CHAPTER V. THE OTHER VERSIONS OF THE NEW TESTAMENT.

The remaining Versions are of less importance in the ascertainment of the sacred text. But some of them have recently received more attention in the general widening of research, and in becoming better known have strengthened their claims to recognition and value. Three of them, at all events, date from the period of the oldest manuscripts of the New Testament now known to be in existence. And the presence amongst us of eminent scholars acquainted with them renders reference to them more easy than it was a few years ago.

Nevertheless, some are of slight service to the critic, being secondary versions, and as such becoming handmaids, not of the Greek, but of some other version translated from the Greek.

In the account of these versions, the Editor of this edition is indebted for most valuable assistance to Mr. F. C. Conybeare, late Fellow of University College, Oxford, who has re-written the sections on the Armenian and Georgian versions; to Professor Margoliouth, who has also re-written those on the Ethiopic and Arabic; to the Rev. Llewellyn J. M. Bebb, Fellow of Brasenose College, who has re-written the account of the Slavonic; and to Dr. James W. Bright, Assistant-Professor of English Philology in the John Hopkins University, who has contributed what is known on the Anglo-Saxon Version.

(1) The Gothic Version (Goth.).

The history of the Goths, who from the wilds of Scandinavia overran the fairest regions of Europe, has been traced by the master-hand of Gibbon (Decline and Fall, Chapters x, xxvi, xxxi, &c.), and needs not here be repeated. While the nation was yet seated in Moesia, Ulphilas or Wulfilas [318-388], a Cappadocian, who succeeded their first Bishop Theophilus in A.D. 348, though himself an Arian and a teacher of that subtil heresy to his adopted countrymen, became their benefactor, by translating both the Old(118) and New Testament into the Gothic, a dialect of the great Teutonic stock of languages, having previously invented or adapted an alphabet expressly for their use. There can be no question, from internal evidence, that the Old Testament was rendered from the Septuagint, the New from the Greek original(119): but the existing manuscripts testify to some corruption from Latin sources, very naturally arising during the occupation of Italy by the Goths in the fifth century. These venerable documents are principally three, or rather may be treated under two MSS. and one group.

1. CODEX ARGENTEUS, the most precious treasure of the University of Upsal, in the mother-country of the Gothic tribes. It appears to be the same copy as Ant. Morillon saw at Werden in Westphalia towards the end of the sixteenth century, and was taken by the Swedes at the siege of Prague in 1648. Queen Christina gave it to her librarian, Isaac Vossius, and from him it was very rightly purchased about 1662 by the Swedish nation and deposited at Upsal. This superb codex contains fragments of the Gospels (in the Western order, Matthew, John, Luke, Mark) on 187 leaves, 4to (out of 330), of purple vellum; the bold, uncial, Gothic letters being in silver, sometimes in gold, of course much faded, and so regular that some have imagined, though erroneously, that they were impressed with a stamp. The date assigned to it is the fifth or early in the sixth century, although the several words are divided, and some various readings stand in the margin _primâ manu_.

2. Codex Carolinus, described above for Codd. PQ, and for the Old Latin _gue_, contains in Gothic about forty verses of the Epistle to the Romans, first published by Knittel, 1762.

3. Codices Ambrosiani, or palimpsest fragments of five manuscripts, apparently like Cod. Carolinus, from Bobbio, and of about the same date, discovered by Mai in 1817 in the Ambrosian Library at Milan, and published by him and Count C. O. Castiglione (Ulphilæ Partium Ineditarum ... Specimen, in five parts, Milan, 1819, 1820, 1834, 1835, 1839). The last-named manuscripts are minutely described and illustrated by a rude facsimile in Horne’s “Introduction,” and after him in Tregelles’ “Horne,” vol. iv. pp. 304-7. They consist of (1) a portion of St. Paul’s Epistles, under Homilies of Gregory the Great (viii); (2) portions of St. Paul, under Jerome on Isaiah (viii or ix); (3) parts of the Old Testament, under Plautus and part of Seneca; (4) under four pages of St. John in Latin part of St. Matt. xxvi, xxvii. The fifth fragment consists of Acts of the Council of Chalcedon with no extracts from the Bible. Mai refers some of the Gothic writing to the sixth century and some as far back as the fourth or beginning of the fifth. Unlike the Codex Argenteus (at least if we trust Dr. E. D. Clarke’s facsimile of the latter), the words in Mai’s palimpsests are continuous: they contain parts of Esther, Nehemiah (apparently no portion of the books of Kings), a few passages of the Gospels, and much of St. Paul(120). H. F. Massmann (Ulfilas, Stuttgart, 1855-57) also added from an exposition a few verses of St. John, and there are fragments at Vienna and Rome(121).

