² IG i² 19 and 20.1–2=Bengtson, Staatsverträge 139=ML 37. Tod i² 31 does not include the two lines from IG i² 20, but mention is made of them (p. 260). It may be noted that each of these editions contains a minor error in line 17 of IG i² 19, for each indicates that the first remaining letter of it, an epsilon, occupies stoichos 20. Actually, the epsilon is located in stoichos 21, and thus is immediately below the first epsilon in line 16, not the lambda to the left of it. Similarly, each edition also places the only two extant letters of line 18 one stoichos too far to the left: the first of the two letters, a sigma, is located in stoichos 26, not 25. The result of these dislocations is that the letters in lines 17 and 18 have been described accurately relative to one another, but have been misplaced in their own lines and thus misplaced relative to the letters which appear in the preceding lines. These misplacements in the more. recent editions, as in the editio minor of IG, may be attributed to the fact that the IG text has '0 in line 17, stoichos 18, whereas it should have h in stoichos 18 and 0 in stoichos 19. These errors do not occur in the text as originally edited by Köhler, U., Hermes ii (1867) 16Google Scholar; nor in the text as given in CIA (IG) i 20.

³ A recent defence of this position is given by Meritt and Wade-Gery, H. T., JHS lxxxii (1962) 67–74Google Scholar and JHS lxxxiii (1963) 100–17.

⁴ See especially Meiggs, R., JHS lxxxvi (1966) 86–98CrossRefGoogle Scholar.

⁵ ADelt vii (1891) 105–6.

⁶ TAPA lxxv (1944) 10 n. 3. In SEG X 7 (which appeared in 1949) G. Klaffenbach went so far as to claim the reading of an undotted alpha before the dotted beta. This reading seems completely out of the question. On this point Pritchett, W. K. (AJA) lix [1955] 58–9)Google Scholar and Meritt, (BCH lxxxviii [1964] 415 n. 2)Google Scholar are in accord with Mattingly, Harold B. (Historia xii [1963] 268)Google Scholar, who writes that the ‘alpha’; in the name ‘cannot be seen on the stone’. The present writer's own examination of the stone also has indicated this, as has careful scrutiny of the squeeze prepared with the aid of Mlle Karapa.

⁷ Op. cit. (note 6) 413–15.

⁸ Professor Meritt expressed this new preference to the writer in a personal letter of the spring of 1971.

⁹ Among them are Meiggs and Lewis (ML pp. 80–1).

¹⁰ Tod (i², pp. 56 and 260) is to be numbered among these.

¹¹ Roos, ErvinOpuscula Atheniensia iv [1962] 9–10)Google Scholar, e.g., concluded that the inscription is to be placed in the 450s, ‘sei es im Archontat des Ariston 454/3 v. Chr., wiees hauptsächlich in der älteren Literatur zu lesen steht, oder des Habron 458/7, wie man die Reste des Archont-namens in IG I² 19, 3 in neuester Zeit mit gleich geringer Berechtigung hat ergänzen wollen’. Piraino, Maria Teresa Manni (Kokalos vi [1960] 69)Google Scholar has evidenced the same attitude, concluding that the document dates to the archonship of Habron, ‘se si esclude l'arcontato di Ariston’.

¹² Op. cit. (note 6) 59.

¹³ Op. cit. (note 6) 269.

¹⁴ Op. cit. (note 6) 413.

¹⁵ Op. cit. (note 6) 267–9. This reading has come to be favoured also by Smart, J. D., JHS xcii (1972) 130–1Google Scholar; and Green, Peter, Armada f rom Athens x and 52Google Scholar.

¹⁶ Annali, suppl. to xii-xiv (1969) 213–17.

¹⁷ Ibid. 205 n. 7.

¹⁸ Op. cit. (note 6) 59.

¹⁹ Op. cit. (note 6) 269 n. 57.

²⁰ Ibid. 269.

²¹ Ibid.

²² Op. cit. (note 6) 415.

²³ Ibid. 414–15.

²⁴ Ibid. 415.

²⁵ Similar observations are made by Pritchett, W. K., Hesperia xxxiv (1965) 132 n. 7Google Scholar; and Smart, op. cit. (note 15) 131 with n. 17.

²⁶ The broken omicron reading earlier suggested by Kohler (op. cit. [note 2] 16–17) and included in CIA i 20 may be noted; but it needs only the briefest mention, since ọον does not permit the restoration of an archon's name.

²⁷ Op. cit. (note 6) 269.

²⁸ Their locations are: line 2, stoichos 37; line 3, stoichoi 40 and 44; line 7, stoichoi 21 and 37; line 12, stoichos 27; line 13, stoichos 31; line 14, stoichos 29; line 15, stoichos 30; line 17, stoichoi 22 and 27. The one partial rho is in line 7, stoichos 37.

²⁹ By the ‘top of a line’ is meant the straight line which results if one marks horizontally through the majority of the upper curvatures of omicrons and uppermost or only horizontal bars of epsilons, taus, and pis which occur in each line.

³⁰ Op. cit. (note 6) 269.

³¹ Ibid.

³² One consequence of the redating of IG i² 19 and 20.1–2 is that it suggests a precise date for IG i² 20.4 ff, which records the alliance between Athens and Halikyai and which was inscribed on the lower part of the stele recording the Athenian alliance with Egesta (as first noted by Raubitschek, , Hesperia xii [1943] 18 n. 29Google Scholar; and discussed more fully by him. op. cit. [note 6] 10–14). If Köhler, U. (AM iv [1879] 32)Google Scholar was correct in his suggestion that the AP which occur in line 5 of that fragment are the first two letters of the archon's name, then the Athenian-Halikyaian alliance would date to 416/5, when Ἀρ[ίμνεστοςἐρχε]. Köhler himself proposed the restoration Ἀρ[ίστων], but the only eponymous archon between 418/7 and 413 (by which point the Halikyaians assuredly had become Athenian allies: Thuc. vii 32.1) whose name begins with Αρ is Arimnestos. A 416/5 date is favoured also by Smart (op. cit. [note 15] 132–3), though for other reasons.