A Plain Introduction to the Criticism of the New Testament, Vol. II.

by Frederick Henry Ambrose Scrivener

First Series. Gospels.

1. Matt. i. 18. Τοῦ δὲ Ἰησοῦ Χριστοῦ ... is altered by Tregelles into Τοῦ δὲ Χριστοῦ, Ἰησοῦ being omitted: Westcott and Hort place Ἰησοῦ between brackets, and Τοῦ δὲ Χριστοῦ Ἰησοῦ of Cod. B in the margin: Tischendorf, who had rejected Ἰησοῦ in his fifth and seventh editions, restored it in his eighth. Michaelis had objected to the term τὸν Ἰησοῦν Χριστόν, Acts viii. 37 (see that verse, to be examined below), on the ground that “In the time of the Apostles the word Christ was never used as the Proper Name of a Person, but as an epithet expressive of the ministry of Jesus;” and although Bp. Middleton has abundantly proved his statement incorrect (Doctrine of the Greek Article, note on Mark ix. 41), and Ἰησοῦς Χριστός[335], especially in some one of the oblique cases after prepositions, is very common, yet the [pg 322] precise form ὁ Ἰησοῦς Χριστός occurs only in these places and in 1 John iv. 3; Apoc. xii. 17, where again the reading is more than doubtful. Hence, apparently, the determination to change the common text in St. Matthew, on evidence however slight. Now Ἰησοῦ is omitted in no Greek manuscript whatsoever[336]. The Latin version of Cod. D (d) indeed rejects it, the parallel Greek being lost; but since d sometimes agrees with other Latin copies against its own Greek, it cannot be deemed quite certain that the Greek rejected it also[337]. Cod. B reads τοῦ δὲ Χριστοῦ Ἰησοῦ, in support of which Lachmann cites Origen, iii. 965 d in the Latin, but on very precarious grounds, as Tregelles (An Account of the Printed Text, p. 189, note †) candidly admits. Tischendorf quotes Cod. 74 (after Wetstein), the Persic (of the Polyglott and in manuscript), and Maximus, Dial. de Trinitate, for τοῦ δὲ ἰησοῦ. The real testimony in favour of τοῦ δὲ Χριστοῦ consists of the Old Latin copies a b c d f ff1, the Curetonian Syriac (I know not why Cureton should add “the Peshitto”), the Latin Vulgate, the Frankish and Anglo-Saxon, Wheelocke's Persic, and Irenaeus in three places, “who (after having previously cited the words ‘Christi autem generatio sic erat’) continues ‘Ceterum potuerat dicere Matthaeus, Jesu vero generatio sic erat; sed praevidens Spiritus Sanctus depravatores, et praemuniens contra fraudulentiam eorum, per Matthaeum ait: Christi autem generatio sic erat’ (Contra Haeres., lib. iii. 16. 2). This is given in proof that Jesus and Christ are one and the same Person, and that Jesus cannot be said to be the receptacle that afterwards received Christ; for the Christ was born” (An Account of the Printed Text, p. 188). To this most meagre list of authorities Scholz adds, “Pseudo-Theophil. in Evang.,” manuscripts of Theophylact, Augustine, and one or two of little account: but even in Irenaeus (Harvey, vol. ii. p. 48) τοῦ δὲ ιυ χυ (tacitè), as preserved by Germanus, Patriarch of Constantinople [viii], stands over against the Latin “Christi.”

We do not deny the importance of Irenaeus' express testimony[338] (a little impaired though it be by the fanciful distinction [pg 323] which he had taken up with), had it been supported by something more trustworthy than the Old Latin versions and their constant associate, the Curetonian Syriac. On the other hand, all uncial and cursive codices (אCΣEKLMPSUVZΓΔΠ: ADFGΦ &c. being defective here), the Syriac of the Peshitto, Harkleian, and Jerusalem (δέ only being omitted, since the Church Lesson begins here), the Sahidic, Bohairic, Armenian, and Ethiopic versions, Tatian, Irenaeus, Origen (in the Greek), Eusebius, Didymus, Epiphanius, Chrysostom, and the younger Cyril, comprise a body of proof, not to be shaken by subjective notions, or even by Western evidence from the second century downwards[339].

2. Matt. vi. 13. ὅτι σοῦ ἐστιν ἡ βασιλεία καὶ ἡ δύναμις καὶ ἡ δόξα εἰς τοὺς αἰῶνας. ἀμήν. It is right to say that I can no longer regard this doxology as certainly an integral part of St. Matthew's Gospel: but (notwithstanding its rejection by Lachmann, Tischendorf, Tregelles, Westcott and Hort) I am not yet absolutely convinced of its spuriousness [i.e. upon much less evidence than is now adduced]. It is wanting in the oldest uncials extant, אBDZ, and since ACP (whose general character would lead us to look for support to the Received text in such a case) are unfortunately deficient here, the burden of the defence is thrown on Φ and Σ and the later uncials EGKLMSUVWfΔΠ (hiat Γ), whereof L is conspicuous for usually siding with B. Of the cursives only five are known to omit the clause, l, 17 (habet ἀμήν), 118, 130, 209, but 566 or hscr (and as it would seem some others) has it obelized in the margin, while the scholia in certain other copies indicate that it is doubtful: even 33 contains it, 69 being defective, while 157, 225, 418 add to δόξα, τοῦ πατρὸς καὶ τοῦ υἱοῦ καὶ τοῦ ἁγίου πνεύματος, but 422 τοῦ πρσ only. Versions have much influence on such a question, it is therefore important to notice that it is found in all the four Syriac (Cureton's omitting καὶ ἡ δύναμις, and some editions of the Peshitto ἀμήν, which [pg 324] is in at least one manuscript), the Sahidic (omitting καὶ ἡ δόξα), the Ethiopic, Armenian, Gothic, Slavonic, Georgian, Erpenius' Arabic, the Persic of the Polyglott from Pococke's manuscript, the margin of some Bohairic codices, the Old Latin k (quoniam est tibi virtus in saecula saeculorum), f g1 (omitting amen) q. The doxology is not found in most Bohairic (but is in the margin of Hunt. 17 or Bp. Lightfoot's Cod. 1) and Arabic manuscripts or editions, in Wheelocke's Persic, in the Old Latin a b c ff1 g1 h l, in the Vulgate or its satellites the Anglo-Saxon and Frankish (the Clementine Vulg. and Sax. add amen). Its absence from the Latin avowedly caused the editors of the Complutensian N. T. to pass it over, though it was found in their Greek copies: the earliest Latin Fathers naturally did not cite what the Latin codices for the most part do not contain. Among the Greeks it is met with in Isidore of Pelusium (412), and in the Pseudo-Apostolic Constitutions, probably of the fourth century: soon afterwards Chrysostom (Hom. in Matt. xix. vol. i. p. 283, Field) comments upon it without showing the least consciousness that its authenticity was disputed. The silence of some writers, viz. Tertullian, Cyprian, Origen, Augustine, Cyril of Jerusalem, and Maximus, especially when expounding the Lord's Prayer, may be partly accounted for by the fact of the existence of the shorter form of the Lord's Prayer as given in St. Luke without the doxology; or upon the supposition that the doxology was regarded not so much a portion of the Prayer itself, as a hymn of praise annexed to it; yet this latter fact would be somewhat unfavourable to its genuineness, and would be fatal unless we knew the precariousness of any argument derived from such silence. The Fathers are constantly overlooking the most obvious citations from Scripture, even where we should expect them most, although, as we learn from other passages in their writings, they were perfectly familiar with them. Internal evidence is not unevenly balanced. It is probable that the doxology was interpolated from the Liturgies, and the variation of reading renders this all the more likely; it is just as probable that it was cast out of St. Matthew's Gospel to bring it into harmony with St. Luke's (xi. 4): I cannot concede to Scholz that it is “in interruption of the context,” for then the whole of ver. 13 would have to be cancelled (a remedy which no one proposes), and not merely this concluding part of it.

It is vain to dissemble the pressure of the adverse case, though it ought not to be looked upon as conclusive. The Διδαχή (with variation) and the Syriac and Sahidic versions bring up the existence of the doxology to the second century; the Apostolic Constitutions in the third; Ambrose, Caesarius, Chrysostom, the Opus Imperfectum, Isidore, and perhaps others[340], attest for it in the fourth; then come the Latin codices[341] f g1 k q, the Gothic, the Armenian, the Ethiopic, and lastly Codd. Φ and Σ of the fifth or sixth century, and the whole flood-tide of Greek manuscripts from the eighth century downwards, including even L, 33, with Theophylact and Euthymius Zigabenus in the eleventh and twelfth. Perhaps it is not very wise “quaerere quae habere non possumus,” yet those who are persuaded, from the well-ascertained affinities subsisting between them, that ACP, or at least two out of the three, would have preserved a reading sanctioned by the Peshitto, by Codd. f k, by Chrysostom, and by nearly all the later documents, may be excused for regarding the indictment against the last clause of the Lord's Prayer as hitherto unproven, in Dr. Scrivener's judgement passed upon much less than the evidence in favour adduced above; and for supposing the genuineness of the clause to be proved when the additional evidence is taken into consideration.

3. Matt. xi. 19. The change of τέκνων of the Received text into ἔργων, as made by Tischendorf, Tregelles (who retains τέκνων in his margin), by Hort and Westcott, is quite destructive to the sense, so far as we can perceive, for Jerome's exposition (“Sapientia quippe non quaerit vocis testimonium, sed operum”) could [pg 326] hardly satisfy any one but himself. The reading ἔργων is supported by אB* (with τέκνων in the margin by the hand B2), 124, the Peshitto Syriac (apparently; for all the older editions we know punctuate ܠܒܕܘ (or ܘܕܒܠ) “doers,” not ܠܒܕܘ (or ܘܕܒܠ) “works”), the Harkleian text (but not its margin), the Bohairic, some copies known to Jerome, Armenian manuscripts, the Ethiopic (one MS. contains both forms), and (after the Peshitto Syriac) the Persic of the Polyglott and its codices. We can hardly question that the origin of the variation arose from the difficulty on the part of translators and copyists to understand the Hellenistic use of τέκνων in this place, and modern editors have been tempted to accept it from a false suspicion that the present passage has been assimilated to Luke vii. 35, where indeed Cod. א and St. Ambrose have ἀπὸ πάντων τῶν ἔργων ἀυτῆς. As we have alleged that Jerome's explanation is unsatisfactory in St. Matthew's Gospel, we subjoin that of Ambrose, which is certainly no less obscure, on the parallel place of St. Luke: “Bene ab omnibus quia circa omnes justitia servatur, ut susceptio fiat fidelium, rejectio perfidorum. Unde plerique Graeci sic habent: justificata est sapientia ab omnibus operibus suis, quod opus justitiae sit, circa uniuscujusque meritum servare mensuram.” In the face of the language of these two great Latin Fathers it is remarkable that all other Latin authorities agree with the Curetonian Syriac and the mass of Greek manuscripts in upholding τέκνων, which is undoubtedly the only true reading.

4. Matt. xvi. 2, 3. The whole passage from Ὀψίας ver. 2 to the end of ver. 3 is set within brackets by Tischendorf in his eighth edition, within double brackets by Westcott and Hort, who holds (Notes, p. 13) that “both documentary evidence and the impossibility of accounting for omission prove these words to be no part of the text of Mt.” Yet it might seem impossible for any one possessed of the slightest tincture of critical instinct to read them thoughtfully without feeling assured that they were actually spoken by the Lord on the occasion related in the Received text, and were omitted by copyists whose climate the natural phenomena described did not very well suit, the rather as they do not occur in the parallel text, ch. xii. 38, 39. Under these circumstances, the internal evidence in favour of the passage being thus clear and irresistible, the witnesses against it are [pg 327] more likely to damage their own authority than to impair our confidence in its genuineness. These witnesses are אBVXΓ, 2, 13, 34, 39, 44, 84, 124 primâ manu, 157, 180, 194, 258, 301, 511, 575. Cod. 482 has the words, but only in a later hand at the foot of the page (Nicholson). Of these cursive codices 157 alone is of the first class for importance, and the verses are explained in the scholia of X (for ver. 3) and of 39. E and 606 have them with an asterisk; but they are wanting in the Curetonian Syriac, the Bohairic according to Mill (but not so other Coptic manuscripts and editions), and the Armenian, as unaltered from the Latin. Origen passes them over in his commentary, and Jerome, in his sweeping way, declares “hoc in plerisque codicibus non habetur.” They are recognized in the Eusebian canons (Tregelles, An Account of the Printed Text, p. 205).

