2

where Cod. D ends, he continues his text without notice to the end of that verse, in order to fill up the page. As all the leaves now missing were lost in his time, except the lower part of Fol. 504, which we have taken from the Whitgift MS., it can be regarded as nothing more than a curious and rather unfavourable specimen of the scholarship and taste of the Elizabethan age.

(2) Patrick Young, the librarian of James I. and Charles I, who first collated the Codex Alexandrinus (A), and published from it the Epistles of Clement of Rome in 1633, seems to have been the next person engaged on Codex D, extracts from which he sent to the brothers Dupuy, through whom they reached Morinus and Steph. Curcellaeus. An unusually full collation of it was made for Walton's Polyglott (Tom. VI. Num. XVI. 1657) by pious Archbishop Ussher, who devoted to these studies the doleful leisure of his latter years. I am grieved that truth compels me to state that I never examined a performance more inaccurate than this. Besides numberless omissions, manifest typographical errors, a looseness and carelessness of citation which is really remarkable, and almost complete inability to distinguish the first from the later hands', its actual misstatements are so many, that I have accumulated a catalogue of 228, with which it is needless to trouble the reader. Some of these are wholly unaccountable, others arise from blindly following Stephens, not a few through confounding the readings of some other copy he had examined with those of Cod. D. Thus Cant. (as he terms it) stands for Gon. (Evan. 59) in Mark xii. 43; xiv. 8; xvi. 9; for Mont. (Evan. 61) Mark xii. 43; 44; Acts i. 20; iv. 8; viii. 12; xv. 7; xvi. 34 &c.; in the Acts sometimes for Lin. (Act. 33).

(3) Largeness of view, critical sagacity, wide and life-long research, comprehend Mill's claims on our gratitude for his great services to textual criticism: those who award to him the humbler praise of an accurate collator can have used his edition of the N. T. (Oxon, 1707) but little. His volume as at first printed, while it produces many new readings from Codex D, retains nearly all the errors of the Polyglott; and though most of these were amended in his Appendix, drawn up many years later as the result of a fresh examination of the manuscript, and though he bestowed (not always with success) some pains and skill on discriminating the changes made by later hands, yet his representation of its contents is too defective and mistaken to be used with any degree of confidence. Not a few additional blunders were made even in the secundae curae of his Appendix: e. g. Mark viii. 29; Acts x. 39; xvi. 35; xxi. 1.

Z. C. von Uffenbach, known as the owner of several manuscripts of the New Testament (Paul. M. 52; Evan. 97. 101), when on a visit to England in 1710, inspected Codex Bezae at Cambridge, and gives a brief but correct account of it in his Reisen durch Reidersachsen Holland und Engelland, T. III. p. 21; Ulm, 17543.

(4) Very superior in character to Mill's collation was the transcript of Cod. D made by J. J. Wetstein in 1716, at the age of twenty-three, when fresh in eyesight as in spirit; yet since he did not use it for the next thirty years, and never appears to have consulted the manuscript afterwards, or much cared for the collations already published (whose heavy faults he would be well aware of), too many of the readings he alleges are even marvellously untrue (e. g. Matth. xi. 30 xpnoтov; xxi. 8 ordo; v. 17+ews before es; Mark xv. 33 i.e. s' is clearly primâ manu; Luke iii. 33 om. TоU papes';

1 Thus he describes the two lines written in sloping uncials of the VIIIth or IXth century at the foot of Fol. 160 b. (John xv. 3, 4) as "eodem ferè charactere."

2 By consulting the original papers of Ussher or his assistants, afterwards used by Mill (N. T. Proleg. § 1505), and now in the Library of Emmanuel College, containing the readings of Cod. D and three others (Evan. 59, 61, 62), which Mr Scott of Westminster School lately

examined for Dr Dobbin (Cod. Montfort. Introd. p. 21), it might be seen what portion of this plentiful crop of errors is due to the printer. Our single purpose is to caution the reader against trusting to the collation as it stands in the Polyglott.

3 My acquaintance with this book, and many other literary favours, I owe to the Rev. J. E. B. Mayor, Principal Librarian of the University of Cambridge.

John iv. 6 yn; xi. 33 dtovs; xxi. 3 ovveẞnoav, K.T. A.), especially in the leaves written by a later hand. In many places, readings washed out or otherwise changed by the original scribe, and utterly overlooked by Kipling, had already been diligently recorded by Wetstein (e. g. Matth. xxi. 1 ẞn@paye; 22 αιτητε; xxvi. 40 αυτουσ; Jchn v. 19 ποιει for ποιηση; xx. 28 και ο θσ for μου και Oo): although his collation on this an more prominent points, copious as it is, by no means exhausts the subject, and his judgement is often wrong in assigning to later times alterations which were really made by the first penman.

