THE TEXT OF THE GREEK BIBLE - A STUDENT'S HANDBOOK by Frederick G Kenyon, late Director & Principal Librarian of the British Museum. First published 1937 New edition 1949 Reprinted 1953. (Prepared for katapi by Paul Ingram 2015).
HOME | contents | << | introduction | syriac | armenian | georgian | coptic | latin | gothic | the fathers | bibliography | >> |


Hardly less important than the actual manuscripts of the Greek New Testament is the evidence of the early translations of it into other languages. Their value depends, first on the earliness of their date, and next on the certainty with which we can ascertain the Greek text from which the translation was made. In the case of the New Testament there are two translations which we know to have been made before the end of the second century, and ethers somewhat later. If we can determine their text with certainty, we shall have evidence earlier by some generations than the earliest extant MSS. in the original Greek. More than that, we shall have evidence showing what sort of text was current at that date in a particular part of the Christian world.
As Christianity spread outwards from Palestine on its world mission, it addressed itself primarily to the Greek-speaking communities of the eastern part of the Roman Empire, as far as Rome itself, where Greek was almost as much a language in common use among the mixed population of the capital as Latin. But also it was passing into and through countries where a considerable part of the population did not speak Greek, and before long a demand grew up for versions of the Christian books in their vernaculars. Of these, three were earliest in urgency and of chief importance: Syriac, for the countries spreading from Antioch to the Euphrates valley; Coptic, for Egypt, which bordered Palestine on the south-west, and where a strong Hellenistic colony was surrounded by a native Egyptian population; and Latin, not only or even primarily for Italy, where much of the population spoke Greek, but certainly for the flourishing provinces of Roman Africa and Roman Gaul. We will take the Eastern versions first.


| syriac | diatessaron | old syriac | peshitta | philoxenian | palestinian |

Syriac is an Aramaic dialect, akin to, but not identical with, that which was in common use in Palestine in the time of our Lord. Aramaic was,the language of the mass of the population of Mesopotamia, whence it spread, with dialectical varieties, over northern Syria and Palestine, where for all except scholars it entirely superseded Hebrew. Palestinian Aramaic was no doubt the language habitually spoken by our Lord; and this gives a special interest to the Syriac Gospels, as coming nearest to the form in which His teaching was originally delivered. There is, however, no direct connection between that original teaching and the Syriac Gospels. The headquarters of Syriac literature and Syriac Christianity was Edessa, an independent principality east of the Euphrates in the great bend of that river; and to this province Christianity must have come (presumably by way of Antioch) in Greek, and there is no doubt that the Syrian Scriptures were translated from Greek. What is doubtful is the exact form in which they first circulated.

(a) The Diatessaron. The most certain fact that we know about the early dissemination of the Christian Scriptures in Syria is that about A.D. 170-180 Tatian, himself a native of the Euphrates valley, but a disciple
the versions and fathers of Justin Martyr at Rome about the middle of the century, made a harmony of the Gospels which circulated extensively in a Syriac form among the churches of Syria. Tatian, having travelled much and having finally been converted to Christianity, appears to have spent many years at Rome as a disciple of Justin; but after the latter's martyrdom (c. A.D. 165) his extremely austere and ascetic views brought him into suspicion of heresy (of the Encratite type), and about 172 he left Rome and returned to his native land, where he died about 180. Whether he had prepared his Harmony of the Gospels before leaving Rome, or did so after his return to Syria, is unknown. In the former case it would no doubt have been made in Greek, in the latter probably in Syriac; and this is one of the questions in dispute among scholars. In favour of the former is its Greek title and the fact that Latin and eventually Dutch translations were made of it; also the fact that it never fell under the suspicion of heresy, and shows no signs of Encratite prepossessions; and that its text, so far as ascertainable, shows strong affinities with D and the Old Latin; of the latter, that its main circulation was unquestionably in Syriac.

The history of the Diatessaron is curious. Until the fifth century, it seems to have been the form in which the Gospel, narrative was principally known in Syria. Theodoret, bishop of Cyrrhus on the Euphrates, writing in 453, says that he found more than two hundred copies of it in reverential use in his district, which he removed, substituting for them copies of the four Gospels, no doubt in the then established Peshitta version. Thereafter it fell into complete oblivion. Victor of Capua, in the early sixth century, found an anonymous Latin Harmony (the existence of which in the West is an argument in favour of a Greek original) which he thought must be that recorded to have been made by Tatian, and edited, substituting a Vulgate text for that which he found; and this is still extant in the Codex Fuldensis (see below, p. 146), written in 541-6. There is also a Dutch Harmony of the Middle Ages, discovered at Liege by Dr. D. Plooij in 1923, which has been shown to have been made from a Latin Harmony closely akin to that in the Fuldensis, but with Old Latin readings instead of Vulgate, and therefore much nearer to the original Diatessaron. ← Edited by D. Plooij and C. A. Phillips, The Liége Diatessaron (Amsterdam, 1929, etc.). There were also other Gospel Harmonies in use in mediaeval times, which may or may not be descended from Tatian. But in the middle of the 19th century all knowledge of the work was so completely lost that the anonymous author of Supernatural Religion (1876), in his attack on the authenticity and early date of the Gospels, could maintain that there was no proof that the Diatessaron was a harmony of the four canonical Gospels, and could even question its existence; while so learned a scholar as Bishop Lightfoot could only answer him with indirect arguments. Yet so long before as 1836 the Mechitarist Fathers at Venice had actually published a commentary on the Diatessaron written by St. Ephraim (d. 373), which conclusively established both the existence and character of the work; but since it was in Armenian, it remained completely unknown, and even when a Latin translation of it was published by Moesinger in 1876, this also remained unnoticed until Dr. Ezra Abbot called attention to it in 1880. This stimulated research, which led to the discovery of two MSS. (one in the Vatican and one in Egypt, whence it was removed to Rome) containing Arabic texts of the Diatessaron itself, which was accordingly edited from them by Ciasca in 1888.

There is therefore now complete evidence of the existence and general character of the Diatessaron; but since we possess it only in late copies of translations, Latin, Armenian, Arabic and Dutch, which have been subject to the universal tendency towards conforming strange texts to that generally received, we are far from certainty as to its original wording, and there is often divergent testimony as to its arrangement. One small additional piece of evidence has come to light quite recently. In 1920 some English officers discovered the remains of a Roman fort at Dura on the Euphrates, once the eastern frontier of the Roman Empire, wherein were some remarkable wall-paintings of the first century and later. Subsequently, when the site had fallen into the French mandated area, excavations were carried out by French and American investigators; and among the debris of houses (including a Christian church and a Jewish synagogue) which had been destroyed by the Romans to strengthen the fortifications just before the final capture of the place by the Persians in 256 a number of vellum and papyrus fragments were found which had been sealed up by these operations. Among these, when they were examined at Yale in 1933, was found a small vellum fragment of the Diatessaron itself, in Greek, in a hand of the first half of the 3rd century (and necessarily before 256). It consists only of fourteen imperfect lines, and contains the narrative of the petition of Joseph of Arimathsea for the body of our Lord, in a mosaic made up of phrases from all four Gospels, with some editorial adjustments. ← Published by C. H. Kraeling (A Greek Fragment of Tatian's Diatessaron from Dura, in Studies and Documents edited by K. and S. Lake, No. Ill, London, 1935). Thus in the first eight words of the fragment, Ζεβεδαίου καὶ Σαλώμη καὶ γυναῖκες τῶν συνακολουθησάντων αὐτῷ, Ζεβεδαίου is taken from Matthew, καὶ Σαλώμη from Mark, καὶ γυναῖκες and αὐτῷ from Luke, while τῶν συνακολουθησάντων replaces the αἱ συνακολουθοῦσαι or -θήσασαι of all our MSS. of Luke. Further down the fragment has ὄνομα Ἰωσήφ (Matthew), ἄγαθος δίκαιος (Luke), ὢν μαθητὴς τοῦ Ἰησοῦ (John). These examples are sufficient to show that Tatian handled his materials with freedom, and that even if we had the original Diatessaron intact, great caution would have to be used to determine the Gospel texts on which he worked.