The united testimony of אB and the Curetonian version suffices to show that the omission was current as early as the second century, while the accordance of CD, of all the Latins and the Peshitto, with the mass of later codices assures us that the words were extant at the same early date. If any one shall deem this a case best explained by the existence of two separate recensions of the same work, one containing the disputed sentences, the other derived from copies in which they had not yet been inserted, he may find much encouragement for his conjecture by considering certain passages in the latter part of St. Luke's Gospel, where the same sort of omissions, supported by a class of authorities quite different from those we have to deal with here, occur too often to be merely accidental.

5. Matt. xix. 17. For Τί με λέγεις ἀγαθόν? οὐδεὶς ἀγαθός, εἰ μὴ εἷς, ὁ Θεός, Griesbach, Lachmann, Tischendorf, Tregelles, Alford, Westcott and Hort read Τί με ἐρωτᾷς περὶ τοῦ ἀγαθοῦ? εἷς ἐστὶν ὁ ἀγαθός. The self-same words as in the Received text occur in the parallel places Mark x. 18, Luke xviii. 19 with no variation worth speaking of; a fact which (so far as it goes) certainly lends some support to the supposition that St. Matthew's autograph contained the other reading [?]. Add to this that any change made from St. Matthew, supposing the common reading to be true, must have been wilfully introduced by one who was offended at the doctrine of the Divine Son's inferiority to the [pg 328] Father which it seemed to assert or imply. Internal evidence, therefore, would be a little in favour of the alteration approved by Lachmann, Tischendorf, and the rest; and in discussing external authority, their opponents are much hampered by the accident that A is defective in this place, while א has recently been added to the list of its supporters [though more recently Φ and Σ have come into the opposite balance]. Under these circumstances we might have been excused from noticing this passage at all, as we are no longer able to uphold the Received text with the same confidence as before, but that it seemed dishonest to suppress a case on which Tregelles (An Account of the Printed Text, pp. 133-8) has laid great stress, and which, when the drift of the internal evidence is duly allowed for, tells more in his favour than any other he has alleged, or is likely to be met with elsewhere[342].

The alternative reading Τί με ἐρωτᾷς περὶ τοῦ ἀγαθοῦ κ.τ.λ. occurs in אBD (omitting τοῦ and ὁ) L, 1 (omitting ὁ), 22, 604. In 251 both readings are given, the Received one first, in ver. 17, the other interpolated after ποίας ver. 18, prefaced by ὁ δὲ ἰησοῦς εἶπεν αὐτῷ. Excepting these seven, all other extant codices reject it, CEFGHKMSUVΓΔ (Γ omits τί με λέγεις ἀγαθόν; Δ omits λέγεις, Π is defective here), even Codd. 33, 69. The versions are more seriously divided. The Peshitto Syriac, the Harkleian text, the Sahidic (Oxford fragments), the Old Latin f q, the Arabic, &c., make for the common reading; Cureton's and the Jerusalem Syriac, the Old Latin a b c e ff1.2 l, the Vulgate (the Anglo-Saxon and Frankish, of course), Bohairic and Armenian, for that of Lachmann and his followers. Several present a mixed form: τί με ἐρωτᾷς περὶ τοῦ ἀγαθοῦ? οὐδεὶς ἀγαθὸς εἰ μὴ εἷς: viz. the margin of the Harkleian, the Ethiopic, and g1 h m of the Old Latin. A few (Cureton's Syriac, b c ff1.2 g1 h l m, Jerome and the Vulgate) add ὁ θεός, as in the common text; but this is unimportant.

Tregelles presses us hard with the testimony of Origen in favour of the reading he adopts: ὁ μὲν οὖν Ματθαῖος, ὡς περὶ ἀγαθοῦ ἔργου ἐρωτηθέντος τοῦ σωτῆρος ἐν τῷ, Τί ἀγαθὸν ποιήσω? ἀνέγραψεν. Ὁ δὲ Μάρκος καὶ Λουκᾶς φασὶ τὸν σωτῆρα εἰρηκέναι, Τί με λέγεις ἀγαθόν? οὐδεὶς ἀγαθός, εἰ μὴ εἷς, ὁ Θεός (Tom. iii. p. 644 d). “The reading which is opposed to the common text,” Tregelles [pg 329] writes, “has the express testimony of Origen in its favour” (p. 134); “might I not well ask for some proof that the other reading existed, in the time of Origen, in copies of St. Matthew's Gospel?” (p. 137). I may say in answer, that the testimony of Origen applies indeed to the former part of the variation which Tregelles maintains (τί με ἐρωτᾷς περὶ τοῦ ἀγαθοῦ), but not at all to the latter (εἷς ἐστιν ὁ αγαθός), and that the Peshitto Syriac version of the second, as also the Sahidic of the third century, uphold the common text, without any variation in the manuscripts of the former, that we know of. Or if he asks for the evidence of Fathers to counterbalance that of a Father, we have Justin Martyr: προσελθόντος αὐτῷ τινὸς καὶ εἰπόντος (words which show, as Tischendorf observes, that St. Matthew's is the only Gospel that can be referred to) Διδάσκαλε ἀγαθέ, ἀπεκρίνατο λέγων, Οὐδεὶς ἀγαθὸς εἰ μὴ μόνος ὁ Θεὸς ὁ ποιήσας τὰ πάντα, citing loosely, as is usual with him, but not ambiguously. Or if half the variation will satisfy, as it was made to do for Origen, Tregelles' own note refers us to Irenaeus 92 for τί με λέγεις ἀγαθόν? εἷς ἐστὶν ἀγαθός, and to Eusebius for the other half in the form above quoted from the Ethiopic, &c. Moreover, since he cites the last five words of the subjoined extract as belonging to St. Matthew, Tregelles entitles us to employ for our purpose the whole passage, Marcos. apud Iren. 92, which we might not otherwise have ventured to do; καὶ τῷ εἰπόντι αὐτῷ Διδάσκαλε ἀγαθέ, τὸν ἀληθῶς ἀγαθὸν θεὸν ὡμολογηκέναι, εἰπόντα Τί με λέγεις ἀγαθόν? εἷς ἐστιν ἀγαθός, ὁ πατὴρ ἐν τοῖς οὐρανοῖς. Jerome and Augustine (for the first clause only, though very expressly: de Consensu Evan. ii. 63) are with the Latin Vulgate, Hilary with the common Greek text, as are also Optatus, Ambrose, Chrysostom, and the main body of later Fathers. Thus the great mass of manuscripts, headed by C [followed by Φ and Σ], is well supported by versions, and even better by ecclesiastical writers; yet, in virtue of the weight of internal evidence [?], we dare not hold out unreservedly against the reading of BDL, &c., now that Cod. א is found to agree with them, even though subsequent investigations have brought to light so close a relation between א and B as to render it impossible, in our opinion, to regard them as independent witnesses[343].

6. Matt. xx. 28. The extensive interpolation which follows this verse in some very ancient documents has been given above (I. 8), in the form represented in the Curetonian Syriac version. It bears the internal marks of evident spuriousness, the first sentence consisting of a rhetorical antithesis as unsuitable as can be imagined to the majestic simplicity of our Lord's usual tone, while the sentiment of the rest is manifestly borrowed from Luke xiv. 8-10, although there is little or no resemblance in the words. The only extant Greek for the passage is in Codd. Φ and D, of which D gives the fullest text, as follows: ὑμεις δε ζητειτε; εκ μεικρου αυξησαι και εκ μειζονος ελαττον ειναι Εισερχομενοι δε και παρακληθεντες δειπνησαι; μη ανακλεινεσθαι εις τους εξεχοντας τοπους μη ποτε ενδοξοτερος σου επελθη και προσελθων ο δειπνοκλητωρ ειπη σοι ετι κατω χωρει; και καταισχυνθηση Εαν δε αναπεσης; εις τον ηττονα τοπον και επελθη σου ηττων ερει σοι ο δειπνοκλητωρ; συναγε ετι ανω και εσται σοι τουτο χρησιμον. The codices of the Old Latin version (a b c e ff1.2 h n and and. em. of the Vulgate[344]) mostly support the same addition, though with many variations: d, as usual, agrees with none; g2has not the first clause down to εἶναι, while g1 m have nothing else. Besides the Curetonian Syriac, the margin of the Harkleian contains it in a shape much like d, noting that the paragraph is “found in Greek copies in this place, but in ancient copies only in St. Luke, κεφ. 53” [ch. xiv. 8, &c.]: Cureton has also seen it in one manuscript of the Peshitto (Brit. Mus. 14,456), but there too in the margin. Marshall states that it is contained in four codices of the Anglo-Saxon version, which proves its wide reception in the West. Of the Fathers, Hilary recognizes it, as apparently do Juvencus and Pope Leo the Great (a.d. 440-461). It must have been [pg 331] rejected by Jerome, being entirely absent from the great mass of Vulgate codices, nor is it in the Old Latin, f l q. No other Greek codex, or version, or ecclesiastical writer, has any knowledge of the passage: while the whole language of the Greek of Cod. D, especially in such words as δειπνοκλήτωρ, ἐξέχοντας, ἥττων, χρήσιμος, is so foreign to the style of St. Matthew's Gospel, that it seems rather to have been rendered from the Latin[345], although in the midst of so much variation it is hard to say from what copy. Cureton too testifies that the Syriac of the version named from him must have been made quite independently of that in the margins of the Harkleian and Peshitto.

No one has hitherto ventured to regard this paragraph as genuine, however perplexing it may be to decide at what period or even in what language it originated. The wide divergences between the witnesses must always dismiss it from serious consideration. Its chief critical use must be to show that the united testimony of the Old Latin, of the Curetonian Syriac, and of Cod. D, are quite insufficient in themselves to prove any more than that the reading they exhibit is ancient: certainly as ancient as the second century.

7. Matt. xxi. 28-31. This passage, so transparently clear in the common text, stands thus in the edition of Tregelles: (28) Τί δὲ ὑμῖν δοκεῖ? ἄνθρωπος εἶχεν τέκνα δύο, καὶ προσελθὼν τῷ πρώτῳ εἶπεν, Τέκνον, ὕπαγε σήμερον ἐργάζου ἐν τῷ ἀμπελῶνι. (29) ὁ δὲ ἀποκριθεὶς εἶπεν, Οὐ θέλω; ὕστερον δὲ μεταμεληθεὶς ἀπῆλθεν. (30) προσελθὼν δὲ τῷ δευτέρῳ εἶπεν ὡσαύτως. ὁ δὲ ἀποκριθεὶς εἶπεν, Ἐγώ, κύριε; καὶ οὐκ ἀπῆλθεν. (31) τίς ἐκ τῶν δύο ἐποίησεν τὸ θέλημα τοῦ πατρός? λέγουσιν, Ὁ ὕστερος. The above is indeed a brilliant exemplification of Bengel's Canon, “Proclivi orationi praestat ardua.” Lachmann in 1842 had given the same reading, with a few slight and unimportant exceptions. The question is proposed [pg 332] which of the two sons did their father's will; the reply is ὁ ὕστερος, the one that promised and then failed! Lachmann in 1850 (N. T., vol. ii. Praef. p. 5) remarks that had he been sure that πρῶτος (ver. 31) was the reading of Cod. C, he should have honoured it, the only word that makes sense, with a place in his margin: “Nihilo minus,” he naïvely adds, “id quod nunc solum edidi ... ὁ ὕστερος veri similius est altero, quod facile aliquis correctori adscribat, illud non item;” and we must fairly confess that no copyist would have sought to introduce a plain absurdity into so beautiful and simple a parable. “Quid vero,” he goes on to plead, “si id quod veri similius esse dixi ne intellegi quidem potest?” (a pertinent question certainly) “corrigetur, si modo necesse erit:” critical conjecture, as usual, is his panacea. Conjecture, however, is justly held inadmissible by Tregelles, whose mode of interpretation is a curiosity in its way. “I believe,” he says, “that ὁ ὕστερος refers not to the order in which the two sons have been mentioned, but to the previous expression about the elder son, ὕστερον δὲ μεταμεληθεὶς ἀπῆλθεν, afterwards he repented and went.” “Which of the two did his father's will! ὁ ὕστερος. He who afterwards [repented and went]. This answers the charge that the reading of Lachmann is void of sense” (An Account of the Printed Text, p. 107). I entertain sincere veneration for the character and services of Dr. Tregelles, but it is only right to assert at once that what stands in his text is impossible Greek. Even granting that instead of the plain answer “the first,” our Lord's adversaries resorted to the harsh and equivocal reply “he who afterwards,” they would not have said ὁ ὕστερος, but ὁ ὕστερον, or (the better to point out their reference to ὕστερον in ver. 29) ὁ τὸ ὕστερον.

