¹ As Visiting Fellow at the British School of Archaeology in Athens I was able to explore part of the Northern Megarid in late May and early June 1953. The Director of the School, Mr. J. M. Cook, accompanied me on one visit, and gave me the benefit of his help and advice. To him and to Professors H. A. Thompson and E. Vanderpool, who read an early draft of this paper, and to the Managers of the Leverhulme Research Fellowships, I wish to express my gratitude. The sherds, to which reference is made, are in the Museum of the British School. The best maps of the area are the 1945 reproduction of the Greek Staff map 1/100,000, sheets 19 (Khalkis) and 18 (Lebadeia), and, of the series showing the civil divisions of Greece, the sheet for the Eparkhia Megaridos, 1/100,000, produced by the Γενικὴ Στατιστικὴ ῾Yπηρεσία τῆς Ἑλλάδος Mr. C. W. J. Eliot kindly read and commented on the Appendix.

² Paus. I 38.8 (after the incorporation of Eleutherae in Attica).

³ See the Appendix at the end of the paper.

⁴ Grundy, G. B., The Great Persian War, 446–7.Google Scholar He writes of a Plataea-Megara road as if it were separate from the Erythrae-Megara road, and in this he is followed to some extent by Kahrstedt, U. in AM (1932)Google Scholar, map facing p. 16. In fact, the routes join just on the south side of the ridge-top of the Cithaeron range. Grundy seems not to have walked south towards Megara. He simply remarks on ‘the troublesome hill region of the Northern Megarid’. To speak of ‘ancient’ wheel-ruts is perhaps misleading; for they are indistinguishable from mediaeval or modern wheel-ruts.

⁵ See the Appendix.

⁶ Described in Perachora I 3 with fig. 2. It is there suggested that in 198 B.C. Philocles brought an army of 1,500 men along this coastal route; but the passage (Livy XXXII 23. 11) to which reference is made does not define the route, and in all probability Philocles used the normal route ‘through Gerania’.

⁷ Paus. I 44. 10; Strabo 391 (the road passed above the cliffs); Gell, W., Itinerary of the Morea (1817), 209Google Scholar; ‘There is a road to Kinetta, passing the sites of Crommyon and Sidus in the way, along the coast toward Megara. Beyond Kinetta is the Kaki Scala, anciently the Scironian rocks, and Megara; but the way is neglected, and travellers usually go over Geranion, by the Derveni, to Megara.’

⁸ Gell, W., Itinerary of Greece (1819), 6–9.Google Scholar His times were Megara to Kháni Dervéni two hours thirty-three minutes, and Kháni Dervéni to Corinth six hours.

⁹ Gell, op. cit. 7 mentions a road turning off near Kháni Dervéni for Aegosthena.

¹⁰ The modern road through Attica touches Mándra, the outskirts of Eleusis, and Megara. In Turkish times a carriageable route, which is more direct, was maintained from Megara to Koundoúra; taking this route, Gell's times from the Isthmus via Megara and Koundoúra to Kriekoúki amount to more than fourteen hours (Gell, op. cit. 11–12; 109).

¹¹ This route has been either overlooked, e.g. by Wallace, W. P., Phoenix, Suppl. Vol. I (1952), 82Google Scholar, who speaks of the ‘roadless hills’ (of the Northern Megarid), or regarded as of minor importance, e.g. by Highbarger, E. L., History and Civilisation of Ancient Megara 7Google Scholar, and by Gomme, A.W., BSA XVIII (1911/1912), 193Google Scholar, who writes: ‘a route from Megara to Plataea, somewhat to the east [this should be west] of the Eleutherae pass; it is mentioned by Pausanias and Xenophon, but was very little known or used’. More attention is given to the route in RE XV 170–1, where it figures as an ‘Abzweigung’ but ‘ebenfalls wichtige Strassenstück’, and by Curtius, and Kaupert, , Karten von Attika 39Google Scholar, who describe it as a ‘Hauptweg aus der Megaris’. In comparing this road with the ordinary mule-track I was helped by some three years of walking in North and Central Greece.

