Obrazy na stronie


[ocr errors]
[ocr errors]
[ocr errors]
[ocr errors]

The Greek uncial characters of Codex Bezae are in the main of the plainest yet most graceful shape. The form of alpha (the lower left limb sometimes passing from the vertex of an angle into a loop or curve, e.g. Facsimile I, 11. 4, 5), beta, mu, nu, upsilon, omega, and more especially delta are of the simple and ancient form: the horizontal line in theta is so fine that it is often scarcely visible: the circle in phi (always a large letter in the oldest copies) sometimes becomes nearly a complete lozenge (0, Facsimile I, 1. 22, and 27 b. 13): in pi the thin horizontal line is always produced slightly beyond the two verticals (which is not much the practice in Codd. Vaticanus and Sinaiticus, nor always in Cod. Alexandrinus, e.g. Gen. i. 1), and frequently terminates in a hook on the left (e.g. Facsimile I, 1. 5): kappa and rho have usually (though not invariably) at their bottom thin horizontal strokes running on to the left, at times barely visible to the eye, occasionally joining them to the preceding letter (Facsimile I, 11. 5, 9 &c.); the vertex of the angle to the right in kappa almost always touching the vertical line. For the rest, the knobs at the upper and lower extremities of epsilon and sigma are rarer than in any save the very oldest manuscripts; eta (H) at times degenerates into something nearer to N (e.g. Fol. 434 b. 4); the horizontal line of tau is at times very thin and plain, though it is oftener strengthened either at its right or left extremity, or at both, by small knobs or hooks. Gamma and xi are among the peculiar letters of Codex Bezae; the upper limb of the former, instead of being straight and horizontal, is often curved much as in an English r (e.g. Facsimile I, 1. 12); the latter is in some places so much like zeta that it only differs from it by having a trifling protuberance in the middle of the descending line (Facsimile I, 1. 12), which now and then is scarcely perceptible (e.g. Foll. 54 b. 28; 148 b. 24 secundo loco)'. All these circumstances (not the less important by reason of their delicacy and minuteness), when taken together, would lead us to assign to this manuscript full as high a date as to the Codex Alexandrinus, which was written early in the fifth century, were not our couclusions somewhat modified by other considerations, of which the debased dialect of the Latin version (on which we shall dwell in Chap. 11I) is the most obvious and weighty: the palaeographical appearance of the Latin character is venerable enough.

The most hasty glance at our Facsimiles I, II of the open leaf of Cod. Bezae suffices to shew the justice of Uffenbach's statement (see p. xi) that the Latin and Greek characters resemble each other so closely that at first sight the one might easily be taken for the other. are not much like the classical fragments written in square uncials, as represented in the second volume of Silvestre’s Paléographie Universelle, such as the Medici Virgil assigned to the third century, the Vatican Virgil of the fourth or the Sallust of the fifth century. They ought rather to be compared with the more round and flowing letters of the versions of the Old Latin, as the Vercellae, Verona and Brixia Gospels (for which, however, we have access only to Blanchini's very poor Facsimiles), the Codd. Palatinus and Claromontanus as published by Tischendorf, and the Cod. Laudianus 35 (E of the Acts). The style of this last is evidently of a somewhat later date, and much heavier in appearance. T, so slender and graceful in Cod. Bezae, in Cod. Laud. is almost always turned up at its lowest extremity with a sort of hook to the right. G, which in our manuscript

G is at times barely distinguished from C (Facsim. II, 11. 8, 12, 22), has a long tail in Cod. Laud.: in Cod. Palatinus the tail or extremity coils inwards in a spiral form, quite peculiar to that copy. Of all these Latin monuments its coeval the Cod. Claromontanus most resembles ours, only that the hand of the former is less firm and regular. M and U are shaped nearly alike in all : the Claromontane alone agrees with Cod. Bezae in having the last stroke of d perpendicular"; the other five 1 Of the compound letters so familiar in the oldest

2 C and G are so much alike that the scribe is apt to documents we find NI 6b. 12; HM 441 6.3; and forming write one for the other (e. g. Foll. 31 a. 4; 41 d. 33 ; 47 a. I letters of separate words N 477 b. 19; 489 b. 19; 506 read garcere; 48 a. 14; 79 a. 17; 179 a. 18; 254 a. 26; b. 29; M 90 6. 12: also very often, whether in the 274 a. 14; 313 a. 18; 462 a. 5; 481 a. 25; +95 a. 23; same word or not (e.g. both in 428 6. 22).

