^{page 5 note 1} Zenob. 6. 5 Δωδωναι̑ον χαλκει̑ον...ἐπὶ τω̑ν πολλὰ λαλούντων καὶ μὴ διαλειπόντων. Diogen. 8. 32 τδ Δωδωναι̑ον χαλκει̑ον· ἐπὶ τω̑ν πολλὰ λαλοὺντων. Greg. Cypr. cod. Mosq. 2. 81 Δωδωναι̑ον χαλκει̑ον· ἐπὶ λάλουΔωδωναι̑ον χαλκει̑ον· ἐπὶ λάλουΔωδωναι̑ον χαλκει̑ον· ἐπὶ λάλουΔωδωναι̑ον χαλκει̑ον· ἐπὶ λάλου Macar. 3. 42 Δωδωναι̑ον χαλκει̑ον· ἐπὶ τω̑ν ἀδολἐσχων καὶ φλυαρω̑ν Apostol. 6. 43 Δωδωναι̑ον χαλκει̑ον· ἐπὶ τω̑ν ἀδολἐσχων καὶ πολλὰ λαλούντων (C. om. καὶ πολλὰ). Suid. Δωδωναι̑ον χαλκει̑ον· ἐπὶ τω̑ν μικρολογούτων (Portus corr. μακρολογούντων). Steph. Byz. s.v. Δωδώνη, Δωδωναι̑ον χαλκει̑ον...ἐπὶ τω̑νπολλὰ·λαλούντων. Eustath. Il. B. 750, Δωδωναι̑ον χαλκει̑ον...ἐπὶ τω̑ν πολλὰ λαλούντων Schol. Philostr. ab Osanno in Auct. Lex. Gr. 14 editus ἐπὶ τω̑ν πολυλόγων, τὸπλέον του̑ ἐν Δωδώνη̢ χαλκείου λέγειν

^{page 5 note 2} Quoted by Steph. Byz. s.v. Δωδώνη. Cp. Zenob. 6. 5 Δωδώναι̑ον χαλκει̑ον· κει̑ται παρὰ Μενάνδρψ ἐν τψ̑ ᾿Αρ᾿ῥηφόρψ (τη̢̑ ᾿Αρ᾿ῥηφόρψ Meineke), Suid. s.v. Δωδωναι̑ον χαλκει̑ον... κέχρηται δὲ τη̢̑ παροιμία̢ Μένανδρος Αὐλητρίδι (Αὐλητρίδι V. C). It appears from Athen. 10. 442D, 13. 559C, that the Αὐλητρίς was another name of the ᾿Αρρηφόρυς. For the text of the fragment I follow Meineke (fr. 3).

^{page 5 note 3} Quoted by Steph. Byz. s.v. Δωδώνη where the MSS. give τὸν ἐν Δωδω̑νι λεγόμενον ον᾿ν ἐκάχαλκυν ἤγειρον Bentley cj. τὸν ἐν Δωδω̑νι λελεγμένον οὕνεκα χαλκὸν | ἤγειρον. Schneider, Call. ii. 526Google Scholar notes that the first line is quoted by Choeroboscus, dictat. in Theodos. p. 418, 18Google Scholar, cp. ib. 104, 5; 116, 4; 273, 35; 290, 30, Bekk, . anecd. p. 1228Google Scholar, Lascar, . gram. p. 112AGoogle Scholar, and is thus enabled to restore the true reading μή με τὸν ἐν Δωδω̑νι λέγοι μόνον οὕνεκα χαλκὸν | ἤγειρον He follows Hecker in supposing that the frag. occurred in the Prologue to the Aetia, and that the poet meant ‘non omnia se quae a Musis aceeperit profusurum, ne garrulum tantum se vocet quisquam.’

^{page 5 note 4} Arist, ὑπὲρ τω̑ν τεττάρων 309 ἐπειδὰν δὲκακω̑ς τινας εἰπει̑ν δέη̙ καὶ καταβαλει̑ν, τψ̑ Δωδωναίῳ μὲν οὐκ ἂ εἰκἀσαις αὐτους χαλκείῳ.

^{page 6 note 1} Procop, . ep. 99Google Scholar, p. 269 Mai ἡμϵῖς καὶ Δωδώνης χαλκϵῖον γϵγόναμϵν.

^{page 6 note 2} Cram., an. 3Google Scholar, 225, 11 τὸ ἐν Δωδώνη̙ χαλκει̑ον ὑπερηχει̑ς.

^{page 6 note 3} Carapanos prints: ὡς ὁ μὲν Δήμων φησὶν ῾ἀπὸ τὸν ναὸν Δωδωναίου Διὸς τοίχους μὴ ἔχοντα, ἀλλὰ—ἑκάστῳ καὶ διαμένειν τὸν ἠ̑χον ἄν τις του̑ ἑνὸς ἐφάψεται.’ But this is doubly ungrammatical.

