Footnotes
[1.]Unfortunately, it did not occur to us till after the work was nearly all in type to transfer the Lithographed Plates to places opposite the pages which they chiefly illustrate, and that in consequence a few expressions in the text ought to be altered. The advantage of this arrangement appears to be so great as to overbalance the slight inaccuracies alluded to, which cannot now be removed. The plates and their references will, it is hoped, be found easily from the explanations here given.[2.]In later manuscripts Proper Names are often distinguished by a horizontal line placed over them, but no such examples occur in these Plates.[3.]The reader will observe throughout these specimens that the breathings and accents are usually attached to the first vowel of a diphthong.[4.]“Remarks upon a late Discourse of Free Thinking by Phileleutherus Lipsiensis,” Part i, Section 32.[5.]I cite from the late Canon Cureton's over-literal translation in his “Remains of a very antient recension of the four Gospels in Syriac,” in the Preface to which (pp. xxxv-xxxviii) is an elaborate discussion of the evidence for this passage.[6.]But see Dean Burgon's “The Revision Revised,” pp. 358-361.[7.]The word ἡτακισμός or ἰτακισμός is said to have been first used by Cassiodorus (a.d. 468-560?). See Migne, Patr. Lat. t. 70, col. 1128.[8.]To this list of examples from the Book of Common Prayer, Dean Burgon (“The last twelve verses of St. Mark's Gospel Vindicated” p. 215) adds the Gospels for Quinquagesima, 2nd Sunday after Easter, 9th, 12th, and 22nd after Trinity, Whitsunday, Ascension Day, SS. Philip and James, All Saints.[9.]Dean Alford (see his critical notes on Luke ix. 56; xxiii. 17) is reasonably unwilling to admit this source of corruption, where the language of the several Evangelists bears no close resemblance throughout the whole of the parallel passages.[10.]The oldest manuscripts seem to elide the final syllable of ἀλλά before nouns, but not before verbs: e.g. John vi. 32, 39. The common text, therefore, seems wrong in Rom. i. 21; iv. 20; v. 14; viii. 15; 1 Cor. i. 17; vi. 11; ix. 27; xiv. 34; 1 Pet. ii. 25; Jude 9. Yet to this rule there are many exceptions, e.g. Gal. iv. 7 ἀλλὰ υἱός is found in nearly all good authorities.[11.]Tischendorf indeed (Nov. Test. 1871), from a suggestion of Granville Penn in loc., says, “ΚΥΡΙΩ omnino scribi solet ΚΩ,” and this no doubt is the usual form, even in manuscripts which have χρω ιηυ, as well as χω ιυ, for χριστῷ ἰησοῦ. Yet the Codex Augiensis (Paul. F) has κρν in 1 Cor. ix. 1.[12.]Especially, yet not always, at the end of a line. Και in καιρός is actually thus written in Cod. Sinaiticus (א), 1 Macc. ix. 7; xv. 33; Matt. xxi. 34; Rom. iii. 26; Heb. xi. 11; Apoc. xi. 18. So Cod. Sarravianus of the fourth century in Deut. ix. 20, Cod. Rossanensis of the sixth (but only twice in the text), the Zurich Psalter of the seventh century is Ps. xcvii. 11; cvi. 3; cxvi. 5, and the Bodleian Genesis (ch. vi. 13) of about a century later. Similarly, καινήν is written κνην in Cod. B. 2 John 5.[13.]My departed friend, Dr. Tregelles, to whose persevering labours in sacred criticism I am anxious, once for all, to express my deepest obligations, ranged various readings under three general heads:—substitutions; additions; omissions. Mr. C. E. Hammond, in his scholarlike little work, “Outlines of Textual Criticism applied to the N. T., 1876, 2nd edition,” divides their possible sources into Unconscious or unintentional errors, (1) of sight; (2) of hearing; (3) of memory: and those that are Conscious or intentional, viz. (4) incorporation of marginal glosses; (5) corrections of harsh or unusual forms of words, or expressions; (6) alterations in the text to produce supposed harmony with another passage, to complete a quotation, or to clear up a presumed difficulty; (7) Liturgical insertions. While he enumerates (8) alterations for dogmatic reasons, he adds that “there appears to be no strong ground for the suggestion” that any such exist (Hammond, p. 17). Professor Roberts (“Words of the New Testament” by Drs. Milligan and Roberts, 1873) comprehends several of the foregoing divisions under one head: “Again and again has a word or phrase been slipped in by the transcriber which had no existence in his copy, but which was due to the working of his own mind on the subject before him.” His examples are ἔρχεται inserted in Matt. xxv. 6; ἰδοῦσα in Luke i. 29; ὑπὲρ ἡμῶν in Rom. viii. 26 (Part 1. Chap. 1. pp. 5, 6).