¹ Part of this essay was awarded the Cromer Prize by the British Academy in 1935: the Appendices and the sections bearing on archaeology (i.e most of Section III on Trade) have been added later, and to make the sutures less apparent much of the original essay has been rewritten. Among many obligations, I must mention particularly that under which I am to Mr. R. H. Dundas for his patient and inspiring encouragement, and to Mr. A. A. Blakeway for much help and many suggestions, which I acknowledge in the course of the essay.

² When the traditional dates of Alcman (Suidas, Ol. 27; Eusebius, Ol. 30.4 and Ol.42) have been scaled down, as they probably should be (see Crusius in RE) he remains a late seventh-century poet.

³ Alcman, fr. 23, 1, 834 (Bergk). The case for the Adriatic and against the Paphlagonian Eneti is argued in Section III. Steph. Byz. sub ᾿Αράξαι ἤ ᾿Αραξοί. They are not mentioned elsewhere.

⁴ See Appendix I for the Diomedes Cult, and references.

⁵ Scylax is quoted in Schol. Apollonius Rhodius, IV, 1215. Νεοταίους τε τούς Σκύλαξ φησιν ἔθνος ᾿Ιλλυρικόν᾿ άπὸ τοίτων παράπλους… There is no reason to reject this fragment: no similar passage occurs in the fourth-century Periplus, and the Scholiast was acquainted with the work of the Caryandean (Schol. Ap. Rhod. I, 1177).

⁶ Hecataeus, frs. 54–71 in FHG I; Jacoby I, frs. 90–108.

⁷ Beloch, , Gr. Gesch. II. 168Google Scholar; Holm, , Gesch. Siziliens, II, 134Google Scholar.

⁸ See the charts in Philippson, Das Mittelmeergebiet.

⁹ It is impossible to be consistent about Illyrian place-names. Where there are familiar Venetian names, I have as a rule used them, as it is pedantic to refer to Crnagora and the Boka Kotorska. However, in the case of little-known places, I have tended to use the Croatian name (e.g. Brač, rather than Brazza; Hvar rather than Lesina), as this form will be found on the best modern maps (e.g. the Jugoslav edition of the Austrian Staff). I have tried to deal with Albanian names on the same principle.

¹⁰ This deterrent was called to my notice by Dr. Cary, who also helped me with a number of other suggestions which I acknowledge in their place.

¹¹ Turrill, W. B., The Flora of the Balkan Peninsula, 42 ff.Google Scholar

¹² E.g. Casson, , Macedonia, Thrace, and Illyria, 320Google Scholar. ‘The complete absence of penetration northward into the Dalmatian archipelago seems explicable on the assumption that piracy was in uncontestable control of these regions.’

¹³ Diodorus, XVI, 5. Cf. Dittenberger (ed. 3), 305.

¹⁴ Polyb. 2, 2.

¹⁵ Diodorus, XV, 14 ff.

¹⁶ Strabo, 214; cf. Zosimus, V, 27. Dittenberger, (ed. 3) no. 305, might be taken to indicate Etruscan piracy in the Adriatic; but (1) before 325 the Etruscan power north of the Apennines had been broken by the Gauls, and (2) Dionysius of Halicarnassus says (I, 29, 2) that the Greeks used to call Λατῖνοι καὶ ᾿Ομβρικοί καί Αὔτονες καὶ συχνοί ἄλλοι by the name Τυρρηνοί.

¹⁷ See Section II below (p. 180).

¹⁸ Id.

¹⁹ Strabo, 217 (Ariminum); 217, 19, 27 (Ravenna: see, further, note 144 of Section II).

²⁰ Pliny, , N.H. III, 113Google Scholar.

²¹ The evidence for the following passage will be found below, Section II, notes 140 ff.

²² Strabo 241. That it was a Picene city is proved by finds from the site, now in Ancona (cf. Randall-MacIver, Iron Age in Italy, ch. IV).

²³ See below, notes 200, 207 of Section III.

²⁴ The suggestion is made in Evans', Through Bosnia and the Herzegovina, 388Google Scholar; cf. Radu Vulpe, Securile de Bronz de tip Albano-Dalmat si domnia lui Cadmos la Enchelei.

²⁵ The references relevant to the Cadmus legend are cited below, in Appendix I.

²⁶ Evans, l.c.

²⁷ Krahe, , Balkanillyrische Geographische Namen, 3Google Scholar.

²⁸ Evans, l.c.

²⁹ Id.

³⁰ Vulpe, op. cit.

³¹ The Beisan axe is described by Couissin, in RA xxvii, 1928, 265.

³² Id.

³³ See Blakeway, in BSA xxxiii, 175Google Scholar (one perhaps Mycenaean).

³⁴ Herod. VII, 170.

³⁵ Plutarch, Quaest. Graec. no. II; cf. BSA xxxvii, 205Google Scholar, note 4.

³⁶ Pseudo-Scymnus, 442–3.

³⁷ Pausanias, V, 22, 2, giving the contemporary epigram.

³⁸ Lycophron, 1043 ff., with Tzetzes ad loc. admittedly not a Homeric tradition (Iliad. II. 549Google Scholar); cf. Paus. V, 22, 3.

³⁹ Pliny, , N.H. II, 204Google Scholar.

⁴⁰ The case for the identification of the ruins on the Pasha Liman with Oricus is conclusively argued by Patsch, , Sandschak Berat in Albanien, 70Google Scholar.

⁴¹ Cf. Hecataeus, fr. 45; Pseudo-Scylax, 26; and Pseudo-Scymnus, 441. The Oricum form appears, e.g., in Pliny, l.c.

