Note. This paper was read on 26 May 1944 at a Joint Meeting of the Oxford Philological Society and the Oxford Branch of the Classical Association. I take this opportunity of thanking critical suggestions, and in particular Drs. F. Jacoby, A. Momigliano, and P. Treves.

^{page 1 note 1} On Polybius' sources for the First Punic War (i. 6–64) see Luterbacher, J., Philol. lxvi, 1907, 396–426Google Scholar; Reuss, F., Phihl. lx, 1901,102–48;Google ScholarPhiiol. lxviii, 1909, 410–27Google Scholar; De Sanctis, , Storia dei Romani, iii. 1 (1916), 224–47Google Scholar(with references to earlier literature); Sisto, L., Atene e Roma, xii, 1931, 176 ff.Google Scholar; and Gelzer, M., Hermes, lxviii, 1933, 133–42Google Scholar.

^{page 1 note 2} Gelzer, M., Hermes, lxviii, 1933, 129–66Google Scholar: ‘Römische Politik bei Fabius Pictor’; Hermes, lxix, 1934, 46–55Google Scholar: ‘Der Anfang römischer Geschichtsschreibung.’ Gelzer's picture of Fabius was to some extent anticipated in Leo, , Gesch. der röm. Literature, i. 85 ffGoogle Scholar.

^{page 1 note 3} See below, Addendum, pp. 15 ff., for a genera criticism of Gelzer's theory.

^{page 1 note 4} The older view of Fabius as an annalist pure simple had been supported by Mommsen's view (Rōmische Forschungen, ii. 272 ff.) that he was the source of Diodorus' annalistic notices (most conveniently consulted in Diodors rōmische Annalen bis 302 a. Chr. samt dem Ineditum Vaticanum herausgegeben Drachmann, von A. B. (Kleine Texte, Bonn, 1912)Google Scholar. But this theory has frequently been exploded. Cf. Klotz, A., Rhein. Mus. lxxxvi, 1937, 206–24Google Scholar, developing a view previously expressed by Ed. Meyer.

^{page 1 note 5} Cf. Jacoby, in Fr. Gr. Hist., Commentary on p. 598Google Scholar; contra Beloch, , Griech. Geseh. iv. 2. 11Google Scholar.

^{page 1 note 6} This is generally accepted; see authorities quoted by Gelzer, , Hermes, lxviii, 1933, 133 II. 5Google Scholar.

^{page 1 note 7} R. Laqueur, P-W (1938), s.v. ‘Philinos (8)’, cols. 2180–93.

^{page 2 note 1} Viz. 16. 1 (263), 17. 6 (262), 20. 4 (261), 21. 4 and 22.1 (260), 24.9 (258), 25.1 (257: one omitted), 26. 11 (256), 36. 10 (255), 38. 6 (254), 39. 1 (253), 39. 8 (251), 39.15 (250).

^{page 2 note 2} Viz. 49. 3 (249: P. Claudius Pulcher), 52. 5(249: L. Iunius Pullus, attributed to 248), 59. 8 (242: C. Lutatius Catulus).

^{page 2 note 3} Cf. Nepos, , Cato, 3Google Scholar; Pliny, , Nat. Hist. viii. 11Google Scholar: listed by Peter, Hist. Rom. Frag. ‘M. Porcius Cato', T. ₂ and F. 88.

^{page 2 note 4} Cf. Beloch, , Römische Geschichte, 132Google Scholar ff.

^{page 2 note 5} e.g. 41. 4 (fourteenth year), 56. 2 (eighteenth year); 56. 11 (σχεδν π τρεῖς νιαντος), 58. 6 (σ ἒτη πλιν),59. 1 (ἒτη σχεδν ἢδη πντε), 63. 4 (ἒτη…εἲκοστ κα ττταρα).

^{page 2 note 6} Peter, Hist. Rom. Frag. F. 19.

^{page 2 note 7} So correctly Klotz, , Rhein. Mus. lxxxvi, 1937, 215Google Scholar. This fragment had already previously been associated with Cato's practice by Schwartz, P-W, s.v. ‘Diodoros (38)’, col. 693.

^{page 2 note 8} Viz. ii. 19. 8 (Lucius Caecilius and Manius Curius; Polybius calls Lucius ‘praetor’, but he was fact consul for 283; cf. Beloch, , Röm. Gesch. 133)Google Scholar; 21.7 (232: M. Aemilius Lepidus); 23.6 (225: L. Aemilius and C. Atilius). C. Flaminius is given as tribune in 232 (21.8); on the date see Mommsen, , Röm. Forsch. ii. 401, n. 23Google Scholar. Peter, Hist. Rom. Frag. F. 23 (Oros. iv. 13. 5 = Eutrop. iii. 5) also mentions consuls by name, and goes on to quote Fabius as a source. But it is safer to draw noconclusions from this, as the reference will have come via Livy. One may ignore the suggestion of Pais, E., Storia critica di Roma, iii. 5. 66 n.Google Scholar, that Polybius' account of the Gallic Wars is derived from Timaeus or Philinus!

^{page 3 note 1} Cf. Beloch, , Romische Geschichte, 100Google Scholar.

