Hostname: page-component-74656f9884-2dmjx Total loading time: 0 Render date: 2026-11-07T23:56:45.454Z Has data issue: false hasContentIssue false
  • English
  • Français

INDOOR AND OUTDOOR: OBSERVATIONS ON THE ORIGINALS OF TERENCE'S ANDRIA AND HECYRA

Published online by Cambridge University Press:  03 April 2017

Benjamin Victor
Benjamin Victor*
Affiliation:
Université de Montréal
*
benjamin.victor@umontreal.ca
Rights & Permissions [Opens in a new window]

Extract

Inquiry will be made here into a specific challenge facing writers of fabulae palliatae, namely the interior scene, which they had not always the same means to display as did the poets of the νέα they adapted. One of them, Terence, will be seen to have reacted by eliminating the interior scenes that he found in his models.

Information

Type
Research Article
Information
The Classical Quarterly , Volume 67 , Issue 1 , May 2017 , pp. 118 - 131
Check for updates
Copyright
Copyright © The Classical Association 2017 

Inquiry will be made here into a specific challenge facing writers of fabulae palliatae, namely the interior scene, which they had not always the same means to display as did the poets of the νέα they adapted. One of them, Terence, will be seen to have reacted by eliminating the interior scenes that he found in his models.

INDOOR AND OUTDOOR IN NEW COMEDY

For the most part, dramatic space in the Greek theatre was a straightforward matter: the stage itself represented an outdoor place, while the skēnē, the building forming its backdrop, stood for a shelter of some sort, on going into which the characters of the play were deemed to be in an enclosed, usually indoor location.Footnote 1 One way around constraints of this scheme was the ekkyklēma (also called exōstra or exōstron), a rolling platform pushed onto the stage through a door from the skēnē.Footnote 2 On seeing the ekkyklēma, the Greek audience understood that the wall of the skēnē had come temporarily down, so to speak: so long as it remained in view there was no more visible barrier between indoor and outdoor.Footnote 3

The ekkyklēma was commonly used at Athens in the staging of tragedy and comedy from the fifth century onwards. In the Hellenistic period, it is known to writers, and even attested in inscriptions, outside Attica as well.Footnote 4 What interests us here is its use in New Comedy. This is most certain at Dys. 689–90 where Cnemon, injured by his fall into the well and unable to walk, is brought onto the stage through the door of his house. At 758 Cnemon asks to be ‘wheeled in’ again (following the supplement εἰσκ]υκλεῖτ’, which is unavoidable), thereby acknowledging that he has been on the ekkyklēma, as Gomme and Sandbach (among others) have realized.Footnote 5 He is visible to Sostratus, who has been outside the house; likewise at 753–4 Sostratus, still outside the house, is visible to him and those who attend him. Nothing in the text invites us to rationalize the action and suppose that Cnemon is literally out of doors, having been moved from the house to be given fresh air; rather we are seeing the interior of his house, or (better) the distinction of interior from exterior has been suppressed or modified.

Now the display of incapacitated persons is well paralleled as a function of the ekkyklēma. Euripides so used the machine at HF 1031 ff., with the same indefiniteness of place as in the Dyscolus.Footnote 6 Likewise in Bellerophon (TrGF 5.311), presumably to show the injured hero after his fall (κομίζετ’ εἴσω τόνδε τὸν δυσδαίμονα, where Aristophanes’ imitation at Eq. 1249 κυλίνδετ’ εἴσω … shows that the ekkyklēma was used). The ekkyklēma is also a strong possibility for the ailing Phaedra's entry at Hipp. 170–1.Footnote 7 As it was a regular method for moving invalids, it is especially pointed that Agathon, a womanish weakling, rides one in Aristophanes’ Thesmophoriazusae (as is made explicit at lines 96 and 265).Footnote 8 Given this association with infirmity, the use of the ekkyklēma should be assumed for an additional passage of Menander, namely Asp. 305–87. There Chaerestratus, on learning of Smicrines’ plans to marry Cleostratus’ sister, is so overcome by grief that he cannot rise on his own, yet somehow he is brought out through the door of his house; he converses with Daos and Chaereas, then is taken back in. It is not at all clear that while he is visible to the audience he is to be deemed out of doors. Not only is there a striking parallelism of motif with respect to other plays (and particularly to Dyscolus), but the poet's method of introducing and closing the Aspis scene, it has been remarked, recalls that of ekkyklēma-scenes in tragedy and in Old Comedy.Footnote 9 It is especially to be noted that Menander avoids language which might imply that Chaerestratus walks with assistance.

Certain other indoor scenes are known to have occurred in New Comedy, but the techniques used to stage them do not admit of reconstruction. Menander's Phasma, of which we have testimonia and fragments including two papyri, and also a Mytilenaean mosaic, must have displayed to the audience action in a passageway communicating between two houses, hence interior.Footnote 10 It has been proposed, but on inadequate grounds, that the girl who appeared in this passageway was displayed on an ekkyklēma;Footnote 11 it has also been suggested that the central stage door was understood to be located within the houses.Footnote 12 In Act I of another Menandrian play, the Philadelphoe or ‘first Adelphoe’, an older man conversed with his two daughters as they worked wool: an ancient mosaic shows them in what looks like an indoor setting,Footnote 13 and Plautus’ imitation in Stich. 87–154 confirms that they were indeed within a house, since to reach the women the man had passed through a door which would normally be shut.

There are also scenes of New Comedy whose ‘interiority’ is itself uncertain. Eating and drinking were frequently enacted on the New Comic stage, to judge from surviving fragments and from imitations in Plautus.Footnote 14 It does not follow, however, that the banquets of New Comedy were always to be understood as taking place outdoors: a convention may have allowed the distinction of outdoor from indoor to be relaxed in this circumstance. In two passages of Roman comedy (Persa 757–8 and Stich. 674–8) it is stated flatly that the party is set outside the house, but has Plautus taken over this detail from the original or added it himself by way of explanation? Or are the entire scenes his composition?Footnote 15 In some of the New Comic fragments (for example, Men. 209 K.–A. [Cecryphalus]), servants are ordered to take away tables and fetch certain items for the party: εἴτ’ εὐθὺς οὕτω τὰς τραπέζας αἴρετε | μύρα, στεφάνους ἑτοίμασον, σπονδὰς ποίει. The apparatus of the banquet was therefore carried from the skēnē onto the stage by actors, not wheeled out on an ekkyklēma.Footnote 16 Indeed, one visual representation of the famous luncheon scene in Menander's Synaristosae actually shows it on the stage.Footnote 17 The ekkyklēma cannot in any case have been suited to representing parties of any notable size, for it had to be small enough to fit through a set of double doors.Footnote 18

New Comedy, then, shows a range of phenomena including:

  • • tableaux on the ekkyklēma, whose fictional space (in so far as it can be defined) is to be understood as indoor;

  • • scenes of feasting and carousing, not normally using the ekkyklēma, which a spectator might possibly understand as indoor;

  • • scenes (such as the father's encounter with his daughters in Menander's Philadelphoe or the sighting of the young girl in the same poet's Phasma) of which the staging cannot be reconstructed but which must be understood as indoor.

