(See Huther, Krit. exeg. Handb. über die drei Briefe des Ap. Johannes, 3te Aufl., 1868, p. 222, note.) It is found in none of the numerous editions printed at Wittenberg before 1596. In the Swiss-German version (not published under Luther's name) printed by Froschover at Zürich in 1529, it was inserted in smaller type, and so in the edition of 1531; in nearly all the later editions from 1534 to 1589 (that of 1561 is said by Ebrard to be an exception), in brackets; in 1597 without brackets, at which time it was also introduced as a proof-text into the Zürich Catechism. The Basle edition by Byrlinger in 1552 is said to have it without brackets. It was still omitted in Meissner's Wittenberg edition of 1607, and in a quarto edition printed at Wittenberg in 1620; also, in Hamburg editions of 1596, 1619, and 1620. Since this last date the interpolation has appeared in the numberless editions of Luther's German Bible without mark of doubt, except that it has been bracketed in the recent authorized" revised edition" of his version of the New Testament (Halle, Canstein'sche Bibelanstalt, 1871), with the following note: "The bracketed words are wanting in Luther's translation, and were not added till later." It should be understood that the words auf Erden, "on earth," in verse 8, are not included in the brackets. They were inserted by Luther in the five editions of his German Bible printed at Wittenberg from 1541 to 1545 inclusive; but this very fact shows that his attention was directed to the passage, and that the omission of the three heavenly witnesses. was intentional.

(Perhaps I may be pardoned for turning aside a moment to correct two errors which have been repeated from Rickli (1828) by a large number of respectable scholars, as De Wette, Tischendorf in his editions of 1841, 1849, 1859, and 1869-72, Bertheau in his edition of Lücke on the Epistles of John (1856), Davidson, Braune in Lange's Commentary, etc. They all speak of Robert Stephens as receiving the passage in his editions of 1546-69, and Beza in his editions of 1565-76. They should have said "Robert Stephens the elder in his editions of 1546-51, and Robert Stephens the

younger in his edition of 1569" (the great Robert died ten years before); also, "Beza in his editions 1565-98." Beza published no edition in 1576: the one of that date erroneously ascribed to him by several writers was edited by Henry Stephens.)

We may observe, finally, that the other early Reformers and friends of Luther generally rejected the passage; so Zwingli, Bullinger, Ecolampadius, Bugenhagen (Rickli, ubi supra, pp. 35, 36). So, also, according to Kettner (Historia dicti Johannei. . . 1 John v. 7, etc., 1713, cap. 13), Melanchthon, Cruciger (or Creutziger), Justus Jonas, Förster, Aurogallus. (See Semler, Hist. u. krit. Sammlungen über 1 John v. 7, I. 248.) Bugenhagen, as we have seen, was especially strenuous against it; see his Expositio Jona, 1550, cited by Rickli, p. 39. It was also omitted in the celebrated Latin version of the Bible by Leo Judæ, Pellicanus, Peter Cholin, Rudolph Gualther, and others, printed at Zürich in 1543, fol., and commonly called the Zürich Bible or Versio Tigurina. A marginal note explains the reasons for its rejection. The passage was received, though with hesitation, by Calvin, and without hesitation by Beza. Both of them, however, explain "these three are one" as relating not to unity of essence, but agreement in testimony.

To trace the history of this gross corruption of the text in modern Translations, Catechisms, and Confessions of Faith, especially in the Greek Church since the sixteenth century, and in modern editions of some ancient versions, as the Peshito Syriac, Armenian, and Slavonic, might be interesting and instructive, psychologically as well as critically; but there is no room for it here.

XX.

ON THE DIVISION OF THE GREEK NEW

TESTAMENT INTO VERSES.

[From the First Part of the Prolegomena to Tischendorf's Editio Octava critica maior.]

THE history of the division of the Bible into our present verses is somewhat obscure, and many erroneous statements are current respecting it. For example, Tischendorf (in Herzog's Real-Encyk. ii. 174, p. 422, 2d ed.) and Reuss (Gesch. d. heiligen Schriften N. T., 5te Ausg., 1874, § 337), after Jahn, Bertholdt, and others, represent it as first introduced by Robert Stephens (Estienne) in his edition of the Vulgate in 1548; De Wette (Einl. in d. A. T. p. 112, 7te Ausg., 1852) and Keil (Einl., 2te Aufl., 1859, p. 518) say 1558. But no edition of the Vulgate was printed by Robert Stephens in either of those years. Others erroneously assign 1545, others still 1557, as the date of the Latin Bible first divided into verses. The best account of the matter that I know of is given by Dr. William Wright, art. "Verse" in Kitto's Cyclopædia of Bibl. Literature (London, 1845); in the third edition (1870) this article is somewhat carelessly abridged. On some points I have supplemented his statements.

