^{page 73 note 2} This paper uses Kenyon's labels for the major Text-types. His terminology has two virtues: first, it claims nothing; second, it is memorable. Kenyon's Alpha equals Hort's Syrian, von Soden's Kappa, Lagrange's A; and it is the Text-type to which the Codex A, Alexandrinus, is most closely related. Kenyon's Beta equals Hort's Neutral plus Alexandrian, von Soden's Eta, Lagrange's B; and it is the Text-type to which Codex B, Vaticanus, is most closely related. Kenyon's Delta equals part of Hort's Western, part of von Soden's Iota, and Lagrange's D; and it is the Text-type to which Codex D, Bezae, is most closely related. Kenyon's Gamma equals Lake and Streeter's Caesarean, von Soden's Iota-alpha-eta-iota, and Lagrange's C for Caesarean; and it is the Text-type to which the Codex Theta is most closely related.Google Scholar

^{page 73 note 3} See Colwell, E. C., ‘Genealogical Method: Its Achievements and Its Limitations’, J.B.L. LXVI (1947), 109–33.Google Scholar

^{page 74 note 1} Sanders, Henry A., ‘The Egyptian Text of the Four Gospels and Acts’, Harvard Theol. Rev. xxvi (1933).Google Scholar

^{page 74 note 2} For a recent review of the literature on this topic, see A. F. J. Klijn, A Survey of the Researches into the Western Text of the Gospels and Acts (Utrecht, n.d.).Google Scholar

^{page 74 note 3} ‘Daraus ergiebt sich als methodisch richtig, zuerst den K-Typ zu fixieren and dann die Aufmerksamkeit den Stellen zuzuwenden, an welchen H sicher oder möglicherweise von K abweicht.’Google Scholarvon Soden, H., Die Schriflen des Neuen Testaments I. Teil: Untersuchungen. II. Abteilung: Die Textformen (Göttingen, 1911), p. 71.Google Scholar

^{page 75 note 1} This is continually assumed in the major part of his work even though he occasionally admits that it is not universally possible; e.g. in his lists of ‘divided’ Eta-readings he includes as a category ‘readings in which the reading of Eta cannot be definitely determined’. Op. cit. pp. 1004 ff.Google Scholar

^{page 75 note 2} Kenyon, Frederic G., The Text of the Greek Bible, 2nd ed. (London, 1949), p. 180.Google Scholar

^{page 75 note 3} Lake, , Blake, , and New, , ‘The Caesarean Text of the Gospel of Mark’, Harvard Theol. Rev. XXI (1928), 207–404.CrossRefGoogle Scholar

^{page 75 note 4} See, for example, Suggs, M. J., ‘The Eusebian Text of Matthew’, Wovum Testamentum, 1 (1956), 233–45.CrossRefGoogle ScholarSuggs presents in Tables I and II agreements of individual manuscripts and groups of manuscripts with Eusebius against the Textu Receptus. Agreements with the Textus Receptus are unsatisfactorily dealt with in the following paragraph:Google Scholar

One other line of evidence must be introduced. There is a marked tendency in Eusebius to reject the “eccentricities” of each of the “Great Texts”. This became apparent in an analysis of Eusebius’ agreements with TR which showed him in frequent disagreement with the special readings of Aleph, D, OL, and “Caesarean” authorities. Lest this should be misinterpreted as suggesting that Eusebius' text was really Byzantine, two further facts may be mentioned. First, most of Eusebius' agreements with TR also are found in the Nestle text. Second, in those infrequent cases where Eusebius' agreement with TR means disagreement with Nestle, his readings are almost without exception found in pre-Byzantine witnesses and have some claim to be primitive.'

Note that Suggs does not mention Vaticanus in this paragraph, nor does he give us data to support or overthrow our suspicion that Eusebius' agreements with the TR are as significant for his relation to other sources as his disagreements. He leaves us with such questions as ‘Wflat would the agreements with TR and Nestle do to the position of Vaticanus in Table I?’

