in D 10 v. 23 and D 15 v. 45.

²¹ As in D 15 v. 45.

²² See Section III below.

^22a A more radical approach to the problem would be to reject the (inferred) 35 letter line from this section of the inscription. Wankel, H. (ZPE xv [1974] 249–54)Google Scholar, in arguing against the restoration of the formula in epigraphical texts of the fifth century, is forced to the same conclusion (p. 253): ‘wird wahrscheinlich die Frage der Zeilenlänge und der Ergänzungen in Z. 15 ff.…überhaupt neu überdacht werden müssen.’ (Wankel does not touch on the problem of the negative at the end of v. 15.)

²³ i²50 (+i²102) = D 18. See quotation under I (c) above.

²⁴ οὐδέ is correctly restored by Meritt, in Athenian Financial Documents of the Fifth Century 54Google Scholar.

²⁵ 1₂12/13a. See now Engelmann, H. and Merklebach, R. edd., Inschriften griechische Städte aus Kleinasien I: Die Inschriften von Erythrai und Klazomenai (Bonn, 1972)Google Scholar Part 1 n.4.

²⁶ I am not concerned here with the problem of whether or not this fragment belongs to the same settlement with Erythrai as i²10. (See ML op. cit. 92–3.)

²⁷ Note that in the Corpus Hiller correctly restores with Kirchhoff's οὐδέ.

²⁸ i²14/15 = ML n. 47.

²⁹ Meiggs and Lewis, loc. cit., who offer a less ambitious text, show basically the same approach to the co-ordination of the passage.

³⁰ Op. cit. ch. V pp. 97–8.

³¹ I deliberately refrain from describing this text as a proxeny decree for Praxiadas of Kos, since the two fragments tentatively published together in the Corpus do not belong together. Michael Walbank, who examined the stones in 1971, informs me per epistulam that there are differences in the stone on which these two fragments are cut and perhaps also in the treatment of the backs (if both are original). Doubtless, however, the honours involved an award of proxenia.

³² The stoichedon pattern is rigid (iota always occupying a full space). The length of line is easily determined from vv. 10 and 11, where the restorations are certain.

³³ Published by McK, John. Camp II, Hesperia xliii (1974) 322–4, ‘Proxenia for Sopatros of Akragas’Google Scholar.

³⁴ During the summer of 1975. My best thanks are due to Dr D. Peppas-Delmousou and her assistant, Miss Ch. Karapa, for their great helpfulness in enabling me to examine this stone, inter alios, in the Epigraphical Museum.

³⁵ Though the paroxytone militates against such a charitable assumption.

³⁶ See my article in CQ n.s. xvi 2 (1966) 295–6. Note that Wilhelm correctly eschews the article in vv. 6, 11 and 16.

³⁷ See note 33 above. Cf. also ii²407 vv. 7–8:

³⁸ The transition is made easier by pointing with a semi-colon after κατάγηται.

³⁹ Not elsewhere (to my knowledge) attested in Attic decrees, but not absolutely unparalleled in Attic prose: see Denniston, op. cit. 193.

⁴⁰ is treated by Croix, G. E. M. de Ste., The Origins of the Peloponnesian War (1972) 47Google Scholar and especially Appendix VIII, p. 314. (I owe this reference to Dr D. M. Lewis). It is perhaps significant that in almost all the examples he quotes the voice is active: B's vessels’, where there can be no doubt that B is acting under coercion. But when the verb is used in the passive, surely a little more is required to show that A is being forced into port, not landing of his own accord. Cf. ii²360 vv. 35–6 (315/4): Wilhelm may well have felt that the force of carried over into the second half of the co-ordination.

⁴¹ See Denniston, op. cit. 511 and K-G. II ii 292.

⁴² Drafters of inscriptions seem frequently to have been careless. But no doubt less accuracy was demanded of them than of their modern counterparts.

⁴³ See K–G. II ii 566–7: ‘So ist aus der Begriff von oder zu entnehmen.’

⁴⁴ I have to thank Professor K. J. Dover and Mr A. G. Woodhead for sharing with me the frustrations of this seemingly innocent stone. I have profited greatly from discussion with them on this text in particular and on epigraphical negatives in general. They are, of course, in no way responsible for the views put forward in this paper.