i. 9, 11. Considerable additions are made to the text after both these verses by Codex 109. This manuscript is numbered 11 in the Library of St. Mark at Venice, and described by Gregory as “haud malae notae.” It contains both a Latin and an Arabic version, and dates from the thirteenth or fourteenth, or, as some suppose, the eleventh, century. After verse 9 we read: μὴ χειροτονεῖν διγάμους μηδὲ διακόνους αὐτοὺς ποιεῖν, μηδὲ γυναῖκας ἔχειν ἐκ διγαμίας· μηδὲ προσερχέσθωσαν ἐν τῷ θυσιαστηρίῳ λειτουργεῖν τὸ θεῖον· τοὺς ἄρχοντας τοὺς ἀδικοκριτὰς καὶ ἅρπαγας καὶ ψεύστας καὶ ἀνελεήμονας ἔλεγχε ὡς θεοῦ διάκονος. After verse 11 we find τὰ τέκνα τοὺς ἰδίους γονεῖς ὑβρίζοντας ἢ τύπτοντας ἐπιστόμιζε καὶ ἔλεγχε καὶ νουθέτει ὡς πατὴρ τέκνα.

Subscription: πρὸς Τίτον (+ τῆς Κρητῶν ἐκκλησίας πρῶτον ἐπίσκοπον χειροτονηθέντα) ἐγράφη ἀπὸ Νικοπόλεως τῆς Μακεδονίας (missa per Arteman: al. per Zenam et Apollo).

Hebrews.

i. 3. Instead of φέρων, the first hand of B wrote φανερῶν, which a second hand altered to φέρων, while a third restored φανερῶν, and wrote in the margin ἀμαθέστατε καὶ κακέ, ἄφες τὸν [? τὸ] παλαιόν, μὴ μεταποίει. A great deal of material might be collected from the margin of old manuscripts, not only for the history of Prayer, as von Dobschütz recently observed, but for other interesting departments of the history of civilisation.

ii. 9. The reading χωρὶς θεοῦ instead of χάριτι θεοῦ is now found only in M and in the second hand of 67. Origen, however, was aware of the various reading: χωρὶς θεοῦ ἢ ὅπερ ἔν τισι ἀντιγράφοις χάριτι θεοῦ. It seems to be a primitive transcriptional error.

x. 34. We have here to choose between δεσμοῖς and δεσμίοις. The latter is manifestly the correct reading. It is attested by A D* and certain minuscules, among which are 37**, 67**. This last is a Vienna manuscript (Vindob. gr. theol. 302), whose marginal readings exhibit a text closely resembling that of the uncials B M, which are defective in Hebrews x. Δεσμοῖς μου is supported by א D^c H K L P, Clem. Alex., Origen (i. 41, where, however, μου is omitted by M* P, according to Koetschau’s new edition), and by d e (vinculis eorum). Zahn (Einleitung, ii. 122) thinks that the connection of the reading δεσμοῖς μου with the tradition of the Pauline authorship of the epistle is suspicious. We find the reading adopted in those regions where the tradition was accepted. It may, however, have been the means of confirming and spreading the tradition, seeing that Clement of Alexandria is actually aware of it. Pseudo-Euthalius, e.g., employs the reading in support of the Pauline authorship (Zacagni 670).

In this same verse א A H have preserved the proper reading ἑαυτοὺς. Ἑαυτοῖς, as given by D E K L, is a would-be correction.

xi. 23. In certain manuscripts (D and three Vulgate codices) an entire verse is inserted after verse 23: Πίστει μέγας γενόμενος Μωϋσῆς ἀνεῖλεν τὸν Αἰγύπτιον κατανοῶν τὴν ταπείνωσιν τῶν ἀδελφῶν αὐτοῦ. Its position shows it to be an interpolation.

xiii. 9. The present tense περιπατοῦντες is exhibited only by א* A D*, all the other witnesses having περιπατήσαντες. The minority are in the right here. A correction is not always an improvement.

xiii. 18. Zahn accepts the καὶ before περὶ ἡμῶν. It is found only in D d and Chrysostom. This combination of witnesses is very rare.