^{page 6 note 4} Müller, F.H.G. iii. 125Google Scholar reads ἀλλήλους with schol. Ven. ap.. Bekk. Bernhardy keeps ἀλλήλοις with A.B.V.C. edd. vett. and schol. Ven. ap. Villoison. ἀλλήλων E. Müller also reads διὰ διαδοχη̑ς and γίγνεσθαι. ψυχη̑ς for ἠχη̑ς is found in A. schol. Ven. ap. Villoison.

^{page 6 note 5} ὅταν δὲ A. Leutsch would read ὡ̑ν ὅταντις.

^{page 6 note 6} A MS. of the fourteenth century containing a compendium of Suidas amplified by important additions ‘imprimis ad proverbia’ (Bast). Schneidewin thought that the author, so far as his proverbs are concerned, was founding upon Aristophanes of Byzantium.

^{page 6 note 7} Presumably Pausanias the lexicographer, a contemporary of Galen, to whom Eustathios was much indebted (Christ, W.Gr. Lit. ³ p. 765Google Scholar).

^{page 7 note 1} Suid. s.v. Φιλόχορος mentions his work πρὸς τὴν Δήμωνος ᾿Ατθίδα and Harpocrat. s.v. Ηετιωνία cites Φ. ἐν τη̢̑ πρὸς Δήμωνα ἀντιγραφη̢̑.

^{page 7 note 2} Dion., Hal., de Din. 3Google Scholar: see Susemihl, A.L.G. i. 595Google Scholar.

^{page 7 note 3} Analogous examples are not wanting. According to Paus. x. 5, 9 the original temple of Apollo at Delphi was made of laurelboughs, and the second temple of wax and feathers (see Frazer ad loc.). Reisch in Pauly-Wissowa i. 1669, 37 suggests that the Κϵρατών of Delian inscrr. was a large platform connected with the famous Κϵράτινος βωμός, which was made from the horns of goats sacrificed to Apollo. Paus. v. 10; 4 states that ‘a gilt caldron is set on each extremity of the roof of the temple at Olympia.’

^{page 7 note 4} Cp. a well-known experiment with two or more tuning-forks.

^{page 7 note 5} See the passages cited in full below.

^{page 8 note 1} Call. h. Del. 286 γηλεχέες θεράποντες ἀσιγήτοιο λέβητος. This, however, may refer to the later and more elaborate λέβης see below.

^{page 8 note 2} Müller, F.H.G. iv. 326Google Scholar reads δύω.

^{page 8 note 3} μάστιγας C.

^{page 8 note 4} Müller, F.H.G. iv. 326Google Scholar reads ὄντας καὶ σϵιομένους.

^{page 8 note 5} Αὐλητρίσι V.C.

^{page 8 note 6} Δήμονα A.

^{page 9 note 1} He was one of the earliest writers of Ἀτθίδϵς. The order, as given by Schwartz loc. cit., is Kleidemos or Kleitodemos, Androtion, Demon, Philochoros, etc. And Christ, W.Gr. Lit. ³ pp. 480, 553Google Scholar, argues that the older Ἀτθίδϵς were one of the main sources of Aristotle's Ἀθ. Πολ. If so, Aristotle may well have had Demon's works before him.

^{page 9 note 2} See further Schneidewin in Par. Gr. i. p. ii f. He concludes: ‘nostrorum Paroemiographorum auctores quin usurpaverint librum Aristotelis non dubito.’

^{page 9 note 3} Read τὴν β´

^{page 9 note 4} Read ὅταν οὖν.

^{page 9 note 5} Delete commas after χαλκοῦς and ἱμᾶσιν.

^{page 9 note 6} Vulg. τψ̑ ᾿Αρηφόρψ Meineke restored τη̢̑ ᾿Αρ᾿ῥηφόρῳ

^{page 9 note 7} Cod. B is a Bodleian MS., Cod. V a Vatican MS. See Schneidewin Par. Gr. i. p. xxx f.

^{page 9 note 8} Cod. A is the Paris MS. of Arsenius (Par. 3058). See Leutsch Par. Gr. ii. p. xiii ff.

^{page 10 note 1} Read καὶ ῾κατὰ μέντοι τοὺς ἡμετέρους, φησὶν ὁ Ταρραι̑ος, ἡ μὲν λαβὴ τη̑ς μάστιγος< >, οἱ δὲ ἱμάντες <ἀπὸ> πεπτώκασιν. κ.τ.λ.᾿ Preller, Polemon p. 61Google Scholar cj. ἕως ἂνὁ ἂνεμος διαμένη̙ τη̑ς μάστιγος ascribing these words to Aristeides. Sehneidewin Par. Gr. i. p. xiv. rightly criticises this and suggests κατὰ μέντοι τοὺς ἡμετέρους, φ. ὁ Τ., ἔστι μὲν λαβὴ τη̑ς μάστιγος, οἱ δὲ ἱμάντες ἀποπεπτώκασιν. κ.τ.λ. Carapanos prints ῾καὶ κατὰ μέντοι τοὺς ἡμετέρους (χρόνους), φ. ὁ Τ., ἡ μὲν λαβὴ τη̂ς μάστιγος (διασέσωσται), οἱ δὲ ἱμάντες ἀποπεπτώκαιν.’

