xv. 26; кρτησ кaι apaẞoι Acts ii. 11; TOUσ aкovovтeo ibid. v. 11; aкovσavтEσ accus. xvii. 8. To the same cause may perhaps be referred the dialysis over the Latin i in iohanne Fol. 57 a 1. 12; üae 82 a 1. 24; iam 427 a 1. 26, which may be observed once in Cod. Laud. 35 (Act. E) igitur ch. vi. 3.

We can say little about Scholz's statement, that the native country of this manuscript may be regarded as the South of France, by reason of its resemblance in style of writing to the uncial Lectionary of the Gospels Evst. 60, whose subscription shews that it was copied A.D. 1022 "in castro de Colonia" for the monastery of S. Denys (Scholz, N. T. Proleg. I. pp. xl; ciii). It is hard to conceive that there can be any striking likeness between codices which differ in age by full 500 years, though there is certainly some affinity between their respective texts; yet the very fact that a Greek Lectionary should be written for a French convent in the eleventh century adds one more link to the chain of evidence that the Churches of Gaul and Asia maintained for many ages the intercourse commenced by Pothinus and his missionaries about A.D. 170, and makes it quite credible that Oriental proper lessons, as well as the Oriental liturgy (Palmer, Origin. Liturg. 1. p. 153, 2nd edit.), were long used in some of the monasteries of those regions: the liturgy, as we know, survived till Pepin's time. Hence we need not transfer our manuscript to Greece in order to account for the liturgical notes scattered throughout its margin in the course of the eighth and three following centuries, or refer them with Beza to the work "indocti cujusdam Graeci Calogeri" (see his Letter, supra p. vi). The very orthography of these notices savours of a Celtic origin (see p. xxvii, note 1); and the only three Saints' Days whose proper lessons are marked are just such as would be specially regarded in the West at their respective dates, viz. the Assumption of the Blessed Virgin (Aug. 15) by a hand of the tenth century (M,, Fol. 229 6), the Festivals of S. George and S. Dionysius the Areopagite, the patron saints of England and France, inserted as late as the twelfth (by O, Foll. 462 b; 488 b)'. Thus all outward appearances point to Gaul as the native country of Codex Bezae (we shall notice internal indications. of the same fact in Chapter III), nor is there any valid reason for thinking that it ever left that country till it was carried into Italy in 1546 (see p. viii).

CHAPTER III.

ON THE LATIN VERSION IN CODEX BEZAE.

THE origin and character of the Latin version standing parallel to the Greek text of Codex Bezae, and known to critics by the name of d, involves questions of considerable difficulty, which have given rise to much discussion. In the present chapter an attempt will be made to prove (1) that it is, on the whole, an independent translation made either directly from the Greek on the opposite page, or from a text almost identical with it; (2) that the translator often retained in his memory, and perhaps occasionally consulted, both the Old Latin version and Jerome's revised Vulgate; (3) that he probably executed his work in Gaul about the close of the fifth century.

I. On the first point, we regret to find our judgement at variance with the decision, or

1 Credner (Beiträge, I. p. 514) who was necessarily dependent on Kipling's edition, is ignorant of the first of these Festivals, which my predecessor passed over in silence, and I presume was unable to read: Mr Bradshaw and I had quite enough to do to decypher the writing between us, though when once made out the sense is unquestionable. The other two are written in the same large, clear, wretched Western hand (wherein v is , and e is e) and are well suited to what we can learn of

the times. The Areopagite had been fully established in the estimation he long held as early as the ninth century, when his works were translated by Hilduin for Louis the Meek; while the glory of S. George (who had been known in Gaul even in the sixth century) was spread throughout the West by Robert of Normandy and his followers on their return from the first Crusade at the close of the eleventh century.

at least with an obiter dictum of Tischendorf. While seeking to account for the absence of the larger chapters (kepáλaia majora) from Cod. Bezae, although they are found in copies of as early or an earlier date, he says "excipiendus tamen est D codex Graecus et Latinus Cantabrigiensis, quod inde explicandum videtur, quia Italus ejusdem codicis textus, a quo ipsum Graecum pependisse certum est, nihil ejusmodi praebuit" (Cod. Sinait. Proleg. p. xxxii, not. 2, edit. min.). By "Italus ejusdem codicis textus" our version d must needs be meant, yet we hold it certain that so far from the Greek text being dependent on or derived from it, the Latin version is little better than a close and often servile rendering of the actually existing Greek.