[14.]This source of variations, though not easily discriminated from others, must have suggested itself to many minds, and is well touched upon by the late Isaac Taylor in his “History of the Transmission of Antient Books to modern times,” 1827, p. 24. So Dr. Hort, when perplexed by some of the textual problems which he fails to solve, throws out as an hypothesis not in itself without plausibility, the notion of “a first and a second edition of the Gospels, both conceivably apostolic” (Gr. Test. Introduction, p. 177).[15.]“Novum Testamentum Textûs Stephanici a.d. 1550 ... curante F. H. A. Scrivener.” Cantabr. 1877 (Editio Major, 1887).[16.]In this manner we propose to indicate the dates of the birth and death of the person whose name immediately precedes.[17.]“Greek and Latin Palaeography,” Chaps. II, III.[18.]“Recent investigations have thrown doubts on the accuracy of this view; and a careful analysis of many samples has proved that, although cotton was occasionally used, no paper that has been examined is entirely made of that substance, hemp or flax being the more usual material.” Maunde Thompson, p. 44.[19.]Tischendorf (Notitia Codicis Sinaitici, p. 54) carried to St. Petersburg a fragment of a Lectionary which cannot well be assigned to a later date than the ninth century, among whose parchment leaves are inserted two of cotton paper, manifestly written on by the original scribe.[20.]“Ten Years Digging in Egypt,” pp. 120, &c.[21.]“Greek and Latin Palaeography,” p. 35; Pliny, Nat. Hist. xiii. 11.[22.]“Nam, quod in palimpsesto, laudo equidem parcimoniam,” Cicero, Ad Diversos, vii. 18, though of a waxen tablet. Maunde Thompson, p. 75.[23.]“Habeant qui volunt veteres libros, vel in membranis purpureis auro argentoque descriptos.” Praef. in Job. “Inficiuntur membranae colore purpureo, aurum liquescit in litteras.” Epist. ad Eustochium.[24.]Miniatures are found even as early as in the Cod. Rossanensis (Σ) at the beginning of the sixth century.[25.]This paragraph which has been rewritten, has been abridged from Mr. Maunde Thompson's “Greek and Latin Palaeography,” pp. 50-52, to which readers are referred for verification and amplification.[26.]“Greek and Latin Palaeography,” p. 49.[27.]Besides the Cod. Sinaiticus, the beautiful Psalter purchased by the National Library from the Didot sale at Paris has four columns (Mr. J. Rendel Harris), and besides the Cod. Vaticanus, the Vatican Dio Cassius, the Milan fragment of Genesis, two copies of the Samaritan Pentateuch at Nablous described by Tischendorf (Cod. Frid.-Aug. Proleg. § 11), the last part of Cod. Monacensis 208 (Evan, 429), and two Hebrew MSS. Cod. Mon. Heb. 422, and Cod. Reg. Heb. 17, are arranged in three columns. Tischendorf has more recently discovered a similar arrangement in two palimpsest leaves of Wisdom and Ecclesiasticus from which he gives extracts (Not. Cod. Sinait. p. 49); in a Latin fragment of the Pentateuch, the same as the Ashburnham manuscript below, seen by him at Lyons in 1843; in a Greek Evangelistarium of the eighth century, and a Patristic manuscript at Patmos of the ninth (ibid. p. 10); so that the argument drawn from the triple columns must not be pressed too far. He adds also a Turin copy of the Minor Prophets in Greek (Pasinus, Catalogue, 1749), and a Nitrian Syriac codex in the British Museum “quem circa finem quarti saeculi scriptum esse subscriptio testatur” (Monum. sacra inedita, vol. i, Proleg. p. xxxi). To this not slender list Mr. E. Maunde Thompson enables us to annex B. M. Addit. 24142, a Flemish Latin Bible of the eleventh century. The late Lord Ashburnham in 1868 printed his Old Latin fragments of Leviticus and Numbers, also in three columns, with a facsimile page; and the famous Utrecht Psalter, assigned by some to the sixth century, by others to the ninth or tenth, is written with three columns on a page.