a quality which they still preserve, when for want of room, which is often the case at the end of a verse, smaller characters are substituted. This has been noted by palæographists as a distinction of the most ancient MSS. It belongs to the Cod. Vatican., Alexandrin., Ephrem. Syr. Rescript., and some other codices of very early date; whereas in books which are known from independent evidence to be not older than the beginning of the sixth century, a change and deterioration in this respect are said to be already perceptible. The purity of the uncial character seems, therefore, to imply a date anterior to the sixth century. Moreover, the Sinaitic Codex is without initial let⚫ters; and this Tischendorf considers a decisive test of age. All who have made palæography their study will agree, he says, with him, that MSS., whether Greek or Latin, which are probably referable to the fifth century, are distinguished by initial letters; while the absence of those letters characterizes papyri and the most ancient codices that we possess, such as the Codex Vaticanus and the celebrated Parisian fragments of the Octateuch. The simplicity and extreme rarity of the interpunction in the Sinaitic MS. is another sign of antiquity. Constantly where a connected series is completed and we should naturally look for a point, either none is inserted, or a blank space is left, and the hand of a later corrector has supplied it. In this respect the Codex Sinaiticus has more affinity with the Cod. Vatican. and the papyri, than with MSS. of apparently later date. As facsimiles have been published of the writing in both these codices, Vatican. and Sinait., the student can readily compare them. In form and general appearance the characters much resemble each other. Those of the Cod. Sinait. as exhibited by Tischendorf at the end of his Notitia, are somewhat taller, and the strokes forming them slenderer and more distinct; but the writing is so clear and uniform and beautiful, that notwithstanding the absence of interpunction and initial letters, even an eye unaccustomed to MSS. after a little practice would read it with

ease.

It has been noticed, as an indication of the probable date of this codex, that the characters in which it is written appear to be in a transition-state between those which are found in the old papyrus rolls and those which became current in the fifth and sixth centuries. A similar conclusion is suggested by the shape and size of the codex itself. When Hug examined the Codex Vaticanus in 1809, at the time that it was deposited in the Imperial Library at Paris, he noticed at once the close resemblance of the expanded codex, with its six parallel columns

A name for the eight first books of the Old Testament, on which a continuous commentary was often written. Heptateuch was used in the same way of the seven first. See Fabric. Biblioth. Græc. edit. Harles. VII. p. 563, sub Procop. Gaz.

on the contiguous pages, and its greater width than height (somewhat in the form, we may presume, of a modern music-book), to the ancient volumen; and he very ingeniously and very probably referred its origin to a period when the old roll was gradually giving way to the modern book. The first books retained much of the form and character of the previous rolls; just as the earliest printing is almost a facsimile of the latest MSS.* This argument from form and appearance applies with peculiar force to the Sinaitic Codex; for when it is opened, it exhibits, except in the stichometric parts, not six, but eight columns on the adjoining pages, and therefore presents a still closer likeness to the ancient volumen. In this respect Tischendorf affirms that it is unique among MSS. hitherto known.

Other tokens of antiquity Tischendorf finds-in the orthography of this MS., which agrees minutely with that of the Codex Vaticanus; in the arrangement of the books of the New Testament; in the remarkable simplicity of their inscriptions and subscriptions; and in the entire absence of accents which even the two earliest correctors, who must have lived some centuries after the original writing of the text, have nowhere employed. Another circumstance is entitled to still greater weight. According to Cæsarius and Epiphanius, the Ammonian sections and the canons of Eusebius began to be generally introduced into Greek MSS. of the New Testament about the middle of the fourth century; and a little later Jerome transferred them to his revision of the Latin version. Now it is remarkable that in all our oldest MSS. of the New Testament these sections are indicated, with the single exception of this Sinaitic monument, which has not a trace of them a primá manu, and of the Cod. Vatican., which has divisions peculiar to itself. "Qua ex re," says Tischendorf, "quod sumitur argumentum quantam in summa codicis vetustate probanda vim habeat, facile cuivis clarum erit." It is further a significant fact, that this codex includes the Epistle of Barnabas and the Shepherd of Hermas along with the canonical books of the New Testament. We know from Clement of Alexandria and Origen, that in the second and third centuries these books were considered by many as having a sort of canonical authority (our modern idea of a

Only a very few Greek MSS. are marked by this transition-character of resemblance to the old roll. Hug had noticed some fragments of Dio Cassius preserved in the Vatican. Tischendorf has added to the number two very ancient copies of the Samaritan Pentateuch at Naplus in Palestine, and one or two other Greek and Latin MSS. p. 12, note.