502 a. 10; 500 a. 17; 508 (I. 23).

The Latin pages




up into a curve, more or less flourished. Codd. Bezae and Claromontanus again conspire with Cod. Augiensis of the ninth century in making 1 with a simple perpendicular line very slightly turned

up in a fine stroke at bottom : the rest have a strong vertical line, sometimes a little curved, though seldom so long as the perpendicular member. FPQR descend below the line in most of these copies (not however FR in that of Vercellae, or PR in that of Verona, or R in the Brixian and Claromontane, but even the angular A in Cod. Palatinus), though our codex alone has the fine strokes running to the left at the bottom of these letters. On the whole, however, the impression conveyed by a careful comparison of them all would suggest the notion that the nice discrimination of their dates by means of the style of writing is not so easy or so certain in regard to Latin manuscripts of the fourth to the sixth centuries as with Greek documents of the same age (see also Tischendorf, Cod. Palatin. Prolegom. p. xv).

The sister bilingual Codd. Bezae and Claromontanus afford the earliest, and, in fact, the two chief extant examples for the New Testament, of manuscripts divided into verses, or otixou, the Latin lines being intended to follow the Greek, and only differing from them by accident. It must be conceded that this division, as applied to the books of Holy Scripture, prevailed much earlier than has been generally supposed. Not only do Athanasius [d. 373], Gregory Nyssen [d. 396), Epiphanius [d. 403] and Chrysostom [d. 407] inform us that in their time the book of Psalms was already divided into orixoi, while Jerome [d. 420 ?] testifies the same for the book of Isaiah (Suicer, Thesaurus Ecclesiast. T. 11. p. 1033), but Origen also [d. 254] speaks of the second and third Epistles of S. John as both of them not exceeding one hundred otixou, of S. Paul's Epistles as consisting of few, S. John's first Epistle of very few (Euseb. Hist. Eccl. L. VI. c. 25), while Eustathius of Antioch in the fourth century reckons 135 such orixou from John viii. 59 to x. 31'. The student, indeed, may see this for himself in regard to the poetical books of the 0.T., as preserved in the two great reliques of the fourth century, Codd. Vaticanus and Sinaiticus; wherein, according to the true principles of Hebrew poetry, the verses do not correspond in metre or in the quantity of syllables, but in the parallelism or mutual relation subsisting between the several members of the same sentence or stanza. In Codex Bezae, this orderly and systematic arrangement, which must have been designed by those who first distributed into orixou the sacred text of the Gospels and the Acts, bas to a certain extent been disturbed and broken up, in some places much more than in others. Now since it will appear clearly hereafter that the manuscript as it stands at present was closely and exactly copied from another, perhaps almost contemporary to itself, similarly divided in respect of orixou though not similarly paged (below p. xxiii), it will follow that the model from which this latter was taken is older still, dating perhap as early as or earlier than the time of Origen. The reader will not doubt that the ancient otixou were being gradually dissolved in course of time by successive transcribers, if he pays any attention to their actual condition in Codex Bezae. In the first Gospel, although many of the clauses are not balanced in the strict and regular fashion which would have satisfied the laws of parallelism as laid down by Bp. Jebb (e.g. Matth. vi. 1; ix. 17; xi. 21; xiii. 40; xvi. 18; xxvi. 51 &c.);


; . though a few lines end in yap (iv. 10; vi. 7; xvii. 15), or in de (xx. 2), or even in the article (vi. 6; xiv. 35 ; xix. 1), yet the close of each otíxos usually coincides with some slight pause in the sense.

With the first page of S. John the dissolution of the verses becomes much more marked (e.g. i. 4; 10; 13), and though only one line (i. 16) ends with the article before ch. vi. 32, yet such irregularity occurs no less than 48 times from that place to the end of the Gospel, while in the

1 I thankfully accept the correction of Tischendorf (Cod. Sinait. Proleg. p. xxi, note 2, edit. min.) to modify the statements of my Plain Introduction, p. 45.




succeeding Gospel of S. Luke an entire breaking up of the stichometry becomes rather the practice than the exception: about Luke viii. the dissolution seems ado; ted almost in preference; prepositions being separated from their cases (e.g. John xiv. 23 ; xxi. 8; Luke vii. 20; viii. 13 ; x. 7), or even words (not always compound words) are divided, whether in the Greek (e.g. John xiii. 36; Luke i. 1 ; vi. 1; 38; vi. 6; xxi. 36), or in the Latin (e.g. Matth. xviii. 33; John


