⁵⁶ L. Quicherat had mistaken them for Isid. Orig. 1.21 (de notis sententiarum). See Bibliothèque de l'Ecole des Chartes 1 (1839/1840), 52Google Scholar. They had, however, been recognised and used by Tassin, R. P. and Toustain, C. F., Nouveau traité de diplomatique (Paris, 1757), iii. 483–7Google Scholar (cf. Villoison, op. cit. [n. 44], proleg. xxii).

⁵⁷ ‘Anecdoton Parisinum’, Zeitschr. f d. Alt. 3 (1845), 81–131Google Scholar (see Mommsen, 's Gesammelte Schriften VII [Berlin, 1909], 217–18Google Scholar and Bergk, 's Kleine philologische Schriften I [Halle, 1884], 580–612Google Scholar). The material is printed by Osann, F., Anecdotum Romanum de notis ueterum criticis, inprimis Aristarchi Homericis et Iliade Heliconia (Giessen, 1851), 327–34Google Scholar, Reifferscheid, A., C. Suetoni Tranquilli… Reliquiae (Leipzig, 1860), 137–41Google Scholar, Nauck, A., Lexicon Vindobonense (Leningrad, 1867), 278–82Google Scholar, Dindorf, W., Scholia Graeca in Homeri Iliadem I (Oxford-Leipzig, 1875), xlvii–xlixGoogle Scholar, Keil, H., Grammatici Latini VII (Leipzig, 1880), 533–6Google Scholar, Funaioli, H., Grammaticae Romanae Fragmenta I (Leipzig, 1907), 54–6Google Scholar (omitting the notae simplices). For Mommsen's pride in his ‘discovery’ and Jahn's congratulations see Mommsen, T. and Jahn, O., Briefwechsel 1842–1868, ed. Wickert, L. (Frankfurt, 1962), 13–15Google Scholar. Some of the material has turned up again in a ninth-century manuscript from Benevento, cod. Rome, Bibl. Casanatense 1086 (see Morelli, C., Rendiconti della Reale Accad. dei Lincei, Cl. Sc. Mor. Stor. e Filol. 5th ser., 19 [1910], 287–328)Google Scholar. The two manuscripts drew upon a common source (see Holtz, L., Stud. Med. 3rd ser., 16 [1975], 142–5Google Scholar, Jocelyn, H. D., CQ n.s. 30 [1980], 394–6CrossRefGoogle Scholar).

⁵⁸ Casaubon, I. gathered most, if not all, of the material available in his time (Notae ad Diogenis Laertii libros de uitis, dictis, et decretis principum philosophorum [Morsee, 1583], 119–22Google Scholar). This was supplemented by de Montfaucon, B., Palaeographia Graeca (Paris, 1708), 186–90Google Scholar, Hexaplorum Origenis quae supersunt i (Paris, 1714), 38–42Google Scholar, Tassin and Toustain, loc. cit. (n. 56), Villoison, op. cit. (n. 44), proleg. xiii–xxii, Gräfenhan, A., Geschichte der klassischen Philologie im Alterthum ii (Bonn, 1844), 92–9Google Scholar.

⁵⁹ 3.65–6 (probably from the Βίοι of Antigonus of Carystus). The Greek text of Diogenes' Φιλισόɸων βίων κα⋯ δογμάτων συναγωγή was printed at Basle in 1533. Related in some way to Diogenes' account are those in cod. Cava dei Tirreni, Arch. S. Trim. 3, fol. 255 (Reifferscheid, A., RhM 23 [1868], 131–2Google Scholar) and a second-century a. d. Florence papyrus (Bartoletti, V., Mélanges E. Tisserant i [Vatican City, 1964 (Studi e Testi, 231), 25–30]Google Scholar).

⁶⁰ Pp. 73–6 Consbruch. The Ἐγχειρίδιον περ⋯ μέτρων, and attached essays were printed at Florence in 1526. Demetrius Triclinius had known and used Hephaestion's work at the beginning of the fourteenth century (see Irigoin, J., Les scholies métriques de Pindare [Paris, 1958], 93–105)Google Scholar.

⁶¹ 1. 21. The Origines was printed at Augsburg in 1472 (‘Etymologiae’).

⁶² ii 102.32, iv 581.21 Adler. The work (‘Suidas’) was printed at Venice in 1499.

⁶³ Pp. 136.13, 599.33, 957.16, 1015.23, 1610.46, 1627.59, 1921.55 of the edition brought out at Rome in 1542 and 1550.

⁶⁴ Cf. schol. Hes. Op. 207–12, 276b, 649a, Theog. 117, 573, Pind. Ol. 2.48 f., 10.78 f., Pyth. 3.18a, 4.305, 318a, 431, 507, Aesch. Prom. 9, Sept. 79, Choeph. 534, Soph. Ai. 962, Ant. 735, 741, 1176, Trach. 402, O.C. 25, 43, 237, 375, 1494, 1740, Phil. 201, 417, Eurip. Hec. 3, 4, 29, 323, Orest. 81, 599, Hipp. 93, 1192, Med. 33, 1346, Andr. 603, 873, 930, Rhes. 41, 239, 716, Aristoph. Av. 76, 107, 204, 302, 1309, 1372, Equ. 721, Lys. 499, 702, Nub. 518, 562, 766, 815, 925, 962, 1178, Pac. 775, 990, Plut. 3, 863, Ran. 35, 153, 557, 575, Thesm. 924, Vesp. 1172, 1282, 1480. Scholia to Aristophanes were printed in 1498 (Venice), to Pindar in 1515 (Rome), to Sophocles in 1518 (Rome), to Euripides in 1534 (Venice), to Aeschylus in 1552 (Venice). Colometrical remarks by Triclinius frequently got mixed with ancient material in these editions. Scholia to Hesiod were printed in 1537 (Venice). The Homeric scholia printed in 1517 (Rome) were those long attributed to ‘Didymus’; on the history of knowledge of more erudite sets see Villoison, op. cit. (n. 44), proleg. xiv n. 1. There was also knowledge of the use of σημεῖα on Demosthenes (Schol. ix 587.25–6 Dindorf; ‘Ulpian’ was printed in 1503 [at Venice]) and on Hippocrates (Galen, xv 110 [= CMG v 9, 1, 58], xvi 800 [= CMG v 9, 2, 154]; the commentaries on Hipopocrates were printed in 1525 [at Venice]), perhaps also of their use on Thucydides (schol. 3.84.1; ‘σχόλια παλαιά’ were printed in 1526 [at Florence]).

