¹ History of Greece II 458 Google Scholar; Busolt, , GG I 519 f.Google Scholar, though disbelieving in any sweeping reform, puts the institutions of Sparta in the same early period. The view of Meyer, Forschungen I (1892)Google Scholar that the evidence is all an invention of c. 400 B.C. seems to me untenable.

² SBB 1918 728 f.Google ScholarPubMed

³ The most thorough advocate of the date c. 600 B.C. is Wade-Gery, whose articles in CQ XXXVII–XXXVIII support his account in CAH III, and of the date c. 550 B.C. Ehrenberg, , Neugründer des Staates (1925)Google Scholar with importan modifications in Epitumbion Swoboda (1927) 19 fGoogle Scholar. and Hermes LXVIII (1933) 288 fGoogle Scholar. Although I disagree with their conclusions, I should express the deep debt which I owe to their works and those of many others.

⁴ Part of this paper was read to the Cambridge Philological Society in January 1949. I am most grateful to Professor Adcock, Professor Wade-Gery, and Mr. R. M. Cook for their generous advice and criticism. The views and the synthesis set out in this paper are my own, but they have been formed in part by some of the many works on this subject. It has seemed best not to cumber this article with too many detailed references to these works, for it is in the nature of the subject that almost everyone agrees with someone at some point.

^4a ημην GL: ιμην Z.

⁵ The emendations καἱ τώς αὐτώς and τούτως (cf. W-G, CQ XXXVII 62 Google Scholar) for οὕτως are unnecessary, because οὕτως is perfectly intelligible, and are not in keeping with the brevity and form of the Rhetra.

⁶ W-G, CQ XXXVIII 117 Google Scholar, argues that ‘the sanctuary of Zeus and Athena, the ordering of the Phylai and Obai, have been prescribed in detail already’ and that ‘the things which are left unspecific had no doubt been specified in earlier enactments’. Neither the first nor the third of the participial phrases are unspecific (at least if my interpretation of the third is acceptable). The clause giving the Phylai and Obai is not specific, and it is true that some specification is required; that may have come later in the same rhetra or in an earlier one (for it is unlikely that any clause could precede Zeus in this rhetra). But the specification is presumably close at hand, not as Wade-Gery would have it in the traditional ‘Dorian Tribes … which purported to be descended from Herakles' three sons’. If these are the Phylai referred to by the Rhetra they are age-old and need no emphasis or mention here. Yet they occupy the emphatic part of the clause and are linked by καὶ to the obes in a rhetra where asyndeton is the rule. This shows that there is a relationship between the two, and the probability is that both are new. This is emphatically stated by W-M, SBB 1918 734 Google Scholar ‘darin kann ich auch jetzt nur die Schaffung von neuen Phylen finden’, although the admission compels him to date the reform to a date later than Tyrtaeus fr. I.

⁷ Aristotle Pol. 1272a, in drawing the similarities between the constitutions of Crete and Sparta, also dwells on the number of the Gerousia. That τριάκοντα is accusative and not (as Treu, Hermes LXXVI Google Scholar supposes) genitive, is clear from the parallelism of the participial clauses; cf. also von Blumenthal, Hermes LXXVII (1942) 212 f.Google Scholar

⁸ Similar phrases such as εἰς ὤρας mean ‘for all time’ (cf. L. & S. ὤρα A3 and B4), their meaning being often emphasised by association with αἰεί or a similar word as in Homer, Od. IX 135 Google Scholar αἰεί εἰς ὤρας, Arist., Thesm. 950 Google Scholar , and Theocr. XV 74 ; the closest and clearest parallel is Isyllus B16 (ed. W-M. Philol. Untersuch. IX Google Scholar) where the phrase may be prompted by the Rhetra. However, W-G, CQ XXXVII 68 Google Scholar states ‘the meetings were surely monthly’, with the note ‘ must mean either monthly or yearly’; so too Ehrenberg, NdS 26 Google Scholar ‘regelmässig jeden Monat’; and Treu, Hermes LXXVI 39 Google Scholar ‘von Zeit zu Zeit (Vollmond zu Vollmond)’, reading , and adducing the Scholiast to Thuc. I 67 , which may be evidence for fifth century Spartan practice but sheds no direct light on . For the essence of the phrase lies in its general character, and it can no more be said to imply a specific date in terms of months, seasons or years than our phrases ‘year in, year out’, ‘in and out of season’, and ‘forever and a day’. I am grateful to Mr. A. S. F. Gow for drawing my attention to examples of this type of phrase and their meaning.

⁹ That it continued into classical times as the Spartiates' place of assembly is clear from Plut., Pelop. XVII Google Scholar .

¹⁰ Aristotle Pol. 1273b and Plato, Laws 772c Google Scholar.