; vi. 18), or in both (Luke v. 19; vi. 9; 48; vii. 20). As the work proceeds from the middle of S. Luke onwards (however we may account for the fact), the arrangement of the orixou becomes less broken and careless, though some of the chief anomalies are met with even to the last (e.g. Mark xiii. 22 Gk. Lat.; Acts iii. 26 Gk.; x. 41 Gk.; xi. 2 Lat.). Although the Latin translator no doubt intended to follow implicitly the lines of the Greek on the opposite page, yet in very many places (full twenty in S. Matthew alone) he departs from them without any visille cause (see Matth. xvii. 25-27; Mark ix. 1); in Luke iii. 1 there is utter confusion between the two, while

; in Luke iv. 16 the Greek has lost something; on Fol. 14 a, 1. 17 of the Greek is dropped altogether (just as a whole otíxos is lost in both after 238 b, 239 a, 1. 30 ; 421 b, 422 a, l. 23), though the lines are set right again by dividing 1. 25 of the Greek; so also Fol. 314 a, 11. 20—31. Both the Greek entirely and the Latin nearly repeat a line Mark xiv. 16; both misplace l. 13 Foll. 566, 57 a, which should precede l. 11; and the Latin transposes lines at Acts xvii. 17 ; so Acts v. 29 in both: in Fol. 259 6 33 kai is even employed as a catch word. Yet on the whole the tendency is (as will be seen hereafter, p. xxiii), for the parallel Greek to keep in check the wandering eye of the scribe when engaged in copying the Latin from its immediate prototype, which, judging from the style of the Latin version (see Chapter iii.), must have been nearly of the same age with himself.

It may be convenient to notice here that citations from the Old Testament are indicated in thirteen places (Matth. xxi. 5; 13; 16; 42 ; xxii. 44 ; xxvi. 31; xxvii. 9, 10 Lat.; Mark i. 2, 3 ; Acts i. 20; ii. 25—28; iv. 25, 26; vii. 49; xiii. 33–35) by throwing the beginning of the orixou

; back two or three letters : which same is done in Mark xiii. 18, though there is no quctation.

Abridgements in the Greek text (those of the Latin will be described hereafter in Chapter iii.) are fewer than in Cod. Sinaiticus and some others : they are - over the last vowel in a line for v (in Latin m), and the letters used for numerals, with a line placed over them, so common in the oldest manuscripts : in Mark vi. 44 e stands for 5000. Codex Bezae is peculiar in its mode of abbreviating inoovo, Xplotoo, and their cases, by writing always ino xpo for the usual forms ισ, χσ, &c. : θεοσ and κυριοσ are shortened into θσ and κσ, as in other copies : πνευμα is usually given as πνα (even πν Acts v. 9; τοισ πνευνα Mark 1. 27); μητηρ, υιοσ, σωτηρ, ουρανοσ, δαυιδ,


v. 9; i. ); , vico, , , lepovoalnu are never contracted, and tatyp not often, yet we see map John xii. 26 ; xiv. 28; xv. 1;


; προσ bid. vi. 65; πρσ, πρι, πρα as elsewhere : we have . . ; , ,

in full, as also OTPU Mark xv. 30; 32; otpv ibid. v. 13; otv v. 14; orn v. 15; plenè v. 20.

; The punctuation is certainly quite as primitive as in many parts even of Codd. Sinaiticus or Palatinus, consisting as it does chiefly of a blank space between the words, or of a middle, sometimes of an upper, very seldom of a lower single point, usually placed in the middle of a verse or oríxos, and found (as in most other copies) much more thickly in some parts than in others: such a point is often set in the middle of a line, in passages where it is hard to see its use. In rare instances, and for special reasons, two stops may occur in one verse (e.g. 104 b, 1. 4): double points also are sometimes placed in the larger spaces, mostly by a later hand, now and then p. m. (e. g. Foll. 79 6 1. 28, 80 a 28; 188 a 29, and occasionally in Cod. Sinaiticus): in this last case we carefully retain them. Nor are capital letters more frequent in this copy than in monuments which all agree to refer to the fifth century : much less so than in Latin copies of that date. Indeed the distinction between

[ocr errors]


and σταυροω



[ocr errors]


capital and other uncial letters is seldom strongly marked in Codex Bezae, so that some have been loth to admit that it employs capitals at all, believing that like Codex Sinaiticus a letter of the same size as the rest is set at the beginning of a line out of range to indicate the commencement of a new clause'. Yet not a few instances will be found where a sentence is begun, even in the middle of a line, with a letter larger than the rest (e.g. 146 6 12; 153 6 23; 198 6 32; 206 a l; 216 b 27; 306 a 17; 324 a 18), so that the use of such capitals in our manuscript cannot be denied, though they are not so conspicuous in size as in one or two others (e.g. Codd. Alexandrinus and Ephraemi) as old or older, which have them only at the beginning of a line.

