1. I am grateful to Mr. I.M. Le M. Du Quesnay for much helpful criticism of this paper.

2. For the generally favourable opinion of mime in pagan antiquity, cf. Reich, H., Der Mimus (1903, repr. 1974) 50–80Google Scholar. In the following paragraphs, I select a mere fraction of the available evidence.

3. On Ecl. 6.11. The story, as Servius records it, is patently apocryphal, since probably none of the Eclogues had been written when Cicero died. It may well, however, be based on fact; note Ecl. 10.2 quae legal ipsa Lycoris.

4. This scene in Horace depends, of course, on a sophisticated literary tradition. The similarities with Lucian, , Convivium 18–19Google Scholar are particularly striking. The incident may or may not be historically true. The important point is that neither Horace nor his friends, including Maecenas, can have objected to having their enjoyment of this type of entertainment advertised and immortalised.

5. 2.3.17-18 posito formose sallat Iaccho,/egit ut euhantis dux Ariadna choros. Saltare, though usually understood as referring to pantomime, can refer equally well to mime; cf. Reich (n.2) 57, n.1. A more important objection to the view that Propertius is referring to a mime is the fact that mimes rarely had mythological subjects; cf. p.75 below.

6. The despondent opening words of the RE article on Mimos, although written almost fifty years ago, are still worth quoting, for they are still true: ‘Geschichte und Theorie des Mimos erfuhren im Altertum nie eine zusammenhängende Darstellung. Wir sind daher auf die zerstreuten, zwar zahlreichen, aber meistens nur zufälligen und nicht immer zuverlässigen Notizen in griechischen und römischen Schriftstellern angewiesen, die in ihrer Gesamtheit ein eindruckvolles Bild von dem hohen Alter, der weiten Verbreitung, der oft verblüffenden Lebenswahrheit, der grossen Lebenskraft und dem – bis etwa in das 1. Jhdt. unserer Zeitrechnung – ständig wachsenden Einfluss des Mimos geben. Auch in der Neuzeit hat sich die Forschung nur in bescheidenem Umfang und mit spärlichem Erfolg dem Mimos zugewendet.’ Most of the surviving remnants of Roman mime are collected and discussed by Bonaria, M., I Mimi Romani (1965)Google Scholar. The majority of these fragments are of little help in literary enquiries, for they have been preserved only by grammarians as illustrations of abstruse linguistic usages.

7. Cf. Reynolds, R.W., CQ 40 (1946) 77–84CrossRefGoogle Scholar.

8. Ovid refers to such mimes also at A.A. 1.501-2; [id.] 3.605-8 (quoted below, p.76), R.A. 755. Further, almost contemporary, evidence for the popularity of the Adultery-Mime is provided by Valerius Maximus 2.6.7 mimis … quorum argumenta maiore ex parte stuprorum continent actus; cf. also Seneca, , Contr. 2.4.5Google Scholar and Suetonius, , Otho 3Google Scholar.

9. To cite just the instances in the Amores, the love-triangle is exploited, alluded to, or at least implied in some way in 1.4, 6, 8,9, 10, 11, 12, 2.2, 3,4, 5, 7, 8, 9, 10, 11, 12, 15, 19, 3.3, 4, [5], 7, 8, 11, 12, and 14.

10. Horace presents a similar scene in the same context at Satires 2.7.56-61 metuens induceris atque/altercante libidinibus tremis ossa favore / quid refert, uri virgis ferroque necari / auctoratus eas, an turpi clausus in arca, / quo te demisit peccati conscia erilis, / contractum genibus tangas caput? Compare also Juvenal 6.237-8 abditus interea latet et secretus adulter / impatiensque morae silet et praeputia ducit. Horace is possibly suggesting a formal link between Satire and mime when he compares Lucilius and Laberius: nec tamen hoc tribuens dederim quoque cetera; nam sic / et Laberi mimos ut pulchra poemata mirer (Satires 1.10.5-6). The existence of such a formal link is supported by Johannes Lydus, who is presumably reflecting a traditional view when he links Roman Satire with the phlyax-playwright Rhinthon of Tarentum, and Persius with Sophron (De Magistratibus 1.41). It is worth adding that the influence of mime has been detected on both the form and the content of Juvenal's ninth Satire; cf. Highet, G., Juvenal the satirist (1954) 274Google Scholar. (I owe this reference to Dr J.G.W. Henderson.) Mime may, therefore, have been an intermediary between (Old) Comedy and Satire. (On the links between mime and Petronius, see below, n.44.) I argue below, p.78, that mime may have been an intermediary between Comedy and elegy.

