¹ Photius, s.v. ἰκρία specifically puts these seats in the agora before the theatre was built (ἐν τῇ ἀγορᾷ … πρὶν ἢ (κατα)σκευασθῆναι τὸ ἐν Διονύσου θέατρον). Further details are supplied by Hesychius and the Suda, though their authority and consistency are matters of controversy. See A. Pickard-Cambridge, The Theater of Dionysus at Athens (1946), 10–15; F. Kolb, Agora und Theater (1981), 26–31; L. Polacco, Il teatro di Dioniso Eleutereo ad Atene (1990), 24–32 (including all the testimonia). The danger of such collapse was not exclusively an archaic phenomenon, as the Hellenistic original of Pl., Curc. 643–7 attests. See W. Beare, The Roman Stage (1963), 241–7.

² Pollux, , Onom. 4.126–7, 130–1Google Scholar, generally supported by Tzetzes, , De. Com. XIA1 125–7Google Scholar (Koster), an addendum to Dübner's Vita Aristoph. XI (not in Koster), and Vitr. 5.6.8, though these sources tend to confuse stage right and audience right. See Rees, K., ‘The parodoi in the Greek theater’, AJP 32 (1911), 377–402Google Scholar and Pickard-Cambridge, op. cit. (n. 1), 234–9. The convention is easily read, for example, into Men., Sa. 95–6 as Moschion exits to the country while Demeas and Nikeratos enter from the harbour. See K. B. Frost, Exits and Entrances in Menander (1988), 103. Attempts to find this convention preserved in Roman comedy are reviewed by M. Johnston, Exits and Entrances in Roman Comedy (1933), 4–12.

³ Liv., Per. 48: ‘inutile et nociturum publicis moribus’. The other sources are Val. Max. 2.4.2 (who names the censors), Vell. Pat. 1.15.3, App., , BC 1.28Google Scholar, and Oros. 4.21.4 (who dates Scipio's intervention to 151). There are discrepancies among these accounts and some textual problems, but the general course of the episode is beyond dispute. See E. S. Gruen, Culture and National Identity in Republican Rome (1992), 206 n. 110. For the possibility that beneath these discrepancies lies evidence for a fourth attempt to build a permanent theatre late in the second century, see n. 32 below.

⁴ Duckworth, G. E., The Nature of Roman Comedy (1952), 79Google Scholar. Cf. Stambaugh, J. E., The Ancient Roman City (1988), 229Google Scholar: ‘the spectators sat on the grass or on wooden bleachers’. No references are given, but the jerry-built quality of the Plautine theatre is often adduced from Tacitus’ comment at Ann. 14.20 that games used to be presented on temporary stages before hastily-constructed grandstands, not the best testimony, especially in its moralizing context, for theatre practices three centuries earlier.

⁵ Contrast the continuing fecundity of Greek scholarship on such issues as reviewed, for example, by Hall, E., ‘Theatrical archaeology’, AJA 101 (1997), 154–8CrossRefGoogle Scholar. For the problems of topography and dramaturgy in Pl., Curc. 462–86, see Moore, T. J., ‘Palliata Togata: Plautus, Curculio 462–86’, AJP 112 (1991), 343–62Google Scholar.

⁶ Gruen, op. cit. (n. 3), 209. For the more traditional view, with its decidedly moralistic cast, see Duck worth, op. cit. (n. 4), 79–82 and Hanson, J. A., Roman Theater-Temples (1959), 18–25Google Scholar.

⁷ Saunders, C., ‘The site of dramatic performances at Rome in the times of Plautus and Terence’, TAPA 44 (1913), 87–97Google Scholar.

⁸ Cic., Har. 24: ‘Nam quid ego de illis ludis loquar quos in Palatio nostri maiores ante templum in ipso Matris Magnae conspectu Megalesibus fieri celebrarique voluerunt?’

⁹ Romanelli, P., ‘Lo scavo al tempio della Magna Mater e nelle sue adiacenze’, Monumenti Antichi 46 (1963), col. 202–330: 223–7Google Scholar discuss the history of excavations on the site. Cf. E. Steinby, Lexicon Topographicum Urbis Romae (1996), 3.206–8. Identification of the figurines remains controversial: see Thomas, G., ‘Magna Mater and Attis’, ANRW 11.17.3 (1984), 1506Google Scholar. The altar has on an adjacent side an inscription dated 27 March A.D. 192, the day of the lavatio of the Magna Mater (CIL 6.4.2: 30967L). For the official cult title, see S. Takács, ‘Magna Deum Mater Idaea, Cybele, and Catullus’ Attis', in Lane, E. (ed.), Cybele, Attis and Related Cults (1996), 372–5.CrossRefGoogle Scholar

¹⁰ Hanson, op. cit. (n. 6), 13–16. Hanson was the first to bring archaeological evidence into the discussion but was probably discouraged from pursuing it far by Romanelli's belief that no trace of the original temple remained.