[28.]“Uncialibus, ut vulgo aiunt, literis, onera magis exarata, quam codices,” Hieronymi Praef. in Job. From this passage the term uncial seems to be derived, uncia (an inch) referring to the size of the characters. Yet the conjectural reading “initialibus” will most approve itself to those who are familiar with the small Latin writing of the Middle Ages, in which i is undotted, and c much like t.[29.]The Cotton fragment of the book of Genesis of the fifth century, whose poor shrivelled remains from the fire of 1731 are still preserved in the British Museum, while in common with all other manuscripts it exhibits the round shapes of Ο and Θ, substitutes a lozenge [symbol] for the circle in phi, after the older fashion ([symbol somewhat like a squared Phi]). Phi often has much the same shape in Codex Bezae; e.g. Matt. xiii. 26, Fol. 42 b, 1. 13, and once in Codex Z (Matt. xxi. 26, Plate xlviii).[30.]Our facsimile is borrowed from the Neapolitan volumes, but Plate 57 in the Paléographie Universelle φιλοδημου περι μουσικη has the advantage of colours for giving a lively idea of the present charred appearance of these papyri.[31.]Cicero de Finibus, Lib. ii. c. 35. The same person is apparently meant in Orat. in Pisonem, cc. 28, 29.[32.]We prefer citing Cod. Frid.-August., because our examples have been actually taken from its exquisitely lithographed pages; but the facsimile of part of a page from Luke xxiv represented in Tischendorf's Cod. Sinaiticus, from which we have borrowed six lines (No. [11 b]), will be seen to resemble exactly the portion published in 1846.[33.]Cod. A is found in the simpler form in the Old Testament, but mostly with the horizontal line produced in the New.[34.]See Maunde Thompson's “Greek and Latin Palaeography.”[35.]Codd. B of Apocalypse, Θa Λ (No. [30]) of the Gospels, and Silvestre's No. 68, all of about the eighth century, slope more or less to the right; Cod. Γ (No. [35]) of the ninth century, a very little to the left. Tischendorf assigns to the seventh century the fragments comprising Leipzig II. (see p. [39]), though they lean much to the right (Monum. sacra ined. tom. i, pp. xxx-xxxiv, 141-176), and those of Isaiah (ibid. pp. xxxvi, xxxvii, 187-199).[36.]The earliest cursive Biblical manuscript formerly alleged, i.e. Evan. 14, on examination proves to have no inscription whatever. “On folio 392, in a comparatively modern hand, is rather uncouthly written ἐγράφη νικηφόρου βασιλεύοντος A. Z. What the initials A. Z. stand for I do not know.” (Dean Burgon, Guardian, Jan. 15, 1873.) The claim of priority for Cod. 14 being thus disposed of (though it must be noted that Dr. C. R. Gregory refers it without doubt to the tenth century), we may note that Cod. 429 of the Gospels is dated 978, Cod. 148 of the Acts 984, Cod. 5pe 994, and Λ, written partly in cursives, and partly in uncials is of the ninth century. But the date May 7, 835 a.d. is plainly visible on Cod. 481, which is therefore indisputably the earliest.[37.]See Maunde Thompson, Greek and Latin Palaeology, chap. x. pp. 130, &c., and chap. viii. pp. 107, &c; Notices et Extracts des MSS. de la Bibliothèque Imperiale, Paris, plate xxiv. no. 21, pl. xlviii. no. 21 ter, xlvi. no. 69, e, xxi. no. 17, xiii. no. 5, xl. no. 62, xviii. 2, pl. xliv; Cat. Gr. Papyri in Brit. Mus. Palaeograph. Soc. ii. pl. 143, 144, Mahaffy, Petrie Papyri, pl. xiv, xxix. &c. (Cunningham Memoirs of R. Irish Academy).[38.]At the end of the Euclid we read εγραφη χειρι στεφανου κληρικου μηνι σεπτεμβριωι ινδ. ζ ετει κοσμου ς τ ω ζ εκτησαμην αρεθας πατρευς την παρουσαν βιβλιον: of the Plato, εγραφη χειρι ιω καλλιγραφου; ευτυχως αρεθη διακονωι πατρει; νομισματων βυζαντϊεων δεκα και τριων; μηνι νοεμβριωι ινδικτιωνος ιδ; ετει κοσμου ςυδ βαςιλειας λεοντος του φιλοχυ υιου βασιλειου του αειμνιστου. It should be stated that these very curious books, both written by monks, and indeed all the dated manuscripts of the Greek Testament we have seen except Canonici 34 in the Bodleian (which reckons from the Christian era, a.d. 1515-6), calculate from the Greek era of the Creation, September 1, b.c. 5508. To obtain the year a.d., therefore, from January 1 to August 31 in any year, subtract 5508 from the given year; from September 1 to December 31 subtract 5509. The indiction which usually accompanies this date is a useful check in case of any corruption or want of legibility in the letters employed as numerals. Both dates are given in Evan. 558, viz. a.m. 6938, and a.d. 1430.