⁴² Schol. Ap. Rhod. IV, 1175: Τήν δέ ἀντικρύ τῆς Κερκίρας χτραν εἴρηκε Μακρίδιην, ἴαως διἀ τὀ ᾠκηκέναι ἐκεἴ τοὐς Εύβοεῖς.

⁴³ Strabo, 269–70.

⁴⁴ Schol. Ap. Rhod. IV, 1216.

⁴⁵ Steph. Byz. Λάδεστα ἤ Λάδεστον (Theopompus); cf. Appian B.C. II, 39 (Liburnians at Epidamnus).

⁴⁶ Thuc. I, 13, 4.

⁴⁷ Appian, B.C. II, 39. Foundation of Epidamnus, Thuc. I, 24, 2.

⁴⁸ See below, Section III, pp. 181–4.

⁴⁹ Cf. Steph. Byz. Δυρρἁχιον ᾿Αλέξανδρος δέ ἐν (cf. Etymologicum Magnum, sub. voc.; Krahe, , Balkanillyrische Geographische Namen, 2Google Scholar) . Strabo 316, . The Strabo quotation reoccurs in Porphyrogenitus, Constantine, On the Themes, 26Google Scholar.

⁵⁰ Some of the coins with the Δυρ- legend may be fifth century. The literary evidence indicates that the Epidamnus name was more common (e.g. Thuc. I, 24; Herod. VI, 127; Paus. V, 22, a late sixth-century epigram).

⁵¹ Paus. VI, 10, 8.

⁵² Herod. VI, 127.

⁵³ Preface to Pind., Schol.Nem. IXGoogle Scholar.

⁵⁴ Strabo, 357.

⁵⁵ CAH III, 546Google Scholar (H. T. Wade-Gery).

⁵⁶ Thuc. I, 24, 5.

⁵⁷ Id.

⁵⁸ Paus. VI, 10, 5.

⁵⁹ Thuc. I, 24, 4.

⁶⁰ Plut., Quaest. Graec. IIGoogle Scholar. The story is undated, but most likely refers to the sixth or fifth century, when the aristocrats were in control.

⁶¹ Thuc. I, 24, 5.

⁶² Thuc. I, 24, 7.

⁶³ Thuc. I, 14, 2.

⁶⁴ Thuc. I, 27, 2.

⁶⁵ See below, note 83 of this Section.

⁶⁶ Thuc. I, 26, 2 (a Corinthian colony) is better authority than e.g. Strabo 316 (a Corcyrean colony) or Pseudo-Scymnus 438 (mixed); it is likely a priori that of the original two hundred all were Corinthians, and that Corcyreans drifted in later.

⁶⁷ No traditional foundation date has been preserved, though 588 is frequently given (e.g by Rey, in Albania, IV, 27Google Scholar, implicitly attempting to date middle Corinthian pottery below this). There is no ancient authority for it. The terminus ante is the destruction of Scillus and Dyspontium. As the site was occupied by natives before the arrival of Greeks, it is more than usually difficult to fix a certain upper limit from the archaeological evidence. Cf. note 69.

⁶⁸ Thuc. I, 26, 2. Cf. p. 170.

⁶⁹ Praschniker (Öjh 1922, Beiblatt, I, pl. 8) published ‘keramische Proben von Apollonia,’ which seem to have been lost in the Austrian retreat when Pojani monastery took fire (l.c. col. 8). From an examination of pl. 8 it is clear that the middle sherd of the top row is not Corinthian, but East Greek; cf. the unbound rosette and three lines (a plant ?) sprouting from the ground on which the animal was standing. He also mentions Protocorinthian sherds, but none has come to light since, as far as one can tell from the collections in Fier town-hall and in the Tirana Museum, so I should be inclined to doubt if it is fair to use this as evidence for dating the foundation before that of Epidamnus. The date is difficult, but most likely c. 600. The East Greek sherd might be a little earlier, but there is certainly no Corinthian of the last quarter of the seventh century. The earliest Corinthian belongs to the first quarter of the sixth.

⁷⁰ Steph. Byz. .

⁷¹ Steph. Byz. l.c. Cf. Гυλάκεια.

⁷² Paus. V, 22, 3.

⁷³ The only other instances with which I am familiar where numbers of settlers are given are Athenian Colonies (e.g. 1000 to Thrace, Plut. Per. 11). Leucas: Pseudo-Scylax, 24.

⁷⁴ Cf. Aelian, , Varia Historia, XIII, 15Google Scholar.

⁷⁵ Herod IX, 93.

⁷⁶ Id.

⁷⁷ Pindar, , Nem. IV, 53Google Scholar. Cf. Appendix IV.

⁷⁸ Herod, IX, 93, . I hope to discuss the evidence for the view here expressed, and the bearing that it has on the date of the composition of Herodotus ‘history, elsewhere.

⁷⁹ Paus. V, 22, 2 ff. . The site of Thronion (Klos, Kanina, or Plloca) awaits discussion.

⁸⁰ Inschr. v. Olympia, 692.

⁸¹ Thuc. I, 38, 3.

⁸² Paus. l.c.

⁸³ Thuc. I, 26, 2.

⁸⁴ Aelian, , Varia Historia, XIII, 15Google Scholar.

⁸⁵ Aristotle, , Politics, IV, 4, 5Google Scholar: .

⁸⁶ Aristotle, l.c.

⁸⁷ See below, Section III, notes 215 ff., for the evidence for these generalisations.

⁸⁸ Herod. I, 163. 6 Ἀδρίης does not mean the northern end of the Adriatic. See Appendix IV.

