We are particularly grateful to Mr T. J. Dunbabin for photos and notes of the sherds in Oxford, to Dr F. Brommer for the photo reproduced as Plate 34, 1, to Professor C. M. Robertson for tracings of the sherds in University College, London, to Professor A. J. Beattie for linguistic advice, to Mr. J. M. Cook for squeezes of the Geneleos inscriptions (to which he called our attention), and to the Ashmolean Museum for permission to publish Plate 34, 4.

¹ The paint of the inscriptions is the standard dark paint of the normal range from near-black to golden-brown.

² The appended list of inscribed sherds is probably not far from complete for finds so far made: excavators usually preserve inscriptions, and the bulk of the most interesting material from Naucratis went to the British Museum.

³ Naucratis has been most prolific, but the finds at the sanctuary in Chios were disappointingly small, so that a statistical comparison is not valid.

⁴ Gardner, E. published a selection rather inaccurately in Naukratis II 63–4 nos. 739–47Google Scholar, and Edgar, C. C. gave the alphabet (BSA V (1898/1899), 51).Google Scholar

⁵ See BSA XLIV (1949), 154–8. To the list there given of exported Chiot pottery additions may be made. Smyrna: many fragments of chalices (J. M. Cook). Erythrae (from Ilica near Ceşme): fragments of chalices (J. D. Beazley). Rhodes: chalice with sphinx in Chalice style (Florence 79244), and small chalice with b.f. ducks (B.M. 67.5–6.42); and from Camirus another plain chalice (Berlin F.1646). Delos: one chalice from the Prytaneion and a b.f. fragment from Rheneia, (Délos XVII, pls. 52, A2; 63, B4).Google ScholarNaxos: perhaps one sherd in Mykonos (J. Boardman). Salonika: sherds reported (AA 1942, 171). Cyprus: phiale in Chalice style, now in Nicosia (A. D. Trendall). Tell Defenneh, Egypt: one sherd from the plain (Tanis II 62—‘Aphrodite bowl’). Zagazig, Egypt: plastic head (CVA Oxford II, pl. 401, 22).

⁶ Nos. 16, 24, 25, 30, 59, 60, 61, 62, 201.

⁷ Though τηφροδιτηι is common on graffiti from Naucratis.

⁸ The Zoilos writer II, to whom the three examples of Ζωιιλος belong, also wrote Ζωιλος (no. 24). The doubled iota is presumably intended to show a glide: cf. e.g. in literature δμομος and in inscriptions ιερηιια, Κεραιιτην and Κεραιιτηι χρηιιзωσιν (SGDI 5495, lines 14, 19, 30, 31, 41: from Miletus: copy of late sixth century original), and Τειιοι (Athenian Tribute Lists II, list 15.I.10: from Athens, 440–39 B.C.). See also Buck, C. D., Greek Dialects ², 31–2.Google Scholar

⁹ On nos. 88 and 201 there may be transposition of the last two surviving letters, and on no. 172 φ may have been omitted or transposed with ρ.

¹⁰ Once, it seems, we have the simple form εθηκε (no. 104, and possibly no. 24). The order of the words is very occasionally changed. In four instances the ethnic Χιος is added (nos. 46, 142, 168, 171; conceivably also no. 170).

¹¹ The η here might be a slip for ε; but more probably ε is omitted or elided, and the η is the end of some such word as καλη (cf. no. 1) or ιερη (cf. Roehl, H., Imagines Inscriptionum Graecarum ³, 31 no. 51).Google Scholar Perhaps on no. 20 also the final η is from a similar formula, if it is not a misspelling of the dative Αφροδιτηι

¹² B.M. 88.6–1.531, from a large bowl decorated in Late Wild Goat style, has inside at the top Αφροδι]τηι τηι ε Ναυκρατι painted in white before firing (Naukratis II, pl. 21, 768). But it is not Chiot (as is said in AM LVIII 28).

¹³ If no. 47 is to be restored as ειμι τωπολωνος: but it could be ειμι των Διοσκουρων

¹⁴ There are so far on Chiot pots no painted dedications to Hera, though she had a popular sanctuary at Naucratis. At Aegina none of the fragments include the name of a divinity; this may be accident.

¹⁵ Other specimens of the same hand give the name Hermomandros (nos. 83, 204), so that there is a presumption that Hermomandros was named in this inscription too, whether as dedicator or γραφεύς. It is impossible to tell whether there was or was not other decoration on this chalice.

¹⁶ The ]κις in front of a woman on no. 25 is by the Mikis writer and so probably is the ending of the dedicator's name.

¹⁷ The quality of the lettering generally shows practice in the use of the painter's brush (and on the awkwardly curved surface of a pot), so that the writer was probably the painter—and the thrower—of the pot. But for precision we prefer the term ‘writer’.

¹⁸ Smyrna—SGDI 5616. Samos, —SEG I 367Google Scholar; SIG ³ 1061. Erythrae—SGDI 5693. Ilium—OGI 444. Cos—SGDI, 3624, 3704. Paros, —IG XII.5, 1043.Google ScholarEretria, —IG XII.9, 249.Google Scholar Athens—J. Kirchner, Prosopographia Attica, nos. 6231–51. These are all of later date.

¹⁹ Gardner, E. unnecessarily supposed that Mikis was a woman (Naukratis II 64 no. 745Google Scholar; cf. Roebuck, C. A., Class. Phil. XLV (1950), 247 n. 62).Google Scholar But there are masculine names in -ις.

