¹ See also B.P.W. 1899, col. 636, No. 9; R.E.G. XII, 1899, p. 172, No. IX. Though from the same source, the texts differ in some respects: L. 2, Στρ[α]τὴ R.E.G.; Στρ[α]τὼ B.P.W. L. 5, ΠΛEIONKE R.E.G.; πλειονηε- B.P.W.

² The texts published by Papageorgiou in Ἀθηνᾶ and B.P.W. and by Contoléon in R.E.G. were revised by Pelekides in Ἀρχ. Δελτ. 8, 1923 (1925), pp. 265 ff.Google Scholar According to Pelekides stone Γ fits exactly to B, so that Papageorgiou's No. 11 is the continuation of No. 8. Pelekides adds also (p. 267) a new fragment of some interest. My friend Mr. J. M. R. Cormack of the University of Reading, England, to whom I am indebted for information on the inscriptions of Edessa, visited the Λόγγς in 1936 and was able to revise the texts of A, which then stood in the bed of a stream partly submerged by flood water; stones B and Γ were not to be found. Another dedication (Μητρὶ θεῶν) from Edessa is published in Dareste, Recueil des Inscriptions jur. gr. II, p. 249, No. 13. Similar documents are known from Beroea (A.B.S. XVIII, 1911–1912, p. 139, No. 2; Ἀρχ. Δελτ. II, pp. 145 and 147) and from Scydra (Dareste, op. cit., p. 250, No. 14; B.C.H. XLVII, 1923, p. 182). The Edessa inscriptions are reprinted and discussed in W. Baege, De Macedonum Sacris, 1913, pp. 113 ff.

³ Preisigke, Wörterbuch, s.v. ἐλάσσων.

⁴ Ibid., Abschnitt 23.

⁵ The references are to the numbers in Ἀθηνᾶ, loc. cit.

⁶ , Contoléon, R.E.G., loc. cit., p. 170;Google Scholar, Dareste, op. cit., p. 254.Google Scholar Cf. B.G.U. I, 96, quoted by , Weiss, Gr. Privatrecht, p. 353, n. 350; B.C.H. XXIII, p. 192, No. 59 (αὐτεξούσιος τέκνων δικαίῳ)Google Scholar.

⁷ , Mitteis, Reichsr. und Volksr., p. 389.Google Scholar

⁸ , Weiss, op. cit., p. 302;Google Scholar, Dareste, op. cit., pp. 261, 283Google Scholar.

⁹ Cf. Weiss, p. 303, n. 202. For χειρόγραϕον see the documents of Amphissa quoted in B.S.A. XVIII, 1911–1912, p. 143, n. 3.

¹⁰ , Weiss, op. cit., p. 305.Google Scholar Professor Nock points out that the intrusion of civil law does not involve the complete abrogation of religious authority; legal and religious sanctions may exist side by side as at Spoleto (Dessau 4911).

¹¹ O.G.I. 383, 11. 171 ff.

¹² La Manomissione, pp. 97, 101; , Weiss, op. cit., p. 298Google Scholar.

¹³ Cf. the mention of payment in earlier dedicatory manumissions, , Weiss, op. cit., p. 299, n. 192Google Scholar.

¹⁴ , Weiss, op. cit., pp. 303,Google Scholar n. 202, and 358 f.

¹⁵ Cf. the δοῦλος of the θεὰ Συρία in B.C.H. XXI, p. 59, No. 68 and the servi Venerii in Cic. Verr. III, 50, etc. It is possible that the terms liberta Veneris Erycinae (Cic. Div. in Q. Caec. 55 f.) and libertus numinis Aesculapi (C.I.L. III, 1079)Google Scholar refer to such slaves after manumission by the god, but in dealing with the Greek example, ἀπελ (ενθέρῳ) τοῦ με[γίστου θεο]ῦ Σαράπιδος (B.G.U. 7, 1564), Viereck and Zucker take it as designating a person manumitted by the sacral process. The use of ἀπελεύθερος in that sense would presumably be due to the influence of the real status on the terminology; the term δοῦλος τοῦ θεοῦ is more in keeping with the fiction of sale or dedication to the god.