[39.]The writer of Burney 21 (rscr) a.d. 1292 (Evan. 571), ὁ ταπεινος Θεοδωρος ἁγιωπετριτης ταχα και καλλιγραφος as he calls himself (that is, as I once supposed, monk of the Convent of Sancta Petra at Constantinople, short-hand and fair writer), was the scribe of at least five more copies of Scripture now extant: Birch's Havn. 1, a.d. 1278 (Evan. 234); Evan. 90, a.d. 1293; Evan. 543, a.d. 1295; Scholz's Evan. 412, a.d. 1301; Evan. 74, undated. To this list Franz Delitzsch (1813-1890) (Zeitschr. f. luth. Theol. 1863, ii, Abhandlungen, pp. 217, 218) adds from Matthaei, Synaxarion in Mosc. Syn. Typograph. xxvi. a.d. 1295, and recognizes Hagios Petros, the country of Theodoros, as a town in the Morea, on the borders of Arcadia, from whose school students have attended his own lectures at Erlangen.[40.]Hence in the later uncials, some of which must therefore have been copied from earlier cursives, Β and Υ (which might seem to have no resemblance) are sometimes confounded: e.g. in Parham 18 (a.d. 980), υ for β, Luke vi. 34; β for υ, John x. 1, especially where β begins or ends a line: e.g. Evan. 59, John vii. 35. Evan. 59 has β for υ very often, yet there is no extra trace that it was copied from an uncial.[41.]The full signature not easily deciphered is ἐτελειώθη τὸ παρὸν ἅγιον εὐαγγέλιον κατὰ τὴν κζ τοῦ ἰαννουαρίου μηνὸς τῆς [?] ω κ ζ ἐγχρονίας. Presuming that ς is suppressed before ω κ ζ this is 6827 of the Greeks, a.d. 1319.[42.]Compare also Buttmann's Greek Grammar (Robinson's translation) p. 467; Bast in (Schaefer's Gregorius Corinthius) tabb. ad fin.; Gardthausen, Palaeographie, p. 248, &c.[43.]Thus the type cast for the Royal Printing Office at Paris, and used by Robert Stephen, is said to have been modelled on the style of the calligrapher Angelus Vergecius, from whose skill arose the expression “he writes like an angel.” Codd. 296 of the Gospels, 124 of the Acts, 151 of St. Paul are in his hand.[44.]Yet Tischendorf (N. T. 1859, Proleg. p. cxxxiii) cites ηιδισαν from Cod. Bezae (Mark i. 34), ξυλωι (Luke xxiii. 31) from Cod. Cyprius, ωι from Cod. U (Matt. xxv. 15) and Cod. Λ (Luke vii. 4). Add Cod. Bezae πατρωιου Acts xxii. 3, Scrivener's edition, Introd. p. xix. Bentley's nephew speaks of ι ascript as in the first hand of Cod. B, but he seems to have been mistaken.[45.]In B-C iii. 10 (dated 1430), the whole manuscript being written by the same hand, we have ι ascript twenty-five times up to Luke i. 75, then on the same page ι subscript in Luke i. 77 and eighty-five times afterwards: the two usages are nowhere mixed. In Evan. 558, subscript and ascript are mixed in the same page, Luc. i. 75, 77.[46.]The invention of breathings, accents, and stops is attributed to Aristophanes of Byzantium, 260 b.c.[47.]See below vol. ii. c. ix. 9. note, end. Dr. Scrivener appears not to have formed a positive opinion, which indeed in some of these cases is hardly possible.[48.]He is speaking (Quaestion. super Genes. clxii) of the difference between ῥάβδου αὐτοῦ and ῥάβδου αὑτοῦ, Gen. xlvii. 31. “Fallit enim eos verbum Graecum, quod eisdem literis scribitur, sive ejus, sive suae: sed accentus [he must mean the breathings] dispares sunt, et ab eis, qui ista noverunt, in codicibus non contemnuntur” (Opera, Tom. iv. p. 53, ed. 1586, Lugdun.); adding that “suae” might be expressed by ἑαυτοῦ.[49.]In the Gale Evangelistarium (Trin. Coll. Camb. O. 4. 22) the interrogative clause is set between two such marks in red. Hence it seems not so much a stop as a vocal note. In the Armenian and Spanish languages the note of interrogation is set before the interrogative clause, and very conveniently too.