^{page 10 note 2} MSS. περὶ, for which Preller restored παρὰ.

^{page 10 note 3} Read ἐπιχωρίων with MSS. and Schneidewin.

^{page 10 note 4} Read ἤχει with Schneidewin.

^{page 10 note 5} Read περιεγένετο with Schneidewin, who adds τὸ χαλκει̑ον unnecessarily.

^{page 10 note 6} Bentley restored ἢ for. ἢν and καλῇ for καλϵῖ. Carapanos prints ἢ τίτθην καλϵῖ.

^{page 10 note 7} Bentley inserted πέρας <οὐ> ποι̂ει

^{page 10 note 8} Meineke, τις for τι.

^{page 10 note 9} Meineke, παράψηθ’ for παρήψαθ’.

^{page 10 note 10} Meineke, καταπαύσαι for καταπαυ̂σαι

^{page 10 note 11} Par. Gr. i. p. xiii.

^{page 10 note 12} Suid. s.v. Ζηνόβιος says ἔγροψϵν ἐπιτομὴν τῶν παροιμιῶν Διδύμου καὶ Ταῤῥαίου κ.τ.λ. Cp. Schol. Ar. nub. 134 Ζηνόδοτος (Herm. corr. Ζηνόβιος) ὁ τὰς Ταῤῥαίου καὶ Διδύμου παροιμίας ἐπιτϵμών. See further Par. Gr. i. p. xxiv. f.

^{page 10 note 13} Par. Gr. i. p. xxv f.

^{page 11 note 1} Par. Gr. i. p. xxx f.

^{page 11 note 2} Sitzb. d. bay. Ak. 1892, p. 644.

^{page 11 note 3} Anth. Pal. ix. 572.

^{page 11 note 4} E.g. by Müller, . F.H.G. iv. 326Google Scholar.

^{page 11 note 5} W. Schmid in Pauly-Wissowa ii. 886, 42.

^{page 11 note 6} See the quotation given above, p. 5.

^{page 11 note 7} Athen. 13. 567 B, cp. Diog. Laert. vii. 188.

^{page 11 note 8} Plut., quaestt. symp. v. 2Google Scholar, 675 B.

^{page 11 note 9} Suid., s.v. Πολέμων, cp. Athen. 13Google Scholar. 574 C.

^{page 11 note 10} Strab. ix. 1. 16.

^{page 11 note 11} Wescher-Foucart, inscr. de Delphes n. 18, 260Google Scholar.

^{page 12 note 1} MS. προσ̣ελθει̑ν Kramer προέλθοι

^{page 12 note 2} Strab. iv. 6. 9.

^{page 12 note 3} Strab. xvii. 3. 7. See Christ, W.Gr. Lit. ³ p. 684Google Scholar n. 3.

^{page 12 note 4} This Apollodoros was Strabo's main source for the geography of Greece (W. Christ ib. p. 684). Schwartz in Pauly-Wissowa i. 2867 ff. enumerates the passages in which Strabo is indebted to him: they include several from bk. 7, but not frag. 3. Nevertheless it is probable that frag. 3 had the same origin. For where Strabo is dealing with places mentioned in the Homeric Catalogue he constantly cites Apollodoros' great work Περὶ νεω̑ν καταλόγου (see Niese, in Rh. Mus. xxxii p. 267 ff.)Google Scholar; and Strabo in frag. 3 is describing Dodona, , which the Catalogue mentions as Δωδώνην δυσχείμεροχ (Il. 2. 750Google Scholar).

^{page 12 note 5} Strabo's wording is a little ambiguous. I assume that ἀνάθημα Κορκυραίων is in apposition to μάστιγα χαλκη̑ν (rather than to ἀνδριάντα or to χαλκει̑ον), partly because μάστιγα is the nearest substantive, partly because the proverb was ἡ Κ ερκυραίων μάστιξ.

^{page 12 note 6} Carapanos Dodone et ses ruines Plate xxiii., nos. 1–2. The λέβης is inscribed in punctured letters ΦΙΛΟΚΛΕΔΑΟΔΑΜΟΦΙΛΟΥΛΕΥΚΑΔΙΟΣΔΙΝΑΙΟΙ The stand is inscribed

^{page 13 note 1} Rich s.v. ‘flagrum,’ Dar.-Sagl. ii. 1155, Fig. 3092.

^{page 13 note 2} Winckelmann, Monum. inedit. i. 8Google Scholar p. 8.

^{page 13 note 3} Dodone et ses ruines p. 168 f.

^{page 13 note 4} ib. Plate iii., no. 8.