^{page 75 note 5} ‘Eusebius’ New Testament Text in the Demonstratio Evangelica', J.B.L. LXXIII (1954), 167–8.Google Scholar

^{page 76 note 1} Streeter, B. H., The Four Gospels: A Study of Origins (London, 1924).Google Scholar Brief summaries of major criticisms of this work are given by Metzger, B. M., Studies and Documents, XVI, Annotated Bibliography of the Textual Criticism of the New Testament 1914–1939 (Copenhagen, 1955), PP. 82ff.Google Scholar

^{page 76 note 2} Gregory's 2322, in the library of the University of Texas.Google Scholar

^{page 76 note 3} ‘Haec vero omnia ita comparata sunt ut textus historiam ex ingenio potius quam cx documentis conficere videantur’— Klijn, A. F. J., A Survey of the Researches into the Western Text of the Gospels and Acts (Utrecht, n.d.), p. 10 and n. 35.Google Scholar

^{page 76 note 4} ‘Pro suspectis habenda sunt quae uni vel altera horum testium prorsus peculiaria sunt, turn uae classium quae videntur esse certam indolem ab homine docto profectam redolent.’—Tischendorf, C., Novum Testamentum Graece, 8th ed., vol. III, Prolegomena by Gregory, C. R. (Leipzig: Hinrichs, 1894), p. 53.Google Scholar Cf. for a similar negative appraisal, ibid. p. 196.

^{page 76 note 5} Colwell, E. C., ‘Genealogical Method: Its Achievements and Its Limitations’, J.B.L. LXVI (1947), 127–9.Google Scholar

^{page 77 note 1} Lagrange, M.-J., Critique Textuelle, II: La Critique Rationnelle (Paris, 1935), p. 41.Google Scholar

^{page 77 note 2} Vogels, H. J., Handbuch der Neutestanuntlichen Text-kritik (Münster, 1923), pp. 228f.Google Scholar

^{page 77 note 3} Vogels, op. cit. p. 231: ‘Reichen doch jedenfalls die lateinische und die syrische Uebersetzung mit ihren Wurzeln bis tief in das 2. Jahrh. hinab, also in eine Zeit, die wohi wenigstens 200 Jahre vor unseren ältesten griechischen Handschriften liegt.’Google ScholarCf. Kenyon, Frederic G., The Text of the Greek Bible: A Student's Handbook, 2nd ed. (London, 1949), p. 111.Google Scholar

^{page 78 note 1} Italics mine.Google Scholar

^{page 78 note 2} Souter, Alexander, The Text and Canon of the New Testament (New York, 1913), p. 14.Google ScholarKlijn, A. F. J., ‘Welke waarde hebben de vertalingen voor de textkritiek van het Nieuwe Testament’, Nederlandsch Theologisch Tijdschrift (1954), pp. 165–8; an English translation, ‘The Value of the Versions for the Textual Criticism of the New Testament’ by Harold H. Oliver, has been accepted for publication in The Bible Translator. See also Wikgren's discussion of the difficulties of using versional evidence and the need for careful discrimination in ‘The Citation of Versional Evidence in an Apparatus Criticus‘—Google ScholarNew Testament Manuscript Studies: The Materials and the Making of a Critical Apparatus, edited by Parvis, Merrill M. and Wikgren, Allen P. (Chicago, 1950), pp. 99–106.Google Scholar

^{page 78 note 3} See Aland, Kurt, ‘Das Johannesevangelium auf Papyrus’, Forschungen und Fortschritte, XXXI (02, 1957), 50;Google ScholarMaldfeld, Georg and Metzger, B. M., ‘Detailed List of the Greek Papyri of the New Testament’, J.B.L. LXIII (1949)., 359–70;Google ScholarMaldfeld, Georg, ‘Der Beitrag ägyptischer Papyruszuegen für den frühen griechischen Bibeltext’, in Mitteilungen aus der Papyrussammlung der Österreichischen Nationalbibliothek N.S., V. Folge, ed. by Gerstinger, H. (Vienna, 1956), pp. 79–84.Google Scholar

