CRITICAL NOTES ON VARIOUS PASSAGES OF THE NEW TESTAMENT.

THE GOSPEL.

Matthew.

With regard to the title, Westcott and Hort say (Introduction, § 423, p. 321): “In prefixing the name ΕΥΑΓΓΕΛΙΟΝ in the singular to the quaternion of ‘the Gospels,’ we have wished to supply the antecedent which alone gives an adequate sense to the preposition ΚΑΤΑ in the several titles. The idea, if not the name, of a collective ‘Gospel’ is implied throughout the well-known passage in the third book of Irenæus, who doubtless received it from earlier generations. It evidently preceded and produced the commoner usage by which the term Gospel denotes a single written representation of the one fundamental Gospel.” Compare Zahn, GK., i. 106 ff.; Einleitung, ii. 172 ff., 178 f.: “Of recent editors, Westcott and Hort have most faithfully interpreted the original idea by setting Εὐαγγέλιον on the fly-leaf, and κατὰ Μαθθαῖον, etc., over the separate books.” I have followed the same principle in the Table of Contents prefixed to the Stuttgart edition of the New Testament. Compare above, pp. 164, 165. On the spelling Μαθθαῖος, instead of Ματθαῖος, compare on the one hand the LXX. manuscripts, which exhibit the forms Μαθανιά, Μαθθανιά, Ματθανιά; Ματταθίας, Ματθαθίας, Μαθθαθίας (see Supplement I. to Hatch and Redpath’s Concordance to the 248Septuagint), and on the other, Blass’s Grammatik des neutestamentlichen Griechisch, § 3, 11 (English Trans. by Thackeray, 1898, p. 11).

i. 16. There are three forms of the text here—

(1) Ἰωσὴφ τὸν ἄνδρα Μαρίας, ἐξ ἧς ἐγεννήθη Ἰησοῦς ὁ λεγόμενος Χριστός: all our Greek uncials and almost all the minuscules.

(2) Ἰωσήφ, ᾧ μνηστευθεῖσα παρθένος Μαρία ἐγέννησεν τὸν Ἰησοῦν (τὸν λεγόμενον) Χριστόν: most of the Old Latin (a d g₁ k q, with b c similarly), Curetonian Syriac, Armenian, and four minuscules—viz., 346, 556, 624, 626, with slight divergencies.

(3) Ἰωσήφ· Ἰωσὴφ δὲ, ᾧ μνηστευθεῖσα (or μεμνηστευμένη?) ἦν παρθένος Μαρία, ἐγέννησεν τὸν Ἰησοῦν Χριστόν: the form underlying the newly-discovered Sinai-Syriac.^[265]

These readings are discussed in the “Additional Note” to Notes on Select Readings, Westcott and Hort, Introduction (1896), p. 140 ff. Reading (2) is dismissed on external grounds as displaying the characteristic features of the “Western” type of text. Reading (3) is regarded as independent of (2), neither confirming it nor confirmed by it. Taken therefore on its own merits, it must yield to the received text (1), as it is easier to suppose that (3) is derived from (1) than vice versa.

Zahn goes fully into these various forms (Einleitung, ii. 291-293). He begins by saying that it is impossible, except on a very loose view of the facts, to conclude that the Sinai-Syriac here preserves the original text, which was gradually displaced for dogmatic reasons by the modified form presented in (2), and ultimately by that given in (1). On the contrary, the Curetonian-Syriac preserves an early form of text, and one that had a pretty wide circulation, so that it cannot be due to an orthodox alteration of the Sinai-Syriac. 249“If it be the case that the latter, like the former, is derived from a Greek original, and that these two earliest versions of the ‘Distinct’ Gospel are not independent of each other but are two recensions of a single version, then it follows that the recension which agrees exactly with a demonstrably old Greek text (in this case the Curetonian Syriac) preserves the original form of the Syriac version; while, on the other hand, the one which deviates from all the Greek, Latin, and other forms of the transmitted text (in this case the Sinai-Syriac) is derived from the other by a process of intentional alteration.” There would be nothing to object to this reasoning were it not that, as it seems to me, there is a flaw in the second of the premises stated above, which of course vitiates the conclusion. In the main, it is true that the Sinai-Syriac and the Curetonian are not independent, but two recensions of a single version, but their common original was, as Zahn himself was the first to suggest, Tatian’s Diatessaron, which did not contain the first chapter of Matthew’s Gospel. So that the Sinai-Syriac may also go back to a Greek text (such as has been discovered in the Dialogue of Timothy and Aquila, see above, p. 99), and be earlier than the Curetonian.

Zahn concludes his examination of this passage by saying: “We may give up all hope of finding in early manuscripts and versions any indication that Joseph was regarded as the natural father of Jesus by the writers of lost Gospels which may have been employed in the composition of the canonical Matthew and Luke. A writer like Matthew, whose purpose was to silence the calumnies raised against the miraculous birth of the Messiah, and who knew how to utilise the smallest details of an intractable genealogy to this end, cannot at the same time have accepted in his narrative statements directly contradicting his view of that occurrence. Any text of Matthew’s Gospel containing such features would be pre-condemned as one that had been tampered with in a manner contrary to the conception of the author.”

i. 18. The reading γένεσις is now supported by the newly-discovered 250Oxyrhynchus Papyrus. It was adopted in the text by Vignon (Geneva, 1574). Origen knew no other reading than γέννησις, which is also attested by L (Codex D is defective here). Westcott and Hort have accordingly given it a place in their Appendix. Weiss explains it as an alteration made in conformity with the verbal forms ἐγέννησε, ἐγεννήθη, occurring in the previous part of the chapter. Zahn (Einleitung, ii. 270, 289) thinks it is probably original. The two oldest and the latest Syriac have a different word here from that in i. 1. These agree with Irenæus in the omission of Ἰησοῦ. Zahn thinks this is probably correct.

i. 25. On πρωτότοκον, see above, p. 166, and the Oxford Debate, p. 4 ff.

v. 25. On ἀντίδικος = בעלדינא, see Lagarde, De Novo, 20 (Ges. Abhdl., 188); quem Matthaei locum quum imitaretur et rideret Lucianus in Navigio 35, ἀντίδικος non ferebat: ἕως ἔτι καθ’ ὁδόν εἰσιν οἱ πολέμιοι, ἐπιχειρῶμεν αὐτοῖς.

vi. 1. δικαιοσύνην, א* B D Syr^sin: ἐλεημοσύνην, most authorities: δόσιν א^a: “your gifts,” Syr^cu. Zahn (Einl., ii. 311) asks whether these variants may not go back to a time when the Aramaic Gospel was interpreted orally in these different ways? The agreement exhibited between א^a and Syr^cu is particularly strange.

vi. 13. There is a considerable amount of unanimity now with regard to the doxology which used to be so much discussed. Among the witnesses supporting its insertion are Syr^cu, which, however, omits καὶ ἡ δύναμις, and the Sahidic, which omits καὶ ἡ δόξα. Syr^sin is unfortunately lost here. In addition to the testimony previously known for the insertion of the Doxology, there is now that of the Teaching of the Apostles, one of the earliest Church writings. But the very fact that the Teaching is a Church work reveals the source of the Doxology—viz. liturgical use. The Conclusion was early added in Church worship from Old Testament analogies; in the First Gospel it is out of place. The Greek manuscripts from which Jerome made his version knew nothing of it, and 251accordingly the Catholic Church omits it to this day. Luther also passed it over in his Catechism, in which the exposition of the Conclusion is limited to the word “Amen,” and says, “it is added that I may have the assurance that my prayer will be heard.” In the Greek Church the Amen was explained as equivalent to γένοιτο, “so may it be.”

viii. 7. Fritzsche (1826) took this verse as a question of surprise. This view has been renewed by Zahn (Einl., ii. 307).

viii. 24. The words “erat enim ventus contrarius eis,” which are found in one manuscript of the Vulgate in W-W after “mari,” and in four after “fluctibus,” are an interpolation from Mark vi. 48. Tischendorf cites two Greek minuscules in support of it. Lagarde’s Vienna Arabic manuscript (see p. 143) mentions it as an addition of the “Roman” version.

xi. 19. Schlottman and Lagarde explain the variation between ἔργα and τέκνα as a confusion of the Aramaic עַבְדָּא (servant: παῖς) and עְַבָדָא (work). See Zahn, Einl., ii. 311 f., and compare also Salmon, Some Thoughts etc., p. 121 f. It is still to be shown, however, that τέκνα is ever used as the equivalent of עַבְדָּא. Hilgenfeld (ZfwTh., 42. 4, p. 629) refers to 4 Esdras vii. 64 (134), where the Latin and the first Arabic version read “quasi suis operibus,” the Ethiopic “quasi filiis suis,” and the Syriac “quia servi eius sumus.”

xii. 36. See on xviii. 7.

xiii. 35. διὰ Ἠσαΐου τοῦ προφήτου is now attested only by א*, two members of the Ferrar group, and some other minuscules, but Eusebius and Jerome found it in several manuscripts, and it was used still earlier by Porphyrius as a proof of Matthew’s ignorance. It is certainly, therefore, genuine, although it is omitted by Syr^sin, Syr^cu, by the “accurate” manuscripts according to Eusebius,^[266] and by the “vulgata editio” according to Jerome. The conjecture of the latter, that Ἀσάφ was the original reading, which was changed to Ἠσαΐου by some unintelligent copyist and then dropped as incorrect, only serves to show what sort of ideas 252he had with regard to textual criticism. The assertion of the Breviarium in Psalmos, p. 59 f., that all the old manuscripts read “in Asaph propheta” is pure fiction. Compare Ἰερεμίου in Matt. xxvii. 9, where one would expect Ζαχαρίου, and where we find that Ἰερεμίου is omitted by some witnesses and replaced in others by Ζαχαρίου or “Esaiam.” “Esaiam” has also crept into the Vulgate manuscript rus (W-W’s R). On the insertion, omission, and interchange of such names, see W-H’s discussion of this passage, and the “Supplementary Note” by Burkitt on Syr^sin in the edition of 1896, p. 143. For an interesting exchange of names (Jonah and Nahum), see Tobit, xiv. 8. Asaph is called ὁ προφήτης in 2 Chron. xxix. 30. Compare Zahn, Einl., ii. 313 f. Weiss⁹ would omit the word on the ground of insufficient testimony as being simply introduced from iii. 3, iv. 14, viii. 17, and xii. 17.

xiv. 3. Zahn (Einl., ii. 309) thinks it extremely improbable that D and certain important Latin witnesses should have removed the (wrong) name, Philip, from this passage on the ground of their better knowledge, while allowing it to stand without exception in Mark vi. 17. He believes rather that they have preserved the original text, and that Φιλίππου is here an interpolation from the passage in Mark. Weiss⁹, on the other hand, sees no reason why it should be either bracketed or omitted. The possibility of its being inserted is shown by the fact that it also crept into six or seven manuscripts of Jerome, collated by W-W. This is one of the passages where Tischendorf in his seventh edition frankly preferred Codex D to all the other Greek witnesses.

xv. 4b. For θανάτῳ τελευτάτω, Syr^cu has נתקטל, evidently in accordance with Exod. xxi. 17. In the Arabic Diatessaron (§ 20, 23) the second half of this verse seems to be replaced by Mark vii. 10b. After “morte moriatur” in this passage, Ephraem adds “et qui blasphemat Deum crucifigatur,” which Zahn (Forsch., i. 157) thinks he must have found in his original. This apocryphal addition, which has no other testimony than that of Ephraem, does not seem to Zahn like 253a passage that had been afterwards removed from the text of the New Testament with complete success (Forsch., i. 241). The correct explanation of the words is given by Harris: they are the Peshitto rendering of Deut. xxi. 23. Compare Driver’s Deuteronomy on the passage, and the reference there made to Lightfoot’s Galatians (Extended Note on iii. 13, ninth edition, p. 152 f.). Symmachus also renders the words: “propter blasphemiam Dei suspensus est,” while Onkelos says על דחב קרם יי אצטליב, and Siphre מביגי שקלל את השם. This should be noted in connection with Matt. xxvi. 65, and still more so with John xix. 7. The only passage usually cited there is Levit. xxiv. 16, according to which Jesus should have been stoned. Our commentators pass too hastily over the question why the Jews insisted on crucifixion instead of stoning.