PLAUTUS

Roman comedy of the second century b.c. derives of course from Greek comedy of the period just before, and shares most of its conventions. The stage itself stands in principle for an outdoor location, the space behind its backdrop (frons scaenae) for an indoor one. Notwithstanding, the representation of indoor scenes on the Roman side does not correspond in quite every particular to that on the Greek. In Plautus we find:

  • • no ekkyklēma: the existence of such a device is never hinted at, which, given the size of the corpus, surely means that Roman theatres of Plautus’ time had none;

  • • some banquet scenes with an expressly outdoor or partly outdoor fictional location (Persa 753–fin. [see lines 757–8]; Stich. 673–fin. [see lines 674–8]);

  • • other banquet scenes with an ambiguous fictional location;

  • • a few scenes played on the stage but to be understood as indoor.

The only clean qualitative break with the Greek tradition is the absence of the ekkyklēma. Indeed, the only mention of an ekkyklēma at Rome comes much later, in 56 b.c., and will be evidence for the elaborate Roman theatres of the late Republic, not the rudimentary ones of the middle Republic.Footnote 19

On other points Plautus may do no more than slightly modify conventions of his Attic models; in any case, one does have the impression of a modification, whose overall direction is toward greater freedom.Footnote 20 Let us first consider the banquets with an ambiguous fictional location. That of Argyrippus, Demaenetus and the courtesan Philaenium in the fifth act of Asinaria should not, in a rational world, be held where passers-by can observe, since Demaenetus’ purpose is to indulge his lust for Philaenium without his wife Artemona's knowledge, and indeed Diabolus and his parasite only become aware of it after going intro (809). Yet, the banquet is secluded from the view of other characters only to a degree (see line 880). Mostell. 308–407 invites similar observations. Where are the symposiasts of this passage to be imagined? If out of doors, why does the text never say that Philematium moves from within the house? She plainly was inside just before, especially at 157–293, where she has chosen her jewellery and made herself up in front of a mirror. If on the other hand the symposiasts are indoors, why must they move into the house when Theopropides’ approach is announced at line 365? As for the scenes that must be understood as interior, the courtesan at her toilette in Mostell. 157–293 has just been mentioned. So, too, has the opening scene of Stichus, which adapts the father-and-daughters scene of Menander's Philadelphoe/First Adelphoe: the young women are within a house (clear at line 87); when their father first approaches they cannot see him, but only hear his voice.Footnote 21 Note also lines 478–632 of Truculentus. There Phronesium appears on a bed—natural only if she is indoors. She is pretending to have recently been through childbirth, because she wants Stratophanes to think that he has fathered a son. As Stratophanes approaches, Phronesium and her servants cannot see him, but, like the daughters in Stichus, only hear him. A maid must be sent to usher him in (502–12), which should mean that Phronesium is not in an intermediate space (uestibulum) but within the house proper. Finally, Mostell. 817–57 represent Theopropides in a part (admittedly an unsecluded part) of Simo's house.

To be sure, it is not always in the direction of vagueness and freedom that Plautus innovated: he moved the eponymous meal of Synaristosae (= Cistellaria) off the stage, substituting a conversation in front of the house as the party breaks up. When, however, Plautus admits interior scenes, he observes few restrictions. His writing then is either admirably bold or unforgivably awkward, depending on how attached one is to illusionistic theatre and how much semiotic consistency one requires. It certainly makes demands on the audience's goodwill. At the end of the Stichus-scene one of the daughters is said to move ‘inside’, however that is to be understood (line 147). At lines 631–2 of Truculentus, Phronesium goes from the stage intro (she does so walking with assistance, so as to maintain her pretence of infirmity). She therefore enters the stage-building, and must be thought to recede either into or further into her house. Where exactly, then, should we imagine her to have been just before? ‘In an interior courtyard’ would account for the facts, but the better answer is ‘We are meant not to ask’. Liapis's concept of de-semiotization (see n. 3, above) is relevant to such passages. It may also be the case that Greek scenes adapted by Plautus benefitted from the ekkyklēma or from the entr'actes, which New Comedy had but Roman Comedy lacked. Entr'actes effect a break in the dramatic sequence. They make discontinuity in the passage of time appear acceptable, and it is easy to imagine them placed so as to smooth over a discontinuity in the fictional location of successive scenes. Plautus chose to stage interior scenes without the means available to his Greek predecessors; in so doing he took on a considerable artistic risk.

TERENCE

Interior scenes, of whatever degree or kind, are entirely absent from Terence's six comedies. There is nothing in his work like the toilette-scene of Plautus’ Mostellaria, nor do we ever have the impression of a barrier, invisible to the audience, shielding his characters from each other's view, as we do in the scenes of Asinaria, Stichus and Truculentus discussed above. Even banquet-scenes, ambiguous as to their fictional setting, are completely absent from Terence. Three possible reasons suggest themselves.

  1. (1) The issue never arose in the Greek plays that Terence adapted.

  2. (2) Terence disliked any fluidity of dramatic space. On finding it in his models he always eliminated it outright, whereas Plautus sometimes kept it (or perhaps even added it on his own) despite limitations imposed by the equipment of his own stage and the absence of act-division.

  3. (3) Though perhaps willing in principle to bend dramatic space, Terence found no way to do so, not even one acceptable to Plautus, in the plays that he adapted, because in these plays means available to the original authors at Athens but not to him at Rome were particularly necessary.

(1) is to be rejected, I hope to demonstrate, for there are strong indications that Terence twice found indoor scenes in his models and eliminated them. (3) accounts quite adequately for the facts. A case can be made for (2), based in part on Terence's fondness for realism in other regards, but the question is not central to our inquiry, and I would rather not go further into it here.

