¹ This essay was awarded the Cromer Prize for 1928 by the British Academy. The author wishes to express his acknowledgments to Wadham College, Oxford for the Richards Studentship, 1926, to the Old Bradfordians' Club of London for their Drummond Studentship, 1926–7, and to Oxford University for the A. M. P. Read Scholarship, 1927. He is also indebted to the late Dr. J. Wells, Mr. M. N. Tod, and Mr. H. T. Wade-Gery for their criticisms and suggestions.

² There is one class of exceptions: navarchs, who were usually private individuals. (Probably Anchimolius, c. 512 B.C., was one of these, for he goes to Athens by sea.) The only individual exception is Euaenetus (Her. VII. 173), who commanded the reconnoitring expedition to Tempe (480 B.C.); he was but not of the royal house, as Herodotus mentions with surprise. It is significant that at the only period before the Peloponnesian war, when Sparta had more commitments abroad than her organisation could meet, we find her traditional methods thus admitting of exceptions.

³ Thuc. III. xciii. 2.

⁴ But we have the precedent of Meyer, , Theopomp. p. 117Google Scholar, for this assumption.

⁵ See Appendix on Thucydides' use of the term ἄρχων when applied to Spartan officers.

⁶ The evidence for (presumably) earlier harmosts in the Perioecis itself is rather doubtful, but their existence may be regarded as probable. Compare Kahrstedt, , Gr. St. I. p. 229Google Scholar, who, however, is apt to be rather too positive on the subject. Contrast Meyer, , Theopomp. p. 269Google Scholar, who errs in the opposite direction. Kahrstedt (ib. p. 23, note 6) does not seem to prove his theory that Heraclea was in effect part of the Perioecis.

⁷ They are an instance of the working of a νόμος, quoted by Thucydides (IV. xxxviii.), whereby under certain circumstances three Spartan officers were nominated, so that on the death of the senior officer, another might be ready at once to take his place.

⁸ When the Peloponnesian garrison was handed, over to the Athenian generals by the men of Nisaea in 424 (Thuc. IV. lxix. fin.), Thucydides implies that the Spartan ἄρχων was not the only Spartan in Nisaea.

⁹ The only previous instance where this phrase is used of soldiers is Thuc. I. lx., where it refers to Aristeus. There is an obvious parallel between the two leaders, the importance of which lies in this:—it shows that Sparta has adopted from Corinth an active policy of offensive warfare by land. One may also compare the Corinthian officer and garrison in Ambracia (c. 425 B.C. Thuc. III. cxiv. 4) with the fully developed Spartan harmost. Diodorus actually converts him into a Spartan (XII. lx. 6).

¹⁰ Thuc. IV. lxxx. 1: since Perdiccas was originally to have given half the τροφή of the army (id. lxxxiii. 6), presumably the Chalcidic league was responsible for the other half.

¹¹ Presumably Brasidas, like Eurylochus, had only two Spartans with him as subalterns; this will, then, have been one of them.

¹² A fuller discussion of this type of harmost is reserved till later.

^12a On this νόμος cf. Isocr. Busiris 18, and Nairn, , Classical Review, Sept. 1929Google Scholar.

¹³ Thuc. V. xxxiv.; here mentioned for the first time.

¹⁴ The customary form used throughout the latter part of Thucydides' narrative is where the ‘allies’ evidently include, for instance, the helots and perioecs. Compare also Thuc. VII. lviii.—Sparta among the list of Syracusen allies—and such later examples of commanders as Pharax (Diod. XIV. lxiii. 4; lxx. i.). Also note Diod. XIV. Ixix. 5: and Plut. Pelop. xxxi.

¹⁵ Thuc. VIII, xxiii. 4. Alcamenes had been killed at Spiraeum: Eteonicus was presumably his successor.

¹⁶ This mention must have occurred in an excursus, as Theopompus did not begin his narrative in the Hellenica till after the battle of Cynossema, by which time Pedaritus was dead: Meyer has cleverly conjectured that it occurred in connexion with the operations of Pasitelidas at Chios (Meyer, , Theopomp. p. 160Google Scholar).

¹⁷ Xen., Hell. I. vi. 18Google Scholar, 8, 26 and 37: cf. Pareti, , Mem. della reale Accad. di Torino, LIX. (1909), p. 119Google Scholar and note 6, who calls him ‘duce straordinario’ exactly =ἁρμοστής.

