(19) The copyist may be tempted to forsake his proper function for that of a reviser, or critical corrector. He may simply omit what he does not understand (e.g. δευτεροπρώτῳ Luke vi. 1; τὸ μαρτύριον 1 Tim. ii. 6), or may attempt to get over a difficulty by inversions and other changes. Thus the μυστήριον spoken of by St. Paul 1 Cor. xv. 51, which rightly stands in the received text πάντες μὲν οὐ κοιμηθησόμεθα, πάντες δὲ ἀλλαγησόμεθα, was easily varied into πάντες κοιμηθησόμεθα, οὐ π. δὲ ἀλ., as if in mere perplexity. From this source must arise the omission in a few manuscripts of υἱοῦ Βαραχίου in Matt. xxiii. 35; of Ἱερεμίου in Matt. xxvii. 9; the insertion of ἄλλου ἐκ before θυσιαστηρίου in Apoc. xvi. 7; perhaps the substitution of τοῖς προφήταις for Ἡσαΐᾳ τῷ προφήτῃ in Mark i. 2, of οὔπω ἀναβαίνω for οὐκ ἀναβαίνω in John vii. 8, and certainly of τρίτη for ἕκτε in John xix. 14. The variations between Γεργεσηνῶν and Γαδαρηνῶν Matt. viii. 28, and between Βηθαβαρᾶ and Βηθανίᾳ John i. 28, have been attributed, we hope and believe unjustly, to the misplaced conjectures of Origen. Some would impute such readings as ἔχωμεν for ἔχομεν Rom. v. 1; φορέσμεν for φορέσομεν 1 Cor. xv. 49, to a desire on the part of copyists to _improve_ an assertion into an ethical exhortation, especially in the Apostolical Epistles; but it is at once safer and more simple to regard them with Bishop Chr. Wordsworth (N. T. 1 Cor. xv. 49) as instances of _itacism_: see class (7) above. (20) Finally, whatever conclusion we arrive at respecting the true reading in the following passages, the discrepancy could hardly have arisen except from doctrinal preconceptions. Matt. xix. 17 Τί με λέγεις ἀγαθόν? οὐδεὶς ἀγαθὸς εἰ μὴ εἶς, ὁ Θεός; or Τί με ἐρωτᾷς περὶ τοῦ ἀγαθοῦ? εἶς ἐστὶν ὁ ἀγαθός: John i. 18 ὁ μονογενὴς υἱός or μονογενὴς Θεός: Acts xvi. 7 τὸ πνεῦμα with or without the addition of Ἰησοῦ: Acts xx. 28 τὴν ἐκκλησίαν τοῦ Θεοῦ or τὴν ἐκκλησίαν τοῦ Κυρίου: perhaps also Jude ver. 4 δεσπότην with or without Θεόν. I do not mention Mark xiii. 32 οὐδὲ ὁ υἱός, as there is hardly any authority for its rejection now extant; nor Luke ii. 22, where τοῦ καθαρισμοῦ αὐτῆς of the Complutensian Polyglott and most of our common editions is supported by almost no evidence whatever.