xvi. 18b, 19. So far as the criticism of the text is concerned, there is no occasion for entering on the discussion whether this passage, like the one resembling it in xviii. 15-18, is original or not. There may, however, be cases in which one cannot overlook the fact that where the “lower” criticism ends the “higher” begins. Compare, on the one side, Zahn, Forsch., i. 244 ff., and on the other, Resch, Logia, p. 55; Paralleltexte, ii. 187-196, 441.

xvi. 22. The peculiar reading, “compatiens,” which is found in the Arabic Tatian (J. H. Hill, p. 137, § 23. 42: Zahn, GK., ii. 546), and which Sellin has also traced in Ephraem, is now explained by the Syr^sin of Mark viii. 32: see my note in Lewis, Some Pages, p. xiii. The very same play upon the words חוס, “to pity,” and חס, “to be far from,” is found as late as in the Histoire de Mar-Jabalaha, de trois autres patriarches, ed. Bedjan, 1895, p. 407, line 14; p. 408, line 4. For a moment I thought of ὀργισθεὶς and σπλαγχνισθείς in Mark i. 41.

xviii. 7. The Dictum Agraphum τὰ ἀγαθὰ ἐλθεῖν δεῖ, μακάριος δὲ δι’ οὗ ἔρχεται, which, according to the Clementine Homilies (xii. 29), ὁ τῆς ἀληθείας προφήτης ἔφη, was known 254also to Ephraem (cf. Zahn, Forsch., i. 241 f. on § 50. 4). An exact parallel to this “harmless expansion of the canonical text” is seen in the form which Matt. xii. 36 assumed in Codex C of the Palestinian Syriac Evangeliarium: that “for every good word that men do not speak they shall give account” (see Lewis, In the Shadow of Sinai (1898), pp. 256-261; and thereon, ThLz., 1899, col. 177).

xviii. 20. On the form in which this saying is found in the Oxyrhynchus Logia, compare Ephraem (Moesinger 165), “ubi unus est ibi et ego sum.” Zahn believes that Ephraem found this in his text, but that Aphraates, who also has it, arrived at it by way of a “spiritual interpretation” of the canonical words. After quoting the comments of Aphraates on these words, Zahn says: “It appears certain, therefore, that Aphraates did not find in his text the apocryphal sentence given in Ephraem, but by way of interpretation reached the same thought that Ephraem found in his text as a word of comfort spoken by Jesus to the lonely. (Ephraem introduces the saying with the words: ‘He comforted them in His saying.’) The interpretation, which may not have been original in Aphraates, became first a gloss and then part of the text of Tatian’s Harmony.” This should be noticed in connection with the Oxyrhynchus Logion. See Burkitt in the Introduction to Barnard’s Biblical Text of Clement (Texts and Studies, v. 5, p. xiv).

xx. 13. The peculiar form of the householder’s reply given in Syr^cu, μὴ ἀδίκει με (Baethgen, μή μοι κόπους πάρεχε) is ignored by Tischendorf. Our commentators also err in not taking note of the variant συνεφώνησά σοι for συνεφώνησάς μοι. Compare the similar variation in John viii. 57; also Luke xviii. 20, τὰς ἐντολὰς οἶδα, read by the Marcionites instead of οἶδας; and Ephes. v. 14, ἐπιψαύσεις τοῦ Χριστοῦ, derived through a presupposed reading, ἐπιψαύσει σοι ὁ Χριστός. Συνεφώνησά σοι in Matt. xx. 13 is also attested by Syr^sin, which agrees with the common text in the first member of the verse. It is also found in the newly-discovered purple 255manuscript in Paris. The Arabic Tatian agrees with the usual text in both members. On the strange mixture of this verse and Luke xvi. 25 in Petrus Siculus (ἑταῖρε, οὐκ ἀδικῶ σε· ἀπέλαβες τὰ σὰ ἐν τῇ ζωῇ σου· νῦν ἆρον τὸ σὸν καὶ ὕπαγε) see Zahn, GK., ii. 445.

xx. 16. The concluding member of this verse is now rightly omitted with א B L Z and the Egyptian versions. All the Syriac versions have it, including the newly-discovered Syr^sin. It is worth observing that the verse with this addition forms the close of a lection in Syr^hier.

xx. 28. Westcott and Hort devote one of their “Notes on Select Readings” to the addition to this verse, and in the edition of 1896 Burkitt adds that it cannot have stood in Syr^sin, because there was not room for it on the leaf that is missing between Matt. xx. 24 and xxi. 20. According to W-H the passage is Western, being attested by D Φ among the Greek manuscripts and by the Latin and Syriac versions. “The first part only, ὑμεῖς—εἶναι, is preserved in m, ger, and apparently Leo, who quotes no more; the second part only, εἰσερχόμενοι—χρήσιμον, in ger₂ and apparently Hilary. The first part must come from an independent source, written or oral; the second probably comes from the same, but it is in substance identical with Luke xiv. 8-10.” Tischendorf states that of the Old Latin, four (f g₂ l q) omit the section, which, however, is found in c d e ff_{1, 2} g₁ h (m) n, two manuscripts of the Vulgate (and. emm.), the Old German, and the Saxon. To these W-W add also the Old Latin r, two manuscripts of the Vulgate not usually employed by them, and, of those forming the basis of their edition, H^mg Θ O—i.e. the Theodulfian Recension. A hand of the tenth century has written on the margin of O, “mirum unde istud additum: cum Lucas parabolam de invitatis ad nuptias et primos accubitus eligentibus decimo canone, ubi M(atthaeu)s sua non communia dicit referat.” This resembles the marginal note attached to the passage by Thomas of Heraclea (not given by Jos. White, but by Adler, from Cod. Assem., 1): 256Haec quidem in exemplis antiquis in Luca tantum leguntur capite 53: inveniuntur autem in exemplis graecis^[267] hoc loco: quapropter hic etiam a nobis adiecta sunt.

The word δειπνοκλήτωρ, which Resch took from this passage into the text of his Logia Jesu, for ὁ σὲ καὶ αὐτὸν καλέσας, found in Luke xiv. 9, should itself have provoked investigation. The only Latin witnesses which render it in a substantive form are d, which has coenae invitator both times, and m, which has invitator the first time. The others give it as a relative clause (qui vocavit, invitavit), so that they may have read it in the form in which it stands in our present text of Luke.^[268] It is impossible not to believe that some connection exists between these substantive expressions and the Syriac מרא אהשמיתא, “master of the feast,” which is found in Syr^cu and Syr^sin, and is also given by Aphraates, for τῷ κεκληκότι αὐτόν in Luke xiv. 12 (Aphr. 388, 12-19; Zahn, Forsch., i. 85, note). Syr^cu has it both times in this passage of Matthew.^[269]

Bengel, like our modern expositors, says nothing of the interpolation in his Gnomon, and his view with respect to it 257has, therefore, to be gathered from his apparatus. “Interjicit cod. Lat. vetustissimus Vos autem, etc. ... Vid. Rich. Simon, Obs. Nouv., p. 31. Et sic fere Cant. (i.e., D) cuius lectio passim exstat. Idem vero Codex Graeca sua ad Latina haec, quae modo exscripsimus, confecit: Latina autem sua, sub manu, vehementius interpolavit, magno argumento licentiae suae. Eandem periocham legit Juvencus, Hilarius: habentque praeterea codd. Lat. aliquot, et inde Sax. Ex. Luc. xiv. 8 f., interveniente forsan Evangelio Nazaraeorum ... Priorem duntaxat partem, ‘Vos autem ... minui’ habet alius cod. Lat. antiquiss. ut si Librarius, cum describere coepisset, non scribendum agnosceret: eandemque Leo M. sic exhibet. Et tamen ... porro ab hoc loco ad Luc. xxii. 28, verbum crescendi protulit Cant. coenaeque invitator ei dicitur δειπνοκλήτωρ.”

The truth is, of course, the very opposite of this, as is shown by the indicative quaeritis and the imperative of the Syriac, which are both derived from the ambiguous ζητεῖτε. There cannot be the slightest doubt of this, seeing that the discovery of Codex Beratinus (Φ) has added a second Greek witness in support of the interpolation. It reads ἐλάττων (cf. minor, c), omits the καὶ before ἐπέλθῃ just as m does with et, has ἄγε in place of σύναγε (accede: d, collige), and the comparative χρησιμώτερον (utilius) for the positive read by D d. The word δειπνοκλήτωρ also occurs in Φ.^[270] It is not found in Bekker’s Pollux or in Schmid’s Hesychius, and the only instance that ancient lexicons are able to cite for its usage is that of Athenæus, who observes (4. 171 B) that Artemidorus calls the ἐλέατρος by that name. The note appended in Hase-Dindorf’s Stephanus was not correct at the time of its publication: Quidam codices Matt. xx. 27, Hesych., Wakef. Eust. Od., p. 1413, 3; nor the quotation from Ducange: Δ. in Lex. MS. Cyrilli exp. ἑστιάτωρ. In the same work δειπνοκλητόριον is cited from Eust., Il., 766, 58, and as an explanation 258of ἑστιατόριον from the Lex. MS. Cyrilli. The word therefore belongs to the later popular language. The question is whether it may not also belong to the vocabulary of Tatian. Moreover, it reminds us of the equally rare word κτήτωρ in Acts iv. 34.

On the occurrence of the passage in Tatian, see Zahn, Forsch., i. 85, 179. On the questions connected with its interpolation see Chase and p. 216 above.

xxii. 23. We have in this verse an illustration of the difference caused by the insertion or omission of the article. If we read οἱ λέγοντες with א^c E F G etc., then the words introduce the creed of the Sadducees (“who say,” Weizsäcker: “members of that sect who deny the resurrection,” Stage); if we omit οἱ with א* B D and Syr^sin, we have then what they actually said to Jesus. But as this would be the only place where Matthew gave an explanation of this sort regarding Jewish affairs, the article should be omitted. See note in loco, Expositor’s Greek Testament, and compare the margin of the Revised English Version.

xxiii. 35. א¹ omits υἱοῦ βαραχίου, which is replaced in the Gospels of the Hebrews by “filium Joiadae.” Zahn (Einl., ii. 308) refers to the view of Hug, adopted by Eichhorn and many others, that the author, or redactor, or translator of Matthew made this Zechariah, who is rightly called the son of Jehoiada in the Gospel of the Hebrews, the son of Barachiah in order to identify him with the Zechariah, son of Baruch, who was murdered by the Zealots (Josephus, Bell., iv. 5. 4). He points out that this would involve a prediction on the part of Jesus, and that, moreover, the scene of the murder is different in the two cases: that the locality in Matt. points to 2 Chron. xxiv. 21, and that Matthew’s mistake in calling him the son of Barachiah is due to a confusion with the Zechariah mentioned in Isa. viii. 2, or that in Zech. i. 1. It should be observed, however, that Lucian alone calls the murdered person in Chronicles by the name of Zechariah; the LXX calls him Azariah.

xxv. 41. See on Luke xx. 35.