¹¹ Examples of the middle use of ἀφίστασθαι are given in L & S; they take their meaning from the context in each case, and the literal sense of the verb is consistent with the translation given above for Plut., Lyc. VI Google Scholar, a passage for which there is no parallel. The middle use of the verb in a similar context but with a different compound occurs in Thuc. I 79, 1 and V 111, 2 , where the meaning is ‘to make to adjourn’. This interpretation is also in harmony with the use of ἀποστατήρ in the Rider (below p. 45). On the other hand W-G, CQ XXXVII 69 Google Scholar translates ἀφίστασθαι in the Rhetra as ‘decline to bring motions forward’ and in Plutarch's phrase as ‘that is they shall not validate it but simply reject it’. He admits that Plutarch cannot be giving this sense to ἀφίστασθαι, but he holds that Plutarch ‘must be wrong’. In support of his view W-G adduces Thuc. IV 118, 9 where he translates ἀποστήσονται as ‘decline to entertain a proposal’. This is however a paraphrase. The literal translation is ‘they will not stand aside from any of the claims you (the Athenians) may make’. Here ἀποστήσονται is passive in meaning and governs a genitive. The same is true of his other analogy in Pindar Ol. 152. Neither is analogous to ἀφίστασθαι in the Rhetra or to in Plutarch's commentary. There is also a further difficulty in W-G's interpretation. In the Rhetra he takes εἰσφέρειν to refer to proposals in the assembly but he refers ἀφίστασθαι in his sense to the ‘preliminary probouleutic process’, that is to the process in the Gerousia. But as is so closely coupled it is unnatural to suppose it to refer to procedure in different places. Another interpretation is given e.g. by Parke, H. W., The Delphic Oracle (1939) 105 Google Scholar who paraphrases the Rhetra as ‘proposals are to be brought forward and divisions taken’. The normal phrase however is διῑστασθαι, cf. Thuc. I 87, 3, and this compound seems essential for such a meaning. For the interpretation ‘Abstimmenlassen’ cf. Ehrenberg, Hermes LXVIII 298 Google Scholar. Cf. below n. 21 for the word ἀφεστήρ.

¹² Cf. W-M, Arist. u. Ath. II 24 Google Scholar n. 24.

¹³ W-G, CQ XXXVII 64 Google Scholar and Treu, Hermes LXXVI 22 f.Google Scholar, who supplies a list of earlier emendations in an interesting article which was accessible to me only when this paper was in the final stage.

¹⁴ Von Blumenthal, in Hermes LXXVII 212 Google Scholar.

¹⁵ The demes of Attica had their agora (e.g. IG II 585 and cf. LS⁹ ἀγορά I) and the word is explained as in Bekker, Anecd. Gr. I 327 Google Scholar. The assembly in Crete was also styled agora cf. Bekker, Anecd. Gr. I 210 Google Scholar, Law of Gortyn X 33 f. and XI 10 f.

¹⁶ This suggestion I owe to Professor Robertson; cf. also Ziegler, Rh. Mus. 76 (1927), 24 Google Scholar. I have also had the advantage of discussing the passage with Mr. Beattie.

¹⁷ In Homeric diction agora is used of the assembly in opposition to the boule, cf. L. & S.

¹⁸ This is apparent in the early use of the word by Homer and by Solon.

¹⁹ Hesych. s.v. ἀπέλλαι; IG V 1, 1144; cf. L. & S. The suggestion, that ἀπελλάӡειν refers to the whole people and ἀγορά to the commons only, seems to dispose of Treu's objection ( Hermes LXXVI 23 Google Scholar) that mention of the former excludes mention of the latter. The derivation of ἀπελλάӡειν is disputed, cf. W-G, CQ XXXVII, 66 f.Google Scholar; Plut., Lyc. VI Google Scholar sees in the word a reference to Pythian Apollo.

²⁰ For instance the alliance of the Eleans and Heraeans ( Tod, GHI I 5 Google Scholar). In the original the participles of Plutarch's text may have been infinitives, but speculation on this matter is hazardous.

²¹ Here I keep the MSS. reading ἔροιτο (for arguments against the emendation to ἔλοιτο cf. Ehrenberg, NdS 20 and 125 Google Scholar and von Blumenthal loc. cit. 213). The derivation of ἔροιτο may be from ἔρομαι or εἴρω (future ἐρῶ); as the latter gives the meaning ‘say, declare’ which best fits the context, it should be preferred. The imperfect middle is used in this sense in Homer Il. I 513 and Od. XI 542. The use of σκολιάν in an adverbial sense is paralleled by its opposite ὀρθήν in the phrases ( Aristoph., Thesm. 1223 Google Scholar) and (Aristoph. Av. 1); the word to be supplied with σκολιάν or σκολιά is originally νὁδόν (the contrast of σκολιαὶ νὁδόν and ὀρθαὶ νὁδόν being common in a metaphorical application in Pindar e.g. Pyth. II 156 Google Scholar), but it falls out in compendious phrases. W-G, CQ XXXVII 63 Google Scholar supplies ῥήτραν ‘for σκολιάν to agree with’; he cites no parallel for ῥήτραν ἔροιτο and it is more probable that ἔπος is the cognate form to ἔροιτο, as in the Homeric phrase ἔπος ἐρέειν Il. III 83 etc. He is supported by von Blumenthal loc. cit. 213. W-G, CQ XXXVII 69 Google Scholar maintains that ἀποστατῆραςᾖμεν has the same sense as ἀποστήσονται in Thuc. IV 118, 9 (cf. above p. 44 n. 11), and although he does not translate the two words he presumably takes them to mean ‘are decliners in the entertaining of the proposal’. This, however, misconceives the meaning of ἀποστήρ; for this form is always active in meaning, i.e. ‘one who makes to go away’, as can be seen in the case of ἀποστήρ Aesch., Cho. 303 Google Scholar and Theb. 1015 and of στατήρ cf. Liddell and Scott s.v. In this connexion the use of ἀφεστήρ at Cnidus should be mentioned (GDI 3505.19, and Plutarch GQ 4). The title is that of the presiding officer at a committee, and Plutarch's comment describes one of his duties. The derivation is most probably from ἀφέӡεσθαι, and the literal meaning is ‘one sitting apart’. This is supported by the title of Boeotian magistrates οἱ ἀφεδριατεύοντες. The word then does not throw any light on the meaning of ἀποστατήρ. Cf. Halliday, , Plutarch's Greek Questions, 49 Google Scholar.

²² This passage is discussed by W-G, CQ XXXVII 71 Google Scholar.

²³ I take the last clause to refer to the whole sentence from down to .

²⁴ MS. οιταδενικαν.

