¹ My thanks are due to D. M. Lewis, who read the penultimate draft of this paper and made helpful suggestions; to P. M. Fraser for help with the geographers; to Dr S. Mitchell for topographical advice; and to Marion Cox for drawing the map.

² For his revised views see his Teubner edition of 1959, where the parallel passages from other authors are most usefully collected.

³ Hereafter references to Diodoros are to Book xiii, and references to Xenophon to Hellenica Book i.

⁴ Bartoletti's mistake was pointed out, and corrected in this sense, by Lotze, D., Abh.Leipz. lvii. I (1964) 21 n.3Google Scholar.

⁵ Maas did not complete the name Ἀ[ν]τ̣[ίοχον, from a doubt whether καταδύω could take a personl object; but cf. Xen., Anab. i 3.17, vii 2.13Google Scholar.

⁶ Since the exact height of these columns is not known, we cannot be more precise.

⁷ This is not what Hell.Oxy. 4 implies, but we do not know for certain that Ephoros followed him exactly. Pausanias does not give a very careful rendering: his battle takes place ‘not far from the city of Kolophon’, by which he no doubt meant from Notion as the harbour of Kolophon, as restored in the papyrus at 26–7, but he has forgotten that Kolophon looks inland from the other side of the mountain.

⁸ Riv.Fil. lix (1931) 222–9Google Scholar = Studi di storia di storiografia greca (Florence 1951) 163–71Google Scholar.

⁹ Meyer, , Gda iv ² 2.335 n. 1, on p. 336Google Scholar.

¹⁰ On Antiochos' standing as κυβερνήτης, cf. Amit, M., Grazer Beiträge iii (1975) 9–11Google Scholar.

¹¹ Op. cit. (n. 9) 335.

¹² Thrasyboulos is here (49.1) designated as ‘the commander of the whole fleet’, but at this stage they have not yet joined up with the main fleet at Kardia. The apparent subordination of Theramenes to Thrasyboulos can hardly be real, and at 50.7 παρεκελεύσατο does not mean that the latter treated Theramenes as a subordinate during the battle. Thrasyboulos owed his generalship to the sailors' assembly at Samos in the previous summer (Thuc. viii 76.2), Theramenes was presumably appointed by a meeting of the Five Thousand at Athens; Diodoros may have misunderstood some comment in his original on the relations between two sets of generals.

¹³ This does not certainly imply that Ephoros left it out, but it is an item that one might have expected to appeal to Diodoros. At Diod. 46.2–3 Alkibiades was in the Hellespont, and it is likely that Ephoros explained at some point how he came to be elsewhere.

¹⁴ At 50.2 Vogel conjectured that the καί found in one branch of the MSS. between ναῦς and μόνας should be taken as κ̅=εἴκοσι, and he referred to Xen. 1. 18. Littman, R.J., TAPA xcix (1968) 267 and 269 n.6Google Scholar, treats this as if it were the established text, but Vogel's conjecture is not very plausible. As to the fact, Plutarch's forty ships would be more likely to draw out the whole fleet of Mindaros, and Littman 269 sensibly prefers this figure.

¹⁵ Ancient writers use Kyzikos for the name of the ‘island’ as well as the city. Pliny, NH v 40.142Google Scholar gives Arctonnesus as an earlier name for Cyzicus; Stephanos s.vv. Ἄρκτων νῆσος and Κύζικος probably means the same by his ἐκαλεῖτο; cf. also Ap. Rhod. i 941. Hirschfeld, , RE ii (1896) 1172Google Scholar, remarked that Arktonnesos for the whole peninsula is a modern usage; but it is convenient to have a distinct name for the peninsula, and the usage is likely to continue, as here.

¹⁶ The question is most thoroughly discussed by Ruge, , RE xii (1925) 228 fGoogle Scholar.

¹⁷ Dr Mitchell confirms my impression, formed without very close inspection.

