in 1862 in four volumes folio, printed in facsimile type, with 19 plates of actual facsimiles of different parts of the manuscript, and on the similar edition of the Codex Vaticanus now publishing at Rome, of which three volumes have thus far appeared, two of them containing the Old Testament as far as the end of Nehemiah, and the other the New Testament part of the manuscript. I have also used Tischendorf's facsimile edition of the Codex Friderico-Augustanus (another name for 43 leaves of the Sinaitic MS.), published in 1846; his "Novum Testamentum Vaticanum" (1867), with the "Appendix" (1869); and his "Appendix Codicum celeberrimorum Sinaitici Vaticani Alexandrini," with facsimiles (1867).

Mr. Burgon's arguments are as follows:-(1.) "The (all but unique) sectional division of Codex B, confessedly the oldest scheme of chapters extant, is in itself a striking note of primitiveness. The author of the Codex knew nothing, apparently, of the Eusebian method."

The Vatican MS. has in the Gospels a division of the text into chapters, which differs from that found in most MSS. from the fifth century onward, and appears, so far as is known, in only one other manuscript, the Codex Zacynthius (E), of the eighth century. It has also a peculiar division into chapters in the Acts and Epistles. Mr. Burgon finds in its scheme of chapters "a striking note of primitiveness." But the Sinaitic

has no division into chapters at all, à prima manu. Is not that quite as primitive? Further, Mr. Burgon's argument appears to be of a circular character. The only proof of the high antiquity of the "scheme of chapters" referred to is its existence in the Vatican manuscript.

It may be worth while, perhaps, to remark that the Roman edition of the Vatican MS. seems to afford evidence (p. 1272, col. 1, and 1299, col. 3) that the division into chapters, noted by numbers in red in the margin, was not made by the original scribe, but by one who preferred in some places a different division into paragraphs. It may have been made, however, by a contemporary hand.

Mr. Scrivener thinks it "very credible that Codex Sinaiticus was one of the fifty volumes of Holy Scripture, written 'on skins in ternions and quaternions,' which Eusebius prepared A. D. 331 by Constantine's direction for the use of the new capital" (Collation of the Cod. Sinaiticus, p. xxxvii. f.; comp. Euseb. Vita Const. iv.36,37). This is possible, though there is no proof of it. Mr. Burgon's argument, that, because the Eusebian sections do not correspond with the paragraphs in the Codex Sinaiticus, Eusebius could have known nothing of the MS. (p. 294), is utterly futile. The object of those sections is

totally different from that of a division into paragraphs. The Eusebian sections are not chapters or paragraphs, but merely serve for a comparison of parallel or similar passages in the Gospels. In not less than 25 instances, there are two of them (in one case three) in a single verse; see, e. g., Matt. xi.27; Mark xiii. 14; Luke vi.21; John xix.6,15,16.

The Eusebian sections are not in the Sinaitic MS. à prima manu, though they may, as Tischendorf supposes, have been added by a contemporary scribe. In that case, the MS. may still be older than the middle of the fourth century; for Eusebius died about A. D. 340. It is curious to see how Scrivener contradicts himself on this matter in a single page (Collation, etc. p. xxxvii.).

(2.) "Cod. (like C, and other later MSS.)," says Mr. Burgon, "is broken up into short paragraphs throughout The Vatican Codex, on the contrary, has very few breaks indeed: e. g. it is without break of any sort from S. Matth. xvii.24 to xx.17: whereas, within the same limits, there are in Cod. as many as thirty interruptions of the context. From S. Mark xiii.1 to the end of the Gospel the text is absolutely continuous in Cod. B, except in one place: but in Cod. & it is interrupted upwards of fifty times. Again: from S. Luke xvii.11, to the end of the Gospel, there is but one break in Cod. B. But it is broken into well nigh an hundred and fifty short paragraphs in Cod. §.

א

"There can be no doubt that the unbroken text of Codex B (resembling the style of the papyrus of Hyperides published by Mr. Babington) is the more ancient. The only places where it approximates to the method of Cod. , are where the Commandments are briefly recited (S. Matth. xix. 18, &c.), and where our Lord proclaims the eight Beatitudes (S. Matth. v.)."