ANDRIA V.4 (904–56)

One suppression of an indoor scene comes in the Andria. In this play Simo has been trying to effect the marriage of his son Pamphilus to Philumena, the daughter of his neighbour Chremes. Pamphilus, however, is in love with Glycerium, who, having come to Athens from Andros, is deemed ineligible for marriage because she is foreign. Glycerium bears a child to Pamphilus, but Simo refuses to believe that the childbirth is real. The tension reaches its high point in Scenes V.2–3 (= 842–903), where the slave Davus tells Simo and Chremes of Crito's arrival in town. This man is Andrian and, says Davus, he knows Glycerium to be in fact an Athenian citizen (859). Thinking the story a fabrication, Simo has Davus bound and taken away for punishment. Pamphilus and Chremes none the less convince Simo to give Crito a hearing (901). Pamphilus enters Chremes’ house, where Crito has gone, and brings him onto the stage. There follows Scene V.4 (= 904–56), where Crito, in the presence of Pamphilus, Chremes and Simo, tells what he knows of Glycerium's life-story. From Crito's information, Chremes recognizes that Glycerium is not only of Athenian birth but also his own lost daughter (933). At once (949–51) he betroths her to Pamphilus.

The recognition and betrothal accomplished, the play ends after just one additional scene of twenty-five lines. Menander will not have so sprinted to the finish. He preferred to follow the resolution with a full act, effecting a gradual decompression.Footnote 22 And it is easy to deduce certain elements of this last Menandrian act. Chremes is eager to be reunited with his daughter Glycerium,Footnote 23 lost early in childhood but just now found again. Having borne a child that same day, Glycerium cannot easily spring from her bed and step outside: instead, Chremes must go in, and Terence has even let stand the sentence where he says as much, propero ad filiam (951).

Now Glycerium must also be formally betrothed to Pamphilus. In Menander the betrothal cannot have stood in the recognition scene, where it does in the Terentian adaptation. For one thing the betrothal should follow, not precede, the reunion of father and daughter.Footnote 24 More seriously, Pamphilus cannot have been present in Menander's recognition scene. Three speaking actors was the absolute limit in Hellenistic Athens,Footnote 25 and three characters other than Pamphilus took part in the Menandrian scene: Chremes and Crito, both necessary for the recognition, and Simo, whose presence is indicated by fr. 48 K.–A. = Donat. in An. 919, however exactly it is to be written.Footnote 26 Pamphilus was therefore not on the stage,Footnote 27 and there can have been no betrothal in his absence.

How then were Pamphilus and Glycerium betrothed in Menander? Chremes will have visited Glycerium as she lay in bed at home. Pamphilus was there already: in Terence's version he enters Glycerium's house at 901, just three lines before the recognition/betrothal begins; all we need assume is that in Menander he did not accompany Crito back out of the house as he does in Terence at 904, but remained behind. In Menander, Chremes will have offered Pamphilus Glycerium's hand beside her bed, and Pamphilus will have accepted. For this bedside scene Menander in all likelihood relied on the ekkyklēma, as he did for the bedside scenes in Dyscolus and Aspis. Terence had no ekkykl ēma to work with (and possibly would not have used it if he had); he certainly refused to bring an invalid out of the house without a strong motive. And so he cut the whole scene out.

J.C.B. Lowe saw a part of the truth: ‘It is likely that he [i.e. Terence] has telescoped the action of Menander's final scenes, including the recognition of Glycerium and her betrothal to Pamphilus (cf. 980 intu’ despondebitur), to make room for the conclusion of the Charinus subplot; it is a plausible hypothesis, if not the only possible one, that this led to early introduction of Pamphilus into the Simo-Chremes-Crito scene.’Footnote 28 Eckard Lefèvre saw another part: a Roman poet ‘eilt gern dem Ende zu’.Footnote 29 But these were not the only considerations weighing on Terence, nor the most pressing.

HECYRA V.3 (799–840)

A solution along similar lines is available to long-standing problems posed by the fifth act of Hecyra. In this play, adapted from an original of the same title by Apollodorus of Carystus, Pamphilus’ wife Philumena and her mother Myrrina have tried to conceal Philumena's pregnancy and childbirth, arousing suspicion and distrust of them by others in the family. Pamphilus seeks to divorce Philumena and not to acknowledge her child. His father Laches and father-in-law Phidippus suspect that Pamphilus acts as he does because he has not broken off relations with the courtesan Bacchis, as he should have upon marrying. During Scene 1 of Act V Bacchis offers to affirm her own and Pamphilus’ innocence by oath (750–6); in the following scene she enters Phidippus’ house to make the same offer to Myrrina and to Philumena (793). Returning to the stage in Scene 3 and finding Parmeno in the street, Bacchis asks him to fetch Pamphilus as soon as possible and to tell him that Myrrina has recognized as her daughter's the ring that Pamphilus once gave her. After Parmeno has gone on his errand, Bacchis relates in a monologue (816–40) the full story: the ring on her finger has just now been identified by Myrrina as one that Philumena had lost to a rapist. Bacchis had been given the ring by Pamphilus ten (lunar) months ago; when giving it to her, Pamphilus had confessed that he had wrested it from a girl during a rape. This is the anagnōrisis that makes the happy ending possible, for it establishes that the child just born to Pamphilus’ wife Philumena is indeed his, not another man's, allowing him to keep Philumena and her child. In the following scene, V.4, Parmeno returns with Pamphilus, who is elated by the news of the ring. Parmeno asks to be told what the excitement is all about, but the play ends with Pamphilus and Bacchis resolved to keep their happy secret to themselves.

Bacchis’ monologue in V.3 will be the centre of our attention:

BA quantam obtuli aduentu meo laetitiam Pamphilo hodie!
quot commodas res attuli! quot autem ademi curas!
gnatum ei restituo, qui paene harunc ipsiusque opera periit;
uxorem, quam numquam est ratus posthac se habiturum, reddo;
qua re suspectus suo patri et Phidippo fuit, exsolui:      820
hic adeo his rebus anulus fuit initium inueniundis.
nam memini abhinc mensis decem fere ad me nocte prima
confugere anhelantem domum sine comite, uini plenum,
cum hoc anulo: extimui ilico: ‘mi Pamphile,’ inquam ‘amabo,
quid exanimatu's obsecro? aut unde anulum istum nactu's?     825
dic mihi.’ ille alias res agere se simulare. postquam uideo,
nescioquid suspicarier magis coepi, instare ut dicat.
homo se fatetur ui in uia nescioquem compressisse,
dicitque sese illi anulum, dum luctat, detraxisse.
eum haec cognouit Myrrina in digito modo me habentem     830
rogat unde sit: narro omnia haec: inde est cognitio facta
Philumenam esse compressam ab eo et filium inde hunc natum.
haec tot propter me gaudia illi contigisse laetor:
etsi hoc meretrices aliae nolunt; neque enim est in rem nostram
ut quisquam amator nuptiis laetetur. uerum ecastor     835
numquam animum quaesti gratia ad malas adducam partis.
ego dum illo licitum est usa sum benigno et lepido et comi.
incommode mihi nuptiis euenit, factum fateor:
at pol me fecisse arbitror ne id merito mihi eueniret.
multa ex quo fuerint commoda, eius incommoda aequum est ferre.    840

