^{page 19 note 1} The speech is published in Reiske's, Oratores Graeci, vol. viiiGoogle Scholar ; Bekker's, Oratores Attici, vol. iv (Oxford), vol. v (Berlin)Google Scholar ; Dobson's, Oratores Attici, vol. iv (with Dobree'S Adver saria, pp. xx–xxi)Google Scholar ; Müller's, K.Oratores Attici, vol. ii: also separately byGoogle ScholarHass, , De Herodis Attici oratione περ πολιτεας (Leipzig, 1880)Google Scholar ; Drerup, ['Ηρώδον]πɛρ πολιτεας, in Studien z. Gesch. und Kult. d. Alt. ii. I (Paderborn, 1908Google Scholar : the most serviceable text); Meyer, , Theopomps HeUenika, pp. 201 ff.Google Scholar (Halle, 1909: his text is substantially a reprint of Drerup's with less apparatus). I refer to the works of Reiske, Dobree, Hass, Drerup, Meyer, by the author's name only: for the discussions by Beloch, Adcock, Knox, Morrison, see p.22, n. I.

^{page 19 note 2} I take this account of the MSS., and all my knowledge of them, from Drerup, pp. 1–2 his apparatus.

^{page 19 note 3} rdquo;Iσως ἢν τις εἴπῃ … πολιτνὺεσθαι. This is a sub-section of the wider question discussed in §§ 28–31: ‘what have we to apprehend from Sparta's known behaviour? She does not bully her neighbours: someone may say that she demands oligarchy, but that is nothing to fear: so her known behaviour gives no cause for apprehension.’ This section is enclosed between its opening and closing sentences; κ ποων χρ παραδειγμτων ρρωδεῖν; … παραδελματα τοιατα παρεχομνονς ρρωδεῖν σὐκ νἰκε. Incidentally this puts the meaning of παραδελματα τοιατα in § 31 beyond question and makes it unlikely that παρδειλμα ibid. is used in another sense.

^{page 20 note 1} Philostratos in his Life of Herodes (vitae sophistarum, ii. 1), § 14: τῷ δ κριτᾳ κα προσετετκει κα παργαγεν αὐτν ς ἤθη 'νλλνων τως μελοὺμενον.

^{page 20 note 2} This seems ‘absurd’ to me. On the general problem of how to distinguish a real from a bogus speech I may refer to Meyer's excellent pages, 209–18. As a ‘work of art’ out speech is incredible. A sophistic speech puts all its cards on the table and is nothing if not explicit: our speech is maddeningly allusive: contrast e.g. the naive ὑ' ὑμν δ' π σοφᾳ in Gorgias’ Palamedes 16 with the gloss-provoking ᾧ γρ ἢμεῖς ἰσχὺομεν of ‘Herodes’ § 32. Among examples of transitional idiom of C. 400 B.C.I note the use of ξενικῷ πολμῳ for ‘foreign war’ in §II, whereas ξνονς in § 15 is ‘mercenaries’.

^{page 20 note 3} Nestle, , N. Jahrb., 1903, pp. 191 ff.Google Scholar ; Köhler, , S B Berlin xxvi, 1893, 504–7Google Scholar : id. ib. 1895, 457. Cf. Meyer, p. 213, n. 1, Diels-Kranz, Vor sokraliker ⁵, n. on 85 B 2.

^{page 20 note 4} Diels-Kranz, , Vorsokratiker ⁵, 85 B 1–2Google Scholar .

^{page 20 note 5} I assume that Thrasymachos' longer fragment is earlier than our speech, whereas the ὑπρ ∧αρισαων is later. I discuss the date of the ὑπρ ∧αρισαων in my text: the occasion of the longer fragment is clearly before the fall of Athens (Athens is still at war, and is ridden by faction: sometime between 413 and 405).

^{page 21 note 1} The fragment shows that the audience cannot be Kyros.

^{page 21 note 2} See below, p.22. (For a summary of the speech,;and for the general situation, see Morrison, pp. 69 ff.)

^{page 21 note 3} Such is clearly the thesis of the opponents of the ‘Herodes’ speaker.