The Dura fragment has been claimed as a proof that Greek was the original language of the Diatessaron. It is certainly remarkable to find a Greek copy of it in the farthest corner of Syria; but it is not decisive, since Dura was a military and commercial post, where there must have been many dwellers who were not Syrians by nationality or language. The manuscripts with which the fragment was found are military documents in Latin or business documents in Greek (with one Syriac). The proof of its Greek origin must still rest on the other considerations indicated above.

The Diatessaron therefore still presents several unsolved problems; its original language, its text, its influence on the mediasval Latin harmonies that have come down to us, still more its influence on the text-tradition as a whole. Von Soden is disposed to attribute to it much of the confusion of the earliest texts, not only in Syriac but in Greek. This view has not generally been accepted; but it is evidently a line of thought which requires investigation, and it is much to be hoped that future discoveries may throw some light upon it. ← The principal English works on the Diatessaron are S. Hemphill, The Diatessaron of Tatian (1880), J. H. Hill, The Earliest Life of Christ ... being the Diatessaron of Tatian (1893), and F. C. Burkitt's Introduction to his edition of the Evangelion da- Mepharreshe (1904).

(b) Old Syriac. It is clear therefore that from the last quarter of the second century a Gospel narrative, made up of interwoven selections from the four canonical Gospels, was in circulation in Syria. The questions remain, Were the four Gospels also known by that time, as separate entities, and were the other books of the New Testament extant in Syriac ? Until about the middle of the last century, no Syriac translation of the New Testament was known earlier than the Peshitta, which was then variously assigned to the fourth, third, or second century. But in 1842, among a large number of Syriac MSS. brought by Archdeacon Tattam and others from the monastery of St. Mary Deipara in the Nitrian desert in Egypt, and acquired by the British Museum, W. Cureton found some eighty leaves of a version evidently different from, and in his view older than, the Peshitta. This MS. was printed and privately circulated in 1848, and definitely published in 1858. Then in 1892 two Cambridge ladies, Mrs. Lewis and Mrs. Gibson, brought back photographs of a palimpsest in the monastery of St. Catherine at Mount Sinai, which proved to contain the same version as the Cureton MS., though with important variants and apparently of earlier date. ← Published by R. L. Bensly, J. Rendel Harris and F. C. Burkitt (The Four Gospels in Syriac transcribed from the Sinaitic Palimpsest, Cambridge, 1894). The standard edition of the O.S. Version, giving the readings of both Cur. and Sin., is Burkitt's Evangelion da-Mepharreshe, Cambridge, 1904, with a very valuable Introduction. These two MSS. constitute for us the Old Syriac version.

Neither MS. is perfect. Of the Sinaitic 142 leaves survive out of 166, containing Matthew i.1—vi.10, viii.3—xvi.15, xvii.11—xx.24, xxi.20—xxviii.7; Mark i.12-44, ii.21—iv.17, iv.41—v.26, vi.5—xvi.8 (omitting the last 12 verses); Luke i.1-16, i.38—v.28, vi.12—xxiv.53; John i.25-47, 16—iv.37, v.6-25, 46—xviii.31, xix.40—xxi.25. Of the Curetonian there are 86 leaves out of 180, containing Matthew i.1—viii.22, x.32—xxiii.25; Mark xvi.17-20; Luke ii.48—iii.16, vii.33—xvi.12, xvii.1— xxiv.44; John i.1-42, iii.5—viii.19, xiv.10-12, 15-19, 21-23, 26-29. The Sinaitic is considered to be of the late 4th century, the Curetonian of the 5th.

With regard to the character of the text of the version, it will be convenient first to quote some of the more noteworthy readings. In Matthew i.16, Sin. has the remarkable reading Ἰωσὴφ δέ, ᾧ ἐμνηστεύθη παρθένος Μαριάμ, εγέννησεν Ἰησοῦν τὸν λεγόμενον Χριστόν, which in the Curetonian appears as (τὸμ Ἰψσήφ), ᾧ μνηστευθεῖσα παρθένος Μαριάμ εγέννησεν Ἰησοῦν τὸν λεγόμενον Χριστόν. Both MSS. have the shorter version of v.44 (with א B, but against D). Cur. gives the doxology to the Lord's Prayer (except ἡ δύναμις καὶ) in vi.13, against א B D; Sin. is defective here. Both omit xii.47 (with K B against D), and read σταδίους πολλοὺς ἀπὸ τῆς γῆς ἀπεῖχε in xiv.24 (with B and fam. 13). They omit xvi.2, 3 with א B and fam. 13, against D; and similarly xvii.21. In xviii.11 their testimony is divided. In xix.17 they agree with B D and fam. 1 in reading τί με ἐρωτᾶς περὶ τοῦ ἀγαθοῦ, but Cur. adds ὁ θεός. In xx.28 Cur. has the additional passage found also in D and Θ (see p. 90); Sin. is defective here. Sin. agrees with א B Din omitting xxiii.14, but Cur. has it. In xxvii.16 Sin. has the reading Ἰσοῦν Βαραββᾶν, found also in fam. 1; Cur. is defective. In Mark Cur. is wholly missing, except a fragment containing xvi.17-20, just enough to show that it had the disputed ending, which Sin. agrees with א B in omitting. Elsewhere in this Gospel Sin. omits ix.44, 46, 49 with א B, against D, but in x.24 has τοὺς πεποιθότας ἐπὶ χρήμασιν with D against א B, and in xiii.33 καὶ προσεύχεσθε with א against B D. In xiv.24 it has καινπῆς against א B D, and in xiv.65 ἔβαλλον with the Old Latin and Sahidic. In Luke ii.14 it has εὐδοκία against all the principal uncials. It is deficient in vi.4 where D has a remarkable addition (see p. 91 above). It omits the last words of vi.48. In ix.35 Sin. supports א B in reading ἐκλελεγμένος, while Cur. follows A D and the bulk of the MSS. in reading ἀγαπητός, as in the parallel passages in Matthew and Mark. Similarly in ix.55-6 Sin. agrees with א B in omitting the words εἶπεν ... σῶσαι while Cur. agrees with D and the Old Latin in retaining them. In x.42 Sin. omits ἑνὸς ἐστιν χρεία with D, while Cur. has them, with A C and most MSS. In the Lord's Prayer (xi.2-4) Sin. gives the shorter version throughout, with א B, while Cur. omits only " Thy will be done, as in heaven, so in earth." In xiv.5, where א has ὄνος ἢ βοῦς, A B υἱὸς ἢ βοῦς, and D πρόβατον ἤ βοῦς, Sin. has βοῦς ἤ ὄνος, and Cur. υἱος ἢ βοῦς ἢ ὄνος. In the institution of the Lord's Supper (xxii.19, 20) Sin. gives the verses in the order 19, 20a, 17, 20b, 18, while Cur. has them in the order 19, 17, 18, omitting 20; which seem to be different attempts to avoid the double mention of the Cup found in א B and most other MSS. In xxii.43, 44, Sin. omits, while Cur. retains, the Vision of the Angel and the Bloody Sweat. Similarly in xxiii.34 Sin. omits, while Cur. retains, the Word from the Cross, " Father, forgive them; " but in 38 both omit " in letters of Greek and Roman and Hebrew." In xxiv, however, both side wholly with D and the Old Latin in their omissions in vv.6, 12, 36, 40, 51, and 52; in 42 Sin. omits, but Cur. retains. In John i.18 Cur. reads ὁ μονογενὴς υὁός, against the μονογενὴς Θεὸς of א B. In iv.9 both have the reading οὐ γὰρ συγχρῶται Ἰουδαῖοι Σαμαρείταις, with A B, which א D omit. In iv.42 Cur. omits ὁ Χριστὸς at the end of the verse, with אB, against A D. In v.3, 4 both omit the stirring of the water (ἐκδεχομένωνv ... νοσήματι). Both omit the pericope adulterae (with א A B, but against D). In ix.35 Sin. supports א B in reading ἀνθρώπου, not Θεοῦ; Cur. is defective. In xi.39 Sin. (Cur. is defective) has a quite singular addition after Κὐριε, " Why are they taking away the stone? " In xviii by an alteration in the order of the verses in Sin. (13, 24, 14, 15, 19-23, 16-18, 25-27) Caiaphas is made to appear as the questioner instead of Annas, and the three denials by Peter are brought together.