^{page 78 note 4} Sec the same warning forcibly presented by Vogels, op. cit. p. 232.Google Scholar

^{page 79 note 1} Riddle, Donald W., ‘Textual Criticism as a Historical Discipline’, Anglican Theol. Rev. XVIII (1936), 220–33;Google ScholarParvis, M., ‘The Nature and Tasks cf New Testament Textual Criticism: An Appraisal’, J. Religion, XXXII (1996), 165–74.Google Scholar

^{page 81 note 1} Cf. Lagrange's sound indictment of Sander's groupings of Manuscript W in Luke i. i–viii. 12 and John v. 12ff. on the basis of minority agreements— op. cit. p. 148.Google Scholar

^{page 81 note 2} In a “family”, the manuscripts are so closely related to each other that their common archetype can be reconstructed with a very slight margin of error.'— Lake, Silva, Family II and the Codex Alexandrinus, Studies and Documents, v (London, 1937), p. 5, n. 12.Google Scholar

^{page 81 note 3} Lake, Kirsopp, Codex i of the Gospels and its Allies, Texts and Studies, VII, no. 3 (Cambridge, 1902).Google Scholar

^{page 81 note 4} Kirsopp, and Lake, Silva, Family 13 (The Ferrar Group). The Text According to Mark with a Collation of Codex 28 of the Gospels: Studies and Documents, XI (London, 1941), ix.Google Scholar

^{page 81 note 5} New, Silva, ‘A Patmos Family of Gospel Manuscripts’, Harvard Theol. Rev. xxv (1932), 85–92. This Family contains Manuscripts 1169, 1173, 1204 and 1385.CrossRefGoogle Scholar

^{page 81 note 6} Colwell, E. C., The Four Gospels of Karahissar, I, History and Text (Chicago, 1936), pp. 170–222. These Families are: Family 574—containing 574, 330, 1815; Family 2327—containing 2327, 1186, 111, 2144, 1148, 1604; Family 179—in von Soden's Iota-Phi-b group; and von Soden's Iota.Beta type—containing in Mark 61, 1216, 1243, 16, 119, 120.Google Scholar

^{page 82 note 1} Of the ten K^r manuscripts available to him, every one read at least 80 % of a list of 180 variants from Stephanus in Mark; and seven of the ten read over 90%. Only four of the ten manuscripts had more than fifteen variants outside of those in the list. And Voss was able to identify ten readings in Mark peculiar to K^r. D. O. Voss, op. cit.; see also the same author's ‘Is von Soden's K^r a Distinct Type of Text?’ J.B.L. LVII (1938), 311–18.Google Scholar

^{page 82 note 2} In the eleventh chapter of Mark, seven of Voss's manuscripts read a total of nine variations from the Textus Receptus; each of these variants has the support of a majority of the K^r witnesses; and no one of these seven manuscripts reads any other variant in this chapter. Six other K^r manuscripts read only one nonK^r variant.Google Scholar

^{page 82 note 3} Op. cit. p. 29. It is true that in that stemma no one of the twenty-one is the father of another one; it is true also that there is enough variation between fidelity to the family text and position in the family tree to require a large number of corrupting intermediaries—for example, the manuscript closest to the archetype in the stemma (Manuscript 114) is ninth in fidelity to the family text, while the least faithful in content (Manuscript 116) has eight manuscripts below it in the tree. Mrs Lake calls attention to this fact with reference to Manuscripts 178 and 116—op. cit. p. 28.Google Scholar