^{page 21 note 4} From a title and one sentence it is rash to conclude very much: but Thrasymachos, like Gorgias and Isokrates, was an educationist, a purveyor of ideas, a ‘Philosopher’, whereas Lysias and Kritias were men of affairs.

^{page 21 note 5} Meyer, p. 280, n. 2, remarks that Thrasymachos' longer fragment ‘operates throughout exactly as our speech does with the notion of νγκ. I speak below in Section E of our author's use of νκη: he welcomes it (§ 2 ναγκῖο κα προσκον § 4 πρτον μν ὡς γαθν οτι … δεὺτερον δ ὡ ναγκαῖον), but Thrasymachos regards it as an evil (e.g. our sedition is the necessary result of folly). The difference appears to me fundamental. Even in the opening argument, so nearly identical (note βονλμην μν in § 3 and in Thrasymachos), the tone has this essential difference: ‘I think some God has had us in his especial care’ our author says: Thrasymachos says ‘since God has brought us to this pass’.

^{page 22 note 1} Beloch, GG² iii. 2. 16–18Google Scholar (400/399), Costanzi, , Studi ital. d. filol. class. vii. 137–59 (410/409), Drerup (404), Meyer (400/399),Google ScholarMorrison, , C.Q. xxxvi. 68 ff. (404)Google Scholar : Adcock, and Knox, in Klio, xiii. 249 ff.Google Scholar, Münscher in Pauly-Wissowa, viii 952–3 (art. ‘Herodes’), Willamowitz, in SB Berlin, 1925, p. 335, n. 5Google Scholar, and others, regard it as a product of Roman times (Herodes Atticus) whose dramatic date is incapable of being fixed. I refer to the works of Beloch, Morrison, Adcock, 's name only: for Drerup and Meyer, see p. 19, n. I.

^{page 22 note 2} See Morrison, p. 68, n. 3: ‘This principle would prove that e.g. the Decrees of Callais and the Pirate Law on the Aemilius Paulus monument at Delphi were spurious’. I might add the psat Xenophon's απολ., or will someone say that this too is by a Roman sophist? Adcok's jeud'esprit, in which he undertakes to prove that no date could be right, is more reasonable: but he has nothing against 404 except the Eleans § 28, for whom see pp. 22–3.

^{page 22 note 3} Meyer, p. 268, n. 1, seeks to meet this by saying that there is no proof that Lykophron was yet supported by Sparta (as he certainly was in 395, Diod. loc. cit.). The hypothesis of Drerup and Morrison, that he was already on the Spartan side in 404, and that his attack was a consequence of our speaker's policy being rejected, seems to me good and economical; but it is not essential to their decisive argument viz. that his action (whatever its nature) could not simply be disregarded, as it is in e.g. § 13. [After the battle, if Lykophron were Sparta's ally, Larisa might still look to Sparta for protection, much as Finland has done to America: this was, I conceive, the plea of the ὑπφ ∧αρισαων (see above)—an unsuccessful plea.]

^{page 22 note 4} Adcock puts the Elis War in 399–397, so that Archelaos is dead before Elis ‘rejoins the League'. Cary, in CAH vi, p. 33Google Scholar , puts it in 401–399. Though it is indifferent to my view of the ‘Herodes’ speech, I think Meyer's date for the war is right (sc. 402–400).

^{page 23 note 1} In § 29 I suggest (p. 31) ὢσπερ οὖν νθάδε Μακεδóνα, which reinforces this point: but argument does not depend upon this correction. The speaker is in either case contrasting Sparta and Archelaos as neighbours, and the cboirep clause either points (ὢσπερ οὖν) or modifies (ὢσπερ οὐδ') the contrast.