¹⁶ , Mitteis, op. cit., p. 387.Google Scholar

¹⁷ , Calderini, op. cit., p. 270.Google Scholar

¹⁸ Edessa: 5 (?), 9 (κατὰ χρηματισμόν); Lairbenos: M.A.M.A. IV, 276A (III) (κατὰ θεῖον ὄνιρον), 276B (κατ᾽ ὄνιρον), 277A (II) (κατὰ τὴν ἐπιταγὴν [τοῦ] θεοῦ), , Oppermann, loc. cit, Nos. 1Google Scholar (κατὰ ὄνιρον?), 8, 11. Calderini (p. 105) and Weiss (p. 305) have mistaken the meaning of κατὰ τὴν ἐπιταγήν. At Edessa, according to Oppermann (col. 525), “Freilassung auf Befehl der Gottheit findet sich nicht,” but κατὰ χρηματισμóν is best taken in this sense; cf.S.I.G. III, 1173. At the Lairbenos shrine the dream seems to have been the common form of revelation of the divine will and χρηματισμóς is to be interpreted in the same way; cf. Matthew, 2, 12: χρηματισθέντες κατ᾽ ὄναρ; (, Powell, Coll. Alex., p. 69):Google Scholar ἐχρημάτισεν κατὰ τòν ὔπνον.

¹⁹ Cf. Sardis VII, Greek and Latin Inscriptions, No. 2.

²⁰ , Mitteis, Grundzüge, p. 135 ff.;Google Scholar, Weiss, op. cit., p. 241Google Scholar.

²¹ For καταγράϕειν see , Mitteis, Grundz., p. 176;Google Scholar, Weiss, op. cit., pp. 267 ff.Google Scholar with notes 79 ff., p. 305, n. 210 and p. 405, n. 174; for its use in the defixiones, Björck, Der Fluch des Christen Sabinus, p. 121. For δωρεῖσθαι and χαρίζεσθαι in foundation records, see Laum, Stiftungen, I, p. 122.

²² , Baege, op. cit., p. 91;Google Scholar, Laum, op. cit., II, p. 41,Google Scholar No. 39. For the gift of vineyards to establish a foundation see also Laum, op. cit., II, Nos. 175A, 175B, 201.

²³ R.-E. VIII, col. 1462, s.v. Hieroduli; , Cardinali, Rendic. Accad. dei Linc., XVII, 1908, p. 175,Google Scholar n. 1, does not commit himself.

²⁴ Op. cit., p. 104, n. 11.

²⁵ Cf. R.-E. VIII, col. 1467; , Cardinali, op. cit., p. 175Google Scholar.

²⁶ , Mitteis, Grundz., p. 272 f.Google Scholar

²⁷ Pfister, R.-E., Suppl. IV, col. 277 ff., passim.

²⁸ Mayser, Gr. d. gr. Pap. II. 1, p. 63 f.; , Radermacher, Neut. Gr. (1911), p. 147 (B.G.U. III, 873, 5: ὁμολογῶ ἐσχηκέναι με)Google Scholar.

²⁹ This is hardly an impersonal passive use; it is an example of the incorrect use of the middle, which is sometimes a literary attempt at Atticism (cf. Jannaris, Hist. Gr. Gr., § 1486; , Hatzidakis, Einleitung, pp. 194 ff.)Google Scholar.

³⁰ , Hatzidakis, op. cit., p. 321;Google Scholar, Thumb, Handb., p. 11Google Scholar.

³¹ , Jannaris, op. cit., § 961;Google Scholar, Psaltes, Grammatik, p. 237,Google Scholar § 352.

³² Cf. n. 57.

³³ , Jannaris, op. cit., § 1498.Google Scholar

³⁴ , Jannaris, op. cit., p. 573,Google Scholar § 13; , Ljungvik, Beitr. z. Syn. d. spätgr. Volksspr., p. 93Google Scholar (ὁμολογῶ); , Hatzidakis, op. cit., p. 215Google Scholar (Gegenwirkung des Atticismus); the use is particularly striking here, since direct speech is common in epiphany description even in literature.

³⁵ , Ljungvik, op. cit., p. 94.Google Scholar

³⁶ , Ljungvik, op. cit., pp. 87 ff.Google Scholar

³⁷ , Ljungvik, op. cit., pp. 54 ff.Google Scholar

³⁸ The stylistic features are the most important for this conclusion, along with the evidence of an attempt at literary usage; poetical elements are not always by themselves proof of literary pretensions (v. Mayser, Grammatik (1906), pp.24 ff.). In our text the combination of pseudo-learned forms, poetical forms and literary style proves clearly that we are dealing with a conscious artistic effort and not merely with popular speech infected by the assimilation of stray poetical elements.

³⁹ Cf. C.R. XLVII, 1, p. 7.

⁴⁰ , Mitteis, Grundz., pp. 72 ff.;Google Scholar, Pettazoni, La Confess. dei Peccati II, 3, p. 89Google Scholar.

