xiv. 17, οὐ γάρ ἐστιν ἡ βασιλεία τοῦ Θεοῦ βρῶσις καὶ πόσις, ἀλλὰ

δικαιοσύνη _καὶ ἄσκησις_ καὶ εἰρήνη. Tregelles (An Account of Printed Text, p. 222) adds 1 Cor. vii. 5; Act. x. 30; Rom. xii. 13 (!) More to their purpose, perhaps, if we desired to help them on, would be the suspected addition of καὶ νηστείᾳ in Mark ix. 29, and of the whole verse in the parallel place Matt. xvii. 21; the former being brought into doubt on the very insufficient authority of Codd. א (by the first hand) Β, of the beautiful Latin copy _k_ from Bobbio, and by reason of the silence of Clement of Alexandria: the latter on the evidence of the same Greek manuscripts (_k_ being defective) with Cod. 33, both (?) Egyptian, the Curetonian and Jerusalem Syriac, the Latin _e_ _ff_1, some forms of the Ethiopic version, and from the absence of the Eusebian canon, which ought to have referred us to the parallel place in St. Mark, whereas that verse is assigned to the _tenth_ canon. In the face of such readings of אΒ it is hard to understand the grounds of Mr. Darby’s vague suspicion that they “bear the marks of having been in ecclesiastical hands.” (N. T., Preface, p. 3.)

255 See (6), (7), (17), (18). The uncial characters most liable to be confounded by scribes (p. 10) are ΑΔΛ, ΕΣ, ΟΘ, ΝΠ, and less probably ΓΙΤ. An article in a foreign classical periodical, written by Professor Cobet, the co-editor of the Leyden reprint of the N. T. portion of Cod. B, unless regarded as a mere _jeu d’esprit_, would serve to prove that the race of conjectural emendators is not so completely extinct as (before Mr. Linwood’s pamphlet) I had supposed. By a dexterous interchange of letters of nearly the same form (Δ for Α, Ε for Σ, Ι for Τ, Σ for Ε, κ for ΙΣ, Τ for Ι) this modern Bentley—and he well deserves the name—suggests for ΑΣΤΕΙΟΣ τῷ θεῷ Act. vii. 20 [compare Heb. xi. 23] the common-place ΔΕΚΤΟΣ τῷ θεῷ, from Act. x. 35. Each one of the _six_ necessary changes Cobet profusely illustrates by examples, and even the reverse substitution of δεκτός for ἀστεῖος from Alciphron: but in the absence of all manuscript authority for the very smallest of these several permutations in Act. vii. 20, he excites in us no other feeling than a sort of grudging admiration of his misplaced ingenuity. In the same spirit he suggests ΗΔΕΙΟΝΑ for ΠΛΕΙΟΝΑ, Heb. xi. 4; while in 1 Cor. ii. 4 for ἐν πειθοῖς σοφίας λόγοις he simply reads ἐν πειθοῖ σοφίας, the σ which begins σοφίας having become accidentally doubled and λόγοις subsequently added to explain πειθοῖς, which he holds to be no Greek word at all: it seems indeed to be met with nowhere else. Dr. Hort’s comment on this learned trifling is instructive: “Though it cannot be said that recent attempts in Holland to revive conjectural criticism for the N. T. have shown much felicity of suggestion, they cannot be justly condemned on the ground of principle” (Introd., p. 277).

256 Thus Canon I of this chapter includes (12), (19): Canon III includes (2), (3), (4), (8), (9), (10); while (13) comes under Canon IV; (20) under Canon VI.

257 “Canon Criticus” xxiv, N. T., by G. D. T. M. D., p. 12, 1735.

258 Dean Burgon cites (Revision Revised, pp. 359, 360) “no less than thirty ancient witnesses.”

259 ’The precept, if we omit the phrase, is in striking harmony with the at first sight sharp, extreme, almost paradoxical character of various other precepts of the “Sermon on the Mount.” Milligan, Words of the N. T., p. 111.

260 Very similar in point of moral feeling is the variation between ὀλιγοπιστίαν, the gentler, intrinsically perhaps the more probable, and ἀπιστίαν, the more emphatic term, in Matt. xvii. 20. Both must have been current in the second century, the former having the support of Codd. אB, 13, 22, 33, 124, 346 [_hiat_ 69], the Curetonian Syriac (and that too against Cod. D), both Egyptian, the Armenian and Ethiopic versions, Origen, Chrysostom (very expressly, although his manuscripts vary), John Damascene, but of the Latins Hilary alone. All the rest, including Codd. CD, the Peshitto Syriac, and the Latins among first class witnesses, maintain ἀπιστίαν of the common text.

261 Perhaps I may refer to my “Textual Guide,” p. 120. The utmost caution should be employed in the use of this kind of evidence: perhaps nowhere else do authorities differ so much.—ED.

262 E.g. Irenaeus, Contra Haereses, v. 30. 1, for which see below, p. 261: the early date renders this testimony most weighty.

263 In deference to Lardner and others, who have supposed that Ignatius refers to the sacred autographs, we subjoin the sentence in dispute. Ἐπεὶ ἤκουσά τινων λεγόντων, ὅτι ἐὰν μὴ ἐν τοῖς ἀρχαίοις εὕρω, ἐν τῷ εὐαγγελίῳ οὐ πιστεύω; καὶ λέγοντός μου αὐτοῖς, ὅτι γέγραπται, ἀπεκρίθησάν μοι, ὅτι πρόκειται. Ἐμοὶ δὲ ἀρχεῖά ἐστιν Ἰησοῦς Χριστός κ.τ.λ. (Ad Philadelph. c. 8.) On account of ἀρχεῖα in the succeeding clause, ἀρχείοις has been suggested as a substitute for the manuscript reading ἀρχαίοις, and so the interpolators of the genuine Epistle have actually written. But without denying that a play on the words was designed between ἀρχαίοις and ἀρχεῖα, both copies of the Old Latin version maintain the distinction made in the Medicean Greek (“si non in veteribus invenio” and “Mihi autem principium est Jesus Christus”), and any difficulty as to the sense lies not in ἀρχαίοις but in πρόκειται. Chevallier’s translation of the passage is perfectly intelligible, “Because I have heard some say, Unless I find it in the ancient writings, I will not believe in the Gospel. And when I said to them, ‘It is written [in the Gospel],’ they answered me, ‘It is found written before [in the Law].’ ” Gainsayers set the first covenant in opposition to the second and better one.

264 Thus Dr. Westcott understands the term, citing from Tertullian, De Monogamia, xi: “sciamus planè non sic esse in Graeco authentico.” Dean Burgon refers us to Routh’s “Opuscula,” vol. i. pp. 151 and 206.

265 Compare too Jerome’s expression “ipsa authentica” (Comment. in Epist. ad Titum), when speaking of the autographs of Origen’s Hexapla: below, p. 263.

266 The view I take is Coleridge’s (Table Talk, p. 89, 2nd ed.). “I beg Tertullian’s pardon; but among his many _bravuras_, he says something about St. Paul’s autograph. Origen expressly declares the reverse;” referring, I suppose, to the passage cited below, p. 263. Bp. Kaye, the very excellence of whose character almost unfitted him for entering into the spirit of Tertullian, observes: “Since the whole passage is evidently nothing more than a declamatory mode of stating the weight which he attached to the authority of the Apostolic Churches; to infer from it that the very chairs in which the Apostles sat, or that the very Epistles which they wrote, then actually existed at Corinth, Ephesus, Rome, &c., would be only to betray a total ignorance of Tertullian’s style” (Kaye’s “Ecclesiastical History ... illustrated from the writings of Tertullian,” p. 313, 2nd ed.). Just so: the autographs were no more in those cities than the chairs were: but it suited the purpose of the moment to suppose that they were extant; and, _knowing nothing to the contrary_, he boldly sends the reader in search of them.

267 I do not observe, as some have thought, that Eusebius (Hist. Eccl. v. 10) intimates that the copy of St. Matthew’s Gospel in Hebrew letters, left by St. Bartholomew in India, was the Evangelist’s autograph; and the fancy that St. Mark wrote with his own hand the Latin fragments now at Venice (_for._) is worthy of serious notice. The statement twice made in the “Chronicon Paschale,” of Alexandria, compiled in the seventh century, _but full of ancient fragments_, that ὡσεὶ τριτὴ was the true reading of John xix. 14 “καθὼς τὰ ἀκριβῆ βιβλία περιέχει, αὐτό τε τὸ ἰδιόχειρον τοῦ εὐαγγελιστοῦ ὅπερ μέχρι τοῦ νῦν πεφύλακται χάριτι Θεοῦ ἐν τῇ ἐφεσίων ἁγιωτάτῃ ἐκκλησίᾳ καὶ ὑπὸ τῶν πιστῶν ἐκεῖσε προσκυνεῖται” (Dindorf, Chron. Pasch., pp. 11 and 411), is simply incredible. Isaac Casaubon, however, a most unimpeachable witness, says that this passage, and another which he cites, were found by himself in a fine fragment of the Paschal treatise of “Peter Bp. of Alexandria and martyr” [d. 311], which he got from Andrew Damarius, a Greek merchant or calligrapher (Pattison, Life of Is. Casaubon, p. 38). Casaubon adds to the assertion of Peter “Hec ille. Ego non ignoro quid adversus hanc sententiam possit disputari: de quo judicium esto eruditorum” (Exercit. in Annal. Eccles. pp. 464, 670, London, 1614).

268 “I have no doubt,” says Tischendorf, “that in the very earliest ages after our Holy Scriptures were written, and before the authority of the Church protected them, wilful alterations, and especially additions, were made in them,” English N. T., 1869, Introd. p. xv.

269 Caius (175-200) in Routh’s “Reliquiae,” ii. 125, quoted in Burgon’s “Revision Revised,” p. 323.

270 “Necdum quoque Marcion Ponticus de Ponto emersisset, cujus magister Cerdon sub Hygino tunc episcopo, qui in Urbe nonus fuit, Romam venit: quem Marcion secutus...” Cyprian., Epist. 74. Cf. Euseb., Eccl. Hist., iv. 10, 11.

271 Dean Burgon attributes more importance to Marcion’s mutilations. _See_ e.g. “The Revision Revised,” pp. 34-35.

272 In 1 Cor. x. 9 Marcion seems to uphold the true reading against the judgement of Epiphanius: ὁ δὲ μαρκίων ἀντὶ τοῦ _κν_ _χν_ ἐποίησεν. Consult also Bp. Lightfoot’s note (Epistle to the Colossians, p. 336, n. 1) on Heracleon’s variation of πέντε for ἓξ in John ii. 20. “There is no reason to think,” he says, “that Heracleon falsified the text here; he appears to have found this various reading already in his copy.”

_ 273 See_ Chap. XI on Acts xxvii. 37.

274 Irenaeus’ anxiety that his own works should be kept free from corruption, and the value attached by him to the labours of the corrector, are plainly seen in a remarkable subscription preserved by Eusebius (Eccl. Hist. v. 20), which illustrates what has been said above, Ὁρκίζω σε τὸν μεταγραψόμενον τὸ βίβλιον τοῦτο, κατὰ τοῦ κυρίου ἡμῶν ἰησοῦ χριστοῦ, καὶ κατὰ τῆς ἐνδόξου παρουσίας αὐτοῦ, ἧς ἔρχεται κρῖναι ζῶντας καὶ νεκρούς, ἵνα ἀντιβάλλῃς ὃ μετεγράψω, καὶ κατορθώσῃς αὐτὸ πρὸς τὸ ἀντίγραφον τοῦτο, ὅθεν μετεγράψω ἐπιμελῶς, καὶ τὸν ὅρκον τοῦτον ὁμοίως μεταγράψῃς, καὶ θήσεις ἐν τῷ ἀντιγράφῳ. Here the copyist (ὁ μεταγραφόμενος) is assumed to be the same person as the reviser or corrector. Mr. Linwood also (_ubi supra_, p. 11) illustrates from Martial (Lib. vii. Epigram. x) the reader’s natural wish to possess an author’s original manuscript rather than a less perfect copy: _Qui vis_ archetypas _habere nugas_. A still stronger illustration of the passage in Irenaeus (v. 30) is Linwood’s citation of a well-known passage in Aulus Gellius, a contemporary of that Father, wherein he discusses with Higinus the corrupt variation _amaro_ for _amaror_ in Virgil, Geor. ii. 247 (Noctes Atticae, Lib. i. cap. 21).