^{page 82 note 4} Op. cit. p. 55.Google Scholar

^{page 83 note 1} This grouping, argued by Ayuso, T. in an article in Biblica, XVI (1935), 369–415, was accepted by the Lakes (in Studies and Documents, vol. v), (1937), p. 4, n. 5; and in Revue Biblique, XLVIII (1939), 497–505,Google Scholar and this revision of Caesarean grouping is reviewed by Metzger, B. in J.B.L. LXIV (1945), 457–89; LXVI (1947), 406–7.Google Scholar

^{page 83 note 2} Kirsopp and Silva Lake, op. cit. pp. ix–x.Google Scholar

^{page 84 note 1} Burkitt, F. C., ‘The Chester Beatty Papyri’, J.T.S. XXXIX (1933), 367.Google Scholar

^{page 84 note 2} Lake, Silva, Family II and the Codex Alexandrinus, p. 5, n. 12.Google Scholar

^{page 84 note 3} James E. McA. Baikie, ‘The Caesarean Text inter pares’. This Cambridge thesis is summarized by Metzger, in J.B.L. LXIV (1945), 475–6. Pasquali's ‘conclusions’ include as no. 10 a clear recognition of process, which is freely summarized as follows: ‘The Greek papyri, the quotations of the Latin tradition—show that in antiquity for popular authors, every copy represents, in some way, a particular edition that is a mixture, graduated variously from preexisting variations, genuine or spurious. The process of contamination, of equalization, had begun in antiquity among the different traditions which culminate in the form of the “Vulgate”. Such conditions explain how papyri that restore, in one point, a genuine reading, contain also particular corruptions.’Google Scholar

^{page 84 note 4} Würthwein, Ernst, The Text of the Old Testament (Oxford, 1957), p. 45.Google Scholar

^{page 85 note 1} Zuntz, op. cit. p. 156.Google Scholar

^{page 85 note 2} Op. cit. pp. 63–5.Google Scholar

^{page 85 note 3} Op. cit. p. 151, n. 1.Google Scholar

^{page 85 note 4} Westcott, B. F. and Hort, F. J. A., The New Testament in the Original Greek: Introduction (London, 1896), p. 145Google Scholar

^{page 85 note 5} Ibid. p. 122.

^{page 85 note 6} Nestle's twenty-second edition (1956), Novum Testamentum Graece, pp. 68–9.Google Scholar

^{page 86 note 1} ‘Apart from the three great recensions there are a large number of early and dissimilar variants that, having been first noticed in the Old Latin texts, are grouped together under the name of Western.’— Vaganay, Leo, Introduction to the Textual Crititism of the New Testament (London, 1937, p. 116). The italics in the above quotation are mine.Google Scholar

^{page 86 note 2} op. cit. p. 69. ‘By far the most probable hypothesis, therefore, is that both A and II were stages in the early development of the Ecclesiastical text, more similar to it than N, B, D or Θ and less similar than EFGH or Vυ).’ See also pp. ix and 68 of that study.Google Scholar

^{page 86 note 3} Harvard Theol. Rev. xxiv (1931), 142.Google Scholar

^{page 86 note 4} J.T.S. XLIII (1942), 184, n. 4.Google Scholar

^{page 87 note 1} The total number of Eta manuscripts listed by von Soden is eleven. He prints EtaGoogle Scholar

With the exception of 3 members 2 times;

With the exception of 4 members 7 times;

With the exception of 5 members 1 time;

With the exception of 6 members 3 times;

With the exception of 7 members 1 time.

When he cites Eta positively, he does it

With 5 members 1 time;

With 6 members 1 time;

With 7 members 2 time;

With 8 members 2 times.