Why then prefer nonsense, for the mere purpose of carrying out Bengel's canon to the extremity? The passage, precisely as it stands in Tregelles' N. T., is sanctioned by no critical authority whatsoever. Cod. B indeed has ὕστερος (which is here followed by Westcott and Hort), Cod. 4 δεύτερος, Codd. 13, 69, 124, 346 (Abbott's four), and 238, 262, 556, 604, perhaps others, ἔσχατος, one or other of which is in the Jerusalem Syriac and Bohairic, the Ethiopic (two manuscripts), the Armenian and two chief Arabic versions; but all these authorities (with tol. of the Vulgate secundâ manu, as also Isidore, the Pseudo-Athanasius, [pg 333] and John Damascene), transpose the order of the two sons in vv. 29, 30, so that the result produces just the same sense as in the Received text. The suggestion that the clauses were transferred in order to reconcile ὕστερος or ἔσχατος with the context may be met by the counter-statement that ὕστερος was just as likely to be substituted for πρῶτος to suit the inversion of the clauses. Against such inversion (which we do not pretend to recommend, though Westcott and Hort adopt it) Origen is an early witness, so that Cod. B and its allies are no doubt wrong: yet as that Father does not notice any difficulty in ver. 31, the necessary inference ought to be that he read πρῶτος[346]. Hippolytus testifies to ἔσχατος in ver. 31, but his evidence cannot be used, since he gives no indication in what order he took the clauses in vv. 29, 30. The indefensible part of Tregelles' arrangement is that, allowing the answers of the two sons to stand as in our common Bibles, he receives ὕστερος in the room of πρῶτος on evidence that really tells against him. The only true supporters of his general view are Cod. D αισχατος (i.e. ἔσχατος), the Old Latin copies a b e ff1.2 g1 h l, the best codices of the Vulgate (am. fuld. for. san. tol. harl.*), the Anglo-Saxon version, and Augustine, though not the Clementine edition of the Vulgate. Hilary perplexes himself by trying to explain the same reading; and Jerome, although he says “Sciendum est in veris exemplaribus non haberi novissimum sed primum,” has an expedient to account for the former word[347], which, however (if am. fuld., &c. may be trusted), he did not venture to reject when revising the Old Latin. On no true principles can Cod. D and its Latin allies [pg 334] avail against such a mass of opposing proof, whereof Codd. אCΦΣLX lead the van. Even the Curetonian Syriac, which so often favours Cod. D and the Old Latin, is with the textus receptus here.

8. Matt. xxvii. 35. After βάλλοντες κλῆρον the Received text, but not the Complutensian edition, has ἵνα πληρωθτῇ τὸ ῥηθὲν ὑπὸ τοῦ προφήτου, Διεμερίσαντο τὰ ἱμάτιά μου ἑαυτοῖς καὶ ἐπὶ τὸν ἱματισμόν μου ἔβαλον κλῆρον. Internal evidence may be about equal for the omission of the clause by homoeoteleuton of κλῆρον, and for its interpolation from John xix. 24, “with just the phrase τὸ ῥηθὲν ὑπὸ (or ἀπὸ) τοῦ προφήτου assimilated to Matthew's usual form of citation” (Alford, ad loc.). External evidence, however, places the spuriousness of the addition beyond doubt. It is first heard of in citations of Eusebius, and is read in the Old Latin codices a b c g2 (not g1) h q, the Clementine (not the Sixtine) Vulgate and even in am. lux., Harl. 2826, lind., in King's Libr. 1. D. ix and the margin of 1. E. vi (but not in fuld. for. tol.* em. ing. jac. san. nor in f ff1.2 g1 l), the Armenian (whose resemblance to the Vulgate is so suspicious), the Frankish and Anglo-Saxon, and as a matter of course in the Roman edition of the Arabic, and in the Persic of the Polyglott. The clause seems to be found in no manuscript of the Peshitto Syriac, and is consequently absent from Widmanstadt's edition and the Antwerp, Paris, and London Polyglotts. Tremellius first turned the Greek words into Syriac and placed them in the margin of his book, whence they were most unwisely admitted into the text of several later editions (but not into Lee's), without the slightest authority. They also appear in the text of the Harkleian, but the marginal note states that 'this passage from the prophet is not in two [“three” Codd. Assemani] Greek copies, nor in the ancient Syriac.' All other versions and Fathers (except Eusebius and the Pseudo-Athanasius), and all Greek manuscripts reject the clause, except Δ, 1, 17, 58 (marg.), 69, 118, 124, 262, 300, 503, 550, Evst. 55: Scholz adds “aliis multis,” which (judging from my own experience) I must take leave to doubt. Besides other slight changes (αυτοις Δ, κλήρους 69 secundâ manu) Codd. Δ, 61, 69, 503 and Eusebius read διά for ὑπό. The present case is one out of many that show an intimate connexion subsisting between Codd. 61 and 69.

9. Mark vi. 20. καὶ ἀκούσας αὐτοῦ πολλὰ ἐποίει, καὶ ἡδέως αὐτοῦ ἤκουε. “ ‘Did many things’ Engl. vers. I think it must have occurred to many readers that this is, to say the least, a very singular expression.” So writes Mr. Linwood, very truly, for nothing can well be more tame or unmeaning. His remedy we can say little for. “I think that for πολλὰ ἐποίει we should read πολλοῦ ἐποίει, i.e. magni faciebat. It is true that classical usage would require the middle voice, sc. πολλοῦ ἐποιεῖτο. But this rule is not always observed by the N. T. writers[348]” (Linwood, p. 11). If, instead of resorting to conjecture, he had opened Tischendorf's eighth edition, he would have found there a reading, adopted as well by that editor as by Westcott and Hort, whose felicity, had it been nothing more than a happy conjecture, he might well have admired. Codd. אBL for πολλὰ ἐποίει[349] have πολλὰ ἠπόρει “was much perplexed,” which the Bohairic confirms, only that, in translating, it joins πολλά with ἀκούσας. This close resemblance between the Bohairic version and Codd. אB (especially Cod. B) is very apparent throughout the N. T.; a single example being their united omission of ἰσχυρόν in Matt. xiv. 30 in company with but one other authority, the great cursive Cod. 33. Hence we do not hesitate to receive a variation supported by only a few first-rate authorities, where internal evidence (Canon II, p. [248]) pleads so powerfully in its favour. Although the middle voice is found elsewhere in the N. T., yet the active in this precise sense may be supported by good examples, even when used absolutely, as here: e.g. ἄλλος οἱ ἀπορέοντι ὑπεθήκατο Herod. i. 191: ὁ δ᾽ ἀπορῶν, ὥς φασι, μόλις κατενόησε τὴν πρόσχωσιν ταύτην τοῦ Ἀχελῴου Thuc. ii. 102.

Another less considerable but interesting variation, occurring just before, in chap. v. 36, παρακούσας “overhearing” instead of ἀκούσας, may be deemed probable on the evidence of א*BLΔ and the Latin e, which must have had the reading, though it is mistranslated neglexit[350]. We gladly credit the same group (אBCLΔ, 473, Evst. 150, 259) with another rare compound, κατευλόγει in ch. x. 16, whose intensive force is very excellent. [pg 336] In ch. xii. 17 a similar compound ἐξεθαύμαζον is too feebly vouched for by אB alone.

[Third Edition. It is only fair to retain unchanged the note on Mark vi. 20, inasmuch as the “Two Members of the N. T. Company” have exercised their right of claiming my assent to the change of ἐποίει into ἠπόρει. I must, however, retract that opinion, for the former reading now appears to me to afford an excellent sense. Herod gladly heard the Baptist, and did many things at his exhortation; every thing in fact save the one great sacrifice which he could not persuade himself to make.]

10. Mark vii. 19. The substitution of καθαρίζων for καθαρίζον, so far from being the unmeaning itacism it might seem at first sight, is a happy restoration of the true sense of a passage long obscured by the false reading. For the long vowel there is the overwhelming evidence of אAB (hiat C) EFGH LSXΔ, 1, 13, 28, 48, 50, 53, 58, 59 (me teste), 61**, 64, 65, 69, 122* 124, 229, 235, 244, 251, 282, 346, 435, 473, 492, 508, 515, 570, 622, Evst. 49, 259, and Erasmus' first edition: his second reads ἐκκαθαρίζων, his third καθαρίζον of ΦΣKMUVΓΠ, 547, 558, and perhaps a majority of the cursives. The reading of D καθαρίζει (καθαρίζειν 61 primâ manu), as also καὶ καθαρίζει of Evst. 222 and the Latin i, seem to favour the termination -ον: purgans of a b c (even d) f ff2 g1.2 l? n q and the Vulgate, is of course neutral. The Peshitto ܕܡܕܒܝܐ (or ܐܝܒܕܡܕ) (qui purgat) refers in gender to the noun immediately preceding, and would require καθαρίζοντα. Will any one undertake to say what is meant by the last clause of the verse as it stands in the Authorized English version, and as it must stand, so long as καθαρίζον is read? If, on the other hand, we follow Lachmann, Tregelles, Tischendorf, Westcott and Hort, we must take the Lord's words to end with ἐκπορεύεται, and regard καθαρίζων πάντα τὰ βρώματα as the Evangelist's comment upon them: “This he said, to make all things clean.” Compare Acts x. 15. This, and none other, seems to have been the meaning assigned to the passage by the Greek Fathers. It is indeed most simply expressed by Chrysostom (Hom. II. in Matt. p. 526 A): Ὁ δὲ Μάρκος φησίν, ὅτι καθαρίζων τὰ βρώματα, ταῦτα ἔλεγεν, where Dr. Field's elaborate note should be consulted. [pg 337] He rightly judges that Chrysostom was treading in the steps of Origen: καὶ μάλιστα ἐπεὶ κατὰ τὸν Μάρκον ἔλεγε ταῦτα ὁ Σωτήρ, καθαρίζων πάντα τὰ βρώματα. Hence Gregory Thaumaturgus designates the Lord as ὁ σωτὴρ ὁ πάντα καθαρίζων τὰ βρώματα. I know not how Tischendorf came to overlook the passage from Chrysostom: Tregelles very seldom uses him. It is obvious how well the elliptical form of the expression suits this Evangelist's style, which is often singularly concise and abrupt, yet never obscure.

