¹ S. pp. 448 f.

² K. pp. 28 f.

³ C. 1, p. 39.

⁴ B. 2, pp. 28 f.

⁵ Conveniently tabulated in C. 1, p. 106.

⁶ The historical library of Diodorus the Sicilian, made English by Booth, G., LondonGoogle Scholar.

⁷ From the context it is clear that this entry is derived from the chronological table used by D.; if we knew how the three histories were dated, we could press the synchronisation as very strong evidence for the validity of D.'s date. As cross-check on D. cf. xiv. 117, 8.

⁸ With this phrase D. introduces his resumption of the events anterior to the declaration of the Sacred War.

⁹ At this point D. reaches in his narrative account the seizure of the temple by Philomelus, an event which he dated in 14, 3 to 357/6; he now proceeds to supply the narrative anterior to his fixed date of 23, 1 (355/4), the declaration of the Sacred War.

¹⁰ οἱ πλησίον οἰκοῦντες, presumably the Locrians of Amphissa.

¹¹ Booth's translation, while stressing the informal nature of the reports, loses the emphasis of αὐ τός; I shall refer to these reports as ‘personal propaganda.’ The phrase recurs in xvi. 89, 2 and in the passive in xvi. 22, 2.

¹² This phrase may explain why the Boeotian forces do not figure in the ensuing operations; they must have found it impossible to storm Delphi, and have then returned to Boeotia.

¹³ It is clear from the action of Philomelus, and the ensuing cessation of hostilities covering the exchange of embassies, that we have reached the end of the campaigning season which followed the seizure of the temple; as we shall see later, the temple was seized in June, and D. gives the year as 357/6;, i.e. by autumn 356 B.C. Philomelus was in control of the Delphi region, and despatched envoys in the winter 356/5.

¹⁴ The resumptive Greek imperfect and the contemporary participle are well turned by Booth's translation.

¹⁵ A condition envisaging the future, i.e. implying that no state was yet at war with the Phocian state. This is, of course, consistent in D.'s narrative with the Boeotian decision (25, 1) to liberate the shrine; for the meaning conveyed by D. is that in reply to Philomelus' personal propaganda Boeotia confined her attentions to the adventurer Philomelus, and had not passed a declaration of war upon the Phocian state. During the winter Philomelus, having secured the approval of the Phocian commons, published the complicity of the Phocian state in the official embassies; and, as he foresaw, war was, in fact, soon declared upon the Phocian state.

¹⁶ It is, of course, clear that this is not a joint declaration of a Sacred War by an Amphictyony, i.e. we have not yet reached the end of the narrative which resumes events anterior to the declaration proper, fixed chronologically by D. at 23, 1 to 355/4. The Thessalians are a significant omission from those powers which took an active part: they are too important to figure under καί τινες ἕτεροι (cf. 28, 4 and 33, 3), and so must be neutral at this stage.

¹⁷ The joint action of the Amphictyons provides the declaration of the Sacred War, announced by D. at 23, 1 and dated there to 355/4; at this point, then, he has reached the end of the resumptive narrative bridging the interval between the seizure of Delphi and the declaration of war. The chronology then fits clearly to the narrative; for after the embassies of Phocis sent out in the winter 356/5, the Locrian operations fall in early 355 B.C. and the Amphictyonic meeting in autumn 355 B.C., i.e. the Attic year 355/4.

¹⁸ The summary of Amphictyonic powers is, of course, separate from the list of powers who answered the Phocian embassies sent out in the winter; the reason for Spartan action is here given more fully than in 23, 2 and 24, 2.

¹⁹ ‘Desperate’ because sacrilege was for the first time committed when temple funds were impounded.

²⁰ It is significant that the Thessalian forces disappear for the remaining operations.

²¹ From the ensuing narrative it is clear that this engagement fell in the autumn at the end of a campaigning season, i.e. late in 354 B.C.; it is elsewhere dignified with the title ‘Battle of Neon’ (Pausanias x. ii. 4).