^{page 3 note 2} It is worth while examining exactly how Polybius uses consul names. Admittedly they referserve to define the year (cf. Beloch, , Griech. Gesch. iv. 2. 11Google Scholar: ‘bis zum Jahre 250 ist die Anordnung annalistisch’), but not in the rigid annalistic way to be seen in Diodorus or Livy. Polybius invariably introduces them in connexion with some historical action; and if, as in 259 (cf. 24.8), there was nothing worth recording, the consuls are omitted. (The omission of the Alinames of the consuls at the beginning of 250 (41. 2) is due to their having been mentioned already in 39. 15.) For the second half of the war the principle is the same; names occur in connexion with important events. Hence because the years between Hamilcar's arrival in Sicily in 247 and the last battle in 242 are passed over in a rapid, schematic fashion, no consul's name is given for that period. Polybius mentions P. Claudius Pulcher (249), his colleague, L. Iunius Pullus, whom he wrongly believed to be one of the pair for 248, and finally C. Lutatius but not his colleague) in 242. The point is this: the omission or inclusion of consul names follows upon the character of the events described, and cannot be magnified into a major criterion for distinguishing Fabius from Philinus. Prima facie any reference to a consul may have come from either source; and the serious result of putting too high an importance upon such references is to be seen in Laqueur's discussion of those passages in Polybius i (22. 1; 26. 11, and 36. 10) where consul names occur in the middle of sections which he admits to be derived from Fabius Pictor. No one will be ready to believe that in each of these cases Polybius made a cross-reference from Philinus.

^{page 3 note 3} Fabius' history, like those of Cincius Alinames mentus and Cato (cf. Gelzer, , Hermes, lxviii, 1933, 129 n. 2)Google Scholar was fairly detailed in discussing the origines, sketchy for the intermediate years, and fuller again from the time of the Punic Wars onward; cf. Dionys, . Hal. i. 6. 2Google Scholar; Corn, . Nep. Cato, 3. 3Google Scholar. See further, below, p. 18, n. 1.

^{page 3 note 4} Cf. Jacoby, , Fr. Gr. Hist, ii D, p. 598Google Scholar.

^{page 3 note 5} Op. cit., col. 2190 following Unger, , Rhein. Mus. xxxiv, 1879, 90 ffGoogle Scholar. Täaubler, E., Vorgesch. d. II. pun. Kr. (1921), 118–19Google Scholar, supports the same thesis, transposing Diod. xxv. 5. 3 and 9, and referring fg. 9 to Matho's last fight (Polyb. i. 87, 8–10); but it clearly refers to Spain. See in general Schwartz, P-W, s.v. ‘Diodoros (38)’, col. 689; De Sanctis, , op. cit. iii. 1Google Scholar. 385, n. 10; and especially Mommsen, , Rὅom. Forsch. ii. 266, n. 60Google Scholar.

^{page 4 note 1} Beloch, , Griech. Gesch. iv. 2. 12Google Scholar; Wendling, , Hermes, xxviii, 1893, 335 ff.Google Scholar, also assumes an intermediate source, later than Poseidonius, between Philinus and Diodorus; contra Jacoby, , Fr. Gr. Hist, ii D, p. 598, line 40Google Scholar.

^{page 4 note 2} On the whole, Polybius’ source for the Mer-cenary War seems to be rather more hostile towards Hanno, and more uncritically pro-Barcine than Philinus. Contrast the picture of Hanno in Polyb. i. 72. 3 with the favourable account of his humane treatment of Heca-tompylos in Diod. xxiv. 10. 1–2 (cf. the vague-ness of Polyb. i. 73. 1).

^{page 4 note 3} Meltzer, , Geschichte der Karthager, ii. 550–1Google Scholar; Beloch, , Griech. Gesch. iv. 2.10–11Google Scholar; cf. De Sanctis, , op. cit. iii. 1. 95, n. 11Google Scholar; von Stauffenberg, S., Kὅonig Hieron der Zweite, 19, 93Google Scholar. Beloch points out that Philinus would hardly have been so favourably disposed towards Hiero, who had deserted the Carthaginians; and also that Poly-bius’ account corresponds very closely with that in Justinus, who summarized Trogus, who in turn used Timaeus. Schwartz, P-W, s.v. ‘Dio-doros (38)’, col. 688 claims that it is impossible to determine Diodorus’ source in book xxii.

^{page 4 note 4} Reuss, F., Philol. Ixviii, 1909, 412Google Scholar; Laqueur, op. cit., cols. 2181–2. Laqueur makes Timaeus the source of Diodorus, xxii, 13, whereas in 1936 (cf. P-W, s.v. ‘Timaios (3)’, col. 108) he was doubtful whether Timaeus went beyond 272, and did not carry his analysis of Diodorus past book xx.

^{page 4 note 5} Laqueur shows that Polybius’ use of βáρ-βαροι to describe the Mamertini in i. 9 may not be used (as De Sanctis, claims, op. cit. ii. 1. 225Google Scholar) to prove that Timaeus was here the source, since the word occurs too in 11. 7, which cannot derive from Timaeus. Equally, however, this use cannot be taken as proof that Philinus was the source of i. 9, since more than one Greek historian may have called the Mamertini barbarians, or indeed Polybius may have repeated the word in i. 11 after using it twice (following Timaeus) in i. 9. Only a very rigid view of Polybius' relation to his sources would deny the possibility of his having used a word which was not in his immediate source, but which he had himself employed two chapters earlier. Hence the verdict must be non liquet.