259xxvi. 73. Ὁμοιάζει was formerly attested by D alone, but has now the further support of Syr^sin. The clause καὶ ἡ λαλιά σου ὁμοιάζει has crept into a great number of manuscripts, including even A, in Mark xiv. 70. There Tischendorf remarks, “Omnino e Mt. fluxit,” in which he is quite right. But he is wrong when he says “ipsum ὁμοιάζει glossatoris est.” Because the glossator must then have been earlier than Tatian (Ciasca, p. 87), and the parent of all those manuscripts. The converse is the truth—viz., that D alone preserves the original reading, and that δῆλόν σε ποιεῖ is the voice of the διορθωτής.

xxvii. 9. The name of the prophet, which was omitted in some manuscripts, according to Augustine, is now omitted only by a b and the two minuscules 33 and 157. Augustine also observes that Matthew himself would have noticed his mistake or had his attention called to it by others. On this compare my notes on ἐβαρύνατε in Acts iii. 14, which I have explained by supposing that the author read כברתם or כבדתם instead of כפרתם (Philologica Sacra, p. 40; above, p. 170). Origen, Eusebius, and Jerome evidently still found Ἰερεμίου in all the manuscripts. Ζαχαρίου is supplied only by 22 and Esaiam by 1. See on Matt. xiii. 35, and compare Expository Times, November 1900, p. 62.

xxvii. 16. Zahn (Einl., ii. 294) points out that Origen also found Jesus given as the prenomen of Barabbas “in very ancient manuscripts,” but that in all probability Tatian did not have it, seeing that Bar-Bahlul cites it expressly as the reading of the “Distinct” (i.e., not harmonised) Gospel. Jerome says that in the Gospel of the Hebrews he was called by a name meaning “filius magistri eorum,” so that he must have been thinking not of Bar-’abbam but of Bar-rabbam.

xxvii. 49. See above, p. 227, and compare Burkitt, Texts and Studies, v. 5, p. xix.

xxviii. 18. Compare Dan. vii. 14b (LXX), καὶ ἐδόθη αὐτῷ ἐξουσία, and also Dan. vii. 13 (= Matt. xxvi. 64), vii. 14 f. (= Matt. xxviii. 18). See the English Revised Version with marginal References (Oxford, 1899).

260According to the subscriptions found in various minuscules, the Hebrew Matthew was translated into Greek “by John,” or “by James,” to which some add “the Brother of the Lord,” or “by Bartholomew, the celebrated Apostle (πανευφήμου), but as others say by John the Theologian, οἳ καὶ ἀληθῶς εἰρήκασιν.” See Tischendorf, and Zahn, Einl., ii. 267.

Mark.

As if to enforce the desire to which I have given expression above (p. 246), there has come into my hands Blass’s Textkritische Bemerkungen zu Markus. If the statements contained in the introductory remarks are correct, and scarcely any other view is possible in the circumstances described, then the textual criticism of the first and second Gospels is a hopeless matter. “An evangelist or teacher who obtained possession of the originally anonymous Commentarius could not feel bound to respect the external form, but considered himself justified in correcting it if it seemed to him to be defective, and even felt called to correct or complete its subject matter.” Blass reminds us that we have whole classes of documents, legends of saints e.g., which were treated with the utmost possible freedom by the copyists, who in fact were in this case editors and revisers. But he says that no one has treated Mark quite so drastically as all this. His summing up of the matter is, that the critic can often do no more than recognise and admit the early multiplicity, and that in such a case it were best to print the text in parallel columns. At the same time he is able to distinguish some of the variants as later falsifications or corruptions. Universally trustworthy authorities there are none; here one group is right, there another, and we no sooner give them credence than they mislead us with some fresh error.

We are far removed, truly, from the confidence displayed by Tischendorf in the treatise he published shortly before his death in 1873 in answer to the question, “Have we the genuine 261text of the Evangelical and Apostolical writings?” All the more urgently, therefore, do we need fresh studies in textual criticism, and their appearance in Germany is the more gratifying on that account. The Markus-Studien of Dr. H. P. Chajes (Berlin, 1899), however, are quite beside the point. They are purely imaginary, having neither substance nor method.

i. 1. On the title, see above; Zahn, Einl., ii. 220 ff., 235; Swete, in loco; and on this last, S. D. F. Salmond, in the Critical Review, April 1899, 206 f.: “We do not see, however, why Professor Swete should regard the opening verses as probably not a part of the original work. One might say the same of the whole paragraph with which the Gospel opens, or, for that matter, the whole chapter. The documentary evidence is substantially the same in each case, and the internal considerations are much too indeterminate.” It may be pointed out, as remotely analogous to this, that before Matt. i. 18 the margin of harl (Z in W-W) contains a note in a hand of the ninth or tenth century to the effect, “genealogia hucusque: incipit evangelium secundum Matthaeum,” while Y has the words “incipit evangelium secundum Matthaeum” in the text, and eight manuscripts begin verse 18 with capital or red letters. Compare Scrivener, I. c. iii., on the divisions of the text in B and other manuscripts.

For the way in which the opening sentences are to be construed, reference must be made to the commentaries. It may be said here, however, that parallels may be cited from the New Testament for each of the three possible constructions. These are (1) Ἀρχὴ ..., καθὼς ... αὐτοῦ, ἐγένετο; (2) Ἀρχὴ ... Καθὼς ... αὐτοῦ. Ἐγένετο; (3) Ἀρχὴ ... Καθὼς ... αὐτοῦ, ἐγένετο. For (1) and (2) compare Luke iii. 1 ff., and for (3) 1 Tim. i. 1 ff. Origen favours the first construction (Contra Celsum, ii. 4; vol. i. p. 131). As regards the text it need only be said that καὶ is read before ἐγένετο (v. 4) by א¹, and that δὲ is found after it, not only in the Coptic, but also in Syr^hier.

i. 2. Origen here read ἐγὼ and ἔμπροσθέν σου (i. 131). But 262the former should be omitted with B D etc., and the latter with all the good authorities. It follows that Matt. xi. 10 is not taken from Mark i. 2 (Zahn, Einl., ii. 316, 332). One can see how important the so-called “lower” criticism may be for the “higher.”

i. 11; ix. 7. See on “Punctuation” above, p. 52.

i. 29. “B here has ἐξελθὼν ἦλθεν, and D b c e q Pesh. have substantially the same. This is not an improvement, because it excludes Peter and Andrew. The reading of Syr^sin is peculiar, ‘and He went out of the synagogue and came into the house of Simon Cephas (Andrew and James and John were with him), and the mother-in-law etc.’” See Zahn, Einl., ii. 252, and below on ix. 14.

i. 41. The remarkable “Western reading” ὀργισθείς is dismissed by Swete with a reference to W-H, who call it “a singular reading, perhaps suggested by v. 43 (ἐμβριμησάμενος), perhaps derived from an extraneous source.” In my Philologica Sacra, p. 26, I have expressed the opinion that it is impossible to suppose a copyist altered σπλαγχνισθείς to ὀργισθείς, even though ἐμβριμησάμενος does follow two verses further down.^[271] Either ὀργή, ὀργίζεσθαι has another meaning in Biblical Greek, which is quite possible, or we have here an instance of a difference in translation. The confusion of the gutturals, e.g., is very common. Compare Ps. xii. 6, יפיח, Gr. יפיע; Ps. xiv. 6, עני = liii. 6, חנך; וישמח in Isa. xxxix. 2 for וישמע in 2 Kings xx. 13; Ps. xxii. 25, ענות, where Gr. has δέησις = תחנות; Ps. xcvii. 11, זרע, Gr. ἀνέτειλεν = זרח; and especially Mark ix. 19 in the recently-published Evangeliarium Hierosolymitanum of Lewis-Gibson, where Cod. B has מרחית for מרעת found in A B. Compare also חד, Matt. vii. 11 (p. 68), and עד (p. 135). A glance at the Thesaurus Syriacus 3953 shows that רעם is used, not only for βροντᾶν, but also for σπλαγχνίζεσθαι, στέργειν, and συμπαθεῖν, while אתרעם stands for χαλεπαίνειν, ἀγανακτεῖν, and γογγύζειν. Payne-Smith gives no instance of ὀργίζεσθαι. The usual Syriac word for it even in Syr^sin and Syr^hier is רנז 263or אתחמת; both verbs are found together in 1 Macc. vi. 59 for the simple ὠργίσθησαν (אתחמתו ורגזו). It is worth noting that in Col. iii. 13, ὀργήν is read by F G, where D* has μέμψιν, and the other authorities μομφήν.

On the reading in Mark i. 41, see Harris, Fragments etc. (1895), p. 6. He shows that Ephraem had ὀργισθείς in his text alongside of σπλαγχνισθείς. The Arabic Diatessaron, in which the pericope does not come till § 22, follows the usual text, and so, too, does Syr^sin.

ii. 14. Zahn (Einl., ii. 263) holds that “Levi son of Alphaeus” is the original reading here and not “James,” and that it was taken from Mark into the Gospel of Peter. The reading “Jacobum” was also taken into the first hand of the Vulgate manuscript G from D 13, 69, 124, a b c d e ff₂ r. In Koetschau’s new edition of Origen, the name is no longer spelt Λεβής but Λευής (i. 113, 19; Cod. P: Λευίς).

iii. 17. Our expositors might tell us where Luther got his “Bnehargem,” which is retained in the German Revised Version. On Daniel ii. 7 Jerome has “Benereem.” I have looked in vain in Lyra, Pole’s Synopsis, Calov, and Wolf.

iii. 31. We have here to choose between καλοῦντες (א B C L etc.), φωνοῦντες (D etc.), and ζητοῦντες (A): Δ leaves a space. I am inclined to think that φωνοῦντες is the original reading, which was improved by the substitution of the more usual word καλοῦντες, just as οὐ φωνεῦντος ἀκούω was altered to λαλέοντος in the Delphic Oracle in Herodotus i. 47. Compare a similar variation in Heb. xi. 13, where the original reading κομισάμενοι (א* P) was thought to be improved by the substitution of λαβόντες (א^c D E K) or προσδεξάμενοι (A). Here, too, A stands alone. Was it never copied?

vi. 16. There is a discrepancy in the Eusebian Canons in this verse which has not been explained. Both Tischendorf and Wordsworth and White number this verse 58/2. But according to the table in TiGr., p. 152, W-W, p. 10, pericope 58 belongs to the tenth Canon as being one that is peculiar to 264Mark. As a matter of fact it is not so, unless Eusebius meant ἀκούσας δέ at the beginning of the verse. It is remarkable that Eusebius did not make the whole of verses 14-20 one pericope of the second canon, but numbered 14, 15 as 57/2 and 17-20 as 59/2. He must therefore have found something peculiar in verse 16 to make it 58.

vi. 20. This passage is very instructive from a textual point of view. Most authorities read that “Herod had put John in prison, heard him and did much,” or “heard much of what he did,” ἀκούσας αὐτοῦ πολλὰ (ἃ) ἐποίει. But in place of this last word א B L and the Bohairic version alone read ἠπόρει, “was much perplexed when he heard him.” The great majority of expositors decide at once in favour of the latter reading, setting aside ἐποίει as the scriptio proclivior. But in that case should it not have been ἠπορεῖτο? In classical Greek it should undoubtedly, but in Biblical Greek we find ἠπόρει in Wisd. xi. 5, 17, for example, and what is specially worth noting, διηπόρει in the parallel passage Luke ix. 7, for which D, it is true, has ἠπορεῖτο. The passage may therefore be taken as showing that the correct reading has been preserved in a very few witnesses. Strict logic, moreover, would lead us to infer that not one of our 1300 manuscripts is derived from any one of these three, but that א B L continued childless. Is that likely? Field, it may be added, decides in favour of ἐποίει (Otium Norvicense; see Expository Times, August 1899, p. 483), and so, too, does Burkitt (Texts and Studies, v. 5, p. xix). In Philo, i. 264, line 8 (ed. Cohn), the manuscripts vary between μετεωροπολειν, —πορειν, —ποιειν, and —λογειν.