²² The discussion of the chronology, as opposed to the narrative, is reserved for section C; I have included some chronological notes in the present section to clarify the narrative account. A full narrative employing independent evidence will be found in the Summary.

²³ PW. gives four instances: ix. 2, 4 = 34 and fr. 4 = 20,4 (each within one section in a fragmentary book), xii. 38, 2 = 40, 2 (single section, probably resumptive), and xiv. 89,2 = 92,3 (limited to four words!). Hampl, , Der König der Makedonen, p. 45Google Scholar, finds a ‘Dublette’ in xiv. 92, 3 = xv. 19, 2 (single section, certainly resumptive).

²⁴ Cf. PW. col. 682 ‘Nun liegt aber die Fuge nicht an einem historisch bedeutsamen Punkt sondern mitten in der Erzählung.’

²⁵ Cf. C. 1, pp. 4–5, e.g. ‘aucun sectionnement du récit n'est done possible si l'on se borne à examiner la couleur et les tendances.’ The hypothesis, that D. was using Ephorus and Demophilus as his sources and that Ephorus ended at just this point, would explain this peculiarity, but most scholars—e.g. K. p. 37—regard the hypothesis as untenable. Beloch (2, p. 28) argues that D. copied the same account twice, but in disposing of one difficulty he seems tome to raise a host of others (even K, p. 30, who does not overestimate Diodorus' intelligence, refuses to accept B's suggestion). Nor do I find Momigliano's attempt to differentiate between the two accounts at all convincing (Rend. Istitut. Lombardo lxv. 1932 pp. 535 f.Google Scholar).

²⁶ The graphic narrative imperfect is rare in Diodorus.

²⁷ B. 2, p. 29.

²⁸ Beloch 2, pp. 29 f., uses this argument as a main reason for supporting the doublet theory.

²⁹ Only in 28, 2 does he assume the offensive; the foray of 25, 3 is limited to the region round Delphi, and could reasonably be represented as defensive.

³⁰ Phocis, having identified herself with Philomelus, could legitimately protest against the action of Thebes in time of peace (25, 1 and 27, 4; in 27, 5 Boeotia declares war on Phocis).

³¹ Untersuchungen über d. Quellen d. gr. u. sik. Geschichte bei Diodor Buch xi–xvi (Kiel, 1868), p. 110Google Scholar.

³² Paus. x. 2, 3 uses the term .

³³ Beloch 2, p. 31, while admitting that ‘beide ergänzen einander,’ under-estimates the interdependence of the two by calling the second ‘subsidiar.’ A passage, which lies outside the supposed doublet narrative, ought to be mentioned; Diodorus, summarising the attitude of the Phocian captains to the temple, states (xvi. 56, 5) , while in his narrative of the early war he describes Philomelus as respecting the temple funds at 24, 5 (a promise), 27.4 (offering examination), and 28, 2 (), but as compelled finally to spoil the treasury at 30, 1. Now, as we have seen, there is no discrepancy between 30, 1 and the three earlier passages: the only discrepancy is between 30, 1 and 56, 5. This discrepancy can be explained in three ways: (1) Diodorus has made a mistake in his short summary at 56, 5 through remembering only the first three instances of Philomelus' attitude to Delphi, (2) Diodorus used a different source at 56.5 from the source he was using at 24–30, and (3) Diodorus used one source at 24, 5, 27, 4, 28, 2, and 56, 5, and another and different source at 30, 1. Of these three explanations only (3) would support belief in a doublet within the narrative 23–31, and that doublet would be comprised by 24, 5 to 28, 2 as one alternative and by 30, 1 f. as the second, with a coupure presumably at the end of ch. 29; such a doublet seems to me untenable, as 30–31 comprises the campaign ending in Neon and is therefore integral to the whole; even if assumed, it can only limit itself to 30, 1–3 ending before εὐθὺς οὖν κ.τ.λ, and therefore falls into the minor doublet school of thought. As regards the doublet theory advanced by Cloché, explanation (3) provides no support; for in Cloché's analysis both doublet narratives contain the statement that Philomelus respected the shrine and the single narrative resumes at ch. 30. Kahrstedt's analysis, however, does meet a variation of explanation (3), for, by taking 23, 1–25, 3 as one alternative and 27, 1–30, 4 as the other, he provides in his first narrative a respect for the temple reiterated in 56, 5, and in his second alternative a notice both of respect and of sacrilege; this doublet theory, however, I have considered impossible on other grounds, and must leave the argument to Cloché. Moreover, explanation (3) is not essential; both (1) and (2) are possible, and I personally incline to (1), adducing as a parallel instance of Diodorus' slovenliness the statement that ‘Onomarchus kindled the Sacred War’ (38, 6, cf. 23, 1).