^{page 13 note 5} As Potter on Clem. Al. loc. cit. observes, ‘solet…Theodoretus Clementem compilare.’

^{page 14 note 1} Ahrens, Buc. Gr. ii. 103Google Scholar, 9 ff. τὸ χαλκεῖον ὠς τάχος ἄχος ἄχει τὸν γὰρ δὲ κεῖον ὠς τάχος ἄχος ἄχει τὸν γὰρ δὲ Gen^b.) χαλκὸν ἐπήφον (so Reinesius, ἐπῆδον 4. 5. Can. Gen^b. Vulc., ἐπῆδον vulg., ἔπλησσον Moriz Schmidt, ἔπαιον Coraes, ἤπειγον Kiessling, ἐπῇδον = ‘fecerunt accinere’ Heyne, ἐπήχουν Hemsterhuis, ἐπῇρον Ahrens) ἐν ταῖς ἐκλειψεσι (so 4. 5. Gen^b., ἐλλείψεσι vulg.) τῆς σελήνησ καὶ ἐν ἐπι Jahn) τοῖς κατοιχομένοις κατηχου μένοισ 4. 5.), ἑπειδὴ ἐνομίζετο καθαρὸς εῖναι καὶ ἀπελαστικὸς ἀποπελαστικὸσ 5.) τῶν μιασμάτων διόπερ πρὸς πᾶσαν ᾶφοσίωσινκαὶ ἀποκάθαρσιν αὐτῷ αὐτὸ Gen^b.) ἐχρῶντο ὤς φησι ὤς φησι om. 4. Gen^b.) καὶ Απολλόδωρος ἐν τῷ περὶ θεῶν (Apollod., fr. 36Google Scholar).—τὸ ἄχει ὠς τάχος ἤχει Gen^b.) ἀντὶ τοῦ ψόφει κροῦε κρούων 5.). ἐπεὶ ὀ τοῦ χαλκοῦ ἠχος οἰκεῖοσ Gen^b.) τοισ κατοιχομένοις κατηχουμένοισ 5.), φησίν Απολ λόδωρος Αθήνησι τὸν ίεροφάντην τῆς Κόρης ἐπι καλουμένης ἐπικρούειν ἐπικρούσειν Gen^b., κπού ειν Dübner) τὸ καλούμενον ἠχεῖον οἰκεῖον 5. Gen^b.)· καὶ παρὰ περι 5. Can. Vulc.) Λάκωσι Λάκωνοσ 5. Can. Gen^b.) βασιλέως ἀποθανανότοσ εἰώθασι κρούειν λέβητα

^{page 14 note 2} Tib. i. 8, 21 f., Ov., met. iv. 333Google Scholar, Liv. xxvi. 5, 9, Tac., ann. i. 28, 1–3Google Scholar, Stat., Theb. vi. 686 f.Google Scholar, Sen., Med. 797Google Scholar, Juv. vi. 442 f., Mart. xii. 57, 16 f., Plin., N.H. ii. 12Google Scholar. So Plut., Aem. Paul. 17Google Scholar.

^{page 14 note 3} See Harduin on Plin., N.H. ii. 12Google Scholar.

^{page 14 note 4} Ruperti on Juv. vi. 442 f.

^{page 14 note 5} Schol. Ap. Rhod. ii. 526 μϵθ’ ὅπλων ἐπιτηρϵῖν τὴν ἐπιτολὴν τοῦ κυνός.

^{page 14 note 6} See Preller-Robert i. 458, n. 2, Schirmer in Roscher i. 549, 33 ff., Pridik, de Cei insulae rebus p. 136Google Scholar f., Gruppe, Gr. Myth. p. 234Google Scholar.

^{page 14 note 7} Ponticus, Herakleidesap. Cic. de divin. i. 130Google Scholar ‘etenim Ceos accepimus ortum caniculae diligenter quotannis solere servare, coniecturamque capere, ut scribit Ponticus Heraclides, salubrisne an pestilens annus futurus sit. namsi obscurior et quasi caliginosa stella exstiterit, pingue et concretum esse caelum, ut eins aspiratio gravis et pestilens futura sit: sin illustris et perlucida stella apparuerit, significari, caelum esse tenue purumque et propterea salubre.’

^{page 15 note 1} Roehl, inscrr. Gr. ant. ² p. 24Google Scholar, ix. 1 figures a votive cymbal with Thessalian inser. on which see Hoffmann, Dial. ii. 52Google Scholar, no. 81, Roberts, Ep. i. 244Google Scholar, n. 237a, and especially Studniczka, in Ath. Mitth. 1896 xxi. 240Google Scholar who reads Καμὸὐνέθυσε τᾶι ΚόρFαι = Καμὼ ἀνέθυσε κ.τ.λ.

^{page 15 note 2} Gruppe, Gr. Myth. p. 54Google Scholar, n. 9.

^{page 15 note 3} See Lenormant in Dar.-Sagl. ii. 576.