Subscription: ἐγράφη (+ ἑβραϊστὶ 31) ἀπὸ τῆς Ἰταλίας διὰ Τιμοθέου: al. ἀπὸ Ἀθηνῶν: al. ἀπὸ Ῥώμης.

CATHOLIC EPISTLES.

The variety in the order of the Catholic Epistles is even more significant than that of the Pauline. When the Syrian Church of Edessa obtained the New Testament, it consisted only of the Gospels, the Pauline Epistles, and the Acts. It contained neither the Apocalypse nor the Catholic Epistles. This is proved among other things by the fact that not a single quotation from these writings is found in the Homilies of Aphraates, the date of which falls between 336 and 345. At a later date the Syrian Church accepted the Epistle of James, 1 Peter, and 1 John, but the four so-called Antilegomena—viz. 2 Peter, 2 and 3 John, and Jude—are to this day excluded from their Canon of the New Testament. Even in the West, James was not reckoned among the books of the New Testament previous to the fourth century. There is no mention made of it in Africa about the year 300, although it was cited at Rome and Carthage at an earlier date. At Alexandria, however, all the seven Catholic Epistles were counted in the New Testament as early as the time of Clement,^[286] and their place in the Canon becomes more and more firmly assured from the time of Eusebius onwards.^[287] At the same time, the order of their arrangement varies very considerably. Indeed, every possible variety occurs, except that Jude seems never to have been placed first, nor 2 Peter last. Thus we find James, 2 Peter, 3 John, Jude; James, Jude, 2 Peter, 3 John; 2 Peter, James, Jude, 3 John; 2 Peter, 3 John, James, Jude; 2 Peter, 3 John, Jude, James, etc.^[288] It follows that in the case of this group of New Testament writings, as well as in that of the preceding, it is necessary and possible to distinguish the three longer from the four shorter epistles in tracing the history of the text. And we see at the same time what justification Luther had in drawing a line between these epistles and the principal books of the New Testament as having been held in quite a different estimation in early times.

iii. 22. After θεοῦ the Vulgate inserts deglutiens mortem ut vitae aeternae haeredes efficeremur, “apparently from a Greek original which had the aorist participle καταπιών; cf. 1 Cor. xv. 54” (W-H, Notes, in loco). See Vetter, Der dritte Korintherbrief.

2 Peter.

i. 1. Zahn considers ἐν δικαιοσύνῃ the original reading, and εἰς δικαιοσύνην a later correction due to taking πίστιν ἐν δικαιοσύνῃ together as “faith in righteousness.” The last two words are to be taken with λαχοῦσιν. Einleitung, ii. 59.

i. 2. Zahn agrees with Lachmann and Spitta in holding that ἐν ἐπιγνώσει τοῦ κυρίου ἡμῶν is the correct text here—that is to say, he omits τοῦ θεοῦ καὶ Ἰησοῦ. Tischendorf’s Apparatus is very diffuse on this verse, and Baljon’s note, which is extracted from it, is accordingly not quite satisfactory.^[289] Like all the other editors, he gives ἐν ἐπιγνώσει τοῦ θεοῦ καὶ Ἰησοῦ τοῦ κυρίου ἡμῶν in the text, but the only variants he mentions are the insertion of Χριστοῦ after Ἰησοῦ, and the omission of ἡμῶν. There is no notice of the omission of the words τοῦ θεοῦ καὶ Ἰησοῦ by any of the witnesses. They are not found in P, the best manuscripts of the Vulgate (am fu dem harl), Philoxenian and Harklean Syriac, nor in minuscules 69, 137, 163. These last, however, the Syriac and the minuscules with m, insert Ἰησοῦ Χριστοῦ after ἡμῶν. Kühl believes that the shorter form is probably due to the fact that in the epistle Christ is everywhere regarded as the object of ἐπίγνωσις. But this is really very improbable. For the scribe could not have been aware of this when he began to write the epistle, so that he must have turned back and deleted the words καὶ θεοῦ καὶ Ἰησοῦ afterwards. At the same time it is a fundamental principle of textual criticism that the lectio brevior is to be preferred. Reference may be made to the Epilogus of Wordsworth and White, ch. vi., De regulis a nobis in textu constituendo adhibitis, where the very title of section 4 implies this principle: “Cum brevior lectio probabilior sit, codices A F H* M Y plerumque praeferendi sunt,” and where the most conspicuous examples of this rule are said to be “Additamenta nominum propriorum, et praecipue sanctorum—e.g. Jesus, Christus, Dominus, Deus.” It is true that in the passage before us we have not simply a case of the insertion of a word or words understood; at the same time, here if anywhere the text is more likely to have been extended than abbreviated. It remains to be seen whether P exhibits a good text in other passages of the Catholic Epistles as well as this, but so far as the minuscules 69 and 137 are concerned, they justly bear a good reputation. Hort calls 69 one of the better cursives, and 137 has a text so closely resembling that of Codices D E as to be of material assistance when these are defective. The minuscules are too often regarded as mere ciphers; as if a cipher more or less behind a number did not make a vast difference. In the very next verse we find 137 supporting א A in reading τὰ πάντα, which is accepted by Tischendorf and Weiss, and preferred also by Kühl. In this instance it contradicts P, which omits τὰ with B C K L.