¹¹ The main publications are Pensabene, P., ‘Area sud-occidentale del Palatino’, Roma, Archeologia del Centro, Lavori e studi di Archeologia 6 (1985), 179–212Google Scholar; ‘“Auguratorium” e tempio della Magna Mater’, Archeologia laziale 2 (1979), 67–74Google Scholar; ‘Scavi nell' area del Tempio della Vittoria e del Santuario della Magna Mater sul Palatino’, Archeologia laziale 9 (1988), 54–67Google Scholar; and ‘Nuovi rinvenimenti nell' area sud-ouest del Palatino (1992–1993)’, Archeologia laziale 12.1 (1995), 13–28Google Scholar.

¹² Liv. 29.10.4–11.8, 29.14.5–14; Ov., , Fasti 4.247–348Google Scholar (arrival of the Magna Mater). Liv. 34.54.3 36.36.3–5 (her games and temple). A fragmentary didascalion to Pseudolus preserves the date and occasion of its performance. For the coming of the Magna Mater, a much discussed event, see in general Thomas, op. cit. (n. 9), 1502–8 and with special attention to its politics, Gruen, E. S., Studies in Greek Culture and Roman Policy (1990), 5–33Google Scholar. For the history of the ludi Megalenses see Taylor, L. R., ‘The opportunities for dramatic performances in the time of Plautus and Terence’, TAPA 68 (1937), 289–91Google Scholar and for their activities, Scullard, H. H., Festivals and Ceremonies of the Roman Republic (1981), 97–101Google Scholar.

¹³ The photographs, taken on 10 April 1997 (the date seemed propitious), are my own. I am grateful to Dotoressa Capo di Ferro of the Soprintendenza Archeologica di Roma for guiding me through the site and to Professor Patrizio Pensabene of the University of Rome for permission to reproduce his elevations of the temples there. An aerial photo of the Palatine, most helpful for showing the steepness of the slope at this point, appears as the frontispiece to Coarelli, F., Roma. Guide Archeologiche Monadori (1994)Google Scholar.

¹⁴ Liv. 10.33.9 (Megillus); 10.29.14 (Fabius). Identification of the Victory temple was first suggested tentatively by Castagnoli, F., ‘Nota sulla topografia del Palatino e del foro Romano’, Archeologia Classica 16 (1964), 186Google Scholar and confirmed by Wiseman, T. P., ‘The Temple of Victory on the Palatine’, Antiquaries Journal 61 (1981), 35–52CrossRefGoogle Scholar = Roman Studies (1987), 187–204 and 380–1, who also identifies the smaller structure as Fabius’ temple of Juppiter Victor. Ziolkowski, A., The Temples of Mid-Republican Rome (1992), 172–9Google Scholar, supports Castagnoli and Wiseman on identification of the Victory temple but believes the smaller structure to be the ‘aedicula Victoriae Virginis prope aedem Victoriae’ dedicated by Cato in 193 B.C. (Liv. 35.9.6).

¹⁵ Romanelli, op. cit. (n. 9), 227–39.

¹⁶ Fire of III B.C.: Obseq. 39 (‘maxima pars urbis exusta cum aede matris magnae.’); cf. Val. Max. 1.8.11; Tac., , Ann. 4.64Google Scholar, and Morgan, M. G., ‘Villa Publica and Magna Mater’, Klio 55 (1973), 231–45CrossRefGoogle Scholar. Augustan restoration: Res Gestae 19; Ov., , Fast. 4.348Google Scholar, with Pensabene, op. cit. (n. 11, 1985), 183–7 and Gros, P., Aurea Templa (1976), 232–4.Google Scholar

¹⁷ Romanelli, op. cit. (n. 9), 232: ‘non si puo pensare che al principìo del II secolo av. Cr. un tempio della importanza di questo, sul Palatino, venisse costruito non piu in opera quadrata di pietra (tufo o peperino) ma in opera incerta…’ Yet Temple D in the Largo Argentina, which Coarelli, F., ‘L'identificazione dell' Area sacra dell' Argentina’, Palatino 12.4 (1968), 365–73Google Scholar, identifies as the Temple of the Lares Permarini dedicated in 179 B.C. by the censor M. Aemilius (Liv. 40.52.4), has a comparable podium. The identification of this temple, but not the dating of its original podium, has since been challenged by Zevi, F., ‘Tempio D del Largo Argentina: Tempio delle Ninfe in Campo?’, Archeologia laziale 12.1 (1995), 135–43Google Scholar.