275 Μακάριοι, φησίν, οἱ δεδιωγμένοι ἕνεκεν δικαιοσύνης, ὅτι αὐτοὶ υἱοὶ Θεοῦ κληθήσονται; ἤ, ὥς τινες τῶν μετατιθέντων τὰ Εὐαγγέλια, Μακάριοι, φησίν, οἱ δεδιωγμένοι ὑπὸ τῆς δικαιοσύνης, ὅτι αὐτοὶ ἔσονται τέλειοι; καί, μακάριοι οἱ δεδιωγμένοι ἕνεκα ἐμοῦ, ὅτι ἔξουσι τόπον ὅπου οὐ διωχθήσονται (Stromata, iv. 6). Tregelles (Horne, p. 39, note 2) pertinently remarks that Clement, in the very act of censuring others, subjoins the close of Matt. v. 9 to v. 10, and elsewhere himself ventures on liberties no less extravagant, as when he thus quotes Matt. xix. 24 (or Luke xviii. 25): πειστέον οὖν πολλῷ μᾶλλον τῇ γραφῇ λεγούσῃ, Θᾶττον κάμηλον διὰ τρυπήματος βελόνης διελεύσεσθαι, ἢ πλούσιον φιλοσοφεῖν (Stromata, ii. 5).

276 In this place (contrary to what might have been inferred from the language of Irenaeus, cited above, p. 262, note 2) the copyist (γραφεύς) is clearly distinct from the corrector (διορθωτής), who either alters the words that stand in the text, or adds to and subtracts from them. In Cobet’s masterly Preface to his own and Kuenen’s “N. T. ad fidem Cod. Vaticani,” Leyden, 1860, pp. xxvii-xxxiv, will be found most of the passages we have used that bear on the subject, with the following from classical writers, “Nota est Strabonis querela xiii. p. 609 de bibliopolis, qui libros edebant γραφεῦσι φαύλοις χρώμενοι, καὶ οὐκ ἀντιβάλλοντες... Sic in Demosthenis Codice Monacensi ad finem Orationis xi annotatum est Διωρθώθη πρὸς δύο Ἀττικιανά, id est, _correctus est_ (hic liber) _ex duobus codicibus ab Attico_ (nobili calligrapho) _descriptis_.” Just as at the end of each of Terence’s plays the manuscripts read “Calliopius recensui.”

277 No doubt certain that are quite or almost peculiar to Cod. D would deserve consideration if they were not destitute of adequate support. Some may be inclined to think the words cited above in vol. I. p. 8 not unworthy of Him to whom they are ascribed. The margin of the Harkleian Syriac alone countenances D in that touching appendage to Acts viii. 24, which every one must wish to be genuine, ος πολλα κλαιων ου διελυ[ι]μπανεν. Several minute facts are also inserted by D in the latter part of the same book, which are more likely to rest on traditional knowledge than to be mere exercises of an idle fancy. Such are απο ωρας ε εως δεκατης annexed to the end of Acts xix. 9: και Μυρα to Acts xxi. 1; the former of which is also found in Cod. 137 and the Harkleian margin; the latter in the Sahidic and one or two Latin copies.

278 Considering that Cod. D and the Latin manuscripts contain the variation in Luke iii. 22, but not in Matt. iii. 17, we ought not to doubt that Justin Martyr (p. 331 B, ed. Paris, 1636) and Clement (p. 113, ed. Potter) refer to the former. Hence Bp. Kaye (Account of the Writings of Clement, p. 410) should not have produced this passage among others to show (what in itself is quite true) that “Clement frequently quotes from memory.”

279 This point is exceedingly well stated by Canon Cook (Revised Version of the first three Gospels, p. 176): “I will not dwell upon indications of Arian tendencies. They are not such as we should be entitled to rely upon.... Eusebius was certainly above the suspicion of consciously introducing false statements or of obliterating true statements. As was the case with many supporters of the high Arian party, which came nearest to the sound orthodox faith, Eusebius was familiar with all scriptural texts which distinctly ascribe to our Lord the divine attributes and the divine name, and was far more likely to adopt an explanation which coincided with his own system, than to incur the risk of exposure and disgrace by obliterating or modifying them in manuscripts which would be always open to public inspection.”

280 “This is possible, though there is no proof of it,” is Professor Abbot’s comment (_ubi supra_, p. 190, but _see_ above, vol. i. p. 118, note 2).

281 In the “Notitia Editionis Cod. Sin.,” 1860. They are Matt. xxvii. 64-xxviii. 20; Mark i. 1-35; Luke xxiv. 24-53; John xxi. 1-25. Other like calculations, with much the same result, are given in Scrivener’s “Cod. Sin.,” Introd. pp. xlii, xliii.

282 And that too hardly to the credit of either of them. “Ought it not,” asks Dean Burgon, “sensibly to detract from our opinion of the value of their evidence to discover that _it is easier to find two consecutive verses in which the two MSS. differ, the one from the other, than two consecutive verses in which they entirely agree_?... On every such occasion only one of them can possibly be speaking the truth. Shall I be thought unreasonable if I confess that these perpetual inconsistencies between Codd. B and 8—grave inconsistencies, and occasionally even gross ones—altogether destroy my confidence in either?” (Last Twelve Verses of St. Mark, pp. 77-8.)

283 Magnus siquidem hic in nostris codicibus error inolevit, dum quod in eadem re alius Evangelista plus dixit, in alio, quia minus putaverint, addiderunt. Vel dum eundem sensum alius aliter expressit, ille qui unum e quatuor primum legerat, ad ejus exemplum ceteros quoque existimaverit emendandos. _Unde accidit ut apud nos mixta sint omnia_ (Praef. ad Damasum).

284 The precise references may be seen in Tischendorf’s, and for the most part more exactly in Tregelles’ N. T. That on Matt. xxiv. 36 is Tom. vii. p. 199, or vi. p. 54; on Galat. iii. 1 is Tom. vii. pp. 418, 487.

285 See our note on Luke xxii. 44 below in Chap. XI. This same writer testifies to a practice already partially employed, of using breathings, accents, and stops in copies of Holy Scripture. Ἐπειδὴ δέ τινες κατὰ προσῳδίαν ἔστιζαν τὰς γραφὰς καὶ περὶ τῶν προσῳδῶν τάδε: ὀξεῖα ᾽, δασεῖα ᾽, βαρεῖα ᾽, ψιλὴ ᾽, περισπωμένη ᾽, ἀπόστροφος ᾽, μακρὰ —, ὑφὲν ᾽, βραχεῖα ᾽, ὑποδιαστολή, Ὡσαύτως καὶ περὶ τῶν λοιπῶν σημείων κ.τ.λ. (Epiphan., De Mensur., c. 2, Tom. iii. p. 237 Migne). This passage may tend to confirm the statements made above, Vol. I. pp. 45-8, respecting the presence of such marks in very ancient codices, though on the whole we may not quite vouch for Sir F. Madden’s opinion as regards Cod. A.

286 “Evangelia quae falsavit Lucianus, apocrypha.” “Evangelia quae falsavit Esitius [_alii_ Hesychius _vel_ Isicius], apocrypha,” occur separately in the course of a long list of spurious books (such as the Gospels of Thaddaeus, Matthias, Peter, James, that “nomine Thomae quo utuntur Manichaei,” &c.) in Appendix iii to Gelasius’ works in Migne’s Patrologia, Tom. lix. p. 162 [A.D. 494]. But the authenticity of those decrees is far from certain, and since we hear of these falsified Gospels nowhere else, Gelasius’ knowledge of them might have been derived from what he had read in Jerome’s “Praef. ad Damasum.”

287 Griesbach rejoices to have Hug’s assent “in eo, in quo disputationis de veteribus N. T. recensionibus cardo vertitur; nempe extitisse, inde a secundo et tertio saeculo, plures sacri textûs recensiones, quarum una, si Evangelia spectes, supersit in Codice D, altera in Codd. BCL, alia in Codd. EFGHS et quae sunt reliqua” (Meletemata, p. lxviii, prefixed to “Commentarius Criticus,” Pars ii, 1811). I suppose that Tregelles must have overlooked this decisive passage (probably the last its author wrote for the public eye) when he states that Griesbach now “virtually gave up his system” as regards the possibility of “drawing an actual line of distinction between his Alexandrian and Western recensions” (An Account of the Printed Text, p. 91). He certainly showed, throughout his “Commentarius Criticus,” that Origen does not lend him the support he had once anticipated; but he still held that the theory of a triple recension was the very _hinge_ on which the whole question turned, and clung to that theory as tenaciously as ever. THIRD EDITION. Dr. Hort (N. T., Introd. p. 186) has since confirmed our opinion that Griesbach was faithful to the last to the essential characteristics of his theory, adding that “the Meletemata of 1811 ... reiterate Griesbach’s familiar statements in precise language, while they show a growing perception of mixture which might have led him to further results if he had not died in the following spring.”

288 It should be also observed that ΦΣ containing SS. Matthew and Mark are probably older than D.

289 E.g. Matt. i. 18; Acts viii. 37 for Irenaeus: Acts xiii. 33 for Origen. It is rare indeed that the express testimony of a Father is so fully confirmed by the oldest copies as in John i. 28, where Βηθανίᾳ, said by Origen to be σχεδὸν ἐν πᾶσι τοῖς ἀντιγράφοις, actually appears in א*ABC*.

290 This view is controverted in Burgon’s “Remains.”

291 Mr. A. A. Vansittart, Journal of Philology, vol. ii. No. 3, p. 35. I suppose too that Mr. Hammond means much the same thing when he says, “It seems almost superfluous to affirm that _every element of evidence must be allowed its full weight_; but it is a principle that must not be forgotten.” (Outlines of Textual Criticism, p. 93, 2nd edition.) Truly it is not superfluous to insist on this principle when we so perpetually find the study of the cursive manuscripts disparaged by the use of what we may venture to call the Caliph Omar’s argument, that if they agree with the older authorities their evidence is superfluous, if they contradict them, it is necessarily false.

292 The evidence of Evan. R, which contains only the decisive letters ΝΗΡΟΥ, is the more valuable, inasmuch as it has been alleged to support the readings of documents of the other class (which no doubt it often does) and thus to afford a confirmation of their authority; it cannot help them much when its vote is against them. On analyzing the 908 readings for which R is cited in Tischendorf’s eighth edition, I find that it sides with A, the representative of the one class, 356 times; with its better reputed rival B 157 times, where A and B are at variance. It is with A alone of the great uncials 101 times, with B alone four, with א alone five, with C alone (but C is lost in 473 places out of the 908) six; with D alone twenty-four. Some of its other combinations are instructive. It is with AC forty-two times and with ACL sixteen; with AD fifty-one and with ADL eighteen; with אB eleven and with אBL twenty-nine; with אL nine times; with AL nineteen; with BL fifteen; with CL never; with DL twice. Cod. R stands unsupported by any of the preceding eighty-nine times, seldom without some countenance (but see Luke xi. 24 ἐκ), such as the Memphitic version, or later codices. In the places where its fragments coincide with those of Cod. Ξ (which is much more friendly to B) they agree 127 times, differ 105.

293 Dean Burgon avers that he is thoroughly convinced that “no reading can be of real importance—I mean has a chance of being _true_—which is witnessed to exclusively by a very few copies, whether uncial or cursive.... Nothing else are such extraordinary readings, _wherever they may happen to be found_, but fragments of primitive error, repudiated by the Church (‘a witness and keeper of Holy Writ’) in her corporate capacity.” (Letter in the _Guardian_, July 12, 1882.) I cannot go quite so far as this. [Dean Burgon has left his reply.]

294 Not that we can in any way assent to the notions of Canon T. R. Birks (Essay on the right estimation of manuscript evidence in the text of the N. T., 1878), whose proposition that “Constant increase of error is no certain and inevitable result of repeated transcription” (p. 33) is true enough in itself, though we cannot follow him when he adds that “Errors, after they have found entrance, may be removed as well as increased in later copies. A careful scribe may not only make fewer mistakes of his own, but he may correct manifest faults of the manuscript from which he copies, and avail himself of the testimony of others, so as to revise and improve the text of that on which he chiefly relies.” Only such a scribe would no longer be a witness for the state of the text as extant in his generation, but a critical editor, working on principles of his own, whether good or bad alike unknown to us.