⁴¹ , Steinleitner, Die Beicht, p. 96, n. 3Google Scholar (Modern Greek κóλασις = hell).

⁴² , Pfister, loc. cit., col. 280.Google Scholar

⁴³ , Steinleitner, op. cit., p. 113,Google Scholar n. 1; P. Oxy. 1066, col. III.

⁴⁴ , Pettazoni, op. cit., p. 195.Google Scholar

⁴⁵ Op. cit., pp. 13; 37, n. 87; 61; 68; 95; 97; 137, n. 51.

⁴⁶ The term is mostly, but not always, used of testamentary manumission; Philologus, LXXXVI, 1931, p. 126Google Scholar.

⁴⁷ , Weiss, op. cit., p. 297,Google Scholar n. 188; cf. ἐλενθέρωσις, P. M. Meyer, Jur. Pap. 6. For ὠνή, , Weiss, op. cit., p. 303, n. 202Google Scholar.

⁴⁸ B.G.U., 326, ii, 11; cf. Suicerus, Thes. Eccl., s.v.

⁴⁹ For καταγράϕειν here, cf. , Weiss, op. cit., p. 305Google Scholar and n. 210.

⁵⁰ See above, n. 14.

⁵¹ , Mitteis, Grundz., p. 271 ff.;Google Scholar see below, n. 88 and n. 90.

⁵² , Lipsius, Att. R., p. 520,Google Scholar n. 2; cf. [Dem.] πρòς Max. § 15: κύριον ἐπιγραψάμενος τòν ἀδελϕòν τòν ἑαυτοῦ.

⁵³ Quoted by , Wenger, Ὅροι Ἀσυλίας, Philologus LXXXVI, 1931, p. 441.Google Scholar

⁵⁴ , Kühner, Gr. Gr. II, i, p. 348,Google Scholar Anm. 5; , Poultney, The Syntax of the Gen. Case in Aristoph. (1936), p. 86Google Scholar.

⁵⁵ , Schulze, Kl. Schr., pp. 160Google Scholar ff., especially p. 162, n. 4.

⁵⁶ The stone is worn at the end of the line, but the photograph seems to confirm ἐμό[ν] and exclude ἐμέ; a further examination of stone or impression would be necessary to put this beyond all doubt. Ἐμέ, in any case, seems stylistically less probable.

⁵⁷ The confusion might have been assisted by the facts that ἐμός was a dying form (, Mayser, op. cit. 1, p. 304)Google Scholar and that for popular speech ἑαντοῦ might also have been ambiguous (ibid., pp. 303 f.).

⁵⁸ For another such formula, see , Diod. Sic. XV, 52, 3:Google Scholar καθάπερ ἦν εἰθισμένον, ἀνηγόρευε κηρύττων μήτ᾽ ἐξάγειν ϴήβηθεν μήτ᾽ ἀϕανίζειν, ἀλλ᾽ ἀπάγοντα πάλιν ἀνασώζειν.

⁵⁹ This use is well illustrated by Holden on Plut. Them., pp. 143, 209.

⁶⁰ E.g. , Menander, Kock III, p. 164;Google Scholar Plut. de Superst., 168d; Tibullus, 1, 3, 30 (below). Cf. Martial, IV, 30, 13.

⁶¹ , Latte, op. cit., pp. 106 f.;Google Scholar, Stengel, Die gr. Kultusalt., p. 80Google Scholar.

⁶² Od. 7, 160.

⁶³ E.g. Thuc. I, 24; 126, 10, 11; 136, 3; 137, 1; III, 28, 2; 70, 5; 75, 3, 5.

⁶⁴ S.I.G. 736, 83.

⁶⁵ , Weiss, op. cit., p. 532 ff.Google Scholar

⁶⁶ E.g. , Schwyzer, Dial. gr. ex. 179, 1, 39 (ναεύειν).Google Scholar

⁶⁷ Urk. der Ptolemäerz., 121, 1. 12 (above).

⁶⁸ Suppl. 268.

⁶⁹ II, 13 (μέχρι ἐμεῦ).

⁷⁰ Knights 1311 ff. Frg. 567 (Pollux 7, 13); cf. frg. 458; Eupolis apud Poll. 7, 13; Plut. Theseus XXXVI.

⁷¹ Paus. II, xiii, 3.

⁷² Diodorus XI, 89, 6.

⁷³ , Chapot, La Prov. rom. d'Asie, pp. 406 ff.Google Scholar

⁷⁴ Verr. 1, 33, 85.

⁷⁵ Ann. III, 60.

⁷⁶ Just. Inst. 1, 8, 2.