A further idea of the character of the version may be obtained from an analysis of 158 passages in the Gospels where there are specially notable variations of text. ← These are the passages given in the Delectus Lectionum notatu dignissimorum in Sanday's exceedingly useful Appendices ad N. T. Stephanicum (Oxford, 1889). The readings of the Sinaitic Syriac, which was not extant in 1889, are added from Souter's Oxford edition. The figures for the Old Latin version must be taken with some reserve, since the MSS. of this version often differ. In 43 passages where there is a clear-cut divergence between א B and D, the Old Syriac sides with the former in 22 cases, and with the latter in 21. Its agreements and disagreements with א, B, D, and the Old Latin are as follows:

With Against
א 47 (Mt. 18, Mk. 6, Lk. 9, Jn. 14)
B 49 (21, 5, 10, 13)
D 48 (19, 3, 11, 15)
OL 68 (30, 6, 14, 18)
76 (Mt. 25, Mk. 11, Lk. 19, Jn. 21)
74 (22, 12, 18, 22)
56 (16, 14, 15, 11)
46 (14, 8, 13, 11)

In 21 passages (not included in the above figures) the testimony of Sin. and Cur. is divided; and this throws some measure of doubt on those other passages where only one of the two MSS. is extant. In general, however, it is evident that, while the version cannot be reckoned wholly with either the א B group or the D group, it shows a preponderance of agreement with the latter.

This conclusion, however, requires further analysis. The significance of the Old Syriac version depends on the answers to the following questions: (1) what is the relation, in age and respective influence, between OS. and the Diatessaron, (2) what is the relation between OS and א B, (3) what is the relation between the two MSS. of OS, Sin. and Cur. ? The conclusions which Burkitt has made most probable (though differences of opinion still exist) are as follows. The Diatessaron was not made from OS, as appears from important differences in text and rendering; e.g., it included the incident of the light at the Baptism (Matthew iii, 16), which is found in two MSS. of the Old Latin (a and g), also Matthew xvi.2, 3, and John v.3, 4, all of which are omitted by both Sin. and Cur., and the last 12 verses of Mark, Luke xxii.43, 44 and xxiii,34, which are omitted by Sin. but included by Cur. It is probable that the Old Syriac version was made about A.D. 200, therefore after the Diatessaron, and that the Western readings in it are largely due to the influence of the latter. In basis, its text was that of the Greek text current in Antioch at that date, which, though often agreeing with the Alexandrian text of א B, represents an independent line of tradition. As between the two MSS. Sin. generally represents the earlier form, Cur. having been to some extent revised from later Greek MSS.

The importance of the Old Syriac, as showing the form of text current in the Church of Edessa at the beginning of the third century, is obvious: but if Burkitt's conclusions are sound it cannot be used as an argument that the Western (D and OL.) type of text prevailed in the East as well as the West, since its Western characteristics are largely due to the influence of the Diatessaron, which was originally made in Rome from MSS. of the Western type.

No MS. of an Old Syriac version of any other part of the New Testament is extant, but there is reason to suppose that the other books were also translated; for Armenian translations of commentaries by St. Eph-raim on the Acts and Pauline Epistles are extant, which imply the use of a Syriac text earlier than the Peshitta.

(c) The Peshitta. This was the version which eventually became the official Bible of the Syriac Church, just as the Vulgate did of the Roman. Before the discovery of the Old Syriac, it was generally assigned to the third, or even to the second, century. But it was shown by Burkitt that the belief that it was used by St. Ephraim was unfounded, and that in fact there is no evidence of its existence before the fifth century. Moreover, he pointed out that Rabbula, bishop of Edessa from 411 to 435, is recorded by his biographer to have translated the New Testament from Greek into Syriac, and that it was consequently not only legitimate but inevitable to identify his translation with the Peshitta, when once the latter was shown to belong to this period. Materially later the Peshitta could not be, since there are extant MSS. of it which are assigned to the latter part of the fifth century; and it would be sheer perversity to maintain that Rabbula's version has perished without trace, while an anonymous production of the same date had captured the Church of Edessa. As the work of this great bishop, on the other hand, its immediate success is quite intelligible. The Diatessaron and Old Syriac dropped out of existence, and the Peshitta became the accepted Syriac Bible. Since, moreover, it was used by both the Monophysite and the Nestorian Churches, .it must have been generally accepted before 431, when the Nestorian secession took place.

Something like 250 copies of the Peshitta are known, of which more than 100 are in the British Museum. Two of these are assigned to the fifth century, and four bear actual dates in the sixth. In character, it shows the influence of the revision which, beginning perhaps at Antioch, ultimately developed into the received Byzantine text. Since this appears in the writings of Chrysostom, who was born and worked at Antioch until 398, it was easily available for Rabbula at the beginning of the fifth century, and the Peshitta shows many readings of this type. On the other hand, it retains traces of the earlier types which are found in the Old Syriac. In the readings analysed above, the Peshitta is found agreeing with the Old Syriac in 54 instances and disagreeing in 65. Where the two Old Syriac witnesses differ from one another, it is almost always Cur. that agrees with the Peshitta, which may be used as an argument that the Curetonian, as compared with the Sinaitic, has been somewhat revised in the direction of the later types of text.

The Peshitta in its original form did not include the Apocalypse or the four minor Catholic Epistles (2 Peter, 2 and 3 John, Jude), which were not accepted as canonical by the Syriac Church. It was first edited at Vienna in 1555 by C. Widmanstadt. The standard edition is that of G. H. Gwilliam (completing the work begun by Philip Pusey), which was published in 1902. Gwilliam used some forty MSS., but he states that they differ little. Syriac scribes, like Hebrew, were very careful copyists.

(d) The Philoxenian and Harkleian Syriac. About a century later than the Peshitta, in 508, a fresh translation was made to the order of Philoxenus, Jacobite bishop of Hierapolis in eastern Syria, by Polycarp, a χωρεπίσκοπος. It was from this version that the four minor Catholic Epistles were added to the Syriac Bible; and a copy of the Apocalypse in this version was discovered by Dr. Gwynn of Dublin in a twelfth-century MS. now in the John Rylands Library at Manchester. From these books it is clear that the Philoxenian version was written in good idiomatic Syriac, and that its text was of the received Byzantine type. For the rest of the New Testament its character is obscured by the fact that in 616 it was drastically revised by Thomas of Harkel from Greek MSS. at Alexandria. Thomas' method was exactly the reverse of that" of Philoxenus, for he was literal without respect for idiom, and the MSS. used by him were of the Codex Bezae type. Other readings of the same type, added in the margin, seem to represent a second stage in this revision.

About fifty MSS. of this version have come down to us, all representing it under the form of the Harkleian revision. The oldest (at Rome) is assigned to the 7th century. The Harkleian Apocalypse was edited by de Dieu in 1627, and the four Philoxenian Epistles by Pococke in 1630, but otherwise the version was unknown until the whole was edited by J. White in 1778-1803, and this remains the only available text.

(e) The Palestinian Syriac. Finally, considerable portions remain of another version in the Palestinian dialect, which Burkitt has shown to have originated at Antioch in the 6th century, probably as a result of the efforts of Justinian and Heraclius to eradicate Judaism in Palestine, and to have become popular on the occasion of a revival of Palestinian Christianity in the nth century. Nearly all the surviving MSS. are in the form of lectionaries, the two most important being a pair of Gospel lectionaries, dated 1104 and 1118, which were edited by Mrs. Lewis and Mrs. Gibson in 1899. Some portions of Acts and Epistles have survived in other fragmentary copies. Its text is mixed in character, but contains early elements, both of the א B and the D type.

Two other Oriental versions deserve brief mention on account of their connection with the Syriac.