What joy have I brought Pamphilus today by coming here! How many good turns have I done him, how many troubles spared him! I'm bringing back his son, whom these women, and Pamphilus himself, almost caused to die. I'm giving him back his wife, whom he expected to lose for good. I've cleared him of his father's and Phidippus’ suspicions. And it was this ring that let it all come about. I recall, you see, how ten months ago he came panting into the house a little after dark, all alone and full of drink, with this ring. I was frightened at the first sight of him. ‘Pamphilus dear,’ I said, ‘won't you tell me: why ever are you out of breath? Where did you get that ring? Do say.’ He pretended not to be listening. When I saw that, I grew more suspicious and pressed him to tell. He confessed that he had forcibly raped a stranger in the street, and said he had taken the ring from her during the struggle. Myrrina just now noticed that I was wearing it on my finger. She asked where I had got it; I told her the whole story. That led to a discovery: it was Philumena that he raped, and her son was conceived then. I am happy that he has got so much joy because of me, however much other courtesans begrudge it. It's not to our advantage, you know, for any of our lovers to find happiness in marriage. But I, you may be sure, shall never let hope of gain set wicked thoughts in my heart. I had the benefit of his kindness, his charm and his generosity as long as it was allowed; his marriage was bad news for me, I admit, but I do believe I never did anything to bring it upon myself. When somebody's been good to you, it's only right to accept the bad, too.

The following comment has come down in Donatus’ scholia on 825:

QVID EXANIMATVS AVT VNDE breuitati consulit Terentius, nam in Graeca haec aguntur, non narrantur.

WHY ARE YOU OUT OF BREATH? WHERE Terence is aiming at concision, for in the Greek comedy these things are played, not narrated.

How to understand Donatus? What he would seem at first to imply—that Apollodorus staged an encounter between Pamphilus and Bacchis just minutes after the rape of PhilumenaFootnote 30 —is of course impossible, since the action of Apollodorus’ play would then extend over nine months, a liberty with regard to time not tolerated by ancient drama. W.E.J. Kuiper came up with a clever alternative: it was from Myrrina, in Phidippus’ house, that Bacchis first heard of the rape; suspecting Pamphilus, she left the house, summoned him and questioned him about it on stage.Footnote 31 Philippe-Ernest Legrand in his later work thought similarly, as did J.C.B. Lowe.Footnote 32 Such a view does make good sense of Donatus’ text as transmitted. On the other hand, it is not the efficient solution one would like, for it only attains its end by requiring several actions, and complicated ones, of Terence. Such an interrogation-scene makes not only Bacchis’ monologue at 816–40 Terentian, but likewise lines 811–12 and practically the whole of scene V.4 (841–80). One reason is Pamphilus’ jubilant mood as he enters in V.4 (and Bacchis’, too, when she greets him), which is comprehensible only if the resolution has already been achieved. The other is the slave Parmeno. In Terence's version of V.4, he is on stage throughout; if that was so in the corresponding scene reconstructed for Apollodorus by Kuiper et al., then he was present when Bacchis got the truth from Pamphilus, making him party to the secret. That is exactly what Pamphilus has worked to prevent (especially in III.4), because Parmeno is a gossip (I.2, especially 109–12), making it unrealistic, if Parmeno is in the know, that the two senes and the matron Sostrata can be kept in the dark. In the Terentian play Pamphilus indeed hopes to withhold the secret from these latter (402), and still hopes to after the discovery (865–8); he must have done likewise in the Apollodoran model.Footnote 33 Pamphilus’ and Bacchis’ humour would therefore have to be changed in Apollodorus’ version of V.4 and Parmeno removed from it.Footnote 34 There would then be nothing more left of Terence's scene. Terence's motive, too, for suppressing Bacchis’ interrogation of Pamphilus would be hazy. He might gain concision, it is true, but he would also forfeit a vivid, tense encounter, such as he usually found congenial. It should especially be borne in mind that Bacchis’ monologue is of a length—25 lines—strongly avoided by Terence, and especially toward the end of a play, where he was at pains to maintain a brisk tempo.Footnote 35 Only once elsewhere, at Ad. 855–81, does Terence allow in the fifth act a monologue of comparable length, and I daresay he had an unusually strong motive there to do so.Footnote 36 To place a stoppage in the action at such a point he must have a reason more fundamental and more powerful than to reduce, by three minutes at most, a play that would still be his shortest without the intervention. Finally, the hypothesis of an interrogation raises bothersome questions about choices made in the composition of the original. Apollodorus gave Parmeno an inquisitive character and showed him very interested in the mysterious quarrel in Laches’ and Phidippus’ households (especially clear in scene I.2). At the same time, he gave Parmeno almost all information—though not quite all—needed to make sense of it.Footnote 37 Now if lines 816–40 replace a dialogue in which Bacchis interrogated Pamphilus, we must assume, as has been noted, that Parmeno was not present in the scene of the original corresponding to V.4. Apollodorus would then have prepared Parmeno for frustration, but never followed through by showing him frustrated.

Maria Rosa Posani pointed out that Bacchis’ monologue, known from Donatus to be a locus of intervention by Terence, includes information, namely the history of the ring, that might well have figured in an omniscient prologue, if the original had such.Footnote 38 In her view the story of the ring received a full narration in Apollodorus’ prologue and hence was passed over quickly in a dialogue between Bacchis and Myrrina;Footnote 39 it was for this dialogue that Terence substituted Bacchis’ monologue, with the complete ring-story, to conpensate his suppression of the prologue. Eckard Lefèvre took up the idea and modified it.Footnote 40 Bacchis, he believes, delivered in the original a monologue like 816–40, but without 822–9, the narrative of the scene between herself and Pamphilus the night of the rape. Apollodorus’ audience, as Lefèvre has it, did not need to hear this narrative at that point in the play, since they already knew the gist of it from the omniscient prologue. Bacchis next summoned Pamphilus, as she does in Terence, sending Parmeno to tell him that Myrrina had recognized her ring as Philumena's. When Pamphilus arrived in the scene corresponding to Terence's V.4, he asked ‘How, Bacchis, did you come to have Philumena's ring?’ She answered, ‘Why, don't you remember what happened on that night ten months ago …?’. Having excised Apollodorus’ prologue, Terence needed to place the information given in the Bacchis-Pamphilus exchange at an earlier point, so that his audience, less prepared than that of Apollodorus, would not be puzzled. At the same time he removed it from its place in V.4, so as not to repeat it. Lefèvre's reconstruction does give Terence a stronger motive for intervention than does that of Kuiper, Legrand and Lowe. But it is no more efficient. Again, Pamphilus’ jubilation in the first part of V.4 cannot be Apollodoran, since in Apollodorus he would not yet have recalled the detail of the ring from the night of the rape and pieced together the whole causal chain. Parmeno's presence creates the same difficulties as we have seen in the views of Kuiper et al. Once more, Terence would have to have done far more than turn some dialogue of V.4 into part of a monologue at V.3: rather he would have to have rewritten V.4 de nouo, replacing a scene or scenes that worked very differently and told a somewhat different story.