⁶⁵ Att. 8.2.4, Fam. 9.10.1. Cf. Pis. 73, Fam. 3.11.5, 9.16.4.

⁶⁶ Orig. Comm. Matth. 15.14 (Origines Werke. Zehnter Band. Origines Matthäuserklärung, hrsg. Klostermann, E. [Leipzig, 1935], 388Google Scholar), Epiphanius, De mens. 2–3, 7–8, 17 (de Lagarde, P., Symmicta ii [Göttingen, 1880], 153–70Google Scholar; Dean, J. E., Epiphanius' treatise on weight and measures. The Syriac version [Chicago, 1935], 16–34)Google Scholar, Jerome, , Praef. interpr. Pent. in Bib. Sacr. ed. Bened. i 64Google Scholar, Iob ix 69–70, Paralip. in Patrol. Lat. xxviii 1393, Epist. 57.11, 106.7, 112.19, Augustin. Ciu, 18.43.

⁶⁷ Jerome, , Praef interpr. Psalm x 3–4Google Scholar, Salom. xi 6, Dan. in Patrol. Lat. xxviii 1359A, Epist. 112.19, Augustin. ap. Jerome, , Epist. 104.3Google Scholar (= Epist. 71.3).

⁶⁸ Inst. 1.1.8, 1.9.3. See also Inst. 1 praef. 9, 1.26.

⁶⁹ Schoppe found in his ‘schedae Fuldanae’ et hinc…parcas’ in Probi adpuncti sunt. See his De arte critica commentariolus (Nuremberg, 1597)Google Scholar, sig. B7. Daniel's reading, et ‘hinc…parcas’ adiuncta sunt, held sway until Bergk made his study of cod. Kassel, Staatsbibl. MS. Poet. fol. 6 (‘Servii Casselani part. III’, Progr. Marburg. 1844, 4).

⁷⁰ See above, n. 35.

⁷¹ See above, n. 44.

⁷² On this passage (fol. 8r [foll. 7 and 9 are missing]) see Cobet, C. G., Mnemosyne 2, 1(1873), 26–34Google Scholar, W. Dindorf, op. cit. (n. 57), 1–2, Scholia Graeca in Homeri Iliadem iv (Oxford-Leipzig, 1878), 394–5Google Scholar, L. Friedländer, Ind. Lect. Königsberg (1876), 4, A. Ludwich, op. cit. (n. 25), 61–4, Erbse, H., Scholia Graeca in Homeri Iliadem i (Berlin, 1969), lxiv–lxviGoogle Scholar.

⁷³ See Villoison, op. cit. (n. 44), proleg. xl. The material was published in the same year by Siebenkees, J. P. in Bibliothek der alten Litteratur und Kunst, Drittes Stück (Göttingen, 1788), 71–2Google Scholar. It is reprinted in Osann, op. cit. (n. 57), 5–8, Reifferscheid, O. cit., 143–4, Nauck, op. cit., 274–6, Dindorf, op. cit., xliv–xlv. Cod. Oxford, Bodl. Libr. Auct. T IV 9 (Gaisford's ‘codex Saibantianus’) is a copy of the Venice manuscript.

⁷⁴ Prefixed to a fifteenth-century text of the Iliad (in a hand of the next century). See Cramer, J. A., Anecdota Graeca iii (Oxford, 1841), 293Google Scholar. The material in question is reprinted in Osann, op. cit. (n. 57), 8, Reifferscheid, op. cit., 144, Nauck, op. cit., 277, Dindorf, op. cit., xlvi.

⁷⁵ Prefixed to an elementary commentary on the Iliad copied in South Italy between 905 and 915. For the material in question see Osann, op. cit., 3–5, Reifferscheid, op. cit., 141–3, Nauck, op. cit., 271–3, Dindorf, op. cit., xlii–xliv, Gardthausen, V., Griechische Palaeographie ii ² (Leipzig, 1913), 411–12Google Scholar, and plate xvi opposite p. 74 of Devreesse, R., Introduction à l'étude des manuscrits grecs (Paris, 1954)Google Scholar. On the relationship between this material and that offered by cod. Venice, Bibl. Marc. gr. 483 see Lameere, W., Aperçus de paléographie homérique (Paris-Brussels, 1960), 244–8Google Scholar.

⁷⁶ Item 102 of the inventory. See Peyron, A., M. Tullii Ciceronis Orationum pro Scauro, pro Tullio et in Clodium fragmenta inedita…idem praefatus est de Bibliotheca Bobiensi, cuius inuentorium edidit (Stuttgart and Tübingen, 1824), 23–30Google Scholar.

