I wish to thank E. Badian and A. B. Bosworth for many valuable discussions and helpful suggestions made during the preparation of this paper; I also wish to thank several anonymous referees for their useful comments. Obviously, none of these persons should be held responsible for those errors which still remain, nor for the arguments presented. I am grateful to Harvard University, the Westinghouse Corporation, and the Winston Churchill Foundation for their financial support during the preparation of this paper.

¹ The extreme nature of Tarn's views is well-demonstrated by a passage relating to the topic of this paper. In his Alexander the Great: sources and studies ii (Cambridge 1948) 260–2Google Scholar, he acquits Alexander of the murder of his brother Karanos by ‘debunking' Karanos’ existence, and closes his account with the words: ‘Alexander did commit two [sic!] murders in his day; there is no need to invent a third which he could not have committed.’ A naive reader is liable to exhaust Tarn's quota of killings in a single sentence of our Alexander sources; and E. Badian forcefully depicts the bloody character of Alexander's later reign of terror in JHS lxxxi (1961) 16–43Google Scholar and Studies in Greek and Roman history (Oxford 1964) 192–205Google Scholar.

² See Ellis, J. R., Philip II and Macedonian imperialism (London 1976) 15–22Google Scholar for the sources and a good discussion of the evidence.

³ References to Amyntas' alleged conspiracy are in Plut., de fort. Alex. 1.3Google Scholar, Curt, vi 9.17, 10.24; while in Arr., An. i 5.4Google Scholar, Alexander offers a foreign king the hand of Kynna, his half-sister and the erstwhile wife (and current widow?) of Amyntas.

⁴ Justin xii 6. 14–15.

⁵ Satyros in Ath. xiii 557, on which see Tronson, A. D., JHS civ (1984) 116–26CrossRefGoogle Scholar.

⁶ Justin ix 7.3.

⁷ Tarn (n. 1) 260 argues this interpretation of the passage, and the same view is held either explicitly or implicitly by Hammond, N. G. L. and Griffith, G. T., A history of Macedonia ii (Oxford 1979) 681 n. 1Google Scholar; Ellis (n. 3) 214; and Fox, R. Lane, Alexander the Great (London 1973) 503Google Scholar. Heckel, W., RFIC cvii (1979) 386–7Google Scholar considers the alternative possibility simply to dismiss it. Berve, H., Das Alexanderreich auf prosopographischer Grundlage (Munich 1926Google Scholar) s.v. ‘Karanos’ argued that Karanos was Phila's son, but his reasoning was very weak (see n. 27 below).

⁸ Justin xi 2.3.

⁹ On this point I am in complete agreement with Tarn (n. 1) 260 and Heckel (n. 7) 387.

¹⁰ Justin xi 7.12.

¹¹ Satyros (n. 5).

¹² Diod. xvii 2.

¹³ Paus. viii 7.7.

¹⁴ Lane Fox (n. 7) 503–5.

¹⁵ With the political aftermath of Chaironeiea occupying his attention, it is unlikely that Philip returned to Macedonia to marry Kleopatra until after the meeting at Korinth; this would place the marriage in spring or summer 337. Diod. xvii 2.3 says that Kleopatra's child was born a few days before Philip's death in summer 336. This would fit well, and two births are impossible. Lane Fox (n. 14) is driven to the wildly implausible conclusion that Kleopatra was already many months pregnant at the time of her marriage to Philip. The general case against two births is well-argued by Heckel (n. 7) 389–93 and the issue of the date of the marriage is discussed by Ellis (n. 2) 301 n. 1, 302 n. 4.

¹⁶ Pausanias' full account is: ἐπὶ δὲ Φιλίππῳ τελευτήσαντι Φιλίππου παῖδα νήπιον, γεγονότα δὲ ἐκ Κλεοπάτρας ἀδελφιδῆς ᾿Αττάλου, τοῦτον τὸν παῖδα ὁμοῦ τῇ μητρὶ ᾿Ολυμπιὰς ἐπὶ σκεύους χαλκοῦ πυρὸς ὑποβεβλημένου διέφθειρεν ἔλκουσα. Nowhere does the actual word ‘son’ (υίός) appear. Only two phrases specify the sex of the child: παι̑δα νήπιον γεγονότα and του̑τον τὸν παι̑δα. Both of these imply a masculine child, but the impression they give is that Pausanias' own confidence in the certainty of his information (or his memory) was not firm. The absence of the word ‘son’ makes it easy to imagine that Pausanias embellished the indeterminantly-sexed παιδίον of Diodorus or some other source into a masculine παίδα νήπιον or τόν παίδα. An anonymous referee was kind enough to point out that a parallel embellishment may have occurred in the case of Julia's baby born in 54 BC, apparently a daughter (Plut., Pomp. 53.5Google Scholar; Dio xxxix 64) but sometimes called a son (Vell, ii 47.2; nepos at Suet., Caes. 26.1Google Scholar and Lucan ix 1049).

¹⁷ Plut. Alex. 10.

¹⁸ Justin ix 8.2–4.

¹⁹ Justin viii 3.10–12.

²⁰ E.g. Justin x 1–2.

²¹ The murder of Attalos, Kleopatra's uncle is well-known (Diod. xvii 2.5). Probably Kleopatra's brother Hippostratos and various other relatives also fell in the purge.

²² Justin xi 5.1–3.

²³ See n. 3 above.

²⁴ Justin xii 6.14–15.

²⁵ E.g. Lane Fox (n. 7) 503, Ellis (n. 2) 306 n. 54, and Hornblower, S., The Greek world, 479–323 BC (London 1983) 262Google Scholar, to name a few recent authors. Various other references to the supporters of the existence of Karanos (whether as a son of Kleopatra or some other wife) are collected by Heckel (n. 7). of Philip's other seven wives; this is so thin as to be non-existent. Strangely enough, Berve also claims that Phila (a sister of Derdas and

²⁶ The existence of other brothers (and sisters) besides those mentioned in the fragment of Satyros may help to resolve another puzzle: Philip's fecundity. As it stands, for Philip to have had only six surviving children after twenty-five years of marriage to a total of seven wives seems implausible. Philip's fertility must have been high, for at the age of forty-seven he obtained a child from his last wife Kleopatra after only about one year of marriage.

²⁷ The suggestion that Karanos was the son of Phila, Philip's second wife, was accepted by Berve (n. 7) and followed by half-a-dozen other (mostly German) scholars in the last hundred years; see Heckel (n. 7) 386 n. 1 for the list. The main argument is that Satyros lists no children for Phila, though he does list children for five Machatas of Illyria) was the only ‘social peer’ of Olympias, and hence only a son of hers could be a rival to Alexander. This is completely untrue (as far as we can judge such things): Audata came from the royal Illyrian house, Meda was a daughter of the Thracian king, Kleopatra was the niece of a leading Macedonian noble, and even Philinna and Nikesipolis (about whose social background we know nothing) are usually assumed to have come from aristocratic Thessalian families.

²⁸ See Ellis (n. 2) 166–7.

²⁹ Satyros (n. 5). The same verb ἐπεισάγω is used in each case.