^{page 88 note 2} Jones and Morey, ii. 4.

^{page 88 note 3} Marti, H., ‘Terenz 1909–1959’, Lustrum vi (1961), 119.Google Scholar

^{page 88 note 4} Most work has been done on the illustrated and related manuscripts. See Webb, R. H., ‘An attempt to restore the γ archetype of Terence's manuscripts’. H.S.C.P. xxii (1911), 55–110; Jones and Morey, ii. 195–221.Google Scholar

^{page 88 note 5} See Wessner's remarks in his edition of Donatus, i (Leipzig, 1902), pp. xliv-xlviii; Mountford, J. F., The scholia Bembina (London, 1934), 119; Leo, 323 ff.Google Scholar

^{page 88 note 6} For the purposes of this article I ignore the ‘mixed’ group of the Calliopians.

^{page 89 note 1} There were certainly manuscripts in antiquity that differed in greater or lesser degree from the Bembinus and the archetype of the Calliopians, but it is impossible to tell what their relationship was with the surviving branches of the tradition. See Prete, S., P. Terenti Afri Gomoediae (Heidelberg, 1954), 35–8; Jachmann, 88.Google Scholar

^{page 89 note 2} Jachmann, 83 ff.. 134–5.

^{page 89 note 3} Marouzeau, J., Térence. Comédies, i (Paris, 1942), 85–7.Google Scholar

^{page 89 note 4} Andrieu, J., Étude critique sur les sigles de personnages et les rubriques de scène dans les anciennes editions de Térence (Paris, 1940), 121. Andrieu, however, tied his conclusions to the acceptance of Marouzeau's view that therewere no significant common errors in A and Σ. Many of his conclusions are compatible with a later separation. Marouzeau never published a detailed study of the common errors to show that they were not errors at all or that they were ‘quasi fatales’, made independently in the two branches of the tradition.Google Scholar

^{page 89 note 5} Craig, J. D., Ancient editions of Terence (Oxford, 1930), 130.Google Scholar

^{page 89 note 6} Jachmann, 111–12, 118.

^{page 89 note 7} Craig was often unconvincing in his explanation of Calliopian readings that are found in antiquity. For criticism see Jones, L. B., ‘Ancient texts of Terence’, C.P. xxv (1930), 318–27.Google Scholar

^{page 90 note 1} See Wessner, P., Gnomon iii (1927), 343;Google ScholarPasquali, G., Storia della tradizione e critica del testo² (Florence, 1952), 363–4.Google Scholar

^{page 90 note 2} Wessner, Gnomon iii (1927), 344; Pasquali, op. cit. 365.Google Scholar

^{page 90 note 3} See especially Byvanck, A. W., ‘Antike Buchmalerei, II: Das Vorbild der Terenzillustrationen’, Mnemosyne vii (1939), 115–35, who showed that stylistically the illustrations are closely related to the miniatures of the Virgil Vaticanus and to the mosaics in the church of Santa Maria Maggiore. He dated the Terentian miniatures to 410–20. Jones and Morey, ii. 45, 117, 200, assigned the hyparchetypes γ² (of C and P) and γ¹ (of F) to c. 500 and c. 450 respectively and thearchetype to the first half of the fifth century. They were reluctant to place the archetype any earlier because of the connections which they believed existed between the illustrations and the Graeco-Asiatic style. Byvanck, 132–4, rejected this association. G. Rodenwaldt, ‘Cortinae. Ein Beitrag zur Datierung der antiken Vorlage der mittelalterlichen Terenzillustrationen’, N.G.G., 1925, Phil.-Hist. Kl., 33–49, assigned 350 as a terminus post quem for the archetype on the basis of the curtains that appear on the miniatures.Google Scholar

^{page 90 note 4} Marti, op. cit. 120; cf. Pasquali, op. cit. 363: ‘A ogni modo l'originale delle illustrazione è divenuto ora almeno un terminus ante quem per la datazione di γ’.

^{page 91 note 1} Kauer-Lindsay in the Oxford Classical Text; Marouzeau in the Budé edition; Pratesi, A., Terenzio. Commedie, ii (Rome, 1952);Google ScholarBianco, O., Terentius. Adelphoe (Rome, 1966).Google Scholar

^{page 91 note 2} Marouzeau and Sargeaunt (in the Loeb) interpreted the stage action quite differently. They believed that Syrus entered alone and handed the provisions inside to the slaves. But if this were so, one might expect Syrus to call on Dromo and Stephanio to come and collect the food. It is true that these two attendants have not been mentioned earlier in the play, but the question of where they came from or how they met in with Syrus is unimportant compared with the evidence of the text itself of this scene.

^{page 91 note 3} For this use of redire alone in the sense ‘come/go home’ cf. Eun. 811.

^{page 92 note 1} Only Dromo is named along with Demea and Syrus in the scene headings at line 355 in the Bembinus, which has no scene division at 364.

^{page 92 note 2} Cf. the figure of Lesbia at And. iii. 2 (481 ff.), who is drawn in a more static fashion but who is pointing and also looking inside the house, and that of Mysis at And. iv. 2 (684 ff.), who is facing the other characters on stage but is pointing inside.

