^{page 32 note 1} In this paper footnotes and postiche references have been reduced to the minimum necessary for readers to follow and check the argument. Full bibliographical citation would in any case be superfluous now that we possess H. Marti's unsurpassable reports of recent work on Terence (Lustrum, vi [1961], 114 ff.Google Scholar and viii [1963], 5 ff.), to which must be added a short addendum of indispensable items that have appeared since 1963, such as McGlynn's, P.Lexicon Terentianum (London and Glasgow, 1963 and 1967)Google Scholar and Ludwig's, W. ‘The Originality of Terence and his Greek Models’ (GRBS ix [1968], 169 ff.).Google Scholar I should like to acknowledge two debts of sincere gratitude: to Mr. R. H. Martin, who was kind enough to read and criticize valuably an earlier draft of this paper, and to the Fondation Hardt, which provided with unstinting generosity its idyllic haven at Vandœuvres for the paper's composition.

^{page 32 note 2} Cf. Lofberg, J. O., CP xv (1920), 61 ff.Google Scholar, and CW xxii (1929), 183 f.Google Scholar

^{page 33 note 1} CW xxiii (1930), 122.Google Scholar

^{page 33 note 2} Cf. Haffter, H., Mus. Helv. x (1953), 80 ff.Google Scholar, in a paper reprinted separately in 1967 (at Darmstadt) under its title ‘Terenz und seine künstlerische Eigenart’.

^{page 33 note 3} Schadewaldt's strictures on this alteration (Hermes, lxvi [1931], 1 n. 2Google Scholar) are answered by Marti, , Untersuchungen zur dramatischen Technik bei Plautus und Terenz (Diss. Zürich, 1959), 81 f.Google Scholar; cf. also Jachmann in RE s.v. Terentius, 615 f.Google Scholar; Ludwig, , op. cit. 175 f.Google Scholar; and Williams, G., Tradition and Originality in Roman Poetry (Oxford, 1968), 289.Google Scholar

^{page 34 note 1} The Art of Terence, 75.Google Scholar

^{page 34 note 2} Godsey, E. R., CW xxii (1928), 66.Google Scholar

^{page 34 note 3} Cf. Harrison, A. R. W., The Law of Athens: The Family and Property, 9 ff., 158 f.Google Scholar

^{page 34 note 4} A delightful irony of this apparently fraudulent claim has not to my knowledge been previously noted. If Phanium had really been an ἐπίκληρος (i.e. if Chremes had been dead), then Antipho would have been the legal claimant for her hand (by the rules of ἐπιδικασία) in the absence abroad of his father Demipho, Chremes' brother and next of kin (Harrison, , op. cit. 132 f.Google Scholar). Thus Phormio's perjury achieved a legally correct result: a fact which many of Apollodorus' Athenian audience would doubtless have appreciated.

^{page 36 note 1} Cf. Greece & Rome xv (1968), 15Google Scholar; and Handley, E. W., Menander and Plautus: A Study in Comparison (Inaugural Lecture, University College, London, 1968), 20 n 8.Google Scholar

^{page 36 note 2} Cf. also 319 f., where Phormio's three broken, self-addressed brevities call for suitable (and different) poses by the actor.

^{page 37 note 1} So Martin in his edition, p. 12; but Abel (Gnomon, xxxiii [1961], 94Google Scholar) rightly notes that the idea of ‘working off a debt’ could also have been Greek, cf. Men. Heros, 28 ff.

^{page 37 note 2} Corruption in the MSS. of Donatus here prevents us from knowing for certain who was the author of the cited iambics. One Vatican MS. (V) gives the source of the quotation as de cen.…, which most editors ‘correct’ to de Enn. Sat., in support of an extremely implausible conjecture by Stephanus (cf. Vahlen's edition of Ennius, pp. xxiv f., 206). But a quotation in iambic senarii on the subject of food first suggests a source from Roman comedy. Hence I propose the conjecture de Cen〈tauro τοῦ δεῐ να〉; Lynceus, a contemporary of Menander, wrote a comedy entitled Κενταυρός, from which one fragment remains (iii. 274 f. Kock), and in it a guest discusses food with a cook in a style very similar to the iambics here quoted by Donatus. Was Lynceus' play perhaps adapted into Latin by the author of the Donatus citation?

^{page 39 note 1} Characteristic of modern demotic Greek too, incidentally. Cf. Haffter, , Untersuchungen zur altlateinischen Dichtersprache, 58 f.Google Scholar; Coppola, , Teatro di Terenzio, 113.Google Scholar

^{page 39 note 2} The problems connected with Stilpo's name in this play were first tackled long ago by Bentley (on 11. iii. 9) and Th. Ladewig (Beiträge zur Kritik des Terenz, Progr. Neustrelitz, 1858, 9Google Scholar); but their solution was not finally achieved until Howe's paper in 1913 (Studies in Philology, xi, 61–3Google Scholar), which modern scholarship has scandalously neglected. If some of the details are less clearly etched in the Terentian picture, this is doubtless due to the Latin poet's rejection of the narrative prologue which must have graced Apollodorus' original (so first Tenney Frank, AJP xlix [1928], 321Google Scholar; cf. particularly Kuiper, ‘Two Comedies by Apollodorus of Carystus’, Mnem. Supp. i [1938], 51 f.).Google Scholar

^{page 40 note 1} Cf. Harrison, , The Law of Athens: The Family and Property, 11.Google Scholar

^{page 41 note 1} Cf. Harrison, , op. cit. 135 n. 2.Google Scholar

^{page 41 note 2} Cf. Carm. Epigr. 1523 Bücheler = CIL ix. 2272.