^{page 23 note 2} Diodoros 14. 17. 5 says тŰς δαπάνας тοῦ πρòς 'Aθηναίους πολέμου καтŰ тò πιβάλλον αὐтοῖς μέρος πῄтουν: Elis is treated like an Athenian tributary who is in arrears [for παιттεῖν see Meritt, , Wade-Gery, , and McGregor, , ATL, pp. 212–13]Google Scholar. The contrast with Thuc. 8. 3. 2 is interesting: Elis was clearly seeking to establish her claim to a religious neutrality (Diod. loc. cit., perhaps preferable to Xen. Hell. 3. 2. 22), and it looks as if Sparta did not insist on ships or service but did insist on payment: or is it (more simply) that after Aigospotamoi Sparta felt stronger? Since our speech is certainly after Aigospotamoi, our argument does not require us to choose between these alternatives: but the problem is interesting. It might be maintained that Sparta's demand in 402 was a bolt from the blue, and that up till then no one could have dreamt that Elis was less independent than Argos: but this seems to me unlike Spartan diplomacy. The spokesman of Elis in most of the negotiations was no doubt Hippias, who appears to have dwelt on the early history of the two states (Plato, , Hippias mai. 281 B, 285 D, E)Google Scholar: I have little doubt that the very remarkable theories about Elis' ‘sacred neutrality’ which Ephoros recorded (F. gr. Hist. 70 F 115) are the theories advanced by Hippias in this context. The thesis is that Sparta had guaranteed the neutrality in old days. Whether (as I believe) these theories were embodied in Hippias' Olympionikai need not be discussed here: I have elsewhere urged that it was in the Olympionikai that the famous diskos was first published (C.Q. xxxviii, p. 5, n. 3); it proved that Lykourgos of Sparta had joined in guaranteeing the Sacred Truce. Elis' claim to neutrality is perhaps further illustrated by IυΟ 30 (= IGA 105 cum addendis) εδοξεν Αθαν[αlι]ον Mελανοπο hυιυν | προξενον και ευεργε|тαν тον Αλειον γραϕσ|αι εν ολυνπιαι εδοξεν. The Αλειοι take their decision in Olympia (the punctuation in 11. 1 and 6 is mine, but it seems necessary): they must surely be the Eleans. Diphilos, of a famous family of kaloi and strategoi, is no doubt Laches' brother, and no doubt the same man as the strategos who was stationed at Naupaktos in 413, Thuc. 7. 34. 3. I suspect his Elean proxeny is connected with that strategia. The arrival of Alkibiades' πλοῖον ϕοπтηγικóν at Kyllene (Thuc. 6. 88. 9) is quite consistent with neutrality: the presence of Korinthian hoplites in Pheia (ib. 7. 31. 1) is perhaps more of a borderline case, but cf. Ephoros loc. cit. тοὺς δι' αὐтς тς χώρας ἰòνтας σтραтοπέδῳ тŰ δπλα παραόνтας πολαμβάνείν μεтŰ тν κтν ρων ἒκβασιν.

^{page 24 note 1} Philostratos says that Kritias could adapt Gorgias' manner to his own style by εὐγλωттία: Epist. 73 (see p. 28, n. 5). If I have rightly understood this phrase, it seems to me a very exact account of what we have here: the antithetical and sometimes purple style, but kept closer to natural speech.

^{page 24 note 2} The acute Dobree wished to correct the title to περ πολέμου: an e compendio πο i.e. πολέμου? Munscher, , in Pauly-Wissowa, , viii, p. 952Google Scholar, speaks of ‘den Titel περ πολιтείας der wohl nicht vom Verfasser stammt’: he maintains Herodes' authorship, but does not explain how the superscription came to be right in one particular but wrong in the other.

^{page 24 note 3} Alkibiades was recalled, presumably on Kritias'motion, immediatelythe 5,000 tookpower (Thuc. 8.97.3): the 400 had refused to recall him (8. 70. 1) and Alkibiades (now a t Samos) had declared against the 400 and for the 5,000 (8. 86. 6). The fortress at Eetioneia was the answer of the 400 to the threats of Alkibiades from Samos (90. 1–3), and it was suspected that they meant not only to keep Alkibiades out but to let the Spartans in (90. 3). Phrynichos was one of the chief authors of this plan (90. 1) and went on the embassy to Sparta which was believed to be treasonable and for which Antiphon was later condemned to death by the 5,000: on his return Phrynichos was assassinated (92. 2), and the rioters demolished the Eetioneia fortress (92. 4–11): to take part in demolishing Eetioneia ranged you with the 5,000 and Alkibiades, against the 400 and Phrynichos (92. 11). It is thus certain that ps.-Demosthenes is wrong in calling Kritias a pro-Eetioneia man in 411: Eetoneia was, as Kritias recognized, the acid test, and in 411 he took one side, in 403 the other. —In view of Thuc. 8.97.3, there can be no doubt that Xenophon is wrong in saying that Alkibiades was still an exile in 407 (or 408?) just before his return to Athens (Hell. 1. 4. 10). Kritias' motion for Alkibiades, like his motion against Phrynichos, clearly belongs to 411.