⁷⁷ For the essay περ⋯ τ⋯ν ⋯μɸερομένων σημείων in cod. Florence, Bibl. Med. Laur. lix 38 (15th cent.), fol. 428^v, see Wachsmuth, C., RhM 18 (1863), 180–1Google Scholar; for the essay περ⋯ τ⋯ν σημείων τ⋯ν κειμένων ⋯ν τοῖς τ⋯ν ⋯ξαπλ⋯ν Ὠριγένους μεταγραɸεῖσι βιβλίοις in cod. Mt Athos, Mon. Vatopedi 507 (12th cent.) and cod. Vatican City, Bibl. Apost. Vat. 2200 (Columnensis 39) (8th-9th cent.) see Serruys, D., Mél. d'Arch. et d'Hist. 22 (1902), 189–93Google Scholar, Diekamp, F., Doctrina Patrum de incarnatione Verbi (Münster, 1907), 248–9Google Scholar; for the essay de notis antiquorum in cod. Cava dei Tirreni, Archivio dell'Abbazia di SS. Trinità 3 (11th cent.), fol. 255^r, see Reifferscheid, A., RhM 23 (1868), 127–33Google Scholar; for the annotated list of signs in cod. Munich, Bayer. Staatsbibl. Lat. 14429, fol. 122^r–v (10th cent. from S. Emmeram) see Kettner, H., ‘Kritische Bemerkungen zu Varro und lateinischen Glossaren’, Progr. d. Klosterschule Rossleben 1868, 33–5Google Scholar, Weber, P., Quaestionum Suetonianarum capita duo (Diss. Halle, 1903), 8–13Google Scholar. Traube (see n. 81) mentions similar material in cod. Boulogne-sur-mer 44.

⁷⁸ See Maass, E., Hermes 19 (1884), 108–9Google Scholar. Cf. Commentariorum in Aratum Reliquiae (Berlin, 1898), 140–1Google Scholar.

⁷⁹ On manuscripts of Homer see Allen, T. W., PBSR 5 (1910), 31–3Google Scholar, Homer, The Origins (n. 25), 314 n. 1, Homeri Ilias i (n. 52), 196–9, Collart, P., RPh 3,7 (1933), 39–40, 3, 13 (1939), 306Google Scholar; of iambic and lyric poetry Fowler, R. L., ZPE 33 (1979), 24–8Google Scholar. On copies of the fifth column of Origen's Hexapla see Metzger, B. M., Manuscripts of the Greek Bible (New York-Oxford, 1981), 38Google Scholar. On Greek manuscripts generally see Devreesse, op. cit. (n. 75), 74–5, 87, 113–14, 133–4, 169, E. G. Turner, op. cit. (n. 25), 146–52 (= Greek Papyri, 112–18).

⁸⁰ Various signs are visible in the margins of a fifty-century Virgil (cod. Florence, Bibl. Med. Laur. xxxix.1 = CLA iii 296; see Ribbeck, O., Prolegomena critica ad P. Vergili Maronis opera maiora [Leipzig, 1866], 158–63Google Scholar), a fifth-to sixth-century Gaius (P. Soc. Ital. 1182 = CLA iii 292), a fifth- to sixth-century Juvenal (JEA 21 [1935], 199–207Google Scholar = CLA Suppl. 1710) and a sixth-century Hilarius (cod. Vienna, Oest. Nationalbibl. MS 2160 = CLA x 1507). Where medieval and renaissance manuscripts are concerned, see on cod. Munich, Bayer. Staatsbibl. Lat. 816a (Lucretius: fifteenth century Italy) Bergk, T., NJbb 83 (1861), 317–20Google Scholar (= Kl. phil. Schr. i [n. 57], 248–52), Sauppe, H., Progr. Göttingen (1864), 11–14Google Scholar (= Ausgewählte Schriften [Berlin, 1896], 433–6)Google Scholar; on cod. Vatican City, Bibl. Apost. Pal. lat. 1615 (Plautus: eleventh century South Germany), Schoell, F., T. Macci Plauti Truculentus (Leipzig, 1881), xxxv–viGoogle Scholar, W. M. Lindsay, op. cit. (n. 4), 82–3. M. D. Reeve informs me, however, that Sauppe was wrong to accept the presence of the χῖ in C.L.M. 816a; what Bergk saw belongs to the annotator's abbreviation of ὡραῖον. On various codices preserving Jerome's signs see Rahlfs, A., ‘Der Text des Septuaginta-Psalters’, Septuaginta-Studien, 2. Heft (Göttingen, 1907), 124–34Google Scholar.

⁸¹ The chapter also occurs in the Cava miscellany (n. 77), foll. 247^v–248^v. For its separate transmission see Traube, L., ‘Textgeschichte der Regula S. Benedicti’, Abh. d. Hist. Cl. d. Kön. Bayer. Ak. d. Wiss. 21, 3(1898), 725Google Scholar (= ed. 2, Abh. d. Kön. Bayer. Ak. d. Wiss., Philos.-philol. und hist. Kl. 25, 2 [1910], 121Google Scholar).

⁸² iv 581.21. The same title is attributed elsewhere in the Suda to a Diogenes or Diogenianus (ii 102.32).

⁸³ Cf. J. Aistermann, op. cit. (n. 4), 10 (arguing that the source of the Paris essay was composed before Probus edited Terence), Brugnoli, G., Atti Acc. Naz. Lincei 8, 6 (1955), 1–16Google Scholar (= Studi Suetoniani [Lecce, 1968], 156–8Google Scholar), N. Scivoletto, op. cit. (n. 4), 114 n. 23 (= Stud. 193 n. 33), G. Pascucci, op. cit. (n. 4), 27 (drawing attention to the conflict between the list of authors given by the Paris essay and the picture of Probus' interests painted in Suetonius' De grammaticis).

