¹ IG iv 29–38. I am most grateful to two Ephors of Antiquities for Aigina: to the late Dr N. M. Verdelis for kindly allowing me to study these inscriptions in 1964, and to Dr B. Ch. Petrakos for similar permission in 1980; also to Miss I. Dekoulakou for her assistance in the latter year. The completion of this study was made possible by the generosity of the Wolfson Foundation. Mr Russell Meiggs has given much patient help and encouragement over the years, for which I am more than thankful, and particularly for his comments on the penultimate draft of this paper. I am most grateful also to Dr D. M. Lewis, who very kindly read this paper, found time to discuss it with me and improved it at many points.

² Hill, G. F., Sources for Greek History², edd. Meiggs, R. and Andrewes, A. (Oxford 1951) 318 f.Google Scholar, B 96. Samos: Barron, J. P., JHS lxxxiv (1964) 35–48CrossRefGoogle Scholar; Tsakos, K., Ἐπιγραϕές Σάμον Ι, ADelt xxxii (1977) 70–9Google Scholar. Chalkis: IG xii.4 934, now lost. Kos: Paton, W. R. and Hicks, E. L., Inscriptions of Cos (Oxford 1891) 160Google Scholar no. 148 Pritchett, W. K., BCH lxxxix (1965) 440Google Scholar, fig. 15; dated to the late fifth century by Sherwin-White, S. M., Ancient Cos, Hypomnemata li (Göttingen 1978) 37CrossRefGoogle Scholar f. See also Meiggs, R., The Athenian Empire (Oxford 1972) 295–8Google Scholar.

³ Thuc. ii 27; Xen. Hell, ii 2.9. See M. Fraenkel, comm. on IG iv 29–32; cf. Kirchhoff, IG i¹ 528, and, more recently, Welter, G., ‘Aiginetica xxix, der Kult der Athena’, AA 1954, 35–6Google Scholar.

⁴ Thuc. iii 50.2. It was of course part of the ritual of any colonial foundation, as for example is implied by the clause of the decree for the foundation of a settlement at Brea c. 445, providing against excessive multiplication of temene: IG i³ 46.9–11, cf. ML no. 49 and p. 131.

⁵ For a comprehensive study of these and other relevant letter-forms, see Meiggs, R., ‘The dating of fifth-century Attic inscriptions’, JHS lxxxvi (1966) 86–98CrossRefGoogle Scholar; Walbank, M. B., Athenian Proxenies of the Fifth Century BC (Toronto/Sarasota 1978) 31–51Google Scholar (a revised version of ‘Criteria for the dating of fifth-century Attic inscriptions’, Φόρος, Tribute to B. D. Meritt, ed. Bradeen, D. W. and McGregor, M. F. [New York 1974] 161–9Google Scholar) in which developing forms identified by Bradeen and McGregor are matched against texts of known date. The tables given by Meiggs, 92, 94, and by Walbank, 39–42 (165–7), which largely supersede that in Larfeld's, W.Handbuch der griechischen Epigraphik (Leipzig 1902Google Scholar) ii Taf. ii, are the authority for stylistic dating adopted throughout the present paper.