⁷⁷ E.g. Od. 7, 160; Ap. Rhod. IV, 693; Plut. Them. XXIV; Herod. 1, 35.

⁷⁸ There is a curious passage relating to the sanctity of gates and doors in Porphyrius, De Antro Nymph. § 27, where it is said: οἶδε δὲ καὶ Ὅμηρος ἱερὰς τὰς θύρας, ὡς δηλοῖ παρ᾽ αὐτῷ ὁ σείων Οἰνεὺς ἀνθ᾽ ἱκετηρίας ‘σείων κολλητὰς σανίδας, γουνούμενος υἱόν’ (Il. IX, 581 ff.). This is no doubt a misinterpretation of Homer, but it may imply familiarity with the idea of supplication at the gate.

⁷⁹ Od. I, 119.

⁸⁰ E.g. Plato, Symp. 203b; Rep. 364b; Luke 16, 21; Arist. Rhet. 1391 all.

⁸¹ , Plato, Phaedr. 245a.Google Scholar

⁸² , Aristoph. Clouds 469.Google Scholar

⁸³ Herod. III, 40; Esther (LXX), 4, 1.

⁸⁴ Cf. Tibullus, 1, 2, 85 (below); , Bingham, The Ant. of the Chr. Ch. (ed. 1856), p. 1059Google Scholar.

⁸⁵ “Das limen des Tempels ist nach alten Glauben beider Völker ein sakraler Begriff” (Norden ad loc). So also the gates of cities are sanctae res and quodam modo divini iuris (, Gaius, Inst. 2, 8)Google Scholar.

⁸⁶ The common use of πύλαι in LXX to denote the gate as the place where justice is administered does not seem relevant to our passage.

⁸⁷ Or without a significant adjective, as perhaps in the phrase σιδηρᾶς θύρας ῥήξω, to which Professor Nock draws my attention (Proc. Brit. Acad. XVII, 1931, p. 251, 1. 12 with the note on p. 263); cf. Kroll, op. cit., passim.

⁸⁸ Cf. , Weiss, op. cit., pp. 442 ff.Google Scholar and notes. Weiss (ibid., n. 46) quotes a case somewhat similar to ours: ἐπανανκάσαι με μετὰ ὑβρέων καὶ πληγῶν ἐγδόσθαι γράμματα χειρογράϕου πράσεως κτλ.

⁸⁹ Cf. , Buckland, Roman Law of Slavery, p. 593Google Scholar (vis and metus).

⁹⁰ The revocability of manumission presents no difficulty in Greek usage; cf. Wyse on Isaeus, p. 379; , Latte, op. cit., p. 101.Google Scholar That the manumission should be revoked by the recall and destruction of a document seems also in keeping with Greek custom; cf. , Weiss, op. cit., p. 443Google Scholar.

⁹¹ For excuses in the confession inscriptions, cf. M.A.M.A. IV, 285 (λημóνησα); 287 (?); 288 (ἔλαθε με).

⁹² A similar appeal to public opinion plays a part in other epigraphic religious documents; cf. , Björck, op. cit., p. 44Google Scholar.

⁹³ , Thumb, Hellen., pp. 144 ff.Google Scholar

⁹⁴ , Psaltes, op. cit., p. 245; Mayser (1906), p. 353.Google Scholar

⁹⁵ Kühner, II, 1, pp. 38 ff.; Stahl, pp. 144 ff.; Jannaris, § 2140; Schmid, Att. III, pp. 112 ff.; Psaltes, p. 230.

⁹⁶ Inc. Scr. Byz. de re mil. (Teubner, 1901); cf. also p. 7,Google Scholar 1. 12 (πετυκνωμέναι οτηκέτωσαν), p. 11, 1. 21 (οτηκέτωσαν ἀπέχοντα).

⁹⁷ In modern Greek στέκω supplies missing forms of εἶμαι, , Thumb, Handb., p. 152Google Scholar.

⁹⁸ Aves 513; cf. ἐκάθητο in 1. 510. See ibid. 515,831; Eccles. 782. The word ἔστηκα in Eccles. 879 (cf. Pax 256) is to be taken similarly and is far too weak to bear the weight of Fraenkel's conclusions as to the mise en scène (Greek Poetry and Life, pp. 262 ff.).

⁹⁹ Schmid, III, p. 114 (Eleganz).

¹⁰⁰ In the classical period εἶναι with the aorist participle appears to be poetic, but it is common in later prose; see the examples in Schmid, loc. cit., and Psaltes, p. 230.

¹⁰¹ Professor , Löfstedt refers me to Stud. Neophil. XI, p. 185, n. 1.Google Scholar