The dialysis, expressed by two points (.., but :: in Fol. 433 6 14) set over iota and upsilon when used alone, and over upsilon or both in the diphthong vi, occurs perpetually in Codex Bezae, as in all the oldest copies, very often primâ manu, sometimes plainly by a later hand (in Acts ii. 17 EVŪTVLOLO is inserted s. m., a straight line being put for the dialysis, as often in Cod. Sinaiticus“), in which case we omit it in our edition: frequently it is quite uncertain whether it was placed by the original scribe, or by some of those who followed him, and then we retain the dialysis, giving it, as it were, the benefit of the doubt. The only other distinctive mark found in this document is the apostrophus ('), in respect of which the same difficulty constantly arises, whether it was written by the first hand or not; a question only to be decided by observing closely the precise shape of the apostrophus (usually thus ) in Codex Bezae), and by comparing the shade of its ink with that on the rest of the page. This mark is judged to be primâ manu, wherever we have retained it in our collation, which is only with the Hebrew Proper names ιερουσαλήμ, αβρααμ, σαμουηλ, δανειο, ιωσηφ, μαριαμ (αβελ Lulke xi. 51 should perhaps have been added): with p final Matth. ii. 9; vi. 2; Acts xi. 24: with & final Matth. xxvi. 41, as in Cod. Sinaiticus vvê' Apoc. xxi. 25 &c.: and once only after an elision, ped' John xvi. 4, which last is possibly by a later hand? Iota subscript of course never occurs in Codex Bezae, but ascript (of the same size as the other letters) in Mark i. 34 p. m. nidioav, while in another place the penman betrays a consciousness of its existence, for in Acts xxii. 3, while writing matpwov, he commenced i immediately after w, though he afterwards altered it into o. Of accents I find no vestige by the first pen, of breathing only one instance (Matth. xxv. 15 w jev: so even Cod. Sinaiticus Gal. v. 21), but a few of each in later hands*.

All that appears in our printed pages, therefore, we judge to have proceeded from the original writer of the manuscript, with a single exception now to be noticed. Codex Bezae, in its primitive state, contained no numerical divisions of the books whatever; neither chapters peculiar (or almost peculiar) to itself like the Codex Vaticanus, nor the Ammonian sections and Eusebian canons in the Gospels like that of Sinai, nor (together with these last) the larger chapters with their appropriate headings, which all other documents exhibit, that date from the fifth century downwards. The sacred

1 We would say once for all that the limited resources of typography hinder us from nicely representing in ordinary characters the varying sizes both of the capitals and of those smaller letters which Col. Bezae, like most others of real antiquity, often crowds into the end of a line. We have done our best, but we are conscious that in many places others would have put capitals or smaller letters where we have not, and vice versa.

2 Even in Cod. Bezae the two points sometimes run almost into one line: e.g. Fol. 456 6 1. 16.

3 The apostrophus after elision seen in our Facsimile, I, 1. 22, ovd' is too faint to be by the first scribe, and we have rejected it with all in several cases. In two places

(81 6. 17 X with ov erased, 302 b. 1 x' for k p. m.) it is plainly the work of the oldest corrector, whom we shall hereafter describe as A. I know not what to make of it in 193 a. 4, 194 6. 2, which look p. m.

b 2, 4 By the corrector we shall call B in 15 6. 15, ñ ultim.; 137 b. 6, eva; 149 6. 8, wv secund. ; 161 b. 23, ; 164 6. 25, în secund. ; 166 6. 20, év; 214 6. 19, óv; 249 6. 19,


, -λώσ; 266 6. 27, εισαγγελοι; 279 6. Ι4, εισ; 320 6. 28, εν; 421 b. 25, ádov; 425 6. 10, óv; 457 6. 16, øv. But 489 b. 26, OUTWO, 500 b. 16, õuws seem later, and G (to be described hereafter) in Acts xvi. 19, 20 has erilaßbuevot... παύλον... αγοράν... άρχοντασ... στρατηγοίσ... εκταράσσουσιν ημών την πόλιν.

