^{page 113 note 1} So these pieces are called by Stobaeus, and this is clearly behind the readings in Schol. Nic. Ther. 377: Welcker, the latter preferable, as one would expect of the person. The compound must be of the copulative type (Debrunner, , Gr. Wortbild., p. 40Google Scholar), meaning ‘both mime and iambos’: an appropriate designation for a genre taking its subject-matter largely from the non-literary mime, its language and metre from the Ionian of Hipponax. , as usual, means more than the metre.

^{page 113 note 2} The titles of the other are completely suitable. The possible interpolations in those of 1 and 6 and Kaibel; against this Groeneboom) are another question.

^{page 113 note 3} So Bücheler (in a lecture summarized in Ztschr. f. d. Gymn. lix [1905], 176–7Google Scholar), except that there, presumably by a mistake or misprint, (2) (i) and (ii) are divided at 40.

^{page 113 note 4} So Orph. H. 67. 6. The stronger tradi tion (see R.-E. ix. 1. 95Google Scholar) was that she was his daughter: die prominent position given to her here suggests however that the variant is being followed.

^{page 113 note 5} Cf. also, for the relative construction in 1 and the participle in 7–8, Norden, E., Agnostos Theos, pp. 166 ff.Google Scholar

^{page 113 note 6} Herzog, R., ‘Aus dem Asklepieion von Kos’, Arch. Rel. Wiss. x (1907), 205 ff.Google Scholar, showed that (91) here means a piece of money, a later substitute for the food offered to the serpent-god. This has been widely accepted: apart from editors of Herodas, see, e.g., Ziehen, , R.-E. xix. 1. 250Google Scholar, Nilsson, , A.J.P. lxviii (1947), 302 sqq.Google Scholar and Gesch. d. gr. Rel. ii. 73.Google Scholar It is rejected by Dodds, , The Greeks and the Irrational, p. 129 n. 66Google Scholar, who, however, does not sufficiently take into account the archaeological argument (which is not invalidated even if the scene is not, as Herzog assumes it is, intended to be specifically Kos, see below).

^{page 114 note 1} Cf. Theocr. 15. 82–83 . Although there are indications that such an unsophisticated attitude to art was normal in antiquity, it is eminently appropriate in these simple women, and it is at least unnecessary to see here Herodas expressing his own views (so, e.g., Dalmeyda, , Les Mimes d'Hér. [1893], 28, 30,Google Scholar and Groeneboom in his edition; this is opposed by Cataudella); the praise of Egypt in 1. 27 sqq., also claimed as Herodas' own, likewise fits its context; 6. 20 is a more likely case (see C.Q. N.s. xiv [1964], 32 n. 3)Google Scholar; 8 is obviously not comparable. With these possible exceptions Herodas maintains a complete detachment (the position of Snell, B., Poetry and Society [1961], 106 sq.Google Scholar, that, because Herodas does not condemn his immoral characters, he must sympathize with them, is untenable).

^{page 114 note 2} See below, p. 120.

^{page 114 note 3} Lippold, , R.-E. xi. 1. 236,Google Scholar thinks it unlikely that the altar-group would have a dedicatory inscription. I am not competent to judge this.

^{page 115 note 1} It is still necessary to state that these have nothing to do with the sculpture of Boethos described by Plin. HM. 34. 84. Boethos belonged to the second century B.C., Herodas to the third (see below).

^{page 115 note 2} For this device see, e.g., Groeneboom, on Aesch. Cho. 22Google Scholar and Fraenkel, on Aesch. Ag. 20,Google Scholar with literature.

^{page 115 note 3} This is a less successful example, since, so to speak, the intentions of author and character do not coincide: Herodas marks the end of the speech by the repetition, whereas Metrotime means to continue, but is forcibly interrupted by Lampriskos, ‘der fürchten mag, dass der Epilog länger werde als die Rede’ (Crusius). There is further a secondary repetition in this passage: shortly after the first appeal to the Muses it is said of the delinquent multi, but see Groeneboom). (8–13), and immediately before the second (53–55; on the text of 55 see Headlam; in 53 the is due to Terzaghi, the papyrus having ).

^{page 115 note 4} This was first seen by Mr. E. W. Hand-ley, who discussed the subject in a seminar at the University of London Institute of Classical Studies. I am much indebted to him in this section.

^{page 115 note 5} The idea (due originally to Weil, , J. des Sea. 1891, 671Google Scholar, ib. 1892, 518; see also Headlam on 1. 11) that the oath by the , which occurs without immediate relevance to their functions only in Herodas (1. 11, 66, 4. 30), is peculiarly Koan, is largely a petitio principii.

^{page 115 note 6} Puccioni's idea that the mention of a feast in 5. 80 shows Miletus to be the scene, the feast being in honour of Neleus, father of and founder of Miletus, is impossible: is always an epithet of Nestor, not of Neleus, cf. the story in Hes. frr. 15, 16; and the feast of Neleus at Miletus was the , cf. Nilsson, , Gr. Feste p. 242.Google Scholar The meaning of this has not been established with certainty. Puccioni rightly rejects Crusius’ connexion (Unters. Z. d. Mim. d. Her. p. 113) with the of Kos who took part in a feast on the (Paton-Hicks, , Inscr. of Cos 37. 47 sqq.Google Scholar), while the feast here is on the fourth day after the Meister's explanation depends on Hsch. , which is probably a late invention (Sornmer, , Beitr. z. gr. Wortf. 119 f.Google Scholar). Possible is the connexion, advocated by Herwerden and Biicheler, with (known from Steph. Byz. 205. 3). Other proposals are too far-fetched. The possibility also remains that (Schulze, , B. ph. W. 1895, 6 = Kl. Schr. 678–9Google Scholar, and Headlam, , C.R. xiii [1899], 154Google Scholar) is correct, (Jevons, , Academy 1891, 337Google Scholar) is improbable: the alteration is considerable, and , rare in Attic poetry (Denniston, , G.P. 200Google Scholar), is even rarer in Hellenistic poetry (only a few times in Theocritus).

