A single good example of this process will make it yet more intelligible to the careful student. It shall be one that has been chosen for the purpose by more than one of the advocates of the system we are on the whole opposing. Of the two forms in which the Lord’s Prayer is delivered to us, Matt. vi. 13 has the clause ἀλλὰ ῥῦσαι ἡμᾶς ἀπὸ τοῦ πονηροῦ in every known authority: in Luke xi. 4 the case is far otherwise. That Tertullian, when citing the words before and after it, should take no notice of it, would of itself prove little. Origen, however, once passes it by in like manner, once more expressly declares that it was not in St. Luke (παρὰ τῷ Λουκᾷ σεσιώπηται), a third time explains in his most happy manner why it was omitted by the one Evangelist, inserted by the other. The question thus raised sets us upon the inquiry what other evidence we have for rejecting the clause in St. Luke. It appears to be wanting in several Greek manuscripts, such as L, 1, 22, 57, 130 both Greek and Latin, 131, 226*, 237, 242, 426, 582, 604, and in the catenas annexed to 36, 237, 239, 253, 259, 426; several of these codices (as 57, 226, 242) not being much found in such company. It is absent from the Vulgate version, and apparently from some forms of the Old Latin, the rather as Augustine says that St. Luke gives five petitions in the Lord’s Prayer, St. Matthew seven, and attributes the omission of our clause to some such reason as Origen had assigned. It is omitted also in the Armenian version, which, except for the later translation by Sahak from Syriac, might be supposed to differ _toto caelo_ from the Latin in country and genius. The list is closed by the younger Cyril, a pure witness from another region, very different lines of evidence thus converging into one. Then comes the probability that if one of the Gospels contained the Lord’s Prayer in a shorter form than the other, nothing was so likely as that a scribe in perfect innocence would supply what he considered an undoubted defect, without staying to reflect with Origen and Augustine that the two were delivered on different occasions, to different classes of persons, with different ends in view. Turning therefore now, with a strong case already made out for the omission of the clause, to א and B, which have been hitherto kept out of sight, we find that B has not the disputed words at all, nor had א by the first hand, but in one three centuries later. The clear result, so far as it goes, is at once to vindicate the claim of אB to high consideration, and to make out a formidable case against the genuineness of the six words involved. We say advisedly a formidable, not necessarily a fatal case, for the counter evidence is still very strong, and comes as much as that alleged above from different quarters, being also as early as widely diffused. It consists of Codd. ACDEFGHKMR(292)SUVΓΔΛΠ, of all cursives not named above, of the Old Latin