^{page 87 note 2} Op. cit. I. Teil: II. Abteilung: A. Die Evangelien, p. 1000.Google Scholar

^{page 87 note 3} Op. cit. pp. 1001–2.Google Scholar

^{page 88 note 1} Klijn, op. cit. pp. 167–71.Google Scholar

^{page 88 note 2} Most recently Zuntz has equalled the force and irony of Housman in indictment of those who would stop with quantites—op. cit. pp. 59–60.Google Scholar

^{page 88 note 3} This has been cogently argued and effectively used by Zuntz in locating in the manuscript tradition—op. cit. pp. 64–6, 95–6.Google Scholar

^{page 89 note 1} Mark i. 32; i. 40 (Soden erroneously i. 46); and xii. 36— op. cit. p. 910.Google Scholar

^{page 89 note 2} Thus the text of W in Matthew is usually classified as Alpha Text-type; in John i. I–V. II, Sanders says it is unclassifiable–Kenyon says it is Alpha—Lagrange suggests it is Delta (equals Latin); in John v. 12 to end, it is Beta Text-type; in Luke i. i–viii. 12 it is Beta; in Luke viii. 13 to end it is Alpha; in Mark i. I–V. 30, it is Delta (equals Latin); in Mark v. 31 to end it belongs to the sub-Text-type of the Caesarean Text-type.Google Scholar

^{page 89 note 3} See my article, ‘The Complex Character of the Late Byzantine Text of the Gospels’, J.B.L. LIV (1935), 211–21; and The Four Gospels of Karahissar: I, History and Text (Chicago, 1936).Google Scholar

^{page 89 note 4} Matthew: (I) One type throughout gospel, a mixture of a Chrysostom text, plus Family 2327, plus 330–1815 of von Soden's Iota-Beta-b. (This type occurs in MS. 2400 kfl Matt. ii. 1–Viii. 25.)Google Scholar

Mark: (2) i. i–vi. 15 Family 574 (574, 330, 1815), a mixture of Family 2327, and von Soden's Iota- Beta.

(3) vi. i6–ix. a MS. is a mixture of von Soden's Iota-Phi-b (as in MS. 179), by correction to von Soden's Kappa¹ (as in MS. V).

(4) ix. 4b–xvi. 20 Family 574 is an indirect ancestor of Family 2327, an identifiable group within von Soden's Kappa^x.

Luke: (5) i. i–viii. 15 MS. 574 predominantly K with some early non-neutral survival.

(6) viii. 16–xvi. 23 MS. is again part of Family 2327.

(7) xvi. 24ff. MS. 574 equals type (5) above.

John: (8) i. I–xviii. n6 MS. 574 equals type (5) above but with more correction to a Stephanus type text.

(9) xviii. 27–xxi. 25 MS. 574 is a mixture of a ‘Caesarean’ type as in MS. 544, by correction to a Kappa^a type as in MSS. 2400 and 482.

^{page 90 note 1} MS. 700 Matt. i. I-Luke x is Gamma Text-type; Luke xi-John xxi is Alpha. MS. Psi Mark is Beta; Luke-John is Alpha. MS. L Matt. i–xvii is Alpha, the rest Beta. MS. Delta Mark iii–xii is Beta, the rest is Alpha. MS. 579 Matt. is Alpha, the rest Beta. MS. 1241 Matt.-Mark is Alpha, the rest is Beta. MS. 2400 Matt. ii. I–viii. 25 is Family 2327 (or Family 2327+ 330); in Rom. x. 19ff. joins MS.330. MS.61 Matt. ix. 15-Mark xvi is Iota.Beta. MS. 485 Mark. i.i–viii. 14 is Family 2327; Mark viii. 14-Luke xxiv joins MSS. 251 ff. MS. 59 Luke xi. 32ff. is Family 2327. For MS. 1204,Google Scholar see Lake's, Silva, Article Harvard Theol. Rev. xxv (1932), 90–1.Google Scholar

^{page 90 note 2} They involve the Latin MSS. 1, c and e, g, a b ff², ff¹, g¹. He seriously urges beginning the study of a mixed text at the end of a book or the codex-op. cit. pp. 230–1, 100f.Google Scholar

^{page 91 note 1} Aland, K., op. cit. pp. 52–3. He applies this label to all early texts: ‘Der Mischtext charakterisiert die Frühzeit.’Google Scholar