⁸⁹ Pseudo-Scymnus 211 ff.: cf. Aristotle, fr. 549; Justin, XLIII.

⁹⁰ Myres, J. L. in JHS 1906, 84CrossRefGoogle Scholar.

⁹¹ Cf. Burn, A. R. in JHS 1927, 146Google Scholar.

⁹² Herod. IV, 152. It is doubtful whether this passage can be accepted at its face value, especially in regard to the ἀκήρατος market of Tartessus, as it is said that a Protoattic amphora, now in Copenhagen, was found in a ‘Punic grave near Cadiz.’ (I am indebted to Mr. J. M. Cook for this information.)

⁹³ See below, notes 224 and 225 of Section III.

⁹⁴ Pliny, , N.H. III, 152Google Scholar.

⁹⁵ Steph. Byz. Μελίτουσα. There is no proof that the -ussa names are specifically Phocaean; yet I find it hard to believe that the very frequent occurrence of such names on the route to Tartessus and beyond is pure coincidence (see Schulten's edition of Avienus, p. 84, note 148).

⁹⁶ Steph. Byz. Βρέττια.

⁹⁷ I am indebted to Dr. Abramić, Director of the Archaeological Museum at Split, for this information and much other kindness.

⁹⁸ For a description of the site, and the Roman finds, see RE sub Brattia.

⁹⁹ See p. 193 of Section III.

¹⁰⁰ See note 69 of this Section, p. 168.

¹⁰¹ See p. 193 of Section III.

¹⁰² Cf. Bilabel, , Ionische Kolonisation, 50Google Scholar for references.

¹⁰³ See below, notes 224, 225 of Section III.

¹⁰⁴ See note 5, Section I.

¹⁰⁵ Cf. Strabo, 654; Vitruvius, 1, 4, 12; Steph. Byz. Ἐλπία. Archaeological evidence is given on p. 193.

¹⁰⁶ The time note given in Strabo 654 is unacceptable ( surely means ‘before the first Olympiad,’ though Hiller von Gaertringen in RE s.v. Rhodos appears to take it to mean ‘before the introduction of dating by Olympiads’).

¹⁰⁷ Thuc. VI, 3, 3 (Gela 690).

¹⁰⁸ See Appendix I for references.

¹⁰⁹ Mayer, Pace, Apulien, 401Google Scholar.

¹¹⁰ See Appendix I.

¹¹¹ Vitruvius, l.c.; Steph. Byz. Ἐλπία.

¹¹² Cf. nos. 4670–71 in Kirchner, Prosopographia Attica.

¹¹³ Vitruvius, l.c.; Steph. Byz. Ἐλπία: Strabo, 654.

¹¹⁴ Cicero, , De Leg. Ag. 71Google Scholar.

¹¹⁵ See p. 193 below on Greek trade with Apulia.

¹¹⁶ See Philipp in RE s.v. Salpia.

¹¹⁷ See p. 192 and note 239 below.

¹¹⁸ Randall-MacIver, (Iron Age in Italy, 237)Google Scholar says that both Rugge and Egnazia were originally Rhodian colonies, but there is no ancient authority for the statement. Mayer, (Apulien, 384 ffGoogle Scholar.) assembles the place-name arguments in favour of extensive Rhodian influence in Apulia, but by themselves they are too weak to be used as evidence.

¹¹⁹ Herod. III, 49; ‘Plutarch’ De Herodoti Malignitate, 860 B ff. . The date of Dionysius is c. 350 (FHG IV, 393Google Scholar), that of Antenor not later than the second century B.C. (Schwartz in RE). It is curious that Antenor tells the story ἐν τοῑς Κρητικοῑς, though natural enough that it should come in Dionysius's Κτίσεις. It is tempting to ask if it could not have been in the . Their version of the story is to be preferred to that of Herodotus: (1) because the latter is inconsistent. The aim of the elaborate religious subterfuge must have been to avoid an open breach with Corinth, and (2) ‘Plutarch's’ point about Ψηφίσματα in favour of Cnidians is a good one. The Cnidians were about this time eager to colonise (Diodorus, V, 9, . Is it possible that the date of the expedition to the Lipari islands is 628–9, as Eusebius says, and that Diodorus has confused it with the Corcyra expedition by putting it c. the 50th Olympiad? There is no more reason for connecting Pentathlus' expedition with the foundation of Acragas (see e.g. JHS 1929, 29Google Scholar) than with the foundation of Selinus, except for the Rhodian connection of Gela. In any case, Diodorus shows that the Cnidians had a motive for colonising in the first quarter of the sixth century. It might be that their bad relations with Alyattes led them to rescue the boys, or, alternatively, that they had acted from humanitarian motives, and that the incident spoilt their standing in Lydia.

¹²⁰ Pseudo-Scymnus, 421; Strabo, 315; Pliny, , N.H. III, 152Google Scholar.

¹²¹ Cf. Strabo, 243, 268 (Cyme and Naxos).

¹²² Apollonius Rhodius, N. 569 ff. .

¹²³ Dittenberger (ed. III), no. 137.

¹²⁴ Implicit in Thuc. 1, 32, 4.

¹²⁵ See Section III below.

¹²⁶ See Section III below, pp. 185–7.

¹²⁷ Arch.-Epigr. Mitt. ix, 33Google Scholar, note 5. I have not traced this pot (a ‘Krug’ = Oenochoe?), but it seems from the description to be Corinthian: it is a priori improbable that it is Italo-Corinthian.