These fragments (for such they still must be called)(122), in spite of the influence of the Latin, approach nearer to the received text, in respect of their readings, than the Egyptian or one or two other versions of about the same age; and from their similarity in language to the Teutonic have been much studied in Germany. The fullest and best edition of the whole collected, with a grammar and lexicon, is by H. C. von der Gabelentz and J. Loebe (Ulfilas Vet. et N. Testamenti versionis Gothicae fragmenta quae supersunt, Leipsic, 1836-46, viz. vol. i. Text, 1836; Pars ii. Glossarium, 1843; Pars ii. Grammatik, 1846), and of the Codex Argenteus singly that of And. Uppstrom (with a good facsimile), Upsal, 1854. This scholar published separately in 1857 ten leaves of the manuscript which had been stolen between 1821 and 1834, and were restored through him by the penitent thief on his death-bed. The Gothic Gospels, however, had been cited as early as 1675 in Fell’s N. T., and more fully in Mill’s, through Francis Junius’ edition (with Marshall’s critical notes), which was printed at Dort in 1665, from Derrer’s accurate transcript of the Upsal manuscript, made in or about 1655, when it was in Isaac Vossius’ possession. Other editions of the Codex Argenteus were published by G. Stiernhielm in 1671 for the College of Antiquaries at Stockholm; by E. Lye at the Clarendon Press in 1750 from the revision of Eric Benzel, Archbishop of Upsal; and (with the addition of the fragments in the Codex Carolinus) by Jo. Ihre in 1763, and by J. C. Zahn in 1805. And also the Gothic and Anglo-Saxon Gospels in parallel columns with the Versions of Wycliffe and Tyndale, London, 1865, and Ulfila, oder die Gotische Bibel (N. T.), E. Bernhardt, Halle, 1875, and St. Mark with a grammatical commentary, R. Müller and H. Hoeppe, 1881, and Skeat, Gospel of St. Mark in Gothic, Clarendon Press, 1882.

(2) The Armenian Version.

The existing Armenian version is a recension made shortly after the Council of Ephesus of a still earlier version, which was based in part upon a Syriac, in part upon a Greek original. This latest recension was made according to “accurate and reliable copies” of the Greek Bible, which, along with the Canons of the Council of Ephesus, were brought from Constantinople about the year 433. One would naturally wish for more details than the above brief statement contains; yet it is all that one can definitely infer from the history of the version as related by three nearly contemporary writers, whose accounts we now subjoin, namely, Koriun, Lazar of Pharpi, and Moses Khorenatzi.

Koriun(123) in his life of St. Mesrop (written between 441 and 452 A.D.) relates as follows:—

In the fifth year of the reign of Vramshapho [i.e. about 397 A.D.], St. Mesrop was first in Edessa, then in Amid, lastly in Samosata, busy all the time about his discovery of the Armenian characters(124). In Samosata, where he was received with great respect by the clergy and bishop, Mesrop met with a Greek scribe, Hrofanos (? Rufinus), in conjunction with whom, and with the help of two pupils named John and Joseph, he undertook a translation of the Bible. They began—and this is noteworthy—with the book of the Proverbs of Solomon; Hrofanos or Rufinus writing down the translation with his own hand. Mesrop next visited the Bishop of the Syrians, who congratulated him on his work. He then returned to Nor Chalach, or new city, as Valarshapat was called by the Romans, in the sixth year of Vramshapho’s reign, A.D. 398. At a later time, Koriun, the writer, was himself sent with Eznik to Constantinople, apparently in quest of books to translate; for they returned with a _sure_ copy of the Scriptures, with works of the Fathers, and with the canons of the Councils of Nice and Ephesus. “Now St. Sahak had long before translated the collection of Church books from Greek into Armenian, as well as much true wisdom of the holy Patriarchs. But he now resumed, and taking with the help of Eznik the former translations made hurriedly and offhand, he confirmed them by the help of the true copies now brought, and they translated much commentary on the books.” The above is the gist of what Koriun has to tell us, though he mentions that scholars were sent to Edessa to translate and bring back the works of the Fathers. Why Mesrop began with the Book of Proverbs, whether he translated more than that, and from which language, we do not learn from Koriun. Lazar of Pharpi(125), who wrote in the last half of the sixth century, is our next authority. He states that up to the last decade of the fourth century, the offices of religion were still read in Greater Armenia in Syriac, a language which the people did not understand. The edicts of the kings of Armenia were also written out in Syriac or Greek characters. But as soon as the Armenian alphabet was discovered, St. Sahak—who was patriarch 390-428 A.D. and an expert in Greek—set himself, in response to the patriotic exhortations of St. Mesrop, of Vramshapho the king, and of the clergy and nobles, to translate the Holy Scriptures. He states that St. Sahak’s version comprised the whole of the Old and New Testaments, and was made from Greek.

Moses of Chorene, bk. iii. ch. 36 ff., copies, confuses, and adds to Koriun’s account. A little before 370 A.D. the Persians overran Armenia, and Meroujah, their leader, burned all the books he could find in the country, proscribed the study of the Greek language, and enacted penalties against any who should speak it or translate from it. At that time, adds Moses, the offices of the church were performed in Greek, because the Armenian alphabet did not yet exist. On the death of Theodosius (Jan. 395 A.D.) there was a partition of Armenia between his successor Arcadius and the king of Persia, by which the latter took undisputed possession of the eastern provinces, including the basin of Ararat, in which lay the new religious centre Valarshapat or Edschmiadzin, the νέα πόλις of the Romans. The new Mesropic alphabet was at first used only in Persian Armenia; for, says Moses, in the parts dependent on the Greeks, all writing had to be in Greek characters, Syriac being forbidden. As soon as Mesrop had elaborated his alphabet with the aid of Hrophanos, he betook himself to the work of translation; and with the aid of his pupils John and Joseph, translated the entire twenty-two authentic books along with the New Testament, taking care to begin with the Book of Proverbs. About the year 406 he returned to Armenia, and found St. Sahak engaged in translating the Syriac Bible. He hints that Sahak would have preferred a Greek original, if Meroujah had not burned all the Greek books nearly thirty years before. This perhaps implies that the version, on which Mesrop had been engaged in Samosata, was made from Greek. Nor is that unlikely; for Rufinus, who helped him, was a Greek, and we learn from Koriun that there were Armenians in Edessa studying both Greek and Syriac. We read in bk. iii. ch. 60 of the History of Moses, about missions sent to Edessa and Byzantium in order to the translation of the works of the Fathers, but we hear nothing more expressly touching the Version of the Bible, save this, that after the Council of Ephesus, Sahak and Mesrop, then in Ashtishat in Taron, received from Byzantium, as aforesaid, the canons of the council recently held, along with accurate copies of the Greek Bible. On receipt of these, Sahak and Mesrop translated afresh what had already been translated, and were zealous in recasting the text. But they were not, it seems, after all, satisfied with their work, and sent Moses to Alexandria to learn the “beautiful tongue” (i.e. Greek), with a view to a more accurate articulation and division (of the Armenian scriptures).