The main facts are as follows. The numbering of the Masoretic verses in the Old Testament is supposed to have been first made by Rabbi Isaac Nathan, for convenience of reference in his Hebrew Concordance, completed A.D. 1448, and first printed at Venice in 1524.* The Quincuplex Psalterium, printed by Henry Stephens the elder at Paris in 1509, and edited by Jacobus Faber Stapulensis (Jacques Le Fèvre d'Estaples), is the first edition of any part of the Old

*The dates 1445 and 1523 are sometimes inaccurately given. These errors are corrected by Bindseil, Concord. Homer. Specimen, Halis, 1867, 8vo, Prolegom. p. xvi., and p. xviii., note.

Testament in which the Masoretic verses are numbered by Arabic figures. The first edition of the whole Bible divided into verses is the Latin translation by Sanctes Pagninus, printed at Lyons in 1528. But in the Apocrypha and the New Testament his division was very different from ours, the verses being twice or three times as long; and it seems to have been followed in no other edition.

The first edition of the New Testament divided into our present verses was printed by Robert Stephens at Geneva in 1551, in 2 vols. 16mo, the Greek text occupying the centre of the page, with the Latin version of Erasmus on one side and the Vulgate on the other. His son Henry tells us that a large part of this verse-division was made "inter equitandum," while the author was on a journey from Paris to Lyons. It was preliminary to his construction of a Greek Concordance of the New Testament, which was completed by Henry Stephens and published in 1594. (See Henry Stephens's preface to this Concordance.) Another reason given by Robert Stephens for the division into separated verses was, “quòd hac ratione vtraque translatio posset omnino e regione Græco contextui respondere" (Pref. to N. T., 1551). The first edition of the whole Bible divided into our present verses was Robert Stephens's edition of the Vulgate, Geneva, 1555 ("VIII. Idvs Aprilis "), 8vo, the division being made for his Latin Concordance, issued the same year. This division also appears in the splendid edition of the Vulgate, accompanied in the Old Testament by the version of Pagninus with the notes ascribed to Vatable, and in the New Testament by that of Beza with his annotations, which was printed in three folio volumes by Robert Stephens at Geneva in 1557. This is the eighth and last edition of the Latin Bible printed by Robert Stephens.* The first French New Testament divided into verses was that printed by Robert Stephens in 1552, in 2 vols. small 8vo, containing the French version of Olivetan revised by

*See LeLong, Bibl. Sacra, ed. Masch, iii. 191 ff.; and for a very full description, Knock, Hist.-crit. Nachrichten von der Braunsweigischen Bibelsammlung, Guelferb. 1754, i. pp. 876-891.

Calvin, and the Latin version of Erasmus, in parallel columns; the first French Bible so divided appears to be the edition of the Genevan version printed by Robert Stephens at Geneva, 1553, in folio.* The first Italian version, so far as I know, which contains the verse-divisions of Stephens was that which was made by the martyr John Louis Paschale, and issued without place, but perhaps at Geneva, in the year 1555.† The first Datch translation which has these verses was published at Emden in 1556, 8vo, by Gell. Ctematius; the Dutch Bible issued at Emden in 1560, 4to, is so divided.‡ The first English New Testament divided into verses was the version of William Whittingham, printed at Geneva in 1557; the first English Bible so divided was the Genevan version, completed in 1560, in which the translation of the New Testament differs widely from that of 1557. Beza followed Stephens's division into verses, with some variations, in the first edition of his Latin translation of the Greek Testament, published at Geneva in 1557 (this is the date at the end of the volume; the title-page is dated 1556), already referred to as the second volume of Robert Stephens's Latin Bible of that year. In his first edition of the Greek New Testament accompanied with his Latin version and notes, Geneva, 1565, fol., and in his numerous subsequent editions, Beza deviated much more frequently from the verse-divisions of Robert Stephens; and his editions had great influence in giving currency to the use of the division into verses, which soon became general. His variations from the division of Stephens were largely followed by later editors, especially the Elzevirs, who also introduced others of their own. Others still will be found in the early modern translations.

The variations of later editions of the Greek Testament from that of Robert Stephens in respect to the division into

* See the account of these editions in Baumgarten's Nachrichten von merkw. Büchern, Halle, 1752, ii. 377 ff., 379 ff.

"Est mihi N. T. Gallico-Italicum, editum a. 1555 ap. Giovan Luigi Paschale, quod est distinctum hodierna distinctione versuum." Leusden, Philol. Hebr.-Gr., Utrecht, 1670, Diss. iii., $ 12, p. 21. Paschale was burned at Rome in 1560 for protestantism.

I See Isaac Le Long, Boek-Zaal der Nederduyt. Bybels, Amst. 1732, p. 716; cf. p. 708 seq., 711 seq., 667 seq.