[50.]The earliest known example of the use of two dots occurs in the Artemisia papyrus at Vienna (Maunde Thompson, p. 69), and other early instances are found in a letter of Dionysius to Ptolemy about b.c. 160, published by the French Institute, 1865, in “Papyrus grecs du Musée du Louvre,” &c. tom. xviii. 2e ptie, pl. xxxiv, pap. 49, and in fragments of the Phaedo of Plato discovered at Gurob. The same double points are also occasionally set in the larger spaces of Codd. Sinaiticus, Sarravianus, and Bezae, but in the last-named copy for the most part in a later hand.[51.]Abbot, ubi supra.[52.]Hoskier, Cod. 604, p. xiii.[53.]Even Codex Sinaiticus has ιηυ and ιυ in consecutive lines (Apoc. xxii. 20, 21), and χρυ Rom. vii. 4.[54.]See below p. [64], note 4.[55.]“Fragmenta pauca evangelii Johannis palimpsesta Londinensia [Evan. Ib or Nb]. In ceteris haec fere tria: Dionis Cassii fragmenta Vaticana—vix enim qui in his videntur speciem majorum litterarum habere revera differunt—item fragmenta palimpsesta [Phaëthontis] Euripidis Claromontana et fragmenta Menandri Porphiriana” (Tischendorf, Cod. Vatic. Proleg. p. xviii, 1867).[56.]The English word paragraph is derived from the παραγραφαί, which are often straight lines, placed in the margin to indicate a pause in the sense. Professor Abbot, ubi supra, p. [195], alleges not a few instances where these dashes are thus employed. A specimen is given in Scrivener's Cod. Sinaiticus, facsimile [3]: see his Cod. Sin., Introduction, p. xl and note. Thus also they appear in Cod. Sarravianus (Tischendorf, Mon. sacra ined. vol. iii. pp. xiv, xx). In Cod Bezae [symbol] is set in the margin forty-nine times by a later hand, and must be designed for the same purpose, though the mark sometimes occurs where we should hardly look for it (Scrivener, Cod. Bezae, Introduction, p. xxviii and note). In Cod. Marchalianus the dash stands over the capital at the beginning of a line, or over the first letter where there is no capital. Lastly, in Codd. Vatic. and Sinait. [symbol] is sometimes set in the middle of a line to indicate a paragraph break, followed by [symbol] in the margin of the next line.[57.]Many other examples of the use of στίχοι and versus in this sense will be found in that admirable monument of exact learning, now so little read, Prideaux Connections, An. 446. Stichometry can be traced back to nearly a century before Callimachus, who (b.c. 260) has been credited with the invention (Palaeography, p. 79). The term στίχοι, like the Latin versus, originally referring whether to rows of trees, or to the oars in the trireme (Virg. Aen. v. 119), would naturally come to be applied to lines of poetry, and in this sense it is used by Pindar (ἐπέων στίχες Pyth. iv. 100) and also by Theocritus (γράψον καὶ τόδε γράμμα, τό σοι στίχοισι χαράξω Idyl, xxiii. 46), if the common reading be correct.[58.]That we have rightly understood Epiphanius' notion of the στίχοι is evident from his own language respecting Psalm cxli. 1, wherein he prefers the addition made by the Septuagint to the second clause, because by so doing its authors ἀχώλωτον ἐποίησαν τὸν στίχον: so that the passage should run “O Lord, I cry unto Thee, make haste unto me || Give ear to the voice of my request,” τῆς δεήσεώς μου to complete the rhythm. This whole subject is admirably worked out in Suicer, Thesaur. Eccles. tom. ii. pp. 1025-37.[59.]In the Epistles of St. Paul, Euthalius seems to have followed a Syrian writer. Gregory, Prolegomena, p. 113; Zacagnius, Collectanea Monumentorum Veterum Ecclesiae, Rome, a.d. 1698, pp. 404, 409.[60.]At the end of 2 Thess., in a hand which Tischendorf states to be very ancient, but not that of the original scribe, the Codex Sinaiticus has στιχων ρπ [180; the usual number is 106]: at the end of Rom., 1 Cor., 1 Thess., and the Catholic Epistles, there is no such note; but in all the other Pauline Epistles the στίχοι are numbered.[61.]So the margin of Gale's Evan. 66 contains readings cited by Mill and his followers, which a hand of the sixteenth century took, some of them from the Leicester manuscript, others from early editions.