11. Mark xvi. 9-20. In Vol. I. Chap. I, we engaged to defend the authenticity of this long and important passage, and that without the slightest misgiving (p. 7). Dean Burgon's brilliant monograph, “The Last Twelve Verses of the Gospel according to St. Mark vindicated against recent objectors and established” (Oxford and London, 1871), has thrown a stream of light upon the controversy, nor does the joyous tone of his book misbecome one who is conscious of having triumphantly maintained a cause which is very precious to him. We may fairly say that his conclusions have in no essential point been shaken by the elaborate and very able counter-plea of Dr. Hort (Notes, pp. 28-51). This whole paragraph is set apart by itself in the critical editions of Tischendorf and Tregelles. Besides this, it is placed within double brackets by Westcott and Hort, and followed by the wretched supplement derived from Cod. L (vide infra), annexed as an alternative reading (αλλως). Out of all the great manuscripts, the two oldest (אB) stand alone in omitting vers. 9-20 altogether[351]. Cod. B, however, betrays consciousness on the scribe's part that something is left out, inasmuch as after ἐφοβοῦντο γάρ ver. 8, a whole column is left perfectly blank (the only blank one in the whole volume[352]), as well as the rest of the column containing ver. 8, which is usual [pg 338] in Cod. B at the end of every other book of Scripture. No such peculiarity attaches to Cod. א. The testimony of L, that close companion of B, is very suggestive. Immediately after ver. 8 the copyist breaks off; then in the same hand (for all corrections in this manuscript seem primâ manu: see p. [138]), at the top of the next column we read ... φερετε που και ταυτα+ ... πάντα δὲ τα παρηνγελμενα τοῖς περι τον πετρον συντομωσ ἐξηγγιλαν+ μετα δὲ ταῦτα καὶ αὐτος ὁ ισ, ἁπο ἁνατολησ καὶ ἁχρι δυσεωσ ἐξαπεστιλεν δι αὐτων το ϊἑρον καὶ ἁφθαρτον κηρυγμα+τησ αἱῶνιου σωτηριασ+ ... εστην δε και ταῦτα φερομενα μετα το ἑφοβουντο γαρ+ ... Αναστὰσ δὲ πρωï πρωτη σαββατου+κ.τ.λ., ver. 9, ad fin. capit. (Burgon's facsimile, facing his p. 113: our facsimile No. 21): as if vv. 9-20 were just as little to be regarded as the trifling apocryphal supplement[353] which precedes them. Besides these, the twelve verses are omitted in none but some old Armenian codices[354] and two of the Ethiopic, k of the Old Latin, and an Arabic Lectionary [ix] No. 13, examined by Scholz in the Vatican. The Old Latin Codex k puts in their room a corrupt and careless version of the subscription in L ending with σωτηρίας (k adding amen): the same subscription being appended to the end of the Gospel in the two Ethiopic manuscripts, and (with ἀμήν) in the margin of 274 and the Harkleian. Not unlike is the marginal note in Hunt. 17 or Cod. 1 of the Bohairic, translated by Bp. Lightfoot above. Of cursive Greek manuscripts 137, 138, which Birch had hastily reported as marking the passage with an asterisk, each contains the marginal annotation given below, which claims the passage as genuine, 138 with no asterisk at all, 137 (like 36 and others) with an ordinary mark of reference from the text to the note, where (of course) it is repeated[355]. Other manuscripts contain marginal scholia respecting it, of which the following is [pg 339] the substance. Cod. 199 has τέλος[356] after ἐφοβοῦντο γάρ and before Ἀναστὰς δέ, and in the same hand as τέλος we read, ἔν τισι τῶν ἀντιγράφων οὐ κεῖται ταῦτα, ἀλλ᾽ ἐνταῦθα καταπαύει. The kindred Codd. 20, 215, 300 (but after ver. 15, not ver. 8) mark the omission in some (τισί) copies, adding ἐν δὲ τοῖσ ἀρχαίοις πάντα ἀπαράλειπτα κεῖται, and these had been corrected from Jerusalem copies (see pp. [161] and note, [193]). Cod. 573 has for a subscription ἐγράφη καὶ ἀντεβλήθη ὁμοίως ἐκ τῶν ἐσπουδασμένων κεφαλαίοις σλζ: where Burgon, going back to St. Matthew's Gospel (see p. [161], note) infers that the old Jerusalem copies must have contained our twelve verses. Codd. 15, 22 conclude at ἐφοβοῦντο γάρ, then add in red ink that in some copies the Evangelist ends here, ἐν πολλοῖς δὲ καὶ ταῦτα φέρεται, affixing vers. 9-20. In Codd. 1, 205 (in its duplicate 206 also), 209 is the same notice, ἄλλοις standing for πολλοῖς in 206, with the additional assertion that Eusebius “canonized” no further than ver. 8, a statement which is confirmed by the absence of the Ammonian and Eusebian numerals beyond that verse in אALSU and at least eleven cursives, with am. fuld. ing. of the Vulgate. It would be no marvel if Eusebius, the author of this harmonizing system, had consistently acted upon his own rash opinion respecting the paragraph, an opinion which we shall have to notice presently, and such action on his part would have added nothing to the strength of the adverse case. But it does not seem that he really did so. These numerals appear in most manuscripts, and in all parts of them, with a good deal of variation which we can easily account for. In the present instance they are annexed to ver. 9 and the rest of the passage in Codd. CEKVΠ, and (with some changes) in GHMΓΔΛ and many others: in Cod. 566 the concluding sections are there (σλδ ver. 11, σλε ver. 12, σλς ver. 14) without the canons. In their respective margins the annotated codices 12 (of Scholz), 24, 36, 37, 40, 41, 108, 129, 137, 138, 143, 181, 186, 195, 210, 221, 222, 237, 238, 255, 259, 299, 329, 374 (twenty-four in all), present in substance[357] the same [pg 340] weighty testimony in favour of the passage: παρὰ πλείστοις ἀντιγράφοις οὐ κεῖται (thus far also Cod. 119, adding only ταῦτα, ἀλλ᾽ ἐνταῦθα καταπαύει) ἐν τῷ παρόντι εὐαγγελίῳ, ὡς νόθα νομίσαντες αὐτὰ εἶναι; ἀλλὰ ἡμεῖς ἐξ ἀκριβῶν ἀντιγράφων ἐν πλείστοις εὑρόντες αὐτὰ καὶ κατὰ τὸ Παλαιστιναῖον εὐαγγέλιον Μάρκου, ὡς ἔχει ἡ ἀλήθεια, συντεθείκαμεν καὶ τὴν ἐν αὐτῷ ἐπιφερομένην δεσποτικὴν ἀνάστασιν. Now this is none other than an extract from Victor of Antioch's [v] commentary on St. Mark, which they all annex in full to the sacred text, and which is expressly assigned to that Father in Codd. 12, 37, 41. Yet these very twenty-four manuscripts have been cited by critical editors as adverse to the authenticity of a paragraph which their scribes never dreamt of calling into question, but had simply copied Victor's decided judgement in its favour. His appeal to the famous Palestine codices which had belonged to Origen and Pamphilus (see p. [55] and note), is found in twenty-one of them, possibly these documents are akin to the Jerusalem copies mentioned in Codd. Evan. Λ, 20, 164, 262, 300, &c.

All other codices, e.g. ACD (which is defective from ver. 15, primâ manu) EFwGH (begins ver. 14) KMSUVXΓΔΠ, 33, 69, the Peshitto, Jerusalem and Curetonian Syriac (which last, by a singular happiness, contains vv. 17-20, though no other part of St. Mark), the Harkleian text, the Sahidic (only ver. 20 is preserved), the Bohairic and Ethiopic (with the exceptions already named), the Gothic (to ver. 12), the Vulgate, all extant Old Latins except k (though a primâ manu and b are defective), the Georgian, the printed Armenian, its later manuscripts, and all the lesser versions (Arabic, &c.), agree in maintaining the paragraph. It is cited, possibly by Papias, unquestionably by Irenaeus (both in Greek and Latin), by Tertullian, and by Justin Martyr[358] as early as the second [pg 341] century; by Hippolytus (see Tregelles, An Account of the Printed Text, p. 252), by Vincentius at the seventh Council of Carthage, by the Acta Pilati, the Apostolic Constitutions, and apparently by Celsus in the third; by Aphraates (in a Syriac Homily dated a.d. 337), the Syriac Table of Canons, Eusebius, Macarius Magnes, Didymus, the Syriac Acts of the Apostles, Leontius, Ps.-Ephraem, Jerome, Cyril of Jerusalem[359], Epiphanius, Ambrose, Augustine, Chrysostom, in the fourth; by Leo, Nestorius, Cyril of Alexandria, Victor of Antioch, Patricius, Marius Mercator, in the fifth; by Hesychius, Gregentius, Prosper, John, abp. of Thessalonica, and Modestus, in the fifth and sixth[360]. Add to this, what has been so forcibly stated by Burgon (ubi supra, p. [205]), that in the Calendar of Greek Church lessons, which existed certainly in the fourth century, very probably much earlier, the disputed verses were honoured by being read as a special matins service for Ascension Day (see p. 81), and as the Gospel for St. Mary Magdalene's Day, July 22 (p. 89); as well as by forming the third of the eleven εὐαγγέλια ἀναστάσιμα ἑωθινά, the preceding part of the chapter forming the second (p. 85): so little were they suspected as of even doubtful authenticity[361].

The earliest objector to vers. 9-20 we know of was Eusebius (Quaest. ad Marin.), who tells that they were not ἐν ἅπασι τοῖς ἀντιγράφοις, but after ἐφοβοῦντο γάρ that τὰ ἑξῆς are found σπανίως ἔν τισιν, yet not in τὰ ἀκριβῆ: language which Jerome twice echoes and almost exaggerates by saying “in raris fertur Evangeliis, omnibus Graeciae libris paene hoc capitulum fine non habentibus.” [pg 342] A second cause with Eusebius for rejecting them is μάλιστα εἴπερ ἔχοιεν ἀντιλογίαν τῇ τῶν λοιπῶν εὐαγγελιστῶν μαρτυρίᾳ[362]. The language of Eusebius has been minutely examined by Dean Burgon, who proves to demonstration that all the subsequent evidence which has been alleged against the passage, whether of Severus, or Hesychius, or any other writer down to Euthymius Zigabenus in the twelfth century, is a mere echo of the doubts and difficulties of Eusebius, if indeed he is not retailing to us at second-hand one of the fanciful Biblical speculations of Origen. Jerome's recklessness in statement has been already noticed (Vol. II. p. [269]); besides that, he is a witness on the other side, both in his own quotations of the passage and in the Vulgate, for how could he have inserted the verses there, if he had judged them to be spurious?

With regard to the argument against these twelve verses arising from their alleged difference in style from the rest of the Gospel, I must say that the same process might be applied—and has been applied—to prove that St. Paul was not the writer of the Pastoral Epistles (to say nothing of that to the Hebrews), St. John of the Apocalypse, Isaiah and Zechariah of portions of those prophecies that bear their names. Every one used to literary composition may detect, if he will, such minute variations as have been made so much of in this case[363], either in his own writings, or in those of the authors he is most familiar with.

Persons who, like Eusebius, devoted themselves to the pious [pg 343] task of constructing harmonies of the Gospels, would soon perceive the difficulty of adjusting the events recorded in vers. 9-20 to the narratives of the other Evangelists. Alford regards this inconsistency (more apparent than real, we believe) as “a valuable testimony to the antiquity of the fragment” (N. T. ad loc.): we would go further, and claim for the harder reading the benefit of any critical doubt as to its genuineness (Canon I. Vol. II. p. [247]). The difficulty was both felt and avowed by Eusebius, and was recited after him by Severus of Antioch or whoever wrote the scholion attributed to him. Whatever Jerome and the rest may have done, these assigned the ἀντιλογία, the ἐναντίωσις they thought they perceived, as a reason (not the first, nor perhaps the chief, but still as a reason) for supposing that the Gospel ended with ἐφοβοῦντο γάρ. Yet in the balance of probabilities, can anything be more unlikely than that St. Mark broke off so abruptly as this hypothesis would imply, while no ancient writer has noticed or seemed conscious of any such abruptness[364]? This fact has driven those who reject the concluding verses to the strangest fancies;—namely, that, like Thucydides, the Evangelist was cut off before his work was completed, or even that the last leaf of the original Gospel was torn away.

We emphatically deny that such wild surmises[365] are called [pg 344] for by the state of the evidence in this case. All opposition to the authenticity of the paragraph resolves itself into the allegations of Eusebius and the testimony of אB. Let us accord to these the weight which is their due: but against their verdict we can appeal to a vast body of ecclesiastical evidence reaching back to the earlier part of the second century[366]; to nearly all the versions; and to all extant manuscripts excepting two, of which one is doubtful. So powerfully is it vouched for, that many of those who are reluctant to recognize St. Mark as its author, are content to regard it notwithstanding as an integral portion of the inspired record originally delivered to the Church[367].