⁶ Historia x (1961) 149Google Scholar. Mattingly puts forward three occurrences of three-barred sigma as later than 446/5. Of these, the Aiginetan horoi are here sub judice, and the Samian are in fact earlier: see Barron (n. 2). The third, the long-lost choregic dedication of Aristokrates son of Skellias (IG i² 772), has been rediscovered since Mattingly wrote, and remains a puzzle: see Shear, T. L. Jr, Hesp. xlii (1973) 173–5Google Scholar no. 1; Amandry, P., BCH c (1976) 19Google Scholar and n. 7, 27–8; ci (1977) 182 with photograph and drawing, 189; Raubitschek, A. E., Hesp. Suppl. xix (1982) 130–2Google Scholar. The dedicator (cf. Pl., Gorg. 472a–b) is prima facie to be identified with the general of the Ionian War (see Andrewes, A. and Lewis, D. M., JHS lxxvii [1957] X 179Google Scholar; Davies, J. K., Athenian Propertied Families [Oxford 1971] 56–9Google Scholar no. 1904). The script, in addition to three-barred sigma, carries a full-scale omikron, obsolete in dated inscriptions after 460 though to be found occasionally in documents dated later on circumstantial grounds, e.g. Kleinias' decree of c. 447 (IG i³ 34; ATL ii pl. 3). It also employs eta, in a curious form with inward-curving verticals. This recalls the form of the aspirate seen in the heading of Quota List 5 of 450/49 (cf. ATL i 23, fig. 23) and also on a horos from Peiraieus (Hill, D. K., AJA xxxvi [1932] 258–9Google Scholar, fig. 7; SEG x 384Google Scholar: I owe the reference to Dr Lewis), likewise dated c. 450. Most curiously of all, the patronymic Σκελίο is spelled in the archaic manner with single consonant for double. A ‘natural’ inscription of the late fifth century would appear to be ruled out: the main possibilities are that it is a deliberately archaizing work (perhaps even recut: so Raubitschek, loc. cit.) or, as Lewis believes, that it is a rather old-fashioned text of c. 440. In addition to Mattingly's three ‘exceptions’ already mentioned, Walbank reports an outlying occurrence of three-barred sigma in the Parthenon accounts of 443/2. I have not so far been able to find it there.

⁷ Sources ² 318, B 96 (a).

⁸ JHS lxxxiv (1964) 44 f.Google Scholar

⁹ Π. Ἡρείωτης, Ἀρχαῖαι Ἐπιγραφαί Αἰγίνης, Προγρ. τοῦ ἐν Αἰγ. Ἑλλ. Σχολείου 1892–93 (Athens 893) 5–10–still the most important first-hand account, and Fraenkel's main source in IG iv.

¹⁰ Jeffery, L. H., The Local Scripts of Archaic Greece (Oxford 1961) 113, pl. 17.19.Google Scholar

¹¹ Guarducci, M., Epigrafia greca iv (Rome 1978) 293–6Google Scholar and fig. 86.

¹² Dr Lewis kindly provided a photograph of this text (and one of IG iv 39); the letter-forms given in IG are largely, but not wholly, reliable.

¹³ For the provenance of stones now in the museum, I depend upon IG iv and the authorities cited there. For the most part, I have not repeated the nineteenth-century bibliography, unless reading or provenance is in dispute.

¹⁴ The chapel, identified by its dedicatory inscription to the right of the west door outside, lies in the level ground to landward of the Perdika road. The stone, concealed by rendering in 1964, is now again visible under whitewash—but not ‘in angulo septentrionali’, as Fraenkel wrote, misunderstanding Iriotis' ἐν τῇ πρὸς Ν. [sc. Νότον] Despite the total confusion of Wordsworth's geography, it seems clear that this is the building in which he saw the stone (Athens and Attica ³ [London 1855] 231Google Scholar): there is no further horos to be found in the church of the Ayy. Asomati on the Oros, nor indeed in the chapel of Ay. Taxiarchis above Marathona at Pakhia Rhakhi. (Note, however, that Wordsworth reports four-barred sigma throughout this text.)

¹⁵ Euthymos: Olympia Mus. 357; L. H. Jeffery (n. 10) pl. 63.19. Eurymedon: Samos, Heraion; Wade-Gery, H. T., JHS liii (1933) 98Google Scholar fig. 3. Egypt, (a) Hegesagores at Memphis: Samos, Heraion; Peek, W., Klio xxxii (1939) pl. opp. p. 289Google Scholar; Jeffery pl. 63.21. Egypt, (b) Inaros' aristeion: Samos, , Heraion, , and Berlin, , Museen, Staatliche; Dunst, G., AthMitt lxxxvii (1972Google Scholar) Taf. 60.

¹⁶ Lygdamis: Museum, British; Roehl, H., Imagines Inscriptionum Graecarum Antiquarum³ (Berlin 1907) 23Google Scholar no. 14 = Inscr. Gr. Antiquissimae (Berlin 1882) 138Google Scholar no. 500; cf. ML 69 no. 32. Lophitis: Chios, Museum; Roehl 24 f. no. 18 = 105 no. 381, side a; SGDI 5653.

