^{page 62 note 2} Busolt, , Staatskunde, i, p. 43Google Scholar (esp. note 1) argues rightly (as I believe) that whereas ‘Rhetra’ means ‘formulation’, in Sparta a formulation only becomes a Rhetra through legislative act. He therefore concludes that our document is a ‘falschlich als Rhetra bezeichneter pythischer Spruch’. I draw the opposite conclusion: it is an act of the assembly, falsely called an oracle, Busolt, ibid. 44 note 2, quotes Wackernagel's confirmation of Wilamowitz's opinion, that the dialect cannot be recognized as specifically either Laconian or Delphic. I state here my view, to avoid ambiguity, but without prejudice: see the second part of this paper; meanwhile, cf. Latte in Pauly-Wissowa, s.v. ‘Orakel’, pp. 842, 843Google Scholar.

^{page 63 note 1} I am not sure that the lacuna indicated in our texts is essential. ‘They now call Babyka-and-Knakion “Oinous”, and Aristotle says Knakion is a river and Babyka a bridge.’ I leave the localizing of these places to the second part: meanwhile, I suggest that this passage does not compel us to assume that Plutarch drew on a second Rhetra-commentary, besides Aristotle's.

^{page 63 note 2} Εἰπεῖν γνώμην is here used in its strict sense ‘to propose a motion’, as Thuc. viii. 68. 1, cf. iii. 49. 1 (where ϒν⋯μαι means ‘proposals’ not ‘opinions’, rogationes not sententiae).

^{page 63 note 3} The principle is stated for Sparta in Plut, . Agis xi. 1Google Scholar.

^{page 63 note 4} ‘Excessive’ is implied in σκολιάν.

^{page 63 note 5} Ehrenberg, , Neugründer, p. 20 and 125Google Scholar.

^{page 63 note 6} Prof. Fraenkel advises me that we do not need to supply a cognate noun, nor any specific noun, to account for the feminine σκολιάν. Wilamowitz on Eur. Herakles line 681 explains the feminine article in τ⋯ν 'Ηρακλέοες καλλ⋯νικον ⋯ε⋯δω by supplying ⋯οιδάν from the verb: but after giving some parallels, he adds that many of his instances may be rather instances of the verwendung des femininunts für unbestimmten abstracta which is common in Greek. See further, to the same effect, Lobeck, , Paralipomena (1837) p. 363Google Scholar; Lohmann, J., Genus und Sexus (1932), p. 17Google Scholar (I owe these references to Prof. Fraenkel). I feel little competence in this matter: but in our present case, where we have ἓροιτο for the verb, and ῥήτραν is so ready to be supplied, I find it hard to doubt that they are cognate. This means that ἓροιτο (or ἓἳροιτο?) has nothing to do with the classical ἔρομαι but is the middle voice of the Homeric Ϝε⋯ρω or εἴρω: see below, p. 70 and note 4 ibid.

^{page 64 note 1} Γヅνώμας, ϒνώμη: sc. rogationes: see p. 63, n. 2 supra.

^{page 64 note 2} Palaeographically, it would seem that the copyist's eye slipped from T to Γ. The corruption of δάμω to ϒαμω is harder to explain: but the necessity of restoring some form (or some derivative) of δ⋯μω is absolute, since Plutarch certainly understood the demos to be spoken of in this clause.

^{page 64 note 3} 'Αντηϒορέω is cited in early editions of Liddell and Scott from Theodorus Studita. See below, p. 71, for a suggestion that ⋯νταγόρησεν be restored for ⋯νταγόρευσεν in Pind, . Pyth. iv. 156Google Scholar: cf. the equally unique ⋯πηϒορέων in Hdt. i. 90. 2, preserved in Hesychios but corrupted in all our MSS. to ⋯πηϒορεύων. ['Άνταγορίαν is, I hear, suggested by Treu, in Hermes lxxvi (1941)Google Scholar.]

^{page 64 note 4} IG xii, fasc. ix, 56, among the lead tablets (defixiones?) from Styra: Αντεϒοριον is no. 19.

^{page 64 note 5} See below, pp. 71 f.

^{page 64 note 6} Republished by Ehrenberg in this no. of C.Q.

^{page 65 note 1} I am tempted to connect a form Βαfκκά with βαὓζω, and to understand it as the Bridge of Shouting. Cf. Thuc. i. 87. 2, κρίνουσι βο⋯, Plut, . Lyc. xxvi. 3–4Google Scholar: and , Aesch.Pers. 574–5Google Scholar, δνσβάνκτον βο⋯τιν—αὐδ⋯ν. Βαζω could be expected to give βαüϒή, as οίμώζω etc. give (or are given by) οίμωϒή, ⋯λολυϒή, ίυϒή. I know no such noun in -κή unless ὑλακή be such. For the digamma, cf αϜυταν in Schwyz. 133 (2) line 3, αϜυτο ib. 760, αϜυταρ ib. app. I. 2, etc.: ‘Dittographie zu υ’, Meisterhans-Schwyzer, , Grammatik ³, p. 4, n. 15Google Scholar.