²⁵ Cf. W-M, Textgesch. 108 Google Scholar ‘das letzte Satzglied fordert eigentlich einen Infinitiv, wenigstens ist es hart, aus ἄρχεινβουλῆς durch Zeugma ein βουλεύεσθαι für das zweite Subject zugewinnen’. Cf. the discussion below of Tyrtaeus' use of δέ.

²⁶ W-G, CQ XXXVIII 1 and 6 Google Scholar maintains that the line can also mean ‘replying to the straight proposals’ (so also Treu, Hermes LXXVI 36 Google Scholar) or ‘replying to the proposals without distorting them’. No examples of such a construction are given in Liddell and Scott, and the root meaning of ἀμείβομαι makes it highly unlikely that it could take such a construction.

²⁷ MS ιχεροεσσα.

²⁸ MS. δε.

₂₉ MS. ευθειην (Maius) ευθειης (Herwerden), cf. Spicil. Vat. 3, 15.

³⁰ MS. δε.

³¹ MS. μηδετι επιβουλευειν.

³² MS. τε.

³³ The text is given in W-G, CQ XXXVIII 3 Google Scholar; his conclusion ‘we need not shrink from correction’ implies a higher degree of corruption in the text than is apparent. In line 8 I read μηδ᾿ ἔτι for Diehl's μηδέ τι because Tyrtaeus does not use the neuter τι and Homer uses βουλεύω with a simple accusative. The form ἐπιβουλεύειν first appears in fifth-century Attic, and L. & S. are mistaken in adopting it here. As W-M remarks, πρεσβυγενεῖς is better than Plutarch's πρεσβύτας. Indeed the remark of Plutarch (789E) suggests that, while γέροντας is adduced from the γερουσίαν of the Rhetra, πρεσβυγενεῑς may have stood in Plutarch's text of the oracle in Tyrtaeus, which Plutarch quotes, and not πρεσβύτας.

³⁴ For the genuineness of line 9 Treu, (Hermes LXXVI 36 fGoogle Scholar.) refers to the use of δῆμος and πληθύς in Tyrtaeus 9, 15 and 8, 3, ἕπεσθαι in 6, 10, and νίκης καὶ κράτεος in Hesiod, Theog. 647 Google Scholar; κράτος is also used in a political sense in Plutarch's version of Solon fr. 5 and in Alcaeus 31. Treu however considers the preceding couplet to be spurious. Yet the diction is equally Homeric, and Tyrtaeus uses ἔρδειν in 8, 27; the poverty of the lines in poetic thought is paralleled in Tyrtaeus (e.g. 9, 37 f.) and the contrast of positive and negative is most common (e.g. 7, 1–4). On my interpretation the vagueness of the lines is intentional, for it conceals a limitation of the people's power under a general phrase.

³⁵ Cf. W-G, CQ XXXVIII 6 fGoogle Scholar. discussing the use of the word in Plutarch, Agis 5 and 8–11 Google Scholar, e.g. (Epitadeus) . The same use occurs in the spurious decree of Byzantium in Dem. XVIII 90. The two meanings—a proposal and an enactment—derive from the original meaning of ῥήτρα, a verbal agreement or bargain (between two parties) as in Homer Od. xiv 393; the word can be used to describe a treaty between two states (e.g. Tod, GHI I 5 Google Scholar), an enactment enshrining an agreement between two parties (e.g. IG xiv 645 95 f. and 145–6), and a proposal for agreement between constituent parts of the state, as here and in Plutarch Agis. The original meaning, a mutual agreement between parties, is emphasised in Tyrtaeus by the compound ἀνταπαμειβομένους, elsewhere unknown, and by the sense of exchange in the verb ἀμείβομαι.

³⁶ The word παρεγγράφω may mean either ‘subjoin’ or in a bad sense ‘interpolate’; the context does not suggest that the latter meaning is to be preferred here (as it appears to be by W-G, CQ XXXVIII 2 Google Scholar).

³⁷ W-G, CQ XXXVIII 1 Google Scholar translates ‘on the grounds that it was part of the god's command’. Plutarch does not mention any ‘part’: προστάσσειν means to enjoin whereas προστάσσειν is necessary for the meaning ‘enjoin in addition’.

³⁸ This is implicit in the hypothesis of Andrewes, CQ XXXII 89 Google Scholar and W-G loc. cit. 2, that the Heracleidai are the subject of ἔνεικαν.

³⁹ W-G, CQ XXXVIII 3 Google Scholar note 1 remarks that ‘the marginale has been unfairly spat upon’. The alternative is to spit upon Plutarch; for one or other must be so treated. Of the two the known Plutarch is a less deserving target than the unknown hand of the marginale.

⁴⁰ The text of column A2 of the papyrus was first published by W-M SBB 1918 728 fGoogle ScholarPubMed. In the text as printed above only some of Diehl's restorations are reproduced.

⁴¹ Homer, Iliad III 362 Google Scholar ; V 655 ; XI 592 . Cf. Lorimer, H. L. BSA XLII 114 Google Scholar.

⁴² Cf. Homer, Iliad XI 592 Google Scholar quoted in the last footnote and Callinus I 10 quoted in the text above.

⁴³ An aorist or imperfect verb in the third person plural can be as easily restored in line 11 as a form of the first person plural.

⁴⁴ I take ἀλλά to repeat the antithesis begun by δέ in the preceding couplet and also to stress [ὄκνου] ἄτερ. The same use of ἀλλά appears in Tyrtaeus VII, 1; VII, 31; VIII, 1; VIII, 21; and in Callinus I 9.

⁴⁵ For instance, in Pind., Schol. Pyth. V 92 Google Scholar .