Some scholars have assumed that the scholium speaking of the Greek original is not in fact a comment on line 825, but rather is misplaced. Line 816, the beginning of Bacchis’ monologue, has been suggested as its true seat,Footnote 41 as have lines 830–2.Footnote 42 Accordingly attempts have been made to understand in Graeca haec aguntur with reference not to Pamphilus’ acquisition of the ring but to the whole narrative section of Bacchis’ monologue. Originals have been reconstructed in which Bacchis told what had transpired indoors not to the audience in a monologue but to PamphilusFootnote 43 or to Parmeno,Footnote 44 placing her report within a dialogue and thus making it a sort of actio; this report would have included a summary of Bacchis’ encounter with Pamphilus just after the rape. The meaning of agere is then stretched uncomfortably, and the problem of Parmeno's knowledge is again raised. It has also been held that Bacchis met Myrrina in front of the house. Flaminio Nencini reconstructed such a scene in some detail,Footnote 45 and has had followers.Footnote 46 He posited that Phidippus went in ahead of Bacchis (as he would need to do, by reason of the three-speaker rule, if Myrrina was to come out after); Myrrina then came out; finally, after the discovery, Myrrina took Bacchis in to meet Philumena. This action would be at odds with the dialogue at 754–5 and 783–92, where Laches and Phidippus have wanted Bacchis not to stay outside, but to go inside. A motive is therefore needed for Bacchis to remain out on the stage; Nencini found one in her shame at showing herself to married women (though in Terence's version she has declared herself ready to do just that at 756–7). A motive, too, is needed for Phidippus, who had agreed that Bacchis should go and meet both women, not to see her in, though less imagination is needed to provide it.Footnote 47 Not to mention Myrrina's entrance onto the stage, for which Nencini himself had no motive to offer. Legrand suggested that Myrrina came out so that she could order Bacchis not to go in.Footnote 48 Walther Ludwig proposed that Phidippus told Myrrina, inside, that Bacchis was about to come in and swear that Pamphilus had not been frequenting her; thereupon Myrrina, who had let Phidippus think Pamphilus’ infidelity her reason for removing Philumena from Laches’ house, feared that her pretence would be exposed; she accordingly came out so as to speak with Bacchis in private.Footnote 49 That Myrrina's thought ran so is indeed sensible in the logic of the story, but how would it have been made clear to the audience? What to do, finally, with Laches’ monologue (794–8) in the scene reconstructed for Apollodorus? It could only have been awkward and inappropriate if Bacchis remained on stage.Footnote 50 Again, the advantage to Terence of making the changes—against his usual policy toward extended monologues in the last act—would be slight. To recapitulate, the hypothesis of a Bacchis-Myrrina dialogue in front of the house in Apollodorus entails stage-action whose motivation is not obvious; it also demands of Terence multiple modifications to V.2 for a paltry overall benefit. After reviewing all attempts to look back to the dialogue underlying Bacchis’ monologue in Terence, one must conclude with Goldberg that ‘no reconstruction has proved satisfactory’.Footnote 51

Now it is not unreasonable in itself to refer in Graeca haec aguntur to a line other than 825. Misplacement of material is never difficult to suppose in the tradition of Donatus, and it is particularly easy when nothing in the scholium itself indicates to copyists or compilers what should be its lemma.Footnote 52 I believe that the scholium belongs to lines 830–2, where it is said that Myrrina noticed the ring and that ‘it led to a discovery’. That is, after all, where Terence's concision is most striking, and the place, if any place there is, to comment breuitati consulit. I propose the following. In Apollodorus Bacchis went into Phidippus’ house toward the end of V.2, just as she does in Terence. Phidippus may have accompanied her or slightly preceded her.Footnote 53 Laches spoke his monologue after her exit, as in Terence. The scene that followed showed only Philumena, bedridden, with Bacchis and Myrrina gathered about her. The recognition came about when Myrrina noticed Bacchis’ ring. Myrrina, and perhaps Bacchis with her, then went elsewhere within the house to find Phidippus, so that Myrrina could declare to him that she was satisfied by Bacchis’ oath.Footnote 54 The rest of the play proceeded as does the Terentian version, except that Bacchis’ monologue in V.3 expressed only her warm thoughts for Pamphilus and left out the narrative of 821–32. In writing the above, Apollodorus made the indoors visible to the audience. That he centred the action about a sickbed suggests that, here too, the ekkyklēma was the means.Footnote 55 Terence rewrote the whole for the same reason as he did Andria V.3-4, because he had no ekkyklēma, and perhaps would have refused one had it been offered. Desire for concision was not a motive for his intervention, though it should not surprise us that a scholiast thought it was. Out of touch with theatrical traditions, the Roman grammatici did not think of the stage's appearance or of its visual conventions: a play was for them mere words, not spectacle. Nor should Hecyra V.3 be taken as evidence that Terence disliked recognitions and sought to keep them off his stage:Footnote 56 the problem for him was not recognition but the fictional space where Apollodorus had set it and the means used to set it there.