¹⁸ This point was put to me by G. S. Kirk after the lecture referred to on p. 15. It applies of course to an ambush of any kind, not only one at Artake.

¹⁹ This is not happily expressed, for it was ‘to the land’ that they retreated from the actual sea-battle. Probably Plutarch's original meant that Mindaros was not yet far enough out to be cut off (cf. Plutarch's πρὸ τοῦ λιμένος), and it was feared that he would escape back safely into the harbour.

²⁰ Ehrhardt, C., Phoenix xxiv (1970) 225–8CrossRefGoogle Scholar, drew attention to the incompatibility of the two accounts, but inevitably failed to make much out of the incomplete data provided by Diodoros.

²¹ In this deer-laden context Meineke's alteration of Stephanos' text to Νεβρίς is attractive, but Pliny's Neuris tells against it.

²² When Anaximenes (FGrH 72 F 26) includes Artake among the colonies of Miletos, that tells us nothing about its status in his own time. Eudoxos, (fr. 336 Lasserre)Google Scholar is more problematic. Strabo xiii 1.4, 582, reports controversy over the extent of Aiolis and the Troad: Homer started the Troad from the Aisepos, Eudoxos ἀπὸ Πριάπου καὶ Ἀρτάκης τοῦ ἐν τῇ Κυζικηνῶν νήσῳ χωρίου ἀνταίροντος τῷ Πριάπῳ, reducing its extent. As it stands this is ‘plain nonsense’, as Leaf, (Strabo 47)Google Scholar robustly put it, for Artake is the wrong side of the Aisepos for Eudoxos' argument. Gisinger, F., Die Erdbeschreibung des Eudoxos von Knidos (Leipzig 1921) 65CrossRefGoogle Scholar, suggested that Artake was brought in to fix the position of Priapos more precisely: Strabo uses ἀνταίρειν to locate a place on the same latitude as another (LSJ s.v. II.2) and Artake is nearly due east of Priapos, but it is not clear how this information would help, and if this was what Eudoxos meant Strabo's wording is very misleading. Lasserre, F., Die Fragmente des Eudoxos von Knidos (Berlin 1966) 244Google Scholar, refers to Strabo vii fr. 58 and suggests that the argument was really about the eastern limit of the Hellespont; but the controversies in vii and xiii appear to be quite distinct and I see no good reason to amalgamate them. More probably καὶ Ἀρτάκης … τῷ Πριάπῳ is an addition by Strabo, or (as Leaf thought) an interpolation, in which case there is no evidence that Eudoxos mentioned Artake or that it was still extant in his time. Artake may have revived in some degree in the Hellenistic period. There are many grave inscriptions said (with various degrees of authority) to come from Erdek in Schwertheim, E., Inschr. gr. Städte aus Kleinasien xviii, Kyzikos I (Bonn 1980)Google Scholar, and if they were a safe guide the revival might have started in the second century BC (his nos 146, 330).

²³ Of Rhodes, an officer of Ptolemy Philadelphos, whose main geographical work was On Harbours: see Fraser, P. M., Ptolemaic Alexandria (Oxford 1972) i 522, 536–7Google Scholar. Wagner, E. A., Die Erdbeschreibung des Timosthenes von Rhodos (Leipzig 1888) 55Google Scholar, contributes nothing for my present question.

²⁴ Even the less plausible MS. readings end in -aeon or -eon, suggesting a Greek genitive plural: does something like Ἀρτακαίων νῆσος lie behind this? The regular ethnic is Ἀρτακηνός, as in the tribute lists, but Stephanos cites Sophokles for Ἀρτακεύς and Demosthenes Bithynos (fr. 6 Powell) for Ἀρτάκιος.

²⁵ It is hard to feel much confidence in Πολυποδουσσαῖος, known only from Stephanos, the next entry after Polydora.

²⁶ Neither Priam's son, nor any other Polydoros in legend or history, appears to have any known connection with this region.