^{page 3 note 1} This description is taken from Die Handschriften-Verzeichnisse der Königlichen Bibliothek zu Berlin, vol. xi. 1890–1897Google Scholar.

^{page 3 note 2} Catalogue des Manuscrits grecs de la Bibliothèque Royale de Bruxelles, par Henri Omont, 1885, p. 25.

^{page 4 note 1} A similar observation (about Vi Z) is made by La Roche, praef. ed. p. xvii.

^{page 4 note 2} Muccioli, Catalogus Malatest. Caesen. bibl. 1780; Martin, Albert, Mélange d'archéol. et d'hist. 1882 ii. 224CrossRefGoogle Scholarsq..; Schrader, , Hermes xxiv. (1894), 25Google Scholarsqq. collated one book (β).

^{page 4 note 3} My notes do not coincide with what Schiader read, ἀ(ν)νοΨώτατοϛ.

^{page 4 note 4} Journal of Theological Studies, 1904, p. 445.

^{page 4 note 5} Cf. E. Maunde Thompson, C.R. 1888, pp. 103, 4.

^{page 6 note 1} I have not been able to find a printed catalogue of the Cracow library; the description is taken from Ludwich, praef. p. xi.

^{page 6 note 2} Bandini, , Catalogus codicum. ……bibl. Mediceae Laurentianae, 1764, ii. 126sqq.Google Scholar

^{page 7 note 1} ‘Indice dei codici greci Laurenziani non compresi nel catalogo del Bandini,’ Studi italiani di filologia classica i. p. 144.

^{page 7 note 2} ‘Indice dei codici greci Riccardiani, Magliabecchiani e Maruccelliani,’ Studi ital, ii. pp. 525, 549. I have not seen these two MSS.

^{page 7 note 3} Cf. Cat. Cod. Graec. Bibl. Ambrosianae dig. Aemidius Martini et Dominious Bassi, 1906.

^{page 7 note 4} On which see Schrader, , Hermes xxii. 346sq.Google Scholar

^{page 8 note 1} On which see Schrader, , Hermes xxii. 346sq.Google Scholar

^{page 8 note 2} Puntoni, V., ‘Indice dei codici greci della biblioteca Estense di Modena,’ Studi ital. ivGoogle Scholar.

^{page 8 note 3} This MSS. is omitted in the Catalogus codicum ……. Bibliothecae Regiae Bavaricae, by I. Hardt, 1812.

^{page 8 note 4} Codices graeci manuscripti regiae bibliothecae Borbonicae descripti a Salvatore Cyrillo, 1826, ii. p. 142Google Scholar.

^{page 9 note 1} Coxe, Catalogi cod.…… bibl. Bodl., pars tertia, 1854, p. 78.

^{page 9 note 2} Omont, H., Inventaire sommaire des manuscrits grecs de la Bibl. Nat. ii. p. 253Google Scholar, iii. pp. 25 sqq.

^{page 9 note 3} On which see Schrader, l.c. p. 347.

^{page 10 note 1} On these cf. Ludwich, Progr. Regimont. 1889, I, p. IGoogle Scholar.

^{page 10 note 2} F4 r. we find five lines in different-coloured inks:

(black).

(brown).

(green).

(red).

(black).

Cf. Estense 245 (iii. G. 12) Χαρτίου μον.

^{page 10 note 3} This description, and also the readings of the MS., are derived, from P. C. Molhuysen De tribus Homeri Odysseae codicibus antiquissimis, 1896, pp. 8 sq. Cf. also Ludwich, , Progr. Regimont. 1888 ii.Google Scholar, Schrader, Porphyrii quaest. Od. pp. 163 sq.

^{page 11 note 1} Seen by Ludwich, , Program. Regim. 1888, 1, p. IGoogle Scholar.

^{page 12 note 1} Codd. MSS.ti Palatini Graeci … rec. et digessit Henricus Stevenson, 1885.

^{page 12 note 2} Codd. MSS.ti Graeci Ottoboniani … rec. E. Feron et F. Battaglini, 1893.

^{page 12 note 3} Codd. MSS.ti Graeci Reginae Suecorum et Pii PP. Il.… rec. et dig. Henricus Stevenson, 1889.

^{page 12 note 4} Codd. Urbinates Graeci… rec. Cosimus Stornaiolo, 1895.

^{page 13 note 1} Omont, H., Centralblatt für Bibliothekswesen, vii. p. 358Google Scholar.

^{page 13 note 2} Graeca D. Marci bibliotheca codicum mstorum, 1740, pp. 245 sqq.

^{page 14 note 1} See Ludwich, Progr. Regimont. 1871, Schrader, Porph. qu. Od. pp. 153 sq.