⁴⁶ Lorimer, loc. cit. 92–3 ‘that from c. 700 B.C. onwards hoplite equipment was general at Sparta is clear, and, in view of the interpretation sometimes put on certain passages in Tyrtaeus or the poetry which goes under his name, the fact is of importance’.

⁴⁷ Ibid. p. 128.

⁴⁸ ὐπὸ πλατέος τελαμῶνος ἀσπίδος εὐκύκλου.

⁴⁹ Ibid. 122 n. 2, the word basin-like is used by Miss Lorimer. Her interpretation is different from mine, because she thinks both epithets refer to hoplite warfare. I agree that both might do so; but in this case if a distinction is drawn, my interpretation seems to fit her description better and also to fit the fragment of Mimnermus quoted below.

⁵⁰ In general Tyrtaeus' evidence on weapons, armour and tactics is, as Miss Lorimer says, perplexing. Onereasonmay be that the Messenian Wars were not struggles between hoplite forces but partook more of guerilla tactics; and, as it takes two hoplite sides to make a hoplite battle, the Spartans may have been compelled by their enemy to modify their equipment and their tactics.

⁵¹ Wyss, , Antimachi Colophonii Reliquiae (1936) 83 and 88 Google Scholar. The restoration ἤἰξαν seems necessary to explain οἱ πὰρ βασιλῆος which should depend on a verb of motion: cf. Xen., Anab. I 1, 5 Google Scholar τῶν παρὰ βασιλέως.

⁵² Paus. IX 29, 4.

⁵³ Lorimer, loc. cit. 120 inclines to put the introduction of hoplite warfare in Ionia generally to a period not much earlier than 600 B.C.

⁵⁴ SBB 1918, 734 Google ScholarPubMed.

⁵⁵ Hermes LVI (1921) 347–8Google Scholar.

⁵⁶ The position of the spear when a hoplite is marching and when he is about to engage is well illustrated by the Chigi Vase, cf. Lorimer, loc. cit. 81. Bowra, , New Chapters in Greek Literature, III 64 Google Scholar shares the view of Gercke and remarks ‘the description of the marching here is reminiscent of some Homeric passages’; he refers to Iliad XV 7 Google Scholar10 and XXI 162 f., both of which concern close combat. He also maintains that the triple division by tribes re-appears in lines 68–71 of the fragment, but four groups are mentioned if one reads further and notes οἴ δέ in line 74.

⁵⁷ CQ XXXVIII 119–20.

⁵⁸ In CAH iii p. 562 Google Scholar Wade-Gery writes that ‘the same date (i.e. shortly before 600 B.C.) is implied in Herodotus' account of the Eunomia, where he relates the Arcadian Wars of the early sixth century as the immediate consequence of the reform…. Herodotus stultifies his narrative by implying that the Reform took place some centuries before those immediate consequences which gave him occasion to mention the matter at all’. W-G here makes Herodotus imply two different things, first that the reform is dated to shortly before 600 B.C., and second that it took place some centuries earlier. What Herodotus says is fortunately more precise: the reform according to the Spartan tradition was in the reign of Leobotes. This statement outweighs any implications which may be held to transfer the reform to a later date. It is true that Herodotus reverts abruptly from the digression to the narrative: the phrase . suggests that he envisaged three stages at Sparta, first expansion after the reform, second recession under defeats (causing them ἡσυχίαν ἄγειν), and third a return to aggression against Arcadia. Now this interpretation of Sparta's development may be good or bad history; but the implication that the reform really took place shortly before 600 B.C. lies with Wade-Gery and not with Herodotus.

⁵⁹ In translating this passage I take the superlatives ἐπὶ πλεῑστον.. χρόνον and ἐκ παλαιτάτου to contrast with one another, emphasising the point that Sparta was longest in distress but earliest to achieve ordered government; the double καί, each being emphasised by strong hiatus, to mean ‘both… and’; and αἰεί, being under the double καί, to take its starting point from ἐκ παλαιτάτου and to mean that consecutively from the time of the early achievement of orderly government Sparta was free from tyranny. This last point is relevant to Thucydides' mention of tyrannies in other states. For a different interpretation of the passage see Andrewes, CQ XXXII (1938) 94 Google Scholar, who holds that ἐκ παλαιτάτου is to be taken only with the first καί clause and who expands the passage to mean ‘yet at a very early time she brought herself to order, without undergoing a tyranny; indeed she never had a tyrant, for it is about four hundred years etc.’. If Thucydides meant to say this, he expressed himself badly and this is possible enough; but the explanatory sentence, which follows, states not that Sparta's power was due to the lack of tyranny but that it was due to the continuity of her constitution over four hundred years. This point makes me reject Andrewes' view that ‘the end of the στάσις, the change to εὐνομία, is not dated, except by the words ἐκ παλαιτάτου and the presumption is that it comes within the four hundred years’. Gomme, , Commentary on Thucydides, I 131 Google Scholar, is critical of Andrewes' interpretation.