^{page 92 note 3} Similar to the striding figure of Syrus here arc the figures of Geta at Phorm. I. 2 and of Syrus himself at Ad. II. 2. The manner in which Syrus is drawn is not confined, however, to characters who are entering. The figure which most closely resembles that of Syrus at Ad. 364 is that of the same character at Ad. 3 (254 ff.) where he is not entering. This latter miniature depicts the end of the scene where Ctesipho hears the sound of the door and Syrus calls on him to wait. Syrus apparently is moving towards Ctesipho to try and prevent him from leaving. That Syrus' posture is typical of anentering character is shown, however, by the fact that in the miniature at Ad. 3 in the codex Bodleianus Auct. F. 2. 13 (O) Syrus is clearly drawn as if he is entering from the house.

^{page 92 note 4} It is also possible that Dromo was not named in the scene heading and that the original illustrator drew him to indicate what was happening inside the house, as at And. iii. 1, where Archylis and Glycerium are drawn behind the door. Even if this were so, the presence of Dromo and the absence of Stephanio would still indicate that the illustrator was depicting the action at 376 ff.

^{page 92 note 5} Jachmann, 16 ff., cited this miniature as a classic example for proof that the illustrations were not based on observation of stage productions of the plays.

^{page 92 note 6} A contributing factor may have been a misunderstanding of ceteros in line 376. The illustrator may have taken this to mean that Dromo was already engaged in cleaning fish.

^{page 93 note 1} Cf. Haut. 1001: ad Menedemum hunc pergam A: ad Menedemum hinc pergam δ: hinc nunc ad Menedemum pergam γ. See Jachmann, 128.

^{page 93 note 2} Jachmann, 99–106.

^{page 93 note 3} Departures from the usual order in the miniatures are discussed by Watson, J. C.,‘The relation of the scene headings to the miniatures in manuscripts of Terence’, H.S.C.P. xiv (1903), 55–172, esp. 71 His conclusion, however, that the scene headings are derived from the miniatures is unconvincing. There are a few departures from the normal order in the scene headings in the Bembinus: cf. And. v. 5; Haut. ii. 3; iv. 7; Phorm. 1; ii. 3; Ad. ti. i; II. 3; 4; v. 2. The view of Kauer, Burs. Jahresb. cxliii (1909), 192–3, that the Bembinus betrayed the influence of the illustrations in some of these cases, was refuted by Jachmann, 90 ff.Google Scholar

^{page 94 note 1} Cf. Watson, op. cit. go, 171.

^{page 94 note 2} Ad loc.: ‘Nihil quicquam vidi absurdius: ut equidem suspicer hominem flore Liberi uvidum hoc scripsisse’.

^{page 94 note 3} Bethe, E., ‘Die antiken Terenz-Illustrationen’, Arch. Jahrb. xviii (1903), 96.Google Scholar

^{page 95 note 1} Leo, G.G.A. 1903, 996, had thought that a ‘Regisseur’ was responsible for the assigning of the words to Clitipho. O. Engelhardt, Die Illustrationen der Terenzhandschriften (Inaug.-Diss. Jena, 1905), pp. 70–1, suggested Calliopius.

^{page 95 note 2} Kauer, op. cit., 199.

^{page 95 note 3} C. Robert, Die Masken der neueren attischen Komōdie (Halle, 1911), 99.

^{page 95 note 4} Idem, Archaeologische Hermeneutik (Berlin, 1919), 189. Here Robert is more accurate(or at least less ambiguous) in his description of the action but less specific in relating the miniature to a particular part of the text.

^{page 95 note 5} Jachmann's criticism here seems to be based on his unwillingness to accept the possibility that some miniatures were combined in the course of transmission with the loss of scene divisions and of figures (such as Syrus here) who appeared in both or all of the pictures so combined. Note also the petitio principii in an argument in a later work of Jachmann: ‘ante Phorm. v. 3, 12 et Ad. v. 5 artificem antiquum in editione sua scaenarum interstitia nulla invenisse eo evincitur quod iam ante Phorm. v. 3, 1 Demiphonem Nausistratam Chremetem depinxit Chremete V. 3, 12 demum scaenam intrante …' (Terentius. Codex Vaticanus latinus 3868 … phototypice editus. Praefatus est G. Jachmann [Leipzig 1929], 13).

^{page 96 note 1} Cf. And. III. 2; IV. 2; IV. 4; Haut. iv. 4; Eun. III. 2; Hec. III. 4; v. 4; Ad. v. 2.

^{page 96 note 2} In the miniatures there are two other examples of a character holding on to another's cloak. At Ad. v. 2 Syrus is trying to hold back Demea and prevent him from entering the house; cf. 780-I: ‘DE. mitte me … non manum abstines?’ At Phorm. v. i Chremes is pulling Sophrona away from the door (cf. 741).

^{page 96 note 3} Kauer, op. cit. 210.

^{page 97 note 1} Op. cit. 13.