11. The debt to Propertius is obvious, that to Horace is more subtle: Ovid means his audience to recall the farcical scene at Satires 1.2.127-34 through his reference to Ilia at lines 39-40. Ilia is not mentioned in Propertius' poem, but she appears in Horace at lines 125-6. For such by-passing of the immediate model, cf. Du Quesnay, I.M. Le M., Papers of the Liverpool Latin Seminar 1 (1976) 55Google Scholar with n.213, and for the use of proper names to signal an allusion, cf. Kroll, W., Studien zum Verständnis der römischen Literatur (1924) 150Google Scholar.

12. The word adulterium occurs at Amores 3.5.44, but Ovid is unlikely to be the author of that poem; cf. Kenney, E.J., ΑΓΩΝ 3 (1969) 1–14Google Scholar. Even if he did write it, that need not weaken the present argument: both the subject-matter and the dialogue form in which it is presented suggest that the Somnium could also have been influenced by mime.

13. The love-triangle is not, of course, confined to the Adultery-Mime. One thinks, for example, of Catullus 17, with its dull-witted husband and beautiful young wife, adservanda nigerrimis diligentius uvis (16).

14. Hubbard, Margaret, Propertius (1974) 151Google Scholar, has already observed in passing that ‘we could not be … more firmly located in the verismo of the mime’.

15. Compare Horace's mock-epic introduction to the contest between the two scurrae at Satires 1.5.51-4.

16. Cf. Hubbard (n.14) 56, n.1.

17. Cf. Hubbard (n.14) 149-52.

18. Greece & Rome n.s.18 (1971) 51–3CrossRefGoogle Scholar.

19. Latomus 32 (1973) 616–22Google Scholar.

20. Cf. Crusius, O., Neue Jahrbücher 25 (1910) 92–8Google Scholar. There are some exceptions; for example, the anonymous Faba, Laberius, ' Anna Peranna (see below, p.76)Google Scholar, and his Paupertas (cf. Crusius 96). Nonius 493M does, in fact, attribute an Odyssia to Laberius, but the text is probably corrupt and he is generally understood to be referring to the poem of Livius Andronicus.

21. This technique is not confined in elegy to the Propertian poems mentioned above; for Propertius in the role of Acontius in 1.18, cf. Cairns, F., CR n.s.19 (1969) 131–4Google Scholar; for Tibullus in the role of Odysseus in 1.3, cf. Bright, D.F., Arethusa 4 (1971) 197–214Google Scholar; for Ovid in the role of Hercules in Amores 3.1, cf. Froleyks, W.J., Der ΑΓΩΝ ΛΟΓΩΝ in der antiken Literatur (1973) 139–41Google Scholar; (Ovid's choice between the two goddesses, the regal virago, Tragedy, and the sly and sexy Elegy, also owes something to the Judgement of Paris). Ovid repeatedly identifies himself with mythological figures, especially Odysseus, in the exile poems; cf. Rahn, H., Antike und Abendland 7 (1958) 115–20Google Scholar. On the possible connection between the Odyssean and mime elements in Propertius 4.8, see also below, p.75 and n.47.

22. I am here dealing only with lines 23-42. I argue below, p.78, that 2.29 is a single elegy.

23. With the detail tunica velata recincta (645) compare, in poems possibly influenced by mime, Horace, , Satires 1.2.132CrossRefGoogle Scholardiscincta tunica fugiendum est and Propertius 4.8.61 tunicis..solutis. For the detail of leaping from the bed, for whatever reason, in such poems, note Horace, , Satires 1.2.129-30CrossRefGoogle Scholar, Propertius 2.29.39-40, Ovid, , Amores 3.7.81-2 (see below, p.79)Google Scholar; also Fasti 1.435, (2.350), 6.343.