⁶⁰ Andrewes, loc. cit. 89 f. has shown correctly that εὐνομία can mean orderliness in the citizens as well as orderliness in government. Aristotle, Politics 1294a Google Scholar defines these two meanings neatly to illustrate his views on constitutional government. Having stated that in an aristocracy the highest posts are assigned to the best citizens, he concludes ‘it seems an impossibility for a city governed not by the aristocracy but by the base to have well-ordered government (εὐνομεῑσθαι), and similarly also for a city that has not a well-ordered government to be governed aristocratically’ (trans. Rackham). His point is that in a well-ordered government the best govern ex hypolhesei, i.e. such a government is an aristocracy. He is dealing here with government and not with the orderliness of citizens, and the verb εὐνομεῑσθαι has a constitutional meaning. In the next sentences he remarks that ‘to have good laws enacted but not obey them does not constitute well-ordered government (εὐνομία). Hence one form of good government must be understood to consist in the laws enacted being obeyed, and another form in the laws which the citizens keep being well enacted (for it is possible to obey badly enacted laws)’. It is clear from this that the noun εὐνομία properly comprises both ideas, that of well-ordered government and that of orderliness in the citizens; either idea without the other is an incorrect usage of the word, whether it be to disobey good laws or to obey bad laws. Myres, CR LXI (1947) 80 fGoogle Scholar. stresses the double significance of the word. Andrewes appears to me to err in taking one meaning of εὐνομία and excluding the other—e.g. p. 93 ‘Then, μετέβαλον ἐς εὐνομίην, they decided to lead better lives. There is here no word of the constitution, though it is clear that there was an important change of some kind.’ Both ideas are implicit in the phrase: orderliness in the citizens in contrast to their being κακονομώτατοι in earlier times and orderliness in government, as is shown by what follows. For Herodotus wrote , and ὧδε ‘in the following manner’ points to the constitutional and social reform of Lycurgus. If further evidence is required, it is supplied by the summary , which resumes . In the passage above, Aristotle uses the verb εὐνομεῑσθαι in the primary sense of having a well-ordered government. Such too is the sense of εὐνομήθησαν in Herodotus; for he has just described the constitutional reform. It cannot be said of this passage that there is here no word of the constitution and the meaning is that they led better lives. I stress this because Andrewes deduces that ‘there are two sets of facts which Herodotus has falsely combined. There is the change from κακονομία to εὐνομία, a change which was effected about 600 B.C. and has not necessarily any bearing on the constitution whatever, and there is the system of Lycurgus, a system of which the Spartan constitution was a part, whose institution was placed, rightly or wrongly, at a very early stage in the history of Dorian Sparta.’ I cannot see any justification in the text of Herodotus for making this separation; the question whether Herodotus' account is due to a misunderstanding of the Spartan tradition is another matter. Equally in the case of Thucydides 1181 see no grounds for separating his statements into two similar sets of facts; the only case when he uses εὐνομία (which is sometimes emended to αὐτονομία) can carry only one meaning, that of a well-ordered government, viii, 64, 5 . Cf. Ps-Xen., Ath. Pol. I, 8–9 Google Scholar.

⁶¹ My preference is for the later date, on the ground that the early chapters of book I show some signs of revision after the outbreak of the Decelean war.

⁶² 1271b 20; cf. 1316a 34, where the tyranny of Charilaus is said to have ended in a change to aristocracy. This latter passage evidently refers to the reform of Lycurgus; the aristocratic nature of his constitution was later modified by changes in the reign of the king Theopompus.

⁶³ 1274a 30.

⁶⁴ 1296a 20; 1273b 33.

⁶⁵ 1270a 1. That the wars precede Lycurgus, is certainly implied by the connexion with the next sentence σχολάσαντεςδέ. It is just possible that Aristotle was referring to later wars in which the laxity of the women was originally of little significance because the wars were on foreign soil.

⁶⁶ 1313a 25; ἐπικαθιστάναι can mean either ‘set over’, cf. Plato, Tim. 72b Google Scholar, or (in later Greek) ‘establish besides’. It does not state that the office of ephor was instituted by Theopompus.

⁶⁷ 1263b 41; 1269b 20; 1270a 20. The context of the last passage shows that Aristotle is thinking of landed property; despite a lacuna in the text the sense is not in doubt. It is probable that a distinction was drawn between the acquired property and the original lot of land, cf. Heracleides Ponticus fr. 2 (7) , probably deriving from Arist. Lac. Pol., and Plut., Inst. Lac. 22 Google Scholar ; cf. Ziehen, Hermes LXVIII (1933) 227 Google Scholar and Meier, Klio Beiheft XLII (1939) 38 fGoogle Scholar. Cf. Hdt. IV 57, 4 on heiresses at Sparta.

⁶⁸ 1270b 1. The word ἄφρουρος is probably Laconian, cf. φρουρὰν φαίνειν, and refers to service in the levy, and is not Attic, referring to garrison service.

⁶⁹ 1270b 18.

⁷⁰ 1271a 13 and 23; 1333b 11.

⁷¹ 1271a 27.

⁷² 1272a 3 ; cf. 1271b 20 for Lycurgus and Crete.

⁷³ Cf. Kessler, Plutarchs Leben des Lykurgos (1910) 104 f.Google Scholar

⁷⁴ Fr. 533 ap. Plut., Lyc. 1 Google Scholar; for the dating of the Olympic truce, cf. Jacoby, , Philol. Untersuch. 16 (1902) 116 Google Scholar. Cf. Heracleides Ponticus fr. 2 (4).

⁷⁵ Fr. 535 ap. Alex., Clemens Strom. I 152 Google Scholar.

⁷⁶ Fr. 536 and 537; cf. Lex Patm. 152 and Heracleides Ponticus fr. 2 (4) (5).

⁷⁷ Fr. 538 ap. Plut., Lyc. 28 Google Scholar. I do not agree with Kessler op. cit. 111, that Rose has not included sufficient of this chapter in the fragment. Cf. Heracleides Ponticus fr. 2 (4).

⁷⁸ As exemplified by Tyrtaeus' poems and as indicated by Thucydides I 4 and 9.

⁷⁹ Hdt. vii 204, viii 131; Tyrtaeus 2; Hellanicus, FGH 4 Google Scholar F 116, Plato, Laws iii 683d Google Scholar; Hellanicus F 188, Hdt. iv 145 f.; Pinder, Pyth. I 65 Google Scholar, Isth. vii 12 f.Google Scholar, Pyth. v 75 Google Scholar, Arist. Lac. Pol. fr. 532; cf. Plato, Laws 682e Google Scholar.