^{page 24 note 4} ‘If you want to praise Kritias,’ says Aristotle (Rhet. 3. 16, 1416b28), ‘you have to tell the whole story, for not many people know it.’ Was his memory kept green in the Academy?

^{page 25 note 1} It seems to me that our speaker has to make out the case (sc. cannot take it for granted) that the sort of constitution which Sparta imposes is the same as the sort he has himself worked for. If Kritias is the author, and the speaker is one of his confederates [is he Prometheus? see next note], this may suggest that Kritias had at least not been a recognized Spartan agent in the years when Athens was still fighting Sparta. But this indication is far from certain: Kritias may have worked openly for Sparta and against Athens: and if he did, it would certainly have been demoralizing for a man of his temper (cf. his judgement on Kimon, , Vorsokr.⁵88Google Scholar B 52 = Plut. Cim. 16. 9).

^{page 25 note 2} Is Prometheus our speaker? is this in fact a reference to the known circumstances of the περ πολιείας?

^{page 25 note 3} Morrison, p. 65 f.

^{page 25 note 4} Aristotle reports the jest, Pol. 3. 2, 1275^b27, and adds тŰ μν ἲσως πορν т δ' εἰρωνευòμενος ‘perhaps in some degree he really didn't know the answer, in some degree was pretending not to’. See Morrison, p. 71. I agree with Morrison that he is mocking the arbitrary choosiness of the demiourgoi, whose function seems to be more or less exactly that of the katalogeis of 411 at Athens: are the politophylakes of Ar. Pol. 1305^b28 perhaps the successors of the demiourgoi? if so, the passage is illustrated by ps.-Lysias, 20.13, cf. 2. [In Larisa, as in oligarchic Boeotia and in the Attic experiments, there was thus created (by exclusion) a new category, neither politai nor xenoi, whom I have proposed to call hypomeiones: cf. § 31 of our speech.] Gorgias (I am assuming) not only taught the Aleuads and their friends, but shared their political outlook, sc. their attitude both to the Larisan bourgeoisie and to Archelaos. The remarkable and serious attack on Archelaos in Plato's Gorgias is surely no casual irrelevance: is it not due to Gorgias having identified himself with Archelaos' friends? ὡς καλς οἶδε Пλάтων ἰαμβίζειν. I feel confirmed in this by the fact that Antisthenes' Archelaos contains an attack on Gorgias (Athen. 220 d). Antisthenes, a cruder moralist than Plato, will surely have harped on Archelaos' enormities. Gorgias no doubt maintained they were irrelevant, and Polos' unskilful argument will be a travesty of this. [I suspect this is the plagiarism of Antisthenes of which Plato was accused by Theopompos, (F. gr. Hist. 115 F 259 = Athen. 508 d)Google Scholar: of minimal importance in the dialogue to us, it is just what Theopompos would notice.]