¹⁸ It seems best to assume that when Xenophon says he means 407–6, since he himself carefully pointed out that Lysander was not allowed to be navarch a second time (Xen., Hell. II. i. 7Google Scholar). Contrast Meyer, who dates Dercyllidas, ' dismissal to 404–3 (Gesch. d. Alt. v, §759)Google Scholar.

¹⁹ Thuc. VIII. lviii.; i.e. in opposition to any ancestral possessions, which had been in Europe. Compare VIII, xliii. Lichas' criticism of the first treaty.

²⁰ Compare above, how Pharnabazus accepted Abydus (Thuc. VIII. lxii. 1), and later resigned Calchedon (Xen., Hell. I, iii, 9Google Scholar).

²¹ Thuc. VIII, lxxxiv.: contrast xxxvi.: just before the second treaty.

²² Thuc. VIII, cviii. 4 seq., Diod. XIII. xlii. 4, who makes his usual error of substituting Pharnabazus for Tissaphernes.

²³ Chios (Thuc. VIII. xxiv. 6, xxxviii. etc.), Thasos (Xen., Hell. I. i. 32Google Scholar), and Byzantium (Xen., Hell. I. iii. 18Google Scholar: where the issue was complicated by famine, but Clearchus had been harsh, Diod. XIII. lxvi. 6).

²⁴ Cf. Busolt-Swoboda, Griechisches Staats und Rechtsaltertümer, p. 1323Google Scholar.

²⁵ E.g. Diod. XIV. xiii, i,

²⁶ E.g. Nepos, C.. Lys. I. 5Google Scholar, in sua teneret potestate. Plut. Lys. XIII.,

²⁷ Probably Theopompus is the original authority favourable to Lysander; for in the 10th book of his Hellenica he took occasion to praise Lysander's industry and self-restraint (Plut., Lys. XXX., and Athen. XII. 543bGoogle Scholar = Frag. 21a and b Oxf.). Ephorus, on the contrary, sought to explain Lysander's conduct by his passion for supreme power, and so brought forward the story of his attempts to overthrow the Heracleid kings by working the oracle (Plut. Lys. XX. and XXV. seq.). Probably therefore, Ephorus is the common source unfavourable to Lysander. Cf. Schwartz, , Quaestiones ex hist. Graec. saec. quart. desumpt. Rostock, 1893Google Scholar.

²⁸ Cf. supra, and compare Isocr. IV. 122, which may be taken literally as referring to the end of the Peloponnesian war.

²⁹ Plut. Lys. xix., i.e. presumably he had freely commandeered supplies for his troops.

³⁰ The figure ‘three months’ is probably not meant as an exact estimate of the period during which the decarchies were in power, but only as a guess at the duration of their first massacres.

³¹ Apparent date about 401.

³² First published by Homolle, , Bull. Corr. Hell. III. (1879), p. 12Google Scholar.

^32a Contemporaneously with the second Spartan empire, Dionysius I of Syracuse was building up a system of φρούραρχοι and garrisons to control Sicily and Magna Graecia. But as he was not at all concerned with maintaining even a nominal independence in his subject cities, outside Syracuse, his φρούραρχοι are not parallel to Sparta's harmosts.

³³ Sparta had even guaranteed to repay to Tissaphernes, as soon as the war was ended, all the subsistence which Persia had provided (cf. the third treaty, Thuc. VIII. lviii. 6). But we do not hear that this claim was raised by Persia in 405–6. On the contrary, Lysander was even given the surplus, presumably by Cyrus' generosity (cf. infra).

³⁴ Cf. Isocr. XII. 67; Polybius. VI. xlix. 10.

³⁵ This figure may be shown to be not too grossly improbable, if one estimates very roughly the minimum Spartan expenditure in some year for which figures are available: e.g. 399–8 B.C.

(i) Wages for 8000 soldiers (Xen., Hell. III, i, 28Google Scholar: i.e. 6000 Cyreans, Xen., Anab. VII. vii, 28Google Scholar + 2000 Ionians, Diod. XIV, xxxvi, 2) at 1 daric per mens. = 96 000 darics = 400 tal. (Attic).

(ii) A fleet of ? 35 triremes (so 402–1, Xen., Anab. I. iv. 2Google Scholar) with wages at 3 ob. each man = not less than 210 tal.

To this minimum total of 610 tal. must be added the other lesser expenses of the war (in addition to wages), and perhaps the cost of maintaining the garrisons in the cities occupied by harmosts (cf. Xen., Hell. II. iii. 13Google Scholar: Athens as an exception). Of course, the Spartans managed largely to recoup the expenses of their Ionian war by the booty taken. But unless the expedition was a sheer gamble, they must have had enough income already to face the expenses. In later years the Ionian war increased still more in cost.