(a) For how else shall we account for the frequent insertion in the Latin of purely Greek words which no other known version ever employed, and for which there are adequate equivalents in Latin? Such are ana Luke ix. 3 only; promeletantes ibid. xxi. 14; aporia ibid. v. 25; machaerae ibid. xxii. 38, but cladius v. 36, gladius in 17 other places; aporior ibid. xxiv. 4; echo Acts ii. 2; allophylus ibid. x. 28; allophoelus ibid. xiii. 19; adynatus ibid. xiv. 8; anetius ibid. xvi. 37; spermologus ibid. xvii. 18; ellada ibid. xx. 2; eremum ibid. xxi. 38; so John vi. 31 only'. On the other hand a few mere technical terms, such as a remote provincial might have heard used, are employed happily enough by d alone of the Latins; e. g. optio carceris (deoμopulaέ) Acts xvi. 23; 27; 36; sestertia docenta (uvpidao TEVTE) ibid. xix. 19, a fair computation. Professio Luke ii. 1-5; Acts v. 37 and decurio (Bovλevrno: cf. Plin. Ep. i. 19) Luke xxiii. 50; Mark xv. 43 are also in the Vulgate. We should notice also one or two ritualistic terms peculiar to d among the versions, which may possibly suggest a somewhat later date than can be assigned to the rest: John xvi. 2 λaтрetav рoopeрew hostiam offerre; Matth. xxvii. 62 Tараσкevn cena pura (but in Mark xv. 42, where alone it occurs besides in Cod. Bezae, parasceue with the Vulgate), although this latter expression was used by Tertullian and Augustine, and has been already explained by Mill (N. T. Proleg. § 1281) after Ducauge: cena pura, however, is found in several of the earlier Latin codices in John xix. 14 (e); 31 (a. b. e); 42 (e); Luke xxiii. 54 (a. b. c. e. ff'1).

(B) Violations of the rules of Latin syntax occur not unfrequently in all extant modifications of the primitive Latin version of the N. T., and are now held to demonstrate the African origin of that venerable work: some of the manuscripts contain them more plentifully than others; none, whether in respect to number or barbarism, to the extent of our d, which indeed is quite unique in its tone and the general current of its diction: nothing is found elsewhere so gross and palpably ungrammatical as many of the following instances, which we commend to the reader's careful examination. The Greek construction of a genitive absolute occurs Luke iii. 15 semel; ix. 43; xix. 11; xxi. 5; 26; 28; xxiv. 31; 36; and not elsewhere; as if the inexperienced translator had been trying an experiment which he saw cause to discontinue: just as o de K.T.A. is rendered qui autem &c. in Matth. ii. 14 and in 53 other passages in that Gospel, but afterwards only in Acts xii. 15; ad (i. e. at) ille, ille vero, ipse vero (Acts xii. 16) being used instead. Notice also the neuter plural noun with a singular verb in Matth. xiii. 26; xviii. 12; Luke xix. 42?; Mark iv. 11: the double negative increasing the strength of the negation, Matth. xxii. 16; John vi. 39; viii. 33; ix. 33; xi. 50; xiv. 30; xvi. 23; Luke xx. 40; xxii. 34; Mark xiv. 60: the genitive is used after a comparative as in Greek (even the Vulgate has it in Acts xvii. 11), Matth. xii. 41; (not v. 42); John v. 20; vii. 31; x. 29; xiii. 16 semel; xiv. 12; xv. 13; Luke vii. 28 semel; Mark xii. 31; the examples being found here again chiefly in the same Gospel: the same remark applies to the following cases