^{page 25 note 5} Philostratos', words (in his Life of Kritias [vitae sophistarum, i. 16], § 2 = Vorsokr. ⁵ 88 A 1), βαρυтέρας ποίει тŰς λιλαρχίας, can hardly be taken as a serious corrective of Xenophon's Theramenes. Philostratos is indeed combating Xenophon's thesis in Mem. 1. 2. 24 that Kritias was corrupted by the Thessalians, and he professes to base himself on what Kritias said in Thessaly: διαλ7egr;γòμενος тοῖς κεῖ δμναтοῖς κα κ7agr;θαπтòμενος μν δημοκραтίας πίσης διαβάλλων δ' ‘Αθηναίους ὡς πλεῖσтα νθρώπων μαρтάνοтας. Has Philostratos been reading Kritias' Thessalian speeches (e.g. the ‘Herodes’ speech)? I doubt it. Just above he has said that Kritias βουλεύμαтος … ξυνελ7aacgr;μβ7agr;νεν ὡςμηλòβοтος ’ Αттικ ποϕανθείη [cf. Isokr. 14. 31]: this seems to me an obvious lie based on school declamations [cf. the subject for debate in Hermogenes, pp. 33–4 Rabe: ‘Kritias has taken sanctuary with the Tyrannicides: shall he be draggedGoogle Scholar away?’] and I suspect a similar origin for these conversations with Thessalian magnates. I think little was known of Kritias' Thessalian visit beyond Xenophon's two passages (Hell. 2. 3. 36; Mem. 1. 2. 24): he was there at the time of Arginousai (Hell. I.e.) though he had been in Athens shortly before (see p. 33, note 1, below), the bulk of his exile was spent in Thessaly (Mem. I.e.), he probably came back when Athens surrendered (Lysias 12. 43 f.; Andokides 1. 80).

^{page 26 note 1} Aristotle, Pol. 1305^b28, may refer to the duration of this hoplite constitution: see p. 25, n. 4, above.

^{page 26 note 2} Cf. p. 25, n. 5.

^{page 26 note 3} Cf. p. 25, n. 5. He would presumably not cut his Thessalian connexions at once.—I follow the chronology set out by Colin, G., Xenophon historien d'apres le livre II des Helleniques, 1933 (Annales de l'Est), pp. 112–16Google Scholar. Surrender of Athens, April; the Thirty take office, September; Thrasyboulos in Phyle, December 404: death of Theramenes, February; of Kritias, May 403.

^{page 26 note 4} He is rebutting the thesis [not that it is by Kritias himself, but] that it is by Herodes; who (he says) would imitate Kritias, whereas this speech does not. More positively he says (p. 67) it is in Gorgias' style (though not of course by Gorgias). I note that Philostratos says that Kritias, like Thucydides, shows Gorgias' influence: they both adapted his grand manner to their own style, Kritias by adding εὐγλωттία. See p. 28, n. 5; p. 31, n. 4; p. 24, n. 1.

^{page 26 note 5} Kritias' fragments are in Diels-Kranz, , Vorsokr.⁵88Google Scholar: the prose fragments are 88 B 31–73. of these, 53–73 are single words: only 31–6, 39–40, 42, 44 contain sentences verbatim.

^{page 27 note 1} Gorgias', fragments are in Vorsokr. ⁵82Google Scholar B: Thrasymachos' ibid. 85 B. Of Isokrates, 18 gives perhaps the best comparison; it is still fifth century: 21 is notoriously exceptional, admitting hiatus freely.

^{page 27 note 2} Plato had been close to Kritias at the time our speech was written (Ep. 7. 324 D), so that perhaps his words in Meno 70 A-B, 76 c-E, are specially in point, on Thessalian literary taste and how you had to cater for it. [I agree with Burnet that Kritias in the Timaeus and Critias is our K.'s grandfather.] See too PhiloStratos Epist. 73: παρ' ο7iacgr;ς т ῤηтορεύειν γοργιάζ6egr;ιν ρωνυμίαν ἒσχεν: on Kritias' adaptation of Gorgias' style see p. 28, n. 5. We do not indeed know that Gorgias ever avoided hiatus as strictly as our speech does. He tolerates it in the Epitaphios, an early work: the Palamedes, where he tends to avoid it, is presumably a little later than Euripides' Palamedes of 415 (Ael. VH 2. 8): so that avoidance may well have been the modern practice in Thessaly by 404; the influence of Thrasymachos may have helped (cf. Vorsokr. ⁶ 85 B 2).