⁸⁴ Cf. A. Reifferscheid, op. cit. (n. 57), 419–20 (arguing that the Περ⋯ τ⋯ν ⋯ν τοῖς βιβλίοις σημείων was part of the De uiris illustribus and included material on shorthand and cryptography), 0. Ribbeck, op. cit. (n. 80), 150, J. Steup, op. cit. (n. 9), 52–3, Leo, F., Plaut. Forsch. ² (n. 3), 32Google Scholar, A. Macé, op. cit. (n. 12), 265–7, Traube, L., ‘Die Geschichte der tironischen Noten bei Suetonius und Isidorus’, Arch. f Stenogr. 53 (1901), 191–208Google Scholar (= Vorlesungen und Abhandlungen iii [Munich, 1920], 254–73Google Scholar), P. Weber, op. cit. (n. 77), 3–24, Gudeman, A., Grundriss (n. 43), 105 n. 1Google Scholar, ‘Krit. Zeich.’ (n. 52), 1917, Wessner, P., Bursians Jahresber. 188 (1921), 80Google Scholar, Funaioli, G., RE ii 4.1 (1931), 630–2Google Scholar, s.v. ‘Suetonius’, Hanslik, R., RE ii 8.1 (1955), 198Google Scholar, s.v. ‘Valerius Probus’, Grisart, A., Helikon 2 (1962), 390Google Scholar n. 49, R. W. Müllner, op. cit. (n. 41), 62, L. D. Reynolds and N. G. Wilson, op. cit. (n. 4), 20, C. O. Brink, op. cit. (n. 4), 36–7, Questa, C., RFIC 102 (1974), 186Google Scholar, J. E. G. Zetzel, op. cit. (n. 4), 15. Fontaine, J., Isidore de Séville et la culture classique dans l'Espagne wisigothique i (Paris, 1959), 74–80Google Scholar, does not name Suetonius. Hillgarth, J. N., ‘The Position of Isidorian Studies: a critical review of the literature since 1935’, Isidoriana (Leόn, 1961), 33 and n. 62Google Scholar, expresses himself obscurely.

⁸⁵ Cf. J. Fontaine, op. cit. (n. 84), 192–4, Holtz, L., RPh 45 (1971), 81–3Google Scholar.

⁸⁶ Cf. Donat. Gramm. Lat. iv 372.15–23, Cledonius, , Gramm. Lat. v 34.1–9Google Scholar, Diomedes, , Gramm. Lat. i 437.10–439.9Google Scholar.

⁸⁷ In 1931 T. W. Allen counted five papyrus texts of Homer out of a hundred with signs (Homeri Ilias i [n. 52], 198). In 1977 P. Oxy. 3224 with its signs (The Oxyrhynchus Papyri vol. xlv, 51–2) stood against 53 other ancient copies of the same set of Hesiod's works lacking signs. Lowe remarks, E. A. (Codices Latini Antiquiores. Suppl. [Oxford, 1971], 13 [on no. 1710])Google Scholar, on the small number of signs to be found in Latin manuscripts. On the tendency of Origen's signs to disappear in the course of the Greek biblical tradition see B. M. Metzger, loc. cit. (n. 79).

⁸⁸ At 1.5.4 positurae and notae appear as two of the thirty divisions of grammar.

⁸⁹ At Adu. Hieron. 2.40 Rufinus refers to the military theta (~ Isidore, Orig. 1.24) in a discussion of Origen's use of critical signs; this could indicate that the treatise used by Isidore was known to Rufinus.

⁹⁰ It is badly muddled. What Epiphanius (n. 66) and the essay published by Serruys (n. 77) call the ὑπολημνίσκος, Isidore calls an antigraphus cum puncto (Orig. 1.21.6).

⁹¹ The use of the same text of Virgil, Aen. 10.88–90, to illustrate the auersa obelismene (referring back to 10.25–9, 55–62) puts this beyond doubt.

⁹² Probus is treated throughout as a recent rather than an ‘ancient’ grammarian. See further below, nn. 133, 140.

⁹³ EXPLICIT is normal in cod. Paris, B.N. lat. 7530 and other South Italian manuscripts, but for FINIT see fol. 145^r. FINIT is said to be characteristic of Irish and Spanish manuscripts (Lindsay, W. M., Palaeografia Latina ii [Oxford, 1923], 5–10, iv [1925], 83–4Google Scholar; but see Lowe, E. A., CQ 22 [1928], 60Google Scholar [= Palaeographical Papers 1907–1965 i (Oxford, 1972), 272]Google Scholar, Oliver, R. P., TAPhA 82 [1951], 239 n. 8)Google Scholar.

⁹⁴ At NJbb 83 (1861), 320 n. 10Google Scholar (= Kl. phil. Schr. i 252 n. 10) he declared the notae simplices to be later than the other twenty-one but did not go into the matter in any detail.

⁹⁵ On the first list see below. Where the second list is concerned, the lack of correspondence in the symbols ∸/H※ and the repetition of ∸ later in the list strongly suggest corruption. graeca metafrasis et bis dictum et repugnans. Φ P graeca metafrasis et repugnans ought to have some correspondence with ⋗ bis dictum and with ∸ repugnans. It is hard to believe that M denoted both malum metrum and aprepes (see below, n. 156).

⁹⁶ Cf. Holtz, L., Stud. Med. 16 (1975), 114 n. 75Google Scholar.

⁹⁷ Weber, op. cit. (n. 77), 19–24, discerned three sections, one concerned with the signs used by Aristarchus and Probus, one concerned with the punctuation of lyric and scenic texts, and one concerned with the ‘uerborum sensus’ (p. 21) or ‘iudicium’ (p. 22). Only the first, in his view, went back to Suetonius.