³⁴ S. I², pp. 495–498; P. pp. 20–25; Homolle, , BCH xxii. pp. 624–627Google Scholar.

³⁵ K. pp. 27–28; B. 2, pp. 265 f.; C.1, p. 106.

³⁶ As is well known, D. is apt, in synchronising the incidence of the Roman year, to leave the Attic year rather ragged at the edges: this, however, does not disturb our reconstruction.

³⁷ A chronological table is appended to this paper.

³⁸ B. 2, p. 263, assuming confusion of Kephisodorus with Kephisodotus 358/7, argues that an error of one year is possible, i.e., he uses it to support 357/6; cf. J. II d p. 697 for discussion.

³⁹ If further argument is desired, consult the works of the Kahrstedt–Beloch–Cloché group, esp. C. 1, pp. 40 f.

⁴⁰ Cf. infra, pp. 58 f.

⁴¹ Cf. infra, pp. 65 f.

⁴² Uhlemann, , Untersuch. über die Quellen der Geschichte Philipps von Makedonien u. d. heiligen Krieges in XVI Buche Diod., diss. Strassburg (1913), p. 69Google Scholar detects a doublet in ch. 35; I must content myself with quoting Momigliano, , Filippo il Macedone, p. 104Google Scholar, n. 1 ‘la presunta reduplicazione… é insussistente.’

⁴³ Dionysius, , Demosth. 13, p. 655Google Scholar.

⁴⁴ Olynth. iii. 4 , i.e. either 352 November or 351 November; cf. Pickard–Cambridge, , Public Orations of Demosthenes, Vol. II. p. 161Google Scholar, in favour of the former year; our argument shows that 351 is not possible.

⁴⁵ Phil. i. 40–41, cf. 17.

⁴⁶ Olynth. i. 13 .

⁴⁷ P.–C. pp. 177 f.; CAH p. 220.

⁴⁸ B. 2, pp. 280–281; he does not mention the two passages I have quoted from Demosthenes, but places Thermopylae in 353 B.C. and Heraeum Teichos in Nov. 351 B.C.

⁴⁹ The rival school supplies further arguments based upon this department of the evidence, esp. P., pp. 24 f.

⁵⁰ Phil. i. 35Google Scholar; Olynth. i. 9Google Scholar.

⁵¹ IG² II. i. 130, first applied in this context by K., p. 42.

⁵² Dem, . Phil. i. 31Google Scholar shows that Philip had been helped by these winds before the famous instance at Olynthus.

⁵³ In section E I hope to prove that D. cannot have made a mistake and that his date-table must have given 354/3. Kahrstedt's suggestion (K. p. 42), that D. here reckons ‘römischer Rechnung,’ is rightly scorned by Beloch (B. 2, p. 269), whose own suggestion ‘die Angabe nach dem wichtigsten Ereignisse, der Einnahme von Pagasae, orientiert ist’ deserves even less respect.

⁵⁴ C. 1, pp. 69–75, with further arguments and bibliography.