¹⁸ Coarelli, F., ‘Public building in Rome between the Second Punic War and Sulla’, PBSR 45 (1977), 10–13Google Scholar; Pensabene, op. cit. (n. 11, 1979), 71.

¹⁹ This possibility was actually raised by Pensabene, op. cit. (n. 11, 1979), 71–3 and echoed by Scullard, op. cit. (n. 12), 98, but both shy away from its consequences. A somewhat cryptic remark by Servius, ap. G. 3.24Google Scholar that stages were once more temporary than seats may offer some indirect support for the idea. Theatre historians, with the partial exception of Hanson, have avoided the question entirely.

²⁰ The basin itself is made of tufa blocks and paved with cotto tiles. See Romanelli, op. cit. (n. 9), 302–6 and figs 72–3, and for the stages of rebuilding, Pensabene, op. cit. (n. 11, 1988), 58–60.

²¹ Thus T. P. Wiseman, ‘Clodius at the theatre’, in idem, Cinna the Poet (1974), 168–9, building on a suggestion of Hanson, op. cit. (n. 6), 14 n. 29.

²² August., , CD. 2.4Google Scholar speaks explicity of the shows ‘for the Berecynthian mother of all, before whose couch on the holy day of her lavatio, such things are sung publicly with unholy performances’ (‘Berecynthiae matri omnium, ante cuius lecticam die sollemni lavationis eius talia per publicum cantitabantur a nequissimis scaenicis…’). Arn., , Nat. 7.33Google Scholar specifies a play on the story of Attis. The lavatio of the Magna Mater was held on 27 March; Augustine's ‘cantitabantur’ indicates a pantomime. For these March rituals, see Fishwick, D., ‘The cannophori and the March festival of Magna Mater’, TAPA 97 (1966), 193–202Google Scholar, Thomas, op. cit. (n. 9), 1517–21, and for juxtaposition of the March and April celebrations, Beard, M., ‘The Roman and the foreign: the cult of the “Great Mother” in imperial Rome’, in Thomas, N. and Humphrey, C. (eds), Shamanism, History, and the State (1994), 164–90Google Scholar. Hanson, op. cit. (n. 6), 14–16, and Wiseman, op. cit. (n. 20), 168–9 do not distinguish sufficiently between them. Pylades' career at Rome dates from the 20s B.C.; Bathyllus was a favourite of Maecenas. For pantomime generally, see Beacham, R. C., The Roman Theatre and Its Audience (1992), 140–9Google Scholar.

²³ Cic., Har. 24: ‘qui sunt more institutisque maxime casti sollemnes, religiosi.’ For Republican constraints on the cult of Cybele, see D.H. 2.19.4–5, Lucr. 2.600–60, the explanatory apologies of Ov., , Fast. 4.191–246Google Scholar, and K. Summers, ‘Lucretius' Roman Cybele’, in Lane, op. cit. (n. 9), 337–65.

²⁴ Ov., , Fast. 4.291–348Google Scholar tells how Claudia drew the goddess’ ship up the Tiber, ‘mira, sed et scaena testificatur loquar’ (326). For the story, see Wiseman, T. P., Clio's Cosmetics (1979), 94–9Google Scholar, and for the praetextae, Flower, H. I., ‘Fabulae Praetextae in context: when were plays on contemporary subjects performed in Republican Rome?’, CQ 45 (1995), 170–90CrossRefGoogle Scholar and further references there.

²⁵ Lenaghan, J. O., A Commentary on Cicero's Oration De haruspicum responso (1969), 125Google Scholar. Scaena can mean ‘performance’ as well as ‘stage’. At Cic., , Leg. 1.47Google Scholar, for example, ‘sensus nostros non parens, non nutrix, non magister, non poeta, non scaena depravat, non multitudinis consensus abducit a vero’, the sequence ‘poet…performance….audience’ is unmistakable. Cavea at Har. 26 (‘in alteram [sc. caveam]…ex altera’) would then by metonymy mean ‘audience’. The one problem with this interpretation lies just above, where Cicero imagines a swarm of bees coming in scaenam caveamve (25). This indeed sounds like a building, but scaenam caveamve is in fact Mommsen's emendation of caenam caveam in the best manuscript and thus presupposes what for us is the point at issue.

²⁶ D.H. 19.5, D.C., fr. 39.6–8 and Zon. 8.2. Val. Max. 2.2.5 draws the moral. Further references in Broughton, MRR, 189–90. The year was 282 B.C.