1 I should have added to this list basiliscus John iv. 46, but that the word is also found in Cod. Vercellensis (a).

of Attic attraction, as it is called, John vii. 31; Luke i. 1; ii. 20; iii. 19; xiii. 17; xix. 37; xxiv. 25; Acts i. 2 (but not in the Greek). Verbs also govern other cases than the rules of Latin syntax demand, and that too in accordance with Greek: the genitive in Matth. ix. 21; x. 31; xx. 25 bis; xxii. 10?; John vii. 40; ix. 29; 31 semel; x. 3; 8; 27; xii. 47; Luke i 53?; 54; xx. 20; 35; xxii. 25; 35?; xxiii. 15?; Mark x. 42; or the dative (all except those in S. Luke with adoro pоσKUVE∞) Matth. ii. 2; 8; 11; ix. 18; xiv. 33; xv. 25; xxviii. 9; John iv. 21; 23 semel; Luke i. 3; xiv. 33; xviii. 13'; Acts vii. 43: add te nocui (adiko σe) Matth. xx. 13 (cf. Mark xvi. 18 in am. fuld., manuscripts of the Vulgate); maledixerunt illum John ix. 28; maledicentes vos [but thus also Cod. Palatin. (e)] Luke vi. 28 (see the Greek); benedixit eos Luke xxiv. 50 (so v. 51) with Augustine in loco; and such a form as puelles Acts xvi. 19. The same inference may be drawn from the varied and luckless shifts made by the Latin scribe to render that great stumblingblock to translators, the Greek article. He begins the attempt with hic in S. John (viii. 26; x. 36; xiv. 19; 22; 30; xvi. 21; 28 bis; xvii. 6; 9; 11 semel; 21, all with o Koopoo), which he resumes in the Acts (iv. 8; 22; 37; xi. 22; xviii. 27; xix. 23; 35); so perhaps Luke xxiv. 17 ista verba haec o Xoyo OUTO: next he tries ille in John xi. 51; Luke xxiv. 9 (with other versions); Acts iii. 10; vii. 43; xii. 7; xx. 25; xxi. 38: or is in Acts iii. 25; xvi. 19: or (as Schulz has partly noticed) ipse in Acts iii. 14; v. 24; 26; vii. 8; 13; 17; 18; 35; 43; 48; 58; viii. 9; x. 16; 21; xi. 12; 23; xiii. 23; xvi. 25; xix. 30; xx. 24: unus seems to be a later expedient. The servility and awkwardness of the translator is especially manifest when he mixes up the constructions of the two languages, thus producing what is neither Greek nor Latin, e. g. Matth. xxiv. 24; John xii. 37; Luke i. 79; xii. 1; 4; 5; xx. 46; xxii. 55; xxiv. 14; 27; 41; Mark ii. 8; v. 4; ix. 14; 42; x. 32; xv. 29; 33. This is most observable in the Acts, whose text, both Greek and Latin, is in so unsatisfactory a state: e. g. iii. 13; 24; vi. 4; viii. 12; x. 25; xi. 1; xii. 20; 25; xiii. 1; 28; xv. 20; 22; xvi. 4; 16; xix. 25; 29; 40; xx. 12. The study of a few of these examples will suffice to shew that they are but halting renderings of the Greek.

(y) More conclusive still are those many instances wherein the Latin has an erroneous rendering which could not have originated in that language, but is plainly derived from following some other Greek reading than that now found in the manuscript, or some false reading of the existing Greek which could not have sprung up in the Latin, or else from a mere misapprehension of the sense of the Greek. Such are Matth. xxii. 40 vopoo verbum (the scribe misread Aoyoo); ibid. xxvi. 6 Tov inv yevoμevov ihu facto"; Luke i. 9 Ovuarai sacrificare (as if

1 Miserere mihi, a doubtful instance, for in this version misereor several times takes a dative; in Matth. xviii. 33 we find both a genitive and dative; a dative in Luke xvi. 24; xvii. 13; xviii. 39; Mark x. 47: a genitive in Luke xviii. 38; Mark x. 48; and six other places.