[ocr errors]

text in both languages is broken up into certain paragraphs of very unequal length, such as we meet with no where else, each commencing in a letter (sometimes but not usually larger than the rest) which encroaches a little on the margin, and is clearly indicated throughout our edition (see also Facsimile Plates I and II, 1. 2): Bp. Marsh counts 153 (I reckon but 148 both in the Greek and Latin) such paragraphs in S. Mark's Gospel'. The Ammonian sections, however, without the Eusebian canons”, are inserted in the side margin of Codex Bezae by a scribe włom we shall hereafter (infra p. xxvii) shew to have lived several centuries later than this manuscript was written, and the beginning of each section is indicated by double points interpolated in the body of the text before or (if space be wanting) over its first letter, very often in both places, referring to similar double points placed in the margin after the Ammonian numerals? The single point by the first hand

?. precedes these latter double points in Fol. 92 b 11. 6, 10, 224 6 1. 16, and follows them Fol. 311 6 20. While banishing to our Adnotationes all other matter found in the margin (infra pp. 448—452), even that written by the same hand as the sections, convenience or necessity have led us to retain these and the corresponding points, which we must once for all request the reader to regard as much more recent than any other writing on the page. The variations in place and number of these sections from these commonly found in manuscripts and editions are recorded in our work (infra p. 453), as they are somewhat curious and interesting; it is also remarkable that as in Codex Sinaiticus (where they cease in S. Luke after ch. ix. 57), they are not carried on complete throughout the Gospels. This arises at the end of S. Matthew (where the last section noted is tvß, ch. xxvii. 62) and the beginning of S. John (i. 1-16) by reason of the side margins of Foll. 102—4 being cut clean away; but this explanation does not hold for the end of S. John (xx. 13—xxi. 25), or of S.

; Mark, where the sections terminate at o, ch. xv. 16.

More recent than the Ammonian numerals and points are the nine leaves supplied in their proper places as Codex Bezae is now bound, but in this edition exiled to an Appendix (pp. 417-—428), and printed in smaller type than the rest of the volume. The first of these additional leaves contains Matth. ii. 21-iii. 7 Latin, jii. 7-16 Greek: the next seven have crowded into them the contents of the eight lost leaves which originally formed quaternion 22 (supra p. xiv), viz. John xviii. 2xx. 1 Latin, xviii. 14—xx. 13 Greek: the ninth leaf has on the first page Mark xvi. 6-15 Latin, on its second page in parallel columns the Greek and Latin of Mark xvi. 15—20 and (the Greek in blue ink) the subscriptions proper to each. All these Latin pages are transcribed from copies of the Vulgate which resembled the Clementine printed edition more closely than do Cod. Amiatinus and the best manuscripts (e. g. John xviii. 12; 16; 19; 31; 36 ter; xix. 6; 16; 24; 28; 36), and


1 In S. Luke I count 136 (143 Lat.); in what remains of S. Matthew 583 (590 Lat.), of S. John 165 (168 Lat.), of the Acts 235. Capricious and irregular as these paragraphs may be, Mr Hansell did right in retaining them in his Texts of the oldest existing manuscripts of the N. T. (Oxford, 1864). It is worth notice that, as in Codd. Siwaiticus and Vaticanus, each Beatitude in Matth. v. forms a separate paragraph.

2 These numerals are so set in some places (e.g. Foll. 34, 91, 92, 127, 163, 335, 339 V) as to leave no room for the Eusebian canons to be placed under them, so that the latter could not have been designed to be subsequently added. Many other copies have the sections without the canons which we might have deemed essential to their completeness: e.g. Evan. Codd. CFHIPQR W Y Z. 5+. 59. 60. 68. 440. iscr. gser., though sscr. contains Eusebius' Epistle to Carpian. What use the Ammonian sec

tions can serve, unless in connection with canons of harmony, those who have studied them most can the least tell. In the uncial fragment of S. Mark [ixth century], discovered by Mr Bradshaw in 1862 (Trin. Coll. Cant. B. VIII. 5), and by him named Wa, the Ammonian sections are placed alone in the margin, and a kind of table of the parallel passages in the other Gospels set at the foot of each column. A similar arrangement is found in Evan. E, which contains the canons besides; and Tischendorf has just informed me that it also appears in the six leaves of S. John, of the sixth century much resembling in style Cod. T of the Gospels), now at S. Petersburg, and by him named Tb.

3 These double points are occasionally misplaced, e. g. 194 6. 1. 18; or even put in the Latin as well as in the Greek, e.g. 93 a. 1. 6.

[ocr errors]
[ocr errors]


« PoprzedniaDalej »