12. Luke ii. 14. If there be one case more prominent than another in the criticism of the New Testament, wherein solid reason and pure taste revolt against the iron yoke of ancient authorities, it is that of the Angelic Hymn sung at the Nativity. In the common text all is transparently clear:

δοξα εν υψιστοισ θεῳ, Glory to God in the highest,

και επι γησ ειρηνη; And on earth peace:

εν ανθρωποισ ευδοκια. Good will amongst men.

The blessed words are distributed, after the Hebrew fashion, into a stanza consisting of three members. In the first and second lines heaven and earth are contrasted; the third refers to both those preceding, and alleges the efficient cause which has brought God glory and earth peace. By the addition of a single letter to the end of the last line, by merely reading εὐδοκίας for εὐδοκία, the rhythmical arrangement is utterly marred[368], and the simple shepherds are sent away with a message, the diction of [pg 345] which no scholar has yet construed to his own mind[369]. Yet such is the conclusion of Lachmann, Tischendorf, Tregelles, Westcott and Hort, although Tregelles and the Cambridge fellow-workers allow εὐδοκία a place in their margins. Of the five great uncials C is unfortunately defective, but א*AB*D, and no other Greek manuscript whatever, read εὐδοκίας: yet A is so inconstant in this matter that in the primitive 14th or Morning Hymn, a cento of Scripture texts, annexed to the Book of Psalms, its reading is εὐδοκεία (Baber, Cod. Alex., p. 569), and such was no doubt the form used in Divine service, as appears from the great Zürich Psalter Od. The rest of the uncials extant (אcB3EGHKLMPSUVΓΔΛΞ, &c.), and all the cursives follow the common text, which is upheld by the Bohairic, by the three extant Syriac (the Peshitto most emphatically, the Jerusalem, and the Harkleian both in the text and Greek margin), by the Armenian and Ethiopic versions. The Vulgate, as is well known, renders “in hominibus bonae voluntatis,” and thus did all the forms of the Old Latin, and after it the Gothic. Hence it follows, as a matter of course, that the Latin Fathers, such as Hilary and Augustine, and the Latin interpreters of Irenaeus (who seems really to have omitted ἐν, as do D and a few cursives) and of the false Athanasius, adopted the reading of their own Bibles. Origen also, in a passage not now extant in the Greek, is made in Jerome's translation of it manifestly to choose the same form. We can only say that in so doing he is the only Greek who favours εὐδοκίας, and his own text has εὐδοκία in three several places, though no special stress is laid by him upon it. But here comes in the evidence of the Greek Fathers—their virtually unanimous evidence—with an authority from which there is, or ought to be, no appeal. Dean Burgon (The Revision Revised, pp. 42-46) affords us a list of forty-seven, all speaking in a manner too plain for doubt, most of them several times over, twenty-two of them having flourished before the end of the [pg 346] fifth century, and who must have used codices at least as old and pure as א or B. They are Irenaeus, of the second century; the Apostolical Constitutions and Origen three times in the third; Eusebius, Aphraates the Persian, Titus of Bostra, Didymus, Gregory Nazianzen, Cyril of Jerusalem (who has been quoted in error on the wrong side), Epiphanius, Gregory of Nyssa four times, Ephraem Syrus, Philo of Carpasus, a nameless preacher at Antioch, and Chrysostom (nine times over, interpreting also εὐδοκία by καταλλαγή) in the fourth; Cyril of Alexandria on fourteen occasions, Theodoret on four, Theodotus of Ancyra, the Patriarch Proclus, Paulus of Emesa, the Eastern Bishops at Ephesus in 431, and Basil of Seleucia in the fifth; Cosmas Indicopleustes, Anastasius Sinaita, and Eulogius of Alexandria in the sixth; Andreas of Crete in the seventh; with Cosmas of Maiuma, John Damascene, and Germanus, Archbishop of Constantinople, in the eighth[370]. Such testimony, supported by all later manuscripts, together with the Bohairic and Syriac versions, cannot but overpower the transcriptional blunder of some early scribe, who cannot, however, have lived later than the second century.

To those with whom the evidence of אBD and of the Latins united appears too mighty to resist, we would fain prefer one request, that in their efforts to extract some tolerable sense out of εὐδοκίας, they will not allow themselves to be driven to renderings which the Greek language will not endure. To spoil the metrical arrangement by forcing the second and third members of the stanza into one, is in itself a sore injury to the poetical symmetry of the passage, but from their point of view it cannot be helped. When they shall come to translate, it will be their endeavour to be faithful, if grammatical faithfulness be possible in a case so desperate. “Peace on earth for those that will have it,” as Dean Alford truly says, is untenable in Greek, as well as in theology: “among men of good pleasure” is unintelligible to most minds. Professor Milligan (Words of the New Testament, p. 194) praises as an interesting form “among men of his good pleasure,” which, not at all unnecessarily, he expounds to signify “among men whom He hath loved.” Again, “among men in whom He is well pleased” (compare chap. iii. 22) can [pg 347] be arrived at only through some process which would make any phrase bear almost any meaning the translator might like to put upon it. The construction adopted by Origen as rendered by Jerome, pax enim quam non dat Dominus non est pax bonae voluntatis, εὐδοκίας being joined with εἰρήνη, is regarded by Dr. Hort “to deserve serious attention, if no better interpretation were available” and for the trajection he compares ch. xix. 38; Heb. xii. 11 (Notes, p. 56). Dr. Westcott holds that since “ἀνθρώποις εὐδοκίας is undoubtedly a difficult phrase, and the antithesis of γῆς and ἀνθρώποις agrees with Rom. viii. 22, εὐδοκία claims a place in the margin” (ibid.): no very great concession, when the general state of the evidence is borne in mind[371].

13. Luke vi. 1. ἐγένετο δὲ ἐν σαββάτῳ δευτεροπρώτῳ. Here again Codd. אB coincide in a reading which cannot be approved, omitting δευτεροπρώτῳ by way of getting rid of a difficulty, as do both of them in Mark xvi. 9-20, and א in Matt. xxiii. 35. The very obscurity of the expression, which does not occur in the parallel Gospels or elsewhere, attests strongly to its genuineness, if there be any truth at all in canons of internal evidence[372]: not to mention that the expression ἐν ἑτέρῳ σαββάτῳ ver. 6 favours the notion that the previous sabbath [pg 348] had been definitely indicated. Besides אB, δευτεροπρώτῳ is absent from L, 1, 22, 33, 69 (where it is inserted in the margin by W. Chark, and should not be noticed, see above), 118, 157, 209. A few (RΓ, 13, 117, 124 primâ manu, 235) prefer δευτέρω πρώτω, which, as the student will perceive, differs from the common reading only by a familiar itacism. As this verse commences a Church lesson (that for the seventh day or Sabbath of the third week of the new year, see Calendar), Evangelistaria leave out, as usual, the notes of time; in Evst. 150, 222, 234, 257, 259 (and no doubt in other such books, certainly in the Jerusalem Syriac), the section thus begins, Ἐπορεύετο ὁ Ἰησοῦς τοῖς σάββασιν: this however is not, properly speaking, a various reading at all. Nor ought we to wonder if versions pass over altogether what their translators could not understand[373], so that we may easily account for the silence of the Peshitto Syriac, Bohairic, and Ethiopic, of the Old Latin b c l q f (secundâ manu) q, and (if they were worth notice) of the Persic and the Polyglott Arabic, though both the Roman and Erpenius' Arabic have δεύτερῳ, and so too the Ethiopic according to Scholz; e “sabbato mane,” f “sabbato a primo:” the Harkleian Syriac, which renders the word, notes in the margin its absence from some copies. Against this list of authorities, few in number, and doubtful as many of them are, we have to place the Old Latin a f* ff2 g1.2, all copies of the Vulgate, its ally the Armenian, the Gothic and Harkleian Syriac translations, the uncial codices ACDEHKMRSUVXΓΔΛΠ, all cursives except the seven cited above, and the Fathers or scholiasts who have tried, with whatever success, to explain the term: viz. Epiphanius, Chrysostom, Isidore of Pelusium, Pseudo-Caesarius, Gregory of Nazianzus, Jerome[374], Ambrose (all very expressly, as may be seen in Tischendorf's note, and in Dean Burgon's “The Revision Revised,” pp. 73-4), Clement of Alexandria probably, and later writers. Lachmann and Alford [pg 349] place δευτεροπρώτῳ within brackets, Tregelles rejects it, as does Tischendorf in his earlier editions, but restores it in his seventh and eighth, in the latter contrary to Cod. א. Westcott and Hort banish it to the margin, intimating (if I understand their notation aright) that it seems to contain distinctive and fresh matter, without deserving a place in the text even as well as Ἰησοῦ in Matt. i. 18. On reviewing the whole mass of evidence, internal and external, we submit the present as a clear instance in which the two oldest copies conspire in a false or highly improbable reading, and of a signal exemplification of the Canon, Proclivi orationi praestat ardua.

14. Luke x. 41, 42. Ἑνὸς δέ ἐστι χρεία. This solemn speech of our Divine Master has shaken many a pulpit, and sanctified many a life. We might be almost content to estimate Cod. B's claim to paramount consideration as a primary authority by the treatment this passage receives from the hand of its scribe, at least if the judgement were to rest with those who are willing to admit that a small minority, whereof B happens to form one of the members, is not necessarily in the right. Westcott and Hort in the margin of their published edition (1881) reduce the whole sentence between Μάρθα ver. 41 and Μαρία ver. 42 to the single word θορυβάζῃ, the truer reading in the place of τυρβάζῃ: in their privately circulated issue dated ten years earlier they had gone further, placing within double brackets μεριμνᾷς καί and from περὶ πολλά downwards. They could hardly do less on the principles they have adopted, while yet they feel constrained to concede that, though not belonging to the original Gospel, the excluded words do not, on the other hand, read like the invention of a paraphrast. They do not indeed: and it is when abstract theories such as modern critics have devised are subjected to so violent a strain, that we can best discern their intrinsic weakness, of which indeed these editors have here shown their consciousness by a change of mind not at all usual with them. For the grave omission indicated above we have but one class of authorities, that of the D, a b e ff2 i l, and Ambrose, the Latins omitting θορυβάζῃ too: while ἑνὸς δέ ἐστι χρεία is not found in c also, and does not appear in Clement. The succeeding γάρ or δέ is of course left out by all these, and by 262, the Vulgate, Curetonian Syriac, Armenian, [pg 350] and Jerome. This testimony, almost purely Western, is confirmed or weakened as the case may be, by the systematic omissions of clauses towards the end of the Gospel in the same books, of which we spoke in Chap. X (see p. [299], note).

We confess that we had rather see this grand passage expunged altogether from the pages of the Gospel than diluted after the wretched fashion adopted by א and B: ὀλίγων δὲ χρεία ἐστιν ἢ ἑνός; the first hand of א omitting χρεία in its usual blundering way. This travestie of a speech which seems to have shocked the timorous by its uncompromising exclusiveness, much as we saw in the case of Matt. v. 22, is further supported (with some variation in the order) by L, by the very ancient second hand of C, by 1, 33, the Bohairic, Ethiopic, the margin of the Harkleian, by Basil, Jerome, Cyril of Alexandria in the Syriac translation of his commentary[375], and by Origen as cited in a catena: ὀλίγων δὲ ἐστι χρεία is found in 38, the Jerusalem Syriac, and in the Armenian (ὧδε being inserted before ἐστιν). This latter reading is less incredible than that of אBL, notwithstanding the ingenuity of Basil's comment, ὀλίγων μὲν δηλονότι τῶν πρὸς παρασκευήν, ἑνὸς δὲ τοῦ σκοποῦ. In this instance, as in some others, the force of internal evidence suffices to convince the unprejudiced reader (it has almost convinced Drs. Westcott and Hort, who have no note on the passage), that the Received text should here remain unchanged, vouched for as it is by AC*EFGHKMPSUVΓΔΛΠ (Χ and Ξ being defective), by every cursive except three, by the Peshitto and Cureton's Syriac (the latter so often met with in the company of D), by the Harkleian text, by f g1 g2? q of the Old Latin, and by the Vulgate. Chrysostom, Augustine (twice), John Damascene and one or two others complete the list: even Basil so cites the passage once, so that his comment may not be intended for anything more than a gloss. No nobler sermon was ever preached on this fertile text than that of Augustine, De verbis Domini, in Evan. Luc. xxvii. His Old Latin copies, at any rate, contained the words “Circa multa es occupata: porro unum est necessarium. Jam hoc sibi Maria legit.” “Transit labor multitudinis, et remanet caritas unitatis” is his emphatic comment.