vii. 33. Codex W^d, published by Harris in facsimile (1896), here exhibits a very peculiar reading which Harnack (ThLz., 1891, p. 356) thinks has affinity with Tatian. It reads: ἔπτυσεν εἰς τοὺς δακτύλους αὐτοῦ καὶ ἔβαλεν εἰς τὰ ὦτα τοῦ κωφοῦ καὶ ἥψατο τῆς γλώσσης τοῦ μογιλάλου. This gives us quite another view of the occurrence than most of the authorities do. It seems much more natural certainly to moisten the 265fingers before putting them in the ears than before touching the tongue. It reads somewhat similarly in Syr^sin, which says that “he put his fingers and spat in his ears, and touched his tongue.”^[272] This manuscript exhibits other noteworthy readings, which will be found most conveniently in Swete.

ix. 14. The singular, ἐλθὼν ... εἶδεν, has the support of D, while Syr^sin takes the side of the plural, ἐλθόντες ... εἶδον. Zahn decides for the latter. He explains the plural by saying that the original narrator was evidently one of the three disciples who were with Jesus on the Mount, in all probability Peter, as tradition has it. Peter, of course, in telling the story, used the first person and the plural number, “When we came down from the mountain we saw, etc.” Mark, reporting the words of Peter, turned the first person into the third, retaining the plural number. Zahn explains in the same way the somewhat peculiar expressions in Mark i. 29. Here Peter said, “we (i.e. Jesus, Andrew, and himself) came into our house with James and John.” In reporting Peter’s words Mark paraphrases “we” and “our,” and says, “they came into the house of Simon and Andrew with James and John.” See Zahn, Einl., ii. 245 f.

x. 30. Neither Tischendorf nor Swete observes that in addition to the readings διωγμῶν and διωγμόν the singular διωγμοῦ is exhibited by D. Has the mysterious reading εἰς που in Clem. Alex. (Quis Dives) anything to do with this? It is worth remarking that the Vienna Arabic manuscript (Lagarde: Storr) has a note after “post persecutionem” to the effect that this is the “Roman” reading.

xiv. 51. καὶ νεανίσκος τις, א B C L; νεανίσκος δέ τις, D; καὶ εἷς τις νεανίσκος, A E etc. This last is rejected by Zahn on the ground that the text has evidently been accommodated to verse 47, under the false impression that another of the disciples is referred to. It is adopted, however, by Tischendorf⁸, and supported by Brandt, Die Evangelische Geschichte etc., Leipzig, 1893, p. 23 ff.

266xiv. 65. ἔλαβον, א A B and most authorities: ἐλάμβανον, D G, 1, 13, 69, 2^pe, al¹⁰: ἔβαλλον, H....: ἔβαλον, E M U etc. The simplest explanation of this variety of readings is that ἐλάμβανον was first, and that it was changed into the more common aorist ἔλαβον, which then became ἔβαλον or ἔβαλλον. The converse is not so likely, viz. that ἔβαλλον or ἔβαλον became first ἔλαβον and then ἐλάμβανον, or that ἔλαβον gave rise directly both to ἐλάμβανον and ἔβαλον or ἔβαλλον. On these and also on internal grounds the reading of D G is to be preferred: “they began to spit upon him, and continued to buffet him.”

xv. 28. Syr^sin is now to be added to the authorities that omit the interpolation. On the interesting names, Zoatham and Chammatha, Dysmas and Gestas, Titus and Dumachus (i.e. Θεομάχος), see Berger in the notice of Wordsworth, and White’s Epilogus mentioned above, and also J. R. Harris in the Expositor, March 1900, p. 162 ff., April, p. 304.

xv. 34. It is extraordinary that no reference is made in Swete’s edition to the very singular reading of Codex D, ὠνίδισας instead of ἐγκατέλιπες. In addition to the testimony of the Old Latin manuscripts c (exprobrasti me), i (me in opprobrium dedisti), k* (maledixisti: see Burkitt in the Journal of Theological Studies, i. p. 278), this reading is attested in Greek by Macarius Magnes. No explanation of it has yet been given that is in all respects satisfactory. See Expository Times, August 1898, and February, March, and April 1900.

xvi. 9-20. The English Revisers had not the courage to omit the conclusion. They print it quite like the rest of the text, only they separate it from the foregoing by a somewhat wider space than usual, and give a note in the margin to the following effect—viz. “The two oldest Greek manuscripts and some other authorities omit from verse 9 to the end. Some other authorities have a different ending to the Gospel.” The German Revised Version has no remark to offer, which is easily accounted for on the principles on which that version is made. The most careful discussion of the passage is now that 267of Swete, pp. xcvi-cv. See also Zahn, Einl., ii. 227-235, 237, 240, and compare the Appendix in Chase’s Old Syriac Element, pp. 150-157, “Note on Mark xvi. 9-20,” and Arthur Wright, The Gospel according to St. Luke, p. xv.

The subscription of several minuscules bears that Mark’s Gospel was written at Rome ten years after the Ascension, and delivered to the brethren there by Peter, the πρωτοκορυφαῖος of the Apostles. Others give Egypt as the place of origin. It is of more importance to observe that Λ 20, 262, 300 contain the note: ἀντεβλήθη ὁμοίως ἐκ τῶν ἐσπουδασμένων. This refers to the subscription to Matthew found in these manuscripts: ἐγράφη καὶ ἀντεβλήθη ἐκ τῶν ἐν Ἱεροσολύμοις παλαιῶν ἀντιγράφων τῶν ἐν τῷ ἁγίῳ ὄρει ἀποκειμένων. A similar subscription occurs in 2^pe, a minuscule of considerable importance for Mark (473 in Scrivener; see above, p. 151, n.).

Luke.

Apart altogether from the question how the numerous and decided peculiarities of Codex D are to be explained, we find a great many problems connected with the text of Luke’s Gospel.

On the supposed title see Zahn, Einl., ii. 383.

i. 26. In place of the definite indication of time, Blass follows certain Latin authorities, especially the Latin Irenæus, in giving: in ipso (or, eodem) autem tempore, ἐν αὐτῷ δὲ τῷ καιρῷ. Zahn points out (Einl., ii. 354) that this is the customary formula for the beginning of a pericope in the Lectionaries, and that while no doubt in the later Greek system the pericope of the Annunciation began with verse 24, 26 is the more appropriate beginning. He adds that in any case the origin of this formula is evident, and that Cod. D, which here parts company with the Latin witnesses, gives other indications besides this of the influence of a pericope-system. See the Introduction to Scrivener’s edition of the Codex, p. li.

i. 46. On the reading Elisabeth, see above, p. 238.

i. 63. The β text inserted the words ἐλύθη ἡ γλῶσσα 268αὐτοῦ before καὶ ἐθαύμασαν πάντες, by way of explaining the astonishment of the people. Zahn thinks this an absurd misplacement, seeing that the mention of Zechariah’s speaking does not come till the following verse, and the people could not know that his tongue was loosed till they heard him speak. Syr^sin accordingly corrects this by putting the mention of the astonishment after that of the speaking, in which it is followed by Blass.

ii. 4, 5. In the β text Blass adopts the reading αὐτοὺς, and transposes the clause διὰ τὸ εἶναι αὐτοὺς ἐξ οἴκου καὶ πατριᾶς Δαυείδ to the end of verse 5. This arrangement is also exhibited by D. Syr^sin reads “both.” One Old Latin manuscript has essent, but as it exhibits the clause in the usual place, Zahn thinks that essent is manifestly a clerical error for esset. The Syriac, he points out, is derived from Tatian. See Einl., ii. 355; Forsch., i. 118; GK., ii. 561; Vetter, Der dritte Korintherbrief (1894), 25.

ii. 7. One Latin manuscript (e) has obvolverunt and collocaverunt, which may be compared with essent in verse 4. Zahn thinks that the plural here is due to the reflection that the mother does not usually herself attend to a new-born infant.

ii. 14. How does the Christmas song of the angels run exactly? Is it ἐν ἀνθρώποις εὐδοκίας, or ἐν ἀνθ. εὐδοκία? The question belongs more to exegesis than textual criticism. The whole matter turns upon a single letter, but it divides Western Christendom in two parts. The Latin Church reads it as in hominibus bonae voluntatis, “among men of goodwill,” or, as modern critics understand it, “among men of God’s good pleasure.” The second reading makes it “goodwill to men.” Which should it be? The former reading, the genitive, is supported by א* A B* D, the Latin, and the Gothic, whereas nearly all the other witnesses, including the Bohairic, the three Syriac, and A itself in the Hymns at the end of the Old Testament Psalter, have the nominative. One thing seems to me decisive in favour of the nominative. Scarcely 269any part of the New Testament is so steeped in the Hebrew spirit as the first two chapters of Luke’s Gospel. As Field points out in the third part of his Otium Norvicense, the Greek ἄνθρωποι corresponds to the Hebrew expression “son of Adam,” which cannot take another genitive after it—“sons of Adam of goodwill.” On the other hand, the word goodwill in Hebrew is always followed by the preposition corresponding to the Greek ἐν. So that, till we have further testimony, I would retain the nominative and the tripartite division, notwithstanding the authority of Tischendorf, Westcott and Hort, Weizsäcker, Stage, and Blass, who, by the way, mentions no variants in the β text.

ii. 40. D here reads ἐν αὐτῷ in place of ἐπ’ αὐτό. The difference is slight, but not unimportant from a theological point of view. It is not accidental, as is shown by the corresponding change of ἐπ’ into εἰς in ch. iii. 22.

iii. 22. Zahn regards this as one of the passages wherein D and its associates have preserved the original reading. They exhibit here ἐγὼ σήμερον γεγέννηκά σε in place of ἐν σοὶ εὐδόκησα. He says, moreover, that “those who hold the former as original need not lament its disappearance from tradition subsequent to the year 300” (Einl., ii. 240, 356). See Burkitt in Barnard’s Biblical Text of Clement, pp. xiii. 38.

iii. 23 ff. May not the peculiar form of the genealogy in D be explained by the Diatessaron, which originally had no genealogy? The index of the Latin edition shows that there was none originally, but we find in the text one compiled from Matt. i. 1-16, Luke iii. 34-37, Matt. i. 17. The first-known manuscript of the Arabic Diatessaron had Matt. i. 1-17 in § 2, and Luke iii. 24-38 in § 9. The better manuscript, discovered later, has no genealogy in the text, but it contains one compiled from Matt. and Luke, inserted between the close of the work and the subscription by way of appendix. See Zahn, GK., ii. 539; J. H. Hill, Earliest Life of Christ etc., p. 3 f.