^{page 4 note 6} Kunz, Margrit, Zur Beurteilung der Prooemien in Diodors historischer Bibliothek, Diss. Zϋrich, 1935, 13–14Google Scholar.

^{page 4 note 7} Cf. De Sanctis, , op. cit. iii. 1. 224Google Scholar. This view is based on the character of Diodorus’ description of the siege of Agrigentum (xxiii. 8) and of the expedition of Regulus (xxiii. 17), and on the general character of the tradition: it seems likely.

^{page 5 note 1} De Sanctis, , op. cit. iii. I. 252Google Scholar; Beloch, , Griech Gesch. iv. 2. 285–6Google Scholar.

^{page 5 note 2} On the Regulus saga and its growth see the excellent article by Klebs in P-W, s.v. ‘Atilius (51)’, cols. 2088–92; De Sanctis, , op. cit. iii. 1. 154–6Google Scholar. I am not convinced by the attempt of Frank, T., Class. Phil. xxx, 1926, 311 ff.Google Scholar, to defend the authenticity of Regulus’ peace-mission; the legend seems to stand or fall as a whole. Pais's, E.attempt to defend the whole legend in Ricerche sulla sloria e sul dirilto pubblico di Roma, iv (1921), 411–37Google Scholar, is uncritical.

^{page 5 note 3} Polyb. i. 35. 5: ἓν σοϕòν βούλευμα τàς πολλàς χρας νικȃ (Eurip., Antiope, fr. 31 Dindorf = Nauck, Tr. gr. fr.² 200). Euripides was in fact referring to the strength of autocracy as against ochlocracy, but Polybius may well be quoting from some collection of proverbs; Themistius, Or. 16, p. 207 D refers to the line as famous, and it is quoted by Plutarch, Sextus Empir., Galen, Eustathius, and others; cf. Wunderer, C., Pofybios-Forschungen, ii (1901), 57–8, 87Google Scholar.

^{page 6 note 1} See JHS, lviii, 1938, 55–68Google Scholar, for a full discus-sion of this tragic treatment of Philip's last years; see especially Polyb. xv. 20. 5.

^{page 6 note 2} e.g. iv. 81. 5 (Cheilon's assassination of the ephors, τς τύχης àρμóςουσαν αὐτοῖς πιθɛίσης δίκην); xx. 7. 2 (the downfall of the Boeotians, σπερ πίτηδες νταπóδοσιν ñ τóχη ποιουμνη βαρως Ӗδοξεν αύτοῖς πɛμβαίνɛιν); cf. i. 84. io where the agent is τó δαιμòνιον.

^{page 6 note 3} e.g. xv. 20. 5 (a person ɛίκóτως τ τύχη μεμϕáμενος πί νθρωπείων πραγμáτων); xvi. 32. 5 (one is justified in blaming Fortune for not saving Abydus); xxxii. 4. 3 (Fortune some-times grants a fine death to the worst men, a blameworthy procedure).

^{page 6 note 4} Polyb. i. 86. 7. This was the aspect of Tyche stressed by Demetrius of Phalerum; cf. Polyb. xxix. 21. 3–6.

^{page 6 note 5} Polyb. xxxix. 8. 2 (τήν τύχην ὡς Ӗστιν γαθή ϕθονσαι τοῖς νθρώποις— particularly when we think that our life has been most blessed and most successful); cf. too Diod. xxvii. 6. 2 (νμεσίς τις θεο); xxvii. 15. 2; xxxi. 11. 3—all passages derived from Polybius.

^{page 6 note 1} Diod. xx. 70; for the latest discussion see Ullman, B. L., TAPhA, lxxiii, 1942, 39Google Scholar. Ullman makes the source Duris, with Schwartz, P-W, s.v. ‘Duris (3)’, cols. 1853–6; but he admits that Schubert, R., Geschichte des Agathokles (1887), 181Google Scholar, and Laqueur, P-W, s.v. ‘Timaios’, col. 1174 (cf. G. Pasquali, Sttid. Ital. Fil. Class, xvi, 1939, 76–7) have good arguments for deriving it from Timaeus, who is known to have been very fond of synchronisms and curious coincidences of time, such as play so large a part in the history of Agathocles (cf. especially Diod. xx. 43. 7). For an example of such a coincidence, specifically attributed to Timaeus, see the story of the statue of Apollo and Alexander's capture of Tyre in Diod. xiii. 108. 4–5.

^{page 7 note 1} Diod. xxiii. 15. 2–6. On Diodorus’ use of Philinus in his account of the First Punic War see Meltzer, , op. cit. ii. 557. His narrative in books xxiii and xxiv contains many details not in Polybius, e.g. Hanno's victories inGoogle ScholarLibya, at Hecatompylos, xxiv. 10. 1–2Google Scholar. In addition he also used a Roman annalist (cf. Meyer, E., Kleine Sckriften, ii (1924), 227, n. 4)Google Scholar, but this can scarcely have been the source of these reflec-tions on Regulus. See also p. 4, n. 1, above.