+ The Ammonian sections with references to the Eusebian canons occur in the Cod. Alexandrin., in the Cod. Ephrem. Syr. and in the Cod. Cantabrig.; in this last, however, without reference to the Eusebian canons, and added, it is thought, by a later hand. See Rev. J. S. Porter's Textual Criticism, p. 202. The Cod. Cantabr., like the Cod. Vatican., seems to have had originally divisions peculiar to itself.

canon did not develop itself till a later period); and even as late as the middle of the fourth, they were still acknowledged to a certain extent by Eusebius as "dubiæ auctoritatis," classed by him among the arriλeyóueva and vóda, not absolutely excluded, but fit to be read. With this decision of Eusebius, the canon, remounting probably to the same date, which is attached to the Cod. Claromontan. is, according to Tischendorf, in entire agreement. Before the close, however, of that century, as we gather from the enactments of the Councils of Laodicea, 364 A. D., and of Carthage, 390 A. D., the number of canonical books was determined by public authority, and both Barnabas and the Shepherd were excluded from them. In the Sinaitic Codex, some leaves are wanting between Barnabas and the Shepherd, as well as at the end of it; and Tischendorf thinks it probable that in these intervals some other of the antilegomena of Eusebius may originally have found a place. All these indications strengthen him in the conclusion, that this document cannot be later than the fourth century. He is aware that a parallel phenomenon in the Cod. Alexandrin., which has the two Epistles of Clement appended to it, and yet by none is placed earlier than the fifth, by some is even brought down as low as the sixth century, may seem at first view to weaken the force of his reasoning; but he meets the objection by replying, that although, as we learn from Jerome, the first Epistle of Clement in some parts of Christendom appears to have been read publicly in the church towards the end of the fourth century, there is not a trace of such a distinction having ever been conferred on what is called the second; and that consequently some other motive than that of supposing them possessed of any degree of canonical authority, must have induced the transcriber of this MS. to place them where they now are.

Certain peculiarities in the text of this codex point, in the judgment of the editor, to the same conclusion as to age which has been rendered probable by other considerations. The best MSS. of the fifth and sixth centuries, A, C, D, with fourteen other uncial MSS., all the cursive, seven copies of the Itala, all of the Vulgate, and all of the Syriac version, with the MæsoGothic and some other ancient versions, contain the twelve last verses of the Gospel of Mark. Nevertheless, Eusebius, who died 340 A.D., declares, in a letter to Marinus, that these verses were wanting in nearly all the accurate copies of his time; and his testimony is confirmed by Jerome, who says, "omnes Græciæ libros pæne hoc capitulum non habere." In its rejection of these verses against the united witness of more than 500 Greek MSS. from the fifth century downwards, though in accordance with

The Shepherd of Hermas is placed in a similar rank to that assigned it by Eusebius in the Muratorian Fragment of a Canon, usually referred to the opening of the third century.

the explicit statements of Eusebius and Jerome-this Sinaitic monument stands alone with the Codex Vaticanus.*

It is well known that critics have found some difficulty in determining to whom the Epistle (generally entitled)" To the Ephesians was actually addressed. Some, after Marcionamong them, Wetstein and Paley-have supposed that it was originally sent to the church of Laodicea; others, such as Michaelis, Hug, Eichhorn, Neander, have regarded it as a kind of circular to be forwarded in succession to different churches; and some have actually imagined that in the original autograph, where the words év 'Epéoy now occur, a space was left vacant where the name might be inserted of the community to which it was to be immediately addressed. The objection to these theories is, that in an overwhelming majority of extant MSS., the words év 'Epéry occur in the opening salutation and determine the specific direction of the Epistle. But here, again, a singular phenomenon meets us. Besides the omission of these words from Marcion's canon, to which in itself no great weight might be thought to attach, Basil attests, about the middle of the fourth century, that they were wanting in old copies in his time, and Jerome indirectly indicates that they were unknown to Origen in the first half of the third century. In support of this very ancient traditional testimony, against the combined evidence of all existing MSS., we can only adduce the Cod. Vatican. and the Cod. Sinaiticus, which again, as represented by their primitive text, stand alone in their united witness.t