270iii. 27. The correct explanation of Ῥησά is that given by Plummer in his Commentary on Luke, and quoted by Bacon in Hastings’ Dictionary of the Bible, ii. 140. “Rhesa, who appears in Luke, but neither in Matt. nor in 1 Chron., is probably not a name at all, but a title which some Jewish copyist mistook for a name. Zerubbabel Rhesa or Zerubbabel the Prince (רֵאשָׁא) has been made into ‘Zerubbabel (begat) Rhesa.’” The interpretation of Rhesa as “prince” is, however, not new. See Pole’s Synopsis: it was not safe to use the proper name Zerubbabel in Babylon, seeing that it meant “ventilatio Babelis,” and the name Sheshbazzar was therefore substituted for it. Sic filii eius Meshullam et Hanania, quia vix ibi tuto aut proprie dici potuerunt Abiud, i.e. patris mei est gloria, et Rhesa princeps (Lightfoot, Horae Hebraicae). Reuchlin (Rudimenta, p. 18) gives the explanation רֶשָׁע (sic) qui cognominatur Mesollam. This interpretation, however, lends no real support to Sellin’s theory.

iv. 34. The exclamation ἔα, which Zahn (GK., i. 682) says is unknown in the New Testament, is omitted by D, eleven Old Latin manuscripts, and also by Marcion. It is supported by a considerable number of witnesses in Mark i. 24. According to Zahn, these witnesses took it from Luke, but of this I am by no means certain. Syr^sin omits it in both places. In Luke it is also omitted by four manuscripts of the Vulgate mentioned by Wordsworth and White.

iv. 34. Marcion invariably omits Ναζαρηνέ. There is, however, no other authority for its omission. See Zahn, GK., i. 685; ii. 456.

iv. 44. Ἰουδαίας is the better attested reading, and on account of the improbability of its being invented, should be regarded as the original. See Zahn, Einleitung, ii. 373.

v. 5. Ἐπιστάτα in the New Testament is peculiar to Luke. In place of it D has διδάσκαλε here, and κύριε in viii. 24. It retains ἐπιστάτα, however, in viii. 45, ix. 33, ix. 49, and xvii. 13.

v. 14. The long interpolation at the end of the verse found 271in D d is derived from Mark i. 45 and ii. 1, though there are slight differences. It is introduced here for harmonistic reasons. Was it taken from Tatian?

v. 27. After the name of Levi, D inserts τὸν τοῦ Ἀλφαίου, which, according to Zahn, is not original. See Einl., ii. 263.

v. 39. Marcion agrees with D in the omission of this verse. Syr^sin and Syr^cu are, unfortunately, both defective here. To the authorities for its omission should be added r, which Weiss does not mention. On the reasons for the omission of the verse, see Zahn, GK., i. 681.

vi. 5. Zahn is of opinion that the narrative of the man working on the Sabbath is taken from the same source as Mark xvi. 9 ff., and the pericope adulteræ, John vii. 53-viii. 11—viz. from Papias, and that it may be historically true. See his Einleitung, ii. 355. Westcott and Hort insert it among their “Noteworthy Rejected Readings,” and Resch puts it among the “Logia Jesu.” The Sinai-Syriac is defective here. For a long time it was thought that D and Stephen’s β were different manuscripts, and they are here cited by Mill as “duo codices vetustissimi.” This was shown to be a mistake by Bengel. Grotius also speaks of “nonnulli codices,” and, according to Mill, thought the words were “adjecta ab aliquo Marcionita.” The narrative seems to have remained quite unknown during the thousand years that elapsed between its relation by D and its publication by Stephen in 1550. According to Scrivener’s edition of Codex Bezae, p. 435, none of the ten or twelve later hands that worked upon the manuscript down to the twelfth century and even later, seem to have touched the page on which this narrative stands (205b). It would seem, therefore, that no copy was ever made of this manuscript either. How much would have been lost had it also disappeared entirely?

vi. 10. Whether ὡς καὶ ἡ ἄλλη is genuine or not is of no material consequence so far as the exposition of the passage is concerned, but it is important in connection with the question of the relationship of Luke to the other Synoptics. The 272words are wanting in Mark iii. 5, but occur in Matt. xii. 13. See Zahn, Einl., ii. 420.

vi. 31. Zahn is not sure if Marcion’s text contained the Golden Rule in this passage in the negative form. GK., i. 680; ii. 462.

vii. 27. Zahn thinks that ἔμπροσθέν σου should perhaps be omitted here (Einl., ii. 316).

viii. 43. The words ἰατροῖς προσαναλώσασα ὅλον τὸν βίον are omitted in B D. Zahn holds it to be an “unworthy insinuation” to suppose that Luke, being himself a physician, toned down the expressions used by Mark as reflecting on the credit of his profession. The words are more likely to be a gloss from Mark. See Einleitung, ii. 437.

ix. 1. This verse is written three times over in codex Ξ. This cannot be a mistake. It might have been written twice by inadvertence, but not three times. The reason lies in the fact related—viz., the conferring of the power over evil spirits.

ix. 16. The reading εὐλόγησεν ἐπ’ αὐτοὺς crept into the Vulgate manuscript called G by Wordsworth and White from the Old Latin. It is now attested also by Syr^sin. See Lewis, Some Pages, in loco. Zahn thinks it is deserving of special attention (GK., i. 682). In this he is quite right.

ix. 18. Marcion here had τὸν υἱὸν τοῦ ἀνθρώπου. See Zahn, GK., i. 686.

ix. 52-56. “It is impossible to suppose that the shorter form of the text is the original, and the longer due to a later interpolation, as this would imply what is incredible—viz., that one of Marcion’s most antinomian readings found its way into a large number of Catholic manuscripts (D, the Peshitto, Harklean Syriac, most Latin witnesses, Chrysostom, etc.). The only probable explanation is that the Catholic writers objected to 54b and 55b on account of the use made of them by the Marcionites, and the apparently Marcionitic character of their contents. They were particularly offensive when taken together. Accordingly, some manuscripts like e and Syr^cu omitted only 54b, others, like A C, only 55b, while 273others again, like B L Syr^sin, boldly omitted both.... The words were written by Luke and not invented by Marcion.” Zahn, GK., i. 681, ii. 468; Einl., ii. 357.

x. 1. Instead of 70, B D, Tatian, the Syriac, and the Latin give 72. According to Zahn, the number has nothing to do with the Jewish enumeration of 70 Gentile nations, languages, or angels, nor the 70 members of the Sanhedrim, the 70 translators of the Old Testament, or with any other number 70. These 70 were not sent to the Gentiles, and Luke gives no hint of the allegorical significance of their number. Any such allegorizing was foreign both to himself and Theophilus, neither of whom was a Jew. See Einl., ii. 392.

xi. 2. On βαττολογεῖν ὡς οἱ λοιποί in D, see my Philologica Sacra, pp. 27-36.

xi. 3. There is a certain amount of probability in Zahn’s view that Marcion was led to insert σοῦ after ἄρτος ἐπιούσιος by thinking of John vi. 33 f., a passage which suggested itself to Origen also in this connection. See Zahn (GK., i. 677, ii. 471), who thinks it probable that Marcion interpreted the words in a spiritual sense (= supersubstantialis).

xi. 53. The text of D here displays several marked variations, which, however, do not affect the sense of the passage. Zahn sees in them the arbitrary alterations of a later time; but Weiss thinks that in some particulars they may preserve the original.

xii. 1. Marcion, seemingly, and Jerome omit πρῶτον, which is attested by most of the Old Latin witnesses, with the exception of b. See Zahn, GK., i. 692, ii. 474.

xii. 14. The words ἢ μεριστὴν were omitted by Marcion (Zahn, GK., i. 682). They are also wanting in the Sinai-Syriac (see Lewis, Some Pages).

xii. 38. The mention of the ἑσπερινὴ φυλακὴ by Marcion and other authorities is due, according to Zahn (GK., ii. 683; Einl., ii. 356), to the “magisterial consideration” that an orderly householder would not come home from the festivities after midnight or in the early hours of the morning, but at the 274latest in the first watch of the night, which was still called the evening. The reading is also found in Irenæus, but not in the Sinai-Syriac.

xii. 51. βαλεῖν (אַרְמֵא) is found here in Syr^sin in place of ποιῆσαι (D e Syr^cu), mittere (b l), δοῦναι (usual text). This is interesting in view of Marcion. See Zahn, GK., i. 604, ii. 476. Tertullian seems to have been mistaken in thinking that μάχαιραν was read in place of διαμερισμόν in this connection (machaeram quidem scriptum est. Sed Marcion emendat, quasi non et separatio opus sit machaerae).

xiii. 8. See above, p. 193 ff. Chase cites this passage as an indication of the laxity of transcription of which D was guilty in introducing what appears to be a common agricultural phrase. In Columella (De Re Rustica, xi. 3) we find “confecta bruma stercoratam terram inditam cophinis obserat.” Chase also cites from the manuscript notes of Hort the reference to Plutarch, Vita Pompeii, 48, αὐτοῦ δέ τις κοπρίων κόφινον κατὰ κεφαλῆς τοῦ Βύβλου κατεσκέδασε. Better than any words of mine are those of Zahn, Einleitung, ii. 346:—No one with any perception of the difference between naïve originality and a regularity due to liturgical, dogmatic, and stylistic considerations can fail to assent to the following propositions—viz., (1) as regards contents and form of expression β (i.e. the text of D and its associates) has preserved much original matter, which from the very first was peculiarly liable to alteration, and which was set aside by the learned revisers from the end of the third century onwards (Lucian, Hesychius, Pamphilus), etc.

xvi. 12. While the common text with Syr^sin reads ὑμέτερον, for which B L have ἡμέτερον, Marcion alone supports 157 e i l in reading ἐμόν. How is this to be explained? Compare above, p. 211, and Zahn, GK., i. 682.

xvi. 19. Zahn denominates the introductory words found in D, εἶπεν δὲ καὶ ἑτέραν παραβολήν, “a liturgical gloss at the beginning of a pericope.” Blass, too, omits them from the β text.

275xvi. 22, 23. א*, most Old Latin witnesses, and the Vulgate omit καί at the beginning of verse 23, and read ἐτάφη ἐν τῷ ᾅδῃ. This conjunction of the words is attested by Tatian and Marcion. The Sinai-Syriac presupposes the form “was buried. And being in Hades he lifted up his eyes.” Attention may be drawn to the detailed notice of the different readings by Wordsworth and White. They say: Asyndeton in Johanne tolerabile, in Luca vix ferendum videtur.... Vix dubium est quin Lucas ipse scripserit καὶ ἐτάφη· καὶ ἐν τῷ ᾅδῃ, sed καί secundum in antiquissimis codicibus ut nunc in א* casu omissum, ex conjectura tribus modis restitutum videtur, sc. καὶ ἐν τῷ ᾅδῃ, et ἐν δὲ τῷ ᾅδῃ et ἐν τῷ ᾅδῃ καὶ; quae lectiones omnes in codicibus Latinis referuntur, et tertiam ab Hieronymo ex traditione codicum suorum servatam magis quam ex ratione praelatam credimus. See Zahn, GK., i. 682, ii. 480.

xvii. 11. “In all likelihood μέσον, without the preposition, as given by D, is the original form. This was variously replaced by ἀναμέσον (Ferrar Group), which is not amiss, by διὰ μέσου (A X, etc.), which is not so good, and by διὰ μέσον (א B L), which is very bad.” Zahn, Einleitung, ii. 391. Compare, also (for μέσον), Jülicher, Gleichnisreden Jesu, ii. 516.

xvii. 21. Marcion inserts ἰδοὺ before ἐκεῖ, which Zahn holds to be original. Syr^sin reads “here it is, or there it is,” and therefore apparently omits the first ἰδοὺ as well. See Lewis, Some Pages. Wordsworth and White omit Tischendorf’s g^{1. 2} from the authorities given by him in support of the omission of the second ecce.

xviii. 20. On the alterations made on the text here by the followers of Marcion, see Zahn, GK., i. 616, ii. 484.

xviii. 25. The evidence in support of the readings τρήματος and βελόνης is very strong (א B D L). The choice of the terms τρῆμα for τρύπημα or τρυμαλιά, and βελόνη for ῥαφίς, betrays the language of the physician. See The Expositor’s Greek Testament, Acts of the Apostles, Introduction, pp. 9-11; Zahn, Einleitung, ii. 427 f., 435 f.