₅₅ Dem., C. Aristoc. 183Google Scholar: Polyaenus iv. 2, 22.

⁵⁶ Cf. section E on this passage.

⁵⁷ C. 1, pp. 77 f. and table, p. 106; D. xvi. 34, 2 is a eulogistic phrase which cannot imply that the date of Pammenes' departure is to be defined exactly by the genitive absolutes; this implication is made by Cloché, but he does not quote the last eight words of the sentence, which make nonsense of his interpretation.

⁵⁸ Cf. B. 1, p. 254.

⁵⁹ Kahrstedt (p. 49), who places the mission of Pammenes in late autumn 354 B.C. and one year after Neon, occupies the same position, though he somewhat naïvely finds Schaefer's explanation ‘köstlich’ (p. 49, n. 100).

⁶⁰ B. 2, p. 269.

⁶¹ Polyaenus iv. 2,22. Cf. P.–C, p. 169, n. 5, whom I follow against B. 2, p. 283.

⁶² B. 2, p. 269, cf. p. 261, simply discards Dionysius as he does Diodorus: his chief argument, that because Chares left Asia in autumn 355 Pammenes took his place in spring 354, seems a remarkable non sequitur. Judeich, , Kleinasiatische Studien, p. 211Google Scholar, dates Pammenes mission to spring 353 B.C. from evidence of Persian chronology, which is, however, itself inconclusive.

⁶³ ii. 131; iii. 148.

⁶⁴ Fr. 1 J. II.b., p. 639.

⁶⁵ Fr. 1 J. II.a, p. 138.

⁶⁶ ix. 6, 1; x. 2, 2; 3, 1; 8, 2.

⁶⁷ x. 3, 1.

⁶⁸ Summarised in C. 1, pp. 40–52; I discuss elsewhere (CQ xxxi. 2, 1937) the references of Diodorus to the duration and nomenclature of the war, for they have no direct bearing on the chronological problem. My view is that Demophilus called his history after the Phocian adventurers, i.e. his war lasted eleven years, from 357/6 seizure of Delphi to 347/6 capitulation of Phalaecus; both Diodorus and Pausanias confused the two traditions of Demophilus and Callisthenes.

⁶⁹ The following dates are confirmed by independent evidence: 360/59 accession of Philip (B. 2, p. 60); 358/7 stasis in Euboea (IG ² II. i. 124 and Foucart, , RA xxxv. pp. 227 f.Google Scholar); 356/5 defeat of the Cetriporis coalition (H. and H., 131; K, p. 40); 356/5 Embata (IG II. ii. 794, col. C, 1. 90, and Kohler, , AM VI. p. 30Google Scholar); 353/2 cleruchies sent to Chersonese (IG II. ii. 795, f. 134); 349/8 Philip attacks Chalcidice (terminus post quem in Dem. Olynthiacs); 348/7 fall of Olynthus (B. 2, p. 279); 346/5 Amphictyonic sentence on Phocis (Naopoioi lists, etc.); 341/0 Phocion ousts Cleitarchus (IG ² II. i. 230); 341/0 siege of Perinthus begun (B. 2, p. 294); 340/39 Athens declares war (Philoch. fr. 135); 338/7 Chaeronea (B. 2, pp. 288 f.); 337/6 declaration of war against Persia (Oxyrh. Pap. I. 26 f.Google Scholar); 336/5 death of Philip (IG ² II. i. 240 and B. 2, p. 60).

