of the ancient versions, and especially of the LXX., have often been appealed to with the view of shewing that they cannot have been taken from pointed MSS.; but much later witnesses point in the same way. Thus in Origen's Hexapla, the second column contained the Hebrew text in Greek letters, and we are thus able to approximate to the Hebrew pronunciation of Origen's time. Making every allowance for corruption of text in the MSS. from which the extant fragments of this second column are derived, a pronunciation is indicated by no means identical with that embodied in the Massoretic pointing. Jerome again (345-420 A.D.), who spent the latter half of his life in Palestine, and has left numerous commentaries on the books of the Hebrew Bible, appears to have been entirely ignorant of any written system of points whatever. He remarks at times on the different ways of pronouncing the consonants of a Hebrew word and the consequent variation in meaning, in a way which would be inconceivable had he any written system of vowels before him. Later still we have the Babylonian Talmud, which was finished probably about 500 A.D. Among the discussions in it as to the meaning of passages of the Law, no appeal seems ever to be made to the vowel-points, though such an appeal would often have been most relevant. The earliest certain references to written Hebrew points are said to occur in the writings of Saadia Gaon († 942 A.D.), but by that time they had evidently been long established.

The Hebrew point-system is so elaborate, being indeed quite out of proportion to the simplicity of the language, that a considerable time must be allowed for its development. It I was in all probability devised by the scholars of Tiberias, and possibly fears as to the rapid growth of Islam may have impressed them with the insecurity of interpretation resting so largely on unwritten tradition.

Besides this system of pointing there exists also another, commonly known as the Babylonian, which was first noticed about forty years ago in MSS. belonging to the Karaite Jews. This is a less elaborate system than that in common use, and in it the points are always written above the line of consonants.

It cannot be too strongly insisted on here that comparatively modern as the point-system is, though we need by no means maintain its infallibility, its history is such that, only on well-established grounds, should it be departed

from.

The history of the text of the Old Testament may conveniently be broken up into three periods:-(1) from the time when the several books were written to the Christian Era,, (2) from thence to the invention of the pointsystem, and (3) from thence to the present day.

For the earlier part of the first period our materials are exceedingly scanty, a brief notice here and there in Scripture testifying to the possession by the Israelites of writings viewed as sacred. Thus Moses writes the book of the Law," and delivers it to the priests and Levites to be kept "in (or, by) the side of the ark" (Deut. xxxi. 26). Long before this time he had been commanded by God to put into writing an account of the victory of Rephidim (Exod. xvii. 14). Joshua again at the end of his life lays up his own record before the Lord (Josh. xxiv. 26). See also 1 Sam. x. 25. In

these passages we have our first glimpses of the formation of the Old Testament. How far and how often during the days of the monarchy the action of editors modified and rearranged the existing materials it is useless to guess, nor does it matter. The divine purpose was the same, whether author or editor was the one to be guided.

An important incident, the true meaning of which has excited much controversy, is the discovery of the "book of the Law" in the eighteenth year of Josiah (2 Kings xxii.). It must be remembered that after Hezekiah's death ensued the fifty-five years, mostly of idolatry and persecution, of Manasseh, the short evil reign of Amon, and the succession of a child of eight. Amid all this the worship of Jehovah might seem almost to die away. The discovery was made at a time and under conditions when it was sure to produce its effect: had it been left to the reign of Jehoiakim, for example, it would, humanly speaking, have been without avail. It is a strong argument that we are here dealing with a case in which God allowed His written Word to be withdrawn for a time, when in the writings of the prophets, who lived two or three generations before the date of the discovery (e.g. Hosea, Amos, Micah), numerous underlying allusions shew abundant knowledge of the details of the Law. Considering the nature of the reforms of Josiah, it cannot be doubted that copies of the Law would be multiplied. And here too the succession of prophets must be noted: Jeremiah in Palestine and Egypt, Ezekiel and Daniel in Babylon, were undoubtedly familiar with the Law.