⁸⁰ Parke, The Delphic Oracle 67 Google Scholar ‘curiously enough, the original and historic oracle about Gela appears to be preserved’, and 70 (of the foundation of Syracuse) ‘in fact there is no reason for denying its claim to be authentic’.

⁸¹ Dunbabin, The Western Greeks 470 Google Scholar ‘it is established that the accepted tradition of the foundation of the colonies, found in Thukydides and Antiokhos and in the main followed by later sources, contains a great deal that is historical and substantially accurate’.

⁸² The evidence is cited by Wide, Lakonische Kulte 281 f.Google Scholar

⁸³ For a summary of views on this topic cf. How and Wells, , Commentary on Herodotus I 85 Google Scholar.

⁸⁴ For example cf. Pind., Isth. vii 12 Google Scholar (of Amyclae).

⁸⁵ E.g. Parke, op. cit. 105 (of the Rhetra) ‘its traces of archaic Spartan diction prove that it is a genuinely ancient traditional document’; Busolt, Staatskunde (1920) 46 Google Scholar.

⁸⁶ The oracle is frequently quoted in later authors, cf. Wide loc. cit. It is cited in Diod. vii 12 (Ephorus) and in Const. Exc. with two additional lines:

.

One explanation is that these lines were added at some later date, in order to support the ascription of the constitution to Delphi alone; another is that the lines belong to another oracle (for Diod. loc. cit. refers to several such oracles) and were wrongly added to the oracle cited by Herodotus. The former is the usual explanation, cf. Parke op. cit. 102 who refers to Pausanias' pamphlet mentioned in Strabo 366 as a possible source for the addition. It should be noted however that Strabo does not say that the oracles cited by Pausanias were not genuine and Pausanias' case would certainly be a better one if they were already accepted as genuine before he wrote. Nor is it clear how the addition of these lines would help the case of Pausanias. The evidence is hardly sufficient to justify any choice between the two explanations.

⁸⁷ Arist. Lac. Pol. fr. 535 ap. Alex., Clemens Strom. I 152 Google Scholar . The same authors also maintained that Lycurgus copied some 01 the institutions 01 Crete: Arist. Politics 1271b 23, 1272a 1, Ps-Plato, Minos 318c Google Scholar etc., Ephorus ap. Strabo 477 and 481. The value of Delphi's sanction is emphasised by Xen., Lac. Pol. VIII 5 Google Scholar.

⁸⁸ Heracleides Ponticus fr. 3 (2) cites Archilochus as an exponent of this view: . Although the fragment is so short, it seems to give more support to the view (probably Aristotle's) than does the citation from Homer.

⁸⁹ Pindar, Pyth. I 62 fGoogle Scholar. Hellanic., FG 4 F 116 Google Scholar stated that Eurysthenes and Procles drew up the constitution; he was censured by Ephorus (70 F 118) on the grounds that he had not mentioned Lycurgus but had attributed the achievements of Lycurgus to others. The fact that Hellanicus did not mention Lycurgus shows that he was not discussing the Lycurgean reform; the burden of Ephorus' censure is that Hellanicus transferred to an earlier date some matters which Ephorus thought were due to Lycurgus. In this dispute one has more confidence in the view of Hellanicus. The passage does not reveal Hellanicus' date for the Lycurgean reform.

⁹⁰ Hellanicus 188, whatever the value of his derivation of the name from Helos.

⁹¹ Hdt. IV 145 Minyans from Lesbos; Arist. Pol. 1270a , which probably refers to the period before Lycurgus; Hdt. iv 149 Aegeidae, cf. vii 15, Arist. fr. 532.

⁹² Tyrtaeus I 12; Pindar, Pyth. I 62 fGoogle Scholar and Schol. ad. loc.; Pindar, Schol. Pyth. v 92 Google Scholar. The 27 phratries survived in the festival of the Karneia, Athenaeus 141 F; the three racial tribes survived as the Ionic tribes did at Athens after Cleisthenes' reform.

⁹³ Adoption at Sparta in the case of the Minyae, Hdt. IV 145 ; at Athens, Philochorus fr. 91.

⁹⁴ Hdt. I 65, 2; Thuc. I 18.

⁹⁵ Cf. Ehrenberg PW s.v. Obai; e.g. IG V 1 688 ωρβ Λιμναεω[ν].

⁹⁶ Paus. III 16, 9–10 . before the time of Lycurgus. Thuc. I 10 of Sparta in his own day. Hdt. III 55 calls Pitana a δῆμος.

⁹⁷ Plut., Lyc. XVI Google Scholar a new-born child is brought before the eldest φυλέται by the father; this suggests that the child belonged by birth to the same tribe as the father.

⁹⁸ IG V 1 564, 4–5 ; 480, 9–10 φυλης Κονοουρεων. Hesychius s.v. ; cf. Hdt. IX 53 for the form . Steph. Byz. s.v. ; cf. IG V 1 515 for the form Μεσοάτης. The Ἀμυκλαῖοι or Ἀμυκλαιεῖς appear in Xen., Hell. iv 5, 10–11 Google Scholar, but whether as members of an obe or a tribe is not clear. The entry of Hesychius s.v. should probably be referred to the Dymanes, one of the three racial tribes; cf. Steph. Byz. s.v. Δυμᾶν, Pind., Schol. Pyth. I 121 Google Scholar , Hesychius s.v. Δύσμαιναι (perhaps Δύμαιναι).

