¹ This research began during the course of three months' work in Greece made possible by the generosity of the Myer Foundation, the Australian Humanities Research Council and Monash University. I am extremely grateful also to the Institute for Balkan Studies and to its director Mr Basil Laourdas for enabling me to attend the Institute's symposium on Ancient Macedonia, at which I was able to discuss various points with several colleagues, and to Professors Ernst Badian, Charles Edson, Christian Habicht and Malcolm McGregor for advice and encouragement. A short, earlier version of my main thesis was published in Ancient Macedonia, 1st International Congress, 1968 (Thessaloniki 1970) 68 ff.

² In that he states that the regency lasted ‘a long time’; this will be discussed below.

³

⁴ xvi 1.3.

⁵ vii 5.9–10.

⁶ Curtius vi 9.17, 10.24, Polyainos viii 60.

⁷ Hammond, , CQ xxxi (1937) 81–2, 85–, 150.Google Scholar

⁸ cf. Sherman in Loeb, vol. vii, 236 f. n. 2, and, on the length of time Philip was held hostage in Thebes, , Aymard, , REA lvi (1954) 15–36.Google Scholar

⁹ cf., for example, Diog. Laert. ii 56, Synkellos p. 500.

¹⁰ For the date, Beloch, Griech. Gesch. iii 1.229.Google Scholar

¹¹ For the date, Hammond, , JHS lvii (1937) 57.Google Scholar

¹² For the date, Ehrhardt, , CQ n.s. xvii (1967) 298–301.Google Scholar

¹³ For the date, Hammond, loc. cit., 57–8.

¹⁴ For the date, Beloch iii, 1.230.

¹⁵ Some (for example, Hogarth, , Philip and Alexander of Macedon [London 1897] 42Google Scholar, Pickard-Cambridge, , CAH vi 203Google Scholar) suggest that Philip assumed the throne immediately after his appointment as regent, while others put it later (Schäfer, , Dem. und seine Zeit ii 16 ff.Google Scholar, at 359/8, but Beloch iii 1.232, Glotz, , Hist. Grec, iii 226Google Scholar, Momigliano, , Filippo il Macedone 53Google Scholar, all date it around the foundation of Philippoi in 356). Niese, , Gesch. des Hell, i 27 f.Google Scholar, Berve, Das Alexanderreich ii No. 61, are noncommittal.

¹⁶ Hist. of Greece, transl. (London 1896) iii 205.

¹⁷ Köhler, , Hermes xxiv (1889) 640–31Google Scholar. IG vii 3055 shows both versions, first published in Pococke, R., Inscriptiones Antiquae, P. 1 c. 5 s. 5 p. 61Google Scholar, and Leake, W. M., Travels in Northern Greece ii 129, 132 and pl. vii no. 32Google Scholar. The two copies are usefully set out in alternating lines by Meister in Collitz, H., Samml. der griechischen Dialekt-Inschriften i 156–9.Google Scholar

¹⁸ Leake, op. cit., 129.

¹⁹ Twenty-six names or fragments of names (including repetitions) are preserved in what appears to be most of the original inscription; for a detailed treatment see my ‘IG vii 3055’, forthcoming in Hermes.

²⁰ Only the first is specifically recorded as having done so, but there are so many gaps in the text that it is impossible to be certain. For details of the ritual observed in consultation of the oracle see Kroll in PW s. ‘Trophonios’ no. 4, and, most important of the ancient sources, Pausanias ix 39. Also Parke, , Oracles of Zeus 232.Google Scholar

²¹ Meister, ap. Collitz, loc. cit.; cf. Keil's emendation, ibid., 159.

²² 393/2–370/69, less two years for Argaios' reign; see my ‘Amyntas III, Illyria and Olynthos’ in Makedonika ix (1969) 1–7, Beloch iii 2.57 ff.

²³ loc. cit., 641.

²⁴ With this apparent confirmation of Justin's notice the only remaining problem was to establish the length of the regency. Just what did Justin mean by his diu? And even if this were not to be taken literally, when, in any case, did Philip become king? See n. 15 above.

²⁵ IG vii 4251, first published by Leonardos in AE iii (1891) 108 no. 51. This and its twin, 4250, first published ibid., no. 50, were variously dated: by Dittenberger, SIG i³ no. 258, pre-338, by Tod ii no. 164 A and B, about 350, by Hicks and Hill, GHI ² no. 142, around the middle of the fourth century, and in IG vii between 366 and 338.