^{page 7 note 2} I am not convinced by Mϋunzer's argument (P-W, s.v. ‘Fundanius (5)’, col. 292) that the Fundanius story is probably ‘von der romischen aristokratischen Geschichtsschreibung in ten-denziṐser Weise ausgebeutet’, as an attack on a member of a plebeian house. The story seems to me to betray a pro-Barcine rather than an anti-plebeian source.

^{page 7 note 3} Cf. Polyb. viii. 3. 3: μία φυχή τς àπáσης στί πολυχειρίας ν νίοις καιροῖς νυστικωτ ρα; 7. 7: ούτως είς νήρ καί μία φυχή δεóντως ρμοσμνη πρòς Ӗνια τν πραγμτων μγα τι χρμα ϕαίνεται γίνεσθαι καί θανμáσιον (echoed by Livy, xxiv. 34. 1: ‘unus homo’). The doctrine is linked up, not very convincingly, by von Scala, , Studien des Polybios, i (1890), 94Google Scholar, with Heracleitus, Fr. d. Vorsokr⁵ (ed. Diels-Kranz), fg. 49: είς μοί μύριοι àν äριστος ἧ. It seems to be a general commonplace of the Hellenistic age, and finds an echo in Ennius’ famous line on Fabius: ‘unus homo nobis cunctando restituit rem’.

^{page 7 note 4} P-W, s.v. ‘Philinos (8)’, col. 2187; cf. Sisto, L., Atene e Roma, xii, 1931, 181Google Scholar; Davin, C., Beiträage zur Kritik der Quellen des I. Punischen Kriegs (Progr. Schwerin, 1889)Google Scholar, who discusses the relationship between Polybius’, Diodorus’, and Dio's accounts of the First Punic War (especially Polyb. i. 29. 3–36. 4), and concludes (p. 16): ‘die ethische und die pragmatische Beurteilung der Vorgange ist bei den drei Autoren die gleiche: die Uebereinstimmung aber lasst sich nur aus einer gemeinsamen Quelle erklaren, und diese kann… nur Philinos sein’. But Davin's method and purpose are different from those of the present essay, and it would be irrelevant to discuss his arguments in detail.

^{page 8 note 1} Cf. Thucyd. i. 22; ii. 48. 3; Isoc. ii. 35; vi. 59.

^{page 8 note 2} Diod. i. 1; this passage is derived from Ephorus by Lorenz, K., Untersuchungen zum Geschichtswerk des Polybius, Stuttgart, 1931, 10Google Scholar; Barber, G. L., The Historian Ephorus (1935), 103Google Scholar; and Ullman, , TAPhA, lxxiii, 1942, 30, n. 31Google Scholar; but the general tone of the passage is hard to recon-cile with a fourth-century source. It is almost certainly Stoic and post-Polybian.

^{page 8 note 1} Cicero, , de oral. ii. 9. 36Google Scholar; Dionys. i. 2.1; Sail, . Jug. 4. 1–2Google Scholar; Strabo, i. 1. 22 ff.; cf. Joseph, . A.J. i. 3; 4Google Scholar; Pliny, , epist. v. 8.11Google Scholar; Lucian, πς δεῖ ίστ. 39 fin., 42, and 43; Herodian, i. 1. 3. See Lorenz, , op. cit. 10Google Scholar.

^{page 8 note 4} Ullman, B. L., TAPhA, lxxiii, 1942, 25–Google Scholar: ‘History and Tragedy’.

^{page 8 note 5} Cf. JHS, lviii, 1938, 55–68. See Polyb. ii. 56; iii. 47. 6–48. 12; 58. 9; vii. 7. 1–2. 6; x. 2. 5–6; xii. 24. 5; 26 b 4 ff.; xv. 34. 1–36. 11 (probably criticizing Ptolemy of Megalopolis); xvi. 12. 7–9; 14.1; 17.9; 18. 2;xxix. 12.1–3.8.

^{page 8 note 6} Lorenz, , op. cit. 11–12Google Scholar. This feature cannot of course be divorced from the question of the historical method pursued by Polybius’ sources, and in particular Fabius; see below.

^{page 9 note 1} The phrase οῢτ ήγωνίσατ' γώνισμα has been generally misunderstood; e.g. ‘neque certamen ullum, quale nostra memoria, certaverat’(Schweighaeuser; cf. Scala, von, op. cit. 172)Google Scholar; ‘or act such a drama’ (Shuckburgh); ‘ever achieved such a triumph’ (Paton). None gets the exact force. An Amorgos inscription of c. 225–175 B.C. (IG xii. 7. 226, lines 4–6) speaks of a κωμφδóς (playwright? or actor? perhaps an actor-manager, owning a troupe of players?) and uses the phrase δρáματα γωνίςεσθαι, i.e. ‘to put on plays for competition’ at a local festival. The meaning is the same here, with the substitu-tion for δρȃμα of γώνισμα ‘a show-piece’ (cf.Thucyd. i. 22: γώνισμα ς τò; used of plays by Aristotle, , Poet. 1451 ^b37)Google Scholar. Polybius uses the same metaphor of Tyche as a play-producer elsewhere, e.g. xi. 5. 8: τς τύχης σπερ πίτηδες πί τήν ξώστραν ναβίβαςούσης τήν ῢμετραν ἂγνοιαν; cf. xxiii. 10. 16; xxix. 19. 2 (with σκήνην for ξώστραν). In xxiii. 10. 12 he (or his excerptor: cf. JHS, lviii, 1938, 64) has the expression τρίτον δ' τύχη δρȃμα… πεισήγαγεν. So too Diodorus, xxxii. 10. 5, in a passage the source of which is doubtful, but is certainly either Polybius or later: τς τύχης σπερ ν δρáμασι τò παρáδοξον τς περιπετείας γοης είς ἕγκλημα.

