104 Timo Flink an apostasy from an earlier state and not of sequence of events31. Taking ἅπαξ with εἰδότας is lectio difficilior, because the sense of the verse flows more naturally when ἅπαξ goes with τὸ δεύτερον. Linked with this is the fact that the author has already used ἅπαξ in verse 3 with its regular meaning “once for all”32. It seems unlikely that he would alter his usage. (II) The τὸ δεύτερον does not necessary require πρῶτον or its equiva- lent as τὸ δεύτερον can mean “afterwards”33. Transcriptionally the lack of πρῶτον would provide a clear motivation to move ἅπαξ to a post-ὅτι position. (III) The post-ὅτι position may well be due patristic conven- tions. Clement, Didymus and Origen quoted the text only partially and quite freely. They extracted and reshaped the text. This naturally called for ἅπαξ to be relocated to go with τὸ δεύτερον, because these authors lacked the εἰδότας-clause or portion of it altogether34. (IV) Transcription- ally the pre-ὅτι position is more difficult, because it creates linguistic difficulties for the ὅτι-clause, which do not exists if the post-ὅτι position is accepted, namely, an independent τὸ δεύτερον and a worse fit with present aspect of εἰδότας than with punctiliar aspect of σώσας instead35. Thus, both transcriptional and intrinsic probabilities favour ἅπαξ with εἰδότας36. (3) Should the reconstructed text read πάντα, πάντας or τούτο? The question is rather straightforwardly answered and does not impose a problem. The πάντας is read only by P72* S:Phmss arabgib, so this might not be a singular reading. However, there are some connections of S:Phmss and P72*, so they may not offer an independent witness but rather two branches of an earlier exemplar, though it is possible that they are not related and the connection is accidental37. The πάντα is read by P72c ) A B C2 Ψ 81 88 442 1739* 2200 // 33c 323 621 623* 630 665 915 1845 // 6 93 1243 1846 1881 2298 2344 2805 // al L:VTR K:Sms>B S:HPh Ä Cyr Ephr Hier Lcf Or. The rival τούτο is read by L 307 326 431 436 453 808 // 18 35 1067 1409 1836 1837 1875 2374 // 5 61 254 468 1292 1735 2186 2818 A.F.J. Klijn, “Jude 5 to 7”, in W.C. Weinrich, New Testament Age. Essays in Honor of 31 Bo Reicke (2 vols.; Macon 1984), I, 237-44; Wikgren, “Some Problems in Jude 5”, 147; Albin, Judasbrevet, 599; Wikgren, “Some Problems in Jude 5”, 147; 32 Bauckham, Jude, 2 Peter, 43; J.N.D. Kelly, A Commentary on the Epistles of Peter and 33 of Jude (London 1969) 254; R.C.H. Lenski, The Interpretation of I and II Epistles of Peter, the Three Epistles of John and the Epistle of Jude (Minneapolis, MN 1966) 617; Wikgren, “Some Problems in Jude 5”, 147; Osburn, “The Text of Jude 5”, 109-110. 34 Wachtel, Der byzantinische text, 351; Wasserman, The Epistle of Jude, 259. 35 Contra Landon, A Text-Critical Study, 76-77, who argues for the post-ὅτι position on 36 the basis of his selection of the subject as anarthrous κύριος with ). Albin, Judasbrevet, 493; J.N. Birdsall, “The Text of Jude in P72”, JTS 14 (1963) 394- 37 99; Wasserman, The Epistle of Jude, 260.