^{page 15 note 4} Cp. Rohde, Psyche 248Google Scholar, 2, and Hoeck 3. 302 ff.

^{page 15 note 5} Lenormant ibid. p. 571 f.

^{page 15 note 6} So et. mag. 180, 36 ὅτι μετὰ κυμβάλων ἠχοῦσα τὴν Κόρην ἐζήτει

^{page 15 note 7} Lenormant ibid. p. 563. Cp. Preller-Robert ii. 792, n. 1.

^{page 15 note 8} Ov., A. A. ii. 609Google Scholar f. ‘condita si non sunt Veneris mysteria cistis, nec cava vaesanis ictibus aera sonant’ is referred by Ruhnken to the rites of Demeter rather than to those of Cybele (Heinsius, Burmann).

^{page 16 note 1} Riess in Pauly-Wissowa i. 1986, 11 s.v. Amulett describes a fibula with a number of small metal knobs found in a tomb and explains them as a prophylactic rattle: see Annali d. Inst. 1882 Pl. Q, 7.

^{page 16 note 2} Bruzza, in Annali delľ Inst. 1875 p. 60Google Scholar (cp. p. 67 f.) cites for the use of bronze in funeral rites Passeri Mem. della Soc. Colomb. vol. i., Lorenzi, de praecom. cyth. fistul. ac tintinnab. Gronov. viii. p. 1469Google Scholar, Magio, de tintinnab. Sallengre ii. p. 1187Google Scholar, Lazzarini, de vario tint. usu Romae 1822Google Scholar.

^{page 16 note 3} See Sartori, Paul ‘Glockensagen u. Glockenaberglaube’ in Zeitschrift des Vereins für Volkskunde vii (1897), 113Google Scholar ff., 270 ff., 358 ff., viii (1898), 29 ff.

^{page 16 note 4} On it see Dyer, ThiseltonThe Ghost World, p. 15Google Scholar.

^{page 16 note 5} Roehl, inscrr. Gr. ant. ² p. 26Google Scholar, x. 8 figures a votive cymbal with Laconian inscr. ‘Οπωρὶς ἀνέθηκε Λιμνάτι, on which see Roberts, Ep. 1, p. 251Google Scholar f., no. 252. Other cymbals inscribed to Artemis Limnatis are I.G.A. 50 and I.G.A. 73 Π[ολυα]νθὶς (?) ἀνέθηκε τᾷ Λιμνάτι

^{page 16 note 6} Jahn, O. in Berichte d. k. Sächs. Ges. d. Wissenschaften Philol.-Hist. 1855, p. 101Google Scholar ff.

^{page 16 note 7} Elworthy, F. T.The Evil Eye, p. 327Google Scholar, enumerates ‘the tympanum on one [hand], bells on one, crotala…on two, cymbals on three.’

^{page 16 note 8} Brit. Mus. Cat. Bronzes, 874–876Google Scholar.

^{page 16 note 9} ib. no. 875, Fig. 22.

^{page 16 note 10} ib. no. 876.

^{page 16 note 11} On the cymbals, tympana, and krotala used in the worship of Cybele see Rapp in Roscher, Lex. ii. 1658, 44 ff.Google Scholar, Decharme in Dar.-Sagl. i. 1682, n. 132, Pottier ib. 1697.

^{page 16 note 12} Firmic., Matern, . de error. profan. relig. 18Google Scholar Halm.

^{page 16 note 13} Fröhner, Cat. 545Google Scholar, Reinach, Rép. Stat. i. 101Google Scholar. Cp. the relief from a Kybele-and-Attis altar in Baumeister, Denkm. p. 801Google Scholar, Fig. 866, which shows two bells slung from a pine.

^{page 16 note 14} Bötticher, Baumkultus, p. 538Google Scholar, Fig. 13.

^{page 16 note 15} Roscher, Lex. ii. 1647Google Scholar, Fig. 2.

^{page 17 note 1} Roscher, Lex. ii. 1613, 15 ffGoogle Scholar. According to Preller-Robert i. 134, ‘aehnliche Gebräuche beobachtete Ross, Kleinas. 7Google Scholar auf der Insel Megiste an der Küste von Lykien.’

^{page 17 note 2} Apollod. ii. 5. 6. Cf. Diod. iv. 13 κατα σκευάσας χαλκῆν πλαταγήν καὶ διὰ ταύτησ ἐξαίσιον κατασκευάζων ψόφον ἐξεφόθει ζῶᾳ καὶ πέρας τῇ συνεχείᾳ τοῦ κρότου ῥᾳδίως ἐκπολε καὶ πέραμήσας καθαρὰν ἐποίησε τὴν λίμνην

^{page 17 note 3} Ap. Rhod. ii. 1049 ff., Hygin., fab. 20Google Scholar. The latter expressly cp. the armed dance of the Curetes.