2 This attempt to make factus answer all the purposes of yevoμevoo disfigures no less than 42 places in the version. In the Acts indeed d has learnt to use a little more licence at times, rendering eyeweтo by nascebatur ii. 43; eyeveтo by respondit xi. 9; by contigit xiv. 1; yevoevoo by conversus xii. 11. In the Acts also, as was quite necessary, much greater freedom is used in translating the particles: thus Te is etiam ii. 44; quoque ii. 46; x. 22; vero iv. 27; xiii. 46; que iii. 10; v. 14; vi. 12 &c.: de is not only autem as usual (vii. 1 &c.), but quoque viii. 13; X. 24; que v. 14; vii. 32; xix. 3; itaque viii. I; vero v. 16; 22; 24; X. 24 &c.: μev ovv quidem xv. 30; ergo xvii.

νομοσ

12; quae xvii. 14; itaquae xvii. 30; ut vero cum xix. 9: yap praeterea i. 15 as well as enim, so constant elsewhere: av is generally made by utique (Matth. xxiv. 43; Luke xii. 39, &c.), butin Acts xvii. 18; 20 by nunc, inxviii. 14 by forsitam. These are stronger reasons than any assigned by Kipling (Cod. Bezae, Praef. p. xii) for supposing that a new hand was employed in the Acts because avaiтioo is made by anetius Acts xvi. 37, but by innocens Matth. xii. 7 (he might have added by sine culpa two verses previously): aduvaтoo by inpossibilis Matth. xix. 26; Luke xviii. 27; Mark x. 27, by adynatus Acts xiv. 8: doğağw by clarifico Acts iii 13; iv. 21; xi. 18; xxi. 20, but no where else. Yet in regard to dosage we meet with just the same variation in the Gospels. In S. Matthew it is glorifico four times, never in S. Luke, but honorifico five times, honoro three times, in the passive gloriam accipio iv. 15: in S. Mark we have honorifico once: in S. John glorifico fourteen times, honori

it were Ovora); xxiii. 12 andia lite; ibid. v. 40 ev тw aνтw крiμaтi in ipso judicio; Mark xii. 38 των τελωνων qui volunt (he translates των θελοντων); xiv. 51 συνδονα επιγυμνου (for επι γ.) sindone nuditatis; Acts vii. 19 катаσоpioаμеvоσ cum justitias coepisset; ibid. xiii. 18 erроT ac si nutrix aluit (he read erpopop., as Mill saw plainly at first, less clearly when he wrote his Appendix to the Ν.Τ.); xiv. 9 υπαρχων εν φοβω possidens in timore; ibid. v. 15 ομοιοιπαηθεισ εσμεν ϋμειν ανθρωποι patientes sumus vobis hominibus (as if he read ανθρωποισ); ibid. v. 16 ειασε sanavit (as if from ιαομαι); xvi. 33 ελυσεν απο των πληγων solvit plagas (ελυσεν for ελουσεν is natural enough, solvit for lavit far less so); xviii. 18 poσevxnv (for evxnv) orationem; xxi. 21 μητε εν τοις εθνεσιν (for εθεσιν) αυτου περιπατειν neque gentes ejus ambulant, in mere blind perplexity. The student may find hundreds of these instances, just as convincing as any we have given.