(5) Richard Bentley, by special indulgence of the University, is said to have had Codex Bezae in his possession at Trinity College Lodge for seven years, while preparing his promised edition of the N. T. If all the use he made of it be represented in his papers published in 1862 by Mr. A. A. Ellis (Bentleii Critica Sacra, pp. 2-26), a single week might have sufficed for his purpose. The readings he gives from Cod. D are few and vague and inexact enough, but no one who has examined his collations of the Codex Augiensis and of the Arundel Lectionary 547, still preserved (Trin. Coll. B. XVII. 8), would expect much in this way from our great Aristarchus. We could have well spared some numerals, &c. set by Bentley in the margin of Cod. D, with the less excuse, inasmuch as it was not, like Cod. Augiensis, his own property.

(6) An unpublished collation made about 1732 or 1733 by John Dickinson of S. John's College, for John Jackson of Leicester, for six pounds sterling, now, with Jackson's other books, in the Library of Jesus College, Cambridge (0 0 2), has enabled us, after Kipling, partially to supply the hiatus in the Latin of Fol. 504 a., and has been consulted with profit in other passages. It is based upon, and aims at supplying and correcting, Mill's very poor representation of Cod. D, and Dickinson has taken laudable care to note the original text, as distinguished from its state as at present existing.

(7) Kipling (Praef. p. xxvi) names three other transcripts: one owned by Richard Simon, the Biblical critic; another sent to Sabatier by Bentley; a third written in 1766 by Thomas Craster, B.A. of S. John's College, for Kennicott, and sent by the latter to J. S. Semler, who published from it the Latin version of S. John in 1771, at the end of his "Paraphrasis Eran. Joann. cum notis." Griesbach also consulted the manuscript, but is said to have taken from it only one reading, and that false, επήγειραν Acts xiv. 2.

(8) In 1793 Dr Thomas Kipling (Senior Wrangler, 1768), afterwards Dean of Peterborough up to his death in 1822, published at the expense of the University of Cambridge, in two splendid folio volumes, his edition of what he was pleased to term " Codex Theodori Bezae Cantabrigiensis," the fruit of five years of toil. The moveable type used for this work in shape resembles as closely as possible the characters of the scribe, and some attempt is made to indicate the varying sizes of the letters, and the relative spaces between them. The present work was undertaken in the hope of producing an edition of Codex Bezae which should be more conveniently read than in Kipling's uncial letters, printed continuously; and be more easily accessible to students than his scarce and costly folios, of which only 250 copies were struck off. In prosecuting my interesting and not unpleasing task I have found the text of my predecessor less inaccurate than some have suspected: the typographical errors detected (83, of which 16 are in his notes &c.), I have recorded (see pp. 452-3) as a matter of duty, not of reproach :-perfect correctness is quite unattainable, yet Kipling has laboured faithfully and not wholly in vain to approach it as near as may be. His most serious fault is one of design and plan, in that he has

1 Happily lettered "MS. Sermons." Mr C. H. Cooper, senior editor of the Athenae Cantabrigienses, kindly informs me that John Dickinson of Sheffield became B.A.

1728-9, M.A. 1732, Assistant Minister of Sheffield 1752-66. This humble and forgotten man must have been a good and early scholar.

placed in the body of his work those numerous changes made by later hands (some of them indeed of very recent date'), which deform the pages of Codex Bezae itself, but which its editor should have been glad to banish into the Notes: nor has he much availed himself of the researches of those who went before him. Respecting his Preface it is enough to say that even seventy years ago it was obviously behind its age, both in respect to its general tone and spirit, and to the then existing standard of critical knowledge'.

I had also prepared full lists of the errors found as well in Ussher's, as in Mill's, Wetstein's, and Bentley's collations of Codex Bezae, but would fain be spared the ungracious office of publishing them. It will probably suffice to say that in whatever place any one or all of them may differ from this edition, I have ascertained by actual comparison with the manuscript, that my reading is the true one.

POSTSCRIPT. After this Chapter was in type Mr E. H. Hansell, Praelector of Theology in Magdalen College, Oxford, favoured me with the copy of a letter written by William Camden the great antiquarian [d. 1623] to Pierre Dupuy, and found by the Revd. J. Stevenson of the Record Office among the Dupuy MSS. (490, fol. 95) at Paris. Camden's account of Codex Bezae, though utterly mistaken, is too curious to be passed over, if only as a specimen of the untrustworthy character of so many of our most promising materials for literary history.