^{page 27 note 3} Some papyri of Thucydides have -тт-: on -σ;σ- in Alkidamas see Luria, , Riv. d. Fil. liv, 1926, p. 220Google Scholar, n. 1. Athenaeus 151 d quotes θαρσ7agr;λέως from Xenophon Anab. 7. 3. 29. Dionysios appears (for his own part) to write γλσσα (Ant. R. 1. 66.1, 67. 3) γλттα (ibid. 1. 68.1; 2. 7. 3–4) θάλσσα (i. 2. 3–4, 3. 3, 9. I, 10. i) θάλαттα (1. 3. 5, etc.) quite indifferently: and in spite of Marcellinus, vit. Th. 52Google Scholar, he quotes Thucydides as writing (1. 22. 4) κтμα ς ε … σύγκειтαι (de comp. 22). Whether himself or his copyists, the result is the same for us: but his habit of translating Herodotos into Attic suggests that he himself thought spelling unimportant to style, and that the ‘scent of antiquity, to be essential, should breathe of something rarer than an odd arrangement of type’ (Quiller-Couch, , Preface to Oxford Book of English Verse, p. viii)Google Scholar.

^{page 27 note 4} Antiphon's spelling has been discussed e.g. by Luria, , op. cit. (previous note), p. 220Google Scholar, n. 1, Rosenkranz, , op. cit., infra (next note), p. 144Google Scholar. The statistics for the orator always include the Tetralogies, but surely should not: whoever wrote them, they are no evidence for the language or style of Antiphon's actual pleadings. [See e.g. Hermes, lviii, p. 104.] Nor do I think much gained by including the single example from I (ἒλασσον in 1. 19); the curiously ‘tragic’ context (Klytaimnestra in 17, κπίνουσιν ὑσтάтην πίσιν in 20, the general tone of 19 itself) makes it rather unsafe to build on. The important statistics are from V and VI, and roughly we may say -σσ- is regular in V, -тт- in VI. This materially alters the look of things: moreover, the common view that VI is later than V (and thus -тт- a modernism) is not certain nor even very probable: Meritt wishes to date VI to 419 (AFD, p. 174) and V is probably later than that. The divergence of V and VI, then, may be due to their having a different Textgeschichte: if it is due to Antiphon himself, it may be that -σσ in V is ‘hellenism’ (see the next note), suited to Antiphon's foreign client. No doubt he often had foreign clients and therefore often ‘hellen-ized’. [It is perhaps worth noting that in Vogel's, statistics of Kurzenmeidung, Hermes, lviii, pp. 87ff.Google Scholar, ‘Herodes'’ 33.5 per cent, disregard is high, but is pretty exactly the same as Antiphon v, Lysias xxiv and xxx.]

^{page 28 note 1} The problem of -тт- and -σσ- is discussed by Rosenkranz ‘Der lokale Grundton und die personliche Eigenart in der Sprache des Thukydides und der alteren attischen Redner’ (sc. down to Andokides; excluding Lysias, , etc.), in Indogermanische Forschungen, xlviii, 1930, pp. 144–5Google Scholar, with statistics from early Attic writers. He concludes (p. 145): ‘wer speziell fur Athen schrieb, wahlte TT: das gilt fur die Inschriften, Reden, politischen Schriften und dergl. Dagegen scheinen die Autoren, die fur ihre Werke auf das gesamte griechische Publikum rechneten, die Schreibweise ΣΣ vorgezogen zu haben, da ΤΤ als Provinzialismus erschien’. Our speech (which he does not take into account) fits well into this. I use the terms ‘hellenize’, ‘Hellenic’ on the strength of Posidippus Comicus fr. 28 Kock, to which Jacoby drew my notice: a Thessalian, blamed for his bad Attic, claims the right of Greeks at large to ‘hellenize’.

^{page 28 note 2} Occasional turns of phrase (e.g. B 44 κα т ἒтι тούтων αἲσχιοον, § 22 κα т μέγισтον) are not very distinctive. [In view of the irony in § 22 fin. should we put a question mark after this fragment?]

^{page 28 note 3} In his Life of Kritias (tritae sophislarum, i. 16), § 4 = Vorsohr ⁵ 88 A 1.

^{page 28 note 4} ‘His attack is formidable when his tone is defensive:’ cf. §§ 1–4, 25–33. This is one facet of the Κριтίου γών (p. 29, with n. 3), if my insertion of <μή> in that passage be correct.