⁹⁸ Nine signs are credited to Aristarchus in our sources. Most modern accounts reduce the number to six. See Villoison, op. cit. (n. 44), Proleg. xii–xxii, Sengebusch, M., ‘Homerica dissertatio prior’ in Dindorf, W., Homeri Ilias ⁴ (Leipzig, 1861), 25–7Google Scholar, Ludwich, op. cit. (n. 25), 1, 19–22, Susemihl, F., Geschichte der griechischen Litteratur in der Alexandrinerzeit i (Leipzig, 1891), 454 n. 105Google Scholar, Cohen, L., RE 2.1 (1895), 865–7Google Scholar, s.v. ‘Aristarchos’, Allen, , Homeri Ilias i (n. 52), 197–9Google Scholar, Davison, J. A., ‘The Transmission of the Text’, in Wace, A. J. B. and Stubbings, F. H., A Companion to Homer (London, 1962), 215–33 (224)Google Scholar, Pfeiffer, op. cit. (n. 25), 218, Fraser, P. M., Ptolemaic Alexandria (Oxford, 1972), i, 464, ii, 672 n. 164Google Scholar.

⁹⁹ This is the term used in the second list in cod. Venice, Bibl. Marc. gr. 483 and the first list in cod. Rome, Bibl. Naz. Vitt. Emm. gr. 6.

¹⁰⁰ The absurdity of the Latin explanation – ponitur quotiens multi uersus inprobantur, ne per singulos obelentur (for series of ⋯βελοί see cod. Venice, Bibl. Marc. gr. 454, foll. 19^r, 19^v, 26^v, 28^r, 29^v etc.) — and the elusiveness of that in the first Rome list — δηλοῖ δ⋯ κα⋯ αὐτ⋯ πολλ⋯ς ζητήσεις πρ⋯ς ταῖς προειρηέναις — suggest that somewhere in their ancestry lay a confession of ignorance. For Aristophanes' use of the sign (to indicate τ⋯ εὐτελές) see schol. Hom. Od. 18.282; for its use in an edition of Plato's Dialogues (πρ⋯ς τ⋯ν ⋯γωγ⋯ν τ⋯ς ɸιλοσοɸίας) Diog. Laert. 3.66. Lehrs, loc. cit. (n. 98), and others deny the κεραύνιον to Aristarchus.

¹⁰¹ Schol. A Hom. Il. 2.192, 203; 8.535–7. Cf. Ammonius ap. Nemesio ap. schol. A Hom. Il. 10.397–9 (verses repeating 10.310–12). For Aristophanes' use of the σίγμα and the ⋯ντίσιγμα see schol. Hom. Od. 5.247–8, schol. Aristoph. Ran. 153.

¹⁰² Cf. the function given to the ⋯ντίσιγμα περιεστιγμένον in Diogenes Laertius' account of an edition of Plato's Dialogues (πρ⋯ς τ⋯ς διττ⋯ς χρήσεις κα⋯ μεταθέσεις τ⋯ν γραɸ⋯ν).

¹⁰³ See Wolf, op. cit. (n. 24), cclvii n. 43 (= ed. 2, 158 n. 43), Lehrs, op. cit. (n. 25), 362 (= ed. 3,340–1), G. G. Pluygers, ‘De carminum Homericorum ueterumque in ea scholiorum…retractanda editione’, Progr. Leyden, 1847, 3, Sengebusch, op. cit. (n. 97), 25. Ludwich, op. cit. (n. 25),i, 20, 22, 209, 318,ii, 139, Susemihl, loc. cit. (n. 98), Gudeman, ‘Krit. Zeich.’ (n. 52), 1923–4, Jachmann, G., Nachr. d. Ak. d. Wiss. in Göttingen, Phil-hist. Kl., 223 n. 1Google Scholar (= Textgeschichtliche Studien, ed. Gnilka, C. [Königstein/Ts, 1982], 882 n. 1)Google Scholar, Nickau, K., Untersuchungen zur textkritischen Methode des Zenodotos von Ephesos (Berlin-New York, 1977), 260–3Google Scholar.

¹⁰⁴ One would talk more accurately of a common source to the source of the Paris list and Isidore on the one hand (see above), and the source of the Rome and Venice lists (see Lameere, op. cit. [n. 75], 42–3, 244–8) on the other.

¹⁰⁵ What Isidore reports makes it clear that ÷ obelus adpunctus in the actual list is corrupt.

¹⁰⁶ It is to be observed that Virgil's imitations of these two episodes come in the reverse order (books 7, 5).

¹⁰⁷ See Meleager, , A.P. 12.257Google Scholar, Martial 10.1.1, P. London, B. M. Inv. 136 (A.D. i). , G. M. Stephen, Scriptorium 13 (1959), 3–14CrossRefGoogle Scholar, Lameere, op. cit. (n. 75), 190–224.

¹⁰⁸ See [Plut.] Vit. Hom. 2.4, p. 25.22–5 Wilamowitz, Eustath. Comm. Horn. Il. 1 proem. p. 5.29–36 ed. Rom.

¹⁰⁹ The statement also occurs in the Venice material (with ηὐδ⋯ντο instead of the correct ἥνωντο). For discussion see Diels, H., SB Berlin Ak. Wiss. (1894), 357 n. 3Google Scholar, Lameere, op. cit. (n. 75), 42–53, Erbse, , Gymnasium 69 (1962), 76Google Scholar, S. West, op. cit. (n. 52), 18–24.

¹¹⁰ See above, nn. 101–3.

¹¹¹ The four signs come 13th-16th in the sequence of explanations given by the Paris codex (15th, 16th, 18th, 19th in the Isidorean sequence), and 10th, 14th, 15th, 16th in the preceding list.

¹¹² One might suppose that at some stage the tradition lost a ⋜ auersa superne obelata.