"Amplissimo viro Petro Puteano Guil. Camdenus S. P. Serius rescribo, V. C., quod serius tuae per Woodfordum mihi redditae, et post aliquot dies quam meae ad te per Porium [sc. John Pory? Calendar of State Papers 1611-1618] dederim. D. Thorii Epicedium ad manus pervenisse spero. Torvallius [sc. Jean l'Oiseau de Tourval? Calendar &c.] Gallicus regis nostri interpres, cui tradidit, ante mensem per Scotum nobilem se transmisisse affirmat. Gratias viro cordatissimo de honorificâ illâ mentione plurimum debeo. Patritio Junio [vid. supra p. xi] aurem vello, et subinde promissi moneo, quod facturum dicit quamprimum exemplar a Cantabrigiensibus nansisci potuit. Quodnam exemplar intelligit nescio; sin illud IIII Evangeliorum quod Theod. Beza in Angliam quondam misit, et ipse ante plures annos in Collegio Trinitatis Cantabrigiae vidi, frustra erit. Ibi enim jam non comparet, et quid de illo factum a sociis nullus dixerit; at dudum rem totam a reverendiss. episcopo Eliensi accepi. Transmisit Beza ad Cecilium, Angliae Thesaurarium, venerandum illud antiquitatis monumentum, ut in Angliâ asservaretur. Ille academiae Cantabrigiensis Cancellarius ad Trinitatis collegium misit, jussitque ut academici gratias agerent, quod factum. Post aliquantum temporis Beza, qui justum pretium potius quam gratias papyraceas (ut opinati sunt nostri) expectavit, nescio quo obtentu utendi in collatione repetiit; et remissum erat. Caeterum in Publica ibi Bibliotheca extat MS. exemplar Evangeliorum et Actorum Apostolorum Graecè et Latinè, sequioris aevi, charactere plane barbaro, et monast. S. Iraenei [sic] Lugduni anno MDLXII delatuin et a Matthaeo Parkero, archiepiscopo Cantuariensi academiae donatum. Hoc juvenes suspicantur esse illud Bezae, sed falluntur, et cum illis Th. Jamesius bibliothecarius Oxoniensis, in Egloga Oxonio-Cantabrigiensi; cujus exemplar, quia forte an non vidisti, cum Eliz. Annalibus, per hunc tabellionem accipies. Cumulato rependas, si Analecta Hibernica, quae ex Cramoysii officinâ nuper prodierunt, cum commodè poteris, mihi impertias (At jam non opus est; nactus sum exemplar). Ama nos et valeas V.C. Saluta mihi quam plurimum ornatiss. Hotomannum, cujus officiosae erga D. Beecherum amicitiae ex animo gratulor. Si initium legum Saxonum et Danorum, quae extant in libro cujusdam Archidiaconi in Bibliothecâ Thuani in 4°, mihi descripseris, longè gratissimum facies. Londini VI Xbris St. V." Since Camden's Annals of Elizabeth were published in 1615, and the Analecta de rebus catholicorum in Hibernia (without the name of the place) in 1616, this strange letter, written when the latter had just appeared, must be dated in the same year: it is easy enough to see why a copy of such a work was not readily procured in England. Thorius' Epicedium on Isaac Casaubon [d. 1614] was not published till 1619, after it had been passed for five years from hand to hand, as we partly gather from Camden's expression, and as was very usual in that age. Thus the Bishop of Ely, whose testimony he alleges, was the great and good Lancelot Andrewes, who held that see from 1609 to 1619. But the story, on whatever authority it rests, involves a simple impossibility. We know from his own letter that Beza had

1 E. g. T (vel T4) Zilea marg. 486 b. 1. He has even inserted by means of manual correction, "qui" foisted into the margin of 272 a. 1. 16, I am afraid in the hand of Bentley himself.