References

1 Most often a building such as a temple, palace or house, occasionally a natural shelter such as a grotto or grove.

2 Scholiasts making of the ekkykl ēma a revolving device (in Aesch. Eum. 64; in Ar. Nub. 184; in Ar. Ach. 408; in Clem. Strom. 4.97) would seem to be victims of a confusion. Sources are collected in Csapo, E. and Slater, W.J., The Context of Ancient Drama (Ann Arbor, 1994), 270–3Google Scholar. Thorough and sound discussions will be found in: Arnott, P.D., Greek Scenic Conventions in the Fifth Century B.C. (Oxford, 1962), 78–88 Google Scholar; Hourmouziades, N., Production and Imagination in Euripides (Athens, 1965), 93–108 Google Scholar; Pöhlmann, E., ‘Zur Bühnentechnik im Dionysos-Theater des 4. Jahrhunderts’, in id. (ed.), Studien zur Bühnendichtung und zum Theaterbau der Antike (Frankfurt, 1995), 155–64Google Scholar. Note, too, the bolder speculations of Dearden, C.W., The Stage of Aristophanes (London, 1976), 50–74 Google Scholar, and Hölscher, U., ‘Schrecken und Lachen. Über Ekkyklema-Szenen im attischen Drama’, in Bierl, A. and von Möllendorff, P. (edd.), Orchestra (Stuttgart, 1994), 84–96 CrossRefGoogle Scholar.

3 ‘Visible’ is to be emphasized:  see Hourmouziades (n. 2), 102–4. It should be noted, too, that the exact value of the ekkyklēma can be unstable, even in the course of a scene: Wiles, D., Tragedy in Athens: Performance Space and Theatrical Meaning (Cambridge/New York, 1997), 163–5CrossRefGoogle Scholar; for such shifting Liapis, V. coins the term ‘de-semiotization’ (‘Genre, space, and stagecraft in Ajax’, in Most, G. and Ozbek, L. [edd.], Staging Ajax's Suicide [Pisa, 2015], 121–58, at 127–30)Google Scholar.

4 Csapo and Slater (n. 2), 270–3.

5 Gomme, A.W. and Sandbach, F.H., Menander: A Commentary (Oxford, 1973)CrossRefGoogle Scholar, ad loc. That Menander there used the ekkyklēma has been contested: Del Corno, D., ‘Spazio e messa in scena nelle commedie di Menandro’, Dioniso 59 (1989), 201–11Google Scholar, at 209; Blume, H.-D., Menander (Darmstadt, 1998), 56–7Google Scholar; id. in DNP s.v. ekkyklēma. Del Corno's objection, that as the poet of realism Menander could not have made such a joke about the stage machinery, is obviated by other instances of humour at the expense of dramatic illusion (Dys. 881–2, 910). Blume argued that the ekkyklēma is not known elsewhere to emerge from any but the central door of the set, whereas Cnemon's door in Dyscolus is to the right of the centre. Allowance must first be made for the state of the evidence on the ekkyklēma. This is skewed towards the fifth century, when the central door was by far the most used, and the set may even have had no other. By Hellenistic times, however, at least one theatre (that of Delos) was equipped with multiple ekkyklēmata (IG XI.199. A 94–6); see G.M. Sifakis, Studies in the History of the Hellenistic Drama (London, 1967), 51. Julius Pollux, admittedly writing at some distance from the facts but nevertheless informed by texts now lost, claimed that each door had its own ekkyklēma (4.128). It is conceivable, too, that the ekkyklēma so suspended normal relations of place that it might emerge from a door other than the one assumed to be Cnemon's.

6 Heracles has been brought, asleep, through a door (1029–30), along with a fallen column and the bodies of his children. A rolling platform is the only means that could be used.

7 Aristophanes of Byzantium (fr. 390 Slater) thought, for what it is worth, that Phaedra came out of her palace on the ekkyklēma. That the place is outdoor (178–80) may or may not present a problem, depending on how strict a consistency one requires in the functioning and semiotic value of the ekkyklēma from one author and passage to another. The issues are well reviewed by Belardinelli, A.M., ‘A proposito dell'uso e della funzione dell’ekkyklema: Eur. Hipp. 170–266, 808–1101; Men. Asp. 309–399, Dsc. 689–758a’, Seminari romani di cultura greca 3 (2000), 243–65Google Scholar, at 245–9.

8 As Rees, K. observed in ‘The function of the πρόθυρον in the production of Greek plays’, CPh 10 (1915), 117–38Google Scholar, at 136–7.

9 Jacques, J.-M., ‘Mouvement des acteurs et conventions scéniques dans l'acte II du Bouclier de Ménandre’, Grazer Beiträge 7 (1978), 37–56 Google Scholar, at 54–5. Similar reasoning in Halliwell, S., ‘The staging of Menander Aspis 299 ff.’, LCM 8 (1983), 31–2Google Scholar. Belardinelli (n. 7), 252–61 provides a detailed review of opinions on this passage.

10 In this comedy, the plot of which is sketched in the scholia to Ter. Eun. 9 (Donatus and also an annotation in Paris, BnF MS lat. 7899), a girl is seen periodically through an opening in the party-wall between two houses. A scene where others look at the girl as she occupies a framed space is shown in a mosaic, MNC3 6 DM 2.11, first published by Charitonidis, S., Kahil, L. and Ginouvès, R., Les mosaïques de la maison du Ménandre à Mytilène (Bern, 1970), 60–2Google Scholar with pl. 8.2. It follows that at least one of her appearances through the wall was not narrated in Menander, but enacted; that implies display to the audience of an unambiguously indoor location.

11 Wiles, D., The Masks of Menander (Cambridge, 1991), 49Google Scholar; Jacques, J.-M., ‘La comédie nouvelle a-t-elle utilisé l'eccyclème?’, Pallas 54 (2000), 89–102 Google Scholar, at 95–6. It is rash to reconstruct matters of staging, as Wiles does, from the Mytilenaean mosaics, poor in detail and iconographically corrupt. Jacques's premise that the ekkyklēma was New Comedy's only means to show interiors (see especially page 98) is very doubtful.

12 Webster, T.B.L., An Introduction to Menander (Manchester, 1974), 81Google Scholar. Webster misleadingly attributed the idea to Lilly Kahil.

13 Gutzwiller, K. and Çelik, Ö., ‘New Menander mosaics from Antioch’, AJA 116 (2012), 573–623 CrossRefGoogle Scholar, at 590–7.

14 Philemon 8 K.–A. is a good example: πιεῖν τις ἡμῖν ἐγχεάτω καὶ ματτύην ποιεῖτε θᾶττον or Diphilus 70 K.–A. Ἀρχίλοχε, δέξαι τήνδε τὴν μετανιπτρίδα … . Such fragments are numerous, running to several dozen. They are catalogued and discussed in Konstantakos, I., ‘The drinking theatre: staged symposia in Greek comedy’, Mnemosyne 58 (2005), 183–217 CrossRefGoogle Scholar.

15 Lowe, J.C.B., ‘Aspects of Plautus’ originality in the Asinaria’, CQ 42 (1992), 152–75CrossRefGoogle Scholar, at 171–3 presents evidence for considering all onstage banquets in Plautus as his own additions.