³⁶ It is assumed here that Beloch is incorrect in identifying I.G. V. i. as a Spartan tribute-list, 404–398 B.C.

³⁷ Cf. Aristotle, , fragm. 544Google Scholar, Rose.

³⁸ Cf. also the penalty on Thorax (infra).

³⁹ Did Sparta use and encourage the issue of coins from newly-restored Aegina? Cf. Xen., Hell. V. ii. 21Google Scholar.

⁴⁰ Xen. Anab. II. vi. i. seq.

⁴¹ Diod., loc. cit.

⁴² Plutarch, , Pelopidas, 15Google Scholar: Poralla, Prosop. d. Lak. No. 585, accepts this identification. Compare Pareti, op. cit. p. 127.

^42a Cf. Polyaen. II, ii, 7, that Clearchus had 4 triremes: perhaps his district was larger (cf. infra).

⁴³ It takes part in the issue of ΣΓΝ coins: contrast Beloch, III. i. p. 77.

⁴⁴ As in the case of Dercyllidas in 394–3, Xen., Hell. IV. viii. 5Google Scholar.

⁴⁵ Eressus and Antissa, Diod. XIV. xciv. 4.

⁴⁶ Cf. Schol. to Aristophanes, Eccl. 102, Xen., Hell. IV. viii. 31Google Scholar.

⁴⁷ Cf. Schol. ad Aeschin. II. 77: (sc

⁴⁸ I cannot follow Beloch (III. i. 78, note 1) in extracting any facts about Melos' relations with Sparta from Isocr., Aegin. 18, 19Google Scholar. For the expulsion of the conjectured harmosts in the restored cities of Chalcidice, cf. infra.

⁴⁹ For Ceos, cf. infra.

⁵⁰ Dem., XX, 59, Schol. ad Aristid. III. p. 85Google Scholar (Dind.): cf. Beloch, , III. 90, note 2, and I.G. II.²1. 24Google Scholar, with Wilhelm, , Eranos Vindobon., p. 241Google Scholarseq. There must have been some harmosts in to account for the ‘harmosts from Europe’ who took refuge with Dercyllidas (cf. infra). Probably Ischolaus (Polyaenus, II. xxii.) was a harmost here, and fought against Chabrias in 390–89. Cf. Schäfer, , Demosthenes (ed. 2), p. 43Google Scholar, note 5, and I.G. II. 22. The alternative date would be about 378.

⁵¹ Xen., Hell. II. iii. 13Google Scholar, and Aristotle, Ath. Pol. xxxvii., who dates the sending even later, after the occupation of Phyle.

⁵² This was the use to which the 100 talents were put: later Xenophon, (Hell. II. iv. 30Google Scholarfin.) speaks of Lysander's force as

⁵³ Xen., Hell. IV. iii. 15Google Scholar: it had been sent through the Isthmus, Plut. Ages. xvii.

^53a Cf. the harmost at Decelea, c. 405 B.C., Dem. XXIV. 128.

⁵⁴ He has to ask for reinforcements from Sparta. Polyaenus, id.

⁵⁵ Contrast Xen., Anab. I. iv. 3Google Scholar: but this may be deliberately reticent. Compare also for further references concerning Clearchus, Meyer, V. iv. §833.

⁵⁶ Compare Meyer, V. iv. § 761, who admits this, and so has to assume that Spartan harmosts were set up in Asia, 405–4, and were withdrawn again by 403. In fact they had never been set up.

⁵⁷ (Xen., Hell. III. i. 3Google Scholar), i.e. in addition to the Greek islands and mainland, which as yet made up the whole of the Spartan empire.

⁵⁸ Xen., Hell. III. i. 3Google Scholar. Contrast the attitude of the rest of Greece, Isocr. V. 95.

⁵⁹ One may compare the reorganisation of Heraclea by the commission under Rhamphias sent to inspect Brasidas (cf. supra).

⁶⁰ Busolt-Swoboda, , Griechisches Staats und Rechtsaltertümer, p. 1325Google Scholar, note 5, accepts a theory of Breitenbach, that ἐπιμελητής is used here because harmosts had been withdrawn by this time from Asia, quoting Xen., Hell. III. ii. 20Google Scholar (cf. infra). But this agreement was not made till a year later, and then was only hypothetical (cf. Judeich, , Kleinasiatische Studien, p. 35Google Scholar, who, however, calls Draco a harmost).