^{page 116 note 1} Southeimer, , R.-E. ivA. 2. 2537–8Google Scholar, to whose list add Ephesus (Hippon. fr. III 12 D³., 78. 12 Masson; Fränkel, M., Inschr. v. Perg. II 268 D 35Google Scholar, with commentary p. 203).

^{page 116 note 2} Unsuccessful attempts to do this have been made by Herzog, , B. ph. W. 1898, 1251Google Scholar, J.O.A.I. vi (1903), 217Google Scholar, Anm. 3 (Ephesus), and Puccioni (Miletus).

^{page 116 note 3} One of Knox's arguments (Philol. lxxxi [1926], 246–7Google Scholar) to show that Herodes [sic] was an Athenian is that the portrayed in 8 was a purely Attic custom and that he says of it (8. 40) . However, although none of the sources gives another locality for this custom, none says that it was restricted to Attica: and the two earliest, Theophr. ap. Porphyr. de abst. 2. 10 (for the Theophrastan origin of this passage see Regenbogen, , R.-E. Suppl. vii. 1512Google Scholar) and Eratosth. fr. 5 D., 22 Pow. , imply at least the possibility that it was known elsewhere. The rest of Knox's case proves nothing (see Herzog, , Philol. lxxxii [1927], 63)Google Scholar, and against him is the Doric name (to which all the evidence points) and some Doric features in the dialect (especially ae and cue contracted to , and the omission of in oaths, for which see Headlam on 5. 77).

^{page 116 note 4} Strab. 14. 2. 19 (of Cos) .

^{page 117 note 1} Some have thought the present tense an anachronism intended to show the women's ignorance (Weil, , J. des Sav. 1891, 670Google Scholar; Reinach, , R.É.G. iv [1891], 214 n. 1Google Scholar; Dal-meyda, , Les Mimes d'Hér. [1893], 135Google Scholar): but this would have point only if the true date were considerably later, which is impossible—Herodas is a Hellenistic writer: besides the general nature of his work, this is indicated by Plin. Epp. 4. 3. 3 where he is set side by side with Callimachus. Attempts to prove the contrary (Ellis, , C.R. v [1891], 457Google Scholar, J. Phil. xxiii [1895], 20–23Google Scholar; Walker, Miâs ii [1910], 378 sqq.Google Scholar) require no refutation.

^{page 117 note 2} In fact it is probably not, cf. Biebei, M.J.D.A.I. xxxviii/ix (1923/1924), 242 ff.Google Scholar (who unfortunately overlooked the fact tha the dating of Herodas 4 depends largely 01 that of the sons of Praxiteles and used it ti date them, thus becoming involved in i circular argument; but this by no mean destroys the force of her other arguments) Byvanck, A. W., Mnem. s. 4, iv (1951), 204 ff.CrossRefGoogle Scholar

^{page 117 note 3} Volkmann, H., R.-E. xxiii. 2. 1621–2.Google Scholar

^{page 118 note 1} Under Ptolemy II Philadelphos (285–247 B.C.). See C.R. N.s. xv (1965), 7–9.Google Scholar

^{page 118 note 2} Before 266 B.C., because the Syrian town , mentioned in 16, changed this name to between 286 and 266 B.C. (Reinach, , Mél. Havet [1909], 452–5Google Scholar); although the objections that the old name might have continued in popular use (Legrand, , Ét. sur Théocr [1898], 126 n. 1Google Scholar) and that might be preferred to for metrical convenience (Herzog, , in Crusius-Herzog, Die Mint, des Her. [1926], 39 f.Google Scholar) have some weight, the case is corroborated by the dating of 1 and 4.

^{page 118 note 3} These spaces are all recorded in the Headlam-Knox apparatus.

^{page 118 note 4} The naming of speakers on their first appearance, found, e.g., in the Dyskolos papyrus, does not occur in Herodas.

^{page 119 note 1} They were certainly not intended for performance by a company of actors, as was once thought: see Hertling, op. cit., Legrand, , R.é.A. iv (1902), 5–35Google Scholar, and the article of Fürst, K. summarized in W.K.P. xxxiv (1917), 180.Google ScholarPasquali, , in Xenia Romana (1907), 15–27Google Scholar, Fürst and Puccioni among others argue that they were pure Buchpoesie, because otherwise they would be incomprehensible and the changes of speaker would lead to confusion. This does little justice to the intelligence of the audience and the talent of the performer. The nature of the mime and of ancient poetry in general favour recitation.

^{page 120 note 1} The assumption of a change before ig depends on Crusius' belief that he could see a paragraphos before this verse: no one else has ever been able to observe this.

^{page 120 note 2} Danielsson's . is an unlikely alteration.

^{page 120 note 3} As indeed the scribe may have intended to divide with the mark above the a, which may be either an apostrophe (Meister) or an acute accent (Knox).

^{page 124 note 1} Cataudella's note, ‘punctum supra o, cum non sit expunctionis signum, potest esse errore adhmctum vel nihil significare’, is an extreme case of the weakness of his edition in critical matters. A dot above a letter is the normal sign of deletion in this papyrus (more common than an oblique stroke through the letter): how then can he simply deny that it is so intended here?