^{page 131 note 2} Made in 1930–1933, in all for seven months, extending geographically from the Aous to the borders of Akarnania.

^{page 131 note 3} Evangelides Πρακτ. 1931, p. 85.

^{page 131 note 4} Id. Πρακτ. 1930, p. 68. The pottery is as yet unpublished and I was unable to see it at the National Museum.

^{page 133 note 1} Still the best map, despite minor errors in the Pindus highlands; referred to below under sheet only.

^{page 134 note 1} Mr. Christos Soules of Janina kindly informed me that pottery was being found.

^{page 135 note 1} Not to be confused with some Hellenistic tombs E. of the Chan.

^{page 135 note 2} I am indebted to Miss Benton of the British School for the description of these sherds.

^{page 135 note 3} Mr. Cook of the British School collected some of these sherds.

^{page 136 note 1} The only obsidian reported from this area consists of two blades in the Finlay Collection in the British School, said to come from Salagóra (S. of Arta on the Ambracian Gulf); I visited Salagóra and am convinced that there was never a prehistoric settlement there.

^{page 137 note 1} As these sites are mentioned for future investigation, I should add that a mound on the right bank of the upper Aous between Mesogephyra and Molývoskepastó Moné and just to the east of the main road lies on an important route; the pottery I found there led to no definite conclusions. Evangelides, Βόρειος Ἤπειρος, p. 24, says he found hand-made prehistoric pottery on the hill of Philátes (lower Thyamis valley), and Rey, Leon, B.C.H, xlii, p. 258Google Scholar, note 2, records the presence of numerous prehistoric sites in the Aous valley. No description of the pottery is given and in neither case have I found anything on the spot to confirm their opinions; Mr. Wace, however, kindly informs me that M. Picard of the French School reported sites in the Aous valley.

^{page 137 note 2} Most of these sherds are now at the British School, c1–c5 were kindly drawn by Miss Benton.

^{page 138 note 1} Evangelides, loc. cit., quotes parallels to Dodona pottery from Heurtley, , B.S.A. xxvii, Pl iiiGoogle Scholar, 9 and xxxix, p. 166, fig. 41, 1 for the former, and from Heurtley, , B.S.A. xxvii, PL. 1, 3 (Vardaróftsa), xxviii, Pl. xiii, b 8 (Boubousti) and B.C.H, xli–xliii, Pl. 1, 1 for the latterGoogle Scholar.

^{page 138 note 2} For plastic decoration cf. ‘rope’ type at Gona in B.C.H. xli—xliii, Pl. x, 1–4 and 8–10, at Dourmouchlon ibid. Pl. x, 5, at Vardaroftsa, in B.S.A. xxvii, Pl. viGoogle Scholar, 5 and Pl. x, a, and at Hagios Mamas, xxix, fig. 13, no. 10. For mammiform decoration cf. Gona and Dourmouchlon B.C.H. xli–xliii, Pl. xii, 1–3, or Molyvopyrgo in B.S.A. xxix, fig. 43, cf. pp. 135–136. Points of similarity are sufficiently obvious to make cross-references from the pottery here illustrated unnecessary.

I must express my gratitude, as one not trained in archaeology, to Mr. Heurtley for his kindness in taking me to Macedonia, a visit to which I owe my acquaintance with Macedonian pottery in the field, and for his interest in my work; in 1931 when I showed him sherds from Koutsoulió he confirmed my opinion that they were related to Bronze Age Macedonian.

^{page 138 note 3} Using Heurtley's table in B.S.A. xxviii, PL xiv. The alternative source, Thessaly, as the secondary centre of dispersion, ca. 1800–1600 B.C., can hardly be considered in view of the lack of painted ware and other characteristics of Thessalian cultures during this period.

^{page 138 note 4} Ibid. p. 165. The Epirote sites in view of their thinness of deposit, lack of building remains, and geographical situation (e.g. Xerakía one hour above the Amphilochian plain on a wooded ridge), can only have been occupied by small shepherd communities.

^{page 138 note 5} Ibid. p. 182 and note.

^{page 139 note 1} Cf. Zeitschrift für Ethnologie xxxiii, pp. 43 ff., figs. 32, 43: similar ornamentation from Sanskimost in Bosnia, dated 600–400 B.C., Wissenschaftliche Mittheilungen aus Bosnien und Herzegovina, vi, pp. 62 ff., figs. 28, 34, 150, I51. A fibula similar to one found at Dodona (Carapanos, Dodone et ses Ruines, Pl. 51, 8) cornes from the same locality. The hammer-axe is apparently peculiar to Dalmatia and Illyria; Mr. Lef Nosi at Elbasan in Central Albania on the Via Egnatia route showed me a similar one found in the locality (for the axes from Skutari cf. illustrations in Ugolini Albania Antica Pl. xv, fig. 16). Ugolini ibid. p. 163 dates the axes to the Bronze Age; Kaer, Wiss. Mitt, aus Bosn. u. Herz, vi, p. 19 to 1250–400 B.C.; Nopcsa ibid, xii, pp. 168 ff. to the Early Roman Empire, and Reinach, , L'Anthropologie xii, pp. xcysw662Google Scholar ff., to 300–500 A.D. Vulpe, , Prähistorische Zeitschrift 1932, pp. 132 ffGoogle Scholar, has reviewed the question and decided (p. 141) for 1000 B.C.—a date which rests upon his explanation of the Cadmus legend as an echo of Phoenician colonisation; the dateable objects found with the axes are ca. 600 B.C. (ibid. p. 139). The pottery associated with these axes I should ascribe to 700–500 B.C. at latest, and connect the spread from the South with the Via Egnatia-Adriatic route which was certainly in use at this time (cf. the rich finds at Trebenischte, which would seem to imply more than a unilateral trade, whether from Macedonia or from Epidamnus and Apollonia). It should also be noted that a pottery of similar ornamentation, technique, and crude manufacture has been discovered in the Salzburg region; with the notable exception of the wish-bone handle, striking parallels are observable, cf. Caroliner-Augusteum Museum, Salzburg, Room xiv., Cases 1 no. 265, 6 nos. 294–5, 7 nos 299–301. Until this pottery can be dated more precisely, the question of an ultimate Northern origin must be shelved. There is one example of the hammer-axe at Salzburg, case 5 no. 6390, but the origin of the find is unknown.