^{page 28 note 5} In Epist. 73 (= Vorsokr. ⁵ 82 B 35, 88 B 17) Philostratos makes a very similar point: Gorgias' grand manner was adapted (cf. p. 31, n. 4) by Thucydides and Kritias, Thucydides converting it into his own by ῥώμη, Kritias by εὐγλωттία. This is no doubt similar to the κωμικ 7egr;ὐγλωттία so much admired in Herodes (Philostr, . vit. soph. 2. 1. 14)Google Scholar: the racy idiom of natural speech, as seen e.g. in good dialogue in a novel. The ‘Herodes’ speech has, I believe, exactly this: an archaist might find it more telling than Gorgias because closer to real speech [asyndeton, ‘Kritias’ own beauty' is, in its simplest form, a mark of comic style: in our speech, §§ 6, 7, 8, 10,16, etc.]. To use κωμική for interpreting what Philostratos meant by Kritias' εὐγλωттία, is not of course to suggest that Kritias copied comedy: both copied life. The word no doubt suggested (to Philostratos) New Comedy rather than the fantastic vocabulary of Old: it is amusing to compare the closing words of § 31 (in a typical comic rhythm) with, e.g., Ar. Knights 878. [See below, on § 31.]

^{page 28 note 6} Aristides, , Ars rhet. 2. 15 = Vorsokr. ⁵ 88 B 46Google Scholar.

^{page 29 note 1} Diels adds <χωρίον>, I have added <μ>.

^{page 29 note 2} Dionysios uses ὢρα of Plato's style (ad Pomp. 2), Plutarch of Herodotos' (de Her. mal. 43).

^{page 29 note 3} I translate γών ‘power’ to give the contrast (which I take to be intended) with ‘beauty’: the means he uses to defeat his adversary, opposed to what he charms his hearers with.

^{page 29 note 4} Compare the taste of George Moore's dictum: ‘No writer ought ever to use either “which” or “that”’ (reported by Nicolson, Harold, Spectator, 17 01 1942)Google Scholar. What is valued is the immediacy of a paratactic style: not unlike what we admire in the narrative style of a ballad (e.g. Clerk Saunders), the stark unexplained juxtaposing of ideas or episodes.

^{page 29 note 5} It would make the flattish point that Kritias is unlike Isokrates; or it might imply an enfant terrible manner, such as Xenophon perhaps adumbrates, e.g. Hell. 2. 3.16 (I do not think this is what Philostratos means).

^{page 29 note 6} As Dionysios does, Thuc. 51: ‘orators like Andokides Antiphon Lysias, Sokratics like Kritias Antisthenes Xenophon’ write a vernacular and prove that Thucydides does not. This list is abridged in Lysias 2 to οῖ тε 'Ανδοκί7dgr;ουλòγοι κα οἱ Κριтίου κα ἂλλοι συχνοί (which likewise use the vernacular, opposed to the ‘Old Attic’ of Plato and Thucydides): λόγοι = literary compositions.

^{page 29 note 7} See the next note.

^{page 29 note 8} Hermogenes, , de ideis, 2. II (pp. 401–2 Rabe)Google Scholar. = Vorsohr? 88 A 19. Hermogenes' words μάλισтα ν тοῖ7sfgr; δημγορικοῖς προοιμίοι7sfgr; are th e only explicit reference which we have to Kritias' speeches. But on p. 403 Hermogenes expressly classes Kritias with the ‘Ten Orators’, as being comparable with Demosthenes rather than with Plato and as practising тò δικ7agr;νικόν and тòσυμβουλευтικόν. Dionysios had classed him with the Sokratics (see note 6, above). Herodes has come between: it looks as if Herodes' ‘publication’ (p. 20, n. 1) had converted Kritias from a Sokratic into an orator.

^{page 29 note 9} тà πολλà λέγων ποϕανтικς (so Rabe: Diels prints ποϕαтικ, but Hermogenes is simply repeating what he has just said of Antiphon, on p. 401, where the meaning is not i n doubt. In both places it is a form of ὂγκος). This ‘apophantic’ manner is illustrated in two ‘fragments’ of Kritias (Vorsokr.⁵ 88 F 46–7) in which the sophist Aristides compares some words of Xenophon with how Kritias would say the same thing. Blunt, downright, authoritative: not urbane like Xenophon or Lysias. In ps.-Lysias xx there are several indications which prove that §§ 1–10 are by a different speaker from §§ 11–36, among them the ‘apophantic’ tone of §§ 1–10. In our speech the tone is tempered to the speaker's youth, yet e.g. §§ 1, 3, 31, 33, are as ‘apophantic’ as Antiphon.

