241. Compare what Westcott and Hort say of Whiston and Bornemann, cited above, p. 222, and particularly the section on the twofold recension of the Acts in Zahn’s Einleitung, ii. § 59, pp. 338-359. See also Burkitt’s Introduction to Barnard’s Biblical Text of Clement of Alexandria (Texts and Studies, v. 5), especially p. xviii: “Let us come out of the land of Egypt, which speaks, as Clement’s quotations show, with such doubtful authority, and let us see whether the agreement of East and West, of Edessa and Carthage, will not give us a surer basis upon which to establish our text of the Gospels.”
242. Attention may be directed in passing to the interesting way in which the witnesses are distributed. Thus we have in verse 51, for the omission of καὶ ἀνεφερ. εἰς τ. οὐρ. א* D Syrsin a b d e ff l*, Aug. ½; verse 52, omit προσκυν· αὐτόν, D Syrsin a b d e ff l, Aug. 1/1; verse 53, αἰνοῦντες for εὐλογοῦντες D a b d e ff l r (Aug.); (Syrsin here has מברכין, not משבחין which represents αἰνοῦντες in Luke ii. 13, 20, xix. 37, and, therefore, must have read εὐλογοῦντες in this passage). Now I ask, is it right to accept the testimony of D and its associates in verse 52, only to reject it in verse 53? And what amount of weight is added to the testimony of D by the addition of that of א*? Schiller says in Tell: “The strong is mightiest alone: united e’en the weak are strong”—how far are both these notions true in textual criticism?
243. So Graefe, but it is not apparent whether the καὶ that belongs to this reading is to be supplied before it or after. Evidently he intends to read αἰν. καὶ εὐλογ. with the great majority of witnesses, and not εὐλογ. καὶ αἰν. with the Ethiopic version. See Th. St. Kr., 1898, i. 136 f. The passage is regarded by Westcott and Hort as a good example of “conflation,” § 146.
244. See now Textkritik der vier Evangelien, pp. 48, 181.
245. Αἰνεῖν is not given in Cremer’s Dictionary among the synonyms of εὐλογεῖν, and is only cited on p. 610 with the reading αἰνοῦντες καὶ εὐλογοῦντες from this passage.
246. See Hilgenfeld, Das Apostel-Concil nach seinem ursprünglichen Wortlaute in the ZfwTh., 42 (1899), 1, 138-149.
247. Harnack, Das Aposteldecret (Acta xv. 29) und die Blass’sche Hypothese, Berlin, 1899. From the Sitzungsberichte der preuss. Akad. der Wiss. Noticed in the Expository Times for June 1899, p. 395 f. See also the Berliner philologische Wochenschrift of the 13th May 1899.
248. See critical note in the Expositor’s Greek Testament (Knowling), in loco.
249. Cf. the passage of Hermas cited above, p. 47.
250. A good example is seen in Ezek. xvi. 3. In the Sixtine edition of 1586 a new page (692) occurs in the middle of the sentence διαμαρτυρον τη Ιερουσαλημ τας ανομιας αυτης ταδε λεγει κυριος τη Ιερουσαλημ, with the result that the eight words from the first Ιερουσαλημ to the second are printed twice by a recessive homoioteleuton, while in Codex 62 they have dropped out altogether owing to a forward error of the same sort. The former mistake is tacitly corrected in all reprints, but the latter could not be detected from the context alone without other testimony. Compare also Mark ix. 10 in codex T of the Vulgate and ff of the Old Latin. In the former the passage from resurrexit to resurrexit is repeated, in the latter it is omitted.
251. I had a teacher once who invariably tried to get over any difficulty in the Greek classics by saying that the text was corrupted by homoioteleuton. We did not always agree with him; he was perhaps a little too ready with this way out of a difficulty, but any one with experience knows how very apt this mistake is to occur.
252. This applies to printed editions as well as to manuscripts. Van Ess’s reprint of the Sixtine Septuagint (1824) is very carefully done, yet five words have dropped out in Joel iii. 9. These are omitted in all the later editions of 1835, 1855 (novis curis correcta), 1868, and 1879, and were only supplied by myself in 1887 on the occasion of the third centenary of the Sixtine edition. They are omitted in Tischendorf’s first edition of 1850, and also in the second of 1856.