²⁶ Attractive as it might be, the restoration of the royal title would leave no room at all for the ethnic, which must surely be a sine qua non of any proxeny-decree; see Klaffenbach, , Griech. Epig. (Göttingen 1966) 80–3Google Scholar, esp. 80. The only other alternative along these lines, that the engraver erased the title and inserted the ethnic in its place before starting the word, is possible but specious. It has been argued, moreover, that the marks remaining on the stone do not support Dittenberger's conjecture; Leonardos, , AE 1919, 64a.Google Scholar

²⁷ As well as the exact coincidence of form and for most of their length precisely the same dispensation of letters in each line, both stones bear erasures under the words ἔλєξє ἔδοξє (line 2). Whatever may have been cut originally—and at so early a stage, still in the preamble, it is unlikely to have been historically significant—exactly the same error or alteration was made with both. Clearly they are exactly synchronous.

²⁸ Arrian i 17.9, D.S. xvii 48.2, Curtius iii 11.18, Plut. Al. 20. He fled, says Berve, (Das Alexanderreich ii 28Google Scholar) ‘vermutlich im Jahre 335’.

²⁹ Curtius vi 9.17, 10.24, Plut. de fort. Al. i 3 (Mor. 327C), Justin xii 6.14.

³⁰ Geyer, PW s. ‘Perdikkas’ no. 3; Aischines ii 28 f. for the regency of Ptolemaios (though his chronology seems completely askew; cf. Aymard, , REA lvi [1954] 19).Google Scholar

³¹ Curtius iii 11.18.

³² ibid., Arrian i 17.9, D.S. xvii 48.2, Plut. Al. 20.

³³ Arrian i 25.

³⁴ ibid., Curtius vii 1.6–7, Justin xi 2.2

³⁵ Arrian i 25.

³⁶ D.S. xvii 32.1.

³⁷ D.S. xvii 80.2, Curtius vii 1.9, Justin xii 14.1.

³⁸ Petrakos, B. Ch., Ἐπιγραφαὶ Ὠρωποῦ; in ADelt xxi (1966) 45–7Google Scholar. I am grateful to Professor C. Habicht for pointing out to me the existence of this stone.

³⁹ ii 13.2.

⁴⁰ Petrakos, loc. cit. with fig. 1 and pl. 23.

⁴¹ D.S. xvi 2.6.

⁴² ibid.

⁴³ Justin vii 4.5, viii 3.10.

⁴⁴ Hammond, CQ, xxxi (1937) esp. 85–9.Google Scholar

⁴⁵ Athenaeus xiii 557B, 560F.

⁴⁶ Plut, de fort. Al. i 3.

⁴⁷ Curtius vi 9.17, 10.24 f.

⁴⁸ Arrian i 7–8.

⁴⁹ The Oropian proxeny-grants (which may have numbered originally more than the two preserved) presumably represent at least moral support for the pretender. DrCoulton, J. J. (BSA lxiii [1968] 147 ff.Google Scholar, esp. 181) has conjectured on the basis of ‘by no means conclusive’ stylistic evidence that the proxenies may have been conferred in recognition of a private benefaction—the donation of the stoa in the Oropian Amphiaraion—but he has kindly informed me that a lowering of the date of these inscriptions to 335 would almost certainly exclude their connection with the stoa, which he would prefer to date higher, rather than lower, than 360. The same motivation, moral support, may be supposed to have inspired the use of the royal title in the Lebadeian inscription. As I have proposed elsewhere (see n. 19), the Lebadeians erred badly in their judgement of the situation, it seems; they may well indeed have been among those in mid-335 who thought Alexander dead (see n. 52) and they may additionally have played safe by their peculiar and enigmatic use of ὑπὲρ αὐτοσαυτῶ (line 9) in the reference to Amyntas' consultation of the oracle. Since their action was in technical violation of the charter of the League of Corinth (IG ii² 236 ll. 11 ff.), we presumably owe the survival of the stone undefaced to the relative obscurity of the oracle and its home.

⁵⁰ Arrian i 1.4.

⁵¹ Arrian i 7 ff.

⁵² Arrian i 7.6.

⁵³ We note from Curtius vi 9.17, 10.24 f. that the genuineness of the Amyntas-plot is unquestioned; the only point under dispute is Philotas' part in it.

⁵⁴ Justin xi 5.1.

⁵⁵ The term is Edson's; see CPh lvi (1961) 198–203.

⁵⁶ There is no inherent difficulty in supposing that the Heeresversammlung could bypass Amyntas Perdikka for Philip in 359—at least, it should not be difficult for those (to my knowledge, everyone nowadays) who can accept the notion that it acted to replace the already acclaimed successor by his uncle at some later date.

⁵⁷ For refs. see Walbank, , Philip V of Macedon 3–4, 295–9.Google Scholar

⁵⁸ The date is disputed; see ibid., 295 f.