The above summary exhausts the evidence of Moses of Khorene(126). It would appear therefrom that the Bible was translated twice into Armenian before the end of the fourth century; by Mesrop from Greek, and by Sahak from Syriac. The circumstance that Mesrop in Samosata began with the Proverbs of Solomon raises a suspicion that the earlier books had already been rendered, when and by whom is unknown. Certainly the reasons given by Koriun and by Moses for Mesrop beginning with Proverbs are insufficient. Moses again in stating that Sahak rendered the entire Bible from Syriac contradicts both Koriun and Lazar. Are we to infer that Sahak and Mesrop after 430 A.D. retranslated according to the Constantinople Bibles what they had already translated from Syriac, and also it would seem from a presumably less perfect Greek text? Anyhow it is unlikely that they would wholly sacrifice their own work, and we should therefore expect to find in the Armenian version a mixture of texts, namely of some old Syriac text, which must have been in vogue as late as 380, of some older Greek text supplied in Edessa or Samosata, and of the Constantinopolitan texts; which last may well have been among the fifty splendid copies which had been prepared under the order of Constantine by Eusebius a century before. If, and how far, these different elements enter into the Version can only be determined by a careful analysis of its readings. It may be that in some MSS. there lurks more of the unrevised text than in others(127). The entire history is an apt illustration of that political see-saw between the Roman and the Persian powers which went on in Armenia during the fourth and fifth centuries, and out of which the patriotic vigour and devotion of St. Mesrop and St. Sahak carved at last a truly national Armenian Church, with an independent life and literature of its own.

The Armenian Version was collated for Robert Holmes’ edition of the Septuagint, though not with desirable accuracy nor from the oldest MSS. For example, the Codex Arm. 3 of the Pentateuch, which Holmes declares, _teste Adlero_, to be of the year 1063, is but an eighteenth century codex. The collation of the New Testament in the eighth edition of Tischendorf’s N. T. is accurate so far as it goes, but is far from being exhaustive or based on a consensus of the oldest MSS. Old codices of the Armenian Gospels are very common, and the present writer knows of as many as eight, none of them later than the year A.D. 1000; of four of these he has complete collations. The rest of the N. T. is only found in codices of the whole Bible, which are rare and always written in minuscules, never in uncials as are the Gospels. He knows of no copies of the whole Bible older than the twelfth century.

Two further questions call for brief answer:—1. Have we the Armenian version as it left the hands of the fifth century translators 2. Did the fifth century version comprise the whole of the Old and New Testament?

In regard to the first question, it must be admitted as probable that changes were subsequently made, at least in the New Testament, in the way both of omission and addition; e.g. in St. Luke xxii. 44, out of four very early uncial codices collated by the writer, the words: ἐγένετο δὲ ὁ ἱδρῶς αὐτοῦ ὡσεὶ θρόμβοι αἵματος καταβαίνοντες ἐπι τὴν γῆν, are found only in one, and that one the earliest, being dated 902 A.D. The words which precede ὤφθη δ—ὲπροσηύχετο are omitted in all four of them. We may infer that ver. 44 was in the original version, and was omitted from the three codices for doctrinal reasons. The additions made to the text after the fifth century are easier to detect; because they only come in some MSS. and not in others, and also because there is so much discrepancy of readings between those codices which add them, that they are at once seen to be lacunas supplied by different hands. This is the case, for example, with the end of St. Mark’s Gospel, which only comes in one of the four codices mentioned, namely in the oldest Edschmiadzin Codex, under the heading “of the Elder Ariston,” which may refer to Aristion, teacher of Papias, or to Ariston of Pella. The case is the same with the episode of the woman taken in adultery. For the settlement of such points there is wanted a careful collation of the oldest codices.

In answer to the other question we may state, without entering into the proof of it, that the fifth century version included all the books of the Old and New Testament save the third book of Ezra, Esther, Tobit, Judith, Wisdom of Solomon, and perhaps the Maccabees. For as we read in Elisaeus that Vartan Mamikonean in the middle of the fifth century inspired his troops to deeds of valour against the Persians by reading to them the Book of Maccabees, we may fairly infer that that also was already then rendered. It may be added that the Psalms were rendered for church use prior to the rest of the Bible, and were translated afresh by Mesrop and his disciples; also that the Book of Revelations was translated twice. The double translation of both these books is a fact which can be traced in various MSS.