[62.]The following subscription to the book of Ezra (and a very similar one follows Esther) in the Cod. Frid.-August, (fol. 13. 1), though in a hand of the seventh century, will show the care bestowed on the most ancient copies of the Septuagint: Αντεβληθη προσ παλαιωτατον λιαν αντιγραφον δεδιορθωμενον χειρι του αγιου μαρτυροσ Παμφιλου; ὁπερ αντιγραφον προσ τω τελει υποσημειωσισ τισ ϊδιοχειροσ αυτου ϋπεκειτο εχουσα ουτωσ; μετελημφθη και διορθωθη προσ τα εξαπλα ωριγενουσ; Αντωνινοσ αντεβαλεν; Παμφιλοσ διορθωσα. Tregelles suggests that the work of the διορθωτὴς or corrector was probably of a critical character, the office of the ἀντιβάλλων or comparer being rather to eliminate mere clerical errors (Treg. Horne's Introd., vol. iv. p. 85). Compare Tischendorf, Cod. Sinait. Proleg. p. xxii.[63.]“Simile aliquid invenitur in codice Arabico epp. Pauli anno 892, p. Chr., quem ex oriente Petropolin pertulimus.” Tischendorf, Cod. Vat. Prolog. p. xxx. n. 3.[64.]Lat. breves, or τίτλοι: but τίτλος means properly the brief summary of the contents of a κεφάλαιον placed at the top or bottom of a page, or with the κεφάλαια in a table to each Gospel. The κεφ. minora = Ammonian Sections.[65.]This full explanation of a seeming difficulty was communicated to me independently by Mr. P. W. Pennefather of Dublin, and Mr. G. A. King of Oxford.[66.]And this too in spite of the lexicographer Suidas: Τίτλος διαφέρει κεφαλαίου; καὶ ὁ Ματθαῖος τίτλους ἔχει ξή, κεφάλαια δὲ τνέ. And of Suicer, s. v.[67.]Ὁ Τατιανός, συνάφειάν τινα καὶ συναγωγὴν οὐκ οἶδ᾽ ὅπως τῶν εὐαγγελίων συνθείς, τὸ διὰ τεσσάρων τοῦτο προσωνόμασεν; ὃ καὶ παρά τισιν εἰσέτι νῦν φέρεται. Euseb. Hist. Eccl. iv. 29.[68.]Ambros. in Prooem. Luc. seems to aim at Tatian when he says “Plerique etiam ex quatuor Evangelii libris in unum ea quae venenatis putaverunt assertionibus convenientia referserunt.” Eusebius H. E. iv. 29 charges him on report with improving not the Gospels, but the Epistles: τοῦ δὲ ἀποστόλου φασὶ τολμῆσαί τινας αὐτὸν μεταφράσαι φωνάς, ὡς ἐπιδιορφούμενον αὐτῶν τὴν τῆς φράσεως σύνταξιν. Dr. Westcott's verdict is rather less favourable than might have been anticipated: “The heretical character of the Diatessaron was not evident on the surface of it, and consisted rather of faults of defect than of erroneous teaching” (History of the Canon, p. 354). From the Armenian version of Ephraem the Syrian's Exposition of Tatian's Harmony, printed in 1836, translated in 1841 by Aucher of the Melchitarist Monastery at Venice, but buried until it was published with notes by Moesinger in 1876, a flood of light is thrown upon this question, and it is now clear “that Tatian habitually abridged the language of the passages which he combined” (Hort, Gk. Test. Introduction, p. 283); and that apparently in perfect good faith.[69.]Ἀμμώνιος μὲν ὁ Ἀλεξανδρεύς, πολλήν, ὡς εἰκός, φιλοπονίαν καὶ σπουδὴν εἰσαγηοχώς, τὸ διὰ τεσσάρων ἡμῖν καταλέλοιπεν εὐαγγέλιον, τῷ κατὰ Ματθαῖον τὰς ὁμοφώνους τῶν λοιπῶν εὐαγγελιστῶν περικοπὰς παραθείς, ὡς ἐξ ἀνάγκης συμβῆναι τὸν τῆς ἀκολουθίας εἱρμὸν τῶν τριῶν διαφθαρῆναι, ὅσον ἐπὶ τῷ ὕφει τῆς ἀναγνώσεως. Ἵνα δὲ σωζομένου καὶ τοῦ τῶν λοιπῶν δι᾽ ὅλου σώματός τε καὶ εἱρμοῦ, εἰδέναι ἔχοις τοὺς οἰκείους ἑκάστου εὐαγγελιστοῦ τόπους, ἐν οἶς κατὰ τῶν αὐτῶν ἠνέχθησαν φιλαλήθως εἰπεῖν, ἐκ τοῦ πονήματος τοῦ προειρημένου ἀνδρὸς εἰληφὼς ἀφορμάς (taking the hint from Ammonius' as Dean Burgon rightly understands the expression), καθ᾽ ἑτέραν μέθοδον κανόνας δέκα τὸν ἀριθμὸν διεχάραξά σοι τοὺς ὑποτεταγμένους. Epist. ad Carpian. initio. I have thankfully availed myself on this subject of Burgon's elaborate studies in The Last Twelve Verses of St. Mark, pp. 125-132; 295-312.[70.]This is the number given for St. Mark by Suidas and Stephen. It is an uncertain point: thirty-four manuscripts give 233, reckoning only to xvi. 8; while thirty-six give 341. See Burgon Twelve Last Verses, p. 311.