15. Luke xxii. 17-20. This passage has been made the subject of a most instructive discussion by Dean Blakesley[376] (d. 1885), whose notion respecting it deserves more consideration than it would seem to have received, though it must no doubt be ultimately set aside through the overpowering weight of hostile authority. He is perplexed by two difficulties lying on the surface, the fact that the Lord twice took a cup, before and after the breaking of the bread; and the close resemblance borne by vv. 19 and 20 to the parallel passage of St. Paul, 1 Cor. xi. 24, 25. The common mode of accounting for the latter phenomenon seems very reasonable, namely, that the Evangelist, Paul's almost constant companion in travel, copied into his Gospel the very language of the Apostle, so far as it suited his design. In speaking of the two cups St. Luke stands alone, and much trouble has been taken to illustrate the use of the Paschal cup from Maimonides [d. 1206] and other Jewish doctors, all too modern to be implicitly depended on. Dean Alford indeed (N. T. ad loc.) hails “this most important addition to our narrative,” which “amounts, I believe, to a solemn declaration of the fulfilment of the Passover rite, in both its usual divisions—the eating of the lamb, and drinking the cup of thanksgiving.” Thus regarded, the old rite would be concluded and abrogated in vv. 17, 18; the new rite instituted in vv. 19, 20. To Dean Blakesley all this appears wholly unsatisfactory, and he resorts for help to our critical authorities. He first gets rid of the words of ver. 19 after σῶμά μου, and of all ver. 20, and so far his course is sanctioned by Westcott and Hort, who place the whole passage within their double brackets, and pronounce it a perverse interpolation from 1 Cor. xi. 24, 25. This much accomplished, the cup is now mentioned but once, but with this awkward peculiarity, that it precedes the bread in the order of taking and blessing, which is a downright contradiction of St. Matthew (xxvi. 26-29) and of St. Mark (xiv. 22-25), as well as of St. Paul. Here Westcott and Hort refuse to be carried further, and thus leave the remedy worse than the disease[377], if indeed [pg 352] there be any disease to remedy. Dean Blakesley boldly places Luke xxii. 19 (ending at σῶμά μου) before ver. 17, and his work is done: the paragraph thus remodelled is self-consistent, but it is robbed of everything which has hitherto made it a distinctive narrative, supplementing as well as confirming those of the other two Evangelists.

Now for the last step in Dean Blakesley's process of emendation, the transposition of ver. 19 before ver. 17, there is no other authority save b e of the Old Latin and Cureton's Syriac, the last with this grave objection in his eyes, that it exhibits the whole of ver. 19, including that τοῦτο ποιεῖτε εἰς τὴν ἐμὴν ἀνάμνησιν which he would regard as specially belonging of right, and as most suitable for, St. Paul's narrative (Praelectio, p. 16), although Justin Martyr cites the expression with the prelude οἱ γὰρ ἀπόστολοι ἐν τοῖς γενομένοις ὑπ᾽ αὐτῶν ἀπομνημονεύμασιν, ἂ καλεῖται, εὐαγγέλια. The later portion of ver. 19 and the whole of ver. 20, as included in the double brackets of Westcott and Hort, are absent from Cod. D, and of the Latins from a b e ff i l, as is ver. 20 from the Curetonian Syriac also: authorities for the most part the same as we had to deal with in our Chap. X. p. [299], note. Another, and yet more violent remedy, to provide against the double mention of the cup, is found in the utter omission of vers. 17, 18 in Evst. 32 and the editio princeps of the Peshitto Syriac, countenanced by many manuscripts of the same[378]. Thus both the chief Syriac translations found a difficulty here, though they remedied it in different ways[379].

The scheme of Dean Blakesley is put forth with rare ingenuity[380], and maintained with a boldness which is best engendered [pg 353] and nourished by closing the eyes to the strength of the adverse case. We have carefully enumerated the authorities of every kind which make for him, a slender roll indeed. When it is stated that the Received text (with only slight and ordinary variations) is upheld by Codd. אABCEFGHKLM (hiant PR) SUXVΓΔΛΠ, by all cursives and versions, except those already accounted for, it will be seen that his view of the passage can never pass beyond the region of speculation, until the whole system of Biblical Criticism is revolutionized by means of new discoveries which it seems at present vain to look for.

16. Luke xxii. 43, 44. ὤφθη δὲ αὐτῷ ἄγγελος ἀπ᾽ οὐρανοῦ ἐνισχύων αὐτόν. καὶ γενόμενος ἐν ἀγωνίᾳ, ἐκτενέστερον προσηύχετο; ἐγένετο δὲ ὁ ἱδρὼς αὐτοῦ ὡσεὶ θρόμβοι αἵματος καταβαίνοντες ἐπι τὴν γῆν. It is a positive relief to know that any lingering doubt which may have hung over the authenticity of these verses, whose sacred words the devout reader of Scripture could so ill spare, is completely dissipated by their being contained in Cod. א[381]. The two verses are omitted in ABRT, 124, 561 (in 13 only ὤφθη δὲ is primâ manu), in f of the Old Latin, in at least ten manuscripts of the Bohairic[382], with some Sahidic and Armenian codices. A, however, whose inconsistency we had to note when considering ch. ii. 14, affixes to the latter part of ver. 42 (πλήν), “to [pg 354] which they cannot belong” (Tregelles), the proper Ammonian and Eusebian numerals for vv. 43-4 (ι)σπγ, and thus shows that its scribe was acquainted with the passage[383]: some Armenian codices leave out only ver. 44, as apparently does Evan. 559. In Codd. Γ, 123, 344, 512, 569, (440 secundâ manu in ver. 43) the verses are obelized, and are marked by asterisks in ESVΔΠ, 24, 36, 161, 166, 274, 408: these, however, may very well be, and in some copies doubtless are, lesson-marks for the guidance of such as read the divine service (cf. sequent.). A scholion in Cod. 34 [xi] speaks of its absence from some copies[384]. In all known Evangelistaria and in their cognate Cod. 69* and its three fellows, the two verses, omitted in this place, follow Matt. xxvi. 39, as a regular part of the lesson for the Thursday in Holy Week: in the same place the margin of C (tertiâ manu) contains the passage, C being defective in Luke xxii from ver. 19. In Cod. 547 the two verses stand (in redder ink, with a scholion) not only after Matt. xxvi. 39, but also in their proper place in St. Luke[385]. Thus too Cod. 346, and the margin of Cod. 13. Codd. LQ place the Ammonian sections and the number of the Eusebian canons differently from the rest (but this kind of irregularity very often occurs in manuscripts), and the Philoxenian margin in one of Adler's manuscripts (Assem. 2) states that it is not found “in Evangeliis apud Alexandrinos, proptereaque [non?] posuit eam S. Cyrillus in homilia ...:” the fact being that the verses are not found in Cyril's “Homilies on Luke,” published in Syriac at Oxford by Dean Payne Smith, [pg 355] nor does Athanasius ever allude to them. They are read, however, in Codd. אDFGHKLMQUXΛ, 1, and all other known cursives, without any marks of suspicion, in the Peshitto, Curetonian (omitting ἀπ᾽ οὐρανοῦ), Harkleian and Jerusalem Syriac (this last obelized in the margin), the Ethiopic, in some Sahidic, Bohairic, and Armenian manuscripts and editions, in the Old Latin a b c e ff2 g1.2 i l q, and the Vulgate. The effect of this great preponderance is enhanced by the early and express testimony of Fathers. Justin Martyr (Trypho, 103) cites ἱδρὼς ὡσεὶ θόμβοι as contained ἐν τοῖς ἀπομνημονεύμασιν ἅ φημι ὑπὸ τῶν ἀποστόλων αὐτοῦ καὶ τῶν ἐκείνοις παρακολουθησάντων (see Luke i. 3, Alford) συντετάχθαι. Irenaeus (iii. 222) declares that the Lord ἵδρωσε θρόμβους αἵματος in the second century. In the third, Hippolytus twice, Dionysius of Alexandria, and Pseudo-Tatian; in the fourth, Arius, Eusebius, Athanasius, Ephraem Syrus, Didymus, Gregory of Nazianzen, Epiphanius, Chrysostom, Pseudo-Dionysius Areopagita; in the fifth, Julian the heretic, Theodore of Mopsuestia, Nestorius, Cyril of Alexandria, Paulus of Emesa, Gennadius, Theodoret, Bishops at Ephesus in 431; and later writers such as Pseudo-Caesarius, Theodosius of Alexandria, John Damascene, Maximus, Theodore the heretic, Leontius of Byzantium, Anastasius Sinaita, Photius, as well as Hilary, Jerome, Augustine, Cassian, Paulinus, Facundus[386]. Hilary, on the other hand, declares that the passage is not found “in Graecis et in Latinis codicibus compluribus” (p. 1062 a, Benedictine edition, 1693), a statement which Jerome, who leans much on others in such matters, repeats to the echo. Epiphanius, however, in a passage we have before alluded to (p. 270, note), charges “the orthodox” with removing ἔκλαυσε in ch. xix. 41, though Irenaeus had used it against the Docetae, φοβήθέντες καὶ μὴ νοήσαντες αὐτοῦ τὸ τέλος καὶ τὸ ἰσχυρότατον, καὶ γενόμενος ἐν ἀγωνίᾳ ἵδρωσε, καὶ ἐγένετο ὁ ἱδρὼς αὐτοῦ ὡς θρόμβοι αἵματος, καὶ ὤφθη ἄγγελος ἐνισχύων αὐτόν: Epiphan. Ancor. xxxi[387]. Davidson states [pg 356] that “the Syrians are censured by Photius, the Armenians by Nicon [x], Isaac the Catholic, and others, for expunging the passage” (Bibl. Critic. ii. p. 438).

Of all recent editors, before Westcott and Hort set them within their double brackets, Lachmann alone had doubted the authenticity of the verses, and enclosed them within brackets: but for the accidental presence of the fragment Cod. Q his hard rule—“mathematica recensendi ratio” as Tischendorf terms it—would have forced him to expunge them, unless indeed he judged (which is probably true) that Cod. A makes as much in their favour as against them. So far as the language of Epiphanius is concerned, it does not appear that this passage was rejected by the orthodox as repugnant to their notions of the Lord's Divine character, and such may not have been at all the origin of the variation. We have far more just cause for tracing the removal of the paragraph from its proper place in St. Luke to the practice of the Lectionaries, whose principal lessons (such as those of the Holy Week would be) were certainly settled in the Greek Church as early as the fourth century (see above, Vol. I. pp. 74-7, and notes). I remark with lively thankfulness that my friend Professor Milligan does not disturb these precious verses in his “Words of the New Testament:” and Mr. Hammond concludes that “on the whole there is no reasonable doubt upon the passage.” Thus Canon Cook is surely justified in his strong asseveration that “supporting the whole passage we have an array of authorities which, whether we regard their antiquity or their character for sound judgement, veracity, and accuracy, are scarcely paralleled on any occasion” (Revised Version, p. 103).