After the returned exiles had once more settled in their own land, we find Ezra (Neh. viii. 1) reading the Law to the people, at their request, as an ordinary and natural thing. To Ezra and his colleagues, "the men of the great Synagogue," has been assigned the setting forth a carefully revised text of the book of the Law, with or without some of the other books of Scripture. It might be assumed from the nature of the case that this editorial work would be continued by Ezra's younger colleague, Nehemiah, and an important confirmation of this is furnished by 2 Macc. ii. 13 which connects his name with books "concerning the kings and prophets," and the writings of David. The former of these is at any rate suggestive of the second of the three volumes, and the name of David of the third.

Besides the Massoretic Hebrew text two other forms of the text have come down to us from pre-Christian times; the Samaritan Pentateuch, which, as the name tells, includes the five books of Moses only, and the so-called Septuagint Greek version, containing the whole of the Old Testament. It is true that the differences between these two and the Massoretic text are considerable, and of this we must speak hereafter; what it is important to note at present is that as compared with the mass that is in agreement, the points of differences are not great, and hardly any important point either of narrative or doctrine is affected in either of the secondary texts.

The period from the time of Ezra to that of our Lord is one of special importance in the history of the text. As has often been pointed out, their religion and the sacred books which enshrined it were all that the Jews had now left. National independence gone, language

well-nigh dead, they clung all the more firmly to what they had left, and their Scriptures became the citadel of the national aspirations. On the careful preservation and interpretation of the text labour was ungrudgingly bestowed, and the Rabbi became a far more important personage than the priest. As the outcome of all this it obviously follows that as regards Jerusalem and Palestine there would be a steady tendency to fixity of the Hebrew text.

In the New Testament a large majority of the quotations from the Old Testament follow the wording of the Septuagint, even when there is a certain amount of divergence between this and the Hebrew. This was due to the almost universal use, spite of its glaring faults, of the Septuagint.

From the Christian Era to the sixth or seventh century A.D. the evidence that meets us, such as it is, tends to bring out more and more the two foregoing points, (1) the fixity and uniformity of the traditional Hebrew text, the variations, when traces of such are found, being slight; and (2) the striking contrast between this text and that of the Greek, and, in a less degree, of other versions.

While the Rabbis of Jerusalem, and later of Tiberias, and of the Babylonian Schools, made, we know, the "word of God of none effect" by their tradition, still as regards the letter of that word, it was guarded jealously enough. The chief outcome of these schools in the above-named centuries clearly shews that. In the quotations from Scripture in the Mishnah (c. 200 A.D.), and the Talmuds of Jerusalem (c. 370 A.D.) and of Babylon (c. 500 A.D.), a certain amount of variations occurs, but within very narrow limits. The tendency clearly was to uniformity, not to variation. Jerome sometimes also refers to readings of the Hebrew not identical with those of the present Massoretic text, but these are not of sufficient weight to touch the main argument.

The first six centuries after Christ saw the rise of a large number of versions of the Old Testament. The majority of these, being taken from the Septuagint, are only of value as evidence for the text of the latter. Made directly from the Hebrew, however, are the Syriac version known as the Peshitto, which yet has some curious unexplained Septuagintal colouring, the Latin Vulgate, and various Greek versions (three of the whole of the O. T., by known authors, Aquila, Symmachus and Theodotion, and three of certain books only and anonymous, commonly known as the Fifth, Sixth, and Seventh Versions). These six Greek versions exist only in fragments, and these are of value rather for the exegesis than the criticism of the text; but little that would come under the head of a variation of text can be deduced from them. The Peshitto, as has been said, is partially ranged with the Septuagint against the Massoretic Hebrew.

Besides these, a group of versions of a different type presents itself, the Targums, paraphrases into Aramaan of different parts of the Old Testament. The date of the ultimate form of these is a matter of much dispute: and though these versions are exceedingly paraphrastic, there is in them a large element of very ancient date; and further, it may broadly be laid down that the Targums support the Massoretic text as against the opposing versions.