2 Kipling's edition was sharply handled by Porson in two notices in the British Critic, Vol. III. 1794, and coarsely abused by one Thomas Edwards, LL.D., who satisfies a certain private grudge in his Remarks on Dr Kipling's Preface to Codex Bezae, Part the First, 1793, in which I notice nothing worth mention save the poor taunt that Kipling's "zealous exertions have not hitherto been

rewarded with the smallest preferment;" as if his book were the worse for that. On the other hand, I read with surprise a statement extracted by Horne (Introduction, Vol. II. Appendix, p. 21, 1834) from a later volume of the British Critic (XI. p. 619, 1819), that "Porson himself collated the printed copy with the original manuscript, and the only fault he could detect was in a single letter of the margin." If this were true (and it is stated "to be well known to many of the Professor's friends"), it would only prove Porson to be a very bad collator.

the manuscript in his possession in 1581; Archbishop Matthew Parker died in 1575. The explanation of the whole matter seems to be that Camden had seen Whitgift's poor transcript of Codex Bezae at Trinity College at some time subsequent to its being placed there in 1604, and came to confound it years afterwards with the original in the University Library. If Bp. Andrewes imagined that Beza looked rather for money than for paper thanks, he did not know the man, whose stern hard nature loved power too intensely to be the slave of a meaner, though it may be a less mischievous passion. Although it seems needless to say more about so plain a matter, the reader may like to know that Lord Burleigh the Chancellor's letter which accompanied the manuscript to Cambridge, dated 9 May 1582, is published in Hartshorne's Book Rarities of the University of Cambridge, p. 13; and that in the Grace for lending the volume to Whitgift "quo illud describat," passed 2 March 1582-3 and enlarged Oct. 10 (Baker MSS. xxiv. 181), it is called "N. T. Graecum quod nuper venerandus Pater Theodorus Beza dono dedit Academie."

CHAPTER II.

ON THE PALAEOGRAPHICAL APPEARANCE OF CODEX BEZAE, ITS PROBABLE ORIGIN AND DATE. THIS invaluable manuscript forms a quarto volume, ten inches high by eight broad, whose margin, though still ample, has been cut down, at least in parts, by the binder. Its material is excellent vellum, perhaps not quite so fine and thin as that of the Codices Claromontanus (Paul. D), Vercellensis (Evan. a), and a few others, but for the most part in good condition; although some of the leaves are falling into holes, while in others the ink has much worn off, or has been washed away or read off on the opposite page through damp, especially on the rougher or outer side of the animal's skin. In some places the mischief has been aggravated by the application of a chemical mixture for the purpose of restoring the faded writing: but, on the whole, every alternate open leaf, as presenting the smooth or inner side of the vellum, is in fair preservation; some being as clear and fresh as if written yesterday. Assuming that Codex Bezae ended with the Acts of the Apostles, it must have originally consisted of 534 leaves, distributed into 67 quires or quaternions of four sheets or eight leaves each, only that the 34th was accidentally made up of only three sheets or six leaves, the innermost sheet of the four being left out: the numeral signatures of the quires, written prima manu, were set at the foot of the last page of each, but so low down that they were often cut away in part or wholly by the binder; we have carefully noted all that remains of them (see pp. 15; 80; 105 &c.). Of these 534 leaves there are lost the first two and seventh of the first quire; all the eight of the third, fourteenth, twenty-second, and fifty-seventh quires; all after the fifth leaf of the forty-fourth to the end of the fifty-second quire; the first and seventh leaves of the sixty-fourth quire; the whole of quires sixty-five to sixty-seven': thus after the loss of 128 leaves, only 406 survive (about twelve of them being more or less mutilated), besides nine added by a much later hand to supply some of the defects, whereof we may better speak hereafter. The manuscript once contained the four Gospels in their usual Western order (SS. Matthew, John, Luke, Mark), the Catholic Epistles and Acts of the Apostles, but on the missing leaves just enumerated we have lost in the Greek, Matth. i. 1-20; iii. 7-16; vi. 20-ix. 2; xxvii. 2-12; John i. 16-iii. 26; xviii. 14-xx. 13; Mark xvi. 15-20; Acts viii. 29-x. 14; xxi. 2-10; 16-18; xxii. 10-20; the text ends after avrov v. 29: in the Latin, Matth. i. 1-11; ii. 21-iii. 7; vi. 8-viii. 27; xxvi. 65—xxvii. 1; John i. 1-iii. 16; xviii. 2-xx. 1; Mark xvi. 6-20; Acts viii. 20-x. 4; xx. 31 -xxi. 2; 7-10; xxii. 2-10; the version ends after consentiens, v. 20. S. Luke's Gospel alone is complete: of the Catholic Epistles nothing remains in either language except twelve lines of the Latin of 3 John 11-152 on Fol. 415, that on which the Acts commence. On the other hand a small

I We calculate that three quires or 24 leaves would be required for the portion of the Acts which follows ch. xxii. 29, auTOU, from observing that it fills 24 pages in the Elzevir N. T. 1624, in which, throughout the former part of the Acts, where Cod. Bezae is full of interpolations, each page corresponds so closely to one leaf of Cod.