Here, apparently, the stress of Mr. Burgon's argument rests on the rarity of paragraphs, indicated by "breaks," in the Vatican MS. as compared with the Sinaitic. If this is so, he has strangely misrepresented the facts in the case. In the first passage referred to, Matt. xvii. 24-xx.17, there are certainly no less than 32 "breaks" in the Vatican MS., designed to mark a division into paragraphs. In 2 instances (Matt. xvii.24, xix.1) the division is made by the projection of the initial letter into the left-hand margin, in the manner usual in the Sinaitic MS.; in 30, by a space between the words, and a dash (-) below the line where the break occurs, projecting into the left-hand margin, after the fashion common in the Herculanean and early Egyptian papyri, and also found, though more rarely, in the Sinaitic MS. Besides these 32 cases, there are 7 in which a paragraph is indicated by a dash simply, the preceding sentence happening to fill the whole line above it. There are also in the passage referred to about 10 places in which the end of a sentence or a

paragraph is indicated by a space simply. (In respect to the representation of these spaces there is a little difference, in two or three places, between the Roman edition and that of Tischendorf.) But dismissing the simple spaces from the account altogether (though they are certainly breaks), we have in the first passage selected by Mr. Burgon a division into paragraphs in the Vatican MS. even more minute than in the Sinaitic. In Mark xiii.1-xvi.8, there are 39 paragraphs in the Vatican MS. marked by the dash and space-or by the dash alone, when the preceding line is full; and in Luke xvii.11-xxiv.53, 129 paragraphs are thus marked, besides 2 in which the initial letter projects into the margin. There are also places in which divisions are marked by spaces alone.

Such being the state of the case, it may perhaps be thought that Mr. Burgon does not mean to argue the superior date of the Vatican MS. from the comparative rarity of its divisions into paragraphs, but merely from the manner in which they are made; and that he intends by "break," the projection of the initial letter of a paragraph into the left-hand margin, which we find in the Vatican MS. in the Beatitudes (Matt. v.), though not in Matt. xix. 18, the only other place, according to Mr. Burgon, in which B "approximates to the method of Cod. " This, however, can hardly be his meaning, for he makes a separate point of that feature of the Sinaitic MS. in his fourth argument, which will be considered in its proper place.

As to the frequency of the division into paragraphs, we find a great difference in different parts of both the Sinaitic and the Vatican manuscripts. For example, in the Sinaitic MS. (vol. ii.) from 1 Macc. v.55 to x.18, 249 verses, there is but one indication of a paragraph besides that with which the passage begins. For 21 entire columns of 48 lines each, viz. from fol. 21*, col. 4, to fol. 26, col. 4, inclusive, there is no break and no sign of a paragraph whatever. In the First Book of Maccabees, which contains 36 pages in the Codex Sinaiticus, there are 16 pages in which there is no indication of a paragraph, and 10 more in each of which but one paragraph is marked. In the Fourth Book of Maccabees the paragraphs are still rarer in proportion to its length. In the Vatican MS., on the other hand, to anticipate a little the answer to Mr. Burgon's fourth argument, there are many pages in each of which from 10 to 20 paragraphs are marked by the projection of the first letter of a word into the left-hand margin; see, e. g., pp. 41, 44, 48, 53, 71, 73-75, 123, 186, 187, 226, 291-294 (vol. i. of the Roman ed.); and a page of the Vatican MS. contains considerably less than a page of the Sinaitic. In respect both to the frequency of the paragraphs, and to the manner of indicating them, much

appears to have depended upon the fancy of the copyist. The books most read would naturally be divided the most.

(3.) "Again," says Mr. Burgon, "Cod. is prone to exhibit, on extraordinary occasions, a single word in a line, as at—

"This became a prevailing fashion in the VIth century; e. g. when the Codex Laudianus of the Acts (E) was written. The only trace of anything of the kind in Cod. B is at the Genealogy of our Lord."