¹²⁸ The names are not without parallel; e.g. (BCH xxxvi, 33Google Scholar), from Rheneia; cf. IG IX (2), 553, 13, Thessalian Larisa. Pullus does not seem to occur in Greece proper, but in Macedon (Dittenberger, ed. III, 674, 28 ff. ), and in Apulia, on the coins of Salpia, suggestive in view of the probability of South Dorian elements in the city.

¹²⁹ Tarentine friendship = Herod. III, 138. For information about sixth-century Tarentum I am indebted to Mr. T. J. Dunbabin, who gave me the opportunity of seeing his work on south Italy and Sicily when it was still in typescript.

¹³⁰ See p. 192 below.

¹³¹ Polyb. X, I. . This might be taken as evidence that Tarentum monopolised Apulian trade in the early period, but Brindisi was certainly founded as early as Tarentum itself (cf. note 102); I incline to think that Polybius meant by the ‘founding’ the sending of the Latin colony. That Pseudo-Scylax 14 mentions Otranto and not Brindisi is an illustration of the latter city's unimportance in the fourth century.

¹³² The archaeological evidence on which this is based is discussed in Section III, p. 190.

¹³³ Pliny, , N.H. III, 18Google Scholar.

¹³⁴ Thuc. VI, 2, 4. Cf. N.H. III, 19 ‘Siculi et Liburni plurima eius tractus tenuere, in primis Palmensem, Praetutianum Hadrianumque agrum. Umbri eos expulere, hos Etruria, hanc Galli.’

¹³⁵ Strabo. 241. Cf. Appendix III.

¹³⁶ See below, note 186.

¹³⁷ Herod. I, 167.

¹³⁸ The Atestine culture is described in Randall-MacIver, , The Iron Age of Italy, 8 ffGoogle Scholar., a work on which the following attempt to give a brief account of the situation in the north-east of Italy before and at the time of the arrival of the Greeks lias been based.

¹³⁹ Theopompus in Athenaeus 526 f. .

¹⁴⁰ Randall-MacIver, op. cit., 119.

¹⁴¹ See below, notes 215 ff. of Section III.

¹⁴² Pliny, , N.H. III, 113Google Scholar (Etruscans destroy 300 Umbrian towns).

¹⁴³ Diodorus, XIV, 113, puts the first Gallic attack not much before the fall of Rome. Livy, V, 33, 6, makes the invasion of a long-drawn-out affair, which is a priori more likely.

¹⁴⁴ Strabo, 214. There is no reason to reject this tradition out of hand, as is always done, though the passage above is worded more confidently than is, I feel, altogether justified. That the tradition fits the historical situation at the end of the sixth century I try to show in the text; internal evidence in favour of the tradition is: (1) It is not a ‘pedigree’ legend, as the Thessalians did not stay in the Po valley. (2) It can hardly have anything to do with a ‘Pelasgian’ myth, as no tradition connecting ‘Pelasgians’ and Umbrians and an Etruscan war is known. (3) Finally, is it likely that so odd a thing as a Thessalian colony in the Po valley would be invented ? Rosenberg's suggestion that the legend grew up round a by-name of Ravenna, Rene (Zosimus, V, 27) and Rhene, mother of Medon, who occurs in some obscure Thessalian myth, is not good enough.

¹⁴⁵ Ceressus, De Mal. Herod. 33Google Scholar, Plut., Camillus, XIXGoogle Scholar. Phocian disaster, Herod. VIII, 27.

¹⁴⁶ BSA xxxii, pp. 139–47Google Scholar.

¹⁴⁷ There is a polygonal wall at Himára which looks to be sixth-century or earlier. There are no Hellenic remains at Onchesmus (Santi Quaranta). Both are obvious sites for Corcyrean forts (cf. Thuc. III, 85).

¹⁴⁸ Dion. Hal. I, 60, 3.

¹⁴⁹ Strabo, 376, . I am indebted to Mr. A. A. Blakeway for pointing out to me the possible relevance of the fact that a considerable percentage of so-called Aeginetan colonial mark coins, not Cydonian (for which cf. Robinson, E. G. S. in Num. Chron. Vth Series, vol. 8Google Scholar) came from the Woodhouse collection made on Corfu.

¹⁵⁰ The identification with Spina of the city the necropolis of which has been excavated in Valle Trebba, the vases being now in Ferrara (see Aurigemma, Il R. Museo di Spina), rests primarily on Strabo 214, , and Pliny, , N.H. III, 120Google Scholar, Hoc ante Eridanum ostium dictum est, aliis Spineticum ab Urbe Spina … Auget ibi Padum Vatrenus amnis ex Forocorneliensi agro. This identification has not been questioned. On the wealth of Spina cf. Beazley, , JHS lvi. 89Google Scholar.

¹⁵¹ Aurigemma, op. cit., 16. ‘L'esempio di Adria e l'assenze di strutture murarie riferibili ad età etrusca, sia in Valle Trebba, sia nelle prossime valli a mezzogiorno della strada per Comacchio insinuano il sospetto che l'abitato dell'antica Spina fosse constituito da una vasta palafitta. … Peraltro, di una palalitta di età romana, sembra, si e constatata l'esistenza quasi alle porte dell'odierna Comacchio.’

¹⁵² Ibid., 12. The most cogent evidence there produced is the Etruscan graffiti. So little of the tomb structure has survived, the wood havingperished, that one is inclined to be sceptical about ‘le analogie fra le necropoli etrusche di Marzabotto e di Felsina e il sepolcreto di Spina.’ Admitting that the size of some tombs, e.g. 555, can be ascerained from the surviving fragments of wood, one has no reason to believe that all the tombs were built on the same principle; the diversity of the burial rites (e.g. tomb no. 506 contains cremated bones and an inhumed child; ratio of burial methods is 686 inhumed, 486 cremated, 41 uncertain, Aurigemma, p. 26. Tomb furniture is found to the right and the left and at both sides of the corpse), tells rather against the idea.

