¹ The following abbreviations are used:

References are to the Teubner text of 1972 ed. O. Seel. Translations are my own. I am most grateful for helpful comments on the first draft of this article by Dr E. Baynham of the University of Newcastle, N.S.W., where it was written. Ellis, J. R., Philip II and Macedonian Imperialism (London, 1976Google Scholar and Cawkwell, G. L., Philip of Macedon (London, 1978)Google Scholar did not discuss the sources of Justin at all. HM 2 (Griffith) mentioned passages from Justin with the addition ‘? Theopompus’ (208f. with the footnotes) or ‘Theopompus’ (560), and ‘the Greek source(s) of Trogus’ (627), but usually without any such qualification. I opted for Theopompus as Justin's source in HM 2.9 and 12 for the oracle about Caranus – in this article I put forward a different view – but I did not express a view in other references to Justin. In THA, pp. 88ff., I argued that Cleitarchus was the source of Justin 9.5.8–9.6.8 and 9.8.1–3, and that Satyrus was the source of Justin 9.7.1–14. I adhere to those arguments in this article.

Most scholars have discussed the sources of Justin (Trogus) in general terms rather than in specific books of Enmann, Justin. A., Untersuchungen über die Quellen des Pompeius Trogus für die griechische und sicilische Geschichte (Dorpat, 1880)Google Scholar was an exception, but he did not deal with books 7, 8 and 9. Of the divide between book 6 and book 7 he was content with the general remark: ‘damit beginnt auch das Gebiet des Theopomp, der von nun an an die Stelle des . Ephorus tritt’ (p. 111). An early theory, that Trogus followed Timagenes, who himself had drawn on various sources, was overthrown by Seel, O., Die Praefatio des Pompeius Trogus (Erlangen, 1955), pp. 18ffGoogle Scholar. The literature on the subject was well reviewed by Forni, E., Valore storico e fonti di Pompeo Trogo (Urbino, 1958), pp. 21–44Google Scholar, to which should be added the bibliography in O. Seel's Teubner text of Justin, pp. xix–xxviii.

² For these and other criticisms of their methods see CQ 31 (1937), 79–81.Google Scholar

³ The MSS. reading ‘Aegeadas’ is preferable to O. Abel's emendation to ‘Argeadas’, because the latter lacks the onomatopoeic element, the sound of ‘bleating’.

⁴ See HM 2 (Hammond), pp. 98f.

⁵ See HM 2 (Hammond), pp. 182f. and Philip of Macedon, p. 168.

⁶ A genealogy is provided in HM 2 (Hammond) facing p. 176; see p. 182.

⁷ The correct relationship of Arybbas to Olympias was paternal uncle; so Cross, G. N., Epirus (Cambridge, 1932), p. 39.Google Scholar

⁸ Philip of Macedon, p. 174.Google Scholar

⁹ The corresponding passages in Diod. 16.2.5–3.6 and 16.4.1–7 provided many more facts but showed less psychological insight. A modern account of the period is given in HM 2 (Griffith), pp. 210–12.

¹⁰ This point was made at greater length by Isocrates in Philippus 43–4 and in Ep. 3.

¹¹ For the building of fortifications at this time see HM 2 (Hammond), pp. 653f. and Hammond, MS, pp. 154–9; and the current excavations at Vergina (Aegeae) and Pella, where Eurydice's shrine to Eukleia and the lay-out of the Pella acropolis fall within Philip's reign. See HM 1.165 for the gateway at Edessa, and my remark in HM 2.670 ‘very large sums were available evidently for the building of new towns and for the fortification and embellishment of old towns in the years after 348.’

¹² The misery of the dispossessed was a favourite topic for rhetorical display, as in D. 19.65 and Livy 32.13.5–9.