One other point must be noticed. From the history of Moses of Chorene, it is not clear what were the imperfections of the Armenian version, to remedy which Moses was sent to Alexandria. We cannot suppose that Mesrop and Sahak and Eznik, and the other doctors who had already translated the Greek codices brought from Byzantium, were incompetent Greek scholars. The object therefore of Moses’ voyage to Alexandria was probably that he might add to the Armenian text the Sections of Ammonius, and also the asterisks and obeli of Origen’s Hexaplaric copy(128). The Ammonian Sections are found in all Armenian New Testaments, and in some copies of the Bible the Origenian marks as well; for instance, in Codex 3270 of the Bibliotheca Vindobonensis. There is no evidence that the Armenians ever used a version of Tatian’s Diatessaron.

The following is a list—not exhaustive—of the oldest known codices of the Armenian Gospels, or “Avetaran”:—

1. In the Library of the Lazareffski Institute in Moscow, written in large uncials on parchment, dated in the year 336 of the Armenian era = A.D. 887. Size, 37.75 × 28 cent.; 229 folios.

2. In the Library of the Mechitarists in the island of San Lazaro, in Venice, an uncial codex, on parchment, written in the year 351 of the Armenian era = A.D. 902.

3. In the same Library, on parchment, in large uncials, dated 1006.

4. In the same Library, in large uncials, on parchment, undated, but evidently older than No. 2.

5. In the Patriarchal Library of Edschmiadzin in Russian Armenia, No. 222 of the printed catalogue of Jacob Kareneantz (Tiflis, 1863). This book is bound in ivory covers, carved, as it would seem, in the Ravennese style in the fifth or sixth century. In large uncials, on parchment, written A.D. 989.

6. In the same Library is No. 223, an uncially written parchment codex. The earliest of the colophons dates from A.D. 1260 and is in majuscule, but the codex itself seems to be at least two centuries and a half earlier.

7. In the same Library, No. 229, written in miniscule, on parchment, A.D. 1035.

8, 9. In the same Library, Nos. 224, 225, in large uncials, on parchment, presumably as old as the eleventh century, but undated.

10. In Tiflis, in an Armenian church. In large uncials, on parchment. Undated, but certainly prior to A.D. 1000.

11. In the Library of the British Museum, in large uncials, on parchment, undated. Probably of the ninth century, but not after the tenth, according to Dr. Baronean, author of the British Museum Catalogue.

12. In Karin or Erzeroum, in large uncials, on parchment. Dated A.D. 986.

13. In the Library of the Fathers of St. Anthony, in Constantinople. Dated A.D. 960.

14. In the island monastery of Sevan, on the lake of that name in Russian Armenia. In large uncials, on parchment. Written during primacy of Vahan, _circa_ A.D. 966.

15. In uncials, on parchment; written in Macedonia, under the Emperor Basil, A.D. 1011. (Carékin, Catalogue des Traductions, omits to specify in what library.)

16. Bibliothèque Nationale in Paris. Codex Armenus VII contains the Four Gospels. Codex Bombyc, litteris uncialibus scriptus.

17. In the same Library, Cod. Arm. VIII. Membranaceus, litteris uncialibus scriptus.

(3) The Ethiopic Version (Eth.).

The Ethiopic translation of the Bible is assigned by Guidi to the end of the fifth, or beginning of the sixth century, the time at which Christianity became the dominant religion in Abyssinia. That religion after a period of decadence began to flourish again in the twelfth century, but in dependence on the Patriarchate of Alexandria. The two principal classes of Ethiopic Biblical MSS. are connected with these periods respectively; the first class being derived from the Greek text before, and the latter after the Alexandrian recension. The corrections, however, vary in different copies, and appear to be the result of desultory rather than of systematic alteration. The MSS. of the Ethiopic N. T. are rarely complete; ordinarily the Gospels, the Epistles of St. Paul, and the Catholic Epistles with the Acts and the Apocalypse constitute separate volumes. The oldest copy of the Gospels would seem to be no. 32 of the Bibliothèque Nationale in Paris, written in the reign of Yekūnō Amlāk; whereas MS. 33 of the same collection represents the later text. Examples of the different recensions are given by Guidi, Atti della R. Academia dei Lincei: Classe di scienze morali &c., iv. 1888, from whom most of the above statements are taken.

Copies of the N. T., especially of the Gospels, are to be found in most collections of Ethiopic MSS.; _see_ especially Wright, Ethiopic MSS. of the British Museum, pp. 23-39, and Zotenberg, Catalogue des MSS. Éthiopiens de la Bibliothèque Nationale (nos. 32-48; in the preface to this latter work a list of other collections are given); also Dillmann, Abessinische Handschriften der Königlichen Bibliothek zu Berlin (no. 20, the four Gospels; 21, the Gospel of St. John); D’Abbadie, Catalogue Raisonné de MSS. Éthiopiens (Paris, 1859; nos. 2, 47, 82, 95, 112, 173, the four Gospels; no. 119, St. Paul’s Epistles; no. 164, Catholic Epp., Apoc., and Acts); Dillmann, Catalogus MSS. Aethiop. in Bibliotheca Bodleiana, nos. 10-15; Fr. Müller, Aethiop. Handschriften der K. K. Hofbibliothek in Wien (Z. D. M. G., xvi. p. 554, no. v, the Gospels; no. vi, St. John’s Gospel); “Bulletin Scientifique de S. Pétersbourg,” ii. 302 (account of a MS. of the Gospels in the Asiatic Institute at St. Petersburg), iii. 148 (account of a MS. of the four Gospels, bearing the date 78 = 1426 A.D., in the Public Library at St. Petersburg, and another of St. John’s Gospel).