¹² Buchon, J. A., La Grèce continentale et la Morée (1843), 555–8.Google Scholar

¹³ BSA XII (1905/6), 101–8.

¹⁴ BSA XII (1905/6), 105, Tillyard says ‘almost every joint is filled with mortar’. This is true only of the vertical joints, many of which are sprung.

¹⁵ Mr. R. V. Nicholls of the British School kindly dated the fragment of ridge-tile.

¹⁶ Perachora I 8, with fig. 3 c.

¹⁷ Scranton, R. L., Greek Walls 81, 176.Google Scholar The round tower at Munychia is pre–337/6 (Scranton, op. cit. 119 and n. 32). In associating round towers with the time of Philip II, I rest on the use of the semi-circular towers at Cytinium, which Philip fortified, and at Ambracia, Paramythià, Goúmini, and Ljábovo in Epirus, where Philip's influence may have been at work.

¹⁸ The site at Pírgos is not marked on Milchhöfer's Übersichtskarte von Attika.

¹⁹ Thuc. V 72. 3 refers to ‘the waggons’ at the battle of Mantinea, assuming it to be known that waggons carried the baggage, etc.; cf. Hdt. VII 40. 1 and Xenophon, HG IV 1.Google Scholar 24 for the Persian use of waggons. The Spartan army had a regular supply corps (Xenophon Lac. 13. 4). On the difficult route from Creusis to Aegosthena Cleombrotus used pack-animals only (Xenophon, HG V 4. 17).Google Scholar This was, I think, exceptional. For during the war, when I moved with forces of guerilla troops in Greece, the problem of carrying supplies even for 300 men by pack-animais was serious. The idea that an ancient army lived on the countryside which it invaded is often fallacious; for, if the flocks had been driven away and it was not harvest-time, the chance of finding much foodstuff was small.

²⁰ Itinerary of Greece 12.

²¹ Plutarch, , Per. 33.Google Scholar

²² Thuc. II 10. 2; 12. 4–5; 13. 1; 18. 1. On the position of Oenoe see the Appendix. Wallace, , Phoenix, Suppl. Vol. I (1952), 82Google Scholar, believing the northern Megarid to be roadless, has suggested that the Spartan force moved by sea across the Gulf of Corinth; a naval operation on this unprecedented scale by the Peloponnesians would surely have been recorded by Thucydides.

²³ Thuc. I 114. 2.

²⁴ Tod GHI 4.1. Cf. Gomme, , Commentary on Thucydides, I 340.Google Scholar

²⁵ The former alternative seems more likely, since it allows time for Pericles to hear of the revolt of Megara and the Peloponnesian preparations, to return from Euboea, and to be in position at Athens before the invasion took place.

²⁶ Xenophon, HG V 4. 14.Google Scholar

²⁷ Xenophon, HG V 4. 16–18.Google Scholar

²⁸ For the meaning of παριόντες, see Xenophon, HG V 19 and LS⁹s.v., I.Google Scholar

²⁹ Xenophon, HG V 4. 36–60.Google Scholar

³⁰ Thuc. I 107–8.

³¹ Thuc. V 57. 2.

³² Xenophon, HG VI 4. 25–6.Google Scholar

³³ Ant. Itin. ed. Wesseling 326; ed. Cuntz 49. Tab. Peut. route 79 (ed. K. Miller, col. 565) gives the stations; Crusa (Creusis)–Pache(Pagae)–Megara–Isthmus, and route 81 (col. 577): Plataea, –Eleusis, –Megara, . Anon. Rav. 198.Google Scholar 9 (ed. Pinder and Parthey) gives Athens–Eleusis–Megara–Pache(Pagae)–Cerusa (Creusis); 199. 1 gives Eleusis–Plataea–Thespiae; and 375. 15 gives Athens–Eleusis–Megara–Isthmus. In this net of main roads the stretch from Creusis to Pagae must have included the Road of the Towers.