^{page 65 note 2} Sc. the sciscendi ratio; how the sovran body i s to enact its acta. Ancient theory did not normally distinguish between acta which prescribed regular routine action (like most of Solon's laws) or action in one special situation (like Solon's amnesty or his seisachtheia): both alike were called νόμος or θεσμός (Solon ap. Plut, . Sol. xix. 4Google Scholar: the locus classicus is , Xen.Mem. i. 2. 41–3)Google Scholar. In Sparta such acta were called ῥ⋯τραι: see the second part of this paper.

^{page 65 note 3} It presumably means the same as in Diod. xxi. 16. 4 (of Agathokles) ⋯κκλησιάσας τ⋯ν λαόν, cf. , Aen.Tact. ix. 12Google Scholar, sc. contionem habere. In the lacuna which follows in Aristotle, Thalheim's supplement [χρόνον μ⋯ν ⋯κκλησί]ασεν μικρόν assumes that ⋯κκλησιάζειν is used here for δημηϒορεῖν: this is improbable, what we want in the supplement is a statement that Peisistratos lowered his voice (e.g. [τ⋯ς ϕων⋯ς ⋯χάλ]ασενμικρόν as Kontos proposed). ‘He set about holding an assembly, and rather lowered his voice: when they said they could not hear, he bade them come up closer, to the Akropolis gate.’ The passage is not noticed in Liddell and Scott s.v. ⋯κκλησιάζω

^{page 66 note 1} See the passages cited in the previous note.

^{page 66 note 2} The cases I quote are all a good deal later than the presumable date of the Rhetra [or of Lykourgos]: at the earlier date the consultant was no doubt more personal and would get a more personal answer. But I am not concerned with the possibility that it is an oracle, so much as with the question how Plutarch's belief that it was will lead him to understand it. Plutarch, who thought it an oracle, may yet have thought the Spartan demos was the subject of ⋯πελλάζειν.

^{page 66 note 3} Other relevant glosses (the bracketed letters in the lemmata are required by the alphabetic order): ⋯πελ(λ)άζειν: ⋯κκλησιάζειν Λάκωνες, and more surprising, ⋯ππαλλάζειν: ⋯κκλησιάζειν ΅Ιωνες: another pair, ⋯πέλλακας: ίερ⋯ν κοινωνούς and (⋯π)άλλακες: ίερ⋯ν κοινωνοί: and ⋯πέλλειν: ⋯ποκλείειν (see Buck, , Gr. Dial. § 75)Google Scholar.

^{page 66 note 4} I use phratry in its Attic sense, for an association comprising more than one Genos. Cf. Pauly-Wissowa, s.v. ‘Labyadai’, xii. 308, lines 8 ffGoogle Scholar.

^{page 66 note 5} Nilsson generalizes this and sees in the Apellai the Dorian counterpart of the Ionian Apatouria, occurring wherever the month Apellaios (see next note) occurs. But are we justified in supposing that the Apellai is usually a phratry festival, just because it is so to the Labyadai? In Tenos the months ‘Απελλαιών’ and Απατουρίων coexist (IG XII. v. 872). Cf. also p. 67, n. 4.

^{page 67 note 1} For Tenos, see the previous note: for Delphi and Epidauros, n. 4 on this page. It is inferred for Sparta from its occurrence a t Herakleia in Italy (Schwyz. 62. lines 2, 95), and further that Panamos (line 101) is there the last and Apellaios the first month of the year: cf. Bischoff in Pauly-Wissowa s.v. ‘Kalender’: further references, Kubitschek, ibid. s.v. ‘Apellaios’. If μηνος Απολλωνος in a Delphic inscription (Collitz-Bechtel 1931, line 1) is the same as Apellaios, this lends colour to the view that there is a real connexion between Apellai and Apollo: or is it merely a lapsus? We read in the Labyadai inscription (Schwyz. 323, D 44–5) ται δε θυσιαι Λαβυαδαν τωπελλαιου μηνος τωι Διονμσωι.

^{page 67 note 2} IG iv,² fasc. 1, 128. Isyllos writes (line 25) ωραις εξ ωραν, and Wilamowitz, (Isyllos, p. 11Google Scholar wished to correct the Rhetra accordingly. The MSS. give two genitives in the singular, ὧρας ⋯ξ ѿρας [not ὡρᾱν, as Wilamowitz implies], and are surely right. The first ѿρας is the same sort of genitive (partitive?) as e.g. Plato, , Phaedo 58 BGoogle Scholar ⋯κάστου ἓτονςθεωρίαν⋯πάξειν. [The ‘partitive’ view of this kind of genitive is given by Meillet-Vendryes, , Traité de grammaire comparée (Paris, 1924), p. 509Google Scholar: a different view is taken by, e.g., Sommer, F., Vergleichende Syntax, 1921, p. 22Google Scholar (who compares λοεσσάμενος ποταμοίο) and Wackernagel, , Vorlesungen über Syntax, ii (1924), pp. 210, 212Google Scholar (‘im Genetiv wird der Bereich gegeben’). I owe these references to Prof. Fraenkel.]