It is recorded by Armenian writers of the fifth century that the Scriptures were translated into Armenian by SS. Mesrop and Sahak, apparently about 400; but that after the Council of Ephesus (A.D. 431), at which Nestorianism was condemned, they received " correct " copies of the Greek Bible from Constantinople, and revised their translation accordingly. The original version was presumably made from Syriac, since ifshows a strong affinity to the Old Syriac in the Gospels, and to the text used by St. Ephraim in the Pauline Epistles; but since a Greek scribe (Hrofanos = Rufinus) is said to have assisted, some use may also have been made of Greek MSS. The revision after 431 would probably have been from MSS. of the Byzantine type, and this seems to be confirmed by the existing MSS. Mr. F. C. Conybeare, who was the first scholar in modern times to study the Armenian version, argues that there was an earlier translation, since a reference to one is found in an Armenian writer, Theodore, of the 7th century, who says also that St. Gregory the Illuminator, who brought Christianity to the Armenians at the beginning of the 4th century, quoted from the apocryphal 3rd Epistle to the Corinthians. This suggests that the early Armenian version contained not only the Gospels but the Epistles, and that it was made from the Syriac, in which alone 3 Corinthians appears. No other trace, however, of this version exists.

There are dated MSS. of the Armenian version of 887, 902, 960, 986, and 989. Three early MSS. examined by Conybeare at Edschmiadzin (the principal library of Armenian literature) omit the last twelve verses of Mark, and a fourth,' which has them, adds a note saying that they are " of the elder Ariston." This has been taken to mean that, the original ending of the Gospel having been lost at a very early date, these verses (which are in quite a different style from the rest of the Gospel) were added by the elder Aristion, who is named by Papias as one of the disciples of our Lord from whom he derived information. It may be so (the difference in the spelling of the name may be ignored), but there is no confirmation of it. Three of Conybeare's MSS. also omit Luke xxii.43, 44, but the passage is in the oldest of them, and according to Theodore it was also in the ancient Armenian version but was subsequently omitted (by Mesrop and Sahak). The early Armenian MSS. also omit the pericope adulterae, and the earliest that has it (that of 989) gives it in a different form from the Greek, with the curious addition that what our Lord wrote on the ground was the sins of the several self-constituted judges.

In general, the value of the Armenian version is for its evidence with regard to the Old Syriac. It was first printed at Amsterdam in 1666, and again in 1805 in a critical edition by the Mechitarist fathers at Venice. The best account of it is by Conybeare in the 4th edition of Scrivener's Introduction (1894); but cf. Armitage Robinson (Euthaliana, 1895), and R. P. Blake (The Caesarean Text of Mark, in Harvard Theological Review, xxi, 307-10, 1928).


The Georgian version has only lately come under notice, and has been studied by few scholars. So far as England and America are concerned, the main guides are Conybeare (in Scrivener, Academy of Feb. I,1896, American Journal of Theology i, 883-912,1897, Zeitschr. f.d. neutestamentliche Wissenschaft, xi, 232-249, 1910) and Blake (op. cit. pp. 286 ff.). According to Blake, the Gospels and Pauline Epistles were always in separate MSS. until modern times, and have distinct textual traditions. The Apocalypse, first translated about 978, was never fully accepted. The earliest and best Georgian Gospels is the Adysh MS., written in 897, which stands by itself. A second group is headed by the Opiza MS. (913) and the Tbet' MS. (995). All others represent a revision in the nth century from Greek MSS. of the Byzantine type, which became the accepted version of the Georgian Church, and is of little importance. The original version was undoubtedly made from the Armenian, but from an Armenian text older than, or better preserved than, any now extant Armenian MS.; and its special interest is that it goes back to a Greek text of the type described below as Cassarean, of which the leading Greek representatives are P46, W (in part of Mark), Θ, fam. 1 and fam. 13. Blake's conclusion is that "the Old Georgian version, represented at present by the Adysh manuscript and by Codices A and B " [Opiza and Tbet'] " was one of the best witnesses to the Caesarean text. It has been corrected by Ecclesiastical " [Byzantine] " copies in the Adysh MS. and somewhat more extensively, but quite independently, in A and B. Further, this Georgian was made from the Armenian. But the Armenian from which it was made has been greatly revised in the MSS. which are the basis of the printed text and in all of the oldest Armenian MSS. which have been the object of special study," as is shown by the fact that one or both of the Georgian recensions contain a number of Caesarean readings which are not in the Armenian as now extant.

It may be observed that the Adysh and Opiza MSS. omit the last twelve verses of Mark, while the Tbet' MS. contains them.

A critical edition of the Georgian Gospels, based upon the Opiza and Tbet' MSS., was published by V. N. Benesevic in 1909-11; the later official vulgate has been published by the Bible Society.


| coptic | sahidic | bohairic | middle egyptian | remainder |

Coptic is the ancient language of Egypt, formerly written in hieroglyphic, hieratic, or demotic, but here written in Greek characters, with the addition of six letters to represent sounds not used in Greek. In this form it was the language and the script of the native Christian Church in Egypt. It seems to have come into existence in the latter part of the second century. In a rather primitive form it appears in a papyrus in the British Museum, written on the back of a horoscope, the date of which is either 95 or 155, but more probably the former. The recently discovered Chester Beatty papyrus of Isaiah (see above, p. 44) has in its margins, in a hand of the early third century* a number of notes in the Coptic language, but written in Greek without any additional letters, which seems to show that the script was in an early stage. From all the evidence available it would appear that round about the end of the second century this form of writing was available to receive the Christian Scriptures.

There are two main dialects of Coptic, viz.: Sahidic, the dialect of Upper or Southern Egypt, and Bohairic, the dialect of Lower or Northern Egypt. Intermediate between these, both geographically and in character, are some Middle Egyptian dialects, of which the more important are known as Akhmimic and Fayyumic, but these are of less importance. The Bohairic eventually became the accepted Egyptian version, but the Sahidic was the earlier. When it was first made is uncertain. At the beginning of the fourth century Pachomius, the great organizer of monachism in Egypt, required his monks (who were common Egyptians, not likely to know Greek) to be diligent in the study of the Scriptures, which implies the existence of Scriptures in the vernacular. Still earlier, St. Anthony, who was born about 250, is recorded to have been greatly affected at the age of twenty by hearing the Scriptures read in church. He did not know Greek, but what he heard might have been an oral paraphrase from the Greek, similar to the Aramaic paraphrases of the Hebrew used in Jewish synagogues. An indication of early date may be found in the fact that the Sahidic text of Job is pre-Hexaplar (see above, p. 59), which points to a date not later than the first quarter of the third century. This may fairly be taken as the lower limit, while the end of the second century is perhaps more probable.

(a) Sahidic. Until the discoveries of papyri in Egypt during the last sixty years, the Sahidic version was only known in relatively few and scattered fragments, some of which were edited by Woide in 1799 and by Zoega in 1810. Now, however, the number of fragments and of substantial MSS. has increased so greatly that Mr. G. Horner, the diligent editor of both the Sahidic and the Bohairic New Testament, was able to produce a Sahidic New Testament, lacking only a few verses, and could say that, at any rate in the Gospels, there were seldom less than three authorities for any particular passage. Since, moreover, the MSS. show little difference among themselves, it may be assumed that we have now an adequate knowledge of the version as a whole. The MSS. mostly came from the White Monastery (south of Thebes), and are divided between Paris, London and other libraries. Several of them are bilingual, Greek and Coptic. Of the larger MSS. the most notable are a nearly complete Acts (with Deuteronomy and Jonah) of the first half of the 4th century (British Museum Or. 7594) and a John, nearly complete, probably of the second half of the same century (in the library of the British and Foreign Bible Society). A complete copy of the Pauline Epistles (9th century) is in Mr. Pierpont Morgan's collection of Coptic MSS., and Mr. Chester Beatty has one MS. of the Pauline Epistles and John and another of Acts and John, datable about 600.

In text the Sahidic used formerly to be claimed as, to some extent at least, an ally of the D and Old Latin type; but increasing knowledge weakens this impression, and shows that it is predominantly associated with the א B or Alexandrian type. Thus in the selected readings in the Gospels referred to above (p. 120) the Sahidic agrees with א B against D in 46 places, and with D against א B in 11: and its total agreements and disagreements with א, B, D and the Old Latin show the following results:

    With Sah. Against

In the more important of these passages, 33 in number, the preponderance of agreement with א B is even more striking. The agreements with B are 28, with א 24, with D 7, and with O L. 7. In 7 other cases the O L. testimony is divided.