^{page 139 note 2} Cf. fig. 2; the best maps are Leake, Travels in Northern Greece, vol. i; Murray's Handy Classical Maps, ed. Grundy, Northern Sheet; Stählin, Das hellenische Thessalien. Cf. especially the Austrian Staff Maps.

^{page 140 note 1} Hesiod fr. 90 (149):

ἐστί τις ῾Ελλοπίη, πολυληϊος ἠδ᾿ εὐλείμων ἀφνειἠ μἡλοιοσι καὶ εἱλιπόδεσσι βόεσσιν, ἐν δ᾿ ἄνδρες ναίουσι πολυρρῆνες πολυβοῦται, πολλοὶ, ἀπειρέσιοι, φῦλα θνητῶν ἀνθρώπων, ἔνθα δὲ Δωδώνη τις κ.τ.λ.

Cf. Semple Geography of the Mediterranean Region p. 323, rainfall of 40 ins.; ibid. pp. 290, 30g, 323, 331 for horses, cattle and sheep; Aristotle Hist. Anim. iii, 21; and Herodotus ix, 92 (Apollonia).

^{page 140 note 2} Cf. Semple, op. cit. p. 281. Livy xxxii, 13: montes Epiri … interiecti Macedoniae et Thessaliae … vestiti frequentibus silvis sunt; iuga summa campos patentes aquasque perennis habent (the plains are those N. of Métzovo). Wace Prehistoric Thessaly, p. 7 suggested that the low western limit of Thessalian sites was set by the forest-belt reaching lower down Pindus than it does to-day. Large forests of holm-oak are found in the upper Kalamas valley and in Amphilochia to-day.

^{page 140 note 3} Cf. Thuc. i, 136 for the conditions at the court of Admetus, and Plut. Pyrrhus 5.

^{page 140 note 4} Heurtley, B.S.A. xxviii, pp. 185–6. For the Vlachs cf. Bourcart, J.L'Albanie et les Albanaises Paris 1921Google Scholar, and Wace and Thompson The Nomads of the Balkans. The Bronze Age people, like the Vlachs to-day, pursued their nomad mode of life for centuries; the homogeneity of their culture leads Heurtley op. cit. p. 187, to assume ‘perhaps a central point of reunion’—was not this reunion formed on the plateau N. of Métzovo, the summer rendezvous of many Vlach shepherds?

^{page 141 note 1} Cf. Hdt. i, 56.

^{page 141 note 2} I have walked over most of the passes here described (part of no. 2 being done by car), and also the upper Aous and Achelous valleys. The times given are my own and are those of a walker, considered fast, carrying a rucksack.

^{page 141 note 3} C.A.H. II, p. 530.

^{page 141 note 4} Who were the Greeks? p. 151.

^{page 141 note 5} Cf. the lower Haliakmon, the upper Aous and the lower Arachthus; limestone mountains usually form deep and inaccessible gorges when cut by a river bed. The photographs shown are no exaggeration. Cf. the illustrations in Woodhouse, Aetolia.

^{page 141 note 6} Cf. Miller Itinera Romana pp. 518 ff.

^{page 141 note 7} Cf. Heurtley, B.S.A. xxviii, p. 160.Google Scholar

^{page 143 note 1} Best map in Woodhouse, Aetolia.

^{page 143 note 2} I am indebted to Mr. Wace for information concerning these routes.

^{page 143 note 3} This account and the times for this pass are taken from the note-books of the late Mr. S.S. Clark, Fellow of Exeter College; he was undoubtedly a fast walker. I have used the pass as far as Korákou and thence via Vitzista to Porta Pazári, a northern variant of the Korákou-Mouzáki section. Cf. Pls. 33, 34.

^{page 143 note 4} I reached Halíki by a high and difficult pass, rarely used now, from Koutsoulió via Prámanda and Sirákou-Kalarýtes. Cf. Pl. 34.

^{page 144 note 1} Cf. for other place-names Fick, Vorgriechische Ortsnamen p. 84, and A.J.A. xxxii, (1928) Pl. 1, p. 146. The most interesting are Θύαμις (modern Kalamás), the main central Epirus route to the Ionian Sea, and Πίνδος. Κάδμος should be withdrawn from A.J.A. as obviously added to deck the Roman localisation of Aeneas' wanderings in the plain of Phoenice. Κόσσος (Carapanos op. cit. I, p. 53 no. 5 should read ἀγ Κόσσῳ) must be added.

^{page 144 note 2} Fixed by Strabo vii, 7, 4 describing the course of the Via Egnatia; cf. Wace, in B.S.A. xviii, p. 167.Google Scholar

^{page 144 note 3} Wace, ibid. p. 168, suggests Pliasa between Kastoria and Biklishta; but it seems possible that Koritsá may provide the site.

