Codex Sinaiticus

Column 1

been. If this superscription be compared with the correction, doubt will be felt as to the identity of the hands, but further study of other superscriptions and corrections will be enough to show that probably both were written by the same hand, though the scribe varied his 'stroke'. This passage is interesting as suggesting that it was usual in a scriptorium for the corrector (original: διορθωτής) A professional proofreader or "reviser" in a manuscript workshop. to correct the text before the titles, subscriptions, and similar details were added; it also supports Dr. Dziatzko's contention that the custom of adding superscriptions began in the second half of the fourth century, and at first were not put in by the original scribe (see p. xiii). A further illustration of the same point is perhaps afforded by the first example classified on Plate II as doubtful. Here the scribe who added the Eusebian apparatus A system of cross-references for the Gospels developed by Eusebius of Caesarea. has placed the appropriate numerals against a correction in the lower margin. It is, however, doubtful whether this correction is from the hand of A² A specific corrector identified by the scholar Tischendorf., and there are in any case two possibilities; (1) that the correction was made before the Eusebian scribe did his work, (2) that the Eusebian scribe noticed that the beginning of his section was missing, and himself added it in the margin, affixing the proper number. I am inclined to think that the second is the correct view, as the writing of the correction seems to me to be identical neither with A² nor with A³, to both of which it has superficial resemblance. In this case we have in this correction a specimen of the writing of the Eusebian scribe: it is doubtful whether there are any others, though it is of course possible that some of the many corrections, which are too short to enable us to identify the hand, may be due to this scribe.

Tischendorf thought that he had found a convincing proof of the identity of A² with the original scribe D in the correction to Matthew 5:45 (folio 3, recto), given in the second column of Plate II. Here the large uncial A script written entirely in capital letters, common in early manuscripts. ends with the incomplete sentence and sends rain on (original: καὶ βρέχει ἐπὶ) and the missing words—the just and (original: δικαίους καὶ) the unjust (original: ἀδίκους)—are supplied in smaller letters. The smaller letters are certainly in the hand of A², and Tischendorf thought that the larger letters are in the hand of D. But there is not, in my judgement, any reason for thinking that he was right; there is no real difference in the style, and in any case D's script is so nearly the same as A's that these words afford no means of deciding between the two hands. Moreover, his argument that no one would break off in the middle of a sentence with on (original: ἐπὶ) is

xxii

Column 2

not sound: scribes were, and are, capable of any mistake, and an example of the same breaking off is to be found on folio 70, verso, column 3 (see Plate II, column 2), where the original scribe wrote the woman (original: ἡ γυνὴ) the un- (original: ἡ ἀπι) and then went to the next line, in the (original: ἐν τῷ) etc., omitting the rest of the word unbelieving (original: ἄπιστος). Nevertheless, that Tischendorf was right in identifying A² with the original scribe D is, I think, suggested clearly enough by the evidence of the superscriptions, and a true example of the kind he sought in Matthew 5:45 can be found in John 19:4 (folio 59, verso, column 3), where the correction no (original: οὐδεμίαν) fault I find (original: αἰτίαν εὑρίσκω) is written in a combination of large and small letters resembling the style of the scribe D and the corrector A² respectively (see Plate II, column 1).

A² must therefore be regarded as a script used by the original scribe D as an alternative to the large regular script of the text. Furthermore it is plain that the same scribe was the corrector (original: διορθωτής) of the manuscript, but if the argument given above be sound, he probably did not add the Eusebian apparatus. Moreover, it will probably be agreed on the ground of general appearance that he was not the scribe who added the line-counts (original: στίχοι) Stichoi: a measurement of the length of a text based on standard line units., and this view is rendered practically certain by the absence of the line-counts (original: στίχοι) from folio 88, which was written by D.

It will be convenient at this point to summarize the scribes who have so far been shown to have been working in the scriptorium, for none of the other early correctors can be proved with certainty to have belonged to it. The list is:—
Scribe A, wrote the greater part of the text, and subscriptions.
Scribe B, wrote the Shepherd of Hermas.
Scribe D, wrote the cancel-leaves, and superscriptions.
Corrector A¹, probably identical with scribe D.
Corrector A², almost certainly identical with scribe D.
Scribe E, the writer of the Eusebian apparatus.
Scribe S, the writer of the line-counts (original: στίχοι).

All these writers worked on the manuscript before it left the scriptorium. It is highly probable that the same is true of some of those which remain to be considered, but evidence is not forthcoming.

The corrector A³. A considerably smaller number of corrections have been made by a hand which cannot for any palaeographical reasons be regarded as later than A², but is clearly distinct from it. It is not marked by any alternation of size in the letters, or by the curious d (original: δ) of A², and is much