295 Very pertinent to this matter is a striking extract from J. G. Reiche (a critic “remarkable for extent and accuracy of learning, and for soundness and sobriety of judgement,” as Canon Cook vouches, Revised Version, p. 4), given in Bloomfield’s “Critical Annotations on the Sacred Text,” p. 5, note: “In multis sanè N. T. locis lectionis variae, iisque gravissimi argumenti, de verâ scripturâ judicium firmum et absolutum, quo acquiescere possis, ferri nequit, nisi omnium subsidiorum nostrorum alicujus auctoritatis suffragia, et interna veri falsique indicia, diligenter explorata, justâ lance expendantur.... Quod in causâ est, ut re non satis omni ex parte circumspectâ, non solum critici tantopere inter se dissentiant, sed etiam singuli sententiam suam toties retractent atque commutent.” In the same spirit Lagarde, speaking of the more recent manuscripts of the Septuagint, thus protests: “Certum est eos non a somniis monachorum undecimi vel alius cujusquam saeculi natos, sed ex archetypis uncialibus aut ipsos aut intercedentibus aliis derivatos. Unde elucet criticum acuto judicio et doctrinâ probabili instructum codicibus recentioribus collectis effecturum esse (?) quid in communi plurium aliquorum archetypo scriptum fuerit” (Genesis, p. 19). Compare also Canon Cook, Revised Version of the First Three Gospels, p. 5.

296 “So extravagant a statement could scarcely be deemed worthy of the elaborate confutation with which Dr. Scrivener has condescended to honour it” (_Saturday Review_, Aug. 20, 1881). Yet this scheme of “Comparative Criticism made easy” has obtained, for its childlike simplicity, more acceptance than the reviewer could reasonably suppose. Dr. Hort, of course, speaks very differently: “B must be regarded as having preserved not only a very ancient text, but a very pure line of very ancient text, and that with comparatively small depravation either by scattered ancient corruptions otherwise attested or by individualisms of the scribe himself. On the other hand, to take it as the sole authority except where it contains self-betraying errors, as some have done, is an unwarrantable abandonment of criticism, and in our opinion inevitably leads to erroneous results” (Introd. p. 250).

297 The textual labours of the Cambridge duumvirate have received all the fuller consideration in the learned world by reason of their authors having been members of the New Testament Revision Company, in whose deliberations they had a real influence, though, as a comparison of their text with that adopted by the Revisionists might easily have shown, by no means a preponderating one. I have carefully studied the chief criticisms which have been published on the controversy, without materially adding to the acquaintance with the subject which nearly eleven years of familiar conference with my colleagues had necessarily brought to me. The formidable onslaught on Dr. Hort’s and Bishop Westcott’s principles in three articles in the _Quarterly Review_ [afterwards published together with additions in “The Revision Revised”] especially in the number for April, 1882, and Canon F. C. Cook’s “Revised Version of the First Three Gospels” (1882), must be known to most scholars, and abound with materials from which a final judgement may be formed. “The Ely Lectures on the Revised Version of the N. T.” (1882), which my friend and benefactor Canon Kennedy was pleased to inscribe to myself, are none the less valuable for their attempt to hold the balance even between opposite views of the questions at issue. The host of pamphlets and articles in periodicals which the occasion has called forth could hardly be enumerated in detail, but some of them have been used with due acknowledgement in Chap. XII.

298 We are concerned not with names but with things, so that Dr. Hort may give his _ignis fatuus_ what appellation he likes, only why he calls it Syrian it is hard to determine. The notices connecting his imaginary revision with Lucian of Antioch which we have given above he feels to be insufficient, for he says no more than that “the conjecture derives some little support from a passage of Jerome, which is not itself discredited by the precariousness of the modern theories which have been suggested by it” (Hort, p. 138).

_ 299 See_ Burgon’s “The Revision Revised,” pp. 271-288.

300 Other examples may be seen in our notes in Chap. XII on Luke ii. 14 for Methodius; Luke xxii. 43, 44 for Hippolytus again; Luke xxiii. 34 for Irenaeus and Origen. Add Luke x. 1 for Irenaeus (p. 546, note 1); xxiii. 45 (Hippolytus); John xiii. 24 (Clem. Alex.); 2 Cor. xii. 7 (Iren. Orig.); Mark xvi. 17, 18 (Hippol.). _See_ also Miller’s “Textual Guide,” pp. 84, 85, where 165 passages on fifteen texts are gathered from writers before St. Chrysostom.

301 For reasons which will be readily understood, we have quoted sparingly from the trenchant article in the _Quarterly Review_, April, 1882, but the following summary of the consequences of a too exclusive devotion to Codd. אB seems no unfit comment on the facts of the case: “Thus it would appear that the Truth of Scripture has run a very narrow risk of being lost for ever to mankind. Dr. Hort contends that it more than half lay _perdu_ on a forgotten shelf in the Vatican Library;—Dr. Tischendorf that it had found its way into a waste-paper basket in the convent of St. Catherine at the foot of Mount Sinai—from which he rescued it on February 4, 1859:—neither, we venture to think, a very likely supposition. We incline to believe that the Author of Scripture hath not by any means shown Himself so unmindful of the safety of the Deposit, as these learned persons imagine” (p. 365). The Revision Revised, p. 343.

_ 302 See_ Appendix of passages at the end of this chapter. Yet while refusing without hesitation the claim of the _monstra_ which follow to be regarded as a part of the sacred text, we are by no means insensible to the fact impressed upon us by the Dean of Llandaff, that there are readings which conciliate favour the more we think over them: it being the special privilege of Truth always to grow upon candid minds. We subjoin his persuasive words: “It is deeply interesting to take note of the process of thought and feeling which attends in one’s own mind the presentation of some unfamiliar reading. At first sight the suggestion is repelled as unintelligible, startling, almost shocking. By degrees, light dawns upon it—it finds its plea and its palliation. At last, in many instances, it is accepted as adding force and beauty to the context, and a conviction gradually forms itself that thus and not otherwise was it written.” (Vaughan, Epistle to Romans, Preface to the third edition, p. xxi.)

303 Thus far we are in agreement with the “Two Members of the N. T. Company,” however widely we may differ from their general views: “The great contribution of our own times to a mastery over materials has been the clearer statement of the method of genealogy, and, by means of it, the corrected distribution of the great mass of documentary evidence” (p. 19). Only that arbitrary theories ought to be kept as far as possible out of sight.

304 So that we may be sure what we should have found in Cod. D, and with high probability in Cod. E, were they not defective, when in Acts xxvii. 5 we observe δι᾽ ἡμερῶν δεκάπεντε inserted after διαπλεύσαντες in 137, 184, and the Harkleian margin with an asterisk; as also when we note in Acts xxviii. 16 ἔξω τῆς παρεμβολῆς before σύν in the last two and in _demid._

305 E.g. Luke xxiv. 3 τοῦ κυρίου ἰησοῦ omitted by D, _a_ _b_ _e_ _ff_2 _l_; ver. 6 οὐκ ἔστιν ὦδε ἀλλὰ ἠγέρθη (comp. Mark xvi. 6), omitted by the same; ver. 9 ἀπὸ τοῦ μνημείου by the same, by _c_ and the Armenian; the whole of ver. 12, by the same (except _ff_2) with _fuld._, but surely not by the Jerusalem Syriac, even according to Tischendorf’s showing, or by Eusebius’ canon, for he knew the verse well (comp. John xx. 5); ver. 36 καὶ λέγει αὐτοῖς, εἰρήνη ὑμῖν omitted by D, _a_ _b_ _e_ _ff_2 _l_ as before (comp. John xx. 19, 26); the whole of ver. 40, omitted by the same and by Cureton’s Syriac (comp. John xx. 20); ver. 51 καὶ ἀνεφέρετο εἰς τὸν οὐρανόν and ver. 52 προσκυνήσαντες αὐτόν omitted by the same and by Augustine, the important clause in ver. 51 by א* also, and consequently by Tischendorf. Yet, as if to show how mixed the evidence is, D deserts _a_ _b_ _ff_2 _l_ when, in company with a host of authorities, both manuscripts and versions (_f_ _q_, Vulgate, Bohairic, Syriac, and others), they annex καὶ ἀπὸ μελισσίου κηρίου to the end of ver. 42. _See_ also Luke x. 41, 42; xxii. 19, 20, discussed in Chap. XII.

306 So of certain of the chief versions we sometimes hear it said that they are less important in the rest of the N. T. than in the Gospels; which means that in the former they side less with אB.

307 Canon Kennedy, whose “Ely Lectures” exhibit, to say the least, no prejudice against the principles enunciated in Dr. Hort’s Introduction, is good enough to commend the four rules here set forth to the attention of his readers (p. 159, note). The first three were stated in my first edition (1861), the fourth added in the second edition (1874), and, while they will not satisfy the advocates of extreme views on either side, suffice to intimate the terms on which the respective claims of the uncial and cursive manuscripts, of the earlier and the more recent authorities, may, in my deliberate judgement, be equitably adjusted.

308 Dean Burgon held that too much deference is here paid to the mere antiquity of those which happen to be the oldest MSS., but are not the oldest authorities. He would therefore enlarge the grounds of judgement.

309 The harmony subsisting between B and the Sahidic in characteristic readings, for which they stand almost or quite alone, is well worth notice: e.g. Acts xxvii. 37; Rom. xiii. 13; Col. iii. 6; Heb. iii. 2; 1 John ii. 14; 20.

310 “The intrinsic evidence seems immoveable against the insertion.” Textual Criticism of the N. T., B. B. Warfield, D.D., p. 135.

311 Yet in Penn’s “Annotations to the Vatican Manuscripts” (1837) “The restoration of this verse to its due place” is described as “the most important circumstance of this [sc. his own] revision.” Its omission is imputed to “the undue influence of a criticism of Origen [ἤδη δὲ αὐτοῦ ἀποθανόντος], whom Jerome followed.”

312 “This gross perversion of the truth, alike of Scripture and of history—a reading as preposterous as it is revolting,” is the vigorous protest of Dean Burgon, The Revision Revised, p. 68, note.

313 “Post enim duodecim apostolos septuaginta alios Dominus noster ante se misisse invenitur; septuaginta autem nec octonario numero neque denario” (Irenaeus, p. 146, Massuet). Tertullian, just a little later (re-echoed by the younger Cyril), compares the Apostles with the twelve wells at Elim (Ex. xv. 27), the seventy with the three-score and ten palm-trees there (Adv. Marc. iv. 24). So Eusebius thrice, Basil and Ambrose. On the other hand in the Recognitions of Clement, usually assigned to the second or third century, the number adopted is seventy-two, “vel hoc modo recognitâ imagine Moysis” and of his elders, traditionally set down at that number. Compare Num. xi. 16. Epiphanius, Hilary (Scholz), and Augustine are also with Cod. B.

314 To enable us to translate “a son, nay even an ox,” would require ἢ καί, which none read. The argument, moreover, is one _a minori ad majus_. Compare Ex. xxi. 33 with Ex. xxiii. 4; ch. xiii. 15.

315 Let me add _ex meo_ Codd. 22, 219, 492, 547, 549, 558, 559, 576, 582, 584, 594, 596, 597, 598, 601, being no doubt a large majority of cursives. So Cod. 662, apparently after correction.

316 But not in the Beirût MS. discovered in 1877 by Dr. Is. H. Hall.

317 A more ludicrous blunder of Cod. B has been pointed out to me in the Old Testament, Ps. xvii. 14 “they have children at their desire”: ΕΧΟΡΤΑΣΘΗΣΑΝ ΫΙΩΝ Cod. A, but ΕΧΟΡΤΑΣΘΗΣΑΝ ΫΕΙΩΝ Cod. B. The London papyrus has ΥΩΝ for ΥΙΩΝ.

318 Codex P is of far greater value than others of its own date. It is frequently found in the company of B, sometimes alone, sometimes with other chief authorities, especially in the Catholic Epistles, e.g. James iv. 15; v. 4; 14; 2 Pet. i. 17 (partly); ii. 6; 1 John ii. 20.

319 We note many small variations between the text of these critics as communicated to the Revisers some years before, and that finally published in 1881. The latter, of course, we have treated as their standard.

320 This precious cursive forms one of a small class which in the Catholic Epistles and sometimes in the Acts conspire with the best uncials in upholding readings of the higher type: the other members are 69, 137, 182, to which will sometimes be added the text or margin of the Harkleian Syriac, Codd. 27, 29, the second hands of 57 and 66, 100, 180, 185, and particularly 221, which is of special interest in these Epistles. The following passages, examined by means of Tischendorf’s notes, will prove what is here alleged: 1 Pet. iii. 16; 2 Pet. i. 4; 21; ii. 6; 11; 1 John i. 5; 7; 8; ii. 19; iii. 1; 19; 22; iv. 19; v. 5.