^{page 145 note 1} Cf. Plut. Titus 3 where it is suggested to Flamininus that he could turn Philip's flank at the Aoi Stena by invading Macedonia via the εὔπορον ὁδὸν καὶ ῥαδίαν διὰ Δασσα- ρητίδος it is possible that the Koritsá-Kastoría route is meant, for the Roman settlement of 229 shows that Dassaretis was on the western side of Pindus as well as possibly on the eastern (cf. Wace, ibid. p. 176).

^{page 145 note 2} Cf. Livy xxxii, 12 ff. for Philip's retreat.

^{page 145 note 3} There are, of course, many other instances. We may mention, for pass 3 Thuc. i, 136, Themistokles from Molossia to Pydna with traders for guides; Thuc. ii, 80, 7, help from Perdiccas for Ambraciote campaign against Stratos. For 2 cf. e.g. Livy xxix 12, from Macedonia to Phoenice (near Santa Quaranta); Polybius xxvii, 16, Perseus finds Aous bridge held on his way to Molossia (i.e. probably by Mesogéphyra). For pass 4 cf. Caesar, B.C. iii, 79Google Scholar, from Apollonia via Epirus to Aeginium. For pass 8 cf. Livy xlii, 55, from Ambracia to Gomphi—asperi acprope invii soli cum ingenti dificultate parvis itineribus aegre Gomphos pervenit; Dionysius Calliphontis 24 reckons the last pass as 3 days. For pass 9 cf. Livy xlii, 53, and xliv, 2; Hdt. vii, 173, 4, κατὰ τὴν ἄνω Μακεδονίην may mean this pass.

^{page 145 note 4} Hdt. iv, 33; Callimachus, Delos, 284 ff. Cf. below.

^{page 146 note 1} While the cardinal point is clear, there is uncertainty concerning the northern half of the route followed by Perseus. The inscription found by Wace, and Thompson, (B.S.A. xvii, pp. 193 ff.)Google Scholar proves that Elimea included the Haliakmon valley down to Velventós, i.e. controlled the northern end of the Volustana pass at Sérvia (pass no. 9); it probably included Kozáne, and may have extended as far up the Haliakmon as Grevená. It is beyond the scope of this paper to discuss the localities of Elimea, Orestis, Parauaea, Paroreia, and Tymphaea; my conclusion is that Elimea and Orestis shared the upper Haliakmon; Parauaea must be the area round the upper Aous (cf. Steph. Byz. s.v.; Plut. Greek Questions 13), and Paroreia the area round the upper Arachthus (cf. Strabo vii, 7), while Tymphaea extended from the sources of the Peneius and the Velemíshti region to Aeginium (by Kalabáka). For the view that Parauaea represents the Haliakmon valley by Grevená cf. Wace, B.S.A. xviii, pp. 181 ff.Google Scholar Alexander's march (Arrian I, vii, 5) from Eordaea to Pelinnaeum clearly crossed the Métzovo plain, where I should hold Parauaea and Tymphaea touched (i.e. = Arrian's words παρὰ τὰ τῆς Στυμφαίας καὶ Παραυαίας ἄκρα).

^{page 146 note 2} Cf. Miller, Itinera Romana p. 559. The Roman coast road Oricum-Buthrotum-Actium is very difficult; cf. Caesar, B.C. iii, 6Google Scholar; the difficulty is due to the mountainous nature of the Epirote coast. For Tepeléni cf. Polyb. ii, 5–6, Scerdilaidas from Illyria to Phoenice via Antigoneia.

^{page 146 note 3} In modern Albania the Greek population has remained strikingly distinct from the Albanian; the former is limited to the Acroceraunian coastal range, and hill villages from the west side of the Gyinokástro plain to the hill villages on the west side of the Aous (Voiousa). The coastal plain of Boutrinto is Albanian. This division according to terrain recurs in Greek Epirus. The situation has caused the frontier commission considerable embarrassment.

^{page 147 note 1} Livy xlv, 29–30: frigida haec omnis diraque cultu et aspera plaga est. Strabo VII vii, 8: Μολοττοί τε καὶ ᾿Αθαμᾶνες καὶ Αἴθικες καὶ Τυμφαϊοι καὶ ᾿Ορέσται Παρωραῖ τε καὶ ᾿Ατιντᾶνες τραχεῖαν οὶκοῦντες χώραν, εἰσὶ . . . ᾿Ηπειρῶται some also assert that Epirus as far as Corcyra is Macedonian, αἰτιολογοῦντες ἄμα, ὄτι καὶ κουρᾷ καϊ διαλέκτῳ καὶ χλαμύδι καὶ ἄλλοις τοιούτοις χρῶνται παραπλησίως For the localities mentioned cf. p. 146 note 1.

^{page 147 note 2} Jason of Pherae, who expanded beyond Thessaly, is said in a laudatory passage (Xen., Hell, vi, 1, 7) to have had some control over Alketas of Molossia, defined as his ὔπαρχος the nature of this control may be judged by the freedom of Alketas in joining the Second Athenian Confederacy.

^{page 147 note 3} This route was used often; cf. e.g. Thuc. ii, 80, 8, iii, 10, 6 and Polyb. v, 3, 6.

^{page 148 note 1} Cf. T. W. Allen, The Homeric Catalogue of Ships, pp. 121–122. Is ᾿Ιθώμη κλωμακόεσσα to be placed by the famous Meteóra pinnacles (near Kalabáka)?

^{page 148 note 2} In addition to the archaeological evidence (p. 131) cf. e.g. Hdt. ii, 52 for Greek tradition.

^{page 148 note 3} Allen, op. cit. p. 133 quoting Leake, , Travels in Northern Greece i, 415Google Scholar; iv, 278 for the phenomenon mentioned by Homer at the junction of the Titaressus and Peneius. Further, the epithet ἀργροδίνης can only refer, whatever the meaning of ἀργυρο- to the upper Peneius, the Tempe region being on other grounds out of the question.