⁵⁹ Arrian succ. 22, Polyainos viii 60, Plut, de fort. Al. i 3, Curtius vi 9.17, 10.24, Justin xii 6.14.

⁶⁰ See n. 34.

⁶¹ Arr. i 20.10.

⁶² Justin ix 5.8.

⁶³ Arr. i 12.7, 14.1, 6, I5–1. He had probably also commanded these prodromoi on the advanceparty; Brunt, , JHS lxxxiii (1963) 27.CrossRefGoogle Scholar

⁶⁴ Arr. i 28.4.

⁶⁵ Das Alexanderreich ii no. 59.

⁶⁶ Arrian i 25.2.

⁶⁷ Phoenix xvii (1963) 244–50, esp. 248.

⁶⁸ See, for a brief treatment, Wilcken, , Alexander the Great (Norton 1967) chap. 2Google Scholar. I have treated the subject in some detail elsewhere (The Unification of Macedonia under Philip II, diss. [Monash University 1970]).

⁶⁹ Wilcken, op. cit., 24.

⁷⁰ de fort. Al. i 3.

⁷¹ ibid.

⁷² The total area of Macedonia as Philip left it was in the region of 16,000 square miles (compared with less than 11,000 in 359); I find Beloch's calculation (iii 1.294, 312 f.) a little low but since he gives no map I am unable to check his location of the borders. Of the area at 336 about 5,500 (or one third) represent Upper Macedonia and 1,550 the Chalkidian peninsula, the latter, as Beloch notes, being the most densely populated area in the now expanded Macedonian state. It emerges from a consideration of the figures for Alexander's forces in 334 (detailed in D.S. xvii 17) and from indications in the Alexander-sources that six taxeis of territorially levied pezetairoi were taken with the expeditionary force (Berve i 112 ff, Tarn, , Alexander the Great ii 53 ff.Google Scholar), while 12,000 of these troops were left with Antipatros—making another six to eight taxeis. Of the twelve to fourteen total, three were from Upper Macedonia (D.S. xvii 57.2; these were all taken by the king), giving this area one quarter, or a little under, of the total levy. In central and eastern Macedonia most of the inhabitants were undoubtedly concentrated on the plains where most of the settlements were located (Casson, , Macedonia, Thrace and Illyria 79–86, 88–90Google Scholar). Upper Macedonia had very few cities and the area of its plains was small. Its economy, to judge by the terrain, must have been largely pastoral rather than agricultural. Therefore it seems incredible that its population-density can have been anywhere near 75 per cent of that for central and eastern Macedonia, which suggests strongly that its levy was heavier than the national average. Further, there appears to have been only one ile of Companion Cavalry for the whole of Upper Macedonia (Arrian i 2.5, Berve i 105), out of a total of at least six territorially recruited (Arrian i 2.5 with iii 11.8, i 12.7, 14.1–6, 15.13, ii 9.3); since the identifiable areas represented are only a relatively small part of the whole state, it seems likely that the 1500 cavalry left with Antipatros represent another six to eight squadrons (Tarn, Alexander ii 156Google Scholar, Brunt, , JHS lxxxiii [1963] 35Google Scholar, Milns, , JHS lxxxvi [1966] 167CrossRefGoogle Scholar concur in putting the complement per squadron at about 200, with perhaps 300 in the Royal Squadron, which was not territorially levied). There may have been political reasons for the low Upper Macedonian representation in the élite cavalry, but the apparent discrepancy between its proportion of the cavalry and the infantry forces again suggests a heavy infantry-levy.

⁷³ Berve i 105.

⁷⁴ D.S. xvii 57.2; cf. Arrian i 2.5.

⁷⁵ Badian, op. cit. (n.67), 248, suggests that Alexander had to delay action until he was well away from Macedonia.

⁷⁶ It is not clear that Alexander was guilty of patricide (through the agency of the assassin Pausanias; D.S. xvi 92), but as Badian shows (op. cit.) he must stand at the head of any list of suspects on grounds of motive and opportunity.

⁷⁷ Attalos' tactless but revealing remark at the celebration of Philip's marriage to the well born Kleopatra (Plut. Al. 9.5:… makes this clear. Although Plutarch claims that Attalos was drunk, it should be stressed that for Attalos to have made this invocation he must have been very confident that Philip had no intention of leaving the kingdom to his hitherto intended heir.

⁷⁸ Badian (loc. cit.) details the known stages of their growing estrangement.

⁷⁹ loc. cit., 248.

⁸⁰ On the significance for Alexander's personal power see Milns, , Alexander the Great (Robert Hale 1968) 143 f., 159 ff.Google Scholar