²⁷ Val. Max. 2.4.2: ‘standi virilitas propria Romanae gentis nota esset’, a moral argument remembered as late as Tac., , Ann. 14.20Google Scholar. There is no evidence, however, that a senatus consultum banning seats ever had any significant effect, nor do we ever hear of an audience unable to sit for a performance. No ban, of course, could have been easily enforced when temple steps stood conveniently by.

²⁸ Jocelyn, H. D., ‘The Roman nobility and the religion of the Republican state’, Jr. Religious Hist. 4 (1966–1967), 103Google Scholar, provides an important, if extreme corrective: ‘For a Roman of traditional upbringing the line between sincere magic-making and clear-eyed deception would have been difficult to draw; likewise that between selfish motives and patriotic ones. A sympathetic Greek observer like Polybius might see political intelligence where in fact there was only the naive piety of blinkered untaught minds.’ Cf. Morgan, M. G., ‘Politics, religion and the games in Rome, 200–150 B.C.’, Philologus 134 (1990), 14–19CrossRefGoogle Scholar.

²⁹ This is the clear implication of Liv. 25.12.14 and Macr., , Sat. 1.17.27–9.Google Scholar Cf. Saunders, op. cit. (n. 7), 91–2, too quickly dismissed by Hanson, op. cit. (n. 6), 12–13. Our evidence for the building activities of the censors is conveniently gathered by Coarelli, op. cit. (n. 18), 3–7.

³⁰ Aug., Res Gestae 21 calls the structure dedicated in memory of Marcellus a ‘theatrum ad aedem Apollinis’. The location had been selected by Caesar (Plin., , NH 7.121Google Scholar; D.C. 43.49.3). The temple, originally dedicated in 431 to Apollo Medicus after a plague (Liv. 4.25.3, 29.7), was extensively rebuilt by the consul C. Sosius in 34 B.C. (Plin., , NH 13.53Google Scholar, 36.28). See La Rocca, E., ‘Der Apollo-Sosianus-Tempel’, in Kaiser Augustus (1988), 121–5Google Scholar, Steinby, op. cit. (n. 9), 1.49–50, and most recently Viscogliosi, A., ‘Ad aedem Apollinis’, Archeologia laziale 12.1 (1995), 79–92Google Scholar. The existing podium of Sosius' temple is strikingly close to the Theatre of Marcellus, but not so oddly close if we recall the traditional association of plays with temple venues.

³¹ The ideological connection between them is stressed by Wiseman, op. cit. (n. 14), 42–6 and ‘The god of the Lupercal’, JRS 85 (1995), 1–22Google Scholar. The post holes of Iron Age structures are visible today just below the steps of the Magna Mater's temple: the most recent report of the area is by Pensabene, P., ‘Casa Romuli sul Palatino’, RPAA 63 (1990–1991), 115–62Google Scholar. The Lupercal awaits modern excavation.

³² North, J. A., ‘Deconstructing stone theatres’, in Apodosis: Essays Presented to Dr W. W. Cruikshank (1992), 76–9Google Scholar, observes that two sources for the events of 154 B.C., Vell. Pat. 1.15.3 and App., , B.C. 1.125Google Scholar, actually identify the opponent to this project as a consul Caepio, not Scipio, who was not consul. No suitable Caepio was on the scene in 154, but Q. Servilius Caepio, consul in 106, could readily have played such a role at the end of the century. North's scenario is attractive, but how our unambiguous sources for the events of 154, Livy and Valerius Maximus, failed to mention this comparable event of 107/6 remains unexplained.

³³ Var., Men. 150B: ‘dum e scaena coronam adlatam imponeret aedilis signo deae.’ The text is in fact corrupt at the key places, but Scaliger's correction of essena hora nam and Madvig's signo deae for signosiae in the MSS. of Nonius must both be right. For the sense, see Wiseman, op. cit. (n. 21), 158–9, and for the date of the satire, C. Cichorius, Römische Studien (1922), 214.

³⁴ Dio 44.6.3 and D.H. 7.72.13 allude to theatrical processions in the first century B.C., but most of the testimony is much later. See Taylor, L. R., ‘The “Sellisternium” and the theatrical “pompa”’, CP 30 (1935). 127–8Google Scholar and Hanson, op. cit. (n. 6), 81–5. The synoptic view of ludi found, for example, in Stambaugh, op. cit. (n. 4), 230–2, cannot be assumed to reflect the early days of any particular festival.