^{page 17 note 4} Roscher, Lex. ii. 1615Google Scholar, 1 ff. Reinach, Rép. Stat. ii. 146Google Scholar, 1 = Caylus v. 50, 1, a bronze which according to Caylus represents ‘Curète frappant sur un tympanon.’

^{page 17 note 5} Cat. 318, Fig. 11.

^{page 17 note 6} E.g. Brit. Mus. Vase Cat. iv. F 58Google Scholar. In F 303 a Maenad carries ‘a tympanum in l. hand, and in r. a bell (?) painted white.’

Roscher, Lex. i. 1085, 16 ffGoogle Scholar.

^{page 18 note 1} Brit. Mus. Cat. Vases ii. p. 70 ff. nos. B 62, B 63, B 65.

^{page 18 note 2} Figured in Dar.-Sagl. ii. 789, Figs. 935, 936.

^{page 18 note 3} Exodus xxviii. 33 ff., xxxix. 25 f. Cf. Zech. xiv. 20 ‘In that day shall there be upon the bells of the horses, Holy unto the Lord.’ Mr.Crooke, W. (Popular Religion and Folklore of N. India p. 108Google Scholar) mentions priests among the Gonds who wear bells for the purpose of scaring demons.

^{page 18 note 4} Cp. Kaibel, , inscrr. Gr. Sic. It. no. 2409, 5Google Scholar.

^{page 18 note 5} The scansion ὖποτέταγμαι metri gratia can be paralleled by the Homeric ολακόμοροι θῖγα τέρες κῦάνεοσ on which see Kühner-Blass i. 308 f.

^{page 18 note 6} Bruzza loc. cit. p. 55, not perceiving the hexameter ending, explains ὄμμα as= ὄραμα, φάντασμα, cp. Xen, . de re eq. 9Google Scholar. 4 ὥσπϵρ ἄνθρωπον ταράττϵι τὰ ἐξαπίναια καὶ ὁράματα.

^{page 18 note 7} O. Jahn loc. cit. p. 31 n. 9 cj. καίνουσαν for βαίνουσαν cp. the grammarians' etymology of βασκαίνϵιν viz. φάϵσι καίνϵιν (schol. Aristoph., Plut. 571Google Scholar, schol. Theocr. v. 12, Etym. mag. 190, 26).

^{page 18 note 8} Cp. C.I.G. 6792.

^{page 18 note 9} I cannot quote another example of this use of the word in Greek. Was it a translation of the Latin ‘subjectus’? Or should we suppose, with Bruzza (p. 55 f. ), that it is the bell which speaks: ‘io campanello sono stato ordinato contro del fascino’? If so, ‘I am subject to’ must presumably mean ‘I am a servant of,’ ‘I am used in the ritual of.’ Bruzza renders ὐποτάσσω by ‘collocare e disporre sotto a una cosa.’

^{page 19 note 1} Mus. Kircher. p. 6, pl. 58, Montfaucon antiq. expliq. iii. pl. 55.

^{page 19 note 2} Dar.-Sagl. i. 902, figs. 1146, 1147, Duruy, Hist. Rome ii. 725Google Scholar, Babelon, Monn. Rép. Rome ii. 291Google Scholar f. The enlargement of the type in Dar.-Sagl. and Duruy is due to De Koehne, , Revue de numism. belge, 5e série, vol. ii (1870), p. 51 f.Google Scholar, pl. iii.

^{page 19 note 3} Suet., Aug. 91Google Scholar ‘ideoque mox tintinnabulis fastigium aedis redimiit, quod ea fere ianuis dependebant.’

^{page 19 note 4} Plaut., Pseud. 326–332Google Scholar.

^{page 19 note 5} Liv. iv. 13, 7.

^{page 19 note 6} Gronovins cj. ‘bove et arvo’!

^{page 19 note 7} Liv. iv. 16, 2.

^{page 19 note 8} Plin., N.H. xxxiv. 5Google Scholar, 11 ‘Item P. Minucio praefecto annonae extra portam Trigeminam unciaria stipe collata’ (a memorial column was erected), cp. xviii. 3. 4 ‘Minucius Augurinus, qui Sp. Melium coarguerat, farris pretium in trinis nundinis ad assem redegit undecimus plebei tribunus: qua de causa statua ei extra portam Trigeminam a populo stipe collata statuta est.’ With this agrees Dion. Hal. περὶ ἐπιβουλῶν p. xxxvi ed. C. Muller.

^{page 19 note 9} See Babelon, Monn. Rép. Rom. ii. 228Google Scholar, Morell, Thesaurus Numism. p. 284Google Scholar f., Stevenson, Rom. Coins p. 559Google Scholar, Dar.-Sagl. i. 1351.

^{page 19 note 10} Babelon loc. cit. says: ‘Les clochettes suspendues au monument servaient à annoncer ľouverture et la fermeture du marché’!

^{page 20 note 1} Ruinart, Acta Martyr, p. 538Google Scholar Veronae 1731 cum velut animal traherent sancti Sisinii orpus exanime, collo aerei testis tinnitum concavum ligaverunt, quod vulgus tintinnabulum vocant.’