Against these may be set two passages in D., which give an incorrect dating: 7, 3 to 8, 7 outbreak of the Social War and Philip's progress until the foundation of Philippi ascribed to 358/7 (Dionys. Lys. 12, p. 480Google Scholar, and B. 2, pp. 258 f.), and 14, 1 murder of Alexander of Pherae ascribed to 357/6 (B. 2, pp. 83–84). But the error in each case seems to lie with D., and not with the date-table. For in 7, 2 it is probable that the phrase ἅμα δὲ τούτοις marks a date-table citation, while 7, 3–4 derives from a narrative source; similarly, 8, 1 limits the chronology by the phrase to Philip's return from Illyria, while the two connections μετὰ δέ (8, 2 and 8, 6), together with the last sentence of 8, 7, suggest that D. has carried his narrative beyond the confines set to this year by his date-table; finally, in the case of Alexander it is possible that D. means to date the intervention of Philip, and not the death of Alexander to 357/6, since he has already (at xv. 61, 2) foreshadowed the death of Alexander as falling in 358/7.

⁷⁰ BCH xxii. pp. 608, 625Google Scholar.

⁷¹ BCH xl. pp. 78 f.Google Scholar, e.g. p. 102.

⁷² B. 2, p. 264.

⁷³ The Greek is Ἡρακλείδου and Ἡρακλείου; but I do not subscribe to Beloch's emendation; it is more probable that Pausanias or his source misquoted the name. Prytanis is also an error for archon.

⁷⁴ Fouilles de Delphes, III. 5, pp. 8 f. (1932)Google Scholar.

⁷⁵ B. 2, pp. 265–266, defines the war not by declaration of war but by the beginning of actual hostilities; this seems to me special pleading.

⁷⁶ Bourguet here overstrains his point; the entry marks not the declaration of war, but the first mention of the war in the history of the meetings. It proves that, if war was declare data regular meeting (as B. 2, p. 266, n.1, argues), then it was declared either in the years when there was no meeting or in spring 353 B.C.

⁷⁷ Bourguet, op. cit. p. 11, n. 8.

⁷⁸ Cf. Cloché, BCH xl. pp. 82–116Google Scholar for the historical significance of the lists in general; Bourguet, p. 10, accepts Cloche's conclusion that the states there represented are ‘de bon gré ou par force’ allied with Phocis.

⁷⁹ Flathe, Reiske, Vogel, Voemel, and Weiske consider Κορινθίους in this passage to be corrupt; the prominence of Corinth in the Naopoioi lists proves their suspicions to be superfluous.

⁸⁰ Incidentally the appearance of Locris and Athens in the list suggests that IG ² II. i. 148, recording an alliance between Locris and Athens, should be dated to 353 B.C., and is evidence not, as P.–C, p. 174, maintains, of a rapprochement between Athens and Thebes, but between Athens and Phocis.

⁸¹ As the reference falls in a passage of bunched narrative in Diodorus, the precise date is uncertain.

⁸² The ultimate source of this passage in Diodorus and also of ch. 29, giving the sides taken by Amphictyonic States at the declaration of war, was probably a quotation of official documents at Delphi.

⁸³ Cf. B. 2, p. 267.

⁸⁴ This overlap in the narrative of Diodorus is understandable, when we reflect that he had compressed under the years 355/4 and 354/3 a narrative of events covering four Attic years (357/3).

⁸⁵ B. 2, pp. 268 f.

⁸⁶ This correction of the MSS. παγάς, accepted by Grote, , History of Greece, vol. xi. p. 261Google Scholar n., and by S. 1², p. 509, n. 2, is certainly right; Momigliano, , Filippo il Macedone, p. 104, n. 3Google Scholar, accepts the correction, but his reason for holding the reference to Pagasae worthless (‘perchè la cronologia è sbagliata e provocata nel cronografo seguito da Diodoro dall' attrazione del nome di Metone’) seems inadequate.

⁸⁷ Cloché (C. 1, p. 74), who concludes from this passage that ‘Pagases n'a done pas succombé avant Pherés,’ not only mistranslates εὐτρεπίσας as ‘occupant,’ but also endeavours to press the sequence of names to a chronological meaning, which is not justified by the third passage (cf. also B. 2, p. 268 n.), and is in itself highly improbable, for why should Pherae capitulate before her port Pagasae and her hinterland Magnesia were captured?