^{page 41 note 3} JRS xlviii (1958), 22 ff.Google Scholar; cf. the same author's Tradition and Originality in Roman Poetry, 378.Google Scholar

^{page 42 note 1} The translation is Lattimore's.

^{page 43 note 1} It is tempting to add that the same technique is adopted in the other play that Terence modelled on Apollodorus, the Hecyra (Pamphilus' monologue, 361 ff.); but here there are some grounds for believing that Terence made changes to his original (cf. especially Schadewaldt, Hermes, lxvi [1931], 1 ff.Google Scholar, and Marti's Lustrum report, 53 f.).

^{page 43 note 2} See the Dziatzko–Hauler commentary, ad loc.

^{page 44 note 1} Pedants, of course, by close examination of the text in their studies, can still detect rough edges in the best-constructed play. Kuiper's list of such inconsistencies in the Phormio (op. cit. 66 ff.) is more useful than some of his explanations for them (cf. Marti, 51 f.).

^{page 44 note 2} Cf. Comfort, CW lii (1959), 248 f.Google Scholar

^{page 44 note 3} But the difficulty is not that of eliminating Antipho from this scene (which is easy), but rather that of imagining what would then have happened in 682 ff., where the dialogue depends upon what Antipho overheard in 606 ff. Yet cf. Nencini, , De Terentio eiusque fontibus (Livorno, 1891), 12 f.Google Scholar

^{page 45 note 1} e.g. Dziatzko–Hauler and Martin on 639–40.

^{page 45 note 2} Cf. also the penetrating remarks of Haffter, , Mus. Helv. x (1953), 97.Google Scholar

^{page 46 note 1} Yet even there his occasional comments bear the hallmark of authority. It is Phormio (not Antipho, note) who brusquely tells Geta at 857 to come to the point after the conventional delaying tactics of the seruus currens.

^{page 47 note 1} There is no warrant for dragging in any reference to Roman funeral orations, as commentators do here. The same proverb occurs twice in Plautus, Bacch. 519 and Poen. 840, in the former passage being so articulated as to preclude any necrological interpretation. It is either translated from or equivalent to the Greek proverb νεκρῷ λέγειν μύθους εἰς ος, which exists in several variants. Cf. Otto, Die Sprichwörter und sprichwörtlichen Redensarten der Römer, 229 f.Google Scholar; McGlynn, s.vv. VERBUM, MORIOR; Handley, , Menander and Plautus, 13 and 20 n. 9.Google Scholar

^{page 48 note 1} Cf. Martin's commentary on 841.

^{page 49 note 1} In 888–9 the punctuation is uncertain, but it seems better to take ingratiis with both clauses: ‘just as this money was given (sc. in the first place) against their will, it'll be given to him equally against their will’. Ingratiis is intentionally positioned at the end of the ut clause and at the beginning of the resumption of the main clause just in order to facilitate the άττὸ κοινοῦ construction. Cf. Sloman's little school edition, ad loc.

^{page 50 note 1} As an illustration of this point, Phormio's fraudulent words at 923 agree precisely with his previous statement as reported by Geta at 661 ff.

^{page 50 note 2} It is one of this play's many delightful touches of irony that at this moment it is Phaedria and his citharistria who are gallivanting at Phormio's house (cf. 835 ff.). Demipho has hit the wrong nail on the head.

^{page 51 note 1} Thus Donatus notes ad loc. (cf. his note on And. 206) on enim uero, ‘principium aliquid per iracundiam dicturi’.

^{page 51 note 2} p. 81. Cf. also Marouzeau's edition, ad loc. (p. 188, n. 4).

^{page 52 note 1} Cf. Haffter, , loc. cit. on p. 33, n. 2.Google Scholar

^{page 53 note 1} The former, depending as it does on an echo possible only in Latin, is as surely Terentian in origin as the latter is Menandrean (cf. Webster, , Studies in Menander, 68).Google Scholar

^{page 54 note 1} It is not accidental that his very first observation is a sentence eight lines long, full of qualification and empty of novelty.

^{page 54 note 2} Cf. Greece & Rome, xv (1968), 9.Google Scholar

^{page 54 note 3} But contrast Siess's more sympathetic analysis of his character in Wein. Stud. xxix (1907), 83 f.Google Scholar

^{page 55 note 1} His Kommentar zu Menanders Dyskolos (Paderborn, 1965), on v. 53.Google Scholar

^{page 56 note 1} The attribution here to Pyrrhias is not completely certain.

^{page 56 note 2} Greece & Rome, xv (1968), 14 f.Google Scholar

^{page 56 note 3} Cf. Webster, , Studies in Later Greek Comedy, 65 f., 194Google Scholar; and Dohm, , Mageiros, 93 ff.Google Scholar

^{page 57 note 1} FCG i. 419 f.Google Scholar