^{page 9 note 2} Sometimes it is men's own excessive conduct which brings its περιπτεια, with results the very opposite of what is contemplated; in such cases they may be said ‘to put themselves, or their folly, on the stage’, i.e. make an Aristotelian tragedy of their lives; cf. v. 15. 2: Apelles and the Macedonian counsellors, inspired by drink, attacked Aratus and so ξεθεáτρισαν αῢς. Their own downfall was the reverse of their intentions. In more general terms (xi. 8. 7), those who try to imitate τѿν εῢτυχούντων (Paton: ‘favoured by fortune’) in unessentials κθεατρί-ςουσι τν αυτѿν κρισίαν— they make a ‘tragic’ display of their own lack of judgement, to their own detriment (μετà βλáβης). In all such cases, Polybius would have said, Tyche was really fulfilling her function as stage-manager.

^{page 9 note 3} Cf. i. 3. 3–6; 4. I–II. For σωματοειδής as a quality of historical events see i. 3. 4; in xiv. 12. 5 it is applied to the description of the end of Ptolemy IV's reign in a single account οίονείσωματοειδ ‘as a unified whole’ (not, as Paton, ‘a life-like picture’). See, for the same idea, Cicero, , ad Jam. v. 12. 4Google Scholar: ‘a principio enim coniurationis usque ad reditum nostrum videtur nuhi modicum quoddam corpus confici posse.’ The idea of the corpus or σѿμα in literature seems to go back to the Platonic-Aristotelian conception of the unity of a literary work; cf. Plato, , Phaedrus, 264 CGoogle Scholar (every λóγος must be like aςѿον); Arist, . Poet. 1459 ^a17 ffGoogle Scholar. (who, however, specifically excepts history from its application!). See Lorenz, , op. cit. 87, n. 92; 99, n. 227Google Scholar. The novelty in Polybius is that, assisted by his conception of Tyche as ‘producer’, he projects the idea of the unity of an historical work on to the objective course of historical events.

^{page 9 note 4} Polyb. vi. 56.6 ff. Cf. Farrington, B., Science and Politics in the Ancient World (1939), 166–8Google Scholar; ‘Examiner’, Greece and Rome, xii, 1943, 59Google Scholar.

^{page 10 note 1} παρεισáγω is used elsewhere by Polybius in the sense of introducing a character or material into a narrative; cf. iii. 20. 3 on the sensational (θαυμáσιον) picture of the solemn sitting of the Senate drawn by Sosylus and Chaereas; iii. 47. 7 on the kind of Hannibal the ‘tragic’ histo-nans introduce into their works; v. 2. 6 on the sons of Aeacus introduced into a poem by Hesiod.

^{page 10 note 2} See above, p. 8, n. 2.

^{page 10 note 3} For the myths of Hades as an ingredient of tragedy cf. Arist, . Poet. 1456 ^a3Google Scholar; 1453 ^b7; cf. Ullman, , op. cit. 31Google Scholar, n. 33. For the function of such myths as an ingredient of ‘political’ religion cf. Arist, ., Metaph. xii. 8. 13. 1074 ^bGoogle Scholar. This idea seems to have been first expressed by Critias in his Sisyphus (cf. Diels, , Fragm. d. Vorsokr. (ed. Rranz, ), ii ⁵, p. 386Google Scholar, frg. 25). See further Farrington, B., op. cit. 87ffGoogle Scholar.

^{page 10 note 4} For instance, in several places (ii. 56. 10; xii. 25 b; xxxvi. 1. 7) Polybius stresses the duty of the historian to record what was actually said (τούς κατ' λήθειαν είϕημνους (λóγους)); but another very revealing passage (xii. 25 i, 4 ff.; on which see Wunderer, C., Polybios-Forschun-gen, ii (1901), 11)Google Scholar shows plainly that he drew no clear distinction between the actual words spoken and τοῢς ρμóςοντας καί καιριους (λóγους), which we shall record εί μλλομε μ βλáπτειν, λλ‘ ώϕελείν τούς ναγινώσκοντας.