^{page 20 note 2} Mai, spicileg. Rom. iii. 312Google Scholar ‘tintinnabulum cum sagmate magnum in collo suspendere, et cum his ad templum suum currere, et Stvltvs Svm clamare praecipiunt.’ Cp. the mediaeval ‘fool’ in ‘cap and bells.’

^{page 20 note 3} De reb. S. Macar. cod. Vat. lxiv, Zoega, Cod. Copt. p. 125Google Scholar Romae 1810.

^{page 20 note 4} Dionys. ii. 70 mentions first λόγχην ἢ ῥάβδον ἤ τι τοιοῦθ᾿ἕτερον but later on τὸν ἐν ταῖς ἀσπίσιν ἀποτελούμενον ὑπὸ τῶν ἐγχειδίων ψόφον.

^{page 20 note 5} ibid.

^{page 20 note 6} See Fowler, WardeRoman Festivals p. 39Google Scholar ff.: ‘the old Latins believed that the Spirit which was beginning to make the crops grow must at this time [March 1] be protected from hostile demons, in order that he might be free to perform his own friendly functions for the community.’

^{page 20 note 7} On this famous tomb see Baumeister, Denkm. i. 608Google Scholar, Martha, Ľ Art étrusque p. 206Google Scholar f., Dar.-Sagl. ii. 836, n. 378. With it should be compared a remarkable object in bronze said to have been found in central Italy and figured by Reinach, S. in ĽAnthropologie vii (1896), 188Google Scholar, fig. 441.

^{page 20 note 8} ἔπαισεν ^*V.

^{page 21 note 1} Christ, W.Gr. Lit. ³ p. 904Google Scholar.

^{page 21 note 2} Krumbacher, Byz. Lit. ²137Google Scholar f., 680.

^{page 21 note 3} Kosmas is described by Suid. s.v. Ἰωάννης ὁ Δαμασκηνός as ἀνὴρ εὐφυέστατος καὶ πνέων μουσικὴν ὅλως τὴν ἐναρμόνιον. οἱ γοῦν ᾀσμαικὸὶ κανόνες Ἰωάννου τε καὶ Κοσμᾶ σύγκρισιν οὐκ ἐδέξαντο,οὐδὲ δέξαιντο ἂν μέχρις ὁ καθ᾿ ἡμᾶς βίος περαιωθήσεται,

^{page 21 note 4} e.g. by Bouché-Leclereq, Hist. de la divination ii. 306Google Scholar.

^{page 21 note 5} Schol. Clem. Al., protr. 11Google ScholarΘεσπρωτία ...ἐνᾗ ἐλέγετο μαντεῖον εἶναι Διός, διὰ λεβήτων ἠχούντων τως γιγνόμενον. This rather hazy account harks back to the Demonian arrangement of a row of tripods or caldrons.

^{page 21 note 6} Call., h. Del. 286Google Scholarγηλεχέες θεράποντες ἀσιγήτοιο λέβητος.

^{page 21 note 7} Probably Herakleides Ponticus (W. Christ ib. p. 586 f.), who is cited in Apostol. x. 99 for another Boeotian story about an oracle.

^{page 21 note 8} The collection of proverbs fathered upon Plutarch goes back to the grammarian Seleukos, who flourished in the time of Augustus and Tiberius (Christ, W.Gr. Lit. ³ p. 605)Google Scholar, according to Crusius, O.Ind. lect. Tüb. 1887Google Scholar and 1895.

^{page 21 note 9} On these MSS. see p. 9, n. 7.

^{page 22 note 1} For the remarkable instance of bell-worship among the Gonds of N. India see p. 28 n. 4.

^{page 22 note 2} Athen. x. 427 D, xi. 479 D, xv. 666 B, 668 B, E, Aristot., rhet. i. 12Google Scholar. 1373a 23, Hesych. s.v. κότταβος.

^{page 22 note 3} Athen. X. 427 D ἦν ἀπ᾿ ἀρχῆς τὸ μὲν σπένδειν ἀποδεδομένον τοῖς, δ δὲ κὸτταβος τοῖς ἐρωμένοις. This is stated on the authority of Theophrastus ἐν τῷ περὶ μέθης.

^{page 22 note 4} Antiphan. Ἀφροδίτης Γοναί fr. 1, 4 ff. Mein.

^{page 23 note 1} Cp. our ‘dumb-waiter’ of similar shape.

^{page 23 note 2} See Dar.-Sagl. iii. 868, n. 4.

^{page 24 note 1} Gusman, Pompéi p. 146Google Scholar: ‘Les clochettes tintaient pendant les éclipses de lune et ľon s'en servait pour conduire les criminels au supplice.’

^{page 24 note 2} See p. 15.

^{page 24 note 3} Vitr. i, 1 and v. 5.