Here, again, Mr. Burgon mistakes the facts in the case. We find this stichometric mode of giving greater distinctness to particulars exemplified in repeated instances in the Vatican MS., besides the striking one of the genealogy in Luke. For example, in p. 211, col. 3 of the MS., the names of the 22 unclean birds in Deut. xiv. 12-18 appear each in a separate line. On p. 247, col. 3, there is a similar stichometry of 6 lines; on p. 254, col. 1, one of at least 25 lines (Josh. xii.10-22, the list of kings), with another example in the same column, and still another in the next; and in p. 485, col. 2, there is one of 11 lines (the "dukes" in 1 Chron. i.51-54). For other instances see p. 71, col. 3; 76, col. 1; 274, col. 2; and 316, col. 3.

We find, moreover, in the Vatican MS., the different branches of the genealogy in Matthew presented in 38 distinct paragraphs; and the beatitudes in Matt. v. and the salutations in Rom. xvi. are similarly treated. This may be regarded as a kind of stichometry, of which we have also examples in the Old Testament: e. g. p. 138, col. 1, 2: 264, col. 1; 272, col. 1; 309, col. 1. All that can be said in respect to the first form of stichi is, that it is much more common in the Sinaitic MS. than in the Vatican, especially in the New Testament. Both MSS. have also another mode of making distinct the items of an enumeration namely, by spaces between the words, with or without dots (the Roman edition of B does not agree with Tischendorf's about the dots); e. g. Rom. i.29-31, both MSS.; and in the Vatican, 1 Cor. vi.9,10; xiii.13; xiv.26; Gal. v.19–23; Phil. iv.8; Col. iii.8. The choice between the modes seems to have been determined by the taste of the scribe; compare, for example, in the Vatican MS., Lev. xi.13-19 with Deut. xiv. 1218 (pp. 111 and 211). It cannot be made a criterion of date. (4.) Mr. Burgon's fourth argument is this:-"At the commencement of every fresh paragraph, the initial letter in Cod.

slightly projects into the margin, beyond the left-hand edge of the column; as usual in all later MSS. This characteristic is only not undiscoverable in Cod. B. Instances of it there are in the earlier Codex; but they are of exceedingly rare occurrence.' The expression "as usual in all later MSS." is likely to mislead. There is a great difference between the style of the Sinaitic MS. and that of the Alexandrine, the Ephrem, and later MSS. generally, in respect to the mode of indicating the beginning of paragraphs. In the Sinaitic, the initial letter, which slightly projects, and often does not project at all, is no larger than the rest, a peculiarity found in but a very few existing MSS., and those the oldest known to us. In the other MSS. referred to, the initial letter, or, when the new paragraph begins in the middle of a line, the first letter of the line following, is very much larger than the others, and stands out wholly in the margin, giving these MSS. a strikingly different appearance from that of the Sinaitic and the Vatican. But the characteristic which Mr. Burgon says is "exceedingly rare," "only not undiscoverable," in the Vatican MS., occurs 10 times on the very first page of that MS.; and in the first 294 pages, viz. from Gen. xlvi.28 (or) to 1 Sam. xix.11 (ayyɛlovo), there are 1441 examples of it. Though less common in the New Testament part of the MS., in the first 8 pages it occurs 31 times. When Codex B was written, the choice between this mode of indicating the beginning of a paragraph and the other, described under Mr. Burgon's second argument, was evidently a matter depending on the taste of the copyist. In the 290 pages following the word ayyeλovo in 1 Sam. xix.11, extending to the end of Nehemiah, there are but two clear examples of it, viz. on pp. 343, 484. (The projecting letter, pp. 578 and 606, is not the first letter of a paragraph or even of a word.) In the two Books of Chronicles, the First Book of Esdras, and the Books of Ezra and Nehemiah together, there is no example of that mode of indicating paragraphs which is usual in the Sinaitic, and so common in the first 294 pages of the Vatican (pp. 41-334). The natural inference is, that we have in the part of the MS. beginning with page 335 the hand of a different scribe; and this inference is confirmed by the striking difference between these pages of the MS. and those which precede in respect to the use of > to fill up a space at the end of a line, and by other peculiarities. Even Mr. Burgon will hardly contend that the scribe who wrote page 334 of the Codex Vaticanus lived 50 or 100 years after the writer of page 335.

Both of these modes of indicating paragraphs are of an antiquity greatly exceeding that of the Sinaitic and Vatican MSS. The use of the space between words and the dash or some other mark to attract attention in the left-hand margin of the column