¹⁵³ Tomb 555.

¹⁵⁴ Aurigemma, op. cit., 14.

¹⁵⁵ Strabo, 421, .

¹⁵⁶ E.g. by Aurigemma, p. 4. ‘L'allontanarsi di Spina dal mare, pel piu lungo corso del flume, fu, con altre cause, l'essenziale ragione per cui Spina decadde e spense.’ But see Dion. Hal. I, 18.

¹⁵⁷ It is hard to date the late Italian from Valle Trebba. It is later than the latest Attic imports, which are c. 360 (e.g. Schefold, Untersuchungen zu den Kertschen Vasen, Nos. 349–50). Beazley, , JHS lvi, 88Google Scholar dates the latest graves to the early third century.

¹⁵⁸ Greek: Justin XXI explicitly, Hecataeus, fr. 58, implicitly, calling it polis without qualification, Etruscan: Livy, V, 33, 7. Plutarch, , Camillus, 14Google Scholar. Pliny, , N.H. IIIGoogle Scholar.

¹⁵⁹ Euripides, , Hippolytus, 735 ffGoogle Scholar. .

¹⁶⁰ Davies, O., Roman Mines in Europe, 239Google Scholar.

¹⁶¹ Ibid.

¹⁶² Δυέσται is an emendation of the text of Strabo, 326; the MSS. run Περισἀδυὲστε συνεστήσαντο… I quote the relevant passage: ἀναμέμικται δὲ τοὐτοις (the inhabitants of Orestis) τἀ ᾿Ιλλυρικἀ ἔθνη τἀ πρὸς τῷ νοτίῳ μερει τῆς ὀρεινῆς καί τἀ ὐπὲρ τοῦ ᾿Ιονίου κόλπου τἤς γὰρ ᾿Επίδαμνου καί τῆς ᾿Απολλωνίας μέχρι τῶν Κεραυνίων ὐπεροικοῦσι Βυλλίονἐς τε καί Ταυλάντιοι καὶ ΠαρΙῖνοι καί Βρῦγοι πλησίον δέ που καί τἀ ἀργυρεἰα τἀ ἐν Δαμαστιῳ περισαδυέστε συνεστήσαντο τὴν δυναστείαν καὶ ᾿Εγχελέους καὶ Σεσαρηθίους καλοῦσι πρὸς δὲ τούτοις Λυγκἤσται τε καὶ ἠ Δευρίοπος καὶ ἡ Τριπολιτῖς Πελαγονία καὶ ᾿´Εορδοι καὶ ᾿Ελίμεια καὶ ᾿Εράτυρα.

¹⁶³ Strabo, 326.

¹⁶⁴ FHG I, 5Google Scholar fr. 67 , cf. fr. 69 (sic) . Fr. 68 links the two together. . Jacoby I, frs. 99–101.

¹⁶⁵ Cf. Thuc. I, 24; Pseudo-Scylax 26; De Mir. Aus. 127; Pliny, , N.H. III, 144Google Scholar (putting them north of the Drin).

¹⁶⁶ Jireček, , Handelstrassen und Bergwerke Serbiens waehrend des Mittelalters (ed. I), 42Google Scholar … ‘ein von Prof. Ljubic juengst im venezianischen Archiv gefundenes Document aus dem Jahre 1595’ mentions three north Albanian silver mines, at Fandi in the Dukajin, Bulgari in the Mirdite country, and ‘in den Bergen oberhalb Alessios.’ I have not found these mines or traced the document, but it is improbable that such a source would be unreliable on the fact of the presence of silver. There is a lump of silver ore in the minerological collection of the Jesuits in Scutari, said to come from Merturi in the Dukajin.

¹⁶⁷ Davies, op. cit. 239. Seltman, , Athens: Its History and Coinage, 128Google Scholar.

¹⁶⁸ Cf. Seltman, op. cit. 129; ‘it is a fair assumption that the Brygoi of Epirus were no more backward than their cousins, the Bryges of Thrace, in working the silver-bearing veins.’

¹⁶⁹ See below, notes 174 and 175.

¹⁷⁰ Strabo, 326. Arrhabaeus, according to Strabo the first Bacchiad to rule in Illyria, was grandfather of Eurydice, mother of Philip of Macedon; she must have been born in the thirties or twenties of the fifth century. This tells against the view that the Bacchiad origin of the house was legendary: had it been, a synchronisation with the Argive Macedonian dynasty would have been natural. But no such high antiquity was claimed.

¹⁷¹ Filow (Die Archaeische Nekropole von Trebenischte) argues that the bronze work is Corinthian (supported, with reservations, in Necrocorinthia, 216). The contents of the new grave (Öjh 1932, 1 ffGoogle Scholar.) have led to the view that the finds are Laconian.

¹⁷² Thuc. 1, 47, 3.

¹⁷³ This was pointed out to me by Mr. A. A. Blakeway, who had discerned and was working out archaeological evidence connecting silver from the north-west with Corinth.

¹⁷⁴ Pliny, , N.H. XXI, 40Google Scholar (Iris laudatissima in Illyrico, et ibi quoque non in maritimis, sed in silvestribus Drinonis et Naronae); Theophrastus, Hist. Plant. IV, 5Google Scholar, 2 (Iris grows best in Illyria and round the Adria). Cf. De Caus. Plant. VI, 18Google Scholar, 2 (Iris scentless in Thrace); Athenaeus, 553 (μύρονἰρίνον used in Athens in fifth century). Cf. Appendix I.