i. 12. Here μελλήσω is given by א A B C P, οὐ μελλήσω by 8 f tol (non differam), and οὐκ ἀμελήσω by K L etc. (“the Antiochean recension and the Syriac versions,” Zahn). “Μελλήσω, with the present infinitive, can hardly be simply a periphrastic future. The idea is rather that the writer will be prepared in the future, as well as in the past and in the present, to remind them of the truths they know, whenever the necessity arises. As they had no evidence of the fulfilment of this promise, the copyists and translators found a difficulty with this expression, and hence the variants.” Zahn, Einleitung, ii. 53 f.

i. 15. The reading σπουδάζω, found in א 31, and the Armenian, is also attested by the Philoxenian Syriac, a fact which Zahn regards as important. “On transcriptional grounds the reading σπουδάσω, preferred by our editors, would appear to be confirmed by the reading σπουδάσατε, exhibited by the Harklean Syriac and a few minuscules. But in reality both these latter readings merely serve to show that a difficulty was felt again in admitting a promise on the part of Peter which he seemed never to have fulfilled.” Einleitung, ii. 54. Compare on μελλήσω above.

i. 21. It is probable that Theophilus of Antioch (Ad Autolycum, ii. 9) read (οἱ) ἅγιοι (τοῦ) θεοῦ ἄνθρωποι, the form exhibited by א and A (“the chief representatives of the Antiochean family”), and also by several Latin witnesses. See Zahn, GK. i. 313; Chase on 2 Peter in Hastings’ Dictionary of the Bible, vol. iii. p. 801.

ii. 13. On this passage Zahn remarks (Einleitung, ii. 53): “The similarity of 2 Peter to the Epistle of Jude was doubtless a source of textual corruption. But it may also aid us in correcting the text. Because, whichever of the two we regard as the original, in any case the one is our earliest witness to the text of the other. If we accept the reading ἀγάπαις in Jude 12, it follows either (1) that Jude read ἀγάπαις in 2 Peter, and that this is the original reading there, or (2) that Peter, supposing he wrote second, altered Jude’s ἀγάπαις to ἀπάταις, which it is hard to conceive, the former being so unmistakable, and the latter much less suitable to the context. In either case, therefore, ἀγάπαις would seem to be the correct reading in 2 Peter ii. 13.” No doubt the alteration of ἀγάπαις to ἀπάταις is “hard to conceive,” but it is not inconceivable. As illustrating how a piece of writing may be misread, it is sufficient to point to Justin’s mistake with regard to “Semoni Sanco Deo Fidio.”^[290] As regards the particular words before us, I may be allowed to cite my Philologica Sacra, p. 47, where I have referred to the frequent confusion of ἀγαπάω and ἀπατάω, ἀγάπη and ἀπάτη in manuscripts of the Old Testament. In Ps. lxxviii. 36, for example, out of more than one hundred manuscripts that have been collated, not one has preserved the correct reading ἠπάτησαν; all have ἠγάπησαν. In 2 Chron. xviii. 2 again only one has the correct text ἠπάτα. From a psychological point of view, therefore, it would seem more natural to suppose that ἀπάταις is the original reading in the passage under consideration, and ἀγάπαις the transcriptional error. The authorities for each are distributed as follows:—