253. In ancient times people always read aloud, even when reading by themselves.
254. Scrivener would explain the “remarkable confusion” of the two prepositions προ and προσ, when compounded with verbs, which we meet e.g. in Matt. xxvi. 39; Mark xiv. 35; Acts xii. 6; xvii. 5, 26; xx. 5, 13; xxii. 25, by saying that the symbol [Symbol: an abbreviation made of a rho character written on top of a block with three lines.] is used indifferently for προ and προσ in the Herculanean rolls, and here and there in Codex Sinaiticus. Seeing that it has become a bad habit in Hebrew Grammars to speak of Aleph prostheticum instead of protheticum, and that the practice is still defended (Gesenius-Kautzsch26, p. 64, n. 3, “rightly so”) after my notice of it (Marginalien, p. 67), I have given some little attention to this confusion, and could cite dozens of examples. Others, of course, have noticed it as well as myself. In his N.T., i. 20, B. Weiss says: “The compounds with προ and προσ are interchanged quite heedlessly,” and he cites in proof of this eight passages from the Acts. He writes similarly in ii. 34. I shall instance only one or two cases in connection with this same word πρόθεσις. Pitra on Apost. Const., 5, 17 (p. 325): πρόθεσιν restituimus cum Vatican. 2, 3, 4, 5, vulgo πρόσθεσιν; Excerpta Περὶ Παθῶν, ed. R. Schneider (Programme of Duisburg, 1895), where the manuscripts deviate in five passages, pp. 5, 14. 20; 6, 5; 13, 7. 13, and we read in § 10, ἀντίκειται δὲ πρόσθεσις μὲν ἀφαιρέσει, etc., and in § 11, πρόσθεσις μὲν οὖν ἐστὶ προσθήκη στοιχείου κατ’ ἀρχήν, οἷον σταφίς, ἀσταφίς καὶ ὀσταφίς. Both times, of course, it should be πρόθεσις, as the better manuscripts have it. Wherever mention is made of the “shewbread,” D invariably turns it into “extra bread,” by reading προσθέσεως instead of προθέσεως. Tischendorf first called attention to this in Luke vi. 4, but it occurs also in Matt. xii. 4. I have no doubt myself that in the case of verbal forms, the σ was inserted in order to avoid the hiatus before the augment. Compare προσέθηκεν for προέθηκεν, Ex. xxiv. 23; προσέθηκας, Ps. lxxxix. 8, Symmachus; προανεθέμην or προσανεθέμην, Gal. i. 16. In Wisdom, vii. 27, the first hand of Sinaiticus even writes προσφήτας for prophets. It is disputed whether the title of one of Philo’s books is προπαιδεύματα or πρὸς [τὰ] παιδεύματα. Etc. etc. Sapienti sat.
255. We find all the possible permutations of the words αὐτοῖς ἐλάλησεν ὁ Ἰησοῦς in John viii. 12. See my note on Codex Purpureus Petropolitanus (N) in Hilgenfeld’s Zeitschrift für wissenschaftliche Theologie, 42 (1899), p. 623.
256. See Harnack, Das Magnificat der Elisabeth (Lukas i. 46-55) in the Berliner Sitzungsberichte of the 17th May 1900, p. 538 ff. A good example of how glosses may creep into the text is afforded by Philo “Quod det.” 11 (Cohn, 1, 266).
257. On the influence of a system of pericopæ on the text of Codex D, see Scrivener’s Introduction to his edition of the manuscript, p. li, and Zahn, Einl., ii. 355.
258. See Semler’s edition of J. J. Wetsteinii libelli ad crisin atque interpretationem N.T., Halae, 1766.
259. (1) Lectio quae in veteribus latinis non apparet probabilior est. (2) Codices qui cum graecis א B L concordant plerumque textum Hieronymianum ostendunt.
260. In view of the frequency with which the witnesses fluctuate between ἡμῶν and ὑμῶν, ἡμῖν and ὑμῖν, etc., it is impossible to adjust their claims on any mere arithmetical principle. Zahn (Einl., ii. 61) calls attention to an important consideration in support of the reading ὑμῖν in 2 Peter i. 4, which applies to other passages as well—viz., “that when the New Testament epistles were read at divine service, ἡμεῖς would very readily and very frequently be substituted for ὑμεῖς, which excluded the reader or preacher.” Compare Acts iv. 12: ἐν ᾧ δεῖ σωθῆναι—, ἡμᾶς or ὑμᾶς?