^{page 67 note 3} Isyllos' law is known only from his verse paraphrase, which gives no date: one naturally assumes the procession is to be annual (rather than monthly or four-yearly). Wilamowitz, speaks of ‘den alljahrlich zu widerholenden bittgang’ (Isyllos, p. 10)Google Scholar.

^{page 67 note 4} In Epidauros the τελεια αϒορα met on the 4th of Apellaios, for appointing the proxenoi and thearodokoi for the coming year (IG iv², fasc. 1, 96): in Delphi, the εννομος εκκλησια met on the 7th or 8th of Apellaios, (Fouilles, III. vi. 31, ii. 102, 103)Google Scholar: these might correspond to the Great Apellai. Ekklesia on the 6th of every month, at Iasos: BCH viii. 219, JHS viii. 101, ix. 340 (Busolt, , Stoatskunde, 447, note 3)Google Scholar. In Chios c. 600 B.C. the Boule demosie met every month τηι τριτηι εξ Εβδομαιων: Schwyz. 687 = Tod, , SGHI 1Google Scholar, B 3–4 [the Ebdomaia are thus a monthly event; in the Molpoi inscription from Miletos, , Schwyz. 726, line 6Google Scholar, they recur in a very obscure context, but in some connexion with the eighth of the month]: we must, I believe, assume that the Demos has met just previously, on or near the Ebdomaia

^{page 68 note 1} 'Ωρας ⋯ξ ὣρας must mean either monthly or yearly.

^{page 68 note 2} Boisacq, Dict. Étym. s.v., while exceedingly doubtful about the etymology of ⋯πέλλαι, suggests that 'Απέλλων (which he regards as the original form of 'Απέλλων: it is frequent in Laconian inscriptions) is derived from ⋯πέλλαι. [Cf. now Glotta, xxvii, p. 32, and Arch. f. Relig. xxxii, pp. 142 ff. I owe these references to Dr. Weinstock.] Bechtel, , Hist. Personenname, p. 61, gives a number of theophoric names in 'Απελλ, e.g. 'Απελλâς, 'Απελλίκων (from 'Απελλικέτης; to this 'Απολλωνικ⋯της is a good parallel): 'Απελλαῖος (the Elean victor at Olympia in 540 B.C.) he regards as named from ⋯πέλλαι not 'Απέλλων (p. 523). Cf. the Spartan πελλ⋯ς inGoogle Scholar, Xen.Hell. iv. 3. 23Google Scholar.

^{page 68 note 3} vi. 57. 2 νεομηνίας δ⋯ πάσας κα⋯ έβδόμας ίαταμένου το⋯ μην⋯ς δίδοσθαι ⋯κ το⋯ δημοσίονίρήιον τέλειον ⋯κατέρω (sc. τ⋯ν βασιλέων) ⋯ς 'Απόλλωνος. It is perhaps legitimate to infer e silentio that there was not a monthly sacrifice to Apollo at the full moon as well.