In the first edition of this work I said it was strange, considering the frequent confusion of ἀγάπη and ἀπάτη, that Tischendorf goes by the majority of his witnesses in the case of 2 Peter ii. 13 (Westcott and Hort in their text, Weiss, Weymouth, and Baljon all do the same), “whereas the same word should be read in both cases, and that ἀγάπαις. Otherwise it would be necessary to suppose that the text was already corrupt when the one writer used the epistle of the other, no matter whether Peter or Jude: quod variat, verum esse non potest.” I cannot understand an argument like that of Kühl (Meyer⁶, on 2 Peter ii. 13, p. 428): “ἀπάταις is presumably original in one of the passages, most likely in 2 Peter, as ἀγάπαις goes better with ὑμῶν in Jude 12 than with αὐτῶν here. B has ἀγάπαις in both places, and C in the same way ἀπάταις, which is explainable on the supposition that originally the one word stood in the one passage and the other in the other. Nearly all recent expositors favour the reading ἀπάταις in 2 Peter.” I am glad now to have the powerful support of Zahn in my dissent from that view. Reference may be made to the excellent article on Jude by F. H. Chase in Hastings’ Dictionary of the Bible, ii. 799-805. His first paragraph is on the “Transmission of the Text,” and the article is a model of what such things should be.^[291] On the Philoxenian Syriac see the work of Merx mentioned above, p. 106 (5). On the rest of the verse, see Zahn, Einleitung, ii. 71. He points out that Tischendorf’s apparatus is misleading here, as it fails to notice the omission of ὑμῖν by the Philoxenian Syriac, the Sahidic version, the Speculum of Pseudo-Augustine (m), and by Pseudo-Cyprian. In his opinion it is an interpolation due to the συν— of συνευωχούμενοι. These pronouns are very liable to be interpolated, as is pointed out by Wordsworth and White in their Epilogus, p. 729, where these “additamenta” come next after “Proper Names”; see above, p. 238.

ii. 15. On Βοσόρ, see p. 243 f.

ii. 22b. In Hippolytus, Refutatio, ix. 7, we find: μετ’ οὐ πολὺ δὲ ἐπὶ τὸν αὐτὸν βόρβορον ἀνεκυλίοντο. On the connection of this with 2 Peter ii. 22, see Zahn, GK. i. 316. Wendland tried to make out that it is a saying of Heraclitus. Compare also Clement, Λόγος Προτρεπτικός, x. 96; ὕες γάρ, φασίν, ἥδονται βορβόρῳ μᾶλλον ἢ καθαρῷ ὕδατι, καὶ ἐπὶ φορυτῷ μαργαίνουσιν κατὰ Δημόκριτον.

iii. 6. The conjectural reading δι’ ὃν for δι’ ὧν Schmiedel thinks well worthy of consideration. See his Winer, § 19.

iii. 10. None of the variants here appears to be the correct reading (κατακαήσεται in various forms: ἀφανισθήσονται: εὑρεθήσεται). What is required is a passive form of ῥέω, or one of its compounds (? διαρρυήσεται).

iii. 16. The article is inserted before ἐπιστολαῖς by א and K L P (“the Antiochean recension”), but omitted by A B C. Zahn, who would omit it, points out that ἐν πάσαις ταῖς ἐπιστολαῖς would imply a complete collection of Paul’s Epistles, and would include all the constituents without exception, whereas without the article the phrase contrasts one epistle known to the readers with those of all kinds that he had written. See Einleitung, ii. 108. Tischendorf admitted the reading now favoured by critics in his seventh edition, but rejected it in the eighth. This same thing occurs not infrequently. See the article on 2 Peter by Chase in Hastings’ Dictionary of the Bible, vol. iii. p. 810.