In dealing with ancient MSS. it is not always easy to decide what has proceeded a primá manu, and what has been added in subsequent revisions. Tischendorf's object in editing the Cod. Sinait. is to exhibit as faithfully as possible the primitive text. He supposes it to have been re-touched at different periods by several hands, at least by four. Some of these appear to have revised only certain books; others, the entire document; others, again, not so much to have meddled with the text, as to have supplied certain additions, as the Ammonian and Eusebian num

Tischendorf observes in a note, that only one copy of the Itala (Codex Taurinensis, formerly Bobbiensis, probably to be referred to the fourth century) excludes these last verses of Mark's Gospel; and that three oriental versions, the Armenian, the Ethiopic and the Arabic, though the great majority of MSS. contain them, seem nevertheless to favour this constitution of the original text. Of modern writers, amidst a perplexing balance of authorities, Griesbach and Lachmann admit them; Tischendorf (edit. stereotyp., Lips. 1850) excludes them. It is curious that, among the Syriac fragments of the Gospels recently published by Dr. Cureton, the only portion preserved of Mark should be the four last verses of this 16th chapter.

+ It ought to be stated that in the Cod. Vatican. the words v 'Epioy have been inserted in the margin, Lachmann says a prima manu; Tischendorf, a secunda. The reading of the Cod. Sinait. confirms the original decision of Tischendorf. He mentions in a note, p. 14, that in a MS. of the Pauline Epistles at Vienna, the words tv 'Epéoy have been marked with points by a later hand, as if on the authority of some more ancient codex.

bers. On this part of the subject, which is very important towards determining the critical worth of the codex, he promises fuller details in his prolegomena (p. 19, note).

In accordance with the practice of former scholars, Tischendorf designates this new source of criticism, and marks its place among the older authorities, by the first letter of the Hebrew alphabet, ; and in his Notitia he has given a tolerably copious specimen of its characteristic readings. From a cursory survey confined to the four Gospels, we should say, that they coincide mostly with B (Cod. Vatican.), with C (Ephrem. Syr. Rescr.), with A (Alexandr.), with D (Cantabr.), with L (Parisiens.), with A (Sangallens.). One or two of these readings it may be curious and instructive to notice.

Mark i. 1, vou tou are omitted, as they are, according to Lachmann and Griesbach, by Irenæus and Origen.

Luke xxiv. 51, καὶ ἀνεφέρετο εἰς τὸν ουρανόν, in the primitive text are omitted, with D (which has ávésn for diésn, the word immediately preceding them), and two very old Latin codices, a and b (Lachmann in loc.). Tischendorf thinks they were inserted in the Cod. Sinait., as it now stands, by a later hand, the same which supplied év 'Epéry, Ephes. i. 1 (p. 19, note).

Luke xxii. 43, 44, from pen to y, though omitted by A and B, and included in brackets by Lachmann, are confirmed a prima et tertid manu. Here, again, this codex coincides with D.

John viii. 1-11. Entire section of the woman taken in adultery is omitted, with A, B, C, and in accordance with the perfect silence of Origen in his commentary. D, on the other hand, has it, though in a somewhat varied form.

Matthew vi. 13, omits the doxology in the Lord's Prayer, with B and D.

Acts xx. 28, reads ἐκκλησ. τοῦ Θεοῦ, instead of του κυρίου, with B, and against A, C, D.

* We may here notice briefly, though it does not properly belong to the special subject of this paper, that besides this biblical codex, Tischendorf brought many other MS. treasures from the East, which have been partly deposited in the Imperial Library, partly consigned to the care of the Academy of Sciences at St. Petersburg. They consist of palimpsests, Greek MSS. uncial and cursive, Syriac, Coptic, Arabic, Turkish, Hebrew (containing a very ancient edition of rabbinical books), Samaritan, Slavonic, Abyssinian and Armenian writings, two rolls of papyrus accompanying a Greek astrolabe (of which he has given a lithographic representation), and some small sepulchral monuments of wood, of Egyptian origin and covered with Greek inscriptions. Among the Greek MSS. are fragments of the Old and New Testament text, which when duly printed and edited will supply fresh materials for textual criticism. The other MSS. promise to increase our knowledge of the early versions of the Scriptures and of the old liturgical and religious literature of the East. Tischendorf speaks of two copies of the Samaritan Pentateuch among these MSS. as very ancient. From the island of Patmos he brought away a fragment containing six books of Diodorus Siculus, with an addition supplying a gap in the extant text, and a copy of the scholia of Origen on the Proverbs of Solomon, which he has printed at the end of his Notitia, with critical notes.