Some results of the criticism of the text by the doctors of the Rabbinic schools may here be noticed. They are usually designated Qri and Kthib, by which is meant that, though a certain word is found in the text (Kthib= written), the direction is given to replace it in reading" by another word (Qri=read). They may for convenience be grouped under three divisions:-(a) modernizing, where certain forms are replaced by those current in later centuries, (B) euphemistic, where coarse or indelicate phrases are softened down, and (y). what may more strictly be called critical, Some have viewed the Qri merely as a critical conjecture of the Rabbis; others, we believe with more justice, as an independent reading resting on definite objective evidence. A familiar instance by way of illustration will be found in Psalm c. 3, where the E. V. follows the Kthib, "and not we ourselves," the margin, with more point, following the Qri, "and His we are."

An extreme case is that of the "Qri without a Kthib," and the "Kthib without a Qri," that is, where a word has to be read which is not in the text (e.g. the word Euphrates in 2 Sam. viii. 3), and where, though in the text, a word has to be passed over.

It has already been pointed out that the date of the invention of the point-system is a very uncertain one, the seventh century A.D. being, however, a reasonable guess. The system of points and the recension of text furnished with them are known as Massoretic, an adjective derived from Massorah, tradition. Under this name is included a mass of details as to the criticism of the text, all the various corrections and variations mentioned above, and others also: such things as the number of verses or letters in the various books, which is the middle verse of each book, and numerous similar things, puerile enough in one sense, and yet testifying to the extraordinary care and zeal lavished on the books.

One set of corrections rather more important than some of those mentioned above is the socalled Tiqqun Sopherim, or "correction of the scribes," a term applied to eighteen places where a certain reading is rejected and replaced by another, whether on critical or subjective grounds we cannot here discuss. An example may be given. In Hab. i. 12 the Massoretic text reads, "we shall not die," which is a "correction of the scribes" for "thou wilt not die," surely a most frigid and suspicious expression.

A short abridgment of the Massorah is printed in the margin of ordinary Hebrew Bibles; the whole body of the Massorah, as may well be imagined, is of very considerable extent.

There were two great centres of Hebrew learning in the centuries of which we have been speaking, Tiberias and the Babylonian Schools. Hence two recensions of text were gradually formed, Western and Eastern. Of these two recensions, two lists of variations may be mentioned:-(1) A list, ordinarily known as "the Western and Eastern readings," 220 in number. These, we may believe, are of a date prior to the rise of the pointsystem, because they all refer to the consonants. None of them is of any importance as regards the sense. It is important to add that not one of them occurs in the Pentateuch, a strong piece of evidence to the exceptional

care taken with the MSS. of the Law. (2) The full establishment of the point-system would in course of time furnish a fresh field for variation. About 1030 A.D. the Western and Eastern MSS. were examined for variations by Aaron ben Asher of Tiberias and Jacob ben Naphtali of Babylonia. This list of variations, which has been several times printed, consists of about 900 readings, which, with one exception, refer to vowels and accents only. The present printed Hebrew text is almost entirely according to the readings of ben Asher. In the multiplication of fresh copies there would always be the desire to follow a carefully revised text. Accordingly various standard copies are mentioned by the medieval Rabbis as those serving for patterns: thus Maimonides († 1204 A.D.) refers to the MS. of the Pentateuch revised by ben Asher, and states that he himself made a transcript of it in Egypt. Another important MS. was the Codex of Hillel, mentioned by the great Rabbi David Kimchi († 1240 A.D.) as seen by him in Spain, and held in high repute as an authority, Others again are the Codices of Jericho and Sinai, both containing the Pentateuch only. The Hebrew Old Testament was printed a considerable number of years before the Greek New Testament, which was not printed till 1516. It appeared first in piecemeal fashion the first portion being the Psalms, published in 1477, unpointed, with Kimchi's Commentary. The place of printing is unspecified. The Law was printed at Bologna in 1482, the Prophets at Soncino in 1485-86, and the Hagiographa at Naples in 1487. The first complete edition of the Hebrew Bible is that of Soncino in 1488; and the first Christian edition is that contained in the Complutensian Polyglott, Alcalà, 1514-17.

The most important editions of the text since that time have been those of Bomberg (Venice, 1518, 1526, 1547-49), to which several commentaries of the great Rabbis were added; the Rabbinic Bible of Buxtorf (Basle, 1611); Athias, Amsterdam, 1667. This last was reproduced by Van der Hooght (Amsterdam and Utrecht, 1705), and on this the great majority of modern editions have been based.