¹⁷ ATL ii D 14, pll. 6–7.

¹⁸ Le Bas, , RA 1844, 102Google Scholar; further bibliography, IG iv 35. Boeckh, who records this text ‘ex schedis Fourmonti et Guil. Gellii’, locates it ‘Athenis [sic] in limine ecclesiae B. virginis Ἀγάνἠ’, but notes that Gell in fact assigned it to Aigina (CIG 527; cf. IG i¹ 528). For the church of the Κοίμησις, at Plasta in the northern part of Palaiochora, see Iriotis (n. 9) 6 f. and n. 1; cf. Anna Yannoutis, Aegina 25. Its most readily identifiable feature, over the door, is a Latin inscription dated 1533 naming Antonio Barbaro. It is to this period that the dedication to ‘St George the Catholic’ relates.

¹⁹ It is worth repeating Fraenkel's observation, IG iv 36, that no. 6 cannot be the same stone as our no. 4, since when Le Bas saw no. 6 (Voyage archéologique [Paris 1855] ii 1681, pl. vi no. 12Google Scholar), no. 4 was still in Spyridon Moraktis' well at Mesokampo.

²⁰ SGDI 5653, lines 6–7.

²¹ The table is based on Walbank (n. 5), omitting the somewhat uncertain evidence of omikron (n. 6 above). I have given 430 rather than 425 as the terminus for epsilon because, though found as late as 425, the form is not used consistently in any one inscription after 431. Here, though occurring only twice on each horos, its use is consistent in three of the four texts.

²² Boeckh, , CIG 526Google Scholar, places this stone ‘circa fines Athenarum in ecclesia Sancti Χαραλαμποῦ’—again ‘ex schedis Fourmonti’ (see n. 18 above). Ross, L., Archäologische Aufsätze (Leipzig1855–61Google Scholar) i 244, avoided this error, but confused the stone with our no. 11, which he saw still in place above the door of St Athanasios' chapel in Aigina. See IG i¹ 528 for the confusion, and IG iv 32 for its resolution. Le Bas, RA 1844, 101 f., carelessly reports the reading of line 3 as Αθεναιας.

²³ Cf. Fraenkel on IG iv 32.

²⁴ The latest examples are in such inscriptions as the signature of the sculptor Alxenor of Naxos, c. 490–75; a marble altar of Zeus Elasteros in Paros, early fifth century; a Samian statue-base at Delphoi, perhaps of 479: respectively Jeffery (n. 10) pll. 55.12, 56.35, 63.17.

²⁵ The only example noted by Jeffery (n. 10) is the late seventh-century dedication by Euthykartides the Naxian on Delos: 290, pointing out an error at this point in the facsimile, pl. 55.3, taken from Roehl (n. 16) 27 no. 30.

²⁶ This stone certainly comes from a church of Ay. Athanasios: Ἀγίου Ἀθανσίου may be read at the beginning of line 3 of the Byzantine inscription which runs along the length of the stone. But the location of the church is in dispute between Wordsworth ([n. 14] 227) and Le Bas ([n. 22] 100; cf. [n. 19] ii 1678,pl. vi no 8). The former sets it a quarter of an hour west of the temple, at ‘Bilikada’, and the record is accepted by Fraenkel. This church, 3 km from the temple as the crow flies, in the area known as Mesagro, is now completely cement-rendered. Le Bas, however, in mentioning a stone which he identifies with Words- worth's, places the church 2 km south of the temple—in an area which is in fact called Vlichada, clearly Wordsworth's Bilikada, a short way north of the cape Peninda Vrachia, at the southern limit of the bay of Ayia Marina: see map by H. Thiersch, ap. Furtwängler, A., Aegina (Munich 1906Google Scholar); repr. by Welter, G., Aigina (Berlin 1938Google Scholar). Wordsworth's topography commands no great respect—cf.n. 14 above—even though all the other provenances for horoi of Athena, including the companion-piece no. 10, are west of the temple, between Palaiochora and the town. Since both he and Le Bas appear to agree on Vlichada as the site, it is certainly possible that the stone comes from one of the two churches of Ay. Athanasios in that neighbourhood. Le Bas read Αθεναιας a Doric form. But since his reading Αθεναιας (for -ες) on our no. 8 proves him capable of error, it is unfortunately unsafe to argue from it the existence of a stone now lost, with an indication of local Aiginetan work in its dialect.