24. Cf. Bömer, on Fasti 3.523ffGoogle Scholar.: ‘Das einzig lesbare Fragment [Bonaria (n.6) no. 10] geht auf eine Liebesszene’. The amatory nature of the mime cannot be asserted with great confidence: the fragment in question reads conlabella osculum.

25. I am not of course denying the existence of the elegiac puella in real life. That is a separate problem.

26. Ovid's words at 1.4.40, dicam ‘mea sunt’ iniciamque manum, are obviously not to be taken at face-value, especially since they follow immediately on fiam manifestus amator.

27. Lucretius' attack on the exclusus amator at 4.1177-84 may help confirm the existence of komastic mimes in Rome. That passage is one of the chief pieces of evidence for the widely, and perhaps correctly, held view that komoi were a feature of real life in Rome. (On this question, cf, most recently, Yardley, J.C., Eranos 76 (1978) 20–1)Google Scholar. Copley, F.O., Exclusus amator (1956) 45–6Google Scholar, argues that ‘Lucretius would never have made himself look ridiculous by attacking with all the scorn of which he was capable something that was only a foolish fancy of the poets and had no counterpart in real life’. This is at least an overstatement. The attack on the exclusus amator is only a very small part of the much broader attack on sex and women and, of the eight lines devoted to the exclusus amator, five are directed not so much against him as against the use of cosmetics. Since, then, the role played by the exclusus amator is a relatively minor one, it does not seem so necessary to insist that he must be drawn from life, and not from literature. A compromise between these two sources is possible. Lucretius' source could have been the contemporary theatre for, if komastic mimes were staged in Rome, the komast could have been familiar to Lucretius and his audience equally well from stage-performances as from real life. Lucretius often illustrates his arguments with exempla taken from the theatre; notice especially lines 978-83 of this same book. The imagery at line 1186, vitae postscaenia, perhaps strengthens the possibility that he is thinking here of the theatre and hence, more specifically, of mime, by this time the most popular type of theatrical performance in Rome.

I make this suggestion diffidently. Even if Lucretius is not drawing directly on real life, he could of course be drawing on, for example, komastic epigram. (For his debt to epigram in his attack on sex generally, cf. Kenney, E.J., Mnemosyne 4th series 23 (1970) 380–8CrossRefGoogle Scholar.) Certainly, although I consider the importance of this passage as evidence for the existence of komoi in real life in Rome to have been exaggerated, I should not wish to argue that a debt to mime (or any other literary model) excludes or even significantly diminishes the possibility that Lucretius is also reflecting reality.

28. Cf. Anderson, W.S., AJP 85 (1964) 1–12Google Scholar and, more recently and in greater detail, Pinotti, P., GIF n.s.8 (19) (1977), 59–71Google Scholar.

29. Cf. Gow, A.S.F., Theocritus II 118Google Scholar.

30. Such a debt to mime can be seen most clearly in Theocritus' other komos poem, Idyll 3, which admits a change of scene between lines 5 and 6.

31. It may be worth adding that Highet (n.10), who detects the influence of mime on Juvenal's ninth Satire, suggests that ‘conceivably the satire is also intended as a parody of the theme of the exclususamator’.

32. Other possible objections are noted and rejected by Cairns, F., Emerita 45 (1977) 337–44CrossRefGoogle Scholar.

33. Cf. Cairns (n.32) 344-9, who treats the whole poem as a komos and the transition after line 22 as a specifically komastic device.

34. The influence of mime on such abrupt transitions in Propertius is noticed by Hubbard (n. 14) 52-3, with particular reference to 1.8, another poem the unity of which has often been disputed.

35. Cf. e.g. Day, A.A., The origins of the Latin Love-Elegy (1938) Chapter 5 andGoogle Scholar, most recently, Yardley, J.C., Phoenix 26 (1972) 134–9CrossRefGoogle Scholar and TAPA 104 (1974) 429–34Google Scholar.