[71.]I subjoin Eusebius' own words (Epist. ad Carpian.) from which no one would infer that the sections were not his, as well as the canons. Αὕτη μὲν οὖν ἡ τῶν ὑποτεταγμένων κανόνων ὑπόθεσις; ἡ δὲ σαφὴς αὐτῶν διήγησις, ἔστιν ἤδε. Ἐφ᾽ ἑκάστῳ τῶν τεσσάρων εὐαγγελίων ἀριθμός τις πρόκειται κατὰ μέρος, ἀρχόμενος ἀπὸ τοῦ πρώτου, εἶτα δευτέρου, καὶ τρίτου, καὶ καθεξῆς προϊὼν δι᾽ ὅλου μέχρι τοῦ τέλους τοῦ βιβλίου [the sections]. Καθ᾽ ἕκαστον δὲ ἀριθμὸν ὑποσημείωσις διὰ κινναβάρεως πρόκειται [the canons], δηλοῦσα ἐν ποίῳ τῶν δέκα κανόνων κείμενος ὁ ἀριθμὸς τυγχάνει.[72.]Something of this kind, however, must be the plan adopted in Codex E (see Plate [xi]. No. 27) of the Gospels, as described by Tregelles, who himself collated it. “[It has] the Ammonian sections; but instead of the Eusebian canons there is a kind of harmony of the Gospels noted at the foot of each page, by a reference to the parallel sections of the other Evangelists.” Horne's Introd. vol. iv. p. 200. Yet the canons also stand in the margin of this copy under the so-called Ammonian sections: only the table of Eusebian canons is wanting. The same kind of harmony at the foot of the page appears in Cod. Wd at Trinity College, Cambridge, but in this latter the sections in the margin are not accompanied by the canons. Tischendorf states that the same arrangement prevails in the small fragment Tb at St. Petersburg; Dean Burgon adds to the list Codd. M. 262, 264 at Paris, and conceives that this method of harmonizing, which he regards as far simpler than the tedious and cumbersome process of resorting to the Eusebian canons (ubi supra, p. [304]), was in principle, though not in details, derived to the Greek Church from early Syriac copies of the Gospels, some of which still survive (p. [306]).[73.]To this list of manuscripts of the Gospels which have the Ammonian sections without the Eusebian canons add Codd. 38, 54, 60, 68, 117; Brit. Mus. Addit. 16184, 18211, 19389; Milan Ambros. M. 48 sup.; E. 63 sup.; Burdett-Coutts I. 4; II. 18; 262; III. 9. Now that attention has been specially directed to the matter, it is remarkable how many copies have the Ammonian sections without the corresponding Eusebian canons under them, sometimes even when (as in Codd. 572, 595, 597) the letter to Carpianus and the Eusebian tables stand at the beginning of the volume. To the list here given must now be added Codd. O, Υ, 185, 187, 190, 193, 194, 207, 209, 214, 217, 367, 406, 409, 410, 414, 418, 419, 456, 457, 494, 497, 501, 503, 504, 506, 508, 518, 544, 548, 550, 555, 558, 559, 564, 573, 575, 584, 586, 591, 592, 601, 602, 620: in all seventy-one manuscripts.[74.]No doubt they do serve, in the manuscripts which contain them and omit the canons, for marks of reference, like in kind to our modern chapters and verses; but in consequence of their having been constructed for a wholly different purpose, they are so unequal in length (as Burgon sees very clearly, pp. 297, 303), that they answer that end as ill as any the most arbitrary divisions of the text well could do.[75.]Sulci in Sardinia is the only Bishop's see of the name I can find in Carol. a Sancto Paulo's Geographia Sacra (1703), or in Bingham's Antiquities, Bk. ix. Chapp. II, VII. Horne and even Tregelles speak of Sulca in Egypt, but I have searched in vain for any such town or see. Euthalius is called Bishop of Sulce both in Wake 12 (infra, note 4), and in the title to his works as edited by L. A. Zacagni (Collectanea Monument. Veter. Eccles. Graec. ac Latin., Rom. 1698, p. 402). But one of Zacagni's manuscripts reads Σούλκης once, and he guesses Ψέλχη near Syene, which appears in no list of Episcopal sees.[76.]Καθ᾽ ἑκάστην ἐπιστολὴν προτάξομεν τὴν τῶν κεφαλαίων ἔκθεσιν, ἑνὶ τῶν σοφωτάτων τινὶ καὶ φιλοχρίστων πατέρων ἡμῶν πεπονημένην.[77.]Αὐτίκα δῆτα is his own expression.[78.]E.g. in Wake 12, of the eleventh century, at Christ Church, the title at the head of the list of chapters in the Acts is as follows: Εὐθαλίου ἐπισκόπου Σουλκῆς ἔκθεσις κεφαλαίων τῶν Πράξεων σταλῆσα (-εῖσα) πρὸς Ἀθανάσιον ἐπίσκοπον Ἀλεξανδρείας.[79.]In Wake 12 certain of the longer κεφάλαια are subdivided into μερικαὶ ὑποδιαιρέσεις in the Acts, 1 Peter, 1 John, Romans, 1, 2 Corinthians, Colossians, 2 Thessalonians, 1 Timothy, Hebrews only. For a similar subdivision in the Gospels, see Evan. 443 in the list of cursive MSS. given below.