¹⁷⁵ Corinth exported scent. Necrocorinthia 5, note 3 (cf. Pliny, , N.H. III, 2Google Scholar, Irinum Corinthii diu maxime placuit). I am indebted to Mr. A. A. Blakeway for pointing out to me the relevance of the iris.

¹⁷⁶ Strabo, 316. .

¹⁷⁷ Earliest mention in Apollonius Rhodius IV, 574. .

¹⁷⁸ Caesar, , B.C. III, 26Google Scholar. … Nymphaeum qui portus … ab Austro non erat tutus.

¹⁷⁹ I hope to discuss elsewhere the evidence for dating the original walls of the fortress at Lesh before the time of Dionysius of Syracuse. Evidence for the Drin valley–Ohrid route to the Aegean in Appendix II.

¹⁸⁰ ‘Published’ in RA XXIV, 1, pl. XV, and mentioned by Ugolini, , Albania Antica, 15Google Scholar, note 2, as having been in the collection of a M. Perrod at Scutari. The drawing in RA is bad; but it is evident that the figure it attempts to represent is now in the Louvre (De Ridder, , Bronze Antiques du Louvre, 27Google Scholar, pl. 16); cf. also de Ridder, op. cit. no. 141, pl. 16: a fifth-century Corinthian Aphrodite, ‘from Albania.’ This latter can be attributed to Corinth with more confidence than can the former, for which I am unable to find a parallel. The latter is partially paralleled by Necrocorinthia, pl. 46, no. 4.

¹⁸¹ Ugolini, op. cit. 35.

¹⁸² On the importance of the position of Medun, see Šufflay, Staedte und Burgen Albaniens. For the date of the wall, cf. note 179.

¹⁸³ Pseudo-Scymnus, 422–3.

¹⁸⁴ See below, notes 185, 186. There is one import that is in all probability earlier than the pots there mentioned. This is the boar, illustrated in Wissenschaftliche Mitteilungen aus Bosnien, VI, 51Google Scholar, and described by Casson, op. cit., p. 310, as ‘a silver jewel representing a boar in relief probably of Ionian workmanship.’ But this boar is quite unlike East Greek boars (cf. Price, , JHS 1925, 199Google Scholar; Clara Rhodos, VI–VII, 492Google Scholar, fig. 19). The most salient characteristic, the deeply hollowed back, is not shared to any marked degree by the Sarajevo example, and its nearest parallel seems to be the boar on a round aryballos with foot (Johansen, pl. XLIV, I), date 650–40 (Necrocorinthia, no. 18, 269). It is most likely Corinthian. This boar is, then, nearly two generations earlier than any pots that have been found on the islands, let alone up the Narenta. As, however, so precious an object would be carefully treated, it is probable that it would survive long after it was made: i.e. it is far less convincing evidence of seventh-century Greek contact with Bosnia than would be a sherd. In this connexion it should be mentioned that this boar cannot ever have stood on the neck of a column crater; it is too thin, and shows no sign of ever having been attached to anything. As an object, it is paralleled by a horse of double ‘Silberblech’ from Trebenischte, as Filow, op. cit., p. 29, notes; the parallel is not exact, as the Sarajevo boar is one thickness of ‘Silberblech’ only.

¹⁸⁵ Fuehrer durch das K.K. Staatsmuseum in S. Donato in Zara, p. 2 (comparative absence of weapons in graves of Nin).

¹⁸⁶ There are certainly two Corinthian pots of Dalmatian provenience: a middle Corinthian convex pyxis, found in a Roman grave at Salona, according to the Spalato Museum register (in connexion with this information I am indebted to the kindness of the Director, Dr. Abramić). An aryballos from Issa is now in Zara. There are several mentions of Corinthian pots from Dalmatia in now-defunct Austro-Hungarian periodicals: e.g. Arch. Epigr. Mitt. IX, 33Google Scholar, note 5 (an aryballos from Gradina and a jug from Korĕula), which sound from the description to be Corinthian (the possibility of their being Italo-Corinthian is a priori remote). Cf. De Pouqueville, , Voyage de la Grèce, 8Google Scholar, note 1: ‘ses villes étaient Lissa et Meo dont on ne connait plus que les mines au milieu desquelles on a trouvé des vases étrusques, des inscriptions et quelques médailles avec la tête de Pallas ayant au révers tantôt une amphore, et tantôt une chèvre.’ There are fragments of a very fine Attic crater (? c. 480) in Spalato, and there is a little late bf. in Zara. These are from Nin. A very late bf. cup ‘angeblich von Antivari’ is now in Sarajevo, (Bos. Mitt. XII, 281)Google Scholar.

¹⁸⁷ The ‘Illyrian’ proveniences are listed in Filow, op. cit. p. 80, note 2. As to date, Furtwaengler (quoted in Bos. Mitt. VI, 149Google Scholar) was of the opinion that the type belonged to the sixth and fifth centuries, hardly to the fourth (cf. Olympia, IV, 171, 1029Google Scholar). Schroeder, (AA 1905, 18)Google Scholar appears to suggest a sixth-century date. The examples from Zeitenlik, near Salonica, published in Albania, II, 40 ffGoogle Scholar., occur in datable contexts; e.g. the presence of the cothons (cf. JHS 1911, 74Google Scholar. The cothons in question are type A 2 of Burrows' and Ure's classification) indicates a date c. 500, though the black glaze cups look a little later.

