John iv. 1. After βαπτίζει we find ἤ omitted in AB* (though it is added in what Tischendorf considers an ancient hand, his B2) GLΓ, 262, Origen and Epiphanius, but appears in אCD and all the rest. Tregelles rejects ἤ in his margin, Hort and Westcott put it within brackets. Well may Dr. Hort say (Notes, p. 76), “It remains no easy matter to explain how the verse as it stands can be reasonably understood without ἤ, or how such a mere slip as the loss of Η after ΕΙ should have so much excellent Greek authority, more especially as the absence of ἤ increases the obvious no less than the real difficulty of the verse.” John vii. 39. One of the worst faults a manuscript (the same is not true of a version) can have is a habit of supplying, either from the margin or from the scribe’s misplaced ingenuity, some word that may clear up a difficulty, or limit the writer’s meaning. Certainly this is not a common fault with Cod. B, but we have here a conspicuous example of it. It stands almost alone in receiving δεδομένον after πνεῦμα: one cursive (254) has δοθέν, and so read _a b c e ff_2 _g l q_, the Vulgate, the Peshitto, and the Georgian (Malan, St. John), the Jerusalem Syriac, the Polyglott Persic, a catena, Eusebius and Origen in a Latin version: the margin of the Harkleian Syriac makes a yet further addition. The Sahidic, Ethiopic, and Erpenius’ Arabic also supply some word. But the versions and commentators, like our own English translations, probably meant no more than a bold exposition. The whole blame of this evident corruption rests with the two manuscripts. No editor follows B here. John ix. 4. Most readers will think with Dean Burgon that the reading ἡμᾶς δεῖ ἐργάζεσθαι τὰ ἔργα τοῦ πέμψαντος (whether followed by με or ἡμᾶς) “carries with it its own sufficient condemnation” (Last Twelve Verses, &c., p. 81). The single or double ἡμᾶς, turning the whole clause into a general statement, applicable to every one, is found in א*BDL, the two Egyptian, Jerusalem Syriac, Erpenius’ Arabic, and Roman Ethiopic versions, in the younger Cyril and the versifier Nonnus. Origen and Jerome cite the passage as if the reading were ἐργάζεσθε, which, by a familiar _itacism_ (_see_ p. 11), is the reading of the first hand of B. The first ἡμᾶς is adopted by Tischendorf, Tregelles, Westcott and Hort: the second by Tischendorf alone after א*L, the Bohairic, Roman Ethiopic, Erpenius’ Arabic, and Cyril. Certainly με of BD, the Sahidic, and Jerusalem Syriac, is very harsh.