(8) The same inference, so far as regards the fact of the independent origin of the Greek text, at least of the Latin which stands on the opposite page in Codex Bezae, may be drawn from those places where the present Latin differs from the Greek in respect to a variation which could have arisen only in the Latin. This process, applied by Wetstein to the case of the Velesian readings (N. T. Proleg. Vol. I. p. 60) has settled the question as to their history and value. Examples of this kind might be adduced from d without limit, but a few of real moment are as good as a thousand. Such are John xii. 43 nyanσav dixerunt (i. e. dilexerunt, cf. xiii. 23; 34); ibid. xiv. 26 Tournoa commovebit (i. e. commonebit); Acts v. 17 nov aepulationem (i.e. aemulationem); ibid. vi. 1 TapelewpovvTo discupiuntur (i. e. dispiciuntur); vii. 28, xvi. 35 ex eo εχθεσ externa (i.e. hesterna) die; vii. 32 eroλpa audiebat (i. e. audebat); ibid. v. 46 evpe referit (i. e. reperit); xii. 20 diα TO Tрepeobai propter ne alienarentur (i. e. propter quod alerentur); xii. 21 conueryope contentionabatur (i. e. concionabatur); xvi. 35, 38 paßdovxovo lectores (i.e. lictores: so fuld. of the Vulgate v. 35); xvii. 4 yuvalkεσ тWV прwтwν mulieres quae morum (i. e. primorum); xx. 9 KаTO zosum (i.e. sursum); ibid. v. 27 Bovλnv volumptatem (i.e. voluntatem).

κατω

II. We shall best investigate the next branch of this discussion-the relation which the Latin version of Codex Bezae bears to the old version of the Western Church, and to the Vulgate revision of it executed by S. Jerome if we select a passage of some length, extant in all the principal manuscripts of the Old Latin, rich in peculiar and idiomatic expressions, and little liable to be corrupted from the synoptic Gospels (Luke xxiv. 1-24); wherein we may compare the translation found in our Codex (d of the critical editions) with that of the Vercelli (a), Verona (b),

fico six. This precarious argument drawn from the use of different words in the several parts of the same work weighs far too much with some critics, and is peculiarly inapplicable in the case of a writer who is apt to change his expression in the self-same verse; e.g. Matth. xix. 12; John xii. 12 (gender of dies); xvii. 1; xx. 19; Mark vii. 15; ix. 37; Acts xx. 13.

1 The following various readings also, the Latin version being quite correct, could only have originated in the Greek: e. g. Matth. xi. 3 epyasoμevoo for eрxoμevoσ (renis, d); John xiv. 21 evpwvnow for eupaviow (ostendam, d); Luke ii. 13 alтovvтwv for avovvTwv (laudantes, d); ibid. vi. 20 ET apaσ for eTapao (TI for II: elevans, d); xiv. 26 πεισει for μεισει (odit, d); Acts iv. 29 αγιασ for amiλao (Fl for III; minacias, d); ibid. v. 4 peσov for μevov (manens, d). Credner (Beiträge, I. p. 463), besides several of the above examples, also cites John v. 39 aμap

τανούσαι for αι μαρτυρούσαι, but this was amended primd manu, as indeed were all except Luke vi. 20 by later hands. We demonstrated above (p. xxiii), by noting certain errors of the transcriber, that the Latin as it now stands was taken from another Latin copy, similarly divided in respect to orixo: we have now further and independent proof that the version (d) was not made direct from the actually existing Greek (D), but from some earlier text, almost though not quite the same, in which such variations as τελωνων, εργαζομενοσ, ενφωνησω &c. had not as yet sprung up.

2 Mr Field has employed this method for the happy elu cidation of that old puzzle in Barnabas c. 3, where the ancient Latin version has "sicut dicit filius Dei resistamus omni iniquitati et odio habeamus:" a saying no where else imputed to the Lord. The Greek in Cod. Sinaiticus. stands ωσ πρεπει υιοισ θυ κ. τ. λ.: sicut decet filiis Dei.