³⁵ Richardson, L. Jr, ‘A note on the architecture of the Theatrum Pompei’, AJA 91 (1987), 123–6CrossRefGoogle Scholar, doubted the existence of a significant temple attached to Pompey's structure because ‘we can see no trace of a massive rear addition, which would have projected into the Piazza Campo de' Fiori’ (125). In fact, the curve of the temple's apse is preserved in the line of a bearing wall within the present Palazzo Pio just off the Piazza. I am grateful to Professor Astra Zarina of the University of Washington Center for showing me this wall and explaining its significance. The Temple of Venus Victrix also finds its echo in the Temple of Venus Genetrix designed for the rival Forum Iulium. For the conceptual relationship between them, see Ulrich, R. B., ‘Julius Caesar and the creation of the Forum Iulium’, AJA 97 (1993), 53–4CrossRefGoogle Scholar.

³⁶ Cat. 5.6–8; Prop. 2.32.11–12, 4.8.75; Ov., , Ars 1–67, 3387–8Google Scholar; Mart. 2.14.10, II.I.II, 11.47.3. Coarelli, F., ‘Il complesso pompeiano del Campo Marzio e la sua decorazione scultorea’, RPAA 44 (1971/1972), 99–122Google Scholar, discusses the textual evidence and surviving decoration of Pompey's opera. For its vistas, see Gleason, K., ‘Porticus Pompeiana: a new perspective on the first public park of ancient Rome’, Jr. Garden Hist. 14 (1994), 13–27CrossRefGoogle Scholar, and more generally, Hanson, op. cit. (n. 6), 43–55. The thematic relation of its constituent parts remains unclear. Coarelli simply notes the presence among its motifs of images recalling Pompey's cult of Venus and others drawn from Hellenistic drama; G. Sauron, ‘Le complexe pompéien du Champs de Mars: nouveauté urbanistique à finalité idéologique’, in L'urbs, espace urbain et histoire (1987), 457–73, sees a more elaborate design inspired by the researches of Varro.

³⁷ Tert., De Spect. 10: ‘…vocans non theatrum, sed Veneris templum nuncupavit, cui subiecimus, inquit, gradus spectaculorum.’ The comment is echoed in Gell. 10.1.7: ‘aedem Victoriae … cuius gradus vicem theatri essent.’ An expensive Augustan renovation (‘Pompeium theatrum … impensa grandi refeci’, Res Gestae 20) may have raised the proscaenium and obstructed this view, but Pompey's original intention is unmistakable. The rebuilt theatre itself then became the model for subsequent structures, but they — and it — belong to the history of imperial rather than republican theatre buildings. See Sear, F. B., ‘The scaenae frons of the theater of Pompey’, AJA 97 (1993). 687–701CrossRefGoogle Scholar.

³⁸ So Johannowsky, W., ‘Osservazioni sul teatro di Iasos e su altri teatri in Caria’, ASAA n.s. 31–32 (1969/1970), 451–9Google Scholar, also noting that the theatre in a city like Iasos did double duty as ekklesiasterion, precisely the kind of political use the Roman aristocracy feared. Temples attached to theatres become common; see Hanson, op. cit. (n. 6), 59–77. His examples all show a temple at the back (or behind) the cavea, allowing the divinity to look out onto the stage as the Magna Mater originally did on the Palatine and as Venus Victrix learned to do in the Campus Martius. So too the Theatre of Marcellus positioned its scaena in notional, if not literal, view of Apollo's temple. Further illustrations and discussion in Frézouls, E., ‘Aspects de l'histoire architecturale du théâtre romain’, ANRW 11.12.1 (1982), 356–65Google Scholar.

³⁹ Rightly observed by Altheim, F., A History of Roman Religion (1937), 290–1Google Scholar. Morgan, op. cit. (n. 28), 27 n. 64 fails to see that the topography of the Palatine would have required an entirely different site for a permanent theatre.

⁴⁰ Gruen, op. cit. (n. 12), 80–4. This does not, of course, preclude the possibility of pre-literary drama at earlier ludi or their continuation into later times as proposed by Wiseman, T. P., Historiography and Imagination (1994), 12–16.Google Scholar

⁴¹ Enn., Ann. 79–81: ‘omnes avidi spectant ad carceris oras.’ See Goldberg, S. M., Epic in Republican Rome (1995), 106–8.Google Scholar

⁴² For Caesar's aedileship, Suet., Caes. 10, Plut., , Caes. 6.1–3Google Scholar; for Scaurus, M., Plin., , NH 36.113–15Google Scholar. The Theatre of Pompey, which Pliny thought sufficient for the city's needs, held perhaps 40,000 (NH 36.115). For the politicization of the theatre, Cic., , Att. 2.19.3Google Scholar, Sest. 117–26; Suet., Caes. 84, and in general, Nicolet, C., The World of the Citizen in Republican Rome (1980), 367–73.Google Scholar

⁴³ Forming a credible estimate for the population of Rome is no easy task. Recent work suggests a population of approximately one million under Augustus, half that in 130 B.C., and no more than c. 200,000 at the beginning of the second century. See Morley, N., Metropolis and Hinterland (1996), 33–9CrossRefGoogle Scholar.