⁹⁹ The fact that the phyle-members and the obe-members of e.g. Konooura are both called Konooureis has led to the belief that the phyle and the obe are ‘identisch’ (Ehrenberg Obai col. 1694) and that the terms are applied in late inscriptions to ‘the same unit’ (W-G, CQ XXXVIII 117 Google Scholar). If they were synonymous and interchangeable the retention of both terms would be surprising, as Kahrstedt, Gr. Staatsrecht 20 Google Scholar has remarked. But the fact that they are homonymous does not mean that they are synonymous; in the same way ‘Jerseymen’ can be used both of men not resident in Jersey but descended from original Jersey families and of men resident in Jersey. The interpretation given in the text accounts also for the phenomenon in IG V 1, 515 , for there is no difficulty in a member of the Mesoa tribe holding office in the obe of Amyclae or vice versa. On the other hand Ehrenberg has to assume that the Epimeletes were appointed by the state. The view of Kahrstedt, Gr. Staatsrecht (1922) 20 fGoogle Scholar. that the obes are subdivisions of the tribes is not warranted by the evidence.

¹⁰⁰ Aristotle fr. 541.

¹⁰¹ As Hdt. IX 53 names the Πιτανάτης λόχος and ὁ Πιτανάτης λόχος (if the MS. reading is retained) and also refers to Πιτάνη as a deme (III 55), he probably derived the name of the λόχος from the obe. Thucydides I 20, 3, however, asserts that no such λόχος ever existed. Despite the view of W-G, CQ XXXVIII 121 Google Scholar, that ‘Thucydides has slipped up badly’, there seems to me little doubt that Thucydides is correct; for he would have been rash indeed to be dogmatic on a point which could be settled by reference to Sparta or Spartan prisoners of war. Thucydides was, I think, drawing attention merely to an error of nomenclature (for Aristotle too includes no λόχος of this name) and not to an error about the method of recruitment. In the fourth century, when the Spartan army was brigaded by μόραι, the evidence of Xen., Hell. IV 5, 10–11 Google Scholar suggests that male members of the same family were in different μόραι and that οἱ Ἀμυκλαῖοι or Ἀμυκλαιεῖς were also in different units, whether the latter were members of an obe or of a tribe named after Amyclae. But this itself may have been an innovation introduced at the time of the new organisation by μόραι.

¹⁰² I take it that the phrase in the Rhetra refers to divisions of the people for different purposes, for instance for purposes of election and military service on the one hand and for purposes of local government and representation on the other.

¹⁰³ In taking separate demes from each of the three regions of Attica to form a new tribe Cleisthenes cut across the local centres of clan organisation; cf. Arist., Ath. Pol. XXI, 3 Google Scholar where it is pointed out that Cleisthenes abandoned the earlier τιττύες.

¹⁰⁴ Cf. Plut., Pelop. XVII Google Scholar quoted on p. 43, n. 9.

¹⁰⁵ The importance of Zeus and Athena in the state-cult of classical Sparta is emphasised by Wide, Lakonische Kulte 8 Google Scholar and 54, citing Zeus Agoraios and Athena Agoraia, Zeus Amboulios and Athena Amboulia, Zeus Xenios and Athena Xenia, all associated in Pausanias' time with the contemporary Agora or the old Agora. Athena Chalcioicos is also named Athena Poliouchos, and besides her shrine stood the statue of Zeus Hypatos; there were also cults of Zeus Lakedaimon, Zeus Boulaios, Zeus Tropaios and Zeus Agetor. The sacrifice to Zeus and Athena, which the king made before leading the Spartan army out of Spartan territory ( Xen., Lac. Pol. 13 Google Scholar), probably dates back to the time of the ξυνοικία. It is possible that the cult title Syllanios is peculiar to the act of the ξυνοικία.

¹⁰⁶ Arist. Pol. 1270a; the sentence concerning the disposal in marriage of the heiress seems to refer back to Lycurgus' legislation.

¹⁰⁷ Ibid. 1271a; and 1272a.

¹⁰⁸ The story of a Lycurgean γῆς ἀναδασμός in Plut., Lyc. VIII Google Scholar is clearly a late invention. When the demand arose in the second MessenianWar, Tyrtaeus opposed it in his ‘Eunomia’ presumably because it was a novelty ( Arist., Pol. 1307b Google Scholar).

¹⁰⁹ Arist., Ath. Pol. XXI 6 Google Scholar.

¹¹⁰ Xen., Lac. Pol. X 7 Google Scholar.

¹¹¹ Plato, Laws 630 Google Scholar pointed out the fallacy of supposing that Lycurgus' aim was primarily military.

¹¹² Pol. 1271a 23.

¹¹³ Pol. 1273b 34.

¹¹⁴ Hdt. I 65, 5; Xen, Lac. Pol. VIII 3 Google Scholar; Arist., Pol. 1270b Google Scholar 20 (discussed above p. 56).

¹¹⁵ Plut., Cleomenes 10 Google Scholar, where the tradition is included in the tendentious speech of Cleomenes. Arist. Lac. Pol. fr. 539 cites the proclamation made to the people by the Ephors on entering office . Cases of disobedience under the agoge were brought before the Ephors, Xen., Lac. Pol. IV 6 Google Scholar. For the general question cf. Busolt, GG I 555 f.Google Scholar

¹¹⁶ Plut., Lyc. VII Google Scholar implying that the Ephorate was then created for the first time. Arist. Pol. 1313a 25, cf. 1270b 20, appears to attribute its origin to Lycurgus and its emergence as a powerful factor to the reign of Theopompus. As Aristotle may still be the ultimate source of Plutarch at the beginning of chapter VII, Plutarch's account may be corrected by reference to Aristotle. The reference to Chiton's institution of the Ephorate in Diog. Laert. I 68 has been convincingly explained as a misunderstanding of Sosicrates, cf. Busolt, GG I 556 Google Scholar, 2. Plato, Laws 692 Google Scholar also puts the constitutional importance of the Ephors later than the reform of Lycurgus.

¹¹⁷ Plut., Lyc. VI Google Scholar.