^{page 10 note 5} Ullman, , op. cit. 30–1Google Scholar. Ephorus wrote Παρáδοξα and was clearly not hostile to the sensational in itself; but according to Strabo, vii. 3.9, he criticized writers who told only of the savagery of the Sauromatae είδóτες τò δεινóν τε καί τ θαυμαστν κπληκτικν ν δεῖν δ τναντια καί λÉγειν καί παραδειγματα ποιεῖσθαι. (παραδειγματα are ‘lessons’ as in Thucyd. iii. 40.) Ullman's case is weakened by his assumption that the preface to Diodorus comes from Ephorus (see above, p. 8, n. 2).

^{page 10 note 6} Cf. Ullman, , op. cit. 41, n. 81Google Scholar.

^{page 11 note 1} On this combination offoresight and courage, and its relation to the factor of Tyche, as de- veloped in Polybius, see Lorenz, , op. cit. 43 ffGoogle Scholar. (who, however, seems to me to overstress the Stoic influence in the notion of óρμή and λóγος his parallels from Thucydides for the τóλμα of the Athenians appear to invalidate this part of his argument); Taubler, E., Täche: historische Studien (1926), 89Google Scholar.

^{page 11 note 2} For Philinus as the source for Drepana and the Aegates Islands see De Sanctis, , op. cit. iii. 1. 228–9Google Scholar; Gelzer, , Hermes, lxviii, 1933, 141Google Scholar; Laqueur, P-W, s.v. ‘Philinos (8)’, cob. 2188–9, who all agree on this point. Reuss, , however, Philol. vlx, 1901, 137Google Scholar, makes Polybius' source for Drepana Fabius Pictor.

^{page 11 note 3} Polyb. i. 14 ff.

^{page 11 note 4} Polyb. i. 19. 15; cf. Diod. xxiii. 9. 1; Oros. iv. 6. 6, who both state that the population was enslaved. This statement is questioned by Beloch, , Griech. Gesch. iv. 1. 653, n. 1Google Scholar, but it was the Roman custom to enslave the population of conquered towns and there is no valid reason to reject the tradition.

^{page 12 note 1} Cf. vii. 7. 6; xxix. 12; and see the other passages quoted by Ullman, , op. cit. 42–3Google Scholar.

^{page 12 note 2} It may also be assumed that Philinus followed the traditional custom of writing and including speeches in his narrative; such may be the explanation of such passages in Polyb. i a s 27. 1 (where the observations of the Punic commanders repeat the arguments of 26.1), and 32. 8; 44. 1; 45. 3; 49. 10; 60. 5, where phrases like παρακαλÉσας τ καιρ τà πρÉποντα may represent a full-dress speech in an original source. However, the phrase occurs elsewhere in Polybius (e.g. ii. 64.1;iii. 71.85108.2:111.11; iv. 80. 15; v. 53. 6; 60. 3; viii. 13. 5; xi. 11. 2), and is perhaps merely Polybius’ substitute for the rhetorical commonplaces which a Timaeus would have introduced at such points (cf. La-Roche, P., Chardkteristik des Polybius (1857), 64)Google Scholar.

^{page 12 note 3} Dion, . Hal. i. 79 ff.Google Scholar; Plut, . Rom. 3 ffGoogle Scholar.

^{page 12 note 4} Plut, . Rom. 8Google Scholar.

^{page 12 note 5} Dion, . Hal. i. 84Google Scholar. also gives a version which he says is Fabian, and after criticizing it for being τν μνθωδεστρων and δραμαικς μεστν τος, he explains it away in euhemerizing terms.

^{page 12 note 6} On this particular crux see most recently Momigliano, A., JRS, xxxiii, 1943, 102Google Scholar, whom I believe to be right in accepting the traditional account. To the authorities he quotes add Rosenberg, P-W, s.v. ‘Romulus’, col. 1085, and De Sanctis, , op. cit. i. 214 ffGoogle Scholar.

^{page 12 note 7} Op. cit. i. 214 ff.

^{page 12 note 8} Cf. Last, H., CAH, vii. 366Google Scholar; Täubler, E., op. cit. 70, 71Google Scholar; the influence of Sophocles' Tyro is rejected by Cary, M., History of Rome, 40Google Scholar.

^{page 12 note 9} Tarn, W. W., JHS, xxvii, 1907, 51, n. 19Google Scholar. Another example may be the stress on the womanly character of Teuta in Polyb. ii. 4.8; 8.12.

^{page 13 note 1} On the corvus (κραξ), alleged to have used at Mylae (Polyb. i. 22) see Tarn, , Hellenisiic Military and Naval Developments, 111–12, 149–50Google Scholar. It is generally agreed that the Carthaginian defeat recorded by Polyb. i. 9–11 is in fact Philinus' version of the battle Mylae, which Polybius failed to recognize because the corvus was omitted (cf. Beloch, , Griech. Gesch. iv. 1. 654, n. 1Google Scholar; Lenschau, P-W, s.v ‘Hannibal (3)’, cols. 2321–2; Tarn, , JHS, xxvii, 1907, 51, n. 19Google Scholar; De Sanctis, , op. cit. 1. 128–9, n. 73; 226)Google Scholar.

^{page 13 note 2} Cf. Pasquali, , Enc. Ital. (1936), s.v. ‘Roma’, 909Google Scholar. ‘La storia di Fabbio Pittore … era, almeno nelle intenzioni, arte, quale la teoria ellenistica esigeva che fosse la storiografia.’ This is true, but it is only half the story; much of Fabius must have reflected his exiguous Roman sources. See below, Addendum, pp. 15 ff.