^{page 148 note 4} Later situated on eastern slopes of Pindus North of the Peneius valley; for references cf. op. cit. pp. 129–130.

^{page 148 note 5} Cf. op. cit. p. 112.

^{page 149 note 1} Prehistoric Thessaly, ch. xvi for Lianokládi, and ch. xiii for earlier East and West connections. This view has been criticised strongly on archaeological grounds, which are not in my opinion conclusive.

^{page 149 note 2} Western Inachus in Hecataeus fr. 72 Jacoby, Eastern in Plut. Greek Questions 13 (cf. Halliday ad. loc.); western Dryopis cf. Dicaearchus v, 30 p. 459 (ed. Fuhr) by Ambracia, and Pliny, , N.H. iv, 1Google Scholar by Dodona; eastern Dryopis in Hdt. viii, 31 and Strabo ix, 5, 10 f., the latter adding that Mt. Tymphrestus was once called Dryopikon; the other names need no references. Pelasgikon may have a purely religious sense (this does not invalidate the evidence), but it is interesting to notice Aesch. Supplices 245 f. where the realm of the Pelasgi stretches from the Strymon to the west as far as the Perhaebi, the western slopes of Pindus, and the mountains of Dodona. Cf. Hesiod (ed. Rzach) frag. 212. The view held by Ridgeway, Early Age of Greece, that the Achaeans came from Dodona is now untenable, but further evidence for east-west connections, which I cannot adduce here, will be found there and also in Thomson, Studies in the Odyssey, pp. 125 ff. Ridgeway's theory may be modified to bear the interpretation that the Achaeans bringing a culture from the South were the overlords of an earlier population coming from the West.

^{page 150 note 1} For iron in Homer cf. Myres, op. cit. pp. 433 ff., where only one reference and one type of iron are noted; both ‘flame-coloured’ and ‘much-belaboured’ are mentioned in the passages quoted above. For the piracy of the Taphians and the possibility of trade passing up the Adriatic Sea cf. Childe, The Bronze Age, p. 193. For the location of Taphians suggested, Od. I, 259 is favourable, Ephyra being not improbably from the poisons-trade in Thesprotia, as in other passages.

^{page 150 note 2} His father was Bouchetos of Bouchetion in Epirus, near Trampyae; cf. Allen, op. cit. p. 97.

^{page 150 note 3} The Thesprotian plain extends to Paramythia in the north, whence it takes 14 hours over wild and deserted country to reach Dodona; to the east the limit is the splendid gorge of the Acheron, turned by a détour South to the eastern end at Gourana, and then to the pass over Mt. Olytziká for Dodona in some 18 hours.

^{page 150 note 4} Excavated by Miss Benton.

^{page 150 note 5} Pace Dörpfeld supra; for Corcyra = Phaeacia note the Illyricised place-name Βαιακή in N. Epirus, Hecataeus frag. 104. I am indebted to Prof. Chadwick for pointing out to me the significance of this name.

^{page 150 note 6} Neither Carapanos nor Evangelides have found any Mycenaean remains.

^{page 151 note 1} Hdt. ii, 56; viii, 47.

^{page 152 note 1} Hdt. v, 61; cf. ix, 43; there seems good reason to suspect Herodotus of credulity here, and Thucydides in his careful distinction of Cadmeans and Boeotians may be tacitly criticising him.

^{page 152 note 2} Apart from the passage quoted, Pindus occurs twice in Herodotus. He states that the mountains West of Thessaly were called Pindus (vii, 129), and he describes the Pelopon nesian members of the fleet at Salamis (viii, 43) as ‘of the Dorian and Makednian race, who had emigrated last (ὔστατα ὁρμηθέντες) from Erineus, Pindus, and Dryopis.’ The former passage strengthens our contention in shewing that Herodotus used the name Pindus in the sense of a mountain chain; the latter passage—Erineus is a town in Doris—refers to the Pindus in Doris, and may appear to weaken our case. But a comparison of i, 56 with viii, 43 shows that Herodotus, in the former case, must mean the range: his actual words are οἴκεε ἐν Πίνδῳ Μακεδνὸν καλεόμενον ὲνθεῦτεν δὲ αὖτις ἐς τὴν Δρυοπίδα μετέβη, καὶ ἐκ τῆς Δρυοπίδος οὔτως ἐς Πελοπόννησον ἐλθὸν Δωρικὸν ἐκλήθη. ἐνθεῦτεν δὲ αὖτις must mean that Pindus is the penultimate stage, and therefore not in Dryopis, while the οὔτως stresses the fact that Dryopis was the ultimate stage. When, therefore, Herodotus says that the Peloponnesian Dorians came ὔστατα from Erineus, Pindus, and Dryopis, he is referring explicitly to the ultimate stage, the Dryopis of i, 56, and means the Pindus ofthat district; for the river (or town) of the name of Pindus was, if not actually within the subdivision Dryopis of the canton of Doris, at least closely adjacent. Cf. Pindar, , Pyth. i, 66Google Scholar, Πινδόθεν ὀρνύμενοι Schol, ad loc., Πίνδος Περραιβίας ὄρος

^{page 153 note 1} If Herodotus is supposed to err in placing Histiaeotis in N.E. instead of in N.W. Thessaly, the route is equally good.

^{page 153 note 2} Heurtley, B.S.A. xxviii, p. 177Google Scholar, chronology fixed by a Mycenaean sherd; in this important article he suggests the identification of the Boubousti people with the Dorians.