⁸⁸ E.g. Theopompus bk. ix. (J. 115, F 78, 81, 82), mentioning Tempe, Pharcadon, and Philip's agent in Perrhaebia, should probably be dated to 354 B.C., since bk. viii., with its reference to the members of the Amphictyony (F 63), reaches autumn 355 B.C., bk. x. refers to Eubulus ( F 99, 100) à propos of the Theoric Fund established 354 B.C., and bk. xi. (F 101) contains a reference to Amadocus dating to after 353 B.C. summer. But restoration of Theopompus' chronology is too conjectural for our purpose.

⁸⁹ Dem. xix. 86: Diophantus was Theoric Commissioner 350–346.

⁹⁰ E.g. Dem. xix. 319 omits to mention the presence of Spartan and Achaean troops at Thermopylae; most scholars infer from Diodorus that they were present.

⁹¹ Despite B. 1, p. 228, n. 1, I consider D.'s date 357/6 more probable than the winter 358/7. The death of Alexander in the latter half of 358 B.C. (D. xv. 61, 2: cf. B. 2, p. 83) was followed by an interval during which Teisiphonus posed as a τυραννοκτόνος, was the proposed recipient of Isocrates' advice, and made an alliance with Athens (Isoc., Ep. vi. 3Google Scholar), whose policy was doubtless to extend her 361/0 alliance with the Thessalian League to cover also the liberated Pherae; but by spring 357 B.C. Teisiphonus had assumed the tyranny, renewed the Theban alliance of Alexander, and acted in Euboea against Athens. The inference, then, is that Athens, and not Philip, was the ally of the Thessalian League in 358/7; only when Athens was engaged in the Social War did the Thessalian League desert Athens and invoke Philip, the ally of Amphipolis and of the Chalcidian League, i.e. in the winter of 357/6 as recorded by D. xvi. 14, 2. To this interpretation the mention of Amphipolis' fall closely followed by a fragment dealing with Thessaly lends some support (Theop. iii. F 42 and 48).

⁹² Xen., Hellen. vi. 4, 37Google Scholar; Schol. Aristeid., p. 298 (Dind.)

⁹³ B. 1, p. 250, argues that they were present; his contention, that otherwise the Boeotian coalition could not field 13,000 men, rests upon the assumption that the 6000 Boeotian hoplites present at Leuctra provide a reliable basis for estimating the Boeotian forces of 355 B.C. But such a figure does not explain the ability of Thebes to send 5000 men to Asia in 354/3 and 4500 men to the Peloponnese in 352/1; we should rather estimate the Boeotian forces at some 9000 as a minimum, and add thereto her Locrian allies. If the aggregate 13,000 contains any Thessalians, it can only be a portion of the Thessalian 6000, and a portion implies a split in the Thessalian ranks.

⁹⁴ Diod. xvi. 31,6; Theop. F. 78 and 81.

⁹⁵ xvi. 35, 5 τυχικῶς παραπλέοντος; whether the mistake arises from D.'s lack of tactical insight or from Athenian attempts to clear themselves of complicity with Phocis, must be a matter of conjecture.

⁹⁶ Theop. J. 115, F. 249.

⁹⁷ B. 2, pp. 270 f.

⁹⁸ Ibid. p. 271 ‘Dionysios’ chronologische Ansätze sind bekanntlich keineswegs unbedingt zuverlässig’!

⁹⁹ The quotation from the chronographic source for the year narrated 32–36 appears clearly in 36.

¹⁰⁰ Pausanias' (viii. 27, 9–10) narrative of the Megalopolis–Sparta war is too short to be of value for either theory.

¹⁰¹ Dem. xxiii. 183.

¹⁰² B. 2, p. 281.

¹⁰³ Mélanges Glotz (1932) I. pp. 215–226Google Scholar.