321 Notice especially those instances in the Catholic Epistles, wherein the primary authorities are comparatively few, in which Cod. B accords with the later copies against Codd. אA(C), and is also supported by internal evidence: e.g. 1 Pet. iii. 18; iv. 14; v. 2; 2 Pet. ii. 20; 1 John ii. 10; iii. 23, &c. In 1 John iii. 21, where the first ἡμῶν is omitted by A and others, the second by C almost alone, B seems right in rejecting the word in both places. So in other cases internal probabilities occasionally plead strongly in favour of B, when it has little other support: as in Rom. viii. 24, where τίς ἐλπίζει; as against τις, τί καὶ ἐλπίζει; though B and the margin of Cod. 47 stand alone here, best accounts for the existence of other variations (_see_ p. 248). In Eph. v. 22, B alone, with Clement and Jerome, the latter very expressly, omits the verb in a manner which can hardly fail to commend itself as representing the true form of the passage. In Col. iii. 6, B, the Sahidic, the Roman Ethiopic, Clement (twice), Cyprian, Ambrosiaster, and auct. de singl. cler., are alone free from the clause interpolated from Eph. v. 6.

322 Viz. Luke i. 1-4, some portion of the Gospel and most of the Acts: excluding such cases as St. Stephen’s speech, Acts vii, and the parts of his Gospel which resemble in style, and were derived from the same sources as, those of SS. Matthew and Mark.

323 Dr. Hort (Introd., Notes, p. 141) confirms the foregoing statements, which we have repeated unchanged from our former editions. “What spellings are sufficiently probable to deserve inclusion among alternative readings, is often difficult to determine. Although many deviations from classical orthography are amply attested, many others, which appear to be equally genuine, are found in one, two, or three MSS. only, and that often with an irregularity which suggests that all our MSS. have to a greater or less extent suffered from the effacement of unclassical forms of words. It is no less true on the other hand that a tendency in the opposite direction is discernible in Western MSS.: the orthography of common life, which to a certain extent was used by all the writers of the New Testament, though in unequal degrees, would naturally be introduced more freely in texts affected by an instinct of popular adaptation.”

324 E.g. Aeschylus, Persae, 411: κόρυμβ᾽, ἐπ᾽ ἄλλην δ᾽ ἄλλος ἴθυνεν δόρυ, or Sophocles, Antigone, 219: τὸ μὴ πιχωρεῖν τοῖς ἀπιστοῦσιν τάδε.

325 Cod. א, for instance, does not omit it above 208 times throughout the N. T., out of which 134 occur with verbs (three so as to cause a hiatus), 29 with nouns, 45 with adjectives (chiefly πᾶσι) or participles (Scrivener, Collation, &c., p. liv). Its absence produces the hiatus in B*C in 1 Pet. ii. 18 (ἐπιεικέσι), and not seldom in B, e.g. 1 Pet. iv. 6, where we find κριθῶσι and ζῶσι, which latter is countenanced by A, and both by אL.

326 Wake 12 (Evan. 492), of the eleventh century, may be taken for a fair representative of its class and date. It retains ν with εἶπεν thirty-three times in St. Matthew, thirteen in St. Mark, as often as 130 in St. Luke. With other words it mostly reserves ν to indicate emphasis (e.g. Luke xxii. 14; xxiv. 30), or to stand before a break in the sense.

327 The terminations which admit this moveable ν (including -ει of the pluperfect) are enumerated by Donaldson (Gr. Gram. p. 53). Tischendorf, however (N. T., Proleg. p. liv), demurs to εἴκοσιν, even before a vowel.

328 With the remarkable exception of those six leaves of Cod. א which Tischendorf assigns to the scribe who wrote Cod. B. In these leaves of Cod. א Ἰωάνης occurs four times: Matt. xvi. 14; xvii. 1; 13; Luke i. 13, in which last passage, however, B has the double _nu_.

329 These last might be supposed to have originated from the omission or insertion of the faint line for ν over the preceding letter, which (especially at the end of a line) we stated in Vol. I. p. 50 to be found even in the oldest manuscripts. Sometimes the anomalous form is much supported by junior as well as by ancient codices: e.g. θυγατέραν, Luke xiii. 16 by KXΓ*Λ, 209, also by 69, and ten others of Scrivener’s.

330 Thus Canon Selwyn cites from Lycophron κἀπὸ γῆς ἐσχάζοσαν, and Dr. Moulton (Winer, p. 91, note 5), after Mullach, ἔσχοσαν from Scymnus Chius.

331 Tregelles presses yet another argument: “If Alexandrian forms had been introduced into the N. T. by Egyptian copyists, how comes it that the classical MSS. written in that country are free from them?” (An Account of the Printed Text, p. 178). But what classical MSS. does he know of, written while Egypt was yet Greek or Christian, and now extant for our inspection? I can only think of Cureton’s Homer and Babington’s papyri.

332 “It is hard to make St. Paul responsible for vulgarisms or provincialisms, which certainly his pen never wrote, and which there can be no proof that his lips ever uttered” (Epistle to the Romans, Preface to the third edition, p. xxi) is Dean Vaughan’s comment on this “barbarism.” He regards the Apostle’s habit of dictating his letters as a “sufficient reason for broken constructions, for participles without verbs, for suspended nominatives, for sudden digressions, for fresh starts.”

333 Dr. Hort, however, accepts the form ἐφ᾽ in this place, aspirating ἐλπίδι, and in the same way favours but does not print οὐχ ὁλίγος eight times in the Acts, adding that although ὁλίγος “has no lost digamma to justify it, like some others, it may nevertheless have been in use in the apostolic age: it occurs in good MSS. of the LXX” (Introd., Notes, p. 143).

334 “A Treatise on the Grammar of New Testament Greek regarded as the basis of N. T. Exegesis. By Dr. G. B. Winer. Translated from the German with large additions and full indices by Rev. W. F. Moulton, M. A., D. D.,” third edition revised, 8vo, Edinburgh, 1882. The forthcoming “Prolegomena” to Tischendorf’s N. T. eighth edition (pp. 71-126), to which the kindness of Dr. Caspar René Gregory has given me access, contain a store of fresh materials on this subject; and Dr. Hort’s “Notes on Orthography” (Introd., Notes, pp. 143-173) will afford invaluable aid to the student who is ever so little able to accept some of his conclusions. See also on the more general subject Dr. Neubauer’s Article in the first issue of the Oxford “Studia Biblica” on “The Dialects of Palestine in the Time of Christ.” He controverts Dr. Roberts’ opinion that “Christ spoke for the most part in Greek, and only now and then in Aramaic.” And he distinguishes between the Babylonian Aramaic, the Galilean Aramaic, and the dialect spoken at Jerusalem, which had more of Hebrew.

335 In Acts ix. 34 Ἰησοῦς Χριστός, the article between them being rejected, is read by Lachmann, Tischendorf, Tregelles, Westcott and Hort, on the adequate authority of אB*C, 13, 15, 18, 68, 111, 180, and a catena (probably also Cod. 36), with one or two Fathers, although against AEP, 31, 61, &c.

336 I know not why Tischendorf cites Cod. 71 (gscr) for the omission of Ἰησοῦ. I have again consulted the MS. at Lambeth, and find _ἰῦ_ in this place.

_ 337 See_ above, I. 130. The precise relation of the Latin Version of Cod. D to the parallel Greek text is fully examined in Scrivener’s “Codex Bezae,” Introduction, chap. iii.

338 Mr. E. B. Nicholson, Bodley’s Librarian, doubts the conclusiveness of Irenaeus’ Latin here “because his copyist was in the habit of altering him into accordance with the oldest Latin version; and because his argument is just as strong if we read _Jesu Christi autem_ as if we read _Christi_. The argument requires _Christi_, but does not in the least require it as against _Jesu Christi_.”

339 “The clearly Western Τοῦ δὲ χριστοῦ,” as Dr. Hort admits, “is intrinsically free from objection, ... yet it cannot be confidently accepted. The attestation is unsatisfactory, for no other Western omission of a solitary word in the Gospels has any high probability” (N. T., Notes, p. 7). He retains ψευδόμενοι, Matt. v. 11.

340 Why should Gregory Nyssen (371) be classed among the opponents of the clause, whereas Griesbach honestly states, “suam expositionem his quidem verbis concludit: [ἀπὸ τοῦ πονηροῦ τοῦ ἐν τῷ κόσμῳ τούτῳ τὴν ἰσχὺν κεκτημένου, οὗ ῥυσθείημεν] χάριτι [τοῦ] χριστοῦ, ὅτι αὐτοῦ ἡ δύναμις καὶ ἡ δόξα ἅμα τῷ πατρὶ καὶ τῷ ἁγίῳ πνεύματι, νῦν καὶ ἀεὶ καὶ εἰς τοὺς αἰῶνας τῶν αἰώνων, ἀμήν”? Griesbach adds indeed, “sed pro parte sacri textûs neutiquam haec habuisse videtur;” and justly: they were rather a _loose paraphrase_ of the sentence before him. _See_ Textual Guide, Edward Miller, App. V.

341 Canon Cook (Revised Version, p. 57) alleges as a probable cause of the general omission of the doxology in early Latin Versions and Fathers, that in all the Western liturgies it is separated from the petitions preceding by an intercalated _Embolismus_. More weighty is his observation that all the Greek Fathers, from Chrysostom onwards, who deal with the interpretation of the Lord’s Prayer, “agree with that great expositor in maintaining the important bearings [of the doxology] upon the preceding petitions.”

342 “Quite a test-passage” Mr. Hammond calls it (Outlines of Text. Crit., p. 76).

343 THIRD EDITION. I would fain side in this instance with my revered friend and Revision colleague Dr. David Brown of Aberdeen, and all my prepossessions are strongly in favour of the _textus receptus_ here. He is quite right in perceiving (Christian Opinion and Revisionist, p. 435) that the key of his position lies in the authenticity of ἀγαθέ ver. 16, which is undoubtedly found in Mark x. 17; Luke xviii. 18. If that word had abided unquestioned here, the form of reply adopted in the other two Gospels would have inevitably followed. As the case stands, there is not considerably less evidence for omitting ἀγαθέ (אBDL, 1, 22, 479, Evst. 5 [_not_ “five Evangelistaria”], _a_ _e_ _ff_1, Eth., Origen twice, Hilary) than for Τί με ἐρωτᾷς κ.τ.λ., although Cureton’s and the Jerusalem Syriac, the Bohairic, and the Vulgate with some other Latin copies, change sides here. It is upon these recreant versions that Dr. Brown must fix the charge of inconsistency. If ἀγαθέ be an interpolation, surely τί ἀγαθὸν ποιήσω is pertinently answered by Τί με ἐρωτᾷς περὶ τοῦ ἀγαθοῦ.

344 Canon Westcott (Smith’s “Dictionary of the Bible,” Vulgate Version) adds Bodl. 857; Brit. Mus. Reg. I B. vii, and Reg. I. A. xviii in part, also Addit. 24,142 by the second hand. Tischendorf also cites _theotisc_.

345 No passage more favours Bp. Middleton’s deliberate conclusion respecting the history of the Codex Bezae: “I believe that no fraud was intended: but only that the critical possessor of the basis filled its margin with glosses and readings chiefly from the Latin, being a Christian of the Western Church; and that the whole collection of Latin passages was translated into Greek, and substituted in the text by some one who had a high opinion of their value, and who was, as Wetstein describes him, ‘καλλιγραφίας quàm vel Graecae vel Latinae linguae peritior.’ ” (Doctrine of the Greek Article, Appendix I. p. 485, 3rd edition.)

346 I see no reasonable ground for imagining with Lachmann that Origen who, as he truly observes, “non solet difficilia praeterire,” did not find in his copy anything between πατρός; and Ἀμήν in ver. 31. On the supposition that he read πρῶτος there was no difficulty to slur over. Moreover, there is not a vestige of evidence for omitting λέγει αὐτοῖς ὁ ἰησοῦς, the existence of which words Lachmann clearly perceived to be fatal to his ingenious guess, although Dr. Hort will only allow that it “weakens his suggestion,” adding in his quiet way “This phrase might easily seem otiose if it followed immediately on words of Christ, and might thus be thought to imply the intervention of words spoken by others” (Notes, p. 17).

347 Jerome conceives that the Jews “intellegere quidem veritatem, sed tergiversari, et nolle dicere quod sentiunt;” and so Canon G. F. Goddard, Rector of Southfleet, believed that their wantonly false answer brought on them the Lord’s stern rebuke. Hilary’s idea is even more far-fetched: viz. that though the second son disobeyed, it was because he _could_ not execute the command. “Non ait noluisse sed non abisse. Res extra culpam infidelitatis est, quia in facti erat difficultate ne fieret.”