¹³ This passage is incompatible with the ‘story’ (λγουσ τινες) in Diod. 16.87 (cf. Plut. Demosth. 20.3) that immediately after the victory Philip and his Friends had a drunken party, Philip became abusive, and he was brought to his senses by Demades, . I attributed the Diodorus passage to Diyllus as source in CQ 31 (1937), 84 and 90.Google Scholar

¹⁴ ‘Parricide’ is appropriate for a Greek writer. Macedonians saw the elimination of a pretender to the throne as the penalty for treason.

¹⁵ The loss of his kingdom by Arybbas had been mentioned already at 7.6.12. The repetition here suggests that the passages belong to different Groups, as we are maintaining.

¹⁶ See Heckel, Marsyas for his career.

¹⁷ For example, Alexander and his Companion ran a race naked in honour of Achilles and garlanded the tomb of Achilles in the Troad (Plut. Alex. 15.8).

¹⁸ Caranus was introduced into the Temenid genealogy by Archelaus; so Parke, H. W. and Wormell, D. E. W., The Delphic Oracle (Oxford, 1956), i.64Google Scholar, HM 2 (Hammond), pp. 5 and 9, and for the head of Caranus on the coins of Archelaus and subsequent kings see Hammond, in Ancient Greek Art and Iconography, ed. Moon, W. G. (Wisconsin, 1983), p. 253Google Scholar. Greenwalt, W. argued in GRBS 26 (1985), 49Google Scholar that Caranus was introduced first by Amyntas III, but he failed to deal with the arguments in favour of Archelaus.

¹⁹ See Hammond, AG ², pp. 264 and 271.

²⁰ See Heckel, 459, who makes similar divisions. I differ from him as regards the meaning of Ἀλεξνδρου ’Αγωγ (T 1) and τ περ Ἀλξανδρου (F 2 and F 3). He thinks on p. 459 that both titles refer to the last two books of Makedonika, whereas I believe that the former was a separate work, as in T 1.

²¹ Heckel attributed F 13 and F 14 to Marsyas of Philippi.

²² For references and further points see CQ 31 (1937), 81Google Scholar. The quotation is from ‘Le fonti della Storia Greca e Macedone nel libro xvi di Diod.’, Rend. 1st. Lombardo 65 (1932), 523–43, p. 530.Google Scholar

²³ Barber, G. L. in The Oxford Classical Dictionary ², edd. Hammond, N. G. L. and Scullard, H. H. (Oxford, 1970), p. 1059.Google Scholar

²⁴ It is important to remember that Polybius was familiar with all the work of Theopompus, whereas we have only fragments by which to judge. Polybius knew that Theopompus was praising Philip highly in this sentence (as Theopompus did also in F 256). The attempt by Connor, W. R. in GRBS 8 (1967), 138fGoogle Scholar. to make the sentence ‘ironic’, to suppose that Polybius failed to realise this, and even to think he (Connor) has a better understanding of the Greek expression τοιοτον ἄνδρα is misjudged. Polybius, after all, knew his author and his own language better than any of us moderns can hope to do.

²⁵ The view of Polybius is supported by that of Dionysius of Halicarnassus who wrote thus of Theopompus. He was uniquely able ‘to examine the hidden motives of actions and of actors, investigate all the states of the psyche which are not easily discovered by most men, and reveal all the mysteries of seeming virtue and latent vice’ (FGrHist 115 F 20a (7)). The heavy drinking of Philip and his Companions in Theopompus F 236 took place when the Athenian envoys were about to negotiate, i.e. some days after the battle. It is not mentioned by Justin.

²⁶ The fact that Theopompus' title, Philippicae Historiae, was adopted by Trogus for his own Universal History is a strong indication that Trogus used the account of Theopompus for the reign of Philip.

²⁷ In particular it is remarkable that neither Trogus nor Justin imposed a uniform style on the material they incorporated. The simple style, almost the ‘running style’ (εἰρομνη λξις) of Marsyas Macedori survives in Justin 7.1.1–7.4.2, whereas the antithetic periodic style of Theopompus appears still in Justin 7.6.3–5 and 8.1.1–4. Judging from the fragments alone Heckel, p. 462, wrote that Marsyas ‘appears to have written in an unadorned straightforward style, complete yet concise’, and he contrasted that style with ‘the flamboyance and verbosity of Theopompus’.