1 John.

iv. 3. Von der Goltz has shown conclusively what was long a matter of conjecture, that Origen not only knew the reading ὁ λύει τὸν Ἰησοῦν, but seemingly preferred it; and that Clement also cites the text in this form in his work on the Passover, which is all but entirely lost. He has also established anew the reliable nature of the Latin version of Irenæus in the matter of Biblical quotations. See Zahn in the ThLbl., 1899, col. 180; Einleitung, ii. 574.

v. 7. The “comma Johanneum” needs no further discussion in an Introduction to the Greek Testament, but its history on Latin soil is all the more interesting. The fact that it is still defended even from the Protestant side is interesting only from a pathological point of view. On the decision of the Holy Office, confirmed by the Pope on the 15th January 1897, see Hetzenauer’s edition of the New Testament, and the notice of it by Dobschütz in the ThLz., 1899, No. 10. On the literature, compare also Kölling (Breslau, 1893); W. Orme’s Memoir of the Controversy respecting the Three Heavenly Witnesses, 1 John v. 7 (London, 1830), New Edition, with Notes and Appendix by Ezra Abbot (New York, 1866); C. Forster, A New Plea for the Authenticity of the Text of the Three Heavenly Witnesses (Cambridge, 1867); H. T. Armfield, The Three Witnesses: The disputed Text in St. John (London, 1893).

An Arabic scholion, attributed to Hippolytus, cites this epistle under the name of Jude. See Zahn, GK. i. 320, 2; 323, 3. In two minuscules cited by Tischendorf, Ἰακώβου is followed by τοῦ ἀδελφοῦ θεοῦ or ἀδελφοθεοῦ, and in one of the subscriptions by τοῦ ἀδελφοθεοῦ. The subscription in ff reads “explicit epistola Jacobi filii Zaebedei (sic).” See Zahn’s Einleitung, i. 75.

ii. 2. συναγωγὴν appears without the article in א* B C and one of Scrivener’s minuscules. This reading is accepted by Zahn, who sees in it an indication that those to whom the epistle is addressed were in possession of several synagogues, that is supposing the word to mean meeting-place, and not simply assembly, as he himself is inclined to believe. See Einleitung, i. 60, 66.

Jude.

5. This verse exhibits an uncommonly large number of variants. Thus εἰδότας occurs with or without ὑμᾶς after it; for πάντα we find both πάντας and τοῦτο; while the position of ἅπαξ varies, the word being found before πάντα, ὅτι, and λαὸν. But even that is not all. Most recent editors read ὅτι Κύριος, but we find also ὅτι Ἰησοῦς: ὅτι ὁ θεὸς: and ὅτι ὁ Κύριος (textus receptus). Tischendorf’s apparatus might lead one to suppose that the witnesses for Ἰησοῦς and ὁ θεὸς omit ὅτι altogether, but that is not so. The ambiguity is due to the loose way in which the note is given. Westcott and Hort think it probable that the original text was ΟΤΙΟ, and that this was read as ΟΤΙΙ̅Ϲ̅, and perhaps as ΟΤΙΚ̅Ϲ̅. Kühl thinks that the easiest explanation of the variants is to suppose that κύριος was the original reading, and that Ἰησοῦς and θεὸς were derived from it. But it seems to me that Zahn has better reason on his side when he argues for ὅτι Ἰησοῦς as the original reading. He first of all eliminates ὁ θεὸς as having no great attestation, and as being found alongside of κύριος in Clement (dominus deus). The choice, therefore, lies between κύριος and Ἰησοῦς. The latter has by far the stronger external attestation, it is the lectio ardua, and is, on internal grounds, also to be preferred. See Einleitung, ii. 88.

22, 23. Zahn has a strong impression that this passage lies at the foundation of Didache, ii. 7: οὐ μισήσεις πάντα ἄνθρωπον, ἀλλὰ οὓς μὲν ἐλέγξεις, περὶ δὲ ὧν προσεύξῃ, οὓς δὲ ἀγαπήσεις ὑπὲρ τὴν ψυχήν σου. If this is really so, we have here a piece of very early testimony, not certainly to the actual words, but to the thought conveyed. See Einleitung, ii. 86.