^{page 13 note 3} Hermes, lxviii, 1933, 140: ‘Die hier vorgebrachten Einzeltatsachen wird er nicht erfunden haben.’ The implication seems to be that had them from Fabius.

^{page 13 note 4} Meltzer, , op. cit. ii. 308Google Scholar; De Sanctis, , op. cit. iii. 1. 158Google Scholar.

^{page 13 note 5} So Laqueur, P-W, s.v. ‘Philinos (8)’, col. 2187; the argument that the reference to Rome's invariable success on land was contradicted by Hannibalic War is less cogent, since if the reflections were those of Polybius himself, it might well be argued that the victories in the east had dimmed the memory of the years 218–216 B.C.

^{page 13 note 6} On this question of terminology see Lorenz, , op. cit. 45Google Scholar, who, however, again claims Stoic influence (see above, p. 11, n. 1).

^{page 13 note 7} The accusation recalls Xerxes flogging the Hellespont to avenge the destruction of his bridge: Hdt. vii. 35–6; Aesch, . Pers. 745 ffGoogle Scholar.

^{page 13 note 8} It may be argued that the didactic attitude towards Rome is inconsistent with Philinus' anti-Roman policy; but lecturing one's enemy has always been a popular enough device of rhetoric. In any case, if the reflections are Polybius' own, it is hard to explain their irrelevance; other Polybian criticisms of Roman conduct are concerned with specific incidents (e.g. i. 83. 11; 88. 8 ff.; iii. 28. 1 (seizure of Sardinia); iii. 26.6 (help given to Mamertines)), and not as here with Roman character.

^{page 14 note 1} Op. cit., col. 2192.

^{page 14 note 2} On the various meanings of αἲρεσις in Polybius see Strachan-Davidson, J. L., Selections from Polybius, 1888, 7Google Scholar.

^{page 14 note 3} I keep Laqueur's terminology, though in fact ‘aristocracy’ is an unsuitable word to describe the ruling class in the Greek states of the third and second centuries, with its stress on wealth rather than blood. The real situation comes out in the Achaean formula πλοετνδα κα ριστνσα (IG, vii. 188, line 8), where πλοετνδα expresses the reality, ριστνσα, the propaganda phrase in which the reality is muffled. See further Aymard, A., Assemblées de la confédération achaïenne, 56, n. 4Google Scholar; 137, n. 4; 335 ff.

^{page 14 note 4} Successive editions are the only way of reconciling the anti-Roman tendency on which Polybius comments with the objective, didactic work which Laqueur's theory postulates (and for which there is no evidence).

^{page 14 note 5} One should note the change in Laqueur's picture of Polybius' development. In his book Polybios (1913), 261–77, he banished the view of Tyche as subject to law, and therefore a proper study for the historian, the concept of world-history, Stoicism, and the didactic-utilitarian approach, to the fifth edition of the Histories; whereas the influence of Philinus came in in editions 2 and 3. Now that Philinus has been shown (rightly, I believe) to have been a didactic-utilitarian writer, Laqueur's famous scheme seems ripe for revision. How its author intends to resolve this contradiction is not yet apparent; but it will be difficult to find a way out which does not offer violence to the original system.

^{page 15 note 1} Susemihl, , Geschickte der griech. Litterat. in der Alexandrinerzeit, i. 634Google Scholar.

^{page 15 note 2} See above, p. 1, n. 2. Gelzer is followed in essentials and amplified by Altheim, F., Epochen der romischen Geschichte, ii (1935), 305 ffGoogle Scholar. For the usual view of the development of annalistic writing at Rome, which Gelzer attacked, see for example, the discussion of Cichorius, P-W, s.v. ‘Annales’, cols. 2255–6; Pais, E., Ricerche sulla storia e sul diritto pubblico di Roma, iv (1921), 177–224Google Scholar.

^{page 15 note 3} Vogt, J., Gnomon, xii, 1936, 524–6Google Scholar, in a review of Altheim's, Epochen der römischen Geschichte, iiGoogle Scholar; see, too, De Sanctis, , Riv. Fil. lxi, 1933, 548Google Scholar.

^{page 15 note 4} F. 1 and 2 (Peter) = Aul. Gell. v. 18. 8.

^{page 15 note 5} Cf. Kornemann, , Klio, xi, 1911, 256Google Scholar; Peter ad F. 2 (HRR, i², p. 179), comparing Polyb. iii. 20. 5.

^{page 15 note 6} The reference to the Greek translation of diarium shows that Asellio was not limiting himself to Roman theory or practice.

^{page 16 note 1} F. 3: in Fabi Picloris Graecis annalibus (Cic. de div. i. 21. 43); F. 24 (Pliny, , n.h. x. 71)Google Scholar; F. 27 (Pliny, , n.h. xiv. 89)Google Scholar; Cic. de orat. ii. 12. 52 (annalium confeclio).