In Acts also, out of 209 readings amounting in length to a στίχος or nearly so, which are marked in Prof. A. C. Clark's recent edition as specially characteristic of the B—Old Latin text, only 25 have Sahidic support. It would therefore plainly be unjustifiable to reckon the Sahidic as forming part of one group with D and the Old Latin. Rather it must be said to be in the main Alexandrian, though with a sprinkling of non-Alexandrian readings, some of which were adopted in the text which had its vogue in the West.

Of particular readings, the following may be mentioned. In the Lord's Prayer (Matthew vi, 13) Sahidic has the doxology (against א B D), except καὶ ἡ βασιλεία. It omits Matthew xii.47 with א B but against Bohairic. It omits xvi.2, 3, with א B and the Old Syriac, against D, the Old Latin, and most Bohairic MSS. It omits xviii.11, with א B and the Bohairic, against D and the Old Latin. In xxiv.36 it omits οὐδὲ ὁ υἱός, with fam. 1, the Old Syriac and Bohairic, against א B D and the Old Latin. In xxvii.49 it omits the piercing of the side, with D and all the versions, against א B. Some MSS. have the shorter ending to Mark, the rest have the longer. It omits Luke ix.55-6 with א B, against D and the Old Latin; the Old Syriac and Bohairic are divided. In Luke xxii.43, 44 it omits the Agony and Bloody Sweat with B and Bohairic against א D, the Old Latin and Old Syriac, also the Word from the Cross in xxiii.34 (with B D Bohairic). In xxiii.53 it has (in a slightly varied form) the curious description of the rolling of the stone to the sepulchre as it appears in D (see p. 92 above); but it agrees with none of the omissions characteristic of D and the Old Latin in xxiv. In John v.4 it omits the descent of the angel to stir the waters, with א B D and most MSS. of Bohairic, against the Old Latin. It omits the pericope adulterae. It omits the baptismal formula in Acts viii.37, with א B against the Old Latin (D is deficient here). Of the many additions in Acts characteristic of D and its allies, it has very few; but in xv.20, 29 it has the " negative Golden Rule " (καὶ ὅσα μὴ θέλουσιν ἑαυτοῖς γίνεσθαι ἑτέροις μὴ ποιεῖν) and in xx.15 the addition καὶ μείναντες ἐν Τρωγυλίῳ. The other additions which it has in common with the D-type are of minor importance and include none of striking character. In the Pauline Epistles the Sahidic is generally in agreement with B.

(b). Bohairic. The Bohairic version has been much longer and more fully known. Readings from it were contributed by T. Marshall to the critical editions of the New Testament by Fell (1675) and Mill (1707), and the whole was published by Wilkins in 1716; but all other editions have been superseded by that of G. Horner (Oxford, 1898-1905). In this 46 MSS. were used for the Gospels, and 34 for the other books. The best Gospels MS. is Huntington MS. 17 in the Bodleian (A.D. 1174), though the Curzon Catena, in which text and commentary are intermingled, is Older (889). For the Acts and Epistles British Museum Or. 424 (1307) is considered the best, and for the Apocalypse Curzon MS. 128 (1320). It will be seen that these MSS., though among the oldest, are relatively late, but the variations among them are slight, and the text shows little sign of having been seriously altered by revision. In character it is definitely to be associated with the א B family, though not so much more so than the Sahidic as was formerly supposed. Thus in the selected readings from the Gospels, where the א B and D readings in the Sahidic are (as already stated) 46 and 11 respectively, in the Bohairic they are 50 and 15; and the total agreements and disagreements are:

    With Bohairic. Against

These figures differ very little from those given above for the Sahidic, and show that both forms of the Coptic Gospels are to be reckoned as equally in the main supporters of the Alexandrian type, with a similar proportion of minor non-Alexandrian variants. In Acts, however, the Bohairic is more exclusively Alexandrian, for it contains none of the D-type variants which were noted above in the Sahidic.

As to its date, Mr. H. C. Hoskier (Concerning the Date of the Bohairic Version, London, 1911) assigns it to about 200-250, and argues that it had a considerable influence on the text of א.

(c) Middle Egyptian. Of these dialects too little is known to be worth discussing here. The surviving fragments are too few and too small, and their interest is rather linguistic than textual. All that can be said is that the underlying text is substantially the same as that of the Sahidic.

The remaining Eastern versions—Ethiopic, Arabic, Persian—can be passed over in a handbook of this scope, as having no importance for the main textual problems of the New Testament. The Ethiopic has one special interest, in that it has preserved the Book of Enoch, which at one time hovered on the verge of
the versions and fathers the Canon, as having been quoted in the Epistle of Jude. It was supposed to be wholly lost, until the traveller James Bruce brought back three MSS. of it from Abyssinia in 1773, from one of which (presented by Bruce to the Bodleian) it was edited by Dr. Laurence, Archbishop of Cashel, in 1821. The original Greek remained unknown (except for some considerable extracts in Syncellus) until 1886, when a small vellum book was found at Akhmim in Egypt, containing the first 32 chapters, with extracts from the Gospel and Apocalypse of Peter, in hands which can probably be assigned to the 6th century. More recently, the Chester Beatty papyri discovered in 1931 include several leaves of a codex (of the 4th or 5th century) containing Enoch (here entitled the Epistle of Enoch) and a homily by Melito of Sardis. ← These will be published shortly by Prof. Campbell Bonner, of the University of Michigan. The portion of Enoch preserved is the end of the book, chapters 97-107. It is interesting to find that this book was kept along with the canonical Scriptures in a Christian community as late as the fourth century.


| latin | old latin | vulgate |

The Latin versions have a double importance and a special interest. In their earlier forms they throw valuable (though somewhat puzzling) light on the original text of the New Testament and its fortunes in the early centuries. In its later form, the Vulgate of St. Jerome, it became the accepted Bible of the Western world for over a thousand years, and is the Bible of the Church of Rome to-day. It was also the text from which all translations of the Scriptures into English were made before the sixteenth century. As an element in the history of religion and of civilization, it is of unique importance.

(a.) The Old Latin. Much obscurity involves the history of the translation of the New Testament into Latin. It seems certain that it did not first take place in Rome. Rome was a cosmopolitan town where Greek was almost as well known as Latin, and it was among the Greek-speaking population that Christianity first took root. St. Paul writes to them in Greek, St. Clement at the end of the first century writes in Greek, so do Justin Martyr and Hermas in the middle of the second century, and Hippolytus at the beginning of the third; and nearly all the bishops of Rome until the end of the second century have Greek names. It was not there that a Latin translation was first needed, but rather in such provinces of the Empire as Gaul or Africa. For both of these districts we have some evidence from the last quarter of the second century. The Scillitan martyrs at Carthage in 180 possessed a copy of the Pauline Epistles, and apparently of the Gospels, and there is no sign that they were acquainted with Greek; and the letter which describes the persecution of the Churches of Lyons and Vienne in 177 shows familiarity with a Latin New Testament. Tertullian (c. 150-220) writing in Africa in Latin, quotes the Scriptures freely, but he is by no means an accurate writer, and he seems often to have made his own translations from the Greek, so that his quotations have to be used with caution.

Firmer ground is reached with Cyprian (c. 200-258), bishop of Carthage, who quotes extensively and accurately, and who certainly used a Latin Bible. Moreover of this Bible we have actual MSS.; for a Bobbio MS. (known as k), of the 4th or at latest 5th century, contains Mark viii-xvi (ending at verse 8) and Matthew i-xv, in a text identical with that of Cyprian, and a Codex Palatinus (e) at Vienna has all the Gospels (though considerably mutilated) in a similar text, while the commentary of Primasius in the 6th century gives an almost complete text of the Apocalypse of this African type. These collectively represent a great part of the New Testament as it circulated in Africa in the first half of the 3rd century, and it is to Africa that we must assign the beginnings of the Latin Bible as we know it, though it is possible that a version was simultaneously in existence in southern Gaul.