⁴⁴ Gruen, op. cit. (n. 3), 188–93.

⁴⁵ Ter., Hec. 4–5: ‘ita populus studio stupidus in funambulo animum occuparat’ (cf. 33–6). The situation was different on the next attempt to produce Hecyra at the funeral games of Aemilius Paullus, when the audience was disturbed by a second crowd arriving for gladiatorial shows scheduled for the same space, probably in the forum. See Gilula, D., ‘Who's afraid of rope-walkers and gladiators?’, Athenaeum 59 (1981), 29–37Google Scholar, and Sandbach, F. H., ‘How Terence's Hecyra failed’, CQ 32 (1982), 134–5CrossRefGoogle Scholar. In 56, Cicero was still complaining about the narrowness of the space before Cybele's temple (Har. 22).

⁴⁶ Cic., , Att. 2.1.5.Google ScholarRawson, E., ‘Discrimina ordinum: the Lex Julia Theatralis’, PBSR 55 (1987), 105Google Scholar, finds even two feet inadequate, ‘given that the toga is a bulky garment, that in a culture so given to gluttony many rich men were doubtless fat, and that real squashing would be felt undignified’. For Athens, see Pickard-Cambridge, op. cit. (n. 1), 140–1; the issue has been reopened (though not quite convincingly) by Dawson, S., ‘The theatrical audience in fifth-century Athens: numbers and status’, Prudentia 29 (1997), 1–14Google Scholar. It is notoriously difficult to calculate crowd capacity in unmarked spaces: the issue involves not just the physical size of human bodies (whether ancient or modern, standing or sitting) but their tolerance for each other. Thus the Elizabethan Rose Theatre would, by modern standards, accommodate c. 400–500 spectators but can be judged from secondary evidence to have held c. 2,000. See I. Mackintosh, Architecture, Actor and Audience (1993), 11–14, 22–5 and M. Bradbrook, The Arts of Performance in Elizabethan and Early Stuart Drama (1991), 200–10. Such considerations vitiate the calculations of C. Huelsen, ‘Il Posto degli Arvali nel Colosseo e la capacità del teatri di Roma antica’, Boll. Communale di Roma (1894), 312–22, that set the capacity of Balbus' theatre at 6–7,000, of Pompey's at 9–10,000, and of the Theatre of Marcellus at 10– 11,000; figures I might otherwise wish I could accept.

⁴⁷ At least part of this prologue is post-Plautine, but still not later than the second century. See Jocelyn, H. D., ‘Imperator histricus’, YCS 21 (1969), 95–124Google Scholar, and most recently Moore, T. J., ‘Seats and social status in the Plautine theatre’, CJ 90 (1994), 114–17Google Scholar. The command at 5 to ‘keep your seats and keep your temper’ (‘bonoque ut animo sedeant in subselliis’) is equally consistent with the temple scene: subsellia ‘benches’ is an entirely appropriate term for temple steps.

⁴⁸ Liv. 34.44.4–5. Cf. Liv. 34.54.4; Cic., pro Corn. ap. Ascon. 55 St., Har. 24. Val. Max. 2.4.3 confirms the date and the earlier de facto practice, though there is some confusion over whether the occasion was the ludi Romani or Megalenses. See Sternberg, J. von Ungern, ‘Die Einführung spezieller Sitze für die Senatoren bei den Spielen (194 V.Chr.)’, Chiron 5 (1975), 157–63Google Scholar, and Rawson, op. cit. (n. 46), 107–10. For the censors' possible motives and public reaction to them, Gruen, op. cit. (n. 3), 202–5.

⁴⁹ Vitr. 5.6.3–5, describing the configuration of stone theatres, contrasts the benches (subsellia) in the cavea and the sedes in the orchestra. Subsellia suggest social inferiority: cf. Pl., Capt. 471; Stich. 93, 488–9, 703–4. The original concession of seats to the Senate in 194 does not imply the existence of an orchestra for them at that time.