The Ethiopic N. T. was first printed in Rome, 1548, cum epistola Pauli ad Hebraeos tantum, cum concordantiis Evangelistarum Eusebii et numeratione omnium verborum eorundem. Quae omnia curavit Fr. Petrus Ethyops auxilio priorum sedente Paulo iii. Pont. Max. et Claudio illius regni imperatore (edition of Tasfā Sion). The remaining thirteen Epistles of St. Paul were printed in 1549. This edition was reproduced in the London Polyglott. Another was issued by T. P. Platt (for the Bible Society) in 1830, reprinted 1844 and 1874. These editions are based on MSS. containing mixed recensions, and are therefore of no critical value.

(4) The Georgian Version (Georg.).

The Church of the Iberians was founded during the reign of Constantine according to tradition; though, if we consider how intimate and frequent had been from a much earlier period their intercourse with the Greeks, we may safely infer that the seeds of Christianity had been long before sown among them. There is no certain evidence of the date at which they translated the Scriptures; but it is probable that their version of the New Testament was made in the fifth and sixth centuries; and that it was made from a Greek text the most perfunctory examination suffices to prove. According to Armenian historians of the fifth century, St. Mesrop, at the same time that he invented the Armenian characters and made the Armenian version for his own countrymen, fulfilled the same service for the Georgians also. In this tradition, however, the Georgians do not concur; and, no doubt, rightly, seeing that their ancient alphabet and their version are alike independent of the Armenian. It is said by some native Georgian scholars that before the tenth century a revision was made of their version, in order to make it more complete.

The present writer knows of no manuscript of the entire Bible in Europe except at Mount Athos, where there is one reputed to be of the tenth century. Others are preserved in the Convents of the Holy Cross at Jerusalem, and of Mount Sinai. In the Vatican Library there is a codex of the New Testament, neatly written on parchment in majuscule, parts of which the present writer has collated with the printed text. This codex is at least as old as the thirteenth century, and in the collations below is referred to as _a_. Beside this codex the writer has examined in the Georgian Library at Tiflis three very ancient codices of the Gospels, written in uncials on parchment. These books were smaller in size than are, as a rule, the copies of the Gospels used in Eastern Churches.

Of the accompanying collations, nos. i-iv are made from them, and the passages collated were photographed by the present writer. These photographs, which represent the originals on a reduced scale, have been deposited by him in the Bodleian Library for the inspection of the curious. The text referred to as _b_ is probably of the tenth century or earlier; the one referred to as _c_ cannot be much later than the eleventh, while that indicated by _d_ must belong to the twelfth, and is the most beautifully written of them.

The Bible was not printed in Georgian until the year 1743 at Moscow in large folio. It is a rare volume, and has never been reprinted. The character is that called ecclesiastical or priestly majuscule, which differs wholly from the civil characters and can, as a rule, be read by the priests only. The New Testament and Psalms have been reprinted at various times from this original edition, both in priestly and civil characters, and of the latter kind very good and cheap copies can be obtained at the British and Foreign Bible Society, printed, however, at Tiflis. It is said that the edition of 1743 was conformed to the Slavonic version of the Bible; and if this were true, it would, of course, impair its value for critical purposes. Of this statement, however, the writer’s collations, so far as they go, afford no proof. Such variations as there are between the printed edition and the manuscript texts are notified in these collations. The point, however, could easily be settled by a thorough comparison of the printed text with the Slavonic.

The MSS. of Tiflis include the last verses of Mark, and the Vatican MS. contains the narrative of the woman taken in adultery, but places it after ver. 44, instead of after ver. 52 of the seventh chapter of John. The printed edition places it after ver. 52, and this uncertainty as to where to insert the narrative, in itself indicates that it is a later interpolation. The printed text also contains the text about the three witnesses; but it is pieced into the context in an awkward and ungrammatical way; and whether it is in any MS. the writer cannot say. The following all too brief collations prove that the printed text fairly represents the MSS.; from which, indeed, it differs very little except in its more modern orthography. It is certain, however, that the most ancient MSS. of this version must be collated and a critical text of it prepared, before it can be quite reliably used as an early witness to the Greek text in regard to any particular points. Where the earliest Greek authorities waver as to the particles by which the parts of the narrative shall be connected—some, e.g. giving καί, others δέ, others οὖν—the Georgian constantly passes abruptly to the new matter without any connecting particle at all—and this, although as a language Georgian is richer in such connecting particles than is Greek. This peculiarity of the version, which is also shared by the old Armenian version, seems to prove that it was made from a primitive text, in which editors had not yet begun to smooth away the sudden transitions.

(5) The Slavonic Version (Slav.(129)).

This version of the Bible is ascribed to Cyril and Methodius, who lived at the end of the ninth century. It is uncertain, however, how much of the New Testament was translated at that date, and how much was the work of a later time. The manuscripts of the version exist in two characters called Glagolitic and Cyrillic: of these it is now generally agreed that the former is the earlier. In considering the version from the point of view of the textual criticism of the New Testament, we need not deal with its later history except in so far as that throws light on its original form. The chief points to which reference will be made will be (i) the different Manuscripts in which the version exists, with their distinctive characteristics, and the evidence they afford as to the earliest form—the _Urtext_—of the version, and (ii) the Greek text presupposed by the version in the form in which we have it.