^{page 97 note 2} On the miniature at Phorm. v. 3 see Watson, op. cit. 151–2. At Haut. v. i and v. 2 the discrepancies between the illustrations in C and P show that conflation has occurred in at least one if not both of the manuscripts, though at a later stage in the transmission.

^{page 98 note 1} Donatus indicates a scene division at And. 722 after the short monologue of Mysis. The Calliopians have combined this brief scene (716–22) with the following one. At Ad. 511 the γ manuscripts do not have a new scene for the short monologue of Hegio (511–6). In this article I have followed the view of Jachmann, pp. 50 ff., that any scene division found in a group of manuscripts is to be regarded as an ‘original’ scene division. Andrieu, J., Le dialogue antique (Paris 1954), 155–6, thought that Jachmann was too dogmatic and believed that scene divisions were added as well as lost. Jachmann excluded too rigorously, perhaps, the possibility of any addition (e.g. with respect to a scene division at Haut. 593, found only in F). But as far as the textual tradition of Terence is concerned I believe that Jachmann is nearer the truth than Andrieu, whose arguments often support rather than contradict Jachmann's thesis.Google Scholar

^{page 98 note 2} At Haut. 562 and And. 533 the scene division does not occur immediately before the first words of the entering character butbefore words spoken by a character who has remained on stage from the preceding scene. The ‘normal’ division would have been at Haut. 563 and in the middle of And. 533: cf. Ad. 81, 364, 636, Haut. 954, Hec. 767. At Haut. 980, where the metre changes, a new scene is indicated in D, G, and p, although a more natural place would have been at line 978 where Chremes leaves the stage and where Kauer-Lindsay place it. At Eun. 909 Thais and Chaerea go into the house, leaving Pythias alone on stage. She speaks to herself for two and a half lines, then Chremes enters with the nurse in the middle of 912. One might expect the scene division to coincide with Chremes' arrival, but the scene begins at 910.

^{page 98 note 3} Note the scene change (γ) at Eun. 943 where Pythias begins to play her trick on Parmeno. It is true that the metre also changes at 943 but the change of direction may have been a factor: cf. Plaut. Aul. 537, where there is a new scene in similar circumstances but no change in metre.

^{page 99 note 1} If there were scene headings at 376 and 381, the former probably contained only the names of Clitipho and Syrus. Suppression of the former would not have affected the scene heading of the previous scene: cf. the apparent loss of the same two names in A at Haut. 980.

^{page 99 note 2} It is possible, however, that neither Clinia nor the two women were drawn.

^{page 99 note 3} Study of the miniatures at Haut. v. i and v. 2 shows, I believe, that later illustrators used the pictures at their disposal for different purposes when pictures were combined. The figure of Chremes in v. 1 (P) is the same as the figure of Menedemus in v. 2 (C). See Jachmann, 138–48.

^{page 100 note 1} There is an interesting exception to this at Phorm. 795: ‘NAYS. faciam ut iubes. sed meum virum ex to exire video? CHR. ehem Demipho’. The Bembinus has the scene division before this line. D, however, places it immediately before Chremes' first words. The position in D would be the more normal. Both manuscripts, however, name Nausistrata first in the heading and this suggests that the scene may have begun originallybefore ‘sed meum virum … video’. Thus ‘regularization’ has occurred in D. Unfortunately, the y manuscripts do not mark a scene division here. Against this one example we should take into account the fact that the different branches of the tradition are usually in agreement, even where the scene begins at an unusual place, as at And. 533 and Haut. 562, or in the middle of the line, as at And. 581, Ad. 81, 364, 635.

^{page 101 note 1} This raises the question of the extent to which the scene division in the first illustrated Γ manuscript was influenced by the presence and position of the miniatures in Ψ. It is possible that scene divisions whichhad been lost in Σ were reinstated.

^{page 101 note 2} This discussion is based on the supposition that Ψ is derived from Φ. It is possible that this is not the case.

^{page 101 note 3} See Marti, op. cit. 120.

^{page 102 note 1} D. Klose, Die Didaskalien and Prologe des Terenz (Diss. Bamberg, 1966), 17 ff., suggests that Eunuchus was composed before Hautontimorumenos but produced third after Andria and Hautontimorumenos. This view is based on the information in the didascaliae that Eunuchus was ‘facta II’ and that Hautontimorumenos was ‘facta III’. Leo, p. 324, thought that this information was the work of a later ‘Grammatiker’.

^{page 102 note 2} Leo, 319.

^{page 102 note 3} The order in the Bembinus seems to be the one followed in the Suetonian life of Terence (see Leo, 320) and may have been the order of plays in Σ as well as in Φ. The order of plays in the δ manuscripts is almost certainly late and probably connected with the strange position of the didascaliae (after the prologues) in the δ manuscripts. Jachmann, 131–2, made the attractive suggestion that the didascaliae in the δbranch were originally taken over from an ancestor of the γ manuscripts. Presumably the model of Δ had no didascaliae and their absence may have caused the ordering of the plays into alphabetical sequence.