17. Luke xxiii. 34. We soon light upon another passage wherein the Procrustean laws of certain eminent editors are irreconcileably at variance with their own Christian feeling and critical instinct. No holy passage has been brought into disrepute on much slighter grounds than this speech of the Lord upon the cross: the words from Ὁ δὲ Ἰησοῦς down to ποιοῦσιν are set within brackets by Lachmann, within double brackets by Westcott and Hort. They are omitted by only BD*, 38, 435, [pg 357] among the manuscripts: by E they are marked with an asterisk (comp. Matt. xvi. 2, 3; ch. xxii. 43,44); of א Tischendorf speaks more cautiously than in the case of ch. xxii. 43, 44, “A [388], and a passage in Arethas of the sixth century. Eusebius assigned the section to his tenth table or canon, as it has no parallel in the other three Gospels. The passage is contained without a vestige of suspicion in אACFGHK (even L) M (hiat P) QSUVΓΔΛΠ, all other cursives (including 1, 33, 69), c e f ff2 l, the Vulgate, all four Syriac versions, all Bohairic codices except the aforenamed two, the Armenian and Ethiopic. The Patristic authorities for it are (as might be anticipated) express, varied, and numerous:—such as Irenaeus and Origen in their Latin versions, the dying words of St. James the Just as cited in Eusebius, Eccl. Hist., lib. ii. cap. 23, after Hegesippus, ἐπὶ τῆς πρώτης τῶν ἀποστόλων γενόμενος διαδοχῆς (Eus.), Hippolytus, the Apostolic Constitutions twice, the Clementine Homilies, Ps.-Tatian, Archelaus with Manes, Eusebius, Athanasius, Gregory of Nyssa, Theodorus of Heraclea, Basil, Ephraem Syrus, Ps.-Ephraem, Ps.-Dionysius Areopagita, Acta Pilati, Syriac Acts of the Apostles, Ps.-Ignatius, Ps.-Justin, Cyril of Alexandria, Eutherius, Anastasius Sinaita, Hesychius, Antiochus Monachus, Andreas of Crete, Ps.-Chrysostom, Ps.-Amphilochius, Opus Imperfectum, Chrysostom often (sometimes loosely enough more suo), Hilary, Ambrose eleven times, Jerome twelve times, Augustine more than sixty times, Theodoret, and John Damascene. Tischendorf adds—valeant quantum—(but only a fraction of this evidence was known to Tischendorf), the apocryphal Acta Pilati[389]. It is almost incredible [pg 358] that acute and learned men should be able to set aside such a silva of witness of every kind, chiefly because D is considered especially weighty in its omissions, and B has to be held up, in practice if not in profession, as virtually almost impeccable. Vain indeed is the apology, “Few verses of the Gospels bear in themselves a surer witness to the truth of what they record than this first of the Words from the Cross; but it need not therefore have belonged originally to the book in which it is now included. We cannot doubt that it comes from an extraneous source” (Hort, Notes, p. 68). Nor can we on our part doubt that the system which entails such consequences is hopelessly self-condemned.

18. John i. 18. ὁ μονογενὴς υἱός, ὁ ὢν εἰς τὸν κόλπον τοῦ πατρός... This passage exhibits in a few ancient documents of high consideration the remarkable variation θεός for υἱός, which however, according to the form of writing universal in the oldest codices (see Vol. I. pp. 15, 50), would require but the change of a single letter, ΥΣ or ΘΣ. In behalf of ΘΣ stand Codd. אBC primâ manu, and L (all wanting the article before μονογενής, and א omitting the ὁ ὤν that follows), 33 alone among cursive manuscripts (but prefixing ὁ to μονογενής, as does a later hand of א), of the versions the Peshitto (not often found in such company), and the margin of the Harkleian (whose affinity with Cod. L is very decided), the Ethiopic, and a host of Fathers, some expressly (e.g. Clement of Alexandria, Didymus “de Trinitate,” Epiphanius, Cyril of Alexandria, &c.), others by apparent reference (e.g. Gregory of Nyssa). The Egyptian versions may have read either θεός or θεοῦ, more probably the latter, as Prebendary Malan translates for the Bohairic[390], the [pg 359] Sahidic being here lost. Their testimonies are elaborately set forth by Tregelles, who strenuously maintains θεός as the true reading, and thinks it much that Arius, though “opposed to the dogma taught,” upholds μονογενὴς θεός. It may be that the term suits that heretic's system better than it does the Catholic doctrine: it certainly does not confute it. For the received reading υἱός we can allege AC (tertiâ manu) EFGHKMSUVXΔΛΠ (D and the other uncials being defective), every cursive manuscript except 33 (including Tregelles' allies 1, 69), all the Latin versions, the Curetonian, Harkleian, and Jerusalem Syriac, the Georgian and Slavonic, the Armenian and Platt's Ethiopic, the Anglo-Saxon and Arabic. The array of Fathers is less imposing, but includes Athanasius (often), Chrysostom, and the Latin writers down from Tertullian. Origen, Eusebius, and some others have both readings. Cyril of Jerusalem quotes without υἱός or θεός,—ὃν ἀνθρώπων μὲν οὐδεὶς ἑώρακεν; ὁ μονογενὴς δὲ μόνος ἐξηγήσατο. C. 7, l. 27, p. 107, ed. Oxon., Pereira.

Tregelles, who seldom notices internal probabilities in his critical notes, here pleads that an ἅπαξ λεγόμενον like μονογενὴς θεός[391] might easily be changed by copyists into the more familiar ὁ μονογενὴς υἱός from John iii. 16; 18; i John iv. 9, and he would therefore apply Bengel's Canon (I. see p. 247). Alford's remark, however, is very sound: “We should be introducing great harshness into the sentence, and a new and [to us moderns] strange term into Scripture, by adopting θεός: a consequence which ought to have no weight whatever where authority is overpowering, but may fairly be weighed where this is not so. The ‘praestat procliviori ardua’ finds in this case a legitimate limit” (N. T., note on John i. 18). Every one indeed must feel θεός to be untrue, even though for the sake of consistency he may be forced to uphold it. Westcott and Hort set μονογενὴς θεός in the text, but concede to ὁ μονογενὴς υἱός a place in their margin.

Those who will resort to “ancient evidence exclusively” for the recension of the text may well be perplexed in dealing with this passage. The oldest manuscripts, versions, and writers are hopelessly divided, so that we can well understand how some critics (not very unreasonably, perhaps, yet without a shadow of authority worth notice) have come to suspect both θεός and [pg 360] υἱός to be accretions or spurious additions to μονογενής. If the principles advocated in Vol. II. Ch. X be true, the present is just such a case as calls for the interposition of the more recent uncial and cursive codices; and when we find that they all, with the single exception of Cod. 33, defend the reading ὁ μονογενὴς υἱός, we feel safe in concluding that for once Codd. אBC and the Peshitto do not approach the autograph of St. John so nearly as Cod. A, the Harkleian Syriac, and Old Latin versions[392].

19. John iii. 13. Westcott and Hort remove from the text to the margin the weighty and doubtless difficult, but on that account only the more certainly genuine, words ὁ ὢν ἐν τῷ οὐρανῷ. Tischendorf rejected them (as indeed does Professor Milligan) in his “Synopsis Evangelica,” 1864, but afterwards repented of his decision. The authorities for omission are אBL (which read μονογενὴς θεός in ch. i. 18) Tb [vi], 33 alone among manuscripts. CDF are defective here: but the clause is contained in AEGHKMSUVΓΔΛΠ, and in all cursives save one, A* and one Evangelistarium (44) omitting ὤν. No versions can be cited against the clause except one manuscript of the Bohairic: it appears in every one else, including the Latin, the four Syriac, the Ethiopic, the Georgian, and the Armenian. There is really no Patristic evidence to set up against it, for it amounts to nothing that the words are not found in the Armenian versions of Ephraem's Exposition of Tatian's Harmony (see Vol. I. p. 59, note 2); that Eusebius might have cited them twice and did not; that Cyril of Alexandria, who alleges them once, passed over them once; that Origen also (in the Latin translation) neglected them once, inasmuch as he quotes them twice, once very expressly. Hippolytus [220] is the prime witness in their behalf, for he draws the theological inference from the passage (ἀποσταλεὶς ἵνα δείξῃ αὐτὸν ἐπὶ γῆς ὄντα εἶναι καὶ ἐν οὐρανῷ), wherein he is followed in two places by Hilary and by Epiphanius. To these add Dionysius of Alexandria [iii], Novatian [iii], Aphraates the Persian, Didymus, Lucifer, Athanasius, Basil, [pg 361] besides Ambrose, Jerome, Augustine, and by John Damascene (thrice), by Cyril of Alexandria, Chrysostom, and Theodoret each four times,—indeed, as Dean Burgon has shown[393], more than fifty passages from thirty-eight ecclesiastical writers; and we then have a consensus of versions and ecclesiastical writers from every part of the Christian world, joining Cod. A and the later manuscripts in convicting אBL, &c., or the common sources from which they were derived, of the deliberate suppression of one of the most mysterious, yet one of the most glorious, glimpses afforded to us in Scripture of the nature of the Saviour, on the side of His Proper Divinity.

20. John V. 3, 4. ἐκδεχομένων τὴν τοῦ ὕδατος κίνησιν. ἄγγελος γὰρ κατὰ καιρὸν κατέβαινεν ἐν τῇ κολυμβήθρᾳ, καὶ ἐτάρασσε τὸ ὕδωρ; ὁ οὖν πρῶτος ἐμβὰς μετὰ τὴν ταραχὴν τοῦ ὕδατος, ὑγιὴς ἐγίνετο, ᾧ δήποτε κατείχετο νοσήματι. This passage is expunged by Tischendorf, Tregelles, Alford, Westcott and Hort, obelized (=) by Griesbach, but retained by Scholz and Lachmann. The evidence against it is certainly very considerable: Codd. אBC*D, 33, 157, 314, but D, 33 contain ἐκδεχομένων ... κίνησιν, which alone A*L, 18 omit. It may be observed that in this part of St. John A and L are much together against N, and against B yet more. The words from ἄγγελος γάρ to νοσήματι are noted with asterisks or obeli (employed without much discrimination) in SΛ, 8, 11?, 14 (ἄγγελος ... ὕδωρ being left out), 21, 24, 32, 36, 145, 161, 166, 230, 262, 269, 299, 348, 408, 507, 512, 575, 606, and Armenian manuscripts. The Harkleian margin marks from ἄγγελος to ὕδωρ with an asterisk, the remainder of the verse with obeli. The whole passage is given, although with that extreme variation in the reading which so often indicates grounds for suspicion[394], in EFGHIKMUVΓΔΠ (with an asterisk throughout), and all known cursives not enumerated above[395]: of these [pg 362] Cod. I [vi] is of the greatest weight. Cod. A contains the whole passage, but down to κίνησιν secundâ manu; Cod. C also the whole, tertiâ manu. Of the versions, Cureton's Syriac, the Sahidic, Schwartze's Bohairic[396], some Armenian manuscripts, f l q of the Old Latin, san. harl.* and two others of the Vulgate (vid. Griesbach) are for omission; the Roman edition of the Ethiopic leaves out what the Harkleian margin obelizes, but the Peshitto and Jerusalem Syriac, all Latin copies not aforenamed, Wilkins' Bohairic, and Armenian editions are for retaining the disputed words. Tertullian clearly recognizes them (“piscinam Bethsaidam angelus interveniens commovebat,” de Baptismo, 5), as do Didymus, Chrysostom, Cyril, Ambrose (twice), Theophylact, and Euthymius. Nonnus [v] does not touch it in his metrical paraphrase.

The first clause (ἐκδεχ ... κίνησιν) can hardly stand in Dr. Scrivener's opinion, in spite of the versions which support it, as DI are the oldest manuscript witnesses in its favour, and it bears much of the appearance of a gloss brought in from the margin. The succeeding verse is harder to deal with[397]; but for the countenance of the versions and the testimony of Tertullian, Cod. A could never resist the joint authority of אBCD, illustrated as they are by the marks of suspicion set in so many later copies. Yet if ver. 4 be indeed but an “insertion to complete that implied in the narrative with reference to the popular belief” (Alford, ad loc.), it is much more in the manner of Cod. D and the Curetonian Syriac, than of Cod. A and the Latin versions; and since these last two are not very often found in unison, and together with the Peshitto, opposed to the other primary documents, it is not very rash to say that when such a conjunction does occur, it proves that the reading was early, widely diffused, and extensively received. Yet, after all, if the passage as it stands in our common text can be maintained as genuine at all, it must be, we apprehend, on the principle suggested above, Vol. I. Chap. I. § 11, p. 18. The chief difficulty, of course, consists [pg 363] in the fact that so many copies are still without the addition, if assumed to be made by the Evangelist himself: nor will this supposition very well account for the wide variations subsisting between the manuscripts which do contain the supplement, both here and in chh. vii. 53-viii. 11[398].