^{page 153 note 3} Hdt. viii, 137 representing Macedonian folk-memory. The claim of the Argeadae, that they came from Peloponnesian Argos, should probably be referred to Argos Oresticum; cf. Appian, Syr. 63 and Strabo vii, frag. 11. Geyer, F., Makedonien bis zur Thronbesteigung Philipps II, p. 12Google Scholar, states without adducing evidence ‘der älteste Sitz der Makedonen war aller Wahrscheinlichkeit nach die Landschaft Orestis.’

^{page 154 note 1} Cf. C.A.H, ii, 530 f.

^{page 154 note 2} Cf. supra, pp. 141 f. and fig. 7.

^{page 154 note 3} This second route would bring tne Thessalians and later the Dorians into Centrai Thessaly, whereas the Thessalians are associated with Arne on reaching the plain, and the Dorians with Doris. So far as the literary evidence alone goes, this second route can only have been subsidiary.

^{page 155 note 1} Evidence from archaeology is lacking for the other border peoples except for the Kephallenes, who must have held Astakós, where Mycenaean pottery has been found.

^{page 156 note 1} Tyrtaeus frag, I ed. Diehl dates the triple division in the case of Sparta to the seventh century. Cf. Powell, and Barber, , New Chapters in Greek Literature, III, p. 63.Google Scholar For Sicyon and Argos before the time of Cleisthenes cf. Hdt. v, 68; Dymanes occur in inscriptions from Thera (I.G. xii, 3, 337) and Cos (Paton-Hicks, Inscriptions of Cos, 367); as Thera was colonised from Sparta, where the tribal system was abolished by the Eunomia, the tripartite division probably dates from the foundation of the colony. The mythical Dyman was invented to give the Dymanes an ancestor; cf. Steph. Byz. s.v., Schol. Pind. Pyth. v, 92, οἰ Αἰγιμίου παἴδες Δύμας καὶ Πἁμφυλος συγκατῆλθον τοἴς ῾Ηρακλείδαις Further references are given by Szanto, R.E. s.v. Dyman, Dymanes. It seems probable (cf. C.A.H. ii, 528) that the epithet τριχἁϊκες used by Homer of the Cretan Dorians refers to the tribal division of Dorian peoples; Hesiod's explanation of the term (Loeb ed., fr. 8), whether correct or not, shows that a threefold division was known in his time.

^{page 156 note 2} The connection of Hyllus with the Hylleis in Central Illyria and Hyllinoi by the Naro I consider far-fetched; if there is any connection, Corinthian influence explains it.

^{page 156 note 3} The name Talaeanes appears in an inscription only (Carapanos op. cit. inscr. no. 13); no doubt there were others. In male names recorded in Epirote inscriptions the male nom. in -αν or -ην is very common. From the Spercheius valley we may add Κυλικρῆνες (F.H.G. iv, 49). For Zeus, Apollo and Pan cf. p. 158 note 7. The Paean song was associated with Macedonia, and Herodotus i, 139 tells us that the Dorians called the letter sigma σόν

^{page 157 note 1} Meteor. I, 14, 532A. Thucydides i, 3, 3 dubs the followers of Achilles ‘the first Hellenes.’

^{page 157 note 2} The earliest use of the term Hellenes and Panhellenes in a wider sense may be inferred from Hesiod Eoiae fr. 18 (Loeb ed.): ῾Ησίοδον ἤδη εἰδέναι καὶ ῾Ελληνας λεγομένους τοὺς σύμπαντας καὶ Πανέλληνας for Sparta cf. Ζεὺς ῾Ελλάνιος in the Rhetra, Plut. Lycurgus 6. For the Graikoi see R.E. s.v.

^{page 157 note 3} Gf. evidence of place-names quoted above p. 149. There may have been others, e.g. the Myrmidons who fought in Doric fashion (ἀγχιμαχηταί) cf. Classical Journal ix, 9, P. 589.

^{page 157 note 4} Hdt. viii, 43, 46, 73; Thuc. vii, 57; Apollodorus ii, 7, 7; in Hdt. i, 146 they are in cluded as non-Ionians in the Ionian migration in company with Gadmeans, Phocians, and Molossians; to Homer they are not known. They recur in Elis (Steph. Byz. s.v. Νεμέα). They were not Dorian, cf. Hecat. fr. 119 (ed. Jacoby) considering them pre-Hellenic.

^{page 157 note 5} Hellopia has a distribution coincident with the Dorian invasion; in the west there are Hellopia next Dodona and Ellopium on Lake Trichonis in Aetolia; in the east, Hellopia on the Thessalian coast, in Dolopia, in Euboea, and at Thespiae in Boeotia.

^{page 157 note 6} Thuc. ii, 102, and i, 98.

^{page 158 note 1} Plut. Greek Questions 13 and Halliday ad. loc. Strabo ix, 5, 22 gives a rationalised (and so suspicious) alternative route. They were an Amphictyonic people.

^{page 158 note 2} Plut. Greek Questions 26 and Halliday ad loc.

^{page 158 note 3} Ibid. 26.

^{page 158 note 4} Mirab. Auscult. 133; Herakles is returning after his quest for the cattle of Geryon, which was sometimes located in S. Epirus. For the Dotian plain see above and Steph. Byz. s.v. ᾿´Αμυρος

^{page 158 note 5} Heliod. ii, 34. Mr. Bachtin kindly drew my attention to this reference. Cf. Scymnus, 615.

^{page 158 note 6} The Vlach memory of Rumanian origin is a striking parallel, if they are indeed Roman colonists expelled from Dacia.