^{page 30 note 1} Cf. the χωρίον in Xen, . Hell. 7. 5. 11Google Scholar.

^{page 31 note 1} Enough to cite Thuc. 3. 82–4 ; Xen, . Hell. 2. 4. 22Google Scholar.

^{page 31 note 2} Cf. Kritias' proclamation εἲ тις тòν ᾈθηναῖον ϕεύγονтα δέξοιтο (Philostr, . Vit. soph. i. 16. I = Vorsokr.⁵ 88 A 1)Google Scholar.

^{page 31 note 3} Sc. π7rgr;οσκόν σтι, cf. e.g. Plato, , Theaet. 196Google Scholar E. The present tense suits the generalizing participle.

^{page 31 note 4} Both perhaps are adaptations (p. 26, n. 4) of the weighty ‘tetrads’ in Gorgias' Epitaphios (Vorsokr. ⁵ 82 B 6): θε7rgr;άπονтες μέν… κολασтαì δέ… αὐθάδεις… εὐόργηтοι:ὑβρισтαί… κόσμιοι… ἂϕοβοι… δεινοί: οὒтε μϕύтου ἂρεος οὒтε νομίμων ρώтων οὒтε νοπλίον ἒρδος οὒтε ϕιλοκάλου εἰρήνης: σεμνοì μέν… ὂσιοι δέ… δίκαιοι… εὐσεβεῖς δέ.

^{page 31 note 5} Ωσπ7egr;ρ οὖν makes explicit something which so far is only implied, e.g. by the protasis of a conditional sentence. See Denniston, , Greek Parti cles, pp. 421–2Google Scholar. The presence of a Macedonian governor has been implied by the contrasted absence of Spartan governors.

^{page 31 note 6} The writer likes to associate εὑρίσκω and ρ §§ 3, 26: the distinction, if any, is fine.

^{page 32 note 1} See p. 28, n. 5. The metre (iambic tetrameter catalectic) is no doubt fortuitous: the words hardly look like a quotation from Old Comedy (cf. e.g. , Ar.Knights, 878)Google Scholar. But the words might be a sort of proverb, since it is a popular and proverbial, as well as a comic, metre: e.g. οù ϕρονтἹπποκλείδῃ, ἂρισтα χωλòς οἰϕεῖ, etc.

^{page 32 note 2} The temporal sense is precluded by πέϕυυκενand by the following γάρ.

^{page 33 note 1} So Plut, . Alcib. 38. 5Google Scholar; Nepos, , Alcib. 10. 1Google Scholar: no doubt all accounts of the murder were tendentious, more or less. Ferguson, , CAH v. 365Google Scholar, in his brilliant account of these events speaks as if Alkibiades had let Kritias down whilst he (Alkibiades) was still in power: ‘his leader either could not or did not protect him—an omission which Critias remembered later.’ But Kritias' poem to Alkibiades (Vorsokr. ¹ 88 B 4–5) was surely written when Alkibiades was actually in Athens, in 407; and it was not written from exile. Alkibiades' own fall is so soon after this that it is unlikely tha t Kritias' exile came first: Kritias in fact was involved in Alkibiades' fall. It may be just this which he could not forgive. I imagin e Kritias was one of those who had urged Alkibiades to attempt a coup d'état before he left Athens on his last campaign (Plut, . Alcib. 34. 7)Google Scholar: without doubt Kleophon would take the first opportunity he could to ruin either of them. von, A. Blumenthal in his essay Der Tyrann Kritias (Stuttgart, 1923), p. 7Google Scholar, quite misconceives the order of events: ‘wird verbannt, kehrt wieder, wirkt fur Alkibiades’.

^{page 33 note 2} тῷ μè οὖν πρώтῳ χρόνῳ Κριтίας тῷ Θηραμένει μογνώμων тε καì ϕίλος ἦν (Xen, . Hell. 2. 3. 15)Google Scholar.