²⁸ Justin meant Amyntas III. But Justin's description of him as ‘Amyntas, the son of Menelaus, the brother of Alexander’ fitted not Amyntas III but Amyntas II; see HM 2 (Hammond), chart facing p. 176. Justin made this error while abbreviating as he shifted from the material of one source to the material from another source in Trogus.

²⁹ The phrase ‘ut est ingressus imperium’ at 7.6.1 refers to Philip embarking on his own reign; for it was to be ‘while one of the sons of Amyntas was reigning’ (regnante) that the oracle was to be fulfilled. Griffith in HM 2.209 n. 1 missed this particular point.

³⁰ See HM 2 (Griffith), p. 225.

³¹ It is most unfortunate that both authors are disregarded in The Spectre of Philip by Ellis, J. R. and Milns, R. D. (Sydney, 1970).Google Scholar

³² That the Macedonians had some claim on the Upper Macedonian tribes was clear from Thuc. 2.99, but commentators did not make the connection between that passage and Justin 7.4.1 until I did so in HM 2.63f.

³³ As I argued in HM 2.651 n. 1, disagreeing with Griffith in HM 2.268f. and 702f., and with Ellis, , Amyntas, pp. 15ff.Google Scholar, who both rejected a period of regency. The ‘graviora bella’ of Just. 7.5.10, which threatened, in early 357 in my opinion, were wars with Athens (Philip having seized Amphipolis), the Chalcidian League, and Grabus the Illyrian – all three being in alliance – while the Paeonians and t he Thracian kings were also hostile as they showed in spring 356.

³⁴ As I pointed out in HM 2.183. There was a shrine to Eukleia dedicated by Eurydice, which M. Andronikos has excavated, and Eurydice was one of the family group of whom statues in gold and ivory were dedicated in the Philippeum at Olympia. To suppose that she was known to have murdered two of her sons and betrayed her husband Amyntas III is little short of absurd. Yet Justin's account has been accepted by Beloch, Geyer, Cloché and others.

³⁵ The allegations in Justin 9.7.1–2 have been taken seriously, for instance, by Beloch and by Badian, E. in Phoenix 17 (1963), 244ff.CrossRefGoogle Scholar, whose arguments were approved by Hamilton C, p. 28.

³⁶ The description which Justin 9.5.1–7 gave of the settlement made by Philip is particularly important – namely the terms of a peace (‘pacis legem’) for all Greece (cf. ‘universae Graeciae’ at 9.3.11), the council with representatives of all states except Sparta (‘consilium omnium’), the fixing of forces in the event of the king being attacked or a war being declared under his leadership (indicating a defensive and offensive alliance between the Greeks of the common peace and Macedonia), and the forces available for war against Persia being those of the Greeks, the Macedonians and the neighbouring non-Greek subject races. Theopompus as a contemporary cannot have given a false account for a contemporary readership. See my discussion in HM 3 (1988), pp. 571ffGoogle Scholar. and especially 573, being in disagreement with the interpretation of HM 2 (Griffith), pp. 623ff. and his view that ‘either Justin or Pompeius Trogus or the Greek source(s) of Trogus must have misunderstood something in the terms of the treaty’. Yet, why should they misunderstand?

³⁷ That Diodorus composed these passages out of his own head was argued in CQ 31 (1937), 91Google Scholar and in THA, pp. 28 and 173 n. 15.

³⁸ It is probable that this comparison was derived from some such comparison by Theopompus; for he wrote an encomium of Philip and of Alexander, and also a censure of Alexander (FGrHist 115 F 255 and 258). For example, Justin 9.8.15 ‘patri mos erat etiam de convivio in hostem procurrere’ may be an echo of Theopompus F 282: ἦν γρ πολυπτης κα ποππκις μεθων ξεβοθει.

³⁹ See CQ 31 (1937), 88Google Scholar for the source of the Proem being Ephorus.