Of this African form of the Old Latin it can be said with certainty that, although it is not identical with the text found in Codex Bezae, it is very closely related to it. It supports D in many of. its most marked divergences from the common type, e.g., in the additional passage at Matthew xx, 28, in the omissions in Luke xxiv, and in many of the additions in Acts. The combination D and OL. (both in its African form and in the forms found elsewhere) constitutes a definite textual family to which the designation " Western " may be properly given.

For the African is not the only form of the Old Latin. A number of other MSS. with pre-Vulgate texts exist, which cannot be associated with Africa. They have much in common with the African text, but also many differences. Whether they represent a revision of the African text or an independent translation is a point upon which scholars differ, and it is not very material. They certainly represent a different type of text from that which we have found domiciled in Egypt and of which the foremost representatives are א B. As a rule, the larger divergences in the Old Latin version are found in the African form, the smaller in the European.
The most convenient way of stating the case and of indicating the available materials will be a short catalogue of the principal MSS. Most of them are fragments, but most are of early date. The following are the MSS. of the Gospels.

a. Codex Vercellensis, at Vercelli, 4th century. Contains the Gospels, somewhat mutilated, in the Western order (Matthew, John, Luke, Mark).

b. Codex Veronensis, at Verona, 5th century. Contains the Gospels, written in silver letters on purple vellum.

c. Codex Colbertinus, at Paris, 12th century. Written in Languedoc, where the Vulgate seems to have been very slow in superseding the Old Latin. Contains the four Gospels, complete, in the OL. with the rest of the New Testament added later from the Vulgate. Matthew and John are much less Old Latin than Mark and Luke.

Codex Bezae

d. The Latin text of Codex Bezae.

e. Codex Palatinus, at Vienna, with one leaf in Dublin and one in the British Museum. 5th century. Contains the Gospels, imperfect, in silver letters on purple vellum. Akin in text to k, and possibly written in Africa.

f. Codex Brixianus, at Brescia, 6th century. Gospels, nearly complete, in silver on purple. Regarded by Wordsworth and White as representing the type of Old Latin on which Jerome based his revision. This, however, is doubted by Burkitt, who regards it as a corrupted Vulgate text. ← Prof. Souter has argued that in Luke at any rate Jerome used a text practically identical with a.

ff1. Codex Corbeiensis I, at Leningrad, 10th century. Contains Matthew only in a mixed OL.-Vulgate text.

ff2. Codex Corbeiensis II, at Paris, 5th or 6th century. Gospels, mutilated, in a text akin to a and b.

g1. Codex Sangermanensis I, at Paris, 8th or 9th century. Contains all Gospels, but only Matthew is Old Latin, and that is of a mixed type.

g2. Codex Sangermanensis II, at Paris, 10th century. Gospels, in an Irish hand, with a mixed OL. - Vulgate text.

Codex Claromontanus

h. Codex Claromontanus, in the Vatican, 6th century. Contains the Gospels, but only Matthew is OL.

i. Codex Vindobonensis, at Vienna, c. 6th century. Fragments of Luke and Mark in silver on purple.

j. Codex Saretianus, at Sarezzano, 5th century. Fragments of John in silver on purple. Said to be akin to a b.

k. Codex Bobiensis, at Turin, 4th or 5th century. See above, p. 136. It has the shorter ending to Mark. Probably written in Africa.

l. Codex Rehdigeranus, at Breslau, 7th century. Gospels, lacking last five chapters of John.

Of the remaining letters, m represents the pseudo-Augustine Speculum, which contains copious extracts from nearly all books (but noticeably omitting Hebrews, which was not accepted as Pauline in the West), in an OL. text, n-v are fragments, of which n is closely akin to a, and p, r1, r2 are in Irish hands and perhaps represent an Irish OL. group.

For Acts there are the following:

d = d of Gospels.

Codex Laudianus

e. The Latin text of Codex Laudianus (E2).

g. Codex Gigas at Stockholm, 13th century. A huge MS. of the whole Bible, but only Acts and Apocalypse are OL. Written in Bohemia, which perhaps explains the appearance of an OL. text in so late a MS.

h. Palimpsestus Floriacensis, at Paris, 6th or 7th century. Fragments of Acts, 1 and 2 Peter, 1 John and Apocalypse, in an African type of text. Much used by Clark in his edition of Acts.

s. Codex Bobiensis, at Vienne, 5th or 6th century. Palimpsest fragments of the last six chapters of Acts, with James and 1 Peter, in a text akin to g.

For the Catholic Epistles, besides h, m, and s, only one MS. need be mentioned.

ff. Codex Corbeiensis, at Leningrad, 10th century. Contains James and the unique Latin text of the Epistle of Barnabas. Predominantly Old Latin, but with Vulgate admixture.

For the Pauline Epistles there are:

d-g. The Latin texts of Codex Claromontanus (D2), Sangermanensis (E3) Augiensis (F2), Boernerianus (G3).

gue. Codex Guelferbytanus, at Wolfenbüttel, 6th century. Fragments of Romans in the palimpsest MS. which also contains P and Q of the Greek Gospels.

r, r2, and r3 are fragments only.

x2. Codex Bodleianus, at Oxford, 9th century. Nearly complete, with a text akin to d.

For the Apocalypse the principal authority is the commentary of Primasius, with the extensive quotations by Cyprian and Tyconius. There are also g and h of the Acts, and the Speculum (m).

Of these, k, e and m represent the African family in the Gospels, h and m in Acts, m and Priscillian (a Spanish writer of the 4th century) in the Pauline Epistles, h, Primasius and Tyconius in the Apocalypse, and Cyprian throughout. For the European family, a and b are the leading representatives in the Gospels, with the Latin version of Irenaeus; in Acts g, s and quotations in Lucifer of Cagliari. The majority of the MSS. present mixed texts, with much variety, for it is one of the characteristics of the Old Latin version that its MSS. differ widely among themselves, so that one frequently finds Old Latin evidence on both sides of a doubtful reading. Augustine, the greatest of all the Latin Fathers, uses both Old Latin and Vulgate. This usage was cleared up by Prof. Burkitt, The Old Latin and the Itala (Cambridge Texts and Studies, iv, 3, 1896). who showed that in short passages, such as an author may easily quote from memory, he often used the Old Latin even to the end of his life; but in longer passages, for which reference to a manuscript would be necessary, he generally in his later years used the Vulgate, which he calls the Itala interpretatio ← Wordsworth and White do not accept this explanation of the phrase, preferring the older interpretation, according to which the Itala is a subdivision of the European class of Old Latin texts.. In particular he refers to the Acta contra Felicem, a report of a heresy trial in 404, in which a long passage from Luke is quoted and another from Acts; the Luke is Vulgate, and the Acts African Old Latin, which throws some light on the usage of the Church at Hippo at the beginning of the 5th century.

(b.) The Vulgate. Such was the state of things when Pope Damasus, about 382, invited Jerome, the leading Biblical scholar of the day, to undertake a revision of the Latin Bible, with a view to putting an end to the confusion caused by the existence of such a multiplicity of conflicting texts. The first sentence of the preface which Jerome eventually prefixed (in the form of a letter to the Pope) to his edition of the Gospels defines the task which was laid upon him.

Novum opus facere me cogis ex veteri, ut post exemplaria scripturarum toto orbe dispersa quasi quidam arbiter sedeam, et, quia inter se variant, quae sint ilia quae cum Graeca consentiant veritate decernam.

Of the disorder then existing in the current Latin Bibles, he makes no doubt:

Si enim Latinis exemplaribus fides est adhibenda, respondeant quibus : tot sunt paene quot codices.

The remedy is to have recourse to the best available Greek manuscripts, since Greek is the original language of the Gospels, and to revise the Latin so as to bring it into accordance with them. At the same time he wishes to do as little violence as possible to the feelings of those who would resent change in the words familiar to them (as we have found when the Revised Version was produced to supersede the Authorized). Therefore he will only make changes when the sense demands it:

Quae ne multum a lectionis Latinae consuetudine discreparent, ita calamo temperavimus ut, his tantum quae sensum videbantur mutare correctis, reliqua manere pateremur ut fuerunt.