²⁷ Jeffery (n. 10) 109; ead. in Φόρος (n. 5) 76 ff. In view of this form, Le Bas' report of a reading Αθεναιας (end of n. 26) becomes the more intriguing.

²⁸ Hdt. v 79-89. See Dunbabin, T. J., Ἔχθρη Παλαιή, BSA xxxvii (1936–1937) 83–91Google Scholar; A. Andrewes, ‘Athens and Aegina, 510–480 BC’, ibid. 1–7; Leahy, D. M., ‘Aegina and the Peloponnesian League’, CPh xlix (1954) 232–43Google Scholar; Jeffery, L. H., ‘The Campaign between Athens and Aegina in the Years before Salamis’, AJP lxxxiii (1962) 44–54Google Scholar; Podlecki, A. J., ‘Athens and Aegina’, Hist. xxv (1976) 396–413Google Scholar; Figueira, T. J., ‘Aeginetan Membership in the Peloponnesian League’, CPh lxxvi (1981) 1–24Google Scholar. I regret that I have not seen the same author's Athens and Aegina in the Archaic and Classical Periods—a Socio-Political Investigation (Diss. U. Penn. 1977).

²⁹ That was the view of Hdt. vi 49, cf. 85–94, esp. 87.

³⁰ Hdt. vii 144.1–2; cf. Thuc. i 14.3, consistent only on the assumption that the Aiginetan war continued after the Persian war.

³¹ Plut. Them. 19.2; cf. Thuc. i 91.1.

³² Cf. Thuc. i 95.4. Aigina's absence would be certain if, as is almost sure, she was a member of the Peloponnesian League at this time. See Leahy (n. 28); de Ste Croix, G. E. M., The Origins of the Peloponnesian War (London 1972) 333 fGoogle Scholar: against, D. M. MacDowell (see n.33); cf. Figueira (n. 28). It should be noted that Aigina sent help to Sparta against the Helot revolt at the time of the great earthquake, Thuc. ii 27.2; but so did Athens, for a time.

³³ Thuc. i 105.2–4, 108.4. We need not here consider the question of the date within the year (autumn, Gomme, HCT ad loc.; spring, ATL iii 178Google Scholar). MacDowell, D. M., ‘Aegina and the Delian League’, JHS lxxx (1960) 118–21CrossRefGoogle Scholar, argued that the events of 457 comprised a revolt followed by reconquest, Aigina having been a member of the Delian league from the outset. He based this on two passages of Diodoros, xi 70.1–4 and 78.3–4. In the former, the historian records under the year 464/3, after a note on the revolt of Thasos and before mentioning the attempt to colonize ‘Amphipolis’, i.e. Ennea Hodoi, roughly contemporary with the revolt according to Thuc. i 100.3, a ‘revolt’ of Aigina which provoked Athens to send a siege force; the outcome is not stated. In the latter passage, under the year 459/8 and with no reference back, he records a ‘war of conquest’ against Aigina following the victories of Halieis and Kekryphaleia (cf. Thuc. i 105.1–2), which resulted in the incorporation of Aigina into the Athenian synteleia. MacDowell's arguments are answered by de Ste Croix (n. 32) 334 f.; cf. Meiggs (n. 2) 51 f., 455 f.; Reece, D. W., JHS lxxxii (1962) 118Google Scholar n. 32. Partly it is a question of the probability of Aigina's adherence to the Spartan bloc in 478 (cf. above, and n. 32), partly a question of Thucydidean usage. For Thucydides refers to πόλεμος, ‘war’, with Aigina as with Karystos (i 98.3) and others, not to ἀπόστασις, ‘revolt’, as he does in the cases of Naxos, Thasos, Euboia, Samos and Byzantion (i 98.4,100.2, 114.1, 115.5). His account is to be preferred; and Diodoros' earlier ‘revolt’ is perhaps to be explained as a reflection of a passage in which Ephoros, his source, compared the revolts of Thasos and Aigina as landmarks in the growth of Athenian power (cf. Meiggs, loc. cit.).