Subscription: At the end of the Armenian Bible of 1698 we find a note to the effect that “this epistle was written in the year 64 by Judas Jacobi, who is also called Lebbaeus and Thaddaeus, and who preached the Gospel to the Armenians and the Persians.”

APOCALYPSE.

Apart from particular passages, the last book of the Bible cannot be unreservedly recommended to the devout laity for special study, but it is peculiarly well adapted as an introduction to the method of textual criticism, and that for two reasons. First of all, because the number of available witnesses to the text is comparatively small, and, secondly, because these are more easily grouped here than in the other divisions of the New Testament. Reference may be made in this connection to the first part of Bousset’s critical studies on the text of the Apocalypse, where the distinction drawn by Bengel between the Andreas and Arethas groups of manuscripts is correctly emphasized. At the same time Bousset himself comes to the rather unsatisfactory conclusion that an eclectic mode of procedure is all that is possible at present. An attempt has been made above (p. 157) with the conclusion of the Apocalypse. We shall now try a few further examples.

In order to ascertain the relationship of the manuscripts we must select passages that exhibit a considerable divergence of meaning with a small variation of form. Such a passage occurs in the last chapter. In Apoc. xxii. 14, after the words “blessed are they,” we read, in the one class of witnesses, “that wash their robes,” in the other, “that do his commandments.” That is to say, we have in the one case ΟΙΠΛΥΝΟΝΤΕϹΤΑϹϹΤΟΛΑϹΑΥΤΩΝ and in the other ΟΙΠΟΙΟΥΝΤΕϹΤΑϹΕ̅ΤΟΛΑϹΑΥΤΟΥ. The difference is exceedingly small, especially when we consider that in early times ΟΙ was frequently written Υ, and ΕΝ at the end of a line Ε̅. I have no doubt that “wash their robes” is the original reading here and that “do his commandments” is the later alteration, though, of course, others will hold the opposite view. For the former we have the authority of א A, for the latter that of Q (i.e., B^apoc; see above, p. 80) with its associates. The question now becomes: Are there any passages where א and A part company, and which are decisive in favour of א? It is impossible to say offhand whether א or A has preserved the correct text. א contains corrections that A does not, and vice versa. Take another example.

The author of the Apocalypse follows the Hebrew idiom, according to which the word or phrase in apposition to an oblique case is put in the nominative.^[292] Thus we have:

ii. 20. τὴν γυναῖκα Ἰεζαβὲλ ἡ λέγουσα. Q makes this ἣ λέγει, and the corrector of א τὴν λέγουσαν. Similarly, iii. 12, τῆς καινῆς Ἱερουσαλὴμ ἡ καταβαίνουσα, where again Q has ἣ καταβαίνει, and א^c τῆς καταβαινούσης. But it is not only the later corrector of א that does this: the first scribe of that manuscript does it himself. For example:

xiv. 12. א has τῶν τηρούντων instead of οἱ τηροῦντες; in verse 14 ἔχοντα instead of ἔχων; in xx. 2, τὸν ὄφιν instead of ὁ ὄφις, etc. In other places A, in this last A alone, it appears, has preserved the correct text.

There are other places, again, where the correct reading is preserved, perhaps, only in a later manuscript, or in none at all. We may compare with the idiom in the Apocalypse what we find at the beginning of the book in the passage about the seven spirits before the throne of God.