It was on these principles that he undertook his revision of the New Testament—a revision, not a new translation; and on these lines he produced the Gospels in 384. The remaining books (published, according to Chapman, in 391) in which there was less confusion, since the influence of harmonistic alterations of the synoptic texts was absent, he treated more summarily. Consequently large elements of the Old Latin remain in the Vulgate, but he selected the variants which agreed with the Greek MSS. which he regarded as the best, or introduced new readings from such MSS. What these were can only be ascertained by comparison with the MSS. that have come down to us, since he rarely gives even the slightest indication of the authorities consulted by him. The conclusion to which Wordsworth and White come with regard to the Gospels, after most careful investigation, is that, while he sometimes followed Greek MSS. differing from any that we know, in the main he used MSS. of the class represented by א B L, and especially a MS. or MSS. closely resembling א. In Jerome's hands, then, the Old Latin version, already considerably modified from its African form in the direction of the Greek MSS., took on a distinctly Alexandrian colour.

With the Old Testament, for which, as described above (p. 59), he eventually deserted the Greek of the Septuagint and made a fresh translation from the Hebrew, we have nothing here to do. When the whole work was completed, about 404, it had to make its way in the face of much hostile criticism, occasioned not so much by the revision of the Old Latin in the New Testament, as by the wholesale changes caused by the abandonment of the Old Latin (and the LXX from which it was translated) in the Old Testament. Consequently its adoption was gradual, and in the course of the process it suffered much contamination with the Old Latin. From our description of the Old Latin MSS. it will have been seen that, for some obscure reason, Matthew in several MSS. remains Old Latin, while the other books are Vulgate; also that in such outlying districts as Languedoc and Bohemia the Old Latin was still being copied in the 12th and 13th centuries. It was thus in a much corrupted form that the Vulgate became the Bible of the Western world in the Middle Ages. From time to time attempts were made to revise it, notably by Alcuin and Theodulf about the beginning of the gth century, by Hartmut at St. Gall towards the end, and by the University of Paris in the 13th; but these rested on no firm basis of textual criticism, and did little to delay the general progress of deterioration. It was consequently in a far from correct form that the Vulgate appeared as the first book produced by the printing-press, the famous Gutenberg or Mazarin Bible of 1456. Other editions were published by Stephanus in 1540 and by Hentenius in 1547, but the first authoritative edition was that which, in pursuance of a decree of the Council of Trent, was produced in 1590 by Pope Sixtus V. This was, however, quickly superseded in 1592 by a new edition prepared under the direction of Clement VIII, which became (and remains) the official Bible of the Church of Rome.

It was not until our own times that a serious attempt was made to recover the true text of Jerome's Vulgate. For the New Testament this was undertaken at Oxford by Bishop John Wordsworth of Salisbury, with the collaboration of H. J. White (afterwards Dean of Christ Church). The Gospels appeared, in a revised text with a full critical apparatus, in 1889-1898, and Acts in 1905, and after Wordsworth's death in 1911 and the interruption caused by the war the work was carried on by White as far as Ephesians before his death in 1934. Arrangements have been made for the completion of the work, which it may be hoped is now in sight. Meanwhile White produced in 1911 a very handy pocket edition of the whole New Testament, with revised text and a select apparatus of the more important various readings from the nine leading MSS. (A, C, D, F, G, H, M, V, Z), and the Sixtine and Clementine editions. This is the most serviceable edition for general use; another useful pocket edition is that produced by E. Nestle in 1906, which gives the Clementine text with the variants of the Sixtine, those of Wordsworth and White as far as Acts and those of Lachmann (1850) and Tischendorf (1854) for the rest, and the MSS. A. and F. The practical needs of students are therefore well provided for.

The extant MSS. of the Vulgate greatly exceed in number those of the Greek New Testament. Gregory in 1909 (appendix to his Textkritik ) enumerated 2472, but White estimated the total as at least 8000. For the Gospels Wordsworth and White used 29 MSS., for Acts 17, for the Epistles 21. A few of the more important may be mentioned here, since they are sometimes referred to in connection with the criticism of the Greek New Testament.

A. Codex Amiatinus, in the Laurentian Library at Florence. A magnificent MS. of the whole Bible, written at Wearmouth or Jarrow under the direction of Abbot Ceolfrid, and taken by him as a present to Pope Gregory in 716. Some leaves of a sister MS. are in England (one in the British Museum and 11 in private hands). Generally regarded as the best MS. of the Vulgate.

C. Codex Cavensis, at La Cava in South Italy, 9th century. Contains the whole Bible, in a small Visigothic hand, and is the leading representative of the Spanish family of MSS.

D. Codex Dublinensis, or Book of Armagh, at Dublin, 8th or 9th century. Contains New Testament in the Irish type of text.

. Codex Epternacensis, at Paris, gth century. Contains Gospels, with a note stating that it was corrected in 558 from a MS. said to have been written by Jerome himself. The note must obviously have been copied from an ancestor of this MS. The text is mixed.

F. Codex Fuldensis, at Fulda, A.D. 541-6. Written to the order of Bishop Victor of Capua. Contains New Testament, the Gospels being in the form of a harmony which is believed to reproduce the arrangement of Tatian's Diatessaron, though thetext has been altered to Vulgate, of which it is one of the best representatives.

G. Codex Sangermanensis, at Paris, 8th or 9th century. Contains the whole New Testament; the Old Testament is Spanish in type, Matthew is Old Latin (see g1 of the Old Latin list), and the rest mixed French and Irish. Especially important for Acts, where Wordsworth and White consider it superior even to A.

H. Codex Hubertianus, in the British Museum, 9th century. Akin to A and Y, with corrections according to the edition of Theodulf.

Codex Karolinus

K. Codex Karolinus, in the British Museum, 9th century. A huge MS. of the whole Bible in the edition of Alcuin, and in Acts the best representative of this edition, which here is almost identical with F.

L, Codex Lichfeldensis, or Gospels of St. Chad, at Lichfield, 7th or 8th century. Contains Matthew, Mark and Luke i.1—iii.9, with illuminations in Anglo-Celtic style. Text akin to D.

O. Codex Oxoniensis, in the Bodleian, 7th century. Gospels, in a mixed text with some Irish influence, akin to X.

Book of Kells

Q. Codex Kenanensis, or Book of Kells, 8th century. Contains Gospels in an Irish hand, with the finest existing Celtic decorations. The text also is Irish in character.

S. Codex Stonyhurstensis, at Stonyhurst College, 7th century. Said to have belonged to St. Cuthbert. A beautiful little copy of John, in a text akin to A Y.

T. Codex Toletanus, at Madrid, 8th- century. Contains the whole Bible in a Spanish type of text.

V. Codex Vallicellianus, at Rome, 9th century. The whole Bible, in the edition of Alcuin, of which it is the best representative in the Gospels.

X. Codex Corporis Christi Cantabrigiensis, at Corpus Christi College, Cambridge, 7th century. Contains Gospels, in a text akin to O, which, like it, formerly belonged to St. Augustine's, Canterbury.

Lindisfarne Gospels. (opens in a new window)

Y. Codex Lindisfarnensis, or Lindisfarne Gospels, in the British Museum, c. 700. Written in honour of St. Cuthbert (d. 687) and illuminated in the finest style of the Northumbrian Anglo-Celtic school. Contains the Gospels, with interlinear Anglo-Saxon translation added in the 9th century. Akin in text to A; and a table of lections for special festivals shows that it must have been copied from a Bible used at Naples, probably one brought over by the Neapolitan abbot Hadrian, who came to England with Archbishop Theodore in 669. A fragment of Luke attached to a later MS. at Durham so closely resembles Y that Prof. C. H. Turner believed it might be the actual exemplar from which Y was copied (see New Pal. Soc., ser. i, pi. 157).

Z. Codex Harleianus, in the British Museum, 6th or 7th century. A beautiful little copy of the Gospels with a text differing from that of A Y.

Δ. Codex Dunelmensis, at Durham, 7th or 8th century. Traditionally said to have been written by Bede. Gospels, in a Northumbrian text, akin to A Y.

Θ. Codex Theodulfianus, at Paris, 9th century. Contains the whole Bible in the edition of Theodulf.