^{page 158 note 7} Cf. Myres, op., cit. p. 169 and p. 563 for Apollo; C.A.H. ii, p. 531 for Zeus of Olympia; Paus, x, 12, 10 Dodona origin of Ζεύς ἤν, Ζεύς ἐστι, κτλ. It is also interesting, in view of the West-Greek origin of the ethnic -αν, to note that the Doric form of Zeus was Ζάν the Boeotian Δήν, and the Cretan on coins ΤΑΝ Apollo's cult name, especially in Thessaly, was Παιάν. The worship of Pan, who shares this ethnic form, was said by Herodotus ii, 145 to have commenced after the Trojan War; he does not occur in Homer, and his mortal grandfather in the Homeric Hymn to Pan is Dryops. Finally we have Poseidon, in Dorian inscriptions Ποτειδᾶν and in early Doric Ποτιδᾶν Odysseus goes to Epirus to appease Poseidon, cf. infra p. 161 and note 3, and inscriptional evidence shows that Poseidon-worship was popular in Epirus. There is not space here to discuss the interesting question whether Poseidon in Epirus was a horse-god or sea-god.

^{page 159 note 1} Hymn to Apollo, 1. 315 f. Cf. Sikes and Allen ad. loc, where the reading Λέκτον is retained; no such name save in Asia Minor is known and Baumeister's emendation Λάκμον is convincing. Otherwise the sound geography noted by the authors for the remainder of the route must lapse, unless there was a Λέκτον on Lakmon. The poem is dated to the early sixth century by F. Jacoby, Die homerische Apollonhymnus p. 53, who recognises the element of cult-tradition from which its material is drawn, and by Dornseiff, Die archaische Mythener Zählung, p. 14. Jacoby, p. 59 considers this route ‘Erfindung der Delphers,’ but it is more likely to be the early tradition preserved at Delphi; he does not explain the devious route from Olympus to the Maliac gulf.

^{page 159 note 2} To avoid repetition ad nauseam of this route I may mention other instances here: Orestes visiting Dodona from Argos stops at Phthia, Eurip. Andromache 886; Io from Argolid to Adriatic via Epirus, Aesch. Prom. 848. Apollo, and Dorians, , cf. C.A.H. iii, 624.Google Scholar

^{page 159 note 3} Aelian, . Nat. Anim. xi, 2.Google Scholar

^{page 159 note 4} It is, however, noteworthy that the Aenianes in Heliodorus do not quote Homer in support of their claim, as one would expect if the claim was a quasi-logical deduction from Homeric evidence.

^{page 159 note 5} His prayer to Zeus of Dodona is interesting in this context; he traced his line to Zeus, and ancient tradition says he was descended ἀπό δρυὸς ἤ ἀπὀ πέτρης Cf. Schol. to Iliad xxii, 126 and Hermes xix, p. 544. Are Zeus and the oak both from Dodona?

^{page 159 note 6} Cf. the Albanian adoption of Alexander the Great, figured on the lekj minted by the Italians, whose political motives obscure the source of adoption.

^{page 159 note 7} Hesychius, s.v. from Arist. Opunt. Resp., and Plut. Pyrrh. I.

^{page 160 note 1} Proclus, Chrestomathia, cf. Apollodorus, , Epit. vi, 12Google Scholar from the Nostoi of Hagias dated to eighth century; evidence neglected by Wilamowitz Pindaros p. 167 on Pindar, , Nemean iv, 84Google Scholar and Nemean vii, 38 (with which cf. Paean vi, no); noted, however, by Gross, Epirus p. 100 n., though he pays no regard to the route bringing Neoptolemos to Epirus. Klotzsch, Epir. Gesch. p. 33 omits both the Cyclic Epics and the Pindaric evidence.

^{page 160 note 2} Allen, Homeric Catalogue, p. 113, adducing no evidence.

^{page 160 note 3} Plut. G.Q. 37 for Tanagra; Anaxagoras in Schol. Ap. Rhod. iv, 814 for Laconia; Paus, with Steph. Byz. s.v. Πρακνόν iii, 24, 5 for Prasiae, and Hesych. s.v. πρἁκιαι. For Elis, Paus, vi, 23, 3. The associations of all these cults suggest great antiquity. Cf. esp. Wide, Lak. Kulte pp. 233 f. (with other evidence for Achilles' name in Laconia); he omits discussion of the Epirote connection.

^{page 160 note 4} The Acheron geography has not been quoted, because the rival claims of the other cults of the dead would demand discussion; there are, however, good grounds for supposing that Hades-cults spread from Epirote Acheron with the Dorians. Thesprotian Acheron and its oracle of the dead first appears in Hdt. v, 92 in the time of Periander, tyrant of Corinth. This suggestion gains greater probability if my point re Achilles cults is conceded.

^{page 161 note 1} Cf. Powell, and Barber, , op. cit. III, p. 43Google Scholar; Herodotus vii, 185, 2; also an Illyrian tribe at the sources of the Erigon, near the Deuriopes (i.e. by Monastir), Strabo vii, 7, 8.

^{page 161 note 2} Proclus, Teleg. fr. 1 and Apollodorus, Epitome 34 (the second version given by the latter is less reliable than the first, confirmed by Proclus); also Nostoi frag. and F.H.G. III, 339. For Musaeus' Thesprotis cf. Alex, Clement. Strom, vi, 2, 25.Google Scholar

^{page 161 note 3} Tzetzes, Ap.ad Lycophron. I. 800Google Scholar, using Arist. Ithac. Resp. It is an expansion of the prophecy of Teiresias in the Nekuia, , Od. xi, 120 f.Google Scholar; cf. Wilamowitz, Homer. Untersuch. pp. 185–198, who concludes that this is part of the oldest Odysseus legends and that the ancients were correct in locating his journey in Epirus. Meyer, E. in Hermes xxx, pp. 241–288Google Scholar, arrives at the same conclusions.