¹⁸⁸ De Mir. Aus. 138 (salt and cattle); Strabo, 317 (salt); Pseudo-Scylax, 24 (boats on the Narenta and a ‘very fertile island’).

¹⁸⁹ The ear of corn on the Κορκυραίων coins attributed to Black Corcyra (e.g. by Head, op. cit. 317, tentatively) tells against the identification; the symbol is hard to reconcile with the known character of the island. (Cf. however, notes 127, 128 of Section II.)

¹⁹⁰ Cf. the fourth-century coinage of Issa (Head, op. cit. p. 318).

¹⁹¹ See above, pp. 173, 174 of Section II.

¹⁹² See Appendix I.

¹⁹³ Etymologicum Magnum . This is a geographical error. Budua lies many miles north of the Drin. The fragment is apparently accepted by Liddell and Scott (ed. 7 sub ναίω; the new edition does not commit itself), though Mueller has his doubts (GGM I, 30Google Scholar ‘nescioquinam Sophocles’). Pearson (Fragments, III, p. 172Google Scholar) regards it as Alexandrian, but recognises that the legend is fifth-century.

¹⁹⁴ N.H. III, 144.

¹⁹⁵ See Patsch in RE sub Epidaurum. The earliest contact is a late archaic Heracles, perhaps from a crater. It was found in Popovo Polje, a few miles inland of Ragusa, and is now in the Prince Paul Muesum, Belgrad.

¹⁹⁶ Casson, op. cit. 315 marks both Melite and the Elaphites as Greek settlements. There is as yet no evidence for this, though I have heard rumours of ancient remains on the island Melite near the church of Sveta Marija.

¹⁹⁷ Cf. Tod, Gr. Hist. Inscr. no. 79, 1, 29. ‘The may be a native of Melitene in E. Cappadocia, or of the Illyrian island of Melite … or of Malta.’

¹⁹⁸ There is, so far as I know, no evidence for ancient exploitation.

¹⁹⁹ Steph. Byz. ; Pliny, , N.H. III, 30Google Scholar, 3.

²⁰⁰ The site of Starigrad (certainly the city Paros or Pharos, on epigraphic evidence) cannot be reconciled with Polybius' account of Aemilius' campaign against Demetrius in 219 (III, 18). Polybius never refers to the city Paros throughout this passage. I hope to deal with this problem more fully elsewhere.

²⁰¹ Steph. Byz. . The being otherwise unknown, the clue to localisation lies in the last four words, Σχερία must be Black Corcyra, by a natural confusion; Corcyra is not in a gulf. So the colony Anchiale must be in the Narent a area.

²⁰² There are fourth-century coins, many found on Hvar. Casson, op. cit., 318, erroneously says that it is not otherwise known; cf., however, Pseudo-Scylax, 22.

²⁰³ It is very tentatively suggested by Wade-Gery, H. T. in CAH III, 551Google Scholar that the Heracleas (e.g. Lyncestis Strabo 326; in Athamania Strabo 323; in Acarnania Pliny, , N.H. IV, 5Google Scholar; cf. Caesar, , B.C. III, 733Google Scholar) in the north-west of Greece and in Illyria may have been Bacchiad foundations. I should doubt if there was any Greek settlement so far north as the Narenta at the time of Bacchiad rule in Corinth, but the colony might have been founded some years after the expulsion of the clan.

²⁰⁴ For a discussion of the modern theory that it was founded by Dionysius of Syracuse, see Appendix III.

²⁰⁵ Dittenberger (3rd ed.), no. 141.

²⁰⁶ See above, note 186. Cf. also Bos. Mitt. XII, 281Google Scholar, fig. 14 (an Attic bf. lekythos of the last quarter of the sixth century).

²⁰⁷ Pseudo-Scymnus, 413–4.

²⁰⁸ See above, notes 133, 134 of Section II.

²⁰⁹ Polybius, XXXII, 18; Strabo, VII, 5, 5.

²¹⁰ Récherches à Salone, 13.

²¹¹ Diodorus, XV, 14.

²¹² Cf. also Brunšmid, , Die Inschriften und Muenzen der Griechischen Staedte Dalmatiens, 16Google Scholar, no. 3.

²¹³ I take it that the source of Pseudo-Scymnus, 423 ff. () is early; at least before c. 235 and perhaps fifth century.

²¹⁴ Pseudo-Scymnus, l.c.

²¹⁵ There is no trace in Ancona Museum of a ‘Protocorinthian balsamry’ said by Randall-MacIver, (Iron Age, 127)Google Scholar to have been found at Belmonte. Throughout this section I am very deeply indebted to Mr. T. J. Dunbabin, without whose help it could hardly have been written. But the errors are my own.

²¹⁶ Necrocorinthia, 163, no. 1178 (early Middle Corinthian). I am indebted to Mr. A. A. Blakeway for pointing out to me the relevance of Omrikos.

²¹⁷ Perhaps a little earlier; cf. note I.

²¹⁸ Illustrated in MonAnt xxxv, pl. 23.

²¹⁹ Necrocorinthia, 189 (Marzabotto, Felsina and the Venice column-crater), and no. 1357 (a Late Corinthian I amphoriskos). From Numana, three aryballoi.

²²⁰ CVA Italy fasc. V and VII.

²²¹ The finds are at Ancona; the most important are: (1) from Rapagnano, ‘a bronze that is clearly Ionic, joined to one that is less certainly Greek, but Ionising’ (I owe this information to Mr. T. J. Dunbabin). (2) From Belmonte, apparently the most important inland Picene settlement, a black figure cylix (last quarter sixth century); another, c. 500; a red figure cylix (middle fifth century). (3) From Castelbellino; nothing before 500. There is also a little red figure from Villamagna; and there is a Chalcidian oenochoe in Triest, provenience unknown (Rumpf, p. 136), and another in Ancona.