It will be convenient to divide the New Testament into three component parts, (i) the Gospels, (ii) the Acts and Epistles, or the _Apostol_ as it is called in Slavonic, (iii) the Apocalypse. There can be little doubt that the Gospels were the earliest part to be translated or that this translation was made for liturgical purposes. This last point explains the great preponderance of MSS. of the version in which the Gospels are arranged in the form of a lectionary(130).

Amongst the earliest manuscripts of the Gospels are the Codex Zographensis, Codex Marianus, and the Codex Assemanicus. The two first Jagić ascribes to the tenth or eleventh century. All these are written for the most part in the Glagolitic character. Besides these, mention must be made of the Ostromir Codex, written in Cyrillic characters, by Gregory, a deacon at Novgorod, and dating from the year 1056-7. In considering the distinctive characteristics of these manuscripts of the version, the first point to notice is that they each preserve certain dialectical forms and expressions by which their place of origin and to some extent their date can be determined. Thus Miklosich regards the Codex Zographensis and Codex Assemanicus as preserving Bulgarico-Slovenish forms, the Ostromir Codex as representative of Russo-Slovenish, and so on. It is mainly in these particulars that the manuscripts differ, though there are also other differences by means of which it has been determined that some Codices, especially those in the Glagolitic character, preserve the version in a more original form than others, as for example the Ostromir Codex. These differences consist(131), (i) in orthography, (ii) in the fact that the later forms of the version translate Greek words left untranslated in the older forms, (iii) in the substitution of later and easier words for archaisms. It may also be noted that alterations are more numerous, as might be expected, in copies of the Gospels made for liturgical purposes than in other copies.

The same remarks would be true of the second part of the Bible, the _Apostol_. This is pointed out by Voskresenski in the book to which reference has been made, but which is known to the writer of these lines only from a review. A very careful examination of the text of the “Apostol,” based on the manuscripts of the Synodal Library, is made by Gorski and Nevostruiev in the work referred to above, pp. 292 ff.

Oblak has examined the Slavonic version of the Apocalypse, of which the manuscripts are fewer and later. The earliest manuscript is ascribed to the thirteenth century, but the textual corruption which it exhibits in comparison with other manuscripts requires that the version which it embodies should be referred at least to the twelfth century. We do indeed find a quotation of the Apocalypse (ix. 14) as early as the Isbornik of Sviatoslav of the year 1073, but in a form so different from the MSS. of the version now extant, that we must regard it as a quotation from memory. The MSS. have many small variations, sometimes merely dialectical, sometimes based on a different Greek text. They also show marks in places of having been corrected with the help of the Latin. But in spite of all their variations Oblak believes that all the manuscripts are to be referred to one common translation made from a Greek text of the Constantinopolitan type, which has been here and there corrupted by Western influence.

It may be noted in conclusion that the earliest dated complete manuscript of the Gospels is dated 1144, the earliest manuscript of the whole Bible, A.D. 1499, and that the earliest printed edition is the famous Ostrog Bible of 1581.

It remains to say something of the Greek text underlying the Slavonic version, for this is the special point of view from which the versions are being here considered. The instances will all be taken from the Gospels, though others might have been added from those collected by Gorski. In the first place it is necessary to draw attention to the fact that for critical purposes a modern edition of the version will be found insufficient. The following are cases(132) where the edition published by the British and Foreign Bible Society, probably based on the Textus Receptus, is misleading as to the real original reading of the version. In St. Matt. xi. 2 Codd. Assem., Zograph., Ostrom., all imply the reading διά, the modern edition δύο: in St. John i. 28 the MSS. have Bethany, the edition Bethabara; in St. John vii. 39 the MSS. insert, the edition omits, δεδομένον; in St. Matt. xxv. 2 the MSS. put μωραί before φρόνιμοι, the edition inverts the order. The Ostromir Codex presents a later form of the version, and so we find instances where the other two MSS., just referred to, preserve what is probably a better reading. Thus in St. Luke ii. 3 they have οἱ γονεῖς αὐτοῦ, the Ostromir Ἰωσὴφ καὶ ἡ μήτηρ αὐτοῦ; in St. John ix. 8 they have προσαίτης, it has τυφλός; in St. John xix. 14 they have τρίτη, it reads ἕκτη; in St. John xxi. 15 they have ἀρνία, it has πρόβατα. Again there are cases where one MS. of the version stands alone. Thus Codex Zogr. stands alone, as against Assem. and Ostrom., in omitting St. Luke xiv. 24, and inserting δευτεροπρώτῳ in St. Luke vi. 1. Again in the choice of Slavonic words for the same Greek original, Cod. Zogr. will agree with Codex Assem. against Codex Ostrom., though where the Codex Assemanicus is freer in its rendering, the Ostromir Codex and Codex Zographensis agree. Sometimes again the Codex Zographensis is alone in curious readings which seem to be conflations of the texts found in the other two manuscripts, or based on a conflate Greek text.

This version and the various manuscripts which contain it have received most attention from Slavonic philologists engaged in examining the earliest monuments of their language; but the readings which have been given will be enough to show that it does not deserve to be dismissed, as summarily as has been sometimes the case, from the number of those versions which have a value for purposes of the Textual Criticism of the New Testament.