^{page 161 note 4} Tzetzes, ibid. I. 1017. Aeas is the correct form of Aous in early authors, cf. Hecat. frag. 102 (ed. Jacoby) and Scylax, Periplus ch. 26. For the route we need hardly recall the obvious parallel of the return of Neoptolemos and the coming of Pythian Apollo.

^{page 161 note 5} Arist. loc. cit.; for Sparta cf. Plut. Greek Questions 48 and Halliday ad loc.

^{page 161 note 6} The view that the Boeotians were in S.W. Thessaly at the time of the Catalogue but were not mentioned therein seems untenable in view of the prominence of the Southern Boeotians. For Dorians in the Homeric world cf. E. Crete, Rhodes, and Cos.

^{page 161 note 7} The weakness of this hypothesis is shown by Allen, Hom. Catal. pp. 45 f.

^{page 162 note 1} Strabo vii, frag. 6, Boion is the general name of Pindus from Orestis to Corax and Parnassus. In vii, 7, 9 Mount Poion (the Hellenised form) is defined as lying near to Macedonia, Thessaly and the country of the Aithikes at the sources of Peneius, i.e. just N.E. of Lakmon. As a town, Boion is one of the tetrapolis in Doris, again associated with a Pindus. The only other name known to me is Boiai in Cape Malea (S.E. point of Laconia).

^{page 162 note 2} If the ethnic -atos is philologically related, one can adduce from Epirote inscriptions Klathiatos, Phoinatos, Kartatos, and possibly Horiatos (Fick, Coll. 1339, 1351, 1356 1367 1346 1366).

^{page 162 note 3} Proclus, , Chrestomathia, Phot. Bibl. cod. 239, p. 990.Google Scholar

^{page 162 note 4} Strabo ix, 2, 4.

^{page 163 note 1} Evidence is mainly inscriptional, and only the safer deductions are here cited; for Aeniania-Malis and Acarnania-Epirus, owing to the later expansion of Aetolia, and the colonisation of Corinth in the case of the latter, the evidence of inscriptions, usually late, is unreliable. I must reserve the discussion of Epirote inscriptions for a later date.

^{page 163 note 2} In Cl. Ph. xxi, p. 18. The following summary is taken from Buck, Greek Dialects (revised edition).

^{page 163 note 3} Corinthia in pre-Dorian times was likewise Aeolic; cf. Thuc. iv, 42, 2.

^{page 163 note 4} Unless Doric can be shown to be regressive in Thessaly from Pelasgiotis to Thessaliotis (in all known cases it proved intrusive), we must assume that the Dorians invaded not direct from Macedonia, nor, in their main stream, into Central Thessaly, but came into South Thessaly from the south-west; in fact pass 5 was used rather than pass 4. As we should have expected, the western Greek Thessaloi, coming from Thesprotia (Herodotus) and giving their name to Thessaliotis, where Arne is situated (Thucydides), spoke a Doric dialect. In the case of the Boiotoi, Thucydides' account of their origin gains support. The evidence of dialect in Boeotia coincides with literary tradition to shew that the Dorian conquest of Boeotia was only partial; the resistance of Boeotia may have saved Attica from the Dorians, as Demosthenes hoped in the case of the Macedonian invasion. Boeotia is turned by the route to Delphi, used by Philip, and that may account for the prominence of Delphi in Dorian tradition.

^{page 164 note 1} A.J. i, p. 201. The wording is partly my own; I must apologise for quoting a statement which Gasson has presumably discarded.

^{page 165 note 1} A. J. i, p. 204; the occupation of the site dates from 950 B.C. onwards.

^{page 165 note 2} Conversely, to identify Geometric pottery of 950 B.C. as the product of early Dorian culture is fallacious for the same reasons.

^{page 165 note 3} Attica preserved its isolation; Casson's arguments op. cit. p. 214 are unconvincing. Did Dorian intercourse come from Aegina and Euboea? Dryopes are located in the south of the latter and colonised Carystos, while others were pushed south to Hermione and Asine. The wanderings of this spear-head tribe may represent a route of Dorian invasion.

^{page 165 note 4} Cytinium has been excavated unscientifically without result; Mr. Skeat and I visited Doris for a brief period of surface exploration without success.

^{page 165 note 5} Especially Marmariane, , B.S.A. xxxi, pp. 1 f.Google Scholar

^{page 165 note 6} Lianokládi, of which the exact chronology is a matter of dispute, was certainly pre-invasion.

^{page 166 note 1} Cf. especially Blinkenberg, in Historisk-Filolog. Meddelelser xiii (1926–1927).Google Scholar

^{page 166 note 2} Myres, op. cit. p. 425.

^{page 166 note 3} E.g. Chauchitsa in B.S.A. xxiii.

^{page 166 note 4} Cf. Lamb, Greek Bronzes, p. 38; figurines are found e.g. at Athens.

^{page 166 note 5} Myres op. cit. p. 447.

^{page 166 note 6} Casson, op. cit. p. 207.

^{page 166 note 7} Ibid. p. 205, map of distribution. On the west coast they have been found at Thermon and Agrinion, as well as Dodona and Leucas.

^{page 166 note 8} E.g. Lamb, op. cit. p. 97 for Peloponnesian influence in regard to the archaic Dodonean bronzes. On the other side note that the Bronze Age culture in Epirus persists into archaic times as regards pottery.

^{page 166 note 9} Athens, Nat. Mus. 223, 298, 326.

^{page 166 note 10} One can suggest as a possibility that it returned by the route it used in reaching the north; the sea-route does not touch Dodona, while the overland Via Egnatia route passes well to the north. There is also a school for Cypriote origin, cf. Thompson L.A.A.A. v, pp. 10–12.