^{page 14 note 2} They are :—

antisigmas β 214–223 m.p.

obeli γ7 232–8, η251–8, λ38, 39, 435–442, μ376–388, ν174–184. 503–6, ο78–85, π247, 249–51.

asterisks and obeli θ 564–7.

asterisks (dotted) ν 430–3, ξ 160–4.

^{page 15 note 1} Catalogus … Bibl. Caesareae Vindobonensis … ed. Daniel de Nessel, 1690.

^{page 16 note 1} Cat. Codd. Graecorum qui in bibliotheca urbica Vratislaviensi adservantur … 1889.

^{page 19 note 1} In this class I mark variants which have a phonetic interest, of whatever value, with an asterisk.

^{page 19 note 2} Or a curious kind of inversion of (e.g. β 191, θ 109, 131, ν 161). A larger omission (Ψ 203–276) is rectified by the insertion of a small leaf between 193 and 194, with this note: οὕτω[202], [277], κατὰ τάξιν

^{page 21 note 1} To judge from Ludwich's and Molhuysen's reports.

^{page 25 note 1} Wrongly ascribed to the scholia of H3 in my text.

^{page 28 note 1} For similar changes apparently to produce metre, cf. ἡβόοντα-εϛ 1446, Ω604, Κόον ≡ 25, κομόονταϛ β8 7, γοόονταϛ 209, στροΦάασθαι for στρωΦᾶσθαι I 463, τροπάασθαι for τρωπᾶσθαι O 666 Π95, γ 119, κνέε for κνῆ Α 639, δηιόοιεν δ 226.

^{page 31 note 1} Br who resisted this, writes in his right margin: . and on the left

^{page 31 note 2} Cf. La Roche, praef. Od. p. xxix.

^{page 32 note 1} There are no obeli in U5, as Dindorf states.

^{page 32 note 2} Not obeli, as Dindorf.

^{page 34 note 1} This reading was omitted by error in my edition.

^{page 44 note 1} λ 38–43, 157–9, 602–4, τ 306, 346–8.

^{page 57 note 1} e.g. vi. 638 Kühn.

^{page 57 note 2} Or perhaps earlier, but the facsimiles ap. Molhuysen do not allow me to appreciate the age of the correctors.

^{page 57 note 3} Some further agreements among the members may be collected: ζ 61 made into H2; η 13 πυρὰν ἔκαιε H2 VI; ο 28 μνηστῆρεϛ H2 OP3; 192 ἀέκοντε H2 O ( + al.); 234 ἐριννὺϛ H2 P3 V3; 297 Φρέναϛ H2 sol.

^{page 60 note 1} I repeat the table given C.R. l.c. of the five principal families of Iliad MSS.

^{page 60 note 2} The rival view, that the Odyssey is moralising, is of course maintained by Aristotle in the Poetics, and elsewhere: Anth. Pal. ix. 522 Ὀδνσσείηϛ δὲ τὸ σῶΦρον γράμμα.

^{page 60 note 3} That is to say in the two Berichte quoted below (p. 65). The real totals I cannot estimate, but they are probably even more in favour of the Iliad.

^{page 65 note 1} See the Comptes rendus of Crönert and Blass, Archiv für Papyrusforschung 1903 and 1904.

^{page 69 note 1} Omissions: β 407 in ք I, γ 487 in ք 3, δ 399 in ք 5, κ 368–72 in ք 8, λ 604 in ք 11, and τ 581 in ք 25. Linguistic survival (?) ζ 298 in ք 6 deterioration γ 472 in ք 3. Additions: ν 51 a, 55 a, 58 a in ք 19, α 92 a in ք 22.

^{page 70 note 1} La Roche, ed. praef. pp. xii, xiii has a simple enumeration.

^{page 70 note 2} Four cases may be distinguished: homoearchon, homoeomeson, homoeoteleuton, and one which I must call ‘heads and tails,’ i.e. where the end of one line affects the beginning of the next, or vice versa. Homoeomeson is exceeding powerful.

(1) Homoearchon: α 381, 2 δ 432,? ε 29 λ 218, 407 ξ 434 ρ 338, 339 τ 458 ν 83 Φ 109, 334. 335 Χ146 ω 276, 533. (2) Homoeomeson: β 408 θ 182, 435, 436 ι 426, 563 λ 517, 518 ξ 476 ο 451 ρ 277, 314, 395, 547 τ 466 ν 46, 152 Φ 122, 123, 302, 318 Χ 317. 329 ψ 178, 179 (?) ω 217, 398. (3) Homoeoteleuton: α 139 8 75, 293, 753 ε 247 η 80, 288 ι 89, 361, 437 λ 513–515 ο 48 σ 119, 120, 115 Φ 109, 189 ψ 241, 242. (4) ‘Heads and Tails’: δ 432 ζ 21 ζ 402 λ 545 ν 258–261 σ 119, 120 τ 18 ω 354. These examples are taken from the Oxford edition. The MSS. exhibit an infinity. For a list from the Hymns, , see J.H.S. xv. 272Google Scholarsq.

