¹ The chronology is that of Hammond, N.G.L. and Griffith, G.T., A History of Macedonia 2 (Oxford 1979). I refer hereafter to this History as HM.Google Scholar

² Diod. 15.13.2 mentions 500 ‘Greek panoplies’, i.e. sets of hoplite equipment, sent by Dionysius of Syracuse to Illyrian allies c.385 B.C. The Molossians seem to have lacked hoplites in the reign of Arybbas (see Frontin. 2.5.19); in the fourth century the training of élite infantry seems to have begun with Alexander II c.369 B.C., to whom I believe that Anaximenes refers (FGrH 72 F 4), not to Alexander III as G.T. Griffith argues in HM 2.705 ff.

³ Bardylis was evidently playing for time, so that he could bring up ‘the great forces’ which were being prepared for the invasion of ‘Macedonia’ (Diod. 16.2.6), i.e. Lower Macedonia. It was from Lower Macedonia and Eordaea presumably that Philip raised his army. Philip moved more quickly than his opponent now and on other occasions.

⁴ The Loeb translation ‘came out to meet the enemy’ implies coming out of the Macedonian cities; but that is not the meaning of

⁵ The verb was used by Xenophon for the violent head-on clash of Agesilaus and the Thebans at Coronea (HG 4.3.19

⁶ The Loeb translation ‘to ride past the ranks of the barbarians’ is incorrect. The verb is used absolutely here, as in Thuc. 7.78.3 and 6. If had been governed by the words would have been with that verb.

⁷ The Loeb translation, ‘as the horsemen pressed on from the flank and the rear’, lacks the precision of the Greek text.

⁸ Paus. 9.40.7–9.

⁹ This passage was not used at all by G.T. Griffith in his account of the battle in HM 2.213 f., although I had referred to it in my account A History of Greece to 322 B.C. (Oxford 1959) 538. He may have thought that it was about a different battle. Beloch, K.J.Griechische Geschichte 3².1 (Berlin 1922) 226Google Scholar with n.2, saw that Diodorus and Frontinus were writing of the same battle; but he miscontrued the course of the battle with his idea of offensive and defensive wings and his obsession with Parmenion.

¹⁰ The original reading was probably ‘Hillyrios’, because the aspirated form occurred in a late fifth century Greek inscription (Meiggs-Lewis, GHI 79.39 and 43) and ‘Hilurii’ in the Latin Fasti of 167 B.C. The original reading was then corrupted into ‘Hyllios’ and ‘yllyrios’. The Loeb translation is imprecise; for C.E. Bennett translated ‘latera’ as ‘flanks’ and ‘sinistrum latus’ as ‘the enemy’s left’.

¹¹ The word (here means originally an oblong frame for making bricks. The corresponding military formation, well described by Asclepiodotus (Tact. 11.6) as saved infantry from being charged by enemy cavalrv but not from their missiles (e.g. in Thuc. 7.78.2). In our battle the rectangular formation was not a protection against the pikemen-phalanx of Philip, as G.T. Griffith suggested in HM 2.213; for Philip’s entire phalanx could contract into close order, charge and overwhelm the Illyrian front with superior numbers.

¹² The Illyrian cavalry either had already engaged and been routed, or it had been withdrawn by Bardylis.

¹³ Asclepiod. Tact. 5.1 for pike and shield and 4.1 and 3 for close order.

¹⁴ The point needs to be emphasised, because it has been denied, e.g. by Markle, M.M. in AJA 82 (1978) 486, who put the invention of the pike much later.CrossRefGoogle Scholar

¹⁵ FGrH 115 (Theopompus) F 348, on which see G.T. Griffith in HM 2.705 f.

¹⁶ These passages do not support the view of Bosworth, A.B. that ‘Upper Macedonia was notoriously deficient in cities’ (A Historical Commentary on Arrian’s History of Alexander 1.252).Google Scholar

¹⁷ SEG 23.471.13, on which see Hammond, N.G.L., Epirus (Oxford 1967) 527 f. For the places mentioned in this and the remaining two paragraphs see the map (fig. 1).Google Scholar

¹⁸ This is evidently the territory held by the Illyrians, in which they were occupying some Macedonian cities.