It might be laid down as a second rule in this connection, that particular importance attaches to those versions in which the distinction of the persons does not depend simply on a single letter but on a separate word (nobis: vobis, etc.). In versions of this sort the original reading is preserved from the first; in the case of the others, the change could be made at any point of the transmission, especially when it was helped by the nature of the writing, which must also, of course, be taken into account.
A glance over the verse enumeration in the margin of one of the modern editions of the text will reveal, perhaps, most clearly how strong is the tendency to interpolation. Of the verses into which Stephen divided the Greek N.T. (1551), the Stuttgart edition omits entirely the following from the Synoptic Gospels—viz., Matt. xviii. 11 (xxi. 44, Tischen.), xxiii. 14; Mark vii. 16; ix. 44, 46; xi. 26; xv. 28; Luke xvii. 36 (xxi. 18, W-H margin); xxiii. 17 (xxiv. 12, 40, Tisch.). Compare also Matt. xx. 28; xxvii. 35, 38, 49; Mark vi. 11; xiii. 2; Luke vi. 5; ix. 55; xii. 21; xix. 45; xxi. 38; xxii. 19 f., 43 f., 47; xxiii. 2, 5, 34, 48, 53; xxiv. 5, 36, 51, 52. In the case of several verses this or that part had to be omitted. Luke xx. 30, e.g., is reduced to the three words, καὶ ὁ δεύτερος, with the result that it becomes the shortest verse in the N.T.
261. The best discussion of the form Ἀπελλῆς will again be found in Zahn, Einl., i. 193.
262. See my note in the Expository Times for July 1900, p. 478, where I have brought forward a new witness for the reading Jonatha—viz., Jerome’s Liber interpretationis Hebraicorum nominum. He explains the word as “columba dans vel columba veniens.”
263. Volck has an article of four and a half pages on Balaam in the PRE3, iii. 227 ff., but he says not a syllable about the form βοσόρ, which is too bad. In Hastings’ Dictionary of the Bible it is at least mentioned though not explained.
264. בער is explained as the Hebrew form of the Aramaic בעור by C. B. Michaelis (De Paronomasia, § 30); Hiller, Onomasticum, 1706, p. 536; and Bernardus (in Marck, In praecipuas quasdam partes Pentateuchi Commentarius, Leyden, 1713, 366). Marck himself makes it the equivalent of פתור. M. M. Kalish, Bible Studies, i. The prophecies of Bileam, London, 1877, contributes nothing to the solution of the question.
265. See Mrs. Lewis, in the Expository Times, November 1900, p. 56 ff., What have we gained in the Sinaitic Palimpsest? I. St. Matthew’s Gospel, where a number of important variants are cited from that manuscript.
266. Corderius (Caten. Psal., ii. 631) substitutes “ancient” for “accurate.”
267. Or “exemplo graeco,” according as the plural points are inserted or not. The passage is printed in Syriac by Cureton, p. xxxvi, who says that it is also found in the margin of the London manuscript of the Peshitto, 14456. He also gives the verses in which Juvencus paraphrases this text.
268. The other variations of the Latin witnesses are extremely instructive—viz.:
| locis eminentioribus | superioribus | g2 emm. | honorificis | m |
| clarior | dignior | d m g2 emm. | honoratior | e |
| deorsum | inferius | g2 emm. | infra | m |
| inferior | humilior | minor | ||
| superius | sursum | in superiori loco. | ||
| utilius | utile | gloriam. |
This variety is an indication of the early age at which the text was translated into Latin.
269. The Thesaurus Syriacus does not contain the word either in col. 1405 under חשמיתא, or in col. 2205 under מרא.
It may also be observed in passing, that the passage is one of those whose sense is entirely changed by the insertion or omission of the negative in this or that witness (see below on Gal. ii. 5). Instead of καὶ ἐκ μείζονος, Syrcu reads καὶ μὴ ἐκ μείζονος. Moreover, it takes ζητεῖτε as imperative, a fact that Tischendorf has failed to notice.