¹¹⁸ Arist. Pol. 1272a 12, 1273a 6–13; Diod. XI 50 and Thuc. I 79 f.

¹¹⁹ Tyrtaeus Eunomia fr. 2 and 3; Hdt. I 65–66; Thuc. I 18.

¹²⁰ The existence of this list of Ephors from 757 is generally accepted both in its own right and because Ephors are found in colonies of Sparta (Thera, Taras by implication from Heraclea, Siris, and Cyrene).

¹²¹ These are all points which weigh heavily against the dating ofthe reform to c. 600 or 550 B.C. The most awkward is the fact that Tyrtaeus III a and b paraphrases part of the Rhetra with its rider. W-G, CQ XXXVIII 115 Google Scholar, taking the Rhetra to be contemporary with Tyrtaeus, argues that Tyrtaeus ‘sought to reinforce its authority by asserting that there was (presumably in the Royal Archives) an ancient oracle enjoining the substance of Clauses II and III [i.e. the latter part of the Rhetra and the rider]’. This hypothesis is far from convincing. Faced with a fundamental reform the Spartans would be swayed not by an oracle of academic antiquity but by the sanction of Delphi at the moment—a sanction which could doubtless be obtained and which the tradition says was in fact obtained. It is necessary to assume that the Kings, the Pythii at Sparta, and Tyrtaeus combined to concoct a falsehood which by the standard of their time was impious, and it is hard to believe that the Spartans would be persuaded that the Rhetra was a revival of an old one never enacted. Ehrenberg, NdS 33 and 49 Google Scholar cuts the Gordian knot more boldly ‘Dass sie den Inhalt der grossen Rhetra widergeben, kann kein ernstlicher Zweifel sein. Dann aber können sie nicht von Tyrtaios stammen’. ‘Der Gesetzgeber Lykurg ist eine Schöpfung des wahren Gesetzgebers von 550.’ It is however difficult to imagine how the lawgiver of 550 was able to convince his contemporaries that the reform was really due to the remote Lycurgus and how the famous poem Eunomia was foisted into the poems of Tyrtaeus after 550 B.C. In Epitumbion Swoboda Ehrenberg has modified his position to the extent that he regards the poem of Tyrtaeus as genuine.

¹²² Plut., Lyc. IX and XXVI Google Scholar, Agis X; cf. Kessler op. cit. 92 f. The tradition of poets and artists visiting Sparta covers the seventh and sixth centuries, cf. Tod-Wace, Catalogue of Sp. pp. 99 f.Google Scholar

¹²³ CR XLIX (1935) 185 Google Scholar.

¹²⁴ Buschor, AM LII (1927) 12 Google Scholar. Desborough, BSA XLIII (1948) 267 Google Scholar notes the wide gulf between the latest local Mycenaean and the Protogeometric style at Amyclae.

¹²⁵ Ibid. and Dawkins, Artemis Orthia 19, 49, 62 Google Scholar.

¹²⁶ AO 62.

¹²⁷ AO 18. The small percentage of slipless pottery at the Artemisium may represent the earliest period of occupation, contemporary with that of the Acropolis and of Amyclae.

¹²⁸ AO 54, 62; AM 14.

¹²⁹ Lane, BSA XXXIV 99 f.Google Scholar

¹³⁰ AO 19 ‘It is not likely that any great error will be made if it (the archaic altar), and the early temple with it, are assigned to a date earlier than 800 B.C.’ AO 239, the oldest ivories being needles for applying kohl to the eyelids: ‘they cannot be dated later than the 9th century B.C.’

¹³¹ La Crète Dédalique (1947); cf. Lane loc. cit. for possible influence of Crete and Thera on Laconian Geometric.

¹³² Cf. Busolt, GG I 338 Google Scholar.

¹³³ Strabo 362, adding that in his own day the declining population lived in thirty πολίχναι.

¹³⁴ The cities of the Perioeci are styled πόλεις by Hdt. VI I 234, Xen., Hell. VI 5, 21 Google Scholar etc.; their independence dated from the pre-Lycurgean period presumably, cf. Isoc., Panath. 176 fGoogle Scholar. That the bulk of them were Dorian is implied in the legends of the conquest and settlement of Laconia and is stated by Thuc. VII 57, 6 in the case of Cythera. The tradition in Plut., Lyc. VIII Google Scholar that the Perioeci held their land by κλῆροι, although misapplied in the context, may reflect a historical fact, that their system of land-tenure was Dorian in character. The tradition that the towns of Laconia were reduced after the Lycurgean reform appears in Paus. III 5–7.

¹³⁵ Dunbabin, The Western Greeks 435 f.Google Scholar

¹³⁶ Thuc. I 12, 4 .

¹³⁷ That these were divisions of Dorian Corinth and of Dorian Megara and that they had constitutional applications is clear from the ὀκτάδες at Corinth after the fall of the tyranny (Nic. Damasc. fr. 60) and from the five strategoi and five demiourgoi at Megara.

¹³⁸ Cf. Demargne op. cit. 148 for a summary of recent views on this controversial issue.

¹³⁹ A term echoed in Plut., Lyc. VI Google Scholar, describing the Great Rhetra as , in Tyrtaeus IIIb μαντζίας and in Isyllus E15 . Plut. 1116 F states that the Lacedaemonians preserved the oracle concerning Lycurgus . The earliest records at Delphi were written on skins (Eur. fr. 629).

¹⁴⁰ Arist. Pol. 1272b mentioning the sussitia as well as the constitution. It is possible that the former survived into the Roman period, cf. Livy XXXIV 61 in circulis conviviisque.

¹⁴¹ No reference has been possible to Chrimes', Ancient Sparta (1949)Google Scholar which was published when this article was in page-proof.