21. John vii. 8. This passage has provoked the “bark” of Porphyry the philosopher, by common consent the most acute and formidable adversary our faith encountered in ancient times [d. 304]. “Iturum se negavit,” as Jerome represents Porphyry's objection, “et fecit quod prius negaverat: latrat Porphyrius, inconstantiae et mutationis accusat.” Yet in the common text, which Lachmann, Westcott and Hort, apparently with Professor Milligan, join in approving, ἐγὼ οὔπω ἀναβαίνω εἰς τὴν ἑορτὴν ταύτην, there is no vestige of levity of purpose on the Lord's part, but rather a gentle intimation that what He would not do then, He would do hereafter. It is plain therefore that Porphyry the foe, and Jerome the defender of the faith, both found in their copies οὐκ, not οὔπω, and this is the reading of Tischendorf and Tregelles: Hort and Westcott set it in their margin. Thus too Epiphanius and Chrysostom in the fourth century, Cyril in the fifth, each of them feeling the difficulty of the passage, and meeting it in his own way. For οὐκ we have the support of א (AC hiant) DKMΠ, 17 secundâ manu, 389: add 507, 570, being Scrivener's pw (two excellent cursives, often found together in vouching for good readings), 558, Evst. 234, the Latin a b c e ff2 l secundâ manu, Cureton's Syriac, the Bohairic, Armenian, and Ethiopic versions[399], a minority of the whole doubtless, yet a goodly band, gathered from east and west alike. In this case no hesitation would have been felt in adopting a reading, not only the harder in itself, but the only one that will explain the history of the passage, had not the palpable and wilful emendation οὔπω been upheld by B: ignoscitur isti, even when it resorts to a subterfuge which in any other manuscript would be put [pg 364] aside with scorn. The change, however, from the end of the third century downwards, was very generally and widely diffused. Besides B and its faithful allies LT, οὔπω is read in EFGHSUVXΓΔΛ, in all cursives not cited above, in f g q, in some Vulgate codices (but in none of the best), the Sahidic, Gothic, and three other Syriac versions, the Harkleian also in its Greek margin. Basil is alleged for the same reading, doubtless not expressly, like the Fathers named above. It is seldom that we can trace so clearly the date and origin of an important corruption which could not be accidental, and it is well to know that no extant authorities, however venerable, are quite exempt from the influence of dishonest zeal.

22. John vii. 53-viii. 11. On no other grounds than those just intimated when discussing ch. v. 3, 4 can this celebrated and important paragraph, the pericope adulterae as it is called, be regarded as a portion of St. John's Gospel. It is absent from too many excellent copies not to have been wanting in some of the very earliest; while the arguments in its favour, internal even more than external, are so powerful, that we can scarcely be brought to think it an unauthorized appendage to the writings of one, who in another of his inspired books deprecated so solemnly the adding to or taking away from the blessed testimony he was commissioned to bear (Apoc. xxii. 18, 19). If ch. xx. 30, 31 show signs of having been the original end of this Gospel, and ch. xxi be a later supplement by the Apostle's own hand, which I think with Dean Alford is evidently the case, why should not St. John have inserted in this second edition both the amplification in ch. v. 3, 4, and this most edifying and eminently Christian narrative? The appended chapter (xxi) would thus be added at once to all copies of the Gospels then in circulation, though a portion of them might well overlook the minuter change in ch. v. 3, 4, or, from obvious though mistaken motives, might hesitate to receive for general use or public reading the history of the woman taken in adultery.

It must be in this way, if at all, that we can assign to the Evangelist chh. vii. 53-viii. 11; on all intelligent principles of mere criticism the passage must needs be abandoned: and such is the conclusion arrived at by all the critical editors. It is entirely omitted (ch. viii. 12 following continuously to ch. vii. 52) [pg 365] in the uncial Codd. אA[400]BCT (all very old authorities) LX[401]Δ, but LΔ leave a void space (like B's in Mark xvi. 9-20) too small to contain the verses (though any space would suffice to intimate the consciousness of some omission), before which Δ* began to write ch. viii. 12 after ch. vii. 52.

Add to these, as omitting the paragraph, the cursives 3, 12, 21, 22, 33, 36, 44, 49, 63 (teste Abbott), 72, 87, 95, 96, 97, 106, 108, 123, 131, 134, 139, 143, 149, 157, 168, 169, 181, 186, 194, 195, 210, 213, 228, 249, 250, 253, 255, 261, 269, 314, 331, 388, 392, 401, 416, 453, 473 (with an explanatory note), 486, 510, 550, 559, 561, 582 (in ver. 12 πάλαι for πάλιν): it is absent in the first, added by a second hand in 9, 15, 105, 179, 232, 284, 353, 509, 625: while ch. viii. 3-11 is wanting in 77, 242, 324 (sixty-two cursive copies). The passage is noted by an asterisk or obelus or other mark in Codd. MS, 4, 8, 14, 18, 24, 34 (with an explanatory note), 35, 83, 109, 125, 141, 148 (secundâ manu), 156, 161, 166, 167, 178, 179, 189, 196, 198, 201, 202, 219, 226, 230, 231 (secundâ manu), 241, 246, 271, 274, 277, 284?, 285, 338, 348, 360, 361, 363, 376, 391 (secundâ manu), 394, 407, 408, 413 (a row of commas), 422, 436, 518 (secundâ manu), 534, 542, 549, 568, 575, 600. There are thus noted vers. 2-11 in E, 606: vers. 3-11 in Π (hiat ver. 6), 128, 137, 147: vers. 4-11 in 212 (with unique rubrical directions) and 355: with explanatory scholia appended in 164, 215, 262[402] (sixty-one cursives). Speaking generally, copies which contain a commentary omit the paragraph, but Codd. 59-66, 503, 526, 536 are exceptions to this practice. Scholz, who has taken unusual pains in the examination of this [pg 366] question, enumerates 290 cursives, others since his time forty-one more, which contain the paragraph with no trace of suspicion, as do the uncials DF (partly defective) GHKUΓ (with a hiatus after στήσαντες αὐτήν ver. 3): to which add Cod. 736 (see addenda) and the recovered Cod. 64, for which Mill on ver. 2 cited Cod. 63 in error. Cod. 145 has it only secundâ manu, with a note that from ch. viii. 3 τοῦτο τὸ κεφάλαιον ἐν πολλοῖς ἀντιγράφοις οὐ κεῖται. The obelized Cod. 422 at the same place has in the margin by a more recent hand ἐν τήσιν ἀντιγράφης οὕτως. Codd. 1, 19, 20, 129, 135, 207[403], 215, 301, 347, 478, 604, 629, Evst. 86 contain the whole pericope at the end of the Gospel. Of these, Cod. 1 in a scholium pleads its absence ὡς ἐν τοῖς πλείοσιν ἀντιγράφοις, and from the commentaries of Chrysostom, Cyril of Alexandria, and Theodore of Mopsuestia; while 135, 301 confess they found it ἐν ἀρχαίοις ἀντιγράφοις: Codd. 20, 215, 559 are obelized at the end of the section, and have a scholium which runs in the text τὰ ὠβελισμένα, κείμενα δὲ εἰς τὸ τέλος, ἐκ τῶνδε ὧδε τὴν ἀκολουθίαν ἔχει, and on the back of the last leaf of both copies τὸ ὑπέρβατον τὸ ὄπισθεν ζητούμενον. In Codd. 37, 102, 105, ch. viii. 3-11 alone is put at the end of the Gospel, which is all that 259 supplies, though its omission in the text begins at ch. vii. 53. Cod. 237, on the contrary, omits only from ch. viii. 3, but at the end inserts the whole passage from ch. vii. 53: in Cod. 478, ch. vii. 53-viii. 2 stands primâ manu with an asterisk, the rest later. Cod. 225 sets chh. vii. 53-viii. 11 after ch. vii. 36; in Cod. 115, ch. viii. 12 is inserted between ch. vii. 52 and 53, and repeated again in its proper place. Finally, Codd. 13, 69, 124, 346 (being Abbott's group), and 556 give the whole passage at the end of Luke xxi, the order being apparently suggested from comparing Luke xxi. 37 with John viii. 1; and ὤρθριζε Luke xxi. 38 with ὄρθρου John viii. 2[404]. In the Lectionaries, as we have had occasion to state before (Vol. I. p. 81, note), this section was never read as a part [pg 367] of the lesson for Pentecost (John vii. 37-viii. 12), but was reserved for the festivals of such saints as Theodora Sept. 18, or Pelagia Oct. 8 (see Vol. I. p. 87, notes 2 and 3), as also in Codd. 547, 604, and in many Service-books, whose Menology was not very full (e.g. 150, 189, 257, 259), it would thus be omitted altogether. Accordingly, in that remarkable Lectionary, the Jerusalem Syriac, the lesson for Pentecost ends at ch. viii. 2, the other verses (3-11) being assigned to St. Euphemia's day (Sept. 16).

Of the other versions, the paragraph is entirely omitted in the true Peshitto (being however inserted in printed books with the circumstances before stated under that version), in Cureton's Syriac, and in the Harkleian; though it appears in the Codex Barsalibaei, from which White appended it to the end of St. John: a Syriac note in this copy states that it does not belong to the Philoxenian, but was translated in a.d. 622 by Maras, Bishop of Amida. Maras, however, lived about a.d. 520, and a fragment of a very different version of the section, bearing his name, is cited by Assemani (Biblioth. Orient, ii. 53) from the writings of Barsalibi himself (Cod. Clem.-Vat. Syr. 16). Ridley's text bears much resemblance to that of de Dieu, as does a fourth version of ch. vii. 53-viii. 11 found by Adler (N. T. Version. Syr., p. 57) in a Paris codex, with the marginal annotation that this “σύνταξις” is not in all the copies, but was interpreted into Syriac by the Abbot Mar Paulus. Of the other versions it is not found in the Sahidic, or in some of Wilkins' and all Schwartze's Bohairic copies[405], in the Gothic, Zohrab's Armenian from six ancient codices (but five very recent ones and Uscan's edition contain it), or in a f l (text) q of the Old Latin. In b the whole text from ch. vii. 44 to viii. 12 has been wilfully erased, but the passage is found in c e (we have given them at large, pp. 362-3), ff2 g j l (margin), the Vulgate (even am. fuld. for. san.), Ethiopic, Slavonic, Anglo-Saxon, Persic (but in a Vatican codex placed in ch. x), and Arabic.

Of the Fathers, Euthymius [xii], the first among the Greeks to mention the paragraph in its proper place, declares that παρὰ τοῖς ἀκριβέσιν ἀντιγράφοις ἢ οὐχ εὕρηται ἢ ὠβέλισται; διὸ φαίνονται παρέγγραπτα καὶ προσθήκη. The Apostolic Constitutions [iii or iv] had plainly alluded to it, and Eusebius (Hist. Eccl. iii. 39. fin.) had described from Papias, and as contained in the Gospel of the Hebrews, the story of a woman ἐπὶ πολλαῖς ἁμαρτίαις διαβληθείσης ἐπὶ τοῦ κυρίου, but did not at all regard it as Scripture. Codd. KM too are the earliest which raise the number of τίτλοι or larger κεφάλαια in St. John from 18 to 19, by interpolating κεφ. ι´ περὶ τῆς μοιχαλίδος, which soon found admittance into the mass of copies: e.g. Evan. 482.

Among the Latins, as being in their old version, the narrative was more generally received for St. John's. Jerome testifies that it was found in his time “in multis et Graecis et Latinis codicibus;” Ambrose cites it, and Augustine (de adult. conjugiis, lib. ii. c. 7) complains that “nonnulli modicae fidei, vel potius inimici verae fidei,” removed it from their codices, “credo metuentes peccandi impunitatem dari mulieribus suis[406].”

When to all these sources of doubt, and to so many hostile authorities, is added the fact that in no portion of the N. T. do the variations of manuscripts (of D beyond all the rest) and of other documents bear any sort of proportion, whether in number or extent, to those in these twelve verses (of which statement full evidence may be seen in any collection of various readings)[407], we cannot help admitting that if this section be indeed the composition of St. John, it has been transmitted to us under circumstances widely different from those connected with any other genuine passage of Scripture whatever[408].