276xx. 35. With reference to this verse, Tertullian makes the following charge against the Marcionites: Nacti enim scripturae textum ita in legendo decurrerunt: “quos autem dignatus est deus illius aevi”; “illius aevi” “deo” adjungunt ... cum sic legi oporteat, “quos autem dignatus est,” ut facta hic distinctione post “deum” ad sequentia pertineat “illius aevi,” etc. Zahn insists, as against Ritschl, Hilgenfeld, and Volkmar, that this requires not only the insertion of ὑπὸ τοῦ Θεοῦ after καταξιωθέντες, but also the active voice instead of the passive, as though the Marcionites had read οὓς δὲ κατηξίωσεν ὁ Θεὸς τοῦ αἰῶνος ἐκείνου, τυχεῖν καὶ τῆς ἀναστάσεως. A similar change of construction occurs in Matt. xxv. 41, where it is quite certain that τὸ ἡτοιμασμένον is a correction of the stronger expression ὃ ἡτοίμασεν ὁ πατήρ μου, found in D, 1, 22, ten Old Latin manuscripts, and the earliest Fathers.

xxi. 30. The insertion of τὸν καρπὸν αὐτῶν may be but a trifling addition, intended to facilitate the sense (Zahn, GK., i. 682), at the same time it is an interesting question how it comes to be in D, 157, 572 (see above, p. 211). Wordsworth and White say that D here is “ex Latinis forsan correctus.” Syr^sin agrees with Syr^cu in inserting the words.

xxii. 16. For πληρωθῇ D reads καινὸν βρωθῇ. On this see my Philologica Sacra, p. 38, where it is suggested that these two readings are due to the confusion of כלה and אכל. This occurs several times in the Old Testament—e.g. 2 Chron. xxx. 22, where ויאבלו is represented in the LXX by συνετέλεσαν. But even apart from the question of a Hebrew foundation for the variant, I am inclined to regard καινὸν βρωθῇ as the original, and πληρωθῇ as the correction.

xxii. 16-21. The narrative of the Last Supper is extant in three forms. There is (1) the common text, (2) that exhibited by the two most important of the Old Latin witnesses (b, e), in which verse 16 is followed by 19a, after which come 17, 18, 21, so that 19b and 20 are wanting altogether. The text of Syr^sin and Syr^cu resembles this. There is further (3) the form exhibited by D and four Old Latins, which has the same order 277as (1), but omits verses 19b, 20. Zahn decides in favour of (2). See his Einleitung, ii. 357 ff. It is to be observed that the last discovered Syriac omits the nominatival clause τὸ ὑπὲρ ὑμῶν ἐκχυνόμενον after τῷ αἵματί μου, which is the only member that seems to be derived, not from 1 Cor. xi. 24 f., but from Matthew and Mark, and that does not agree in construction with the rest. This confirms the supposition that these two verses are not part of the original text. See Westcott and Hort, Notes on Select Readings, p. 63 f.; Plummer, Commentary on St. Luke in the International Series (T. & T. Clark). Compare also the article by the latter in Hastings’ Dictionary of the Bible (Lord’s Supper).

xxii. 36. On ἀράτω Basil the Great (d. 379) remarks: ἀράτω ἤτοι ἀρεῖ· οὕτω γὰρ καὶ τὰ πολλὰ τῶν ἀντιγράφων ἔχει ... ὡς μὴ εἶναι πρόσταγμα ἀλλὰ προφητείαν προλέγοντος τοῦ Κυρίου. At present D is quite alone in exhibiting the reading ἀρεῖ, which is worth noting in view of τὰ πολλὰ above.

xxii. 43, 44. These verses, with their mention of the Bloody Sweat and the Strengthening Angel, are omitted in A B R T, one Old Latin (f), the Bohairic, Sahidic, and Armenian versions, and the Sinai-Syriac. On the other hand, they are read by the Curetonian Syriac and the Peshitto, by the first and third hands of א (the second hand enclosed them in brackets and cancelled them by means of dots), by D, as also by most of the Old Latin witnesses and the Vulgate. In the Greek Lectionaries they are omitted at the place where one would naturally expect them, but are found in the text of Matthew xxvi., together with portions of John xiii., in the Liturgy for Holy Thursday. This explains their insertion after Matt. xxvi. 39 in the Ferrar Group, at least in 13, 69, 124. The first of these, moreover, repeats the first two words of verse 43 (ὤφθη δὲ) in Luke, but no more. The necessary inference is that these verses are no part of the original text of Luke. They go back, however, to a time when extra-canonical traditions from the Life and Passion of Jesus were 278in circulation either orally or in writing. Zahn holds that D here has preserved what Luke wrote.

xxiii. 2. Zahn (GK., i. 668) expressly points out that Marcion did not invent the additional words καὶ καταλύοντα τὸν νόμον καὶ τοὺς προφήτας, but found them in his exemplar. They occur in eight Old Latin and at least five Vulgate manuscripts, among which are four of the early codices collated by Wordsworth and White. One of them omits et prophetas, while some others have nostram after legem. Weiss takes no notice of this addition, nor of the further addition in verse 5 of the words ἀποστρέφοντα τὰς γυναῖκας καὶ τὰ τέκνα, supported by at least two Old Latin manuscripts, both of which add non enim baptizantur sicut et nos, while one of them exhibits the still further extension nec se mundant. If the addition were really made by Marcion, it would be all the more deserving of attention. The omission of the additional words in verse 2 is conceivably due to homoioteleuton, the eye of the scribe passing from καταλύοντα to κωλύοντα, In the case of verse 5, the mention of the women and children is quite consistent with what is said elsewhere in the narrative of the Passion, but the reference to baptism and purification is not so clear. Codex c has the singular baptizatur, but this is merely a clerical error.

xxiii. 34. The case of the First Word from the Cross is remarkable. This verse, containing the words, “Father, forgive them, for they know not what they do,” is bracketed in א by an early corrector, and then restored; it is omitted by B without being replaced; it is inserted in D by a hand not earlier than the ninth century, and omitted by two Old Latin manuscripts, by two Bohairic codices, by the Sahidic version, and by the newly-discovered Sinai-Syriac. Is it possible to suppose that a Christian would have cancelled these words in a Bible manuscript like א, unless he had valid reasons for doing so in the tradition of the Gospel text? Zahn thinks they were omitted from D by mistake. On the Order of the Seven Words see my note in the Expository Times for June 1900, p. 423 f.

279xxiii. 38. The notice of the three languages in which the Inscription on the Cross was written is taken from the text of John, and is read by all the Latin authorities with the single exception of codex Vercellensis (a). Syr^sin is now to be added to the witnesses supporting the omission of the clause. Its use of the word בטקא reveals the ultimate affinity of this version with the Curetonian Syriac. The interpolation, as Zahn rightly asserts (GK., i. 675), points to the estimation in which John’s Gospel was held at an early date. Its insertion in Luke is undoubtedly erroneous.

xxiii. 43. The insertion of τῷ ἐπιπλήσσοντι in D, as well as the other variants found in this manuscript, viz. ἔλευσις, which is also read by D in Luke xxi. 7, and θάρσει, which is inserted by others in Luke viii. 48, is attributed by Zahn (Einleitung, ii. 356) to some preacher who sought in this way to contrast the penitent thief with his comrade. With the substantival expression ἔλευσις, compare δειπνοκλήτωρ exhibited by D in Matt. xx. 28. On the somewhat rare verb ἐπιπλήσσειν, compare the new edition of Origen, i. 5,8; also Clement Alex, (ed. Dindorf), i. 186, 188.

xxiii. 53. After κείμενος U, with a few minuscules, reads καὶ προσεκύλισεν λίθον μέγαν ἐπὶ τὴν θύραν τοῦ μνημείου, while three Vulgate manuscripts have a similar addition et inposito eo inposuit monumento lapidem magnum. On the other hand D, with its Latin, reads καὶ θέντος (leg. τεθέντος) αὐτοῦ ἐπέθηκεν τῷ μνημείῳ λίθον ὃν μόγις εἴκοσι ἐκύλιον. The same thing is found in the Old Latin manuscript c, et cum positus esset in monumento, posuerunt lapidem quem vix viginti volvebant. The Sahidic and Tⁱ exhibit a similar expansion of the text. In this addition, which Scrivener thought was “conceived somewhat in the Homeric spirit,” Harris detects a Latin hexameter which the scribe of Codex Bezae “deliberately incorporated into his text and then turned into Greek.” See his Study of Codex Bezae in Texts and Studies, ii. 1, 47-52. Chase, on the other hand, adduces Josephus, Bell. Jud., vi. 5, 3 (Syro-Latin Text, p. 62 ff.). Compare my Philologica Sacra, pp. 39, 58.

280xxiv. 6. The reading ὅσα (D, c, Marcion, etc.) in place of ὡς is now attested also by the Sinai-Syriac.

xxiv. 32. In place of καιομένη (α text) and κεκαλυμμένη (D), Blass inserts βεβαρημένη in the β text on the authority of the old Syriac versions, the Armenian, and the Sahidic. But in the Syriac this last reading is due to a transcriptional error of יקיר for יקיד (see Blass himself, p. 120, and compare the variants מוקר and מוקד in Rahmani’s Testamentum D. N. Jesu Christi, p. 112, 6); and as the Armenian is derived from the Syriac, the only question becomes whether the Sahidic reading is due to the same error. Κεκαλυμμένη in D, which has hitherto baffled explanation, is shown to be a purely clerical error by comparison with Heb. xii. 18, where also κεκαυμένῳ becomes κεκαλυμμένῳ in the Greek of D and in Pseudo-Athan. 57.

xxiv. 34. For λέγοντας D reads λέγοντες, which is simply a clerical error arising easily from the influence of the Latin, which would be the same in either case. For the conclusions drawn from this reading by Resch, see his Aussercanonische Paralleltexte, iii. 779 f. Other examples of the same mistake (—ες for —ας) occur in Matt. xxii. 16; Acts vi. 11, xvi. 35; Rom. vi. 13. It is interesting to observe that Origen had Σίμωνος καὶ Κλεόπα (i. 184, ed. Koetschau).

xxiv. 37. Zahn (GK., i. 681) rejects the supposition that the reading φάντασμα for πνεῦμα was coined by Marcion and taken from a Marcionite Bible into D. That he is right in doing so appears from Chase, who shows that φάντασμα here is the same as δαιμόνιον ἀσώματον in Ignatius (Ad Smyrnaeos, iii. 2). See my Philologica Sacra, p. 25. The Semitic equivalent of φάντασμα as well as of δαιμόνιον is שֵׁאד, שִׁאדָא,^[273] which is used in both the earlier Syriac versions, 281the Curetonian and the Lewis, to represent φάντασμα in Matt. xiv. 26 and Mark vi. 49. I find that שאדא is used for πνεῦμα in the translation of Eusebius (Eccles. Hist., v. 16, ed. Wright-Maclean, p. 289).

xxiv. 39. All the authorities agree in saying that Marcion omitted the words ψηλαφήσατέ με καὶ ἴδετε, while Tertullian and Epiphanius state that he also omitted σάρκας καὶ. See Zahn, GK., ii. 495, who adds that “the longer clause—i.e. ψηλαφήσατέ με καὶ ἴδετε—is also omitted by D, it (with the exception of Colbertinus), vg, but not Syr^cu, as Tischendorf wrongly states.” This however is a misapprehension. The om in Tischendorf refers only to με after ψηλαφήσατε. This is omitted by Syr^cu as well as by D and also by Syr^sin. It would be more exact to say, however, that the καὶ before ἴδετε is also omitted by Syr^cu. Moreover, Syr^sin agrees with Syr^cu in reading ὅτι ἐγώ εἰμι αὐτός after ἴδετε.

The subscription of certain minuscules states that Luke’s Gospel was written fifteen years after the Ascension. Some say εἰς Ἀλεξανδρείαν τὴν μεγάλην, others ἐv Ῥώμῃ, while one says very strangely, ἐv τῇ Ἀττικῇ τῆς Βοιωταίας, “for Theophilus, who became bishop after divine baptism.” Λ, 262, 300 also contain here the notice of careful collation. The chapter enumeration in these manuscripts is not the same, being 342, 349, and 345 respectively.

John.

In this Gospel the attention of textual critics was long confined to the passage vii. 53-viii. 11. They failed to observe that in other places there are clauses and whole verses whose omission or interpolation has to be investigated in connection with vii. 53 ff., as, for example, iv. 9, v. 3, 4, and that interesting questions of textual criticism are raised in other parts of the book as well.