¹⁰⁴ Kahle, F., De Demosthenu orationum Androtioneae Timocrateae Aristocrateae temporibus (Diss. Göttingen, 1909), pp. 10–11Google Scholar.

¹⁰⁵ Demosthenes, in the interest of his client, naturally interprets the whole as a treacherous plot against Athens, aided and abetted by Aristocrates, but it would be highly uncritical to accept Demosthenes' tendencious and far-fetched account.

¹⁰⁶ We know little of Philip in 355 B.C. Neapolis appealed to Athens for help in 355 B.C. (IG II. i. 66), and by 354 B.C. he had reached Maroneia. It is probable that he advanced over the are a between the two in 355 B.C., for in 354 B.C. he was busy besieging Methone and Pagasae.

¹⁰⁷ Cloche's arguments against 353 B.C. (Mélange Glotz I, p. 223Google Scholar) depend upon his doublist chronology of the Sacred War, and the false deduction that, because Sestos was captured in 353/2 and Chares reported the meeting at Maroneia, there must have been an Athenian fleet in the Hellespont during the winter months, January–March 353 B.C. His own arguments (p. 218) for 356 B.C. strain the Greek to too precise a translation, and his contention that the Athenians had no fleet in the Hellespont in 356 B.C., the very height of the Social War with Athens cut off from the naval base at Samos, is untenable. Pickard–Cambridge, , CAH p. 218Google Scholar, places the capture of Sestos in 352 B.C.; I see no reason for changing the date from the position it holds in Diodorus' narrative, quite apart from our reconstruction. Theop. F 101 in bk. xi. refers to Amadocus joining Philip, but the chronology of bk. xi. cannot be fixed precisely.

¹⁰⁸ B. 2, p. 259.

¹⁰⁹ H. & H. 130.

¹¹⁰ IG II. ii. 794 C. 90.

¹¹¹ D. xvi. 21, 4 Chares did not return to Athens after Embata and needed Persian pay to support his full forces. Ibid. 22, 1.

¹¹² D. xvi. 21, 2, cf. 7, 3, and Dionysius, Lys. 12, p. 480Google Scholar, both end the war in 356/5: the date mid-summer is more probable than earlier in 355 B.C., as it accounts for one of Diodorus' remarks about the duration of the war, for the confusion could arise that the war lasted three years if peace was signed midsummer: the negotiations with Persia push the victory of Chares back to late winter or early spring on either date.

¹¹³ E.g. B. 1, pp. 239–248; Momigliano, , Filippo il Macedone, pp. 93–99Google Scholar. C. 2, pp. 167 f. is an exception.

¹¹⁴ Xen., Hellenica, VII. v. 4Google Scholar and Diod. xvi. 23, 3; Aristotle, , Politics, V. iii. 4Google Scholar (Loeb) regards stasis in Phocis as the main cause of the Sacred War; it is possible that the separatist party intrigued with Athens in 357 B.C.

¹¹⁶ B. 2, p.260 suggests April 356 B.C., and Marshall, , Second Athenian Confederacy, p. 110Google Scholar, suggests autumn 357 B.C. The terminus ante quern is provided by the presence of a garrison on Andros in May; since it was in the interest of Athens to act with all speed, I should place Chios at the beginning of the sailing season, say in March.

¹¹⁶ For membership cf. Theopompus F 63, Aeschines, ii. 116, Pausanias, x. 8, 2; the list given by P.–C, p. 172, and Kahrstedt, , Griechisches Staatsrecht, I. p. 383Google Scholar, reads Thessalois, Perrhaiboi, Magnetes, Achaioi Phthiotai, Malieis, Ainianes, Dolopes-and-Oitaioi, Dorieis, Iones, Phokeis, Boiotoi, Lokroi. Each member held two equal votes: Athens had one Ionian vote, and Sparta one Dorian; the Thessalian influence, which extended over the small tribes of north Greece (cf. Diod. xvi. 69,8), controlled fifteen votes out of the twenty-four; Phocis, as friend of Athens and Sparta, commanded four or possibly five, and Thebes with Locris commanded four.