36. The similarities between the plot of Menander's Epitrepontes and that of the third century A.D. mime PLit. Lond. 52 (= Page, D.L., Greek literary papyri (1942) no.79Google Scholar) are particularly striking.

37. Cf. Yardley, , TAPA (n.35), who cites (pp.432–4)Google Scholar evidence for such relationships in both Comedy and mime.

38. It may be worth adding that the image which Ovid uses in his erotodidaxis at A.A. 1.59 quot caelum Stellas, tot habet tua Roma puellas, finds its closest parallel at Herodas 1.32-3 (ἔστ' ἐν .

39. Cf. TLL. s.v. carissa.

40. One may also compare Horace's acknowledgement in the Odes of his debt to his archaic, but not to his Hellenistic, models. The elegists may have been considerably influenced by such humble handbooks as Parthenius' Erotica Pathemata, but they naturally preferred to lay claim to having poets such as Philetas and Callimachus as their models; cf. Griffin, J.JRS 67 (1977) 19Google Scholar.

41. The other instance, Tibullus 1.5.39-42, is much briefer and more delicately expressed: saepe aliam tenui: sed iam cum gaudia adirem. / admonuit dominae deseruitque Venus. / tune me discedens devotum femina dixit, / a pudet, et narrat scire nefanda meam.

42. It is worth noting, in this context, the hypothesis proposed by Abel, W., Die Anredeform bei den römischen Elegikern (1930) 70 n.16Google Scholar, that the introduction of mimetic dialogue to erotic epigram was an innovation made by Philodemus. The general influence of mime on epigram is of course much older; cf. Gow, and Page, on Asclepiades 25Google Scholar (= A.P. 5.181).

43. Martial refers to this Catullus, (RE 2)Google Scholar at 5.30.3 facundi scaena Catulli; cf. Lib. Spect. 7.4. It should be said, however, that in the present passage he may be referring to the aeschrological poems of the poet from Verona. The influence of mime on the mockery of impotence and other such defects in other genres has already been suggested by Brecht, F.J., Philol. Suppl. 22.2 (1930) 96Google Scholar. It may be worth adding that Pomponius wrote a Hirnea Pappi in the closely related genre of Atellane farce; for mime as a more modern equivalent of Atellane farce, cf. Cicero, , Ad Fam. 9.16.7Google Scholar. It may also be worth adding that mime- (pantomime-?) actors (sometimes?) wore phalloi; cf. Schol., to Juvenal 6.65-6Google Scholar.

44. The exact extent of the influence of mime on the Satyricon and on the novel genre in general is uncertain; for a bibliography on this problem, cf. Astbury, R., CPh 72 (1977) 30 n.41Google Scholar, adding Möring, F., De Petronio mimorum imitatore (1915)Google Scholar.

45. Petronius, by his imitation of Ovid's poem, acknowledges his debt only to the higher genre. This is yet another example of the practice discussed at the end of the preceding paragraph.

46. Mime is not, of course, the only possible source for this theme in novels. Its much earlier exploitation in Old Comedy and, more significantly, in Lysias' speech on the murder of Eratosthenes, shows it to be widespread and traditional.

47. The Satyricon is heavily influenced not only by mime, but also by the Odyssey; cf. Cameron, A.M., Latomus 29 (1970) 397–425Google Scholar. Petronius' exploitation of Odyssean and mime elements involves much the same technique as was suggested above for Propertius 4.8 (p.75).

48. The possible influence of tragic pantomime on the Metamorphoses is noted by Galinsky, G. Karl, Ovid's Metamorphoses (1975) esp. 68–9Google Scholar. Cairns (n.32) 353, suggests pantomime as the source of Propertius' mythological paradeigmata in 1.3.1-6.

49. The text is uncertain, but the meaning is clear.

50. Isidore is referring to pantomime, not to mime. The term mimus was often used in this inexact manner, especially in later antiquity.

51. Cf. RE s.v. Pantomimus coll. 847-9.

52. For the extensive influence of pantomime on Nonnus' Dionysiaca, cf. Weinreich, Otto, Epigramm und Pantomimus (1948) 161–72Google Scholar.