270. I see that Chase, who discusses the passage in pp. 9-14 of his Syro-Latin Text, has the same impression: “the compound Greek word in D, ὁ δειπνοκλήτωρ, seems intended to represent the Syriac expression ‘the lord of the supper.’”
271. The case is quite different in 1 Macc. v. 2, where the first hand of א wrote ὠργίσθησαν for ἐβουλεύσαντο. Here ὠργίσθη occurs immediately before it.
272. So given in Merx’s edition, but not in Lewis.—Tr.
273. See also von Dobschütz, Das Kerygma Petri, p. 82, where he cites the passage of Origen relating to the Doctrina Petri, which is also quoted by Tischendorf on Luke xxiv. 39, and insists rightly that in the LXX δαιμόνιον is never employed to represent דוּחַ. Conybeare’s articles on “The Demonology of the New Testament” in the Jewish Quarterly Review (1896) I have unfortunately been unable to consult. Joh. Weiss never mentions φάντασμα in his article on Dämonen und Dämonische in the PRE3, iv.
274. This description is elsewhere understood as applying to Theodore of Mopsuestia.
275. The same variation occurs in Luke vi. 7, where א* B S X read κατηγορεῖν (κατηγορῆσαι D), while אc A E F have κατηγορίαν.
276. When examining Codex B I took occasion to look at certain other passages, and discovered some strange mistakes in Tischendorf’s statements with regard to that manuscript, as I did previously in the case of Codex D. In 3 John 13 B has ἀλλὰ for ἀλλ’, on which Tischendorf has no note. Westcott and Hort mention the passage in their Notes on Orthography, ii. p. 153, but say nothing about B. On Jude 5 we find Tischendorf saying in his Apparatus: ειδοτας sine υμας cum A B C2 ... ϛ (Gboo) add. υμας cum א K L. But υμας stands quite plain in B. Had they known this, Tischendorf and Westcott and Hort would certainly have printed their text differently. How far back this false testimony with regard to B extends I am unable to certify. It is found in Tischendorf’s seventh edition of 1859, and in Huther’s Commentary of the same date. I repeat my Ceterum censeo, that two or three sharp eyes should really revise the statements current about B. This one is repeated from Tischendorf by Baljon, Weiss, and all our Commentators. At the same time, Weiss has quite properly inserted υμας in his text, on the ground that while there was no occasion for its interpolation, its omission is quite conceivable. He will, no doubt, be gratified to see his reasoning confirmed by this weighty testimony afforded by Codex B.
277. Further instances of changes requiring investigation are:—ερωταν and παρακαλειν; οραν and θεασθαι; αγειν and φερειν; ερχεσθαι and υπαγειν; υπαρχειν and ειμι; συν and μετα; εις and εν; εως, μεχρι and αχρι; ενωπιον and εμπροσθεν; ετερος and αλλος; οικος and οικια; παις and παιδιον; πολις and κωμη; λαος and οχλος; ναος and ιερον; φεγγος and φως; active and middle voice, αρχεσθαι, etc.
278. See, however, Romans i. 14, and compare Zahn, Einl., i. 263.—Tr.
279. I have already (p. 37) referred to the frequency with which mistakes, often quite incredible mistakes, in translation occur. A few additional instances may be cited here.
There is, for example, that of Ephraem in John ii. 2, mentioned above, p. 287.
According to Aphraates 41, 16, Jesus promised to the mourners דלהין נתבשפון, i.e. that they should be entreated. The writer of the text, therefore, that Aphraates used, must have taken παρακαλεῖν here in the sense of “to entreat.” See Zahn, Forschungen, i. 78.
The same writer (383, 16) renders the words in Luke xvi. 25 νῦν δὲ ὅδε παρακαλεῖται in the form יומנא דין בעית מנה, i.e. “but to-day thou entreatest of him”, where παρακαλεῖν is again taken in the sense of “to entreat”, though a different word is used for it. See Zahn, ibid.
Again, Aphraates (390, 4) renders παράκλησιν (αὐτῶν) in Luke vi. 24 בעותהין, “their prayer, their request.” Zahn, ibid.