348 His sole example is ὁδὸν ποιεῖν Mark ii. 23, which seems not at all parallel. The phrase may as well signify to “clear away” as “make their way.”

349 πολλὰ ἂ ἐποίει is the reading of Abbott’s four and of Codd. 28, 122, 541, 561, 572, Evst. 196.

350 Which is certainly its meaning in Lucian, Tom. ii. p. 705 (Salmur. 1619); I know no example like that in St. Mark.

351 I have ventured but slowly to vouch for Tischendorf’s notion, that six leaves of Cod. א, _that containing_ Mark xvi. 2-Luke i. 56 _being one of them_, were written by the scribe of Cod. B. On mere identity of handwriting and the peculiar shape of certain letters who shall insist? Yet there are parts of the case which I know not how to answer, and which have persuaded even Dr. Hort. Having now arrived at this conclusion our inference is simple and direct, that at least in these leaves, Codd. אB make but one witness, not two.

352 The cases of Nehemiah, Tobit, and Daniel, in the Old Testament portion of Cod. B, are obviously in no wise parallel in regard to their blank columns.

353 Of which supplement Dr. Hort says unexpectedly enough, “In style it is unlike the ordinary narratives of the Evangelists, but comparable to the four introductory verses of St. Luke’s Gospel” (Introduction, p. 298).

354 We ought to add that some Armenian codices which contain the paragraph have the subscription “Gospel after Mark” at the end of ver. 8 as well as of ver. 20, as though their scribes, like Cod. L’s, knew of a double ending to the Gospel.

355 Burgon (_Guardian_, July 12, 1882) speaks of seven manuscripts (Codd. 538, 539 being among them) wherein these last twelve verses begin on the right hand of the page. This would be more significant if a space were left, as is not stated, at the foot of the preceding page. In Cod. 550 the first letter α is small, but covers an abnormally large space.

356 Of course no notice is to be taken of τέλος after ἐφοβοῦντο γάρ, as the end of the ecclesiastical lesson is all that is intimated. The grievous misstatements of preceding critics from Wetstein and Scholz down to Tischendorf, have been corrected throughout by means of Burgon’s laborious researches (Burgon, pp. 114-123).

357 The minute variations between these several codices are given by Burgon (Appendix E, pp. 288-90). Cod. 255 contains a scholion imputed to Eusebius, from which Griesbach had drawn inferences which Burgon (Last Twelve Verses, &c., Postscript, pp. 319-23) has shown to be unwarranted by the circumstances of the case.

358 Dr. C. Taylor, Master of St. John’s College, Cambridge, in _The Expositor_ for July, 1893, quotes more evidence from Justin Martyr—hinting that some also remains behind—proving that that Father was familiar with these verses. Also he cites several passages from the Epistle of Barnabas in which traces of them occur, and from the Quartodeciman controversy, and from Clement of Rome. The value of the evidence which Dr. Taylor’s acute vision has discovered consists chiefly in its cumulative force. From familiarity with the passage numerous traces of it arose; or as Dr. Taylor takes the case reversely, from the fact of the occurrence of numerous traces evident to a close observer, it is manifest that there pre-existed in the minds of the writers a familiarity with the language of the verses in question.

359 It is surprising that Dr. Hort, who lays very undue stress upon the silence of certain early Christian writers that had no occasion for quoting the twelve verses in their extant works, should say of Cyril of Jerusalem, who lived about A.D. 349, that his “negative evidence is peculiarly cogent” (Notes, p. 37). To our mind it is not at all negative. Preaching on a Sunday, he reminds his hearers of a sermon he had delivered the day before, and which he would have them keep in their thoughts. One of the topics he briefly recalls is the article of the Creed τὸν καθίσαντα ἐκ δεξιῶν τοῦ πατρός. He must inevitably have used Mark xvi. 19 in his Saturday’s discourse.

360 Several of these references are derived from “The Revision Revised,” p. 423.

361 Nor were these verses used in the Greek Church only. Vers. 9-20 comprised the Gospel for Easter Monday in the old Spanish or Mozarabic Liturgy, for Easter Tuesday among the Syrian Jacobites, for Ascension Day among the Armenians. Vers. 12-20 was the Gospel for Ascension Day in the Coptic Liturgy (Malan, Original Documents, iv. p. 63): vers. 16-20 in the old Latin _Comes_.

362 To get rid of one apparent ἀντιφωνία, that arising from the expression πρωῒ τῇ μιᾷ τοῦ σαββάτου (_sic_), ver. 9, compared with ὀψὲ σαββάτων Matt. xxviii. 1, Eusebius proposes the plan of setting a stop between Ἀναστὰς δέ and πρωΐ, so little was he satisfied with rudely expunging the whole clause. Hence Cod. E puts a red cross after δέ: Codd. 20, 22, 34, 72, 193, 196, 199, 271, 345, 405, 411, 456, have a colon: Codd. 332, 339, 340, 439, a comma (Burgon, _Guardian_, Aug. 20, 1873).

363 The following peculiarities have been noticed in these verses: ἐκεῖνος used absolutely, vers. 10, 11, 13; πορεύομαι vers. 10, 12, 15; τοῖς μετ᾽ αὐτοῦ γενομένοις ver. 10; θεάομαι vers. 11, 14; ἀπιστέω vers. 11, 16; μετὰ ταῦτα ver. 12; ἕτερος ver. 12; παρακολουθέω ver. 17; ἐν τῷ ὀνόματι ver. 17; κύριος for the Saviour, vers. 19, 20; πανταχοῦ, συνεργοῦντος, βεβαιόω, ἐπακολουθέω ver. 20, all of them as not found elsewhere in St. Mark. A very able and really conclusive plea for the genuineness of the paragraph, as coming from that Evangelist’s pen, appeared in the _Baptist Quarterly_, Philadelphia, July, 1869, bearing the signature of Professor J. A. Broadus, of South Carolina. Unfortunately, from the nature of the case, it does not admit of abridgement. Burgon’s ninth chapter (pp. 136-190) enters into full details, and amply justifies his conclusion that the supposed adverse argument from phraseology “breaks down hopelessly under severe analysis.”

364 “Can any one, who knows the character of the Lord and of His ministry, conceive for an instant that we should be left with nothing but a message baulked through the alarm of women” (Kelly, Lectures Introductory to the Gospels, p. 258). Even Dr. Hort can say: “it is incredible that the Evangelist deliberately concluded either a paragraph with ἐφοβοῦντο γάρ, or the Gospel with a petty detail of a secondary event, leaving his narrative hanging in the air” (Notes, p. 46).

365 When Burgon ventures upon a surmise, one which is probability itself by the side of those we have been speaking of, Professor Abbot (_ubi supra_, p. 197) remarks upon it that “With Mr. Burgon a conjecture seems to be a demonstration.” We will not be deterred by dread of any such reproach from mentioning his method of accounting for the absence of these verses from some very early copies, commending it to the reader for what it may seem worth. After a learned and exhaustive proof that the Church lessons, as we now have them, existed from very early times (Twelve Verses, pp. 191-211), and noting that an important lesson ended with Mark xvi. 8 (_see_ Calendar of Lessons); he supposes that τέλος, which would stand at the end of such a lesson, misled some scribe who had before him an _exemplar_ of the Gospels whose last leaf (containing Mark xvi. 9-20, or according to Codd. 20, 215, 300 only vers. 16-20) was lost, as it might easily be in those older manuscripts wherein St. Mark stood last.

366 The Codex lately discovered by Mrs. Lewis is said to omit the verses. But what is that against a host of other codices? And when the other MS. of the Curetonian includes the verses? Positive testimony is worth more than negative.

367 Dr. Hort, however, while he admits the possibility of the leaf containing vers. 9-20 having been lost in some very early copy, which thus would become the parent of transcripts having a mutilated text (Notes, p. 49), rather inconsistently arrives at the conclusion that the passage in question “manifestly cannot claim any apostolic authority; but it is doubtless founded on some tradition of the apostolic age” (_ibid._ p. 51).

368 Dr. Hort will hardly find many friends for his division (Notes, p. 56),

Δόξα ἐν ὑψίστοις θεῷ καὶ ἐπὶ γῆς, Εἰρήνη ἐν ἀνθρώποις εὐδοκίας.

369 I am loth to sully with a semblance of unseasonable levity a page which is devoted to the vindication of the true form of the Angelic Hymn, and must ask the student to refer for himself to the 470th number of the _Spectator_, where what we will venture to call a precisely parallel case exercises the delicate humour of Addison. “So many ancient manuscripts,” he tells us, concur in this last reading, “that I am very much in doubt whether it ought not to take place. There are but two reasons which incline me to the reading as I have published it: first, because the rhyme, and secondly, because the sense, is preserved by it.”

370 This torrent of testimony includes ninety-two places, of which “Tischendorf knew of only eleven, Tregelles adduces only six” (R. R., p. 45, note).

371 Every word uttered by such a scholar as Dr. Field (d. 1885) is so valuable that no apology can be needed for citing the following critique from his charming “Otium Norvicense,” Part iii. p. 36, on the reading εὐδοκίας and the rendering “among men in whom he is well pleased.” “To which it may be briefly objected (1) _that it ruins the stichometry_; (2) that it separates ἐν from εὐδοκία, the word with which it is normally construed; (3) that ‘men of good pleasure’ (אנשי רצון) would be, according to Graeco-biblical usage, not ἄνθρωποι εὐδοκίας, but ἄνδρες εὐδοκίας; (4) that the turn of the sentence, ἐν ἀνθρώποις εὐδοκία, very much resembles the second clause of Prov. xiv. 9: ובין ישרים רצון rendered by Symmachus καὶ ἀναμέσον εὐθέων εὐδοκία.” But this is almost slaying the slain.

372 Κυριακὴ δευτεροπρώτη is cited by Sophocles in his Lexicon from “Eustr. 2381 B” in the sense of _low Sunday_ (McClellan, N. T., p. 690). Canon Cook conjectures that it may mean the first sabbath in the second month (_Iyar_), precisely the time when wheat would be fully ripe (Revised Version, p. 69). [More probably it is “the first sabbath after the second day of the Passover.”] On the other hand, “If the word be a reality and originally in the text, its meaning, since in that case it must have been borrowed from something in the Jewish calendar, would have been traditionally known from the first.” (Green, Course of Developed Criticism, p. 56.) But why would it? The fancy that δευτεροπρώτῳ had its origin in numerals of reference (B A) set in the margin will most commend itself to such scholars as are under the self-imposed necessity of upholding Codd. אB united against all other evidence, of whatever kind.

373 Just as Jerome, speaking of the latter part of 1 Cor. vii. 35, says, “In Lat. Codd. OB TRANSLATIONIS DIFFICULTATEM hoc penitus non invenitur.” (Vallars. ii. 261, as Burgon points out.)

374 Dr. Hort and the _Quarterly Reviewer_ (October, 1881, p. 348) almost simultaneously called attention to the question put by Jerome to his teacher Gregory of Nazianzus as to the meaning of this word. “Docebo te super hac re in ecclesia” was the only reply he obtained; on which Jerome’s comment is, _Eleganter lusit_ (Hier. _ad Nepotianum_, Ep. 52). Neither of these great Fathers could explain a term which neither doubted to be written by the Evangelist.

375 Cyril applies the whole passage to enforce the duty of exercising with frugality the Christian duty of entertaining strangers: “And this He did for our benefit, that He might fix a limit to hospitality” (Dean Payne Smith’s Translation, pp. 317-20).

376 Praelectio in Scholis Cantabrigiensibus habita Februarii die decimo quarto, MDCCCL, quâ ... Lucae pericopam (xxii. 17-20) multis ante saeculis conturbatam vetustissimorum ope codicum in pristinam formam restituebat, Cathedram Theologicam ambiens, J. W. Blakesley, S. T. B., Coll. SS. Trinitatis nuper Socius (Cambridge, 1850).

377 “Intrinsically both readings are difficult, but in unequal degrees. The difficulty of the shorter reading [that of pure omission in vers. 19, 20] consists exclusively in the change of order, as to the Bread and the Cup, which is illustrated by many phenomena of the relation between the narratives of the third and of the first two Gospels, and which finds an exact parallel in the change of order in St. Luke’s account of the Temptation” (iv. 5-8; 9-12). Hort, Notes, p. 64.

378 Adler says “in omnibus codicibus,” and _guelph. heidelb._ Dawkins iii and xvii in Jones, and cod. Rich are specified. Lee sets the verses in a parenthesis. But the Curetonian has them after ver. 19 in words but little differing from his or Schaaf’s.