TEXTUAL EVIDENCE for the OLD TESTAMENT. We must now view the matter from the other side. The direct evidence for the text of the Old Testament comes under three heads, that of Hebrew MSS., of Ancient Versions, and of the Quotatious from the Old Testament in the Talmud and other ancient Jewish writings.

Two points at once strike the student of the Hebrew MSS. of the Old Testament: first, the comparatively late date of the earliest known MSS.; and secondly, the exceedingly narrow limits within which the variations of these MSS. fall; not that these variations are few in number, but as a rule most trifling in importance.

There are but few MSS. as yet known which can be assigned with probability to an earlier date than the year 1000 A.D.

The Hebrew MSS. of the Old Testament may be divided into Synagogue-rolls and MSS. for personal use. As to the former, the Talmud lays down elaborate rules, as to the nature of the skins and fastenings, the number of columns in each, the size of each column and the title.

MSS. other than Synagogue-rolls are in book form of various sizes, from folio downwards. Sometimes these are dated, but when no date is given, it may be more or less approximated to by various internal characteristics. The MSS. fall into various families, distinguished by different sorts of characters, as well as certain differences of text. These are known as (a) the Spanish, written in square, elegantly-formed letters. These represent the text of the Codex of Hillel, and give the Massoretic text in its closest form. With these on the whole agree the Oriental family of Hebrew MSS. (8) the German. These are written in a comparatively rude and inelegant character, and also display a certain amount of divergence from the Massoretic text. Indeed, the books of the "prophets" and Hagiographa are arranged in a slightly different order from that in the Massoretic text. (7) Midway between these come the Italian MSS., both as regards the shape of the letters and the character of the text.

The first attempt to examine the text of a large number of Hebrew MSS. was that of Kennicott, in whose edition of the Hebrew Bible (Oxford, 1776-1780) a collation is given of 634 Hebrew MSS., besides printed texts of the Bible, copies of the Talmud, &c. Subsequently, De Rossi published (Parma, 17841798) a collation of 825 MSS.

In the present century, Hebrew MSS. of the Bible have been collated by Pinner at Odessa, and others have come to light in India and China.

The great mass of them date from the beginning of the eleventh century onwards. The following may be mentioned: (1) a Cod. Laudianus, no. 1 of Kennicott, and by him assigned to the tenth century, though placed later by subsequent critics. (2) De Rossi's Cod. 634, assigned by him to the eighth century. (3) Pinner's Cod. 1 of the Pentateuch. This has a subscription, stating that it was corrected in 580 A.D., and if this be true, it would be the oldest MS. of the Hebrew Bible known. Some internal phenomena, however, point to a much later date. (4) Pinner's Cod. 5, an imperfect copy of the Pentateuch, the date of the subthe Prophets, the date of the subscription scription being 843 A.D. (5) Pinner's Cod. 3 of being 916 A.D. In this MS. the vowels and accents are not those in common use, being written above the line of the letters, as we have mentioned above. (6) Pinner's Cod. 13, an imperfect copy of the "former prophets,' claiming to have been written not later than 938 A.D. To these may be added (7) a MS. in the University Library at Cambridge (Kennicott 89), whose subscription would assign it to the year 856 A.D. Opinions are much divided as to the truth of this statement. The great majority of Hebrew MSS. do not contain the whole Bible, but only parts thereof.

The scrupulous care bestowed upon the Hebrew text, as far as our evidence extends, contrasts strongly with the free handling to which the other texts were subjected, and this contrast in itself furnishes a strong prima facie case in favour of the Hebrew.

When we come to view the matter in detail, we find, on throwing out those versions which, being taken from the Greek and not from the Hebrew, have no independent value, that there remain the Septuagint and the fragments of the other Greek versions, the Chaldee

Targums, the Peshitto Syriac, and the Latin Vulgate; and also, though not a version, the Samaritan Pentateuch. Of these, the Targums and the later Greek fragments cannot be said to point to a text differing materially from the Massoretic. Nor, broadly speaking, can a different result be said to be yielded by the Latin Vulgate, made as it was in Palestine and under the influence of Jerome's Palestinian Jewish teachers; though where the Septuagint did not differ too materially from the Hebrew, he follows the former rather than the latter, so as not to break with old associations more than could be helped.