16 Much the same reasoning in Konstantakos (n. 14), 202–3.

17 A mosaic (c.third century a.d.) of Zeugma in Anatolia: Abadie-Reynal, C. and Darmon, J.-P., ‘La maison et la mosaïque des Synaristôsai (Les femmes au déjeuner) de Ménandre’, in Early, R. (ed.), Zeugma: Interim Reports (Portsmouth, RI, 2003), 79–99 Google Scholar. Other representations of this scene do not provide evidence of a method of staging, though some have found such in them. A Pompeian mosaic of it inspired Bulle, H.'s reconstruction of a skēnē equipped with a broad opening between the doors (θύρωμα), through which interiors were displayed: Untersuchungen an griechischen Theatern (Munich, 1928), 277–83Google Scholar. Among critics of Bulle, note especially Moretti, J.-C., ‘Formes et destinations du proskenion’, Pallas 47 (1997), 13–39 CrossRefGoogle Scholar, at 35–7, and Hourmouziades (n. 2), 94.

18 The only source pointing to the ekkyklēma in symposiastic scenes belongs to the Roman milieu in the late Republic, and perhaps to a different dramatic genre: see note 19, below.

19 Cic. Prov. cons. 14 (the man spoken of is L. Calpurnius Piso Caesoninus, cos. 58): itaque ille alter aut ipse est homo doctus et a suis Graecis subtilius eruditus, quibuscum iam in exostra helluatur, antea post siparium solebat, aut amicos habet prudentiores quam Gabinius … ‘and so that other one is perhaps a man of education, well trained by his Greek tutors, with whom he now feasts on the exostra, whereas he used to do so behind the onstage curtain, or perhaps he has wiser friends than does Gabinius …’. Exostra is simply another term for ekkyklēma. We do not know if Cicero had in mind the staging of comedy or of mime.

20 It has even been proposed that most of the scenes at issue (namely all those involving banquets) are free Plautine composition: Lowe (n. 15).

21 Petersmann, H. (ed.), T. Maccius Plautus, Stichus (Heidelberg, 1973), 40–1Google Scholar rescued the illusion in Stich. 87–154 by placing the sisters in the doorway, with the doors swung open toward the outside. But normally Roman house-doors opened inward, as did Greek ( Marquardt, J., Das Privatleben der Römer [Leipzig, 1886], 229Google Scholar; D.–S. 3.607–8; Klenk, H., Die antike Tür [Giessen, 1924], 15–20 Google Scholar), and Roman stage-doors can hardly have done otherwise. The idea of D. Wiles (n. 11), 57 is no better.  He has the father enter the scene-building at orchestra-level (that is to say one storey below the stage), then emerge from inside his daughter's house through the onstage door; parallels, however, are lacking for any action whatever at orchestra-level in Rome, let alone for the orchestra-level exit and offstage climb.

22 See Webster, T.B.L., Studies in Later Greek Comedy (Manchester, 1970), 185–6Google Scholar; Lefèvre, E., ‘Menander’, in Seeck, G. (ed.), Das griechische Drama (Darmstadt, 1979), 307–53Google Scholar, at 342; Hunter, R.L., The New Comedy of Greece and Rome (Cambridge, 1985), 40–2CrossRefGoogle Scholar; and especially Holzberg, N., Menander: Untersuchungen zur dramatischen Technik (Nuremberg, 1974), 121–73Google Scholar, who analyses the function of fifth acts in detail.

23 To enhance readability, the Terentian character-names will be used when their Menandrian equivalents are meant. It is likely that the characters were named differently in Menander.

24 See Fantham, E., ‘Plautus in miniature: compression and distortion in the Epidicus ’, PLLS 3 (1981), 1–28 Google Scholar, at 20.

25 There is no certain case of four speakers on stage in extant New Comedy. See Gomme and Sandbach (n. 5), 16–19; Hourmouziades, N., ‘Menander's actors’, GRBS 14 (1973), 179–88Google Scholar; Sandbach, F.H., ‘Menander and the three–actor rule’, in Bingen, J. et al. (edd.), Le monde grec: hommages à Claire Préaux (Brussels, 1975), 197–204 Google Scholar; Frost, K.B., Exits and Entrances in Menander (Oxford, 1988), 2–3 Google Scholar, 11.

26 Otoy artoc et tin is transmitted, οὕτως οὗτός ἐστιν the likeliest restoration yet proposed; Terence's translation at 919 guarantees that Simo is meant.

27 Similar conclusions in Webster, T.B.L., Studies in Menander (Manchester, 1960 2), 81Google Scholar, Lowe, J.C.B., ‘The Eunuchus: Terence and Menander’, CQ 33 (1983), 428–44CrossRefGoogle Scholar, at 431, and Lefèvre, E., Terenz’ und Menanders Andria (Munich, 2008), 128–9Google Scholar. Lowe thought further that Terence's lines 952–6 were adapted from a Menandrian exit-monologue. It should be supposed rather that the pardon of Davus was elaborated at greater length in Menander (Lefèvre [this note], 161), perhaps taking up a scene in itself. Webster attributed to Menander an exit by Crito in the middle of the scene corresponding to V.4, followed by a lightning-change and entry as Pamphilus. This reconstruction has been rightly criticized for too much movement of characters onto and off the stage (Lefèvre [this note], 128 n. 228).

28 Lowe, J.C.B., ‘Terence's four-speaker scenes’, Phoenix 51 (1997), 152–69CrossRefGoogle Scholar, at 160.

29 Lefèvre (n. 27), 129; see also 161.

30 Once again, the Terentian character-names will be used for the corresponding figures in the model.

31 Kuiper, W.E.J., Two Comedies by Apollodorus of Carystus, Terence's Hecyra and Phormio (Leiden, 1938), 29–35 Google Scholar.

32 Legrand, P.-E., ‘À propos du dénouement de l'Hécyre’, REA 43 (1941), 49–55 CrossRefGoogle Scholar, 132, at 52–3; Lowe, J.C.B., ‘Terentian originality in the Phormio and Hecyra ’, Hermes 111 (1983), 431–52Google Scholar, at 438–42.