Colbert (c), Palatine (e), and Brixia (f), copies or modifications of the elder Latin; and with the common printed or Clementine Vulgate as amended by collation with its three best manuscripts, Codd. Amiatinus (am.), Fuldensis (fuld.) and Forojuliensis (for.)'. Single verses may readily be found which might serve to shew either that d is completely independent of all other known translations and made exclusively from the Greek on the opposite page; or, on the contrary, that it is a mere modification of the Old Latin, differing no further from other copies of it than e (for example) does from f. The careful study of d in many such long passages as that here subjoined leads us to believe that neither of these views presents us with the whole truth. The Latin of Cod. D was really constructed immediately from its Greek text, servilely following it (as we have just seen) to the violation of the simplest rules of Latin syntax, and thus contains much, both in respect to words and phrases, that is quite peculiar to itself: while on the other hand, inasmuch as it was the work of a Western scribe on whose memory the diction of his native version was firmly imprinted, like that of King James' Bible is on our own, the translator unconsciously and habitually imitated it, sometimes for whole verses together, even in places where the Greek original might have taught him to render otherwise. The parallel columns containing the several versions cover pp. xxxvi, xxxvii.

The general independence and occasional conformity of d appear equally clear throughout these verses. As our attention is directed at present solely to the diction of the several translations, we only note in passing the remarkable identity in reading between Codex Bezae (Dd) and c in v. 1 (ελογίζοντο δε εν εαυταισ τισ αρα αποκυλίσει τον λιθον, where yet the Latin words of c and d are as far apart as they can be), in v. 5 (ot de eiTav) and elsewhere (e. g. vv. 6, 7); as also between Dd and e in v. 24 (eudoper): Dd is opposed to all the Latins in v. 5 Ta πроσшжα, v. 10 in omitting nv de. δε. Our d will be seen to stand quite alone in v. 3 introeuntes; v. 4 aporiarentur (one of the barbarous Greek words catalogued above) and amictu scoruscanti; v. 5 timore factae (here again from the Greek, yevoueva); v. 6 mementole; v. 11 paruerunt in conspectu and derisus; v. 13 abeuntes and iter habentis; v. 14 horum, in heedless misapprehension of the final τούτων ; 16 ut non for ne; V. v. 18 advena and nescisti (to which a comes nearest); v. 20 potentes; v. 21 incipiebat (ueλAwv, a word similarly rendered by d in Matth. ii. 13, and 22 other places, by coeperat Acts xvi. 27, by volente Acts xx. 4, by habeo Luke x. 1; xix. 4); v. 22 seduxerunt and matutinae (mane in a); v. 24 de his qui erant nobiscum and sic sicut (ovTwo wo): most of these instances being highly characteristic. Just as visible, however, is the resemblance between d and some or all of its fellows: in v. 23 it approaches very close to a, and has with it vivum v. 5; stadios v. 13: in v. 9 Toto evdeкa is rendered illis undecim by all except c, with which alone d has reliquis v. 9: in v. 13 Kwμηv is castellum, in v. 15 opeλew is fabulari (neither of them very happy translations) in all but a, as d renders opeλnoao in Acts xx. 11 and with bee has fabulabantur here in v. 14, though f and the Vulgate rightly change it into loquebantur: in v. 16 all save e employ tenebantur (detinebantur, e). Above all, there occurs sometimes for verses together (e. g. vv. 7, 8) such a similarity in the tone and rhythm of the sentences as cannot be deemed accidental; yet may perhaps be sufficiently accounted for on the part of the scribe called d by imperfect recollections of the primitive Latin version still fondly cherished in his mind.

1 All necessary information respecting these MSS. will be found in Scrivener's Plain Introd. pp. 256-269, and in Mr Westcott's noble article on the Latin versions in Smith's Dictionary of the Bible.

2 The version in e is often very rude and free, and its

κωμην

various readings full of interest: e. g. in v. 17 et steterunt tristes, with NA p. m. B. Besides v. 24 Tischendorf (Cod. Palat. Proleg. p. xviii) cites John x. 3; xvii. II; Luke xxiii. 29?; xxiv. 49 as passages where D and e agree against all others.