Σ. Codex Sangallensis 1395, at St. Gall, 6th century. Consists of leaves used in bindings. About half the Gospels text survives. The oldest MS. of the Vulgate Gospels, but not used by Wordsworth and White. According to Turner, its text is of the same class as Z.

These MSS. fall into several distinguishable groups or families. That which Wordsworth and White (with general acceptance) regard as the best in the Gospels is the Northumbrian group headed by A S Y Δ, written at the time when, under the leadership of Bede and Ceolfrid, Northumbria led the world in Biblical scholarship. It has been argued by Dom Chapman that it descends from the edition known to have been prepared by the scholar-statesman Cassiodorus (d. about 580). Closely akin to this group is F, which is not Northumbrian in origin. A separate and less good group is headed by Z; these are characterized by Turner as non-Cassiodorian texts current in Italy in the 6th century. An Irish group is formed by D L Q; and O X form a mixed group between this and the Z-group. C T represent the text of Spain, K V the edition of Alcuin, and Θ and the corrections in H that of Theodulf. In Acts several of these MSS. drop out, and the editors' preference is for G C A F D, in that order, the agreement of them representing a combination of all the principal lines of text. For the Epistles the principal MSS. are the same, but White had not at his death produced an estimate of their respective quality.

The Vulgate is so important for the history of the Bible in the West, that it has seemed worth while to describe the authorities for its text at some length; but owing to its very mixed character it is of less value for the recovery of the original Greek than the earlier Syriac, Coptic, and Old Latin versions.


Only one other version need be briefly mentioned. This is the translation made by bishop Ulfilas in the 4th century for the Goths in Moesia. It was made from the Greek in both Testaments; in the Old Testament from a Lucianic text of the LXX, and in the New Testament from a text predominantly of the Byzantine type, with a sprinkling of earlier readings. It exists only in fragments, the principal MS. being the splendid Codex Argenteus at Upsala, of the 5th or 6th cent., containing rather more than half the Gospels written in silver upon purple vellum.


One other important source of evidence remains to be mentioned, that of the Scriptural quotations found in the early Christian writers. Most of those that are important in this respect lived at dates earlier than those of most of our oldest MSS., and if we could be sure that we had them as they wrote them, we should have, for the passages quoted, the evidence of older MSS. than those which we possess, and we should know that MSS. of that particular type were in circulation in a particular part of the Christian world at a known date. Unfortunately we cannot always be sure that the quotations have come down to us intact. Copyists were apt, when they came to a Scriptural passage, to write it down in the form with which they were familiar, and so the later MSS. of the Fathers not infrequently show a different text from the earlier ones. It is necessary therefore first of all to secure scientifically edited texts of the Fathers themselves; and in this direction considerable progress has been made in the series of Greek Fathers undertaken by the Academy of Berlin, and of the Latin Fathers by that of Vienna. ← An index to patristic quotations was compiled by Dean Burgon, and exists in 16 huge manuscript volumes in the British Museum (Add. MSS. 33421-36). When used with the proper precautions, however, and remembering that in short quotations a writer is apt to rely on his memory and that some writers are inaccurate quoters, the evidence of the Fathers is- of the greatest value in determining the time and place at which the principal types of text came into existence or were current.

A brief summary, therefore, of the Christian writers in the earliest centuries will be useful. In the sub-apostolic age quotations are rare and inexact, and their chief value is to prove that certain books were then in existence. Such are the few passages which appear to show a knowledge of the Gospels and Pauline Epistles in the Epistles of Clement, Barnabas, Ignatius, and Polycarp. From the second half of the second century we have the Apology of Justin Martyr, the Diatessaron of Tatian (so far as its text is recoverable), and the extensive quotations from Marcion preserved in the writings of those who sought to confute his heresies. Of all of these it can be said that the evidence, so far as it goes, seems to show that they used texts which show a considerable amount of deviation from the Alexandrian type. The same is the case with Irenaeus, whose principal work, a Confutation of Heresies, was produced between 181 and 189, written in Greek but extant mainly only in a Latin translation, the date of which is doubtful. Irenaeus in his youth heard Polycarp in Asia Minor, but lived and worked mainly at Lyons, so that the MSS. at his service must have been those current in southern Gaul.

From Egypt our earliest patristic evidence is that of Clement of Alexandria (c. 150-215), a priest of the Alexandrian Church, and from 190 to 202 head of the catechetical school in that city. He was an industrious writer, widely acquainted with pagan as well as Jewish and Christian literature. His quotations are plentiful, and it is a noteworthy fact, in view of his place of residence, that they also are generally not of the א B family. That family makes its first appearance in the writings of Origen, the greatest scholar among the early Greek Fathers. Born about 185, he succeeded Clement, when barely eighteen, as head of t,he catechetical school at Alexandria, and in that city he worked, save for a visit to Rome in 213 and a four years' residence at Caesarea in 215-9, until in 231 he was forced to leave Alexandria finally, when he took up his abode at Caesarea until his death in 253. Origen was essentially a scholar, with some conception of the principles of textual criticism. His labours on the Septuagint have been described in Chapter II. Of the New Testament he made no edition, but he wrote commentaries on nearly every book, in which he not only quoted copiously, but repeatedly refers to varieties of readings in different manuscripts. In his earlier writings he used MSS. predominantly of the א B type, though readings of a different character are not wanting; but after his final departure to Caesarea he seems to have used MSS. of another type, as to which more will be said when we are examining the later history of textual criticism (see p. 177 below). At his death his manuscripts remained at Caesarea, and became the nucleus of a famous library formed by his disciple Pamphilus (d. 309), to which references occur in later literature and in notes in various MSS. (e.g. א, see p. 48).

Of the Latin writers after Irenaeus, Tertullian (c. 150-220) and Cyprian (c. 200-258) have been mentioned above (p. 136) in connection with the Old Latin version. From Rome itself there is only Hippolytus, who flourished about 220, and wrote (in Greek) commentaries on Matthew, John and the Apocalypse in addition to his great Refutation of all the Heresies. Of all these works only portions remain, and no special study seems to have been made of his quotations. For the Syriac Church the most important writers are Aphraates (fl. c. 340) and Ephraim (d. 373), both of whom are of great value for the reconstitution of the Old Syriac version. Eusebius of Cassarea (c. 270-340), the historian of the early Church, the friend of Pamphilus and bishop of Caesarea, carries on the tradition of Origen and had access to his library. In Asia Minor there were during the fourth century Basil and the two Gregories; in Palestine and Syria, Cyril of Jerusalem (bishop 351-386), and especially Chry-sostom (c. 347-407), who worked at Antioch until 398 and thereafter at Constantinople, and in whose voluminous writings we find the first stage of that revised text which eventually became the accepted Bible of the Byzantine Church.

The two great Latin Fathers of the late fourth and early fifth century, Jerome (c. 345-420) and Augustine (354-430) have already been mentioned (pp. 141-3). To them may be added Lucifer of Cagliari (d. 371) and Ambrose (bishop of Milan 374-397), both of whom are useful for the Old Latin version. Finally, mention has already been made of the value of Tyconius and Primasius for the Old Latin version of the Apocalypse, and of Priscillian for the Pauline Epistles.

After the first quarter of the fifth century the Fathers lose much of their value for our present purpose. The Byzantine revision was establishing itself in the East, and the Vulgate in the West, and the writings of the Fathers throw less and less light on the subject with which we are concerned, the recovery of the earliest form of the Greek Scriptures.


The same general works as have been given in the bibliography appended to Chapter III. The principal works relating to each Version have been cited in the descriptions of them. Reference may also be made to the articles on Versions in Hastings' Dictionary of the Bible and Cheyne and Black's Encyclopaedia Biblica, though these are not now fully up to date. For the Fathers, Harnack, Geschichte der altchristlichen Litteratur bis Eusebius (1893-1904); Smith and Wace, Dictionary of Christian Biography.

[katapi ed: In recent years, much attention has been given to digitizing MSS., and a good place to search on-line is the CSNTM website. The source MSS websites can accessed through their Manuscripts list. They deal in the main with NT mss., but they are worth checking out none-the-less. (Your might also try the Wikipedia website for more background on individual MSS.)