379 “Si fides habenda A. F. Gorio ‘in Conspectu Quattuor Codicum Evangeliorum Syriacorum mirae aetatis’ apud Blanchini Evangelium Quadruplex p. DXL, et hi quattuor Codices cum Veronensi [_b_] faciunt.” Blakesley, _Schema_ facing _Praelectio_, p. 20.

380 Especially mark his mode of dealing with ἐκχυννόμενον ver. 20, which by a little violence (not quite unprecedented) is made to refer to ποτήριον instead of to αἵματι: “Ex Matthaeo vel Marco accessit clausula ista τὸ ὑπὲρ ὑμῶν ἐκχυννόμενον, fraude tamen ita piâ accessit, ut potius grammaticis legibus vim facere, quam vel literulam demutare maluerit interpolator. Ita fit ut vel hodie male assutus pannus centonem prodat. Postulat enim sermonis ratio, ut cuivis patet, τῷ ὑπὲρ ὑμῶν ἐκχυνομένω, non τὸ ὑπὲρ ὑμῶν ἐκχυνόμενον, quod tamen in Matthaeo Marcoque optime Graece dicebatur, cum subjectum de quo praedicabatur non ἡ διαθήκη verum τὸ αἷμα esset” (_Praelectio_, p. 22).

381 Very undue stress has been laid on Tischendorf’s statement, “Hos versus A corrector uncis inclusit, partim etiam punctis notavit; C vero puncta et uncos delevit,” and אa has sometimes been spoken of as only a little less weighty than א itself. I had the satisfaction, through Dean Burgon’s kindness, of showing some of our critics, Dr. Hort included, a fine photograph of the whole page. The points are nearly, if not quite, invisible, the unci are rude slight curves at the beginning and end of the passage only, looking as likely to have been scrawled fifty years since as fourteen hundred. Yet even now Dr. Hort maintains that Tischendorf’s decision is probably right, strangely adding, “but the point is of little consequence” (Notes, p. 65).

382 Bp. Lightfoot’s Codd. 2, 4, 8, 9, 16, 17, 19, 22, 26 omit them altogether: they are in the margin of 1, 20. They stand in the text of 3, 14, 21, and so in 18 _primâ manu_, but in smaller characters.

383 Yet Dr. Hort contends that “The testimony of A is not affected by the presence of Eusebian numerals, of necessity misplaced, which manifestly presuppose the inclusion of vv. 43, 44: the discrepance merely shows that the Biblical text and the Eusebian notation were taken by the scribe from different sources, as they doubtless were throughout” (Notes, p. 65). It is just this readiness to devise expedients to meet emergencies as they arise which is at once the strength and the weakness of Dr. Hort’s position as a textual critic. These sections and canons illustrate the criticism of the text in some other places: e.g. Matt. xvi. 2, 3; xvii. 21; ch. xxiii. 34; hardly in Luke xxiv. 12.

384 Ἰστέον ὅτι τὰ περὶ τῶν θρόμβων τινὰ τῶν ἀντιγράφων οὐκ ἔχουσιν: adding that the clause is cited by Dionysius the Areopagite, Gennadius, Epiphanius, and other holy Fathers.

385 Thus in Evst. 253 we find John xiii. 3-17 inserted _uno tenore_ between Matt. xxvi. 20 and 21, as also Luke xxii. 43, 44 between vers. 39 and 40, with no break whatever. So again in the same manuscript with the mixed lessons for Good Friday.

386 “Upwards of forty famous personages from every part of ancient Christendom recognize these verses as part of the Gospel; fourteen of them being as old, some of them being a great deal older, than our oldest manuscripts” (The Revision Revised, p. 81).

387 The reader will see that I have understood this passage, with Grotius, as applying to an orthodox tampering with Luke xix. 41, not with xxii. 43, 44. As the text of Epiphanius stands I cannot well do otherwise, since Mill’s mode of punctuation (N. T., Proleg. § 797), which wholly separates καὶ γενόμενος from the words immediately preceding, cannot be endured, and leaves καὶ τὸ ἰσχυρότατον unaccounted for. Yet I confess that there is no trace of any meddling with ἔκλαυσε by any one, and I know not where Irenaeus cites it.

388 Lightfoot’s Codd. 22, 26. The clause stands in the margin of 1, 20, in the text of 2, 3, 8, 9, 14, 16, 17, 19, 21, 23.

389 Dean Burgon (Revision Revised, p. 83), who refers to upwards of forty Fathers and more than 150 passages (_see_ also Miller’s Textual Guide, App. II), burns with indignation as he sums up his results: “And _what_ (we ask the question with sincere simplicity), _what_ amount of evidence is calculated to inspire undoubted confidence in any given reading, if not such a concurrence of authorities as this? We forbear to insist upon the probabilities of the case. The Divine power and sweetness of the incident shall not be enlarged upon. We introduce no considerations resulting from internal evidence. Let this verse of Scripture stand or fall as it meets with sufficient external testimony, or is clearly forsaken thereby.”

390 “Gospel according to St. John from eleven versions,” 1872, p. 8. Dr. Malan also translates in the same way the Peshitto “the only Son of God” and its satellite the Persic of the Polyglott as “the only one of God.” With much deference to a profound scholar, I do not see how such a rendering is possible in the Peshitto: it is precisely that which he gives in ch. iii. 18, where the Syriac inserts ܒܪܚ ܕ (or ܕ ܚܪܒ). Bp. Lightfoot judges θεός the more likely rendering of the Bohairic, though θεοῦ is possible.

391 We are not likely to adopt Tischendorf’s latest reading and punctuation in Col. ii. 2, τοῦ Θεοῦ, Χριστοῦ.

392 Hence we cannot think with Prebendary Sadler (Lost Gospel, p. 48) that μονογενὴς θεός is very probably the original reading, and must even take leave to doubt its orthodoxy. The received reading ὁ μονογενὴς υἱός is upheld by Dr. Ezra Abbot in papers contributed to the American _Bibliotheca Sacra_, Oct. 1861, and to the _Unitarian Review_, June, 1875; it is attacked with characteristic vigour and fullness of research by Dr. Hort in the first of his “Two Dissertations” (pp. 1-72) written in 1876 as exercises for Theological degrees at Cambridge.

393 The Revision Revised, p. 133. Also Miller’s “Textual Guide,” App. VI.

394 To give but a very small part of the variations in ver. 4: δέ (_pro_ γάρ) L, _a_ _b_ _c_ _ff_, Vulg. -γάρ Evst. 51, Boh. + κυρίου (_post_ γὰρ) AKLΔ, 12, 13, 69, 507, 509, 511, 512, 570 and fifteen others: at τοῦ θεοῦ 152, Evst. 53, 54.—κατὰ καιρὸν _a b ff_ ἐλούετο (_pro_ κατέβαινεν) A (K), 42, 507. Ethiop.—ἐν τῇ κολυμβήθρᾳ a b ff. ἐταράσσετο τὸ ὕδωρ C3GHIMUVΛ*, 440, 509, 510, 512, 513, 515, 543, 570, 575, Evst. 150, 257, many others. + in piscinam (_post_ ἐμβάς) _c_, Clementine Vulg. ἐγένετο FL, 69, at least fifteen others.

395 Either Dean Burgon or I have recently found the passage in Codd. 518, 524, 541, 560, 561, 573, 582, 594, 598, 599, 600, 602, 604, 622.

396 Of Lightfoot’s list of manuscripts, the passage is omitted in Codd. 2, 4, 5, 7, 8, 16, 17, 18, 19, 21, 23, 25, 26. It stands in the text of 3, 9, 14, in the margin only of 1, 20.

397 “Both elements, the clause ἐκδεχομένων τὴν τῶν ὑδάτων (sic) κίνησιν, and the scholium or explanatory note respecting the angel, are unquestionably very ancient: but no good Greek document contains both, while each of them separately is condemned by decisive evidence” (Hort, Introd., p. 301).

398 Dean Burgon has left a long vindication of the whole passage amongst his papers not yet published.

399 Add from Dr. Malan (_ubi supra_, p. 97), the Georgian, Slavonic (text, not margin), Anglo-Saxon, and Persic. His Arabic (that of Erpenius) agrees with the Peshitto Syriac. The Armenian version of Ephraem’s Tatian also reads _non_.

400 Codd. AC are defective in this place, but by measuring the space we have shown (p. 99, note 2) that A does not contain the twelve verses, and the same method applies to C. The reckoning, as McClellan remarks (N. T., p. 723), “does not preclude the possibility of small gaps having existed in A and C to mark the _place_ of the Section, as in L and Δ.”

401 Yet Burgon’s caution should be attended to. “It is to mislead—rather it is to misrepresent the facts of the case—to say (with the critics) that Codex X leaves out the ‘pericope de adulterâ.’ This Codex is nothing else but a _commentary on the Gospel, as the Gospel used to be read in public_. Of necessity, therefore, it leaves out those parts of the Gospel which are observed _not_ to have been publicly read” (_Guardian_, Sept. 10, 1873).

402 The kindred copies Codd. Λ, 215 (20 has an asterisk only against the place), 262, &c., have the following scholium at ch. vii. 53: τὰ ὠβελισμένα ἔν τισιν ἀντιγράφοις οὐ κεῖται, οὐδὲ Ἀπολ[λ]ιναρίῳ; ἐν δὲ τοῖς ἀρχαίοις ὅλα κεῖ[ν]ται; μνημονεύουσιν τῆς περικοπῆς ταύτης καὶ οἱ ἀπόστολοι, ἐν αἷς ἐξέθεντο διατάξεσιν εἰς οἰκοδομὴν τῆς ἐκκλησίας. The reference is to the Apostolic Constitutions (ii. 24. 4), as Tischendorf perceives.

403 Yet so that the first hand of Cod. 207 recognizes it in the text, setting in the margin τὸ δὲ λοιπὸν ζήτει εἰς τὸ τέλος τοῦ βιβλίου (Burgon, _Guardian_, Oct. 1, 1873).

404 A learned friend suggests that, supposing the true place for this supplemental history to be yet in doubt, there would be this reason for the narrative to be set after Luke xxi, that a reader of the Synoptic Gospels would be aware of no other occasion when the Lord had to lodge outside the city: whereas with St. John’s narrative before him, he would see that this was probably the usual lot of a _late_ comer at the Feast of Tabernacles (ch. vii. 14). Mr. J. Rendel Harris thinks that the true place for the _pericope_ is between ch. v and ch. vi, as for other reasons which we cannot depend upon, so from our illustrating the mention of the Mosaic Law in ch. viii. 5 by ch. v. 45, 46.

405 Yet on the whole this paragraph is found in more of Bp. Lightfoot’s copies than would have been anticipated: viz. in the text of 3, 8, 14, 16, 17, 18, 23, 24, in the margin of 1, and on a later leaf of 20. It is wanting in 2, 4, 5, 7, 9, 19, 21, 25, 26.

406 “Similiter Nicon ejectam esse vult narrationem ab Armenis, βλαβερὰν εἶναι τοῖς πολλοῖς τὴν τοιαύτην ἀκρόασιν dicentibus.” Tischendorf _ad loc._ Nicon lived in or about the tenth century, but Theophylact in the eleventh does not use the paragraph.

407 Notice especially the reading of 48, 64, 604, 736 (_primâ manu_) in ver. 8 ἔγραφεν εἰς τὴν γῆν ἑνὸς ἑκάστου αὐτῶν τὰς ἁμαρτίας.

408 We are not surprised in this instance at Dr. Hort’s verdict (Introd. p. 299): “No interpolation is more clearly Western, though it is not Western of the earliest type.” Dean Burgon has left amongst his papers an elaborate vindication of this passage, from which however the Editor cannot quote.

409 The form τὸν Ἰησοῦν Χριστόν, objected to by Michaelis, is vindicated by Matt. i. 18, the reading of which cannot rightly be impugned. _See_ above. Compare also ver. 12.

410 ὡς αὐτὸς ὁ εὐνοῦχος πεισθεὶς καὶ παραυτίκα ἀξιῶν βαπτισθῆναι, ἔλεγε, Πιστεύω τὸν υἱὸν τοῦ θεοῦ εἶναι Ἰησοῦν Χριστόν. Harvey, vol. ii. p. 62.

411 Such are αὐτῷ with or without ὁ Φίλιππος in E, 100, 105, 163, 186, 221, the Harkleian with an asterisk: σου added after καρδίας in E, 100, 105, 163, 186, _tol._, the Harkleian with an asterisk, the Armenian, Cyprian; but _ex toto corde_ the margin of _am._ and the Clementine Vulgate: τόν omitted before Ἰησοῦν in 186, 221 and others.