¹¹⁷ Paus. x. 2, 1; Aesch. ii. 140; Demophilus, fr. 1 (Mueller, , FHG II. 86aGoogle Scholar).

¹¹⁸ I doubt whether Thebes had the power to call an extraordinary meeting, and also whether she would wish, by taking so prominent a part, to parade her animosity towards Phocis. My argument above is still valid for an extraordinary meeting though more valid for an autumn meeting.

¹¹⁹ In point of equity, Philomelus could advance in his propaganda a strong case against the political use of the Amphictyony, but I think equity was only effective as a rhetorical stimulant in fourth-century politics.

¹²⁰ It may seem an anachronism to relate northern and southern Greek politics at this date; but the inscription, published in Trans. A.P.A. LXV. (1934), p. 105, reminds us that Philip and Chalcidice consulted Delphi in the winter of 357/6 B.C.

¹²¹ Thebes was at liberty to turn a blind eye upon Philomelus' position as strategos autokrator, if she wished; the party strife in Phocis made it expedient to regard his action as personal, and there is no doubt that, if Philomelus had been ousted from Delphi, the opposite party in Phocis would have applied the principle . The eagerness of Philomelus to secure the approval of the Phocian commons in the winter shows that his position was still equivocal.

¹²² They evidently represent the party hostile to Philomelus; they may have had leanings towards Thebes, the champion of an extreme democracy at this time.

¹²³ Possibly including Acarnania and Byzantium, who contributed funds later (H. & H., 135).

¹²⁴ Proposed by Hegesippus, an ardent imperialist (Aesch. iii. 118); Xen., Poroi v. 9Google Scholar gives the terminus ante quem as summer 355 B.C.

¹²⁵ We know only of Achaea, which later sent troops to assist Phocis.

¹²⁶ I believe there is a close analogy between the actions of Timotheus and Chares; Timotheus bluffed the Great King into recognising Athens by supporting Ariobarzanes; Chares hoped for the same result but Artaxerxes Ochus called Athens' bluff.

¹²⁷ Xen. Poroi and Isocrates, De Pace date just before the final peace: cf. Marshall op. cit. p. 115, and Jebb, , Attic Orators, ii. p. 183Google Scholar; Miltner, , in Mitt. Ver. Klass. Phil. Wien I. (1924), pp. 45–46Google Scholar, seems to me to neglect the internal evidence, esp. Isocrates, ch. 16. Aesch. ii. 71–72 depicts the loss of Athenian prestige abroad.

¹²⁸ Dem. xiv. 33 and 40 (354/3) betrays the fear that Theban and Persian co-operation against Athens was possible as late as 354 B.C.

¹²⁹ Diod. xvi. 31 gives a list of Amphictyonic powers on the side of Thebes.

¹³⁰ Reflected in the resumption of the Naopoioi meetings with six foreign states represented in spring 353 B.C.

¹³¹ IG II. ii. 795 f. 134 dates the sending of cleruchs to 353/2.

¹³² Dem. xvi. 31: the speech is dated by Dionysius to 353/2.

¹³³ Dated by Dionysius, Dem. 13, p. 655 to 353/2.

¹³⁴ Paus. x. 2, 5 states that after the death of Onomarchus Philip joined the Theban συμμαχία; it reflects a possibility, if not a fact. Momigliano, suggestion (Filippo il Macedone, p. 106)Google Scholar, that Thebes could have helped Philip at Thermopylae, but would not, because she was pledged in the Peloponnese and afraid to admit Philip into central Greece, seems to me to underestimate the weakness of Thebes and overestimate her sense of honour.

¹³⁵ Dem., Olynth. iii. 4Google Scholar dates the siege to November 352 B.C.

¹³⁶ Dated by Dem. xix. 59.