¹¹³ Isidore makes clear the function that the source attributed to ⋝⋜ (…ponitur finita loco suo monade, significatque similem sequentem quoque esse). See also Bassus, Caesius, Gramm. Lat. vi 266.18–267.2Google Scholar.

¹¹⁴ Pp. 73.12–76.16.

¹¹⁵ On Heliodorus see Thiemann, C., ΗΛΙΟΔΩΡΟϥ ΑΡΙΣΤΟΦΑΝΕΙΟΣ ΚΩΛΟΜΕΤΡΙΑ (Diss. Halle, 1868), 21–43Google Scholar, Heliodori colometriae aristophaneae quantum superest una cum reliquis scholiis in Aristophanem metricis (Halle, 1869 [pp. 95–136 on the signs])Google Scholar, Hense, O., Heliodorische Untersuchungen (Leipzig, 1870), 35–72Google Scholar, RE 8.1 (1912), 28–40 (32), Consbruch, M., ‘De ueterum περ⋯ ποιήματος doctrina’ Bresl. Phil. Abhandl. v, 3 (Breslau, 1890), 52–67Google Scholar, Conradt, C., ‘Ueber die Semeiotik des Heliodoros’, NJbb 151 (1895), 273–7Google Scholar, W. G. Rutherford, op. cit. (n. 50), 88–92, White, J. W., The Verse of Greek Comedy (London, 1912), 384–421Google Scholar, Holwerda, D., ‘De Heliodori conunentario metric in Aristophanen’, Mnemosyne 4, 17 (1964), 113–39CrossRefGoogle Scholar, 4, 20 (1967), 247–72 (247–8, 254–6, 263), Pfeiffer, op. cit. (n. 25), 189. A somewhat simpler set of colometrical signs is described in the old scholia to Pindar (see Irigoin, op. cit. [n. 60], 49–52).

¹¹⁶ Despite the differing names for the sign, there are striking similarities between the Latin note ⋝ diple superne obelata ponitur ad condicionem locorum uel temporum uel personarum mutatam and Greek accounts of the scenic κορωνίς, most particularly with one in the Byzantine treatise on tragedy published from cod. Oxford, Bodl. Libr. Barocci 131, fol. 415, by Browning, R., ΓΕΡΑΣ. Studies presented to George Thomson (Prague, 1963), 67–81Google Scholar (§ 10) ⋯ δ⋯ κορων⋯ς μέρους ⋯στ⋯ σημεῖον, ὅταν οἱ ὑποκριτα⋯ ⋯ξελθόντες τ⋯ς σκην⋯ς μόνον τ⋯ν χορ⋯ν καταλείπουσι, κα⋯ ⋯πεισέλθωσι π⋯λιν, ⋯πόταν κα⋯ τ⋯ν τόπον ἔστιν ⋯λλ⋯ξαι, κα⋯ τόπον κα⋯ χορ⋯ν, κα⋯ ὅλον τ⋯ν μ⋯θον, ⋯ρχῇ δ⋯ ⋯πεισοδίον ἣ τελευτῇ). It may be that a desire to keep this sign distinct from the narrative coronis (ῖ coronis tantum in fine libri posita inuenitur caused a compiler of the Latin list to seek a new name. The use of obelus rather than * paragraphus may also evidence a desire to harmonise.

¹¹⁷ This point is made by Gudeman, ‘Krit. Zeich.’ (n. 52), 1926.

¹¹⁸ See Aesch. Eum. 234, 488, P. Oxy, 20.2257 (on the Αἴτνη and the Ἀχιλλέως ⋯ρασταί of Aeschylus), Soph. Ai. 815, Aristoph. Ach. 394, Equ. 722 (the δύο διπλαῖ mentioned in schol. ΕΓ³ have often been thought to indicate the change of scene), Ran. 180, Thesm. 277 (P.S.I. 1194 [A.D. ii] has a κορωνίς in the margin). For avoidance of scene change by the Latin dramatists see Jocelyn, H. D., The Tragedies of Ennius (Cambridge, 1967), 165Google Scholar, Entretiens Fond. Hardt 17 (1972), 57Google Scholar.

¹¹⁹ Isidore also puts these signs last, and in their company the coronis.

¹²⁰ The term crux (see Asconius ap. Philarg. Virg. Buc. 3.105) may have been used by some Latin writers to denote the ἄλογος.

¹²¹ Schol. A Hom. Il. 16.613…⋯ν δ⋯ τῇ δευτέρᾳ ἄλογος αὐτῷ παρέκειτο (from Didymus). Lehrs, op. cit. (n. 25), 362 (= ed. 3, 341), changed ἄλογος here to ⋯βελός; cf. Ludwich, op. cit. (n. 25), 19 n. 20, 21 n. 23.

¹²² See P. Oxy. 15.1788, add. P. Oxy. 21, p. 142 (= ad Alcaeus F 3 b. 40^c, p. 167 L–P).

¹²³ Heliodorus ap. schol. Aristoph. Vesp. 1283e expresses bewilderment about τόποι ⋯πτ⋯ ἔχοντες στιγμ⋯ς κα⋯ ⋯λόγους.

¹²⁴ Virg. Aen. 10.444.

¹²⁵ Cassiodorus used a ‘chresimon’ to mark orthodox opinions in biblical commentaries which he read (see Inst. 1.9.3); an ‘achresimon’/‘achriston’ to mark the unorthodox (1.1.8, 1.9.3).

¹²⁶ The material relates to historici as well as to poets. For the use of signs in the margins of prose writers see above, n. 64.

¹²⁷ Cf. P. Oxy. 13 (1919). 1611.56, where marks the beginning of a quotation of Acusilaus in a grammatical treatise.

¹²⁸ indicates ‘dogmata ualde necessaria’; see Corpus Christianorum ser. lat. XCVII (Turnhout, 1958), 2Google Scholar. Cf. above, n. 125.