^{page 73 note 1} Perhaps in obedience to the scholion.

^{page 74 note 1} For instance, to provide a pendant verb with an accusative.

^{page 75 note 1} C.R. 1899, p. 432.

^{page 76 note 1} This is often invoked by Galen xvi. 202 ; cf. also 909. xvi. 634 The meaning of ἔδαΦοϛ is ‘text’ κείμενον, not as in the Lexx. ‘original’: as ib. xvi. 837 μέλανα [margin] , or simply ‘manuscript’ ib. 468 cf. also 468, 634, xviii. 2. 863, 909. Another word for ‘text’ is : xvi. 80 . Schol. Pind. Ol. v. I Cf. further for μέτωπον Galen xvii. 2. II, μετώπια xvii. 2. 194, μετώπιον and νῶτον (tergum) xv. 624.

^{page 77 note 1} Wattenbach, Das Schrifiwesen im Mittelalter, pp. 265 sqq. has some patristic examples.

^{page 77 note 2} Repeated by Plutarch, Sulla 26.—In the anecdote above Diels (Rh. Mus. xxx. II n.) missing the point thinks a ‘later addition.’ As if such additions were made! Strabo, who is full of ‘readings,’ , etc., has at least two references to ἀντίγραΦα: the MSS. of Thucydides (374) which gave the form Μεθώνη for Μεθάνα, and the ancient MSS. of Homer (550). The earliest occasion in antiquity when a graphical error enters into consideration appears to be Polybius xii. 4 a, 4 (a numerical discrepancy in Ephorus).

^{page 77 note 3} Some of the following passages were utilised by Lehrs, Aristarchus³ p. 348, and before him by Villoison, Diatriba, pp. 135 sqq. Few have made their way into the manuals.

^{page 78 note 1} Under Гεδρωσία Stephanus has the usual expression:

^{page 78 note 2} ‘scratches,’ are probably accents. Vat. 1553 (theol. s. x–xi) has an index; (sc. rough, unlevigated; one of the eight remains, in this condition). Στίγματα could me.in letters in general: Aelian ap. Suid. in .

^{page 78 note 3} Scribes' eyes suffered from the white material: Galen iii. 776

^{page 78 note 4} Even the ‘interlinear gloss’ finds its expression in literature: schol. I. ycophr. 275 καὶ αἱ μὲν

^{page 79 note 1} The director of Laur. 32. 9 apologises for exactly this error in the case of his scribe: (250 v.) (sc. the marks to rectify the text)

^{page 79 note 2} , however, which occurs so frequently among his prescriptions in vol. xii., means ἐν etc., as is shown by the alternatives τινέϛ 893, 779, 838 836.= prescription, ordonnance, as well as = reading, is unknown to our lexica.

^{page 81 note 1} Equivalents for λείπει are numerous: in the Odyssey MSS. the condemnatory word is split and one syllable is placed before, the other after, the superfluous line: e.g. va … cat α 87 Br, α 93, β408 O, cf. vacat L8 ρ 233, 577 al., vacant 22l U 3, βακατ ρ 91 Rio, ἀρ… γόϛ θ 501 UI, πα….έλκει π 179, Ψ 48 O, ὑπέρ ….εστι B 637 U 13. ὑπέρ ….εστιΠερισσόϛ is common.

The reverse, a space left accidentally blank, is apologised for with λήθη (λήθη Roe 18 a. 1349f. 90r.); λάθοϛ Ven. ix. 16 (s. xv.) f. 246 V. excuses a repetition. A bolder tone is sometimes taken: Vat. 1347 (s. xv.)

^{page 81 note 2} So the s. x. Paris Demosthenes 2934: ἀντεβλήθη at the end of the Παραπρέσβεια Διώρθωμένον (-ωται) which does not necessarily imply collation is more common; see the facsimile of the Bodleian Plato praef. p. v; add Ven. 454 f. 17r. . Such expressions as Rarocci 121 f. 57 v. are probably attributable to a reviser.

^{page 82 note 1} Such as those of Hippocrates by Artemidorus and Dioscurides: Galen xv. 23

^{page 83 note 1} The uncertainty is not unlike that as to Chrysippus' teaching upon phlebotomy: Galen xi. 151 The explanation also seems the same, Chrysippus' works were lost: ib. 221 .

^{page 83 note 2} . Suidas in v.

^{page 85 note 1} I said the contrary l.c. p. 171; but in the light of the experience gained by combining collations into apparatus, I take the statement back.