(6) The Arabic Version (Arab.).

Arabic versions (Arab.) are many, though of the slightest possible critical importance; their literary history, therefore, need not be traced with much minuteness. A notice is quoted from Bar-hebraeus (Assemani, Bibl. Or., ii. 335) to the effect that John, Patriarch of the Monophysites from 631-640, translated the “Gospel” from Syriac into Arabic; and some scholars have believed in the existence of a pre-Mohammedan version of parts at least of the New Testament on other grounds; from such a version (written in the “Hebrew” character) in the opinion of Sprenger (Das Leben und die Lehre Muhammads, i. 131) come the verses of St. John’s Gospel (xv. 23-27, xvi. 1), cited by Ibn Ishaq (ob. 768) in his “Life of Mohammed” (ed. Wüstenfeld, i. 150)(133). These verses are evidently translated from the (Jerusalem?) Syriac; but the translation of the Gospel, from the Syriac into Arabic, existing in a Leipzig MS. brought by Tischendorf from the East and described at length by Gildemeister (De evangeliis in Arabicum e simplici Syriaco translatis, Bonn, 1865) is shown by internal evidence to be posterior to Islam (pp. 30 sq.). The Arabic versions of the Gospel existing in MS. are divided by Guidi (Atti della R. Academia dei Lincei, classe di scienze morali &c., 1888, 1-30) into five sorts: (1) those made directly from the Greek; (2) made directly or corrected from the Peshitto; (3) made directly or corrected from the Coptic; (4) MSS. of two distinct eclectic recensions made in the Alexandrian Patriarchate in the thirteenth century; (5) MSS. (chiefly derived from the Syriac) which are distinguished by their style; being in rhymed prose or elegant Arabic. MSS. of the first sort can all, he says, be traced to the convent of St. Saba near Jerusalem, and are preceded by the lives of its founders, St. Eutimius and St. Saba; the version they contain is to be ascribed to the time of the Caliph Mamun (ninth century). Of the MSS. of class 4, one set represents a recension made by Ibn El-Assāl, circ. 1250; while another represents a less elaborate recension made shortly afterwards, in which the passages omitted in the other were restored, while marginal notes recorded their omission in other versions. Versions of the fifth class were made in the tenth, fourteenth, and seventeenth centuries. A list of MSS. containing the different recensions of all these classes is given by Guidi, _l. c._, pp. 30-33.

The printed texts all represent varieties of the second eclectic recension of class 4, of which five editions are enumerated by Gildemeister(_l. c._, pp. 42, 3, and iv). 1. Roman edition of the Gospels from the Medicean Press, 1591 (ar.r), edited by J. Baptista Raymundi, some copies having a Latin translation by Antonius Sionita. The MS. on which this edition was based is unknown. 2. Edition of Thomas Erpenius (1584-1624), Leyden, 1616, containing the whole New Testament (ar.e). This edition was based on the Leyden MS., Scaliger 217, written in Egypt in the year of the Martyrs 1059 (A.D. 1342-3); two other manuscripts also employed by Erpenius for the Gospels are now in the Cambridge University Library (G. 5. 33, and G. 5. 27, written A.D. 1285). A third MS. employed for this edition was in the Carshunic character. The Acts and Pauline Epistles, the Epistles of St. James, St. Peter 1 and St. John 1 in this edition are translated from the Peshitto; the remaining Catholic Epistles and the Apocalypse are from some other source; the latter shows some remarkable agreement with the Memphitic (Hug, Einleitung in das N. T., pp. 433-5). 3. Edition of the whole N. T. in the Paris Polyglott (ar.p), 1645, reprinted with little alteration in the London Polyglott (1657). Gildemeister, _l. c._, proves against Lagarde (_l. c._, xi) that this recension in the Gospels is not an interpolated reprint of the Roman edition, but is based on a MS. similar to Paris Anc. f. 27 (of A.D. 1619) and Coisl. 239 (new Suppl. Ar. 27) described by Scholz, “Bibl. Krit. Reise,” pp. 56, 58. The Acts, Epistles, and Apocalypse follow the Greek, but are by another translator. 4. Edition of the whole N. T. in the Carshunic character (Rome, 1703), edited by Faustus Naironus, for the use of the Maronites, from a MS. brought from Cyprus, reprinted Paris, 1827; the Acts, Epistles, and Apocalypse represent the same version as that of Erpenius, but in a different recension. 5. Edition of the four Gospels from a Vienna MS. (previously described by S. C. Storr, Dissertatio inauguralis critica de evangeliis Arabicis, Tübingen, 1875, p. 17 sq.), by P. de Lagarde (Die vier Evangelien Arabisch, Leipzig, 1864). The MS. contains various readings from the Coptic, Syriac, and Latin (according to Lagarde, Gildemeister more naturally renders rūmī by Greek). The editor has prefixed a table of variants between his text and that of Erpenius, but regards the relation of the former to the original as involving questions too complicated for immediate discussion (p. xxxi).

Extracts from MSS. of Arabic versions in French and Italian libraries are given by J. M. A. Scholz, Biblisch-Kritische Reise, Leipzig and Sorau, 1823; a description of several others, some of great antiquity, is to be found in Tischendorf’s “Anecdota Sacra et Profana,” pp. 70-73 (2nd ed.); and Professor Rendel Harris, in “Biblical Fragments from Sinai” (Cambridge, 1890) has published a facsimile of a fragment of an Arabic version from a bilingual MS. of the ninth century; the version whence it is derived agrees with none of those that have been published, and was probably older than any of them.