¹²⁹ Cf. the way in which the names of Crates and Aristarchus are added to the discussion of the περιεστιγμένη διπλ⋯ in the second Venice list (~ the second Rome list).

¹³⁰ Not sensibly, since two distinct Greek uses had been described.

¹³¹ Not sensibly, since the use of the sign to mark disagreement with Zenodotos had been described. Cf. Gudeman, ‘Krit Zeich.’ (n. 52), 1920.

¹³² A discussion which we have shown, above, to muddle the Greek facts.

¹³³ Probus is certainly made the last in a line of sign-using Latin grammarians, but neither the distance between him and the others nor that between him and the person responsible for the statements about Latin practice is made clear.

¹³⁴ Bergk rightly recognised Ennius and Lucilius. L. Mueller was wrong to change historicorum to scaenicorum (NJbb 87 [1863], 176)Google Scholar; see Leo, , Geschichte der römischen Literatur i (Berlin, 1913), 359 n. 1Google Scholar. Aristarchus is now known to have written an ὑπόμνημα on Herodotus I (Amherst, P. 2.12 [London, 1901]Google Scholar), although that does not necessarily mean that he was also responsible for an ἔκδοσις of the text. Schol. Thuc. 3.84.1 shows that there existed annotated texts of the other classical Greek historian.

¹³⁵ The presence of Aelius (Stilo) may be accepted, although Bergk preferred Laelius (Archelaus). All else is in doubt. For conjectures and discussion see Corte, F. Della, La filologia latina dalle origini a Varrone (Turin, 1937), 74 n. 3Google Scholar, Bonner, S. F., Hermes 88 (1960), 358–9Google Scholar, Timpanaro, S., Contributi di filologia e di storia della lingua latina (Rome, 1978), 84–5Google Scholar, Zetzel, op. cit. (n. 4), 15–17.

¹³⁶ As Aistermann did, op. cit. (n. 4), 10.

¹³⁷ As Bergk did, op. cit. (n. 57), 85 (= Kl. phil. Schr. i 585), Ribbeck, op. cit. (n. 80), 150, Leo, , Pl. Forsch. ² (n. 4), 32 n. 2Google Scholar.

¹³⁸ As Questa, C. does, RCCM 7 (1965), 925 n. 19Google Scholar, QUCC 1 (1966), 20Google Scholar, Due cantica delle Bacchides (Rome, 1967), 70 n. 19Google Scholar, RFIC 102 (1974), 179, 186, 188Google Scholar, in La critica testuale greco-latina oggi, metodi e problemi ed. Flores, E. (Rome, 1981), 160Google Scholar.

¹³⁹ Gramm. 24.2–3. See part I.

¹⁴⁰ Just when the auctores studied by Varro, Cicero and Caesar were replaced in the schools of the capital is not at all clear from our evidence. It was no ordinary school in which Q. Caecilius Epirota lectured on Virgil and alii poetae noui (Sueton. Gramm. 16.3). Virgil, Lucretius and Horace would have been in a sense antiqui by the time they were commonly read in preference to Ennius and Lucilius. The first clear sign of Virgil's classical status is in anecdotes like those about Caligula (Sueton. Cal. 34.2) and Remmius Palaemon (Sueton. Gramm. 23.4). Persius' detailed knowledge of Horace's poetry and the story told in his biography (51–3) that he read Lucilius only after leaving the school of Remmius Palaemon provide the earliest evidence for the rise of Horace and the fall of Lucilius. Tacitus separated Lucretius from Virgil and Horace, putting him with Lucilius, Sisenna, Varro and Calvus (Dial. [a work with a mid-seventies dramatic setting] 23.2). On the other hand, Seneca, no lover of the poets of the old Republican syllabus (see Dial. 5.37.5, Epist. 58.5, Gell. 12.2.1–8), cited Lucretius quite often. Quintilian cited him twice and mentioned him with respect as the next best hexameter poet after Virgil (Inst. 10.1.87; 12.11.27).

¹⁴¹ Suda iv 729.6.

¹⁴² Suda i 356.31–2.

¹⁴³ P. Oxy. 2.221, col. xv.16–17 (Ammonius on Hom. Il. 21.290 [= Schol. Il. v 107 Erbse]).

¹⁴⁴ Even if Gellius' stories about Probus are fictitious (see nn. 5–7) they attest a considerable reputation for him in the second century. Schol. Ver. Virg. Aen. 9.369 suggests that C. Sulpicius Apollinaris thought his opinions worth citing.

¹⁴⁵ On the dubious authenticity of some of the literary manuscripts referred to by Gellius and others see Goold, G. P., HSCPh 74 (1970), 160–2Google Scholar, Zetzel, , HSCPh 77 (1973), 235–43Google Scholar, Holford-Strevens, L. A., LCM 7, 5 (1982), 67Google Scholar.

¹⁴⁶ With Sueton. Gramm. 24.5 contrast Suda i 351, 24–5 (s.v. 3892 Ἀρίσταρχος).

¹⁴⁷ See part III.

¹⁴⁸ As Scivoletto and Büchner (above n. 4) seem to argue.

¹⁴⁹ See above, n. 98.

¹⁵⁰ How much Aristarchus was concerned with purely aesthetic judgements is unclear. The Paris list mentions verses marked with the obelus ipsius Homeri proprios sed non eo dignos (for the notion of Homer occasionally slipping see Lucil, 344–7, Hor. Ars 359–60, [Longin.], Subl. 9.14, 33.4, 36.2).

¹⁵¹ Cf. Aistermann, op. cit. (n. 4), 11–13 (arguing that the notae simplices had to do with rhetorical exercises), Scivoletto, op. cit. (n. 4), 114–15 (= Studi, 194–6).