412 “Non reperi in graeco codice, quanquam arbitror omissum librariorum incuria. Nam et haec in quodam codice graeco asscripta reperi, sed in margine.” Erasmus, N. T., 1516.

413 They plead, besides the confessed preponderance of manuscript evidence for Ἑλληνιστάς, that “A familiar word standing in an obvious antithesis was not likely to be exchanged for a word so rare that it is no longer extant, except in a totally different sense, anywhere but in the Acts and two or three late Greek interpretations of the Acts; more especially when the change introduced an apparent difficulty” (Hort, Notes, p. 93). _Judicet lector._

414 Cambridge Paragraph Bible, Introduction, pp. lvi and lxxxii.

415 But with the same lack of accuracy which so often deforms this great copy: ως ετροφοφορησεν σε _κς_ ο _θς_ σου ως ει τις τροποφορησει _primâ manu_ (Vercellone).

416 Witness too Lucian’s ὑπερμεγέθη ναῦν καὶ πέρα τοῦ μέτρου, μίαν τῶν ἀπ᾽ Αἰγύπτου εἰς Ἰταλίαν σιταγωγῶν (Navig. seu Vota, c. 1) which was driven out of its course to the Piraeus. Mr. Smith, of Jordan Hill, cannot bring its dimensions under 1,300 tons.

417 Dr. Field, however, says that “this is a mistake.” The Syriac is ἔχωμεν and nothing else. For ἔχομεν this version (and all others) would put ܐܬ ܥܢ (or ܢܥ ܬܐ): but “when the word is in the subjunctive mood, since ܐܬ (or ܬܐ) is indeclinable, it is a peculiarity of the Harkleian to prefix the corresponding mood of ܗܘܐ (or ܐܘܗ), here ܢܗܘܐ (or ܐܘܗܢ)” (Otium Norvicense, iii. p. 93). For this strange phrase he cites Rom. i. 13; 2 Cor. v. 12, and to such an authority I have but _dare manus_.

418 It is simply impossible to translate with Jos. Agar Beet, in the [Wesleyan] _London Quarterly_, April, 1878, either “Let us then, justified by faith, have peace with God,” or “Let us then be justified by faith and have peace with God.” Acts xv. 36 will help him little: the other places he cites (Matt. ii. 13, &c.) not at all.

419 Dr. Vaughan (Epistle to the Romans) has ἔχωμεν in his text, and compares Heb. xii. 28, ἔχωμεν χάριν, “where there is the same variety of reading.” B is lost in this last place, but ἔχομεν, which is quite inadmissible, is found in Codd. אKP, the Latin of D, 31 and many other cursives, the printed Vulgate, and its best manuscripts. In Rom. xiv. 19 even Dr. Hort is driven by the versions and the sense to adopt in his text διώκωμεν of CD and the mass of cursives, rather than διώκομεν with אABFGLP, &c. The like confusion between ο and ω appears in the text we shall examine next but one (1 Cor. xiii. 3) and in the subjoined note (p. 384). See also φορέσομεν and φορέσωμεν, 1 Cor. xv. 49. We must confess, however, that in some of our oldest extant MSS. the interchange of ο and ω is but rare. In Cod. Sarravianus it is found in but twenty-three places out of 1224 in which itacisms occur, 830 of them being the mutation of ει and ι. On the other hand, ο stands for ω and _vice versâ_ very frequently in that papyrus fragment of the Psalms in the British Museum which Tischendorf, perhaps a little hastily, judged to be older than any existing writing on vellum.

420 Dr. Hort (Notes, p. 116) observes that διαθρύπτω is specially used in the Septuagint (Lev. ii. 6; Isa. lviii. 7) for the breaking of bread.

421 Few things are too hard for Dr. Hort, yet one is almost surprised to be told that “The text gives an excellent sense, for, as ver. 2 refers to a faith towards God which is unaccompanied by love, so ver. 3 refers to acts which seem by their very nature to be acts of love to men, but are really done in ostentation. First the dissolving of the goods in almsgiving is mentioned, then, as a climax, the yielding up of the very body; both alike being done for the sake of glorying, and unaccompanied by love” (Notes, p. 117).

422 Tyler compares _shoushou_ also in 2 Cor. vii. 5, 9; Ps. v. 11 (12).

423 Neither Winer nor his careful translator, Professor Moulton, seems disposed to yield to Lachmann’s authority in this matter. “In the better class of writers,” says Winer, “such forms are probably due to the transcribers (Lobeck on Phrynichus, p. 721), but in later authors, especially the Scholiasts (as on Thucydides iii. 11 and 54), they cannot be set aside. In the N. T., however, there is very little in favour of these conjunctives” (Moulton’s “Winer,” p. 89 and note 4, p. 361 and note 1). Yet Tregelles thinks “there would be no difficulty about the case, had not one been made by grammatical critics” (An Account of the Printed Text, p. 211, note †). But in his own example, John xvii. 2, ἵνα ... δώσῃ is read by אcACGKMSX, 33, 511, 546, and (so far as I can find) by no other manuscript whatever. On the other hand δώσει (read by Westcott and Hort; _see_ Introd., Notes, p. 172) is supported by BEHUYΓΔΛΠ (א has δωσω, D εχη, L δωσ), and (as it would seem) by every other codex extant: δώσῃ came into the common text from the second edition of Erasmus. Out of the twenty-five collated by myself for this chapter, δώσει is found in twenty-four (now including Wake 12 or Cod. 492 and Cod. 622), and the following others have been expressly cited for it: 1, 10, 11, 15, 22, 42, 45, 48, 53, 54, 55, 60, 61 (Dobbin), 63, 65, 66, 106, 118, 124, 127, 131, 142, 145, 157, 250, 262, Evst. 3, 22, 24, 36, and at least fifty others, indeed one might say all that have been collated with any degree of minuteness: so too the Complutensian and first edition of Erasmus. The constant confusion of ει and η at the period when the uncials were written abundantly accounts for the reading of the few, though AC are among them. In later times such itacisms were far more rare in careful transcription, and the mediaeval copyists knew their native language too well to fall into the habit in this passage. In Pet. iii. 1 ἵνα κερδηθήσονται is read by all the uncials (אABCKLP), nearly all cursives, and the Complutensian edition, in the place of -σωνται of Erasmus and the Received text; just as we have ἵνα γινώσκομεν in אAB*LP, 98, 99, 101, 180, 184, 188, 190 in 1 John v. 20. The case for ἀρκεσθησόμεθα 1 Tim. vi. 8 is but a shade less feeble.

424 Tischendorf, however, boldly interposes a comma between the words (_see_ p. 359, note), and is followed by Westcott and Hort and by Bp. Lightfoot, whose note on the passage (Coloss. p. 318) is very elaborate. This mode of punctuation would set χριστοῦ in apposition to μυστηρίου, in support of which construction ch. i. 27 (ὅ); 1 Tim. iii. 16 (ὅς) are alleged. This, however, is not the sense favoured by Hilary (_in agnitionem sacramenti dei Christi_, and again _Deus Christus sacramentum est_), and would almost call for the article before χριστοῦ. In meaning it would be equivalent to D*, &c., ὅ ἐστιν _χσ_.

425 In Dr. Swete’s edition, vol. ii. p. 11, Theodore expounds thus in the old Latin version: _sed facti sumus quieti in medio vestro_, hoc est, “omni mediocritate et humilitate sumus abusi, nolentes graves aliquibus videri.”

426 A like combination is seen in Cod. 37 in 1 Tim. vi. 19 τῆς αἰωνίου ὄντως ζωῆς.

427 Dean Burgon has just presented me with the photographed page in Cod. G, respecting whose evidence there can be no remaining doubt.

428 The true reading of the Codex Alexandrinus in 1 Tim. iii. 16 has long been an interesting puzzle with Biblical students. The manuscript, and especially the leaf containing this verse (fol. 145), now very thin and falling into holes, must have been in a widely different condition from the present when it first came to England. At that period Young, Huish, and the rest who collated or referred to it, believed that _ΘΣ_ was written by the first hand. Mill (N. T. _ad loc._) declares that he had first supposed the primitive reading to be _ΟΣ_, seeing clearly that the line _over_ the letters had not been entirely made, but only thickened, by a later hand, probably the same that traced the coarse, rude, recent, horizontal diameter now running through the circle. On looking more closely, however, he detected “ductus quosdam et vestigia satis certa ... praesertim ad partem sinistram, qua peripheriam literae pertingit,” evidently belonging to an earlier diameter, which the thicker and later one had almost defaced. This old line was afterwards seen by John Berriman and four other persons with him (Gloucester Ridley, Gibson, Hewett, and Pilkington) by means of a glass in the bright sunshine, when he was preparing his Lady Moyer’s Lecture for 1737-8 (Critical Dissertation on 1 Tim. iii. 16, p. 156). Wetstein admitted the existence of such a transverse line, but referred it to the tongue or _sagitta_ of Ε on the reverse of the leaf, an explanation rejected by Woide, but admitted by Tregelles, who states in opposition to Woide that “Part of the Ε on the other side of the leaf _does_ intersect the Ο, as we have seen again and again, and which others with us have seen also” (Horne, iv. p. 156). This last assertion may be received as quite true, and yet not relevant to the point at issue. In an Excursus appended to 1 Timothy in his edition of “The Pastoral Epistles” (p. 100, 1856), Bp. Ellicott declares, as the result of “minute personal inspection,” that the original reading was “indisputably” ΟΣ. But the fact is, that the page is much too far gone to admit of any present judgement which would weigh against past judgements, as any one who examines the passage can see for himself. Woide could see the line in 1765, but not in 1785.

429 Yet how can it be _precarious_ in the face of such testimony as the following (_Quarterly Review_, Oct. 1881, p. 363)? Τὸ δὲ θεὸν ὄντα ἄνθρωπον θελῆσαι γενέσθαι καὶ ἀνασχέσθαι καταβῆναι τοσοῦτον ... τοῦτό ἐστι τὸ ἐκπλήξεως γέμον. Ὅ δὴ καὶ Παῦλος θαυμάζων ἔλεγε; καὶ ὁμολογουμένως μέγα ἐστὶ τὸ τῆς εὐσεβείας μυστήριον; ποῖον μέγα? θεὸς ἐφανερώθη ἐν σαρκί; καὶ πάλιν ἀλλαχοῦ; οὐ γὰρ ἀγγέλων ἐπιλαμβάνεται ὁ θεός (Chrysostom, i. 497). It is necessary to study the context well before we can understand the strength or weakness of Patristic evidence.

430 Twenty-three times in all, as Ward (_see_ p. 394, note) observes, adding that “nothing can be more express and unquestionable than his reading.” The _Quarterly Reviewer_ speaks very well (_ubi supra_), “A single quotation is better than many references. Among a multitude of proofs that Christ is God, Gregory says: Τιμοθέῳ δὲ διαρρήδην βοᾷ ὅτι ὁ θεὸς ἐφανερώθη ἐν σαρκί, ἐδικαιώθη ἐν πνεύματι” (ii. 693).

431 Bentleii Critica Sacra, p. 67, ’Σχόλια Photii MSS. (Bib. Pub. Cant.) _ad loc_. ὁ ἐν ἁγίοις Κύριλλος ἐν τῷ _ιβ_ κεφαλαίῳ τῶν σχολίων φησίν, ὃς ἐφανερώθη ἐν σαρκί.’ Photius also quoted Gregory Thaumaturgus (or Apollinarius) for θεός.

432 Dr. Swete, in his masterly edition of the Latin translation of Theodore’s commentary on St. Paul’s Epistles, after citing the Latin text as _qui manifestatus est in carne_, adds “Both our MSS. read _qui_, here and [15 lines] below and use the masculine consistently throughout the context.... Thus the present translation goes to confirm the inference already drawn from the Greek fragment of Theodore, de Incarn. xiii (Migne, P. G. 66, 987), that he read ὃς ἐφανερώθη” (vol. ii. p. 135 n.): pertinently observing that if Theodore used ὅς, he was in harmony with the Syriac versions.

433 “Conspectum lectionis hujus loci optime dedit in sermone vernaculo William H. Ward, V. D. M. in Bibliotheca Sacrâ Americanâ, anni 1865,” Tregelles N. T. _ad loc_. For a copy of this work I am indebted to the kindness of A. W. Tyler of New York. Mr. Ward wonders that neither Tregelles nor I have noticed a certain pinhole in Cod. A, which was pointed out to Sir F. Madden by J. Scott Porter, made by some person at the extremity of the sagitta of the Ε on the opposite page, and falling exactly on the supposed transverse line of the Θ. I cannot perceive the pinhole, but the vellum is fast crumbling away from the effects of time, certainly through no lack of care on the part of those who keep the manuscript.