²⁰ As Roebuck asserts (ibid., n. 65). The name is found also at Halicarnassus (SIG ³ 46) and Rhodes, (IG XII.1, 764).Google ScholarCf. also JHS LXVIII (1948), 148. Homer, of course, knew the name (Od. II 15).

²¹ Residents, unless returning from Chios, must have placed a special order through a trader or friend.

²² The gross totals of dedications may be a little too high, since a few fragments reckoned separately may be from the same pots as other fragments.

²³ Another explanation, that the manufacturers themselves exported their pottery, implies a commercial organisation that is probably too elaborate for the archaic Greek world.

²⁴ On Ionic graffiti from Naucratis of much the same date crossed θ is fairly common.

²⁵ Inscription—Roehl, H., Imagines Inscriptionum Graecarum ³, 24 no. 21Google Scholar; E. S. Roberts, Introduction to Greek Epigraphy I, no. 152: Schwyzer, E., Dialectorum Graecarum Exempla Epigraphica, 715 (3).Google Scholar Statue and date—Buschor, E., Altsamische Standbilder II 29 and figs. 86–9Google Scholar; Richter, G. M. A., Archaic Greek Art, 103–4 and fig. 160.Google Scholar

^25a Buschor, op. cit. I 18 and figs. 57, 59–60; Richter, , Kouroi, 143 and figs. 201–3.Google Scholar (Roehl, op. cit., 25 no. 22 is misleading.)

²⁶ Inscription—Roehl, op. cit., 20 no. 5; Roberts, op. cit., no. 138; Schwyzer, op. cit., 723(3). Statue—Winter, F., Kunstgeschichte in Bildern, 202, 1.Google Scholar Date—Richter, op. cit., 108.

^26a Buschor, op. cit. II 26–9 and figs. 90–101 (the inscriptions on figs. 90–2, 99, 101). In the forms of the letters the inscriptions of Geneleos are remarkably like that of Chares and may well be by the same hand; the statues too are fairly closely related, but their resemblances and differences are not sufficient to prove or disprove a common authorship.

²⁷ Inscription—Roehl, op. cit., 21 no. 9; Roberts, op. cit., no. 42: Schwyzer, op. cit., 731. Historical dating—early, e.g. Roberts, op. cit., 334–5; Brouwers, A., REG XLI (1928), 107–13CrossRefGoogle Scholar, XLII (1929), 1–10: late, e.g. Beloch, K. J., Gr. Gesch. I 1, 388Google Scholar; Guarducci, M., ASA n.s. III/V (1941–1943), 135–40.Google Scholar

²⁸ Roehl, op. cit., 19 no. 2; Roberts, op. cit., no. 133; Schwyzer, op. cit., 723(1), dating it in the seventh century.

²⁹ Roehl, op. cit., 19 no. 3; Roberts, op. cit., no. 134; British Museum Inscriptions, no. 931.

³⁰ Inscription—Roehl, op. cit., 26 no. 26; M. N. Tod, Greek Historical Inscriptions ², no. 7; Austin, R. P., Stoichedon Style, 13–15.Google Scholar Statue and date—Buschor, op. cit., II 40–1 and figs. 141–3.

³¹ Roehl, op. cit., 18–9 no. 1; Roberts, op. cit., no. 130; Tod, op. cit., no. 4.

³² AJP LVI (1935), 291–301: ω ‘can hardly have been invented in the Ionian mother-country much before 575 B.C.’. (p. 297). Cf. also AJA XXXVII (1933), 8–29 (especially 22–3).

³³ For references to the earliest dated material from Naucratis see JHS LVII (1937), 228.

³⁴ Cf. Roberts, op. cit., 157. The inscriptions of Abu Simbel have received too much attention in the study of early Greek alphabets: epigraphically they are little more than curiosities.

³⁵ Carpenter, , AJA XXXVII 22.Google Scholar

³⁶ Note the graffito of Dolion, recently found at Old Smyrna (JHS LXXI (1951), 37 fig. 9): this comes from a context dated by Transitional and Early Ripe Corinthian and so should be seventh century (and not of its very end). It shows ω as well as closed η) and koppa.

Even so, our specimens of the Ionic alphabet do not go far back and it is impossible to disprove Carpenter's ingenious explanation of the origin of ω: the Ionians had no aspirate when they received the alphabet, found themselves (on the acrophonic principle) with two letters for the vowel ‘e’, differentiated them for long and short values of that vowel, and by analogy created ω to differentiate the long and short ‘O’.

³⁷ Incidentally Naukratis II, pl. 21, 728 is also in Berlin: the sherd is from a chalice with decoration at the top much as our Plate 34, 13–16, and the tail and back probably of a lion below.

³⁸ Close to but probably earlier than Louvre A.330(1) = S.682 (S. Zervos, Rhodes, Capitale du Dodécanèse, figs. 38 and 117; E. Pfuhl, MuZ III, fig. 120).

³⁹ The painted Chiot inscriptions published in Naukratis II are numbered 739–47 on pp. 63–4 and pl. 21: the drawings are inaccurate and the commentary unhappy.

⁴⁰ The inscription is abnormally on the lower part of the bowl.

⁴¹ For the subject cf. Athens, Acr. 450a (Graef, pls. 15 and 24; E. Pfuhl, MuZ III, fig. 119, 1); B.M. 88.6–1.482; 88.6–1.559: 1924.12–1.1229; 1924.12–1.1235.

⁴² The final letter, though resembling α δ, is probably α curtailed because of the rosette below. In any case, the left-hand sherd was found in the sanctuary of Aphrodite.

⁴³ Mr. T.J. Dunbabin kindly informed us that this might be from the same pot as the preceding item.