^{page 68 note 4} The Greeks had no hard-and-fast division of days, as at Rome, into fasti, nefasti, comitiales, etc. The Boule at Athens is ordered to summon the Ekklesia ‘two days after the return of the armed force’ and thereafter to have as many consecutive sessions as are needed, and no calendar obstacles are envisaged: IG i². 63 = Tod 66 = ATL A 9, lines 34 ff.: cf. Aeschines ii. 60, iii. 67. In the last instance, Aeschines protests against the disregard of the Dionysia: in the first, it is likely that the meeting was, in fact, adjourned because of a holiday spirit which had nothing to do with the calendar (, Plut.Nic. vii. 7Google Scholar = Theopompos, , Fgr. Hist. 115 F 92Google Scholar: cf. AJP lix. 130). Alkibiades came home on the Plynteria, a dies nefaslus, ⋯ποϕρ⋯ς ⋯μέρα: the narratives are not quite specific, but it looks as if the Boule and Ekklesia met the same day, though many thought it ominous: , Xen.Hell. i. 4. 20Google Scholar, Diod. xiii. 69. 1, , Plut.Alcib. 33. 2Google Scholar. Tribute is fixed at the Great Panathenaia, the allies being then conveniently assembled: 1Gi². 63 = Tod 66 = ATL A 9, lines 27 ff., IGi². 57 = Tod 61 = ATL D 3, lines 8, 31. There was no doubt a prejudice against public business on such occasions: worst on the ⋯μέραι μιαραί or ⋯ποϕράδες, Plynteria or Anthesteria (, Xen.Hell. i. 4. 12Google Scholar, Photius s.v. ⋯ποϕράδες ⋯μέρα and θύραζε Κ⋯ρες, Suid. s.v. ⋯πο ϕραάδες ⋯μέραι Athen. 437 c) when the sublunary order stood in precarious balance: the prejudice on other festivals sprang perhaps from the dislike of work, since a festival was a ‘holiday’ in our sense, cf. the oracle in Dem. xxi. 53, έλινύειν μίαν ⋯μέραν, and the sarcasms of ps.-Xenophon, 'Αθ. π. iii. 2 and 8. Yet almost all occasions, in an archaic state, are in some sense religious, and religious business was legitimate, e.g. on the Kronia, (Dem. xxvi. 29)Google Scholar: the Labyadai certainly took decisions at their Apellai. On the whole, I would conclude that the monthly apellai was a ‘religious occasion’ involving a θνσία (see previous note), but not a holiday. Cf. Nilsson, Primitive Time-Reckoning, ch. xiii. For the monthly Ebdomaia at Chios, see p. 67, n. 4. The connexions of Apollo with the Seventh are well known: Hesiod, , WD 767Google Scholar, Proklos ad loc., id. ad Plat. Tim. 200 CD, 233, , Aesch.Suppl. 800Google Scholar, , Kallim.Hymn. Del. 251–5Google Scholar, , Plut.Qu. Gr. 9Google Scholar.

^{page 69 note 1} I can make nothing of ɦοπε νομος αποστατο at the end of an early Laconian inscription, IG v. 1, 1155 (= Schwyz. 51).

^{page 70 note 1} This, and the contemporary Ol. x. 40, are the earliest appearances of the word ἇπορος.

^{page 70 note 2} For Pyth. iv. 145, see p. 71, n. 2. Comparable uses of ⋯ϕίσταμαι in good Attic: , Aesch.Choeph. 872Google Scholar, , Eur.Med. 742Google Scholar, Isokr. viii. 81, iv. 83.

^{page 70 note 3} In the Argive αϜρετενε, αρητευε (Schwyz. 85 line 14, 91 line 3, 92 line 2, 96 (I) line 4, 99 line 4, cf. 83 B line 25), the α- is usually explained as augment: if so, it presupposes the nomen agentis ῥήτας. This ‘Speaker’ is the eponymous president of the Argive Boule: he is perhaps the ‘formulator’, the man who formulates the motions before they are voted on. Alternatively the ⋯(Ϝ)ρήτας (contracted from ⋯να(Ϝ)ρήτας) might be the ‘announcer’ of the votes.

^{page 70 note 4} I am most unwilling to correct to ἔροι or εῖροι (-το being a scribe's dittography before τούς) or to assume any corruption graver than ερ- for ειρ-. I hope I have made it plain that I am not claiming that the classical verb ἓρομαι (cognate with ἓρωτάω, ἓρευνα) could bear the sense ‘I say’: rather, that the totally different archaic verb fε⋯ρωcould be used in the middle voice.

^{page 70 note 5} Φυλάζω, ὠβάζω, formed (like ⋯πελλάζω) from ϕυλά and ὠβά, do not necessarily imply the creation of the bodies in question. This clause, vital for the historical context, is discussed in the second part of this paper.

^{page 71 note 1} Pindar would write ΑΝΤ ΑΓΟΡΕΣΕΝ, which might well provoke correction. Cf. Hdt. i. 90. 2, quoted in p. 64, n. 3.

^{page 71 note 2} I would emphasize the constitutional idiom in this poem (cf. 110, 153, 265), which has a political aim, as seldom in Pindar: the Sparta-Kyrene connexion through Euphamos is pressed, as well as the further might-have-been connexion (43–51, echoed 254–62, just before the serious politics begin; cf. 22, 175: it recurs in the companion poem Pyth. v. 69–81). Here then the words ἓσομαι τοῖος ἄλλα, etc., suggest a formula of amendment (τα ἃλλα, etc.). Note, finally, lines 145–6: if we accept (as both Wilamowitz, , Pindaros, 388 note 3Google Scholar, and Schroeder, Pind. Pyth. erklärol, ad loc, are inclined to) Chairis' emendation ⋯ϕίαταιντ', we get a good sense from the technical word: Let the Moirai rule it out of order, if kinsmen should quarrel.

^{page 72 note 1} 'Οπερ is presumably avrayopla rather than κράτος, sc. refers to καί τῷ βουλομένῳ…⋯ντειπεῖν ἔξεστιν rather than to [that plus] κύριοι κρίνειν είσί.

^{page 72 note 2} The indicative has better MS. authority than the subjunctive.