¹ For kind assistance in securing the two photographs which have made this reconstruction possible I am indebted to the friendliness of Reverend A. V. Valentine Richards of Christ's College, Cambridge.

² This appears not to be the manuscript of “Variae Lectiones of the New Testament,” the loss of which in 1645 or 1646 is described by Dr. Parr (C. R. Elrington, Life of James Ussher [Ussher's Works, vol. I], 1847, pp. 244 f.), but rather that referred to in Ussher's letter to Reverend Dr. Hammond, January 14, 1650 (Works, vol. XVI, p. 174), which, so far as he there states, may not yet have been in existence in 1646.

³ It is possible that the little triangle containing μην τινι κυπριω perished between the date of Mill and that of Dickinson.

⁴ This writer seems to display a similar tendency to refer to ‘the day,’ in his liking for the Old Testament phrase ἡ ἡμέρα τοῦ σαββάτου (τῶν σαββάτων) Lk. 4, 16; 13, 14; 13, 16; Acts 13, 14; 16, 13, which no other New Testament writer uses; and in his probably quite incorrect expression, ἡ ἡμέρα τῶν ἀζύμων, Lk. 22, 7.

⁵ Sardinia belonged to the Byzantine empire from the middle of the sixth century; Greek monks are known to have been there in the period of the monothelite controversies of the seventh century; Chapman, J., Notes on the Early History of the Vulgate Gospels, 1908, p. 158.Google Scholar

⁶ The close relation of Codex E to Bede's Greek text was mentioned by Richard Simon, ‘Dissertation critique sur les principaux actes manuscrits,’ p. 60, appended to his Histoire critique des principaux Commentateurs du Nouveau Testament, Rotterdam, 1693.

⁷ Turner, C. H., Art, ‘New Testament, Text of,’ in Murray's Illustrated Bible Dictionary, 1908, p. 586Google Scholar; Souter, A., The Text and Canon of the New Testament, 1913, p. 29.Google Scholar

⁸ Jülicher, A., ‘Kritische Analyse der lateinischen Übersetzungen der Apostelgeschichte,’ Zeitschrift f. d. neutestamentliche Wissenschaft, XV, 1914, pp. 182 f.Google Scholar

⁹ Of course an occasional blunder creates a difference which was not in the exemplar, and was not intended in E e.

¹⁰ Tischendorf, in the Prolegomena to his edition of Laudianus, Codex, Monumenta sacra inedita, Nova collectio, vol. IX, 1870, p. XVIIGoogle Scholar, gives a number of other instances of difference, chiefly minor variation in conjunctions. In several of his cases the Latin e agrees with D.

¹¹ Jülicher (op. cit. p. 182) speaks of the “impossibility” that the text of D or of E should anywhere have been altered to make it agree with d or e, but the whole context of his sentence raises the strong suspicion that “Unmöglichkeit” is a printer's error for ‘Möglichkeit.’

¹² Zahn, , Die Urausgabe der Apostelgeschichte des Lucas, 1916, p. 226Google Scholar, remarks that “der gewissenhafte und nicht ungebildete Schreiber den Lc in einem anständigen, eines solchen Schriftstellers würdigen Gewand vor den Leser treten lässt, recht im Gegensatz zu dem unsauberen, ohne Geschmack und ohne ernstes Nachdenken aus allerlei verschiedenfarbigen Lappen zusammengeflickten Anzug, in welchen Dd die AG und ihren Verfasser gekleidet hat.”

¹³ The Old Syriac Element in the Text of Codex Bezae, 1893, pp. 132–138.

¹⁴ Acta apostolorum, 1895, pp. 28 f.

¹⁵ Philologica Sacra, 1896, pp. 43–45.

¹⁶ Such Latin influence on the Greek text must be admitted in certain cases in Codex Bezae (for instance 2, 11 αραβοι; 16, 12 κεϕαλη), even by those who do not accept Rendel Harris's far-reaching contentions. Of Codex G (Boernerianus) of Diehl, Paul E., Zeitschrift f. d. neutestamentliche Wissenschaft, XX, 1921, p. 107Google Scholar, writes: “Die Vergewaltigung des griechischen Textes durch den Lateiner in G ist so offenkundig, dass, wo immer G als einziger Grieche mit dem Lt. übereinstimmt, die lateinische Lesart als aus dem Griechischen nicht bezeugt zu gelten hat.”

¹⁷ It may be observed in passing that Gregory's statement, Prolegomena, p. 411, footnote, that the Greek codex 218 (now 1522; e^scr α464) has a text like that of E, is not borne out by a closer examination; but see Scrivener, An Exact Transcript of the Codex Augiensis, pp. lviii f.

¹⁸ Codex 81 (formerly 61^ac; α162; p^scr), while of course a minuscule, contains a text which approves itself in Acts as at least equally true to type with any of the old uncials, and it may properly be treated for practical purposes as one of them; see Hort, Introduction, p. 154. In fact, Codex 81 probably furnishes a better example of the standard Alexandrian text than any of the four uncials (B א A C), for it shows fewer peculiar and erratic readings. It rarely stands by itself in the group, except where it shows (as do also A and C, and in less degree א) Antiochian influence.

¹⁹ The figures stated in these investigations cannot be guaranteed as more than approximate, but they are very nearly accurate, and the margin of possible error does not affect their value for the purposes for which they are employed. Mere variations of spelling have been neglected.

²⁰ Taylor, C., Hebrew-Greek Cairo Genizah Palimpsests from the Taylor-Schechter Collection, 1900, p. 94.Google Scholar

²¹ For other illustrations of the several types of reading to be found in E see the elaborate analysis and lists in Von Soden, Die Schriften des Neuen Testaments, I. Teil, 3. Abteilung, pp. 1709, 1717–1720, 1811–1814.