⁵⁰ These are the only dates noted in surviving calendars, leading Hadzsits, G. D., ‘The dates of the Megalesia’, TAPA 61 (1930), 165–74Google Scholar, to argue that the festival occupied only these two days. The Romans' tendency in this period, however, to expand the holiday calendar at every opportunity makes so literal a reading of the sources hard to accept. See Taylor, op. cit. (n. 12), 284–304. Scullard, op. cit. (n. 12), 99, is discreetly non-comittal: ‘whether it had always been a seven-day festival remains uncertain.’

⁵¹ Thus Liv. 44.37.8 dates to 3 September 168/7 B.C. an eclipse that by modern reckoning occurred on 21 June 168, meaning that the Megalesia in the 160s took place in late January. See Briscoe, J., Commentary on Livy Books XXXIV–XXXVII (1981), 17–26.Google Scholar

⁵² Balsdon, J. P. V. D., Life and Leisure in Ancient Rome (1969), 268Google Scholar, a valid conclusion even without Huelsen's estimate of theatre capacities (n. 46 above), which Balsdon wrongly accepts. Cf. Morgan, op. cit. (n. 26), 33–5. Even among the aristocracy, not every one was at the games or even in the city. During the Megalesia of 56, for example, Crassus, Caesar, and Pompey (and their retinues) were out of town. Cicero was defending Caelius in the Forum before a jury of equites. See Wiseman, op. cit. (n. 20), 162–3. Milo's trial for the death of Clodius likewise took place during the Megalesia of 52.

⁵³ Suet., Vita T. says that the play was acted twice and earned the unprecedented fee of 8,000 nummi. For the amount, see Gilula, D., ‘How rich was Terence?’, SCI 8/9 (1989), 74–8Google Scholar. This is our only known case of an encore, but the record is both unreliable and incomplete. Mattingly, H. B., ‘The Terentian Didascaliae’, Athenaeum 37 (1959), 168–9Google Scholar, traces this particular statement not to Varro but to a misunderstanding of Eun. 19–22. The likelihood of multiple performances at Roman ludi strengthens the case for understanding Cicero's two scaenae at Har. 25 as referring to performances rather than theatres.

⁵⁴ See Hunter, R. L., The New Comedy of Greece and Rome (1985), 35–42CrossRefGoogle Scholar. Roman tragedies were apparently also performed entirely from the stage and without choruses: see Jocelyn, H. D., The Tragedies of Ennius (1967), 18–21, 29–38.Google Scholar

⁵⁵ Cheiron: London, BM F 151. Auletris: Melbourne, National Gallery of Victoria D14/1973. Musical scenes include the famous Bari Pipers in a private collection and the St. Petersburg Obeliaphoroi, Hermitage Mus. inv. 2074 (W.1 122). All are well illustrated in Taplin, O., Comic Angels (1993)Google Scholar, pl. 12.6, 15.13, 14.11, 14.12 respectively. If, as Taplin suggests (93–4), these are the stages of travelling players, there would be even less reason to envision an orchestral space before them. The ubiquitous staircase is in any case an indication of ready communication between the stage and the space below, though it is not in itself a hallmark of temporary stages. See Sifakis, G., Studies in the Hellenistic Theatre (1967), 130–2.Google Scholar

⁵⁶ Livius Andronicus, the first Roman dramatist, was a Greek from Tarentum: Cic., Brut. 72–3. For the influence of Magna Graecia on Roman popular culture — of which comedy was certainly a part — see now N. Horsfall, La cultura delta plebs romana (1996), 21–32, and for Italian theatre outside Rome and Graecia, Magna, Rawson, E., ‘Theatrical life in Republican Rome and Italy’, PBSR 53 (1985), 97–113Google Scholar.

⁵⁷ Slater, N., Plautus in Performance (1985), 165 n.Google Scholar 17. See the useful survey of L. Benz, ‘Die römischitalische Stegreifspieltradition zur Zeit der Palliata’, in the significantly titled volume, Plautus und die Tradition des Stegreifspiels (1995), 139–54.

⁵⁸ The eyeline is defined as five degrees above the horizontal of a standing actor of average size. For this principle of theatre design and its consequences in performance, see Mackintosh, op. cit. (n. 45), 135–8. For the Greek acting style, see Pickard-Cambridge, A., The Dramatic Festivals of Athens (2nd edn, 1968), 167–76Google Scholar. Corresponding testimony for Republican Rome is lacking until Cicero's time. There is a good sampling of this in E. Csapo and Slater, W. J., The Context of Ancient Drama (1995), 275–85Google Scholar. The description of a Plautine performance by Beacham, op. cit. (n. 22), 86–116, is also helpful.