³⁴ Hdt. i 64.2; Thuc. iii 104. The decline of the panegyris which Thucydides notes no doubt dated from the middle of the century, when Athens replaced Delos as centre of the League, and work on the great temple was abandoned. For the Delian temples, see Courby, F., Les Temples d'Apollon, Délos xii (Paris 1931) esp. ch. iv, 217 ff.Google Scholar

³⁵ Thuc. i 95.1; Plut. Arist. 23.2–6. Cf., among other occasions, the appeals of Aristagoras and Themistokles, Hdt. v 97.2, viii 22.1.

³⁶ Hdt. i 141.4, 170.1, vi 7. On the cult see Hdt. i 148.1, Strabo 639. During the fifth century it came to be neglected in favour of the Ephesia, perhaps after the revolt of Samos in 440: Hornblower, S., ‘Thucydides, the Panionion Festival and the Ephesia (III 104)’, Hist. xxxi (1982) 241–5Google Scholar. The devaluation of Delos (n. 34) and Panionion would of course set a terminus ante quem for an Athenian initiative in favour of their gods, such as the Aiginetan horoi represent. Poseidon Helikonios was also worshipped at Athens: Kleidemos FGrH 323 F 1.

³⁷ Thuc. i 109–10, implying the loss of the greater part of a fleet of 250 ships, 50,000 men. On the Egyptian campaign and the credibility of Thucydides' view of the scale of this disaster, see Meiggs (n. 2) 101–8.

³⁸ Meiggs (n. 2) 109–24, with detailed discussion of the evidence of the quota-lists for revolts in Miletos, Erythrai and elsewhere. Cf. id., JHS lxiii (1943) 22–34; HSCP lxix (1963) 3–6Google Scholar; Barron, J. P., JHS lxxxii (1962) 1–6CrossRefGoogle Scholar; Bradeen, D. W. and McGregor, M. F., Studies in Fifth-Century Attic Epigraphy (Norman, Okla. 1973) 24–70.Google Scholar

³⁹ Plut. Per. 12.1. The date is deduced from that of the first quota-list of sums paid to Athena in Athens: see ML 83 ff., on no. 39. For an attempt to dissociate the removal of the treasury from the Egyptian disaster, see Pritchett, W. K., Hist, xviii (1969) 17–21.Google Scholar

⁴⁰ The tribute record, given in full in ATL i, is most conveniently summarized by Meiggs (n. 2) 538–61, App. 14.

⁴¹ Arist. Rhet. 1411a15; cf. Plut. Per. 8.7.

⁴² Cf. Meiggs (n. 2) 295.

⁴³ For what follows, see ATL i 218Google Scholar, iii 38 f, 53–8, 303; Meiggs (n. 2) 183, adducing Pindar's prayer for the freedom of Aigina in an ode of 446, Pyth. viii esp. 140–2.

⁴⁴ Thuc. i 67.2, cf. 139.1, 140.3. See Meiggs, loc. cit.; and, for a more sceptical view, de Ste Croix (n. 32) 293 f.; Gomme, , HCT i 225–6.Google Scholar

⁴⁵ BSA xlix (1954) 21–5Google Scholar; photograph, Hondius, J. J. E., Novae Inscriptiones Atticae (Leiden 1925) no. 1Google Scholar, pl. i. Mattingly, H. B., ‘Athens and Aigina’, Hist, xvi (1967) 1–5Google Scholar, uses the intrusive Ionic eta and a probable Ionic gamma in this text to urge a date for it later than 431.

⁴⁶ What may have been the relation of the Aiginetan horoi to those found in Samos and elsewhere, and what the significance of the Ionian hands at work on the naming of Apollo and Poseidon, raise wider issues which will be discussed at length in my forthcoming study of the propaganda of Athenian imperialism.