The repeated revision and correction which these translations have undergone (Gildemeister, _l. c._, 1-3), while they give evidence of the industry and zeal of the Arabic-speaking Christians, have made scholars despair of employing them for critical purposes; “they rather serve,” says Gildemeister, “to illustrate the history of biblical and Christian studies.”

(7) The Anglo-Saxon Version (Sax.).

There is but one known version of the four Gospels (the only portion of the N. T. that was translated into A.-S.); this version was made, probably in the South-West of England at or near Bath, in the last quarter of the tenth century. It is preserved in four MSS.: (Corp.) Corpus Christi Coll. Camb. MS. 140; (B) Bodleian Lib. MS. 441; (C) Cotton MS. Otho C. I (seriously injured by fire), and (A) Camb. Univ. Lib. MS. Ii. 2. 11. Of these the first three may be dated, in round number, about the year 1000; the fourth (A) belongs to the following half-century. The Bodl. Lib. has also recently acquired a fragment of four leaves of St. John’s Gospel, which agrees closely with A. [Published by Napier in “Archiv f. n. Sprachen,” vol. lxxxvii. p. 255 f.]

It may also be mentioned that there are in the Brit. Mus. two additional copies of this version (Bibl. Reg. MS. I. A. xiv, and Hatton MS. 38). These belong to a period after the Conquest and have no critical value, for the first is copied from B, and the second is copied from the first.

This version is based upon a type of the Vulgate MSS. that has not yet been definitely determined. Old Latin readings make it certain that the original MS. was of the mixed type.

Next in importance to this version are the two following Latin MSS. of the four Gospels, with an interlinear Anglo-Saxon gloss. (1) MS. Nero D. 4 (the Lindisfarne MS., also known as the Durham Book). The Latin was written by Eadfrith, bishop of Lindisfarne 698-721; the interlinear gloss being about two and a half centuries later, made near Durham about the year 950. (2) The Rushworth MS. (Bodl. Lib. Auct. D. ii. 19). The Latin was written by the scribe Macregol, probably in the eighth century. The gloss, by the scribes Farman and Owun, is referred to the latter half of the tenth century. These two Latin texts differ but slightly; they are also of the Vulgate types.

All the MSS. that have now been mentioned are published in one volume (of four parts) by Professor W. W. Skeat: “The Holy Gospels in Anglo-Saxon, Northumbrian, and Old Mercian Versions, synoptically arranged, with collations exhibiting all the readings of all the MSS.; together with the Early Latin Version as contained in the Lindisfarne MS.; collated with the Latin Version in the Rushworth MS. Cambridge: University Press, 1871-1887.” Dr. James W. Bright has published an edition of St. Luke’s Gospel of the A.-S. Version, Oxford, 1892, and has in preparation a critical edition of the entire Version [which has been published recently]. The earlier editions of the Anglo-Saxon Gospels are by Archbishop Parker, 1571; Dr. Marshall (rector of Lincoln College), 1665; Benjamin Thorpe, 1842; Dr. Joseph Bosworth, 1865.

(8) The Frankish Version (Fr.).

A Frankish version of St. Matthew, from a manuscript of the ninth century at St. Gall, in the Frankish dialect of the Teutonic, was published by J. A. Schmeller in 1827. Tischendorf (N. T., Proleg., p. 225) thinks it worthy of examination, but does not state whether it was translated from the Greek or Latin: the latter supposition is the more probable.

(9) Persic Versions (Pers.).

Persic versions of the Gospels only, in print, are two: (1) one in Walton’s Polyglott (pers.p) with a Latin version by Samuel Clarke (which C. A. Bode thought it worth his while to reconstruct, Helmstedt, 1750-51, with a learned Preface), obviously made from the Peshitto Syriac, which the Persians had long used (“yet often so paraphrastic as to claim a character of its own,” Malan, _ubi supra_, p. xi), “interprete Symone F. Joseph Taurinensi,” and taken from a single manuscript belonging to E. Pocock(134), _probably_ dated A.D. 1341. This version may prove of some use in restoring the text of the Peshitto. (2) The second, though apparently modern [xiv?] was made from the Greek (pers.w). Its publication was commenced in 1652 by Abraham Wheelocke, Professor of Arabic and Anglo-Saxon and University Librarian at Cambridge, at the expense of Sir Th. Adams, the generous and loyal alderman of London. The basis (as appears from the volume itself) was an Oxford codex (probably Laud. A. 96 of the old notation), which Wheelocke, in his elaborate notes at the end of each chapter, compared with Pocock’s and with a third manuscript at Cambridge (Gg. v. 26), dated 1014 of the Hegira (A.D. 1607). On Wheelocke’s death in 1653 only 108 pages (to Matt, xviii. 6) were printed, but his whole text and Latin version being found ready for the press, the book was published with a second title-page, dated London, 1657, and a short Preface by an anonymous editor (said to be one Pierson), who in lieu of Wheelocke’s notes, which break off after Matt. xvii., appended a simple collation of the Pocock manuscript from that place. The Persians have older versions, parts of both Testaments, still unpublished. There is another copy of the Persian Gospels at Cambridge, which once belonged to Archbishop Bancroft, and was brought from Lambeth in 1646, but was not restored in 1662 with the other books belonging to the Lambeth Library.