¹⁵² Cf. Sueton. Gramm. 4.6–10.

¹⁵³ Cf. for the phraseology schol. A Hom. Il. 20.269–72 μάχεται δ⋯ σαɸ⋯ς τοῖς γνησίοις, schol. Ver. Virg. Aen. 9.369 adnotant…contrarium illi esse ‘sepsit se…habenas’ (7.600).

¹⁵⁴ For δισσολογία as a reason for deletion see schol. A. Hom. Il. 1.474. For the topic of περισσότης see Ammonius, Hom. Il. 21.290 (P. Oxy. 2.221, col. xv.25–6), schol. T Hom. Il. 1.189–93, schol. A Il. 1.295 et al., Serv. Dan. Virg. Aen. 1.21 hi duo si eximantur nihilominus sensus integer erit, 4.418 Probus sane sic adnotauit ‘si hunc uersum omitteret melius fecisset’, 8.627 et quibusdam uidetur hunc uersum omitti potuisse, 8.731 hunc uersum notant critici quasi superfluo et humiliter additum. The symbol Θ is elsewhere associated with different topics: Sidon. Carm. 9.334–5 (talking of his cousin Probus as a critic of bad verse) isti qui ualet exorationi destrictum bonus applicare theta; Auson. Epigr. 87.12. The Cava essay de notis antiquorum (see above, n. 77), lists a theta in amputandis.

¹⁵⁵ For the ⋯λλότριος στίχος see schol. A Hom. Il. 1.365. On literary κλοπή see Theopompus ap. Athen. 11.508c–d (FGrH 115 F 259), Porph. ap. Euseb. Praep. eu. 10.3, Clem. Alex. Strom. 6.2, Diog. Laert. 2.97, 3.57, Vitruv. 7 praef. 4–7, Donat. Vit. Verg. 45–6. The topic is not always easily separable from that of μετάɸρασις (see below, n. 159). The sign → if correctly explained (alienus et superuacuus), could have had to do only with interpolation.

¹⁵⁶ It seems likely that Probus used two signs to mark two quite different phenomena and that our list has suffered corruption. For metrical criticism in commentaries see Sisenna, ap. Rufin. Gramm. Lat. vi 561.8–10Google Scholar (on Plaut. Aul. iv 9 according to Hermann) haec scaena anapaestico metro est. sed concisa sunt, ut non intellegas, Serv. Virg. Buc. 2.50 non slat uersus (cf. Georg. 2.82, Aen. 3.535 [citing Donatus], 4.132, 5.299, 10.129, 11.243), Aen. 2.778 hic uersus caret scansione, 3.336 contra metrum (cf. 3.636, 4.22, 10.778), 5.481 pessimus uersus, 12.144 animaduertendum autem uersum hunc sine caesura esse. For criticism of τ⋯ ⋯πρεπές and the like see comm. anon. P. Oxy. 8.1086, col. ii.63–73 (on Aristarchus' athetesis of Il. 2.791–5), schol. A Hom. Il. 1.29–31, 3.423, 4.345–6, 14.1, 24.130–2, schol. Hom. Od. 6.244, Cornutus ap. schol. Ver. Virg. Aen. 5.488, Probus ap. Donat. Ter. Phorm. 1005, Serv. Virg. Aen. 2.592, Serv. Dan. Virg. Aen. 8.731.

¹⁵⁷ For the explanation of Π (contra historiam) cf. schol. Pind. Ol. 4.31 b παρ' ἱστορίαν (Pyth. 7.9b, schol. Soph. Trach. 633, Phil. 425, 445, schol. Eurip. Androm. 107, 224, 616, Rhes. 508). The phrase παρ' ἱστορίαν got into common parlance (see Cic. Att. 13.10.1 and Housman, A. E., CR 15 [1901], 155Google Scholar [= Class. Pap. ii 536–7]). For the duty of the grammaticus to comment on historia see Cic. De orat. 1.187, Quintil. Inst. 1.8.18. ‘History’ included the events and persons of the heroic period. For concern about Virgil's departures from the common tradition see Hyginus ap. Gell. 10.16, Macrob. Sat. 5.17.5, 5.22.9 (Probus notat), Serv. Virg. Aen. 1.267, 474, 8.493.

¹⁵⁸ The explanations of — (praepositum sine consequente) and ∸ (consequens sine praeposito) are not entirely clear. For the general topic see schol. A Hom. Il. 7.443, 20.269–72, 24.614–17, Hor. Ars 41–3, Serv. Virg. Aen. 2.668, 7.601, 8.40, 9.365, 10.157, 331 (Donatus), 12.120, 124, 357.

¹⁵⁹ For the topic of dependence on a previous author see, where Greek literature is concerned, Glaucus in hyp. Aesch. Pers. (on the beginning of the Aeschylean tragedy and that of one by Phrynichus), schol. Pind. Pyth. 4.507 (τ⋯ χῖ ὅτι ⋯κ τ⋯ν Ἡσιόδου Ἔργων εἴληπται), Theon, Progymn. in Rhet. Gr. ii 62.22–63.30). On passages of Virgil dependent on Greek literature see Macrob. Sat. 5 passim, Serv. Dan. Virg. Aen. 1.44 et al. For dependence on older Latin authors see Macrob. Sat. 6.1–2, Serv. Dan. Virg. Aen. 1.44 et al. For dependence on both Greek and Latin predecessors see Macrob. Sat. 6.3, Serv. Virg. Aen. 6.625. The topic was considered important both in διόρσωσις (cf. Gell. 1.21) and in κρίσις (cf. Gell. 17.10, Serv. Dan. Virg. Aen. 8.625).