Bezae, that 94 pages of the printed book answer to 951 leaves of the manuscript.

2 Followed by the subscription epistulae iohanis III explicit &c., where epistulae seems to be the genitive, not the plural as Credner supposes (Beiträge, I. p. 456-7).

fragment of Fol. 96 which contained Matth. xxvi. 65-xxvii. 1 Latin, xxvii. 2-12 Greek, though overlooked by Kipling, is bound up between Foll. 89 and 90, and its contents will be found below in their proper place; we have also recovered from the previous collations nearly all the readings of the last ten or eleven lines of Fol. 504 (Acts xxi. 7-10 Latin, 16-18 Greek), which though evidently damaged when Whitgift's copy was made (see p. xi), were not cut off from the rest of the leaf till after Dickinson's time (see p. xii). Our only difficulty is with the Catholic Epistles, which could hardly have covered more than fifty of the missing sixty-six' leaves between the end of S. Mark and the beginning of the Acts, even though we suppose that S. Jude was inserted, as in some catalogues, otherwise than in the last place. Since the superfluous sixteen leaves would suffice neither for the Epistle to the Hebrews, nor for the Apocalypse, nor for any other book at all likely to occur in such a position, but would take up exactly two quires, we venture to suggest that the original penman may have miscounted his quires by two at some place in the portion that is lost; just as we know that one of two later scribes must have done in Cod. Sinaiticus, inasmuch as they differ by unity in numbering the quires from the commencement of S. Paul's Epistles (Quat. 82 or 81) down to the end of the manuscript.

Like its younger contemporary the Cod. Claromontanus (Paul. D), Codex Bezae has the Greek text in the post of honour on the left hand page of each open leaf, the Latin version on the right hand or second page: in this respect differing from the other bi-lingual copies, Codd. Laudianus (Act. E) and Augiensis (Paul. F), which exhibit the Latin on the left, the Greek on the right, in two parallel columns of the same page, as also does Cod. Sangermanensis (Paul. E) only with the Greek on the left. Both the Greek and Latin of Cod. Bezae are written in bold, regular, and elegant uncial characters with the words undivided, arranged not as most others, in lines containing nearly the same number of letters, but in verses or orixo determined by the sense, the Greek and Latin closely corresponding with each other. Every page exhibits thirty-three such verses or lines (kept regular by the usual means of a bodkin-acus-and a ruler), except at the end of a book, when the scribe breaks off to fill up the rest of each page with simple arabesque ornaments and a brief subscription, partly written and partly adorned with bright red colours, but in a style not more elaborate than is seen in corresponding parts of Codd. Sinaiticus and Alexandrinus. Such an open leaf at the end of S. John's Gospel (Foll. 181 b., 182 a.: infra p. 159). is represented in our first two facsimile pages (Plates I, II), to which the reader will please to make frequent reference as he examines our statements throughout the present chapter. In the titles and subscriptions of the several books the words appear separated, and a tendency to the same practice may be observed here and there in the body of the Latin version itself (e.g. Fol. 138 a.). The first three lines of each book whose beginning is extant, are in the rich red or vermillion paint we have just mentioned, still perfectly bright and fresh, while the ink of the rest of the manuscript is on the smooth side of the vellum of the yellowish brown colour which (as well as the red) our Facsimile tries to imitate, but of a darker or more ashy hue where the surface is rougher, or the leaf otherwise in worse condition: so precarious is any argument that may be drawn from mere difference in the shade of ink as to change in the hand which used it.

1 S. Mark must have ended on Fol. 349 a., the Acts begin Fol. 415 b.: the Catholic Epistles occupy 49 pages in the Elzevir N. T. 1624 (see the last note but one), but possibly some little allowance should be made for the larger space taken up by Cod. Bezae at the end of the several books.

2 Fol. 251 a. contains 34 lines, to make it correspond with its parallel 250 b., which had lost a verse by ouocoτέλευτον between ειπεν and εκατον 1. 29.

3 In this reprint of the text of Codex Bezae, since the words are divided for the convenience of the reader, it has not been so easy as it was to Kipling to represent the spaces found at times in the original between the several words and letters; but we have done so as far as was possible. We have given απαρτι and ουκέτι, but δια τι always. In such compounds as num quid we follow the spaces left by the scribe.