³⁴ Even for a maritime state such as Athens transport by land was cheaper than transport by sea (Thuc. VII 28. 1).

³⁵ The direct route from Megara to Pagae, which the modern road follows, passes well below Tripodiscus, and it presumably did so in ancient times.

³⁶ Plutarch, , Greek Question 17.Google Scholar Many ancient and modern place-names in the Megarid are descriptive of natural features or products: Pagae, Crommyon, Schoenus, Aigeiros, Plátanos, Vathikhóri, Psátha, Cynosoura, Koróna, Kérata, Karídhi (walnut), Kaliakoúdha (crow), Mítikas (nose).

³⁷ Aeschylus, Ag. 302Google Scholar; λίμνην δ᾿ ὑπὲρ Γοργῶπιν ἔσκηψεν φάος The word Γοργῶπις may be a descriptive epithet or a place-name, similar to those mentioned in the last note. That the ‘goat-traversed’ mountain is Mt. Gerania seems most probable; for it lies on the line towards Mt. Arachnaeus, and in line 360 Aeschylus mentions a headland jutting into the Saronic Gulf, which fits the stage of the beacon-signal from Gerania to Arachnaeus. Such signals from Cithaeron to Gerania and the Peloponnese may have been employed in the Persian Wars during the campaign of Plataea. It has been suggested that Aeschylus' lake should be identified with Lake Mavrolímni or Lake Vouliagméni. But the former is north of Mt. Gerania, and the latter west of Mt. Gerania, so that neither is on the line of fire from Mt. Cithaeron. For recent discussions see Perachora I 94; RE XXI (2nd series) 1371 and XV 169; Fraenkel, , Aeschylus Ag. II 160.Google Scholar

³⁸ Other locations have been suggested (cf. Halliday, W. R., Plutarch's Greek Questions 219Google Scholar). Near Lake Vouliagméni there are remains of a very small village, and it may have been possible to reach it by waggon from Thermae (Loutráki); but it is off the route to Delphi, and from early times this area was Corinthian and not Megarian. The Lake Mavrolímni is near a coastal route, but this route was probably impracticable for a waggon. A supposed lake near Pagae would have been described as near Pagae and not as near Aigeiroi (assuming Aigeiroi to be a place-name). Lakes in other places were called ‘Gorgopis’, the most famous being at Corinth (Cratinas in Hesych. s.v. ‘Γοργῶπις’). The passage in Etym. Magn. 384. 32 refers probably to three or more lakes.

³⁹ FGrH 115 F 241.

⁴⁰ The Boeotian town, whose territory marched with that of Oenoe, was probably Hysiae. Oenoe was thus contiguous with Aegosthena, Plataea, Hysiae, and Eleutherae. In the civil divisions of today Víllia is contiguous with Porto Yermenó, Plataea, Kriekoúki, and Mázi (see the map cited at the end of note 1 above). In Klio XI (1911) 436 f. Beloch identified Gyphtokastro with Panactum, Eleutherae with Myupolis, which lies east of Gyphtokastro, and Oenoe with a site to the south on the edge of the Eleusinian plain; in addition, he made the territory of Eleutherae contiguous with the Megarid. His arguments are not convincing, and his conclusions seem to be impossible; for Thucydides (V 3. 5, and II 18. 1) placed both Panactum and Oenoe on the Boeotian border. This is inconceivable if the territory of Eleutherae marched with Megaris; for, if Eleutherae was Attic, Boeotia had no access to Oenoe, and, if Eleutherae was Boeotian, Panactum was not on the border of Attica at all. The suggestion made by Leake, , NG II 408Google Scholar, that the Megarian town Ereneia (Paus. I 44. 8) might be placed at Koundoúra, seems to me improbable on geographical grounds.