^{page 24 note 4} Vitr. i. 1 ‘in cellis sub gradibus,’ y, 5 ‘inter sedes theatri constitutis cellis.’

^{page 24 note 5} Vitr. i. 1 ‘uti vox scaenici sonitus conveniens in dispositionibus tactu cum offenderit, aucta cum incremento clarior et suavior ad spectatorum perveniat aures,’ v. 5 ‘vox a scaena … excitaverit auctam claritatem et concentu convenientem sibi consonantiam.’

^{page 24 note 6} See ib. n. 8.

^{page 25 note 1} Dar.-Sagl. ii. 449, fig. 2594.

^{page 25 note 2} Liv. xxxviii. 17.

^{page 25 note 3} Aesch., sept. 386Google Scholar.

^{page 25 note 4} Frazer, Pausanias iii. 314Google Scholar, Golden Bough ² i. 344 ff.

^{page 25 note 5} Following W. Mannhardt, Mr. Frazer has abundantly proved that whipping is a frequent form of ceremonial purification, the underlying idea being that it will drive out evil influences of all sorts: Golden Bough ² iii. 127–133, 215–219, esp. 218 n. 1 for the whipping of inanimate objects. To the examples there cited add the statement of Philo Alex., that farmers on the approach of a storm used once to beat the air with whips and rods (Bruzza loc. cit. p. 63, Lumbroso, nuovi studi alessandrini p. 41Torino 1872)Google Scholar. Agrippa beating the surface of the Avernian lake to dissipate its miasmas, and Xerxes laying stripes upon the Bosphorus, may have a similar signification.

^{page 25 note 6} Golden Bough ² iii. 60–93.

^{page 26 note 1} ibid. iii. 91 f.

^{page 26 note 2} Brit. Mus. Cat. Bronzes nos. 875, 876. Elworthy, The Evil Eye p. 327Google Scholar mentions the whip on four votive-hands known to him.

^{page 26 note 3} In Roscher, Lex. ii. 1671Google Scholar, fig. 6, Kybele, drivng her lion-car, holds a two-thonged whip (?) in her right hand and a tympanum in her left. Was the combination accidental or designed ?

^{page 26 note 4} Ap. Rhod. i. 1184 ff.

^{page 26 note 5} id.ib. i. 1139.

^{page 26 note 6} Cp. Theocr. ii. 30 ῥόμβος ὁ χάλκεος.

^{page 26 note 7} Fritzsche on Theocr. ii. 17, Jahn, Berichte d. k. sächs. Ges. d. Wiss. Philol.-Hist. 1854 p. 257Google Scholar. Lévesque, in Histoire et mémoires de ľinstitut royal de France Hist.-Litt. anc. iii. (1818), 5Google Scholar ff. concludes that ‘il avoit le plus souvent la forme du jonet nommé parmi nous sabot ou toupie.’

^{page 26 note 8} Fröhner Cat. de la coll. Van Branteghem, no. 67 = Dar.-Sagl. ii. 1154, fig. 3087.

^{page 26 note 9} Zonaras vii. 21 goes off on a wrong tack: καὶ κώδων ἀπήρτητο καὶ μάστιξ τοῦ ἃρματος, ἐνδεικτικὰ τοῦ καὶ δυστιχῆσαι αὐτὸν δύνασθαι,ὤστε καὶ αἰκισθῆναι ἢ και δικαιωθῆναι θανεῖν. τοὺς γὰρ ἐπί τινι ἀτοπήματι καταδικασθέντας θανεῖν νενόμιστο κωδωνοφορεῖν, ἵνα μηδεὶς βαδίζουσιν αἰτοῖς ἐγχριπτόμενος μιάσματος ἀναπίμπληται.

^{page 27 note 1} Frazer, Golden Bough ² iii. 63Google Scholar.

^{page 27 note 2} G. F. Angas Savage Life and Scenes in Australia and New Zealand; front. to vol. ii shows the pahu or war-gong being beaten.

^{page 27 note 3} Handbuch d. kl. Alt. V. ii. 1 p. 355 n. 7.

^{page 27 note 4} He probably was not Lucian: see Croiset, Hist. lit. grecque v. 590Google Scholar. But cp. Christ, W.Gr. Lit. ³ p. 747Google Scholar.

^{page 27 note 5} Frazer, Golden Bough ² iii. 463Google Scholar.

^{page 28 note 1} E.g. Reinach, Rép. Vases i. 23Google Scholar, 114, 175, 332, 363, 403, ii. 4, 46, 287, Dar.-Sagl. i. 1353, fig. 1794.

^{page 28 note 2} Cp. the two gilded eagles perched on a couple of columns before the altar of Zeus Lukaios (Paus. viii. 38. 7).

^{page 28 note 3} Crooke, Popular Religion and Folklore o N. India p. 60Google Scholar f.

^{page 28 note 4} The same is true of the bell. Id.ibid. p. 108: ‘The Gonds have elevated the bell into a deity in the form of Ghagarapen.’