The Peshitto Syriac not unfrequently agrees with the Septuagint against the Hebrew, and this is doubtless due to the high position held everywhere in the Christian Church by the former; though it is by no means easy to see whether in these points subsequent interpolations have caused the difference, or whether the Septuagint was an influence on the mind of the original translator. Probably the first cause alone would be insufficient to account for the phenomenon.

There remain then the Septuagint and the Samaritan Pentateuch. The variations of the latter from the Massoretic text are often, but very far from invariably, supported by the Septuagint. At one time, the most extravagant claims were put forward on its behalf, scholars such as Morinus, Houbigant and Kennicott, placing it on a decidedly higher level than the Massoretic text. But Gesenius in his work on the subject (1815) gave a systematic arrangement to the variations between the two texts and shewed that the mass of variations are due to attempts at grammatical emendation, to glosses, to corrections of apparent deficiencies in the sense, to adaptations in favour of the Samaritan view of things and the like.

There remains finally the Septuagint. Real as the differences often are between this and the Hebrew, several points may be alleged to lessen them. The current text of the Septuagint is in a woeful state; the most recent editions leaving untouched a multitude of passages where the correct Greek of the original

translation has obviously become corrupted. Very many variations are also due to difference of interpretation and not of text, unpointed Hebrew MSS. giving great latitude to interpretation. Very many of those which remain hinge on the difference between similar Hebrew letters. Some of the variations of the Septuagint are doubtless correct, though probably no two scholars will exactly agree as to which they are; nor will any important difference be made, on any view, either as to history or theology.

It may sufficè here to mention one where we cannot doubt that the Massoretic text is corrupt. In Psalm xxii. 17 (16 E. V.) for "they pierced' the Massoretic text has "like the lion," the difference in the Hebrew being simply that between the two similar letters, Vay and Jod. Every known ancient version makes the word a verb, the Vav (which renders the word a verb) is found in many Hebrew MSS., and there is a certain amount of traditional evidence supporting that view. On all these grounds we cannot hesitate to believe that the word is a verb (meaning probably “to pierce").

The Talmud, as we have already said, consists of the Mishnah, completed at Tiberias about the end of the second century A.D., on which two commentaries called Gemaras (or "completion") arose in Palestine and Babylonia, and were finished about 370 A.D. and 500 A.D. respectively. The Midrashim are expositions or expansions of the various books of the Bible, and are probably very composite as regards dáte.

In all these a considerable quantity of the Bible is quoted or embodied; yet it cannot be said that the various readings thus detected are of any noteworthy importance. There appear to be 14 various readings known from the Mishnah, and in the collation of the quotations in the Talmud for Kennicott about 1000 variations were noted. None, however, was of any special moment. Such evidence, therefore, as is afforded by this third head is strongly and absolutely confirmatory of the Massoretic text. On the Apocrypha and other Apocryphal books see above, p. 9.

2.

Object. The object of Textual Criticism is the restoration of the exact words of a writing, when the original copy of it has been lost.

Materials. The materials available for the determination of the exact words of the lost originals of the different books of the New Testament, and the skill to use them, have varied in different ages. At the present day three kinds of documentary sources are recognised, and have been made available for the purposes of criticism by the self-denying labours of generations of students. They are Greek Manuscripts, or written copies of the Greek Text of the whole or parts of the different books;

Versions, or translations from the Greek into some other language;

'Fathers,' i.e. the testimony borne by ecclesiastical writers, directly or indirectly, to the readings of MSS. in their own day.

GREEK MANUSCRIPTS. These are divided into two classes, according to the characters in which they are written. Some are written entirely in capital letters, and are called Uncials. Others are written in smaller letters and a running hand, and are called Cursives or Minuscules. Uncial MSS. are, as a class, older than Cursives. No Uncial is later than the eleventh century; no Cursive earlier than the ninth. The oldest MSS. are written with no breaks between the words and very few stops. For purposes of reference an Üncial MS. is denoted by a capital letter (A, B, A, ) X); a Cursive by an Arabic numeral (1, 2, 3....).