¹⁹ For this pass and for its probable use by the Illyrians in 423 B.C. see HM 1.42 f. and 107.

²⁰ Greek writers or recorders used not the local tribal name but the general name, which was intelligible to Greek readers. The point is of importance, because Papazoglou, F. in Historia 14 (1965) 143 ff.Google Scholar and again in Les Illyriense et les Albanais,ed. Garašanin, M. (Belgrade 1988)189Google Scholar has argued for the existence of an Illyrian State and has identified Bardylis as king of that State. In BSA 61 (1966)239ff. I disagreed, my view being that there were a number of Illyrian kings, each ruling over his own cluster of tribes, and that Bardylis was the strongest such king c.400-358. Another view has been advanced by Carlier, P. in L’lllyrie meridionale et I’Epire dans I’antiquité, ed. Cabanes, P. (Adosa 1987) 42,Google Scholar that the term ‘the king of the Illyrians’ meant the supreme hegemon of the Illyrians, recognised as military commander by the majority of the Illyrians. However, there is no suggestion of such a system in our sources; we hear rather of the tribes fighting one another (Strabo C 316 init. and 317 fin.), and we do not see any example of such a hegemony, e.g. in 335 B.C. when the kings Glaucias and Cleitus acted as equals and the Autariatae were intending to make a separate attack (Arr. An. 1.5.1). It was the very lack of such a unification of ‘the majority of tribes’ that enabled Philip to defeat them one by one — Bardylis, Grabus, ‘Dardano ceterosque finitimos’, Pleuratus, and Pleurias (see HM 2.21 for their tribes).

²¹ The literary evidence which I am citing now shows that the Dardanians were Illyrians according to writers of the fourth century B.C. and not that the Dardanians and the Illyrians were ‘deux nations distinctes’, as Papazoglou, F. maintains in Les Illyriens et lesAlbanais 174 and 179.Google Scholar

²² In HM 2.471 G.T. Griffith wrote thus: ‘ “finitimos” could referto “neighbours” either of the Dardanians or of Macedonia’. But this is refuted by ‘ceteros’ which must mean ‘the other neighbours’ of Macedonia and cannot mean ‘the other neighbours’ of the Dardanians, for whom no neighbour has been mentioned.

²³ The other two tribes were far from the borders of the Macedonia of 359: the Ardiaei centred on Kotor and the Autariatae operating towards the Danube basin (see BSA 61 [1966] 249).

²⁴ They were neighbours of the Paeonians (see BSA 61 [1966] 250).

²⁵ ‘Illyrian Bardylis’ is the accepted emendation of ‘Balalirem Illyricum’. The Dardanii were a cluster of tribes with that common name, and their king added other Illyrian tribes to enlarge his power-block (we may compare the Macedones and the Molossi and their kings’ actions).

²⁶ G.T. Griffith in HM 2.471, ‘Hammond’s identification of them [the Dardanii] as the tribe of Bardylis seems certain’, appeals to me more than the view of Hatzopoulos, M.B in L’lllyrie méridionale et l’pire dans l’antiquité 84 ff.,Google Scholar that Bardylis ruled not over Dardania but only over the area from the Lakes to the area north of Molossis (i.e. to Parauaea?). In the opinion of Strabo C 326 the tribes of that area were mainly Greekspeaking and not Illyrian. Ellis, J.R. accepted Bardylis as king of Dardania in Philip II and Macedonian Imperialism (London 1976) 44; he referred (250 n.3) to my article in BSA 61 (1966) 239 ff.Google Scholar

²⁷ The view from Philip’s frontier towards Albania is magnificent. I am fortunate to have travelled through all the country mentioned in this article. I am grateful to Dr T. Ryan of the University of Newcastle for help in our discussion of the passages in Trogus and Justin mentioned in this article.