^{page 17 note 2} The Latin version (cf. Peter, , HRR, i ², pp. 112–13)Google Scholar was probably a translation, not the work of the later jurist; cf. Schanz-Hosius, , Gesch. der röm. Lit. i (1927), 172Google Scholar; Beloch, , Romische Gesch. 98Google Scholar, against Mfinzer, P-W, s.v. ‘Fabius (128)’, col. 1843.

^{17 note 3} In his commentary to i. 2. 8 (vol. v (1792), pp. 125–30): cf. especially: ‘Ubi Postumium Albinum ait graece πρ ἱστ. scripsisse, non cogitavit de peculiar! quadam ratione, qua scripta erat ilk historia; nihil quidquam amplius significat nisi quod Cicero in Bruto (21. 81 = T. 2) ait, Albinus qui Graece scripsit historiam, ait quod Gellius (xi. 8. 2 = F. 1) Albinus res Romanas oralione graeca scriptitavit aut Plutarchus in Catone mai. (12 = T. 4) ‘Aλβῖνος ἱστοραν λληνισρ γρψας.’ Schweighaeuser's analysis of the meaning of πραγματικ ἱστορα will be found repeated in Strachan-Davidson, , Selections from Polybius, 3–5Google Scholar.

^{page 16 note 4} See Schweighaeuser, loc. tit., and Strachan-Davidson, , op. cit. 5–6Google Scholar.

^{page 16 note 5} There is the same distinction in Plut, . Galba, 2 3Google Scholar: τ μν οὐν καθ' ἒκαστα τν γενοννων παγγλλειν κριβς τς πραγματκς ἱστορας σιν (as opposed to biography). It is noteworthy that the same meaning is expressed elsewhere {Alex. 1), by the word ἱστορας γρφομεν, λλ βονς.

^{page 16 note 6} That Postumius' history, like Fabius’, went back to the origines of Rome (cf. F. 3 (Peter), which will scarcely be from a separate work: HRR, pp. cxxv–cxxvi), and so will have ranked technically among the accounts of the κτσεις of town (as Gelzer and Altheim admit: cf. too Cato's title, Origines), is another reason for not pressing the meaning of πραγματικ, which excludes this branch of history.

^{page 16 note 7} Macrobius, iii. 20. 5, referring to Postumiu' work, speaks of annali primo (F. 2 Peter).

^{page 17 note 1} Gell. v. 18. 7 contrasts res gestae which per annos scribuntur (annales) with those written per dies (diarium or φημες). Asellio of course distinguishes annales, written after the manner those, qui diarium scribunt, from res gestas Romanis, which show quo consilio quaque ratione gesta essent.

^{page 17 note 1} Verrius Flaccus’ view won a certain following; it is also recorded by Servius, , ad Aen. i. 373Google Scholar: ‘inter historiam et annales hoc interest: historiaest eorum temporum quae vel vidimus vel videre potuimus, dicta π το ἱστορεῖν i. q. videre. annales vero sunt eorum temporum, quae aetas nostra non novit.’

^{page 17 note 3} F. 77 = Gell. ii. 28. 6.

^{page 17 note 4} Cf. Momigliano, , JRS, xxxiii, 1943, 102Google Scholar.

^{page 17 note 5} Scullard, , History of the Roman World from 753 to 146 B.C., 417Google Scholar.

^{page 17 note 6} See above, p. 2, n. 3.

^{page 17 note 7} Römische Forschungen, ii. 363, n. 113; cf. Schanz-Hosius, , op. cit. i. 171Google Scholar.

^{page 17 note 8} Cf. Polyb. xiv. 12. 1 ff., giving the reasons why Polybius made the reign of Ptolemy Philopator an exception to his normal practice writing κατ' νιατν: xxxii. 11. 3 ff. His method was thus a practical compromise between the Thucydidean διαρεσις κατ' νιατς (in fact κατ θρη κα χειμνας) and the arrangement κατ γνος, which Ephorus adopted in polemical opposition to his predecessor. On this see Bloch, H., Athenian Studies presented to W. S. Ferguson (1940), 308–16Google Scholar.

^{page 17 note 9} See too Zimmerman, , Klio, xxvi, 1933, 257 ff.Google Scholar; Kornemann, , Hist. Zeit. clxv, 1932, 287Google Scholar.

^{page 18 note 1} Cf. Vahlen, , Enn.² praef. clxxivGoogle Scholar. Ennius devoted the first three of his eighteen books to Androthe kings, and by the sixth book had reached Pyrrhus.

^{page 18 note 2} Cf. Peter, , HRR, lxxivGoogle Scholar. See further F. Jacoby, P-W, s.v. ‘Hellanikos (7)’, cols. 138–42; Pearson, L., The Local Historians of Attica (1942), 8Google Scholar; Jacoby, P-W, s.v. ‘Ktesias (1)’, cols. 2040ff.; R. Laqueur, P-W, s.v. ‘’, especially cols. 1092 ff., who stresses the emphasis given to epic and myth in the Atthides. Androthetion and Philochorus (whom Peter mentions) are in this respect hardly typical; they pay much less attention to the earlier period, especially Androtion. See also Bloch, H., op. cit. 344–6Google Scholar.

^{page 18 note 3} Cf. Beloch, , Römiscke Geschichte, 96–8Google Scholar.