Chapter xxi., which the last two verses of the preceding chapter clearly show to be an Appendix, is equally well 282attested by all the authorities, while the omission of xx. 31 by the first hand of G is just one of those unaccountable phenomena which make their appearance so frequently in the domain of textual criticism. The same thing is probably to be said of the omission in א of the last verse of chapter xxi. Tischendorf was of opinion that this last verse in א, together with the concluding ornament and subscription, was not by the same hand (A) as had written the Gospel of John, but by another (D) who had acted as corrector, and had written part of the Apocrypha and six leaves of the New Testament. Tregelles, on the other hand, who examined the passage in Tischendorf’s presence, thought the difference was due simply to the scribe having taken a fresh dip of the ink: that at all events the scribe who wrote the Gospel (A) did not intend it to conclude with verse 24, otherwise he would have added a concluding ornament and subscription as in the case of Matthew and Luke. The verse is found in all the other manuscripts and versions with which we are acquainted, and the question with regard to א is interesting only from the fact that a few manuscripts do contain a scholium to the effect that the verse is an addition (προσθήκη) inserted in the margin (ἔξωθεν) by one of the scholars (τινὸς τῶν φιλοπόνων),^[274] and afterwards incorporated in the text by another without the knowledge of the former (καταγέντος (?) δὲ ἔσωθεν ἀγνοίᾳ τυχὸν τοῦ πρώτου γραφέως ὑπό τινος τῶν παλαιῶν μέν, οὐκ ἀκριβῶν δέ, καὶ μέρος τῆς τοῦ εὐαγγελίου γραφῆς γενόμενον). This entire note, however, is evidently no more than an inference drawn from the contents of the verse, as the Syriac Commentary of Theodore shows. See further, Zahn, Einleitung, ii. 495, and the reference to the Commentary of Ishodad in Sachau’s Verzeichnis der syrischen Handschriften in Berlin, p. 307.

With respect to the pericope adulteræ, on the other hand, we may be quite certain that it did not originally stand in the 283position it now occupies (vii. 53-viii. 11), nor indeed in John’s Gospel at all, although the decision of the Holy Office of the 13th February 1897, which was confirmed by the Pope two days later, obliges Catholic exegetes to hold it as genuine. It is omitted in a great many manuscripts and versions—e.g. in א B L T. A and C are defective here, but the amount of space shows that they could not have contained it. It is omitted in the Syriac and Egyptian versions, in the Armenian and the Gothic, in some Old Latin codices, and in the earliest of the Greek and Latin Fathers. On the other hand it is found in all the manuscripts of Jerome and in Codex D, which is the only one of the earlier Greek manuscripts to contain it. In some minuscules and later Armenian manuscripts it stands at the end of the fourth Gospel, where now Westcott and Hort put it. In minuscule 225, written in the year 1192, it follows vii. 36; in the Georgian version it comes after vii. 44; while in the Ferrar Group—i.e. in minuscules 13, 69, 124, 346, 556—it is inserted after Luke xxi. 38. Its insertion after vii. 36 is probably the result of an accidental error. In the Greek Lectionaries the liturgy for Whitsunday begins at verse 37 and extends to verse 52, followed by viii. 12, so that the pericope was, by mistake, inserted before instead of after this lection. Its position in the Georgian version is the more remarkable, seeing that in the Old Latin Codex b, which contained the pericope by the first hand, the entire passage from vii. 44-viii. 12 has been erased. As a probable explanation of its position in the Ferrar Group after Luke xxi. 38, it has been suggested that the scribe inserted it there owing to the resemblance between Luke xxi. 37 and John viii. 1, and also between Luke xxi. 38 (ὤρθριζε) and John viii. 2 (ὄρθρου). Harris thinks that its proper place is in John between chapters v. and vi., because reference is made in v. 45, 46 to the Mosaic Law, which is also mentioned in viii. 5.

But the remarkable thing is that here again the text of D differs in a conspicuous manner from that of the other witnesses. In viii. 2 the words καὶ καθίσας ἐδίδασκεν αὐτούς are 284wanting: in verse 4 we meet the sentence ἐκπειράζοντες αὐτὸν οἱ ἱερεῖς ἵνα ἔχωσιν κατηγορίαν αὐτοῦ, which does not come till after verse 5 in the other text: for μοιχείᾳ D has ἁμαρτίᾳ: in verse 5 it reads, Μωυσῆς δὲ ἐν τῷ νόμῳ ἐκέλευσεν τὰς τοιαύτας λιθάζειν, for which the other text has ἐν δὲ τῷ νόμῳ Μωσῆς ἐνετείλατο τὰς τοιαύτας λιθοβολεῖσθαι: in verse 11, D has ὕπαγε where the other text has πορεύου. Now, if two persons got such an easy sentence as “Moses in the Law commanded to stone such” to translate from Latin, Hebrew, or any other language into Greek, one of them might quite well use κελεύειν and λιθάζειν, and the other ἐντέλλεσθαι and λιθοβολεῖν. And so the question is suggested whether the two forms in which the text exists were not derived from different sources, that of D, e.g., from its Latin. But on closer examination the latter supposition is seen to be impossible. For the Latin corresponding to ἔχωσιν κατηγορίαν αὐτοῦ is “haberent accusare eum,” showing that the Latin translator read κατηγορεῖν in his original,^[275] and for ὥστε πάντας ἐξελθεῖν he has “uti omnes exire,” where again the infinitive speaks for the priority of the Greek. On the other hand, it is to be observed that, according to Eusebius (Eccles. Hist., iii. c. 39, sub fin.), Papias knew and recorded an incident περὶ γυναικὸς ἐπὶ πολλαῖς ἁμαρτίαις διαβληθείσης ἐπὶ τοῦ Κυρίου, ἣν τὸ καθ’ Ἑβραίους εὐαγγέλιον περιέχει. So that the Gospel according to the Hebrews (i.e. the Palestinian Jewish Christians) contained a narrative similar to this, we may say quite confidently, contained this narrative. From that Gospel it was taken and inserted in some manuscripts after Luke xxi., in others after John vii. By the time of Augustine it was so widely propagated in the Latin that he thought it had been removed from certain manuscripts by people of weak faith, or rather by enemies of the true faith, “credo metuentes peccandi immunitatem dari mulieribus suis.” The pericope is no part of John’s Gospel, though it belongs to the oldest stock of evangelic tradition. 285On the question whether it may not originally have stood between Mark xii. 17 and xii. 18, and so between Luke xx. 26 and xx. 27, see Holtzmann in the ThLz., 1898, col. 536 f. Vide supra, p. 66.

i. 5. Zahn raises the question what word Ephraem found in his copy of the Diatessaron corresponding to κατέλαβε, seeing he gives vicit. In this connection I might (with the proviso that the reading may be more easily explained from the Armenian) point out that the Syriac word תשתלט corresponds to καταλαβέτωσαν in Sirach xxiii. 6. This stands elsewhere for ἄρχω, δεσπόζω, ἐξουσιάζω, κατακυριεύω, κυριεύω, κρατῶ. כבשׁ also frequently represents the Greek καταλαμβάνειν. The Sinai-Syriac for John i. 5 is unfortunately lost.

i. 13. The reading ὃς ... ἐγεννήθη is, so far as is known at present, attested by Latin witnesses only, “qui natus est.” But as Zahn is careful to point out (Einl., ii. 518), it did not originate on Latin soil, for Justin presupposes it, and, moreover, Irenæus constantly applies the passage to the Incarnation, while the Valentinians, who had the usual text, were accused by Tertullian of falsification. And it is not proved that the two last-mentioned used anything but a Greek Bible.

i. 17. According to early testimony, this verse, so frequently quoted since the time of Ritschl, once ran: “The Law was given by Moses, but its truth came by Jesus.” See Zahn, Forsch., i. 121, 248.

i. 18. Zahn agrees with Hort in holding that the originality of the reading μονογενὴς θεὸς (without the article) is established. See Westcott and Hort, Notes on Select Readings; Zahn, Einleitung, ii. 544, 557; Westcott, Commentary on John, in loco. It may be mentioned here that Codex Monacensis of Origen’s Commentary on the Gospel of John has ὁ/μονογενὴς υἱός/θεός with ὁ and υἱός both written above the line in a later hand. This gave rise in the Codex Regius to the reading ὁ μονογενὴς υἱὸς θεός.

i. 28. Is it βηθαβαρά or βηθανίᾳ? The former is exhibited by the Sinai-Syriac, the Curetonian, and the margin of the Harklean, 286and the latter by the other three Syriac, and the Arabic Diatessaron. With regard to the former, is it the case, as is supposed by many, that it is due simply to a conjecture of Origen, and that Syr^cu and Syr^sin took it from him? According to Zahn (GK., i. 406), Hilgenfeld pointed in this direction in the ZfwTh., 1883, 119. See also Lagrange, Origène, la critique textuelle et la tradition topographique (Revue Biblique, iv., 1895, pp. 501-524). Origen explains βηθανία as οἶκος ὑπακοῆς, and the Syriac as “place of praise.” Compare on this the much-discussed passage in the Gospel of Peter (ὑπακοὴ ἠκούετο, c. xi.). Βηθαβαρά, on the other hand, he interprets as οἶκος κατασκευῆς, so that he must either have spelt it Bethbara, בית ברא, as in Jud. vii. 24, or taken it as Beth-ha-bara. It is spelt βηθααβαρά in Lagarde’s Onomastica Sacra, 240, 12, and Bethabara in 108, 6 (Bethbaara, Codex B). Jerome (see Onomastica Sacra) interpreted the name as “domus humilis (= ?) vel vesperae” in Joshua xv. 6, as “domus multa vel gravis” in xv. 59, and as ἀοίκητος in xv. 61, following Symmachus. Luther had Betharaba, but in three impressions of the New Testament and in three of the Postils he had Bethabara (according to Bindseil-Niemeyer), and in the margin Bethbara, with a note in which reference is rightly made to Jud. vii. 24, “ut mysterium consonet.” See my German or Greek-German New Testament. It may be asked if “Ainon” in John iii. 23 has any connection with Bethania. Compare בֵית עְַנות in Jos. xv. 59. For ἐν Αἰνών e has in eremo and f has in deserto. How is this to be explained? Compare Zahn, Einleitung, ii. 561.

i. 34. For υἱὸς א*, Syr^cu, Syr^sin, and e read ἐκλεκτός. D is here defective. Zahn thinks the latter reading is original, and the former an example of an early and widely current alteration. Westcott and Hort insert ἐκλεκτός among their Noteworthy Rejected Readings. The two readings are combined in some manuscripts “electus filius Dei.” See Zahn, Einl., ii. 515, 544, 557.

i. 41. Zahn here decides for the nominative πρῶτος. Both the disciples of John who attached themselves to Jesus found 287their brother, but Andrew was the first to do so. See Einleitung, ii. 477 f.

ii. 2. In his Commentary on the Gospels, extant in the Armenian only, we find Ephraem saying, “Graecus scribit recubuit et defecit vinum” (§ 53), which shows that he had a Greek exemplar before him containing the itacism ἐκλίθη for ἐκλήθη. See Zahn, Forsch., i. 62, 127, and compare Luke xiv. 8, where Antiochus, Homil., iii., has κατακλιθῇς for κληθῇς.

ii. 3. Zahn is perhaps right when he says that no critic need doubt for a moment that the original reading is the longer, genuinely Semitic text exhibited by א*, the Harklean Syriac, and the best Latin manuscripts. D is defective, as also Syr^cu and Syr^sin.

iii. 5. βασιλείαν τῶν οὐρανῶν is attested only by א*, a few minuscules, by c m, and certain early Fathers, in place of βασιλείαν τοῦ θεοῦ, which has now the support of the Sinai-Syriac. Zahn thinks the former reading to be correct (Einleitung, ii. 294). If that is so, this will be the only place where the expression is found in the New Testament outside the Gospel according to Matthew, where it occurs some thirty-three or thirty-four times. See note on “The Kingdom of Heaven” in the Expository Times for February 1896, p. 236 ff.