²²² This happened before Pindar wrote, e.g., Nem. IV, 69Google Scholar.

²²³ Geographical Journal, 1925, 485Google Scholar.

²²⁴ Cary, M. in JHS 1924, 166 ff.CrossRefGoogle Scholar

²²⁵ Herod. III, 115 (tin and amber are connected in such a way as to leave no doubt that Herodotus thought of them as coming to Greece the same way).

²²⁶ Pseudo-Scymnus, 392–393.

²²⁷ See note 216.

²²⁸ Cf. Geographical Journal, 1925, 482Google Scholar. ProfGlotz, (Le travail dans la Grèce ancienne, 150)Google Scholar holds that this was the most important factor in the supposed Greek failure to exploit the Adriatic. The thesis will not stand, as amber was not the only, or the most valuable, commodity obtainable at the northern end of the sea.

²²⁹ FHG I, fr. 58; cf. Aristotle, , Hist. Anim. VIGoogle Scholar, I.

²³⁰ Polemo in Schol. on Euripides, , Hippolytus, 231Google Scholar.

²³¹ Iliad, 2, 851–2Google Scholar. .

²³² De Mir. Aus. 69. .

²³³ In Odyssey V, 371Google Scholar it certainly means a horse, know of no case where it definitely means a mule.

²³⁴ Nothing has been said, in this survey of Greek influence in the Po valley, about the Novilara stelae. I have not been able to come to any conclusion about their alphabet and date. It should be added that in citing evidence for Greek influence in Picenum, I have no intention of denying the strength of Etruscan; e.g. the graves at Fabriano (MonAnt xxxv, 273 ffGoogle Scholar.) are as Etruscan as many at Numana are Greek; and at Belmonte, neutral in position as compared with Fabriano on the Picene side of the pass to Perugia, and Numana on the sea, not far from Spina, bucchero is more frequent than Attic, and is also imitated in the local black or grey ware. Despite this foreign influence, it is clear that the style of the situlae (MonAnt xxxv, pl. 2–8) is Picene, and neither Greek nor Etruscan. The extraordinary vigour of this art is not without relevance to the character of the Picenes and the failure of the Greeks to establish themselves inland. Randall-MacIver, op. cit. p. 127, greatly exaggerates the strength of East Greek influence at Belmonte when he speaks of the ‘strongest possible Greek affinities and connections … with the Ionic art of the sixth and fifth centuries.’ The small bone and ivory figures, published in MonAnt xxxv, pl. 24 ff., are most likely Etruscan, as I was kindly informed by Mr. H. G. Payne. The resemblance of the situlae to Rhodian is limited to a common fondness for deer. I know, for instance, of no East Greek parallel for the weird objects (hardly birds) above the antlers and the backs of the animals.

²³⁵ Strabo 282.

²³⁶ Necrocorinthia, nos. 1141, 1186, 1199 (from Noicattaro, first quarter sixth century); 1346 from Monte Sannace; 1347, 1402, 1459, from Bari; 1451 (from Noicattaro second quarter of sixth century).

²³⁷ Necrocorinthia, 225. The earlier is probably developed Protocorinthian or Transitional, a generation earlier than the pots.

²³⁸ Illustrated in Mayer, Apulien, pl. XXIII. It is possible that the beginnings of Greek trade with Apulia go back much earlier than the second half of the seventh century, for, as Mr. T. J. Dunbabin pointed out to me, there is a Protocorinthian oenochoe in the Czartoryski Museum (no. I, p. 6 in CVA) which must be not far in date below the year 700, and the provenience of which is given as Apulia; cf. also Johansen, p. 89 (subgeometric aryballoi in Bari). If exact proveniences were known, it could mean that the Greeks reached Apulia at much the same time as they were colonising the south of Italy.

²³⁹ Cf., however, a Chalcidian pyxis from Ruvo (Rumpf, no. 197, pl. CLXVIII). Fifth century Attic imports occur, e.g. column crater, c. 480, published in the first fascicule of the Lecce CVA, pl. V, 3 and 4; there are a few Corinthian and many Attic pieces in the Jatta Collection at Ruvo. This does not alter the fact that Daunia in general did not import from Greece, as Ruvo (Mayer, op. cit. 83) became the centre of Peucetia early in the fifth century. Caution is indicated in accepting the date, owing to the still unsolved problems presented by the chronology of Daunian and Peucetian Geometric. On the Ruvo problem see BSR xi, 48Google Scholar.

²⁴⁰ I am indebted to Mr. T. J. Dunbabin for bringing the two latter to my notice.

²⁴¹ See Appendix I.

²⁴² Mayer, op. cit. 167.

²⁴³ Dittenberger (3rd ed.) no. 305, 11, 55 ff. .

²⁴⁴ Cf. Pliny, , N.H. III, 120Google Scholar; Pseudo-Scylax 16.

²⁴⁵ Thuc. VII, 33, 4.

²⁴⁶ Strabo, 284.

²⁴⁷ Necrocorinthia, 352.

²⁴⁸ Now in Spalato, Zara and Vienna.

²⁴⁹ It is here assumed, though not with entire confidence, that the Nesactium sculptures (see e.g. Hoernes, , Urgeschichte der Bildenden Kunst in Europa, 472 ff.Google Scholar) do not show signs of Hellenic influence. The most fundamental question, that of their date, seems, as in the case of the Novilara stelae, to be quite uncertain.