^{page 271 note 1} Cf. V. Martin, Papyrus Bodmer II, Evangile de Jean, chap. 1–14. (grec). Bibliothèque Bodmer (Cologny-Genève, 1956). Also V. Martin and J. W. B. Barns, Papyrus Bodmer II, Supplément. Evangile de Jean, chap. 14–21. Nouvelle édition augmentée et corrigée, avec reproduction photographique du manuscrit (chap. 1–21). Bibliothèque Bodmer (Cologny-Genève, 1962).

^{page 271 note 2} Cf Martin, V. (1956), p. 31, ‘Nous avons déjà laissé entendre que la réparation de ces très nombreuses omissions était selon toute probabilité due au scribe original. En tout cas rien n'oblige à les attribuer à une autre main.’Google Scholar

^{page 271 note 3} Boismard, M.-E., ‘Le Papyrus Bodmer II’, Revue Biblique, LXIV, 3 (07 1957), 363–98, esp. P. 365.Google Scholar

^{page 272 note 1} Colwell, E. C., ‘Scribal Habits in Early Papyri: A Study in the Corruption of the Text’, The Bible in Modern Scholarship, ed. Hyatt, J. P. (Nashville-New York: Abingdon Press, 1965), pp. 370–89, esp. p. 382.Google Scholar

^{page 272 note 2} Deletion is usually by erasure, but diagonal lines and superior dots (with or without parentheses) are also found.

^{page 272 note 3} This scheme of analysis is indebted to Colwell, op. cit.

^{page 272 note 4} These corrections accord with the high incidence of αι for ɛ which is characteristic of the scribe of ⁶⁶.

^{page 272 note 5} Cf. the review of Martin, V.–Barns, J. W. B. (1962) by Boismard, M.-E. in Revue Biblique, LXX, 1 (01 1963), 120–33, esp. p. 131.Google Scholar

^{page 273 note 1} Usage elsewhere in John is ηγαπα at xi. 5, xiii. 23, xix. 26, xxi. 7, 20, and ηγαπησεν at xiii. I, xv. 9.

^{page 273 note 2} Cf Boismard, (1963), p. 121.Google Scholar

^{page 273 note 3} Deletion is by diagonal lines, superior dots and erasure, with α written above the erased ε. The different methods of deletion seem to offer no clue to any ‘later corrector(s)’!

^{page 273 note 4} With προσκυνω, John's usage is divided (accusative iv. 22bis, 24, and dative iv. 21, 23, ix. 38), while usage elsewhere favours the dative, cf. Bauer (Arndt–Gingrich), s.v., and references.

^{page 273 note 5} Boismard, (1963), p. 123 suggests an arrested ει ουν.Google Scholar

^{page 274 note 1} Twelve of these are from ι to ει (e.g. λεγι to λεγει iii. 4, υμις to υμεις v. 33, etc.), while there are two examples each of ει to η (iv. 10, vi. 40) and α to αι (vii. 4, 14). There is one example each of η to ει (iv. 10), αι to ε (xii. 29), ει to ε (xiii. 10), ι to υι (xii. 36), ιει to ει (xiii. 14), ου to ο (xi. 54), and ω to ο (v. 29).

^{page 274 note 2} α v. 36, xi. 51; η iii. 2; ο i. 15, iv. 6.

^{page 274 note 3} ɛν vii. 9, xi. 6; ου xiv. 10.

^{page 274 note 4} iii. 17 κοσμον … κοσμον iv. 39 vi. 54 τρωνων = xiii. 19

^{page 274 note 5} σ ii. 11, iv. 48, vi. 9; α ix. 41; λ viii. 17; ν ii. 22.

^{page 274 note 6} τον i. 27; ουτος iii. 2; και xiii. 32; μαρθας corrects a repeated μαριας xi. I; ουδɛνι viii. 33 conflates D with rel (?). Phrases are repeated at xii. 26 ɛμοι τις διακονη, and xiv. 3 παλιν ɛρχομαι.

^{page 274 note 7} παντα i. 9; βηθσαιδα v. 2; μητɛρα vi. 42; λɛγω xiii. 21.

^{page 274 note 8} This may have been caught by the scribe when he proceeded to the following προς αυτον!

^{page 275 note 1} I.e. the transcription of the text published in 1956, cf. n. 3, p. 302.

^{page 275 note 2} E.g. Boismard (see n. 5, p. 303); Aland, K., ‘Neue Neutestamentliche Papyri II’, New Testament Studies, X, 1 (10 1963), 62–79;Google ScholarMerk, A., Novum Testamentum Graece et Latine, editio nona (Rome, 1964), pp.862–7.Google Scholar

^{page 275 note 3} United Bible Societies' edition (UBS) reading.

^{page 276 note 1} UBS reading, indicated as editorial preference in Nestle's margin.

^{page 277 note 1} UBS reading.

^{page 277 note 2} UBS bracketed reading.

^{page 278 note 1} UBS reading.

^{page 279 note 1} The scribe usually drops the article (e.g. i. 46, ii. 25, iii. 19, iv. 12, vi. 10, vii. 22, viii. 25, x. 36 xi. 35, xii. 9 bis, 12, 16, xiii. 21, 23), but twice it is the corrector who removes it (x. 33, xii. 22).

^{page 279 note 2} I.e. agreeing in the addition of a word, though not in the place of its insertion (v. 6, vi. 52, vii. 12, viii. 48, ix. 39, x. 29, xii. 2, 9). Nestle supports the original against the corrector in this sense twice (vii. 52, xi. 32), and is independent of both three times (xi. 45, xii. 22, xiii. 24).

^{page 279 note 3} The UBS adopts the corrector's reading at ii. 12, vi. 2, vii. 46, viii. 28 (bracketed), xi. 54, xiii. 21. It is interesting to observe that of the twenty-six instances where the corrector of differs from the UBS, only three readings are without some form of Caesarean support, apart from the five instances of peculiar readings.

^{page 280 note 1} Cf. n. 1, p.310.

^{page 281 note 1} The corrections eventually made in the manuscript average three to a page. Errors clearly caught by the scribe while copying average one to every fourth page, but they are distributed with surprising evenness through the manuscript. The incidence is slightly higher in chapters iii–v, but only chapters ii, xi and xii are without examples. It may also be noticed that codices א D and W show a higher incidence of peculiar readings than does .

^{page 281 note 2} Accuracy is as difficult in reconstruction as in transcription, and for much the same reasons. Slips overlooked in antiquity include ii. 11 τη(ς) γαλɛιλαιας, iii. 6 iv. 23 προσκυνου(ν)τας, v. 30 δυνο ( = α)μαι, vi. 52, 53 , viii. 20 ρη(μα)τα, ix. 5 κοσμω(ω), ix. 16 σχι(σ)μα, ix. 31 ix. 40 υ( = η)μɛις…ɛσμɛν, xi. 35 xi.38 xiii. 19 ɛγω(ɛι)μι. Transcriptional slips more recently perpetrated include i. 17 αΛηθɛια for η αληθɛια, vi. 58 εκ for εκ του, ix. 4 δυναται for δυνη[σε]ται, and ix. 34 αμαρτιαις for αμαρτια.