The last clause of John v. 14 is rendered “that thou mayest have need of nothing else,” where χρεία must have been read instead of χεῖρον. Zahn, Forschungen, i. 161 f. Compare also the Syriac text of Apoc. ii. 13; viii. 13, etc.
280. On the custom of dictating letters, see Norden, Die antike Kunstprosa (1898), p. 954 ff. On the autograph additions to the letters of the Emperor Julian, see Bidez and Cumont, Recherches etc., p. 19 (see above, p. 174).
281. What an amount of perplexity would have been avoided had Paul been in the way of dating his letters exactly, or had the copyists preserved the dates, supposing they were there originally! One, but only one, of the epistles of Ignatius bears a date—viz. that to the Romans: ἔγραψα ὑμῖν ταῦτα τῇ πρὸ ἐννέα καλανδῶν Σεπτεμβρίων (x. 3).
282. On these two verbs compare Exod. xxxi. 17, where the LXX has ἐπαύσατο and Aquila ἀνέψυξε; Isa. xxxiv. 14, LXX ἀναπαύσονται, Aquila ἀνέψυξε; Isa. xxviii. 12, Aquila ἀνάψυξις; compare also ἀνάπαυσις, Matt. xi. 29; καιροὶ ἀναψύξεως, Acts iii. 20. Weiss, in his Commentary, ignores the reading of B in Rom. xv. 32; in his discussion of the text he supposes that the text was mutilated by a translator, and that D E F G “sought to restore it in their own way.”
283. See 1 Cor. v. 8, where the Syriac has קדישותא—i.e. ἁγιότητος—for ἀληθείας.
284. To these examples, gathered quite incidentally, one might add as many from the Old Testament and other books if one paid any attention to them in reading. Take, for example, Herodotus i. 24. Was the votive offering that Arion set up at Taenarum μέγα or οὐ μέγα? In the Germania xv. 1 did Tacitus say of the Germans “non multum venationibus, plus per otium transigunt,” or “multum venationibus, etc.”? In the new edition of Origen (i. 87, 16) Koetschau reads ἀχρήσιμα where the manuscript and the earlier editions have χρήσιμα, and he lets an οὐκ stand which others omit, etc.
285. In his Lucian Lagarde gives examples of his being deceived by certain letters shining through from the opposite side—e.g. Esther v. 22 and 27. This latter is a most interesting case. The following verse begins with μη, and Lagarde thought that the first scribe had added another μη by mistake and afterwards erased it, whereas it turned out that what he took to be ΜΗ was nothing else than ΗΝ shining through from the other side.
286. Cf. Westcott, Canon, Part II. ch. ii. § 1, p. 354 ff.; Bible in the Church, p. 125 f.
287. Cf. Westcott, Bible in the Church, p. 153 ff.
288. See Article on The Catholic Epistles, by Salmond, in Hastings’ Dictionary of the Bible, i. p. 359 f.
289. It is certainly difficult to construct an Apparatus which shall be concise and yet clear. In Jude 22 Baljon adopts ἐλεᾶτε in the text, and yet he leaves the apparatus arranged in such a way as to suggest that he intended to read ἐλέγχετε with Tischendorf.
290. The inscription on a column at Rome dedicated to a Sabine god which Justin read as “Simoni Sancto Deo,” and understood as referring to Simon Magus. See Kurtz, Church History (Macpherson), i. p. 97; Neander, Church History (Bohn), ii. p. 123, note.
291. Compare also the articles on 1 and 2 Peter by the same writer in vol. iii.
292. See Blass, Grammar of N. T. Greek, § 31, 6, Eng. Tr., p. 80 f. Compare the similar German idiom used in the titles of books, “von X. Y. ordentlicher Professor.” How naturally this comes to a Hebrew is shown by the fact that Sal. Bär, in his translation of the Massoretic note at the end of the books of Samuel (Leipzig, Tauchnitz, 1892, p. 158), among other lovely things has “ad mortem Davidis rex Israelis.”
293. In this (independent) suggestion I am glad to find myself in agreement with Lachmann (Studien und Kritiken, 1830, p. 839), and Westcott and Hort (ii., App., 137). I see that Baljon and Zahn too follow it. But Bousset still writes ἡμέραις αἷς.