33 This is one point on which there is near-general agreement, and for excellent reason. The plot of Hecyra has been crafted so as to place Laches and Phidippus in a state of misunderstanding favourable to Pamphilus, who is of course the focus of our sympathy—a deft inversion of the common ἄγνοια theme, making of error not the obstacle but the means to a happy ending. If Apollodorus had cleared up all misunderstanding at the end, then he would have carefully prepared this brilliant play on expectations only to abandon it at the last minute. In general, Apollodorus seems to have been eager to subvert convention: Gilula, D., ‘Terence's Hecyra: a delicate balance of suspense and dramatic irony’, SCI 5 (1979/1980), 137–57Google Scholar, at 145 n. 15 points out the hushing of Philumena in place of her cries of pain (318), the ring that goes from girl to rapist rather than the other way round, and the ‘running’ of Parmeno, for reasons that are the opposite of those usual. The last of these features is part of a larger inversion: the lead slave who dupes no one, but rather is duped by his master.

34 So Lowe (n. 32), 441. Kuiper (n. 31), 35–47 ascribed to the original an extended fifth act, which he fleshed out with much fancy (Bacchis was recognized as the lost daughter of Laches’ cousin Phania!).

35 Denzler, B., Der Monolog bei Terenz (Zurich, 1968), 103Google Scholar.

36 I have argued elsewhere (‘ Terentius orator an poeta: the endings of Eunuchus and Adelphoe ’, CQ 62 [2012], 671–91CrossRefGoogle Scholar) that Terence intended the monologue at Ad. 855–81 as a dazzling rhetorical showpiece; whether my views on it are accepted or not, this monologue must still be of greatest importance to an understanding of the ending.

37 Note especially vv. 320–1 of Terence's version, where Parmeno has come within a hair of hearing the key to it all, 327–35, where again he has just missed becoming party to the whole secret, and 823, where it is specified that on the night of the rape Pamphilus had no servant with him, thereby excluding Parmeno and any slave who might have informed Parmeno. All this must have been done for a purpose.

38 Posani, M.R., ‘Sui rapporti fra l’« Hecyra » di Terenzio e l’Ἑκυρά di Apollodoro di Caristo’, Atene e Roma 44 (1942), 141–52Google Scholar, at 147–9.

39 Posani (n. 38) followed Nencini's reconstruction of V.2, on which see below.

40 Lefèvre, E., Die Expositionstechnik in den Komödien des Terenz (Darmstadt, 1969), 75–9Google Scholar.

41 The possibility was already raised by Ladewig, T., Beiträge zur Kritik des Terentius (Neu-Strelitz, 1858), 9Google Scholar.

42 Nencini, F., De Terentio eiusque fontibus (Livorno, 1891), 61–2Google Scholar.

43 Könighoff, J., De ratione quam Terentius in fabulis Graecis Latine convertendis secutus est commentatio (Cologne, 1843), 29Google Scholar with n. 39; Stavenhagen, K., ‘Menanders Epitrepontes und Apollodors Hekyra ’, Hermes 45 (1910), 564–82Google Scholar, at 581.

44 Saekel, A., Quaestiones comicae de Terenti exemplaribus Graecis (Berlin, 1914), 86–7Google Scholar.

45 Nencini (n. 42), 61–3.

46 Most notably Legrand, P.-E., Daos: tableau de la comédie grecque pendant la période dite nouvelle (Lyon and Paris, 1910), 480Google Scholar; Lafaye, G., ‘Le modèle de Térence dans l’Hécyre’, RPh 40 (1916), 18–32 Google Scholar, at 30–2; Ludwig, W., ‘review of O. Bianco, Terenzio. Problemi e aspetti dell'originalità (Rome, 1962)’, Gnomon 36 (1964), 152–60Google Scholar; Posani (n. 38); Denzler (n. 35), 13–18.

47 Phidippus has been disdainful of Bacchis throughout the scene: see Goldberg, S. (ed.), Terence: Hecyra (Cambridge, 2013)Google Scholar, ad loc.

48 Legrand (n. 46), 480.

49 Ludwig (n. 46), 157.

50 If Bacchis and Laches were on stage together, Laches’ speech could not resemble what it is in Terence. He could hardly have gloated over the advantages of Bacchis’ friendship while she was still within earshot, or have congratulated himself on the success of his plan while Bacchis was still not sure whether to go through with her part of it (as Nencini had her not be). Nor will Laches have remained on stage through Bacchis’ dialogue with Myrrina; had he done so he would have learned what in the logic of the play he must not—that there had been questions about his daughter-in-law's chastity and that his son was a rapist; there is a similar conclusion in Denzler (n. 35), 18.

51 Goldberg (n. 47), on lines 824–9.

52 Donatian scholia attached to the wrong words or lines, marked in Wessner's edition by cruxes around the lemmata, are not rare, and there must be at least as many more that Wessner never noticed. When a scholium is marked as confused or obscure in the commented electronic edition of Donatus by Bruno Bureau et al. (http://hyperdonat.huma-num.fr/editions/html), it usually becomes comprehensible and clear on the assumption of an erroneous lemma. Nor is it even necessary in this case that fault reside with Donatus’ copyists and adapters. The scholium attached to Hec. 825 and speaking of Terence's model was taken over from an earlier commentator, for the wording ad Hec. praef. I.1, haec fabula Apollodori dicitur esse Graeca excludes direct knowledge of Apollodorus by Donatus, as does the wording in Vita Terenti 10, duae ab Apollodoro translatae esse dicuntur comico (these words are part of the addition by Donatus, not of the ‘Life’ by Suetonius). Not knowing Apollodorus at first hand, Donatus need not have been clear on what the scholium he excerpted meant by ‘these things’, and could easily have set this scholium at the wrong place in his compilation.

53 The Terentian text is itself ambiguous on this point.

54 Ter. Hec. 870–1 Myrrina ita Phidippo dixit iureiurando meo | se fidem habuisse et propterea te sibi purgatum.

55 The three-actor rule will have posed a challenge, which Apollodorus may have met by one of the following means: (1) Laches’ monologue, corresponding to Terence's lines 794–8, took up the necessary stage-time, while the other actors changed and took their places on the ekkyklēma. When this was rolled out, only Myrrina and Philumena were seen, talking to each other. The actor who had played Laches changed costume, then entered as Bacchis, notionally from another room of the house, while the ekkyklēma-scene was in progress. The close of the ekkyklēma-scene coincided with that of the act. (2) Line 798 was the last of its act. The ekkyklēma rolled out onto the stage as the next act opened, with Myrrina, Bacchis and Philumena already in conversation. After the discovery Myrrina exited the stage to search for Phidippus elsewhere in the house, as dialogue between Bacchis and Philumena continued. The actor who had played Myrrina changed to become Parmeno. The ekkyklēma was rolled back into the stage-building as he entered.

56 As it has at times been taken, beginning with Haffter, H., Terenz und seine künstlerische Eigenart (Darmstadt, [1953] 1967), 36–7Google Scholar.