^{page 167 note 1} Op. cit. p. 414. Epirote type = Blinkenberg Class V. Childe, Bronze Age, p. 115 supports Myres.

^{page 167 note 2} The further fibulae found by Evangelides at Dodona (published loc. cit.) are in Blinkenberg's classification of class II two, of class V two, of class VI one, and of class VII one fine example.

^{page 167 note 3} Athens Nat. Mus. nos. 640, 645, 646.

^{page 167 note 4} At Glasinač in Bosnia; cf. finds at Kumani by Skutari of zoomorphic figures on fibulae in Ugolini, Albania Antica, p. 53 and Pl. xxxiv fig. 41. Mr. Lef Nosi of Elbasan on the Via Egnatia route in Albania snowed me a Geometric horse, with legs and belly forming a semicircle, of the type found at Dodona.

^{page 167 note 5} The origin of the duck in Geometric art, if not from the islands, where the duck-shaped vase is early, may perhaps be sought in the lake-lands of Western Macedonia, Thessaly, and Epirus; duck are common in the last region, and Leake comments on the good shooting available. With regard to the horse, objections have been raised on the ground that the Dorians fought as infantry; this argument is invalid, for the earlier use of the animal is that of the pack-horse, and cavalry is a later evolution dependent upon settlement in plains, where that arm can operate. Cf. previous note for horses in Epirus.

^{page 167 note 6} Of later finds cf. Πρακτικἁ 1931, p. 87 figs. 4 and 5; cf. Blinkenberg, loc. cit.

^{page 168 note 1} B.S.A. xxviii, pp. 158 ff.

^{page 168 note 2} A.J. i, p. 209; finds were not published by the original excavators.

^{page 168 note 3} Op. cit. pp. 456 ff., from which I draw the above account in conjunction with Heurtley's article.

^{page 168 note 4} The Vardar valley has connections with the north not only through the Morava, but also through the western Via Egnatia route. It is, moreover, noteworthy that in the Homeric period the Taphian pirates show that there was Adriatic trade (whether they or others blocked Odysseus' trade by that route), for piracy implies trade; if these pirates were as strong a bar to the north Ionian sea as Homer suggests, the overland Via Egnatia route becomes more valuable.

^{page 168 note 5} Unfortunately not described more closely; Heurtley maintains that they coincide chronologically with the latter period of the Boubousti site, in which case we must qualify our statement by saying that they are objects of peoples left behind in the invasion period but known at an earlier period in Central Macedonia.

^{page 169 note 1} Op. cit. p. 448.

^{page 169 note 2} Childe, Bronze Age, p. 23, dates it to L.M. III b.

^{page 169 note 3} Cf. Wade-Gery, , in C.A.H. II, pp. 524–525.Google Scholar The strongest evidence for the view that the introduction of iron into the Aegean was due to the Dorians, is provided by the discovery of the earliest iron weapons in East Crete, the locality of Dorians in Homer. On the other side cf. Myres, op. cit. pp. 433 f.

^{page 169 note 4} Two objects at Dodona, an axe-head and a straight-edged iron sword, show traces of northern or Lausitz influence: the latter is paralleled by those found at Halos (Myres, op. cit. p. 433); the former is tanged and of the Hallstatt type, cf. Carapanos, op. cit. Pl. liv.

^{page 170 note 1} The alternative view, that Geometric pottery is a natural development from Late Mycenaean, does not, of course, invalidate our theory; if anything, the crudity of Epirote pottery lends that view some support. Personally, I find it difficult to believe that the Dorian peoples are completely dissociated from a pottery expressive of artistic principles which are for me alien to Late Mycenaean culture. It is perhaps profitable to compare the influence of Orientalising art upon early Sparta with the influence of Late Mycenaean upon the Dorians of the invasion and post-invasion period; the Spartans adopted whole heartedly an art technically far superior to their own but later transformed it by the application of their own artistic canons.

^{page 170 note 2} A further point lies outside the scope of this paper. The Dorians according to our theory were for many centuries adjacent to Mycenaean Greece; while this helps to explain their speaking a dialect of Greek akin to Ionic and Aeolic, it does not cast any positive light upon their earlier derivation. It does, however, push back their entry into north Greece beyond the Mycenaean period; they cannot, therefore, be connected with the northern invasions which may have reached Macedonia in this period: cf. Childe, Bronze Age, p. 246.

^{page 171 note 1} For Epirus cf. fibula class v and Lamb, op. cit. p. 97.

^{page 171 note 2} The Herakles legends, for obvious reasons, have been omitted in the present inquiry.

^{page 172 note 1} E.g. in the case of Bouboústi, where the dating of the site mainly rests upon one Mycenaean sherd. We must look to the works of Mr. Heurtley and Mr. Skeat shortly to be published to extend our knowledge in this direction.

^{page 174 note 1} If Μακεδνός is held to be an early synonym, note that Hesiod, Eoiae fr. 3, uses Μακηδών

^{page 176 note 1} Cf. W. R. Halliday, Indo-European Folk Tales and Greek Legends, p. 65, for the ‘respectable antiquity’ of Hesiodic versions; and p. 78: the Hesiodic writings represent ‘the unification of historical traditions of the Greek tribes.’

^{page 177 note 1} Among the Thessaloi, for instance, the tribe of the Aleuadae, as Mr. Westlake has pointed out to me.

^{page 178 note 1} Cf. Halliday, op. cit. pp. 59 ff. for general remarks on Greek legend.

^{page 178 note 2} Cf. Thuc. iv, 53, 3 for position of Laconia.