i. 4. ἀπὸ τῶν ἑπτὰ πνευμάτων ... ἐνώπιον τοῦ θρόνου αὐτοῦ. In the space indicated by the dots Erasmus has ἅ ἐστιν, Codex 36 has ἅ εἰσιν, Q and C have ἃ, which is adopted by Tischendorf and Westcott and Hort, א and A have τῶν, which is adopted by Lachmann, Tregelles, and by Westcott and Hort in their margin, while Codex 80 has nothing at all. All these variants are explainable on the supposition that the original reading was τὰ. Exception being taken to this construction, one copyist made it τῶν, the other ἃ, the third supplied the copula, and the fourth dropped the offending word altogether. Similarly, in chap. v. 13, א alone has preserved the correct reading τὸ, for which the others have ὃ or ὅ ἐστιν. Another case is ii. 13, where the writer wished to say, “in the days of Antipas, my faithful witness, who was slain.” According to the idiom mentioned above, while Ἀντίπα was in the genitive, ὁ μάρτυς would be in the nominative of apposition. But owing to the influence of this nominative, Ἀντίπα was made nominative so as to agree with it, and the sentence then ran, ἐν ταῖς ἡμέραις Ἀντίπας ὁ μάρτυς μου ... ὃς ... which could not be construed. The consequence was corrections of all sorts. The boldest expedient was simply to drop the ὃς, but other means were adopted to relieve the construction. After ἡμέραις some inserted αἷς or ἐν αἷς, Erasmus read ἐμαῖς, א has ἐν ταῖς, and some Latin witnesses illis. But read Ἀντίπα in the genitive and all is in order.^[293]

The Apocalypse presents quite a number of passages enabling us to distinguish the manuscripts. There is very little difference in form between λύσαντι and λούσαντι (i. 5), ἀετοῦ and ἀγγέλου (viii. 13), λίθον and λίνον (xv. 6), but it makes a great difference whether we read “who redeemed us” or “who washed us,” an “eagle flying” or “an angel flying,” “wearing pure linen” or “wearing pure stone.” These variations are the result of accidental errors in transcription. But we meet an instance of intentional alteration in xiii. 18, where the number of the beast is variously given as 666 and 616.

Grouping the witnesses for the former variants we have—

The two readings are combined not only by certain commentators, but in some manuscripts, ἀγγέλου ὡς ἀετοῦ.

Tregelles and Westcott alone accept the reading λίθον; all the other editors regard it as an early transcriptional error. Holtzmann refers to the parallel passages i. 13, iv. 4, vii. 9, 13, xvii. 4, xviii. 16, xix. 8, 14, in support of λίνον, but they point rather the other way. For “fine linen” Apocalypse has βύσσινος five times, but never once λίνος, which means only the material, and not the garment made of it. Moreover, we find a parallel in the Old Testament, though in another connection, in Ezekiel xxviii. 13, where we read πάντα λίθον χρηστὸν ἐνδέδεσαι, so that λίθον here must not be so confidently rejected. Λίθον was more liable to be changed to λίνον than vice versa, as the Vulgate shows, in which the authorised printed edition has linteo where the manuscripts read lapidem. At the same time one cannot but admit that primitive transcriptional errors do occur. The reading ἀγγέλου in viii. 13, to which certain manuscripts prefix ἑνὸς, seems to me to be corroborated by ἄλλον ἄγγελον πετόμενον in xiv. 6. Or are we to read ἀετόν there in the face of all the witnesses?

v. 1. The correct text here is that adopted by Zahn: γεγραμμένον ἔσωθεν καὶ ὄπισθεν κατεσφραγισμένον. Grotius, though mistaken as to the true text, was the first to give the right interpretation of the words by taking ἔσω (ἔσωθεν) with γεγραμμένον, and ἔξωθεν (ὄπισθεν) with κατεσφραγισμένον. “Locus sic distinguendus γεγραμμένον ἔσω, καὶ ἔξωθεν κατεσφραγισμένον.” This combination of the words (“haec nova distinctio”) was combated for the reason among others that it deprived them of all their force and rendered them superfluous, for who ever saw a roll that was written on the outside and sealed on the inside. See Pole’s Synopsis, where it is said of Grotius, “tam infelix interpres Apocalypseos est magnus ille Hugo in rebus minusculis.” Zahn (Einleitung, ii. 596) improves the text of Grotius, but retains his connection of the words. He holds that ἔσωθεν and ὄπισθεν are not correlative terms, and that the idea of a papyrus roll written on both sides (ὀπισθόγραφον) must be abandoned; compare above, p. 43, n. 2. The book was, in fact, not a roll but a codex. Two things point to this. There is, first, the fact that is said to be ἐπὶ τὴν δεξιὰν. Had it been a roll it would have been ἐν τῇ δεξιᾷ. Moreover, the word used for opening the book is ἀνοῖξαι, and not, as in the case of rolls, ἀνελίσσειν, ἀνειλεῖν, or ἀναπτύσσειν. That it was not written on the outside is also shown by the fact that it was sealed with seven seals, the purpose of which was to make the reading of the book impossible. Not till the seventh seal is broken is the book open and its contents displayed. This βιβλίον is quite different from the βιβλαρίδιον mentioned in chapter x. 2, 9. See also E. Huschke, Das Buch mit 7 Siegeln (1860), to which Zahn refers (lib. cit. 597).