iii. 24. Zahn (Einl., ii. 515) thinks that the omission of the article before φυλακὴν shows that there was some uncertainty regarding the fact mentioned by John. This, however, is open to question. That the insertion or omission of the article may be of importance is shown by such examples as John v. 1; Matthew xxii. 23; Acts xvi. 6; James ii. 2.

iii. 34. Our knowledge of the text of the Sinai-Syriac here rests solely on the last reading of Mrs. Lewis: “Not according to his measure gave (or, gives) God the Father.” This rendering, as well as the insertion of ὁ Θεός in many Greek texts, is due to the fact that πνεῦμα was not taken as the subject of the sentence.

iv. 1. For ὁ κύριος Tischendorf reads ὁ Ἰησοῦς, which is probably correct. Ὁ κύριος is elsewhere found only three times 288in John—viz., vi. 23, xi. 2, xx. 20. According to Zahn (Einl., ii. 391) the first two passages are explanations outside the Gospel narrative interjected by the evangelist, while the words in the last passage are spoken from the point of view of the disciples.

iv. 9. The words οὐ γὰρ συγχρῶνται Ἰουδαῖοι Σαμαρείταις were retained by Lachmann and by Tischendorf in his seventh edition, because the only authorities known at that time for their omission were D a b e. But when the first hand of א appeared in confirmation of the testimony of these witnesses, the words were dropped by Tischendorf and bracketed by Westcott and Hort. Syr^cu and Syr^sin insert them, and perhaps Tatian. Zahn is inclined to admit them. “The classic brevity of the interjected explanation speaks for its genuineness.” See his Einleitung, ii. 549.

v. 1. ἡ ἑορτή is supported by א C etc., and ἑορτή by A B D etc. On the chronology, see Zahn, Einleitung, ii. 516.

v. 3b, 4. After ξηρῶν D alone inserts παραλυτικῶν, and then adds, with A² C³ I Γ Δ Λ Π (this last, however, with asterisks), the clause ἐκδεχομένων τὴν τοῦ ὕδατος κίνησιν. The shorter text is given by א A* B C* L. The whole of the fourth verse is omitted by א B C* D, 33, 157, 314. In this case D and A change sides. Within the limits of the verse there are a great many variations, which show that it is a very early addition. Some of the words are hapax legomena, like δήποτε, ταραχή, νόσημα. Zahn thinks the gloss may have been one of the “expositions” of Papias. According to the Commentary of Ishodad, Theodore of Mopsuestia did not consider this verse as part of the Gospel of John (Sachau, Verzeichnis der syrischen Handschriften, p. 308). See Zahn, Einleitung, ii. 557. Cyril says the incident occurred at Pentecost.

v. 36. Zahn (Einl., ii. 557) calls μείζων a difficult reading, and one that could not have been invented: “I, as a Greater than John, have the witness of God.”

vii. 8. οὔπω has taken the place of οὐκ in all the uncials except א D K M P, a fact which reveals its antiquity. Οὐκ 289is retained also by Syr^cu and Syr^sin. The change was introduced to obviate the inconsistency between vii. 8 and vii. 10. Porphyry (apud Jerome, Contra Pelagium, ii. 17) on the ground of οὐκ, accused Jesus of “inconstantia et mutatio,” and Schopenhauer (Grundprobleme der Ethik, 2nd edition, p. 225) cited this passage as justifying an occasional falsehood, saying that “Jesus Christ himself on one occasion uttered an intentional untruth.” See Zahn, Einleitung, ii. 547.

vii. 15. See Addenda, p. xvi.

viii. 57. According to the authority cited in the ThLz., 1899, p. 176, the first hand of Codex B is supposed to have written εορακεσε: “the final ε has been erased, and the ε preceding it changed into α.” I have examined the photograph of B in the Stuttgart Library, and can find no trace of an ε ever having stood after σ. The blank space of the size of two letters is meant to divide the sentences. It is the case, however, though neither Tischendorf, Fabiani, nor the pamphlet of 1881 mentions it, that the first hand wrote εορακες, which was then made into εωρακας by means of a stroke drawn through the ο. The matter is not insignificant in view of what is said in Westcott and Hort’s Notes on Orthography, Appendix, p. 168. Burkitt supposes that εορακεσε was the reading of the ancestor of א B (Texts and Studies, vol. v. 5. p. ix).^[276]

290xii. 7. τετήρηκεν, without ἵνα, has the support of a comparatively large number of manuscripts. Peerlkamp and De Koe read ἵνα τί ... τετήρηκεν; Zahn (Einl., ii. 518) has no doubt that the correct reading is ἵνα ... τηρήσῃ, and that it was replaced by τετήρηκεv (without ἵνα) on the ground that this Mary was not among the women who came to the sepulchre to anoint the body of Jesus. He says that the true text presupposes that Mary would like to use the remainder of the ointment to anoint the body of Jesus after his death, and that the words of Jesus were intended to prevent Mary and the disciples afterwards following the suggestion of Judas.

xiii. 2. The change of a single letter here is important from a harmonistic point of view. א* B L read δείπνου γινομένου, i.e., “during supper,” but א^c A D have δείπνου γενομένου, which means “after supper.” Compare Zahn, Einleitung, ii. 520.

xiii. 34. On the form in which this saying was cited by the Marcionites, see Zahn, GK. i. 678.

xviii. 12 ff. The Sinai-Syriac, probably following Tatian, gives the following arrangement of the verses—viz., 12, 13, 24, 14, 15, 19-23, 16-18, 25-28. On this see Zahn, Einl., ii. 521. Spitta would arrange the verses, 12, 13, 19-23, 24, 14, 15-18, 25b, 27, 28. See Zahn, Einl., ii. 558. It does sometimes happen that a leaf of a manuscript is misplaced, but it is hard to account for such transpositions as these. Compare the Journal of Theological Studies, October 1900, p. 141 f.

xix. 5. Though not properly connected with the criticism of the text, the question may be asked here, by way of a contribution to a subject much discussed of late, whether the expression ἰδοὺ ὁ ἄνθρωπος may not be connected with בר נשא. Compare ὁ ἄνθρωπος and ὁ υἱὸς τοῦ ἀνθρώπου in Mark ii. 27, 28. In this passage of John, B omits the article before ἄνθρωπος, reading ἰδοὺ ἄνθρωπος simply. See the Expository Times, November 1899, p. 62 ff., “The Name Son of Man and the Messianic Consciousness of Jesus,” where Schmiedel’s article with the same title in the Protestantische Monatshefte is noticed.

291xix. 37. The quotation, according to Zahn, is made from the Hebrew. The LXX has ἐπιβλέψονται πρὸς μὲ ἀνθ’ ὧν κατωρχήσαντο. The later Greek versions all seem to have kept the first three words as in the LXX but to have variously corrected the second clause, for which Aquila gives σὺν ᾧ ἐξεκέντησαν, Theodotion εἰς ὃν ἐξεκέντησαν, Symmachus ἔμπροσθεν ἐπεξεκέντησαν. Compare with this Apoc. i. 7, οἵτινες αὐτὸν ἐξεκέντησαν; Barnabas vii. 9, ὄψονται αὐτὸν ... κατακεντήσαντες; Justin, Dial. 32, ἐπιγνώσεσθε εἰς ὃν ἐξεκεντήσατε. It has accordingly been supposed that John in the Gospel and Apocalypse followed some unknown Greek version which exhibited the characteristic forms ὄψονται (found only in John and Barnabas) and εἰς ὃν ἐξεκέντησαν (given by John, Justin, Theodotion, and partly by Aquila). But this supposition is simply a proof of unwillingness to admit a palpable fact—viz., that in the Gospel and Apocalypse John gives an independent rendering of the original text of Zechariah xii. 10, and that Barnabas and Justin follow John. See Zahn, Einleitung, ii. 563.

The subscriptions state that the fourth Gospel was written thirty or thirty-two years after the Ascension, at Ephesus, in the reign of Nero, or, as some say, of Domitian. It is also said to have been published by Gaius, the host of the Apostles (διὰ Γάϊον τὸν ξενοδόχον τῶν ἀποστόλων). Others say that it was dictated to Papias of Hierapolis the disciple of the Apostle. On the alleged autograph (ἰδιόχειρον) preserved at Ephesus, see above, p. 30.

ACTS.

It would unduly enlarge the extent of this work were I to go on mentioning all the passages in the Acts that are more or less striking from the textual point of view. This book has already been more frequently referred to than the others. I would again refer the student to Zahn’s Introduction. I agree with that writer in thinking it impossible in many cases to 292suppose that a scholiast manufactured the text we now find in Codex D with no other material before him save the usual text and his inkhorn. At the same time there is undoubtedly room for much diversity of opinion with respect to many matters of detail. Ι would instance such a simple narrative as that of Acts iii. 1-5, and ask what reasonable ground a copyist could have had for altering ὃς into οὗτος, ἀτενίσας into ἐμβλέψας, βλέψον into ἀτένισον, ἐπεῖχεν into ἀτενίσας or vice versa, or for omitting or inserting ὑπάρχων or λαβεῖν.^[277] Such changes might, however, be introduced by an author who writes a passage twice over. Without himself being fully conscious of his reasons for doing so, he might substitute a final construction for a participle, introduce or remove an asyndeton, replace one word by its synonym, and make all the striking linguistic changes which a comparison of the two texts reveals.

Time will show whether I am right in my conjecture that ἐβαρύνατε in iii. 14 is due to an error in translation. In illustration of the interchange of λαοῦ and κόσμου in ii. 47, I have cited in Philologica Sacra, p. 39, a number of instances of the confusion of עם or עמא with עלם or עלמא, to which I would now add Daniel viii. 19, Sirach xlv. 7; xlvii. 4; Matt. i. 21 (in the Curetonian Syriac). Compare also Eusebius, Eccles. Hist., iv. 15, 26; History of Mary, ix. 17; xiv. 11 (ed. Budge). Whether the change in the passage in question is really to be explained in this way, or by the supposition of an “anti-Judaic tendency,” as Corssen prefers (Göttinger gelehrte Anzeigen, 1896, vi. 444), may be left an open question for the present. I would just point to one thing in favour of my view, and in answer to what Zahn says in his Einleitung, ii. 423. He says there: “Linguistic considerations are against the supposition that a pure Greek like Luke, the physician of Antioch, was 293able to read a Hebrew book. For a thousand Jews (Syrians and Copts) who were able at that time to read, write, and speak Greek, there would be at most a single Greek possessed of a corresponding knowledge of Hebrew or Aramaic. And I confess that I have hitherto sought in vain for this rara avis.” Quite true, but how do we know that the physician of Antioch was a pure Greek? All the Prologues to the Gospels unanimously call him “natione Syrus.” I have pointed out in my Philologica Sacra, p. 13, what is very generally admitted, that in the New Testament Ἕλληνες denotes simply the “heathen,” whether they speak Greek or not.^[278] The woman mentioned in Mark vii. 26 was a Ἑλληνὶς Συρο-Φοινίκισσα τὸ γένος, and in the same way Luke was a Ἕλλην of Antioch (Acts xi. 20), but Σύρος τὸ γένος. He is one of the thousand who could read, write, and speak Greek, though he was not above making such a mistake in translation when using a Hebrew or Aramaic book as I think he certainly does in Luke xi. 41, and as I am inclined to think he does in Acts iii. 14, till I find a better explanation of the reading ἐβαρύνατε than has yet been given.^[279] I am glad to see from Zahn that more than seventy years ago, in his dissertation entitled De Codice Cantabrigiensi (1827), p. 16, Schultz suggested that the text of D may perhaps be 294derived from a Syriac version. According to what I have said above on Tatian, this view must certainly be admitted as possible, and I see that it has been revived by Chase.