⁵⁹ The point is made and the evidence gathered by Bain, D., Actors and Audience (1977), 186–9Google Scholar. For the theatre of Lycurgus, see Pickard-Cambridge, op. cit. (n. 1), 138–41 (cavea), 146–47 (orchestra). Its monumental skene as reconstructed by Townsend, R. F., ‘The fourth-century skene of the theater of Dionysos at Athens’, Hesperia 55 (1986), 421–38CrossRefGoogle Scholar, would also have further discouraged intimacy. The fifth-century theatre, which may have had a rectilinear orchestra and certainly had a smaller auditorium, presented a significantly different performance space, which precludes direct comparison between Old and New Comedy. I therefore limit discussion here to Plautus and his Greek models. For the proportions of that Periclean theatre and its lingering problems, see Polacco, op. cit. (n. 1), 170–4 and further references there.

⁶⁰ Oenopion and Maron are legendary drinkers. Karpelos is unknown, Timocles presumed to be a contemporary Athenian. See Arnott, W. G., Alexis, The Fragments (1996), 304–6.Google Scholar

⁶¹ cf. the similar direct address at 45–6. This style is largely formulaic, paralleled in the first case by Hen., fr. 5.6–8 and repeated verbatim in the second by Men., Sik. 24–5.

⁶² For Plautus and his model, see Stockert, W. (ed.), Plautus, Aulularia (1983), 8–16CrossRefGoogle Scholar and Arnott, op. cit. (n. 59), 859–64. Not everyone thought such restraint a virtue in a prologue. The hetaira Gnathaena once raised a laugh at Diphilus' expense by saying that she had chilled his wine by pouring one of his prologues into it (Athen. 580). A whiff of that Ψυχροτής survives in the prologue to Plautus'Rudens.

⁶³ For the embedded comic routine, see Gilula, D., ‘The crier's routine’, Athenaeum 81 (1993), 283–7Google Scholar. Other good examples of such prologues are Amph. 1–7, Mil. 79–85 (delayed), Poen. 1–10, Truc. 1–9. See Slater, op. cit. (n. 56), 149–53.

⁶⁴ Vitr. 5.6.2. For the Hellenistic stages, raised perhaps four metres above the orchestra, see Pickard Cambridge, op. cit. (n. 1), 190–4; Sifakis, op. cit. (n. 55). 133–5.

⁶⁵ The word's only other appearances are at Amph. 68 (the same context) and Truc. 931. Theatrum appears just once, referring to the genre, not the place: ‘nugas theatri’ (Ps. 1081). Terence, so much more restrained in his metatheatrical effects, uses no vocabulary drawn from the stage.

⁶⁶ Varr., Men. 561 refers to an instrumentalist playing in the orchestra (‘priusquam in orchestra pythaules inflet tibias, domi suae ramites rumpit’). The pythaules, as distinct from the choraules, accompanied solo singers (Diom. GLK 492). For Polyb. 30.22.11, describing Anicius' musical extravaganza in the Circus Maximus, ‘orchestra’ had its Greek sense of a performance space where dancers performed to musical accompaniment (…ὀρχησταὶ δύο εἰσήγοντο μετὰ συϕωνίας εἰς τὴν ὀρχήστραν). On all this, see Scamuzzi, U., ‘Studio sulla Lex Roscia theatralis’, RSC 17 (1969), 289–91Google Scholar.

⁶⁷ Scamuzzi, op. cit. (n. 66), 270–9. For the lex Roscia theatralis: Liv., Per. 99 and the scene at Cic., , Att. 2.19.3Google Scholar. Further references in Broughton, MRR 145 and a good, general treatment by Rawson, op. cit. (n. 46), 102–6.

⁶⁸ Vitr. 5.6–7 contrasts the different geometry of Greek and Roman theatres. Cf. Frézouls, op. cit. (n. 38), 365–9. Vitruvius‘ orchestra was strictly a seating area for senators (5.6.2). For a survey of Republican theatre buildings, see Beacham, op. cit. (n. 22), 56–69, and for their design, Brothers, A. J. in Barton, I. (ed.), Roman Public Buildings (1989), 99–104Google Scholar. Curio's ‘temporary’ theatre, built in 52 B.C., was still in use in June 51 (Plin., , NH 36.116–20Google Scholar; Cic., , Fam. 8.2.1Google Scholar).

⁶⁹ Gopnik, Adam, ‘Letter from London’, The New Yorker, 7 July 1997, 31Google Scholar. Equally apposite to the Latinist's experience is the comment of Mackintosh, op. cit. (n. 46), 14: ‘The Rose excavation provides early evidence that great theatres which nurture creative drama are usually very small’. Our evidence for this truth is much earlier still.