ix. 17. For ὑακινθίνους Primasius has spineas (= ἀκανθίνους), a reading which neither Bousset nor Baljon, strange to say, think worth recording. Bousset rightly observes that in the following verse πῦρ corresponds to πύρινος, and θεῖον to θειώδης, so that καπνός lets us see what the writer understood as the colour of hyacinth—viz. the colour of smoke. But the ideas of “thorns” (spineae) and “smoke” are even more closely related.

xiii. 18. Irenæus found 616 given as the number of the beast in some manuscripts, which he could only explain as a transcriptional error: “hoc autem arbitror scriptorum peccatum fuisse ut solet fieri quoniam et per literas numeri ponuntur, facile literam Graecam quae sexaginta enuntiat numerum in iota Graecorum literam expansam.” In reality, however, the change from ξ to ι would be a contraction rather than an expansion, and the alteration would seem to be intentional, seeing that 666 in Hebrew characters gives the Greek form Neron Kesar, and 616 the Latin Nero Kesar. Irenæus himself, however, appeals to the fact that the number 666 was found ἐν πᾶσι τοῖς σπουδαίοις καὶ ἀρχαίοις ἀντιγράφοις, μαρτυρούντων αὐτῶν ἐκείνων τῶν κατ’ ὄψιν τὸν Ἰωάννην ἑωρακότων (v. 30, 1-3). The opening words in the Latin translation run, “in omnibus antiquis et probatissimis et veteribus scripturis.” The subscription which he himself appended to his own principal work (see above, p. 149) shows how scrupulously exact he was with respect to ἀντίγραφα, so that we may give him credit for having consulted old and reliable manuscripts of the Apocalypse. The erroneous reading (616) is now found only in C and two minuscules (5 and 11).

xxii. 11. The only authorities cited by Tischendorf in support of the reading δικαιωθήτω (in place of δικαιοσύνην ποιησάτω) are the two minuscules 38 and 79 and the Clementine Vulgate. But we find the passage alluded to in the epistle which the Church of Lyons wrote giving an account of the Martyrdom of the year 177: ἵνα πληρωθῇ ἡ γραφή· ὁ ἄνομος ἀνομησάτω ἔτι, καὶ ὁ δίκαιος δικαιωθήτω ἔτι (apud Euseb., Eccles. Hist., v. 1, 58). This lends such support to the reading δικαιωθήτω in Apoc. xxii. 11, that Zahn not unnaturally speaks of it as “certainly the original text” (GK. i. 201). E. A. Abbott places the date of the Epistle of the Church of Lyons as early as 155 (see Expositor, 1896, i. 111-126). Another aspect would be given to the question if the Greek form of the Epistle were derived from a Latin, or if, as Resch supposed, the words were a quotation of a saying of Jesus (Agrapha, § 133, p. 263 ff.).

I take the opportunity of appending to Resch’s work the fine saying which Zahn cites from Augustine’s Contra Adversarium Legis et Prophetarum (ed. Bassan. x. 659 ff.) as an otherwise unknown Apocryphum. The disciples asked Jesus “de Judaeorum prophetis, quid sentire deberet, qui de adventu eius aliquid cecinisse in praeteritum putantur.” And He, “commotus talia eos etiam nunc sentire, respondit: Dimisistis vivum qui ante vos est et de mortuis fabulamini.” A similar saying from the Acta Petri Vercell. 10 is cited by Harnack in connection with the third of the Oxyrhynchus Logia: “Qui mecum sunt, non me intellexerunt.”