434 “As the Apostle here applies to _Christ_ language which in the Old Testament is made use of with reference to Jehovah (_see_ Isa. viii. 13), he clearly suggests the supreme godhead of our Redeemer,” as Dr. Roberts puts the matter (Words of the New Testament, p. 170). Not, of course, that our critical judgement should be swayed one way or the other by individual prepossessions; but that those who in the course of these researches have sacrificed to truth much that they have hitherto held dear, need not suppress their satisfaction when truth is gain.

435 This translation of 2 Peter, 2, 3 John, and Jude, printed by Pococke from Bodl. Orient. 119, well deserves careful study, being totally different in style and character both from the Peshitto and the Harkleian, somewhat free and periphrastic, yet, in our paucity of good authorities just here, of great interest and full of valuable readings. Thus, in this very verse it reads ἀδικούμενοι (“being wronged as the hire of their wrong-doing”) with א*BP and the Armenian, difficult as it may seem to receive that word as genuine: in ver. 17 it omits εἰς τὸν αἰῶνα with אB and some other versions: in ch. iii. 10 it sides with the Sahidic alone in receiving οὐχ εὑρεθήσεται (apparently correctly) instead of εὑρεθήσεται of אBKP, of the excellent cursives 27, 29, 66 _secundâ manu_, of the Armenian and Harkleian margin, where the Received text follows the obvious κατακαήσεται of AL and the rest, and C hits upon ἀφανισθήσονται in pure despair.

436 Bp. Chr. Wordsworth speaks as though there were a _paronomasia_, a play on the words ἀγάπη and ἀπάτη, comparing (after Windischmann) 2 Thess. ii. 10. “The false teachers called their meetings ἀγάπαι, _love feasts_, but they were mere ἀπάται, _deceits_. Their _table_ was a _snare_” (Ps. lxix. 22). This view might be tenable if St. Peter, with whom the _paronomasia_ must have taken its rise, were not the earlier writer of the two, as the Bishop of Lincoln believes he was, as firmly as we do. Perhaps Dr. Westcott’s notion that 2 Peter is a translation, not an original, at least in ch. ii, will best account for the textual variations between it and St. Jude.

437 See the Cambridge Paragraph Bible, Introduction, pp. xxxv, xxxvii.

438 “Restitui in Grecis hoc membrum ex quatuor manuscr. codicum, veteris Latini et Syri interpretis auctoritate. sic etiam assueto Johanne istis oppositionibus contrariorum uti quam saepissimè.” Beza, N. T., 1582.

439 Horne (Introduction, vol. ii. pt. ii. ch. iii. sect. 4), and after his example Tregelles (Horne, iv. pp. 384-8), give a curious list of more than fifty volumes, pamphlets, or critical notices on this question. The following are the most worthy of perusal: Letters to Edward Gibbon, Esq., by G. Travis, Archdeacon of Chester, 1785, 2nd edit.; Letters to Mr. Archdeacon Travis, &c., by Richard Porson, 1790; Letters to Mr. Archdeacon Travis, &c., by Herbert Marsh [afterwards Bp. of Peterborough], 1795; A Vindication of the Literary Character of Professor Porson, by Crito Cantabrigiensis [Thomas Turton, afterwards Bp. of Ely], 1827; Two Letters on some parts of the Controversy concerning 1 John v. 7, by Nicolas Wiseman, 1835, for which _see_ Index. For Dr. Adam Clarke’s “Observations,” &c., 1805, _see_ Evan. 61. Add F. A. Knittel on 1 John v. 7. Professor Ezra Abbot’s edition of “Orme’s Memoir of the Controversy on 1 John v. 7,” New York, 1866, has not fallen in my way. As elaborate works, on the verses are “A new plea for the authenticity of the Text of the Three Heavenly Witnesses, or Porson’s Letters to Travis eclectically examined,” Cambridge, 1867, being the performance of a literary veteran, the late Rev. Charles Forster, whose arguments in vindication of the Pauline origin of the Epistle to the Hebrews, published in 1838, modern Biblical writers have found it easier to pass by than to refute; and “The Three Witnesses, the disputed text in St. John, considerations new and old,” by the Rev. H. T. Armfield, Bagster, 1883.

440 That the Codex Montfortianus was influenced by the Vulgate is probably true, though it is a little hasty to infer the fact at once from a single instance, namely, the substitution of χριστός after that version and Uscan’s Armenian for the second πνεῦμα in verse 6: “quae lectio Latina Graece in codicem 34 Dublinensem illum Montfortianum recepta luculenter testatur versionem vulgatam ad cum conficiendum valuisse” (Tischendorf _ad loc._).

441 It is really surprising how loosely persons who cannot help being scholars, at least in some degree, will talk about codices containing this clause. Dr. Edward Tatham, Rector of Lincoln College, Oxford (1792-1834), writing in 1827, speaks of a manuscript in his College Library which exhibited it, but is now missing, as having been once seen by him and Dr. Parsons, Bishop of Peterborough (Crito Cantabrigiensis, p. 334, note). Yet there can be no question that he meant Act. 33, which does not give the verse, but has long been known to have some connexion with the Codex Montfortianus, which does (_see_ Act. 33).

442 Of the two Spanish MSS. one _leon._2 contains the passage only in the margin, the other _leon._1 adds at the end of ver. 8, _in __xpo__ __ihu_. Canon Westcott cites a manuscript in the British Museum (Add. 11,852), of the ninth century, to the same effect, observing that, like _m_ and _cav._, it contains the Epistle to the Laodiceans. This MS. runs “quia tres sunt qui testimonium dant _sps_ et aqua et sanguis, et tres unum sunt. Sicut in caelo tres sunt pater verbum et _sps_ et tres unum sunt.” Westcott’s manuscript is, in fact, _ulm._, and had already been used by Porson (Letters, &c., p. 148).

443 Mr. Forster (_ubi supra_, pp. 200-209) believed that he had discovered _Greek_ authority of the fourth century for this passage, in an isolated Homily by an unknown author, in the Benedictine edition of Chrysostom (Tom. xii. pp. 416-21), whose date Montfaucon easily fixes by internal evidence at A.D. 381. As this discovery, if real, is of the utmost importance in the controversy, it seems only right to subjoin the words alleged by this learned divine, leaving them to make their own way with the reader: (1) εἷς κέκληται ὁ Πατὴρ καὶ ὁ Υἱὸς καὶ τὸ Πνεῦμα τὸ Ἅγιον: (2) δεῖ γὰρ τῇ ἀποστολικῇ χορείᾳ παραχωρῆσαι τὴν Ἁγίαν Τριάδα, ἢν ὁ Πατὴρ καταγγέλλει. Τριὰς Ἀποστόλων, μάρτυς τῆς οὐρανίου Τριάδος.

444 The “Prologus Galeatus in vii Epistolas _Canonicas_,” in which the author complains of the omission of ver. 7, “ab infidelibus translatoribus,” is certainly not Jerome’s, and begins to appear in codices of about the ninth century.

445 The writer of a manuscript note in the British Museum copy of Travis’ “Letters to Gibbon,” 1785, p. 49, very well observes on the second citation from Cyprian: “That three are one might be taken from the eighth verse, as that was certainly understood of Father, Son, and Holy Ghost, _especially when Baptism was the subject in hand_” [Matt. xxviii. 19].

446 It will be seen upon examination of our collations on p. 402 that the points of difference between Codex Montfortianus (34) and Erasmus’ printed text are two, viz. that 34 omits καί after πνεῦμα in ver. 8, and with the Complutensian leaves out its last clause altogether; while, on the other hand, Erasmus and Cod. 34 agree against the Complutensian in their barbarous neglect of the Greek article in both verses. As regards the omission in Cod. 34 of the last clause of ver. 8 (καὶ οἱ τρεῖς εἰς τὸ ἕν εἰσιν), it is obvious to conjecture that the person, whosoever he was, that sent the transcript of the passage to Erasmus, who never saw the MS. for himself, might have broken off after copying the disputed words, and neglected to note down the further variation that immediately followed them. After the foregoing explanation we must leave the matter as it stands, for there is no known mode of accounting for the discrepancy, whereof Mr. Forster makes the very utmost in the following note, which, as a specimen of his book, is annexed entire: “Bishop Marsh labours hard to identify the Codex Britannicus used by Erasmus, with the Codex Montfortianus. Erasmus’s own description of the Codex Britannicus completely nullifies the attempt: ‘Postremo: Quod Britannicum etiam in terrae testimonio addebat, καὶ οἱ τρεῖς εἰς τὸ ἕν εἰσι, quod non addebatur hic duntaxat in editione Hispaniensi.’ Now as this clause is also omitted in the Montfort Codex, it cannot possibly be the same with the Codex Britannicus. In this as yet undiscovered MS., therefore, we have a second and independent Gr. MS. witness to the seventh verse. The zeal of the adversaries to evade this fact only betrays their sense of its importance” (p. 126). Alas! _Hi motus animorum._

447 I side with Porson against Travis on every important point at issue between them, and yet I must say that if the former lost a legacy (as has been reported) by publishing his “Letters,” he was entitled to but slender sympathy. The prejudices of good men (especially when a passage is concerned which they have long held to be a genuine portion of Scripture, clearly teaching pure and right doctrine) should be dealt with gently: not that the truth should be dissembled or withheld, but when told it ought to be in a spirit of tenderness and love. Now take one example out of fifty of the tone and temper of Porson. The immediate question was a very subordinate one in the controversy, namely, the evidence borne by the Acts of the Lateran Council, A.D. 1215. “Though this,” rejoins Porson, “proves nothing in favour of the verse, it proves two other points. That the clergy then exercised dominion over the rights of mankind, and that able tithe-lawyers often make sorry critics. _Which I desire some certain gentlemen of my acquaintance to lay up in their hearts as a very seasonable innuendo_” (Letters, p. 361, quoted from “A Tale of a Tub” p. 151). As if it were a disgrace for an Archdeacon to know a little about the laws which affect the clergy.

448 Gaussen (Theopneustia, pp. 115-7) has still spirit remaining to press the masculine forms οἱ μαρτυροῦντες ver. 7 and οἱ τρεῖς ver. 8 as making in favour of the intervening clause: “Remove it, and the grammar becomes incoherent:” a reason truly, but one not strong enough to carry his point.

449 We are compelled to draw a sharp distinction between γεγεννημένος and γεννηθείς in the same context, and, with all deference to the _Quarterly Reviewer_ (April, 1882, p. 366), we do not think his view of the matter more natural than that given in the text: “St. John,” he suggests, “is distinguishing between the mere recipient of the new birth (ὁ γεννηθεὶς ἐκ τοῦ Θεοῦ),—and the man who retains the sanctifying influences of the Holy Spirit which he received when he became regenerate (ὁ γεγεννημένος ἐκ τοῦ Θεοῦ).” [The distinction given between the perfect and aorist, as I have altered it in the text, is perfectly just, and explains the passage. The effects of regeneration if continued are indefectible, but the mere fact of regeneration entails constant watchfulness.]

450 So it certainly seems to me after careful inspection of Cod. A, although it may be too bold to say, as some have, that there are in it no corrections by later hands. Above in ver. 10 ἐν ἀυτῷ is supported by ABKLP and a shower of cursives in the room of ἐν ἑαυτῷ of א and the Received text, but here there is no difference of sense between the two forms. Dr. Hort (Introd., Notes, p. 144) has an exhaustive and cautious note on the breathing of αυτου, αυτῳ, &c., and ultimately declines to exclude the aspirate from the N. T.

451 The Revision Revised, pp. 247-8.

452 For a very full and clear account of a MS. of this class, the reader may consult an article by Prof. Isaac H. Hall in the “Journal of the American Oriental Society,” vol. xi, No. 2, 1885.

453 It is not meant that these terms occur as titles. _Apostolos_ (ܫܥܝܐ or ܐܝܥܫ) as applied to a book means the fourteen Epp. of St. Paul. _Evangeliom_, in the sense of _Evangelistary_ in a title, is quoted in “Thesaurus Syriacus.”

But many liturgical terms were borrowed from the Greeks, especially by the Maronites. For a succinct account of Greek and Latin Service Books, _see_ Pelliccia’s “Polity” (tr. Bellett, 1883), pp. 183-8: for the Syriac system, _see_ Etheridge’s “Syrian Churches,” pp. 112-6.