[80.]Διὰ τὴν τριμερῆ τῶν εἴκοσι τεσσάρων πρεσβυτέρων ὑπόστασιν, σώματος καὶ ψυχῆς καὶ πνεύματος. See Matthaei, N. T. Gr. et Lat. vii. 276, note 4.[81.]Many manuscripts indicate passages of the Old Testament cited in the New by placing > (as in Codd. Vatican. Wd, &c., but in Sinait. more rarely), or [symbol with two greater-than marks], or some such mark in the margin before every line. Our quotation-marks are probably derived from this sign, the angle being rounded into a curve. Compare the use of [symbol like right double quotation mark] in the margin of the Greek Testament of Colinaeus, 1534, and Stephen's editions of 1546, -49, -50, &c. Evan. 348 and others have [symbol]. In Codd. Bezae, as will appear hereafter, the words cited are merely thrown a letter or two back in each line.[82.]The whole mystery is thus unfolded (apparently by Cosmas) in Lamb. 1178, p. 159: Καὶ γὰρ τὰ χερουβὶμ τετραπρόσωπα; καὶ τὰ πρόσωπα αὐτῶν εἰκόνες τῆς πραγματείας τοῦ υἱοῦ τοῦ Θεοῦ; τὸ γὰρ ὅμοιον λέοντι, τὸ ἔμπρακτον καὶ βασιλικὸν καὶ ἡγεμονικὸν [John i. 1-3] χαρακτηρίζει; τὸ δὲ ὅμοιον μόσχωι, τὴν ἱερουργικὴν καὶ ἱερατικὴν [Luke i. 8] ἐμφανίζει; τὸ δὲ ἀνθρωποειδές, τὴν σάρκωσιν [Matt. i. 18] διαγράφει. τὸ δὲ ὅμοιον ἀετῶι, τὴν ἐπιφοίτησιν τοῦ ἁγίου πνεύματος [Mark i. 2] ἐμφανίζει. More usually the lion is regarded as the emblem of St. Mark, the eagle of St. John.[83.]N. B. The στίχοι of the Acts and of all the Epistles except Hebr. are taken from the Codex Passionei (G or L), an uncial of the ninth century.[84.]
The numbers of the Gospel στίχοι in our Table are taken from the uncial copies Codd. GS and twenty-seven cursives named by Scholz: those of the ῥήματα from Codd. 9, 13, 124 and seven others. In the ῥήματα he cites no other variation than that Cod. 339 has 2822 for St. Matthew: but Mill states that Cod. 48 (Bodl. 7) has 1676 for Mark, 2507 for Luke (Proleg. N. T. § 1429). In Cod. 56 (Lincoln Coll.) the ἀναγνώσματα of St. Matthew are 127, of St. Mark 74, of St. Luke 130 (Mill).
In the στίχοι, a few straggling manuscripts fluctuate between 3897? and 1474 for Matthew; 2006 and 1000 for Mark; 3827 and 2000 for Luke; 2300 and 1300 for John. But the great mass of authorities stand as we have represented.
I am indebted for the following Memoranda on Cod. א to the kindness of the Dean of Derry and Raphoe.
i. It is demonstrable that the Eusebian Sections and Canons on the margin are contemporaneous with the text. For they are wanting from leaves 10 and 15. Now these leaves are conjugate; and they have been (on other grounds) noted by Tischendorf as written not by the scribe of the body of the N. T., but by one of his colleagues (“D”) who wrote part of the O. T. and acted as Diorthota of the N. T. It thus appears that, after the marginal numbers had been inserted, the sheet containing leaves 10 and 15 was cancelled, and rewritten by a contemporary hand. The numbers must therefore have been written before the MS. was completed and issued.
ii. The exemplar whence these numbers were derived, differed considerably from that which the text follows. For, in some cases, the sectional numbers indicate the presence of passages which are absent from the text. E.g. St. Matt. xvi. 2, 3, which is sect. 162, is wanting; and 162 is assigned to ver. 4, while the wrong canon (5 for 6) betrays the presence in the canonizer's exemplar of the passage omitted by the scribe. The same is true of St. Mark xv. 28 (in which case the scribe is “D”).
iii. The scribe who wrote the text was unacquainted with the Eusebian sections. For the beginning of a section is not marked, as in A and most subsequent MSS., by a division of the text and a larger letter. On the contrary the text is divided into paragraphs quite independent of the Eusebian divisions, which often begin in the middle of a line, and are marked merely by two dots (:) in vermilion, inserted no doubt by the rubricator as he entered the numbers in the margin. The fact that the numbers of the sections as well as of the canons (not as in other MSS. of the Canons only) are in vermilion, points the same way.