⁶¹ Xen., Hell. III. ii. 12Google Scholar:

⁶² Probably his chief motive in recording the previous commission was because it provided an opportunity to vindicate his Cyreans.

⁶³ Kahrstedt, (Gr. St. I. p. 250Google Scholar, foot) does not seem to have proved his theory, that the harmosts mentioned by Xenophon in this passage must be harmosts of the Hellespontine region. He argues that the major and the minor harmosts cannot be found in the same area. But did not Dercyllidas control the Hellespont also? He had commanded, e.g. at Lampsacus (Xen., Hell. III. ii. 6Google Scholar).

⁶⁴ Cf. Isocr, . Panegyr., 142Google Scholar, and Meyer, , Theopomp. p. 19Google Scholar, note 2.

⁶⁵ Xen, . Hell. III. iv. 10Google Scholar, and compare and contrast Plut, . Ages. 8Google Scholar.

⁶⁶ We cannot be certain that Xenophon was settled at Scyllus quite as early as 393–2. Diogenes may, however, be writing loosely; and in any case Xenophon's presence there at that date does not seem impossible, if he was exiled for fighting against Athens at Coronea (394).

⁶⁷ Meyer, (Theopomp. p. 108Google Scholar) prefers to connect Alexander with Thibron's first campaign; but his objections to the later date do not seem sufficient.

⁶⁸ παραπλησίως, (i.e. φρουρὰν ἐκβαλόντες), μετέαλον, XIV. lxxxiv. 3Google Scholar. Cf. Ditt., Syll. 126Google Scholar for a decree of Erythrae in praise of Conon at this time, and Dem. XX. lx. for a statue in his honour at Ephesus.

⁶⁹ Xen., Hell. IV. viii. 39Google Scholar: for the form ἁρμοστήρ cf. I.G. V. 937 and Plato Comicus, πρέσβεις (Kock, ), frag. 126Google Scholar.

⁷⁰ Philodicus is mentioned by Diod. (XIV. xcvii. 3), in company with Ecdicus (Eudocimus) and Diphridas, as a Spartan commander, but is omitted by Xenophon. Was he appointed as harmost of Rhodes—a position never achieved? So also Judeich, , Kleinas. Stud. p. 90Google Scholar.

⁷¹ E. von Stern seems to go too far in assuming, on the strength of Polybius, IV. xxv. 5, that Sparta actually failed to withdraw her harmosts. Xen., , Hell. V. i. 33Google Scholar, iv. 10, 46 does not prove anything, except the existence of pro-Spartan oligarchies in Thespiae and Plataea: while Isocr., Plat. xiii., shows that the harmost was in Plataea only during the subsequent war; cf. infra.

^71a Cf. Autocles' speech, Xen., Hell. VI. iii. 7Google Scholar seq.

⁷² On the harmosts in Thebes and Oreus, and Sparta's plans against Jason, , cf. Class. Quarterly, Oct. 1927Google Scholar.

⁷³ Cf. Xen, . Hell. V. iv. 56Google Scholar, for a famine in Thebes.

⁷⁴ Isocr. XIV. xiii.; he is probably to be identified with Gerandas, Plut. Pel. XXV.

⁷⁵ Compare the situation of Clearchus or Hippocrates after the battle of Cyzicus.

⁷⁶ Xen., Hell. V. iv. 46Google Scholar, and cf. the inscription from Thespiae, (I.G. VII. 1904Google Scholar) of a Spartan

^76a Polyaen. III. ix. 48; cf. Arist, . Rhetor. III. 10Google Scholar. For an alternative dating cf. Beloch, III. i. 92, note 2.

⁷⁷ For a late instance of a set up by Cleonymus for a short time in Troezen, , c. 277Google Scholar B.C., compare Polyaen. II. xxix. 1.

⁷⁸ Compare Walker, C.A.H., V. App. iv.

^79a Compare, however, the commanders and garrisons of Hellenic monarchs, e.g. Cassander.

⁷⁹ Clearchus, Thuc. VIII. viii. etc. Xenophon, , Hell. I. iii. 15Google Scholar: Pedaritus, Thuc. VIII, xxviii. Theopompus, Frag. viii. (Oxf.).

⁸⁰ E.g. (Xen, . Hell. VI. 1Google Scholar. 18) (Thuc. I. cxi.), or (Xen. Hell. passim), which never occurs in Thucydides.