THE NEW SCHAFF-HERZOG
gint which confirmed the Hebrew, he made the two the ground text, and marked changes by diacritical signs. It is pardonable that he took his Hebrew text — whence he got it is not known — as the original text; but it was contrary to sound criticism to take those readings of the Septuagint which agreed with the Hebrew for the true ones, instead of those which differed from it (cf. the third axiom of Lagarde for the restoration of the Septuagint, Mitiheilungen, i, 21). Nevertheless we should be extremely thankful if the work of Origen had been preserved. Until 1896 it was known only from the descriptions of Eusebius, Jerome, Epiphanius, and some later writers, and by specimens preserved in scholia of Biblical manuscripts, a great part also by a literal Syriac translation (see below, § 6). In 1896 Giovanni Mercati discovered in a palimpsest of the Ambrosian Library of Milan the first continuous fragments of a copy of the Hexapla, and in 1900 another and much older piece was found by C. Taylor among the Greek palimpsests from the Cairo genizah in the Taylor and Schechter collection. These fragments show that Origen put generally only one Hebrew word, or at the most two, in one line; the extent of the work, therefore, must have been much greater than was previously supposed. The later fate of the original is unknown. Jerome saw and used it in the library at Cffisarea; it may have been destroyed there during the invasion of the Arabs.
Origen arranged his work in six columns, the first containing the Hebrew text in Hebrew letters, the second the same in a Greek transcription, the third the translation of Aquila, the fourth that of Symmachus, the fifth the Septuagint, the sixth the translation of Theodotion. For some books, especially the Psalms, Origen had a fifth, sixth, and even a seventh translation at his disposal (see below, 2, 5 3). In the Septuagint column he used the system of diacritical marks which was in use with the Alexandrian critics of Homer, especially Aristarchus, marking with an obelus — under different forms, as — , called lemniscus, and — , called hypolemniscus — those passages of the Septuagint which had nothing to correspond in Hebrew, and inserting, chiefly from Theodotion under an asterisk (*), those which were missing in the Septuagint ; in both cases a metobelus ( y ) marked the end of the notation. This column was copied afterward with additional excerpts from the other versions on the margins; and, if it had been copied with all its critical marks, it would have been well, but later copyists neglected these completely and produced what we may call kryptohexaplaric manuscripts, completely spoiling by this carelessness the value of the Septuagint for critical purposes. Such a copy, for instance, is, for Kings, the Codex Alexandrinus ; and it is but a poor defense of these copyists that the same process has been repeated in the nineteenth century by the Moscow and Athens reprints of Grabe's edition of that codex.
After Origen, Eusebius and his friend Pamphilus were careful to continue or disseminate his exegetical labors. Copies of the Pentateuch arc known which were compared with the Samaritan text
(cf. S. Kohn, Samareitikon und Septuaginta, in Monatsschrift fur Wissenschaft des Judenthums, new series, i, 1894, pp. 1-7, 49-67; ZDMG, 1893, p. 650). Jerome mentions besides Eusebius and Pamphilus, Lucian and Hesychius, the text of the former being used from
5. Lucian and
Hesychius Constantinople to Antioch, that of the latter in Alexandria and Egypt, while the provinces between, especially Palestine, kept to the copies of Origen as published by Eusebius and Pamphilus (Prwfatio in parallpomena ; Adv. Rufinum, ii, 27). About neither the work nor the person of Hesychius (see Hesychius, 1) is there complete certainty. He may have been the martyr bishop mentioned by Eusebius (Hist, eccl., viii, 13) together with Phileas of Thmuis. The result of his labors is sought now for the Octateuch in the manuscripts 44, 74, 76, 84, 106, 134; for the prophets, especially Isaiah and the Twelve, in the Codex Marchalianus and its supporters 26, 106, 198, 306 (cf. N. McLean, in JTS, ii, 1901, p. 306, and A. Ceriani, De Codice Marchaliano, Rome, 1890, pp. 48 sqq., 105 sqq.). Lucian was a deacon of Antioch, who died a martyr at Nicomedia 312 (see Lucian the Martyr) . He must have known a Hebrew text which showed many peculiarities, especially in the historical books, and perhaps used for his purposes the Syriac version. The first part of his work has been edited by Lagarde in Librorum Veteris Testamenti canonicorum, pars prior, grace (Gottingen, 1883; cf. his Mittheilungen, ii, 171). But this revision must not be confounded with the original Septuagint any more than the English Revised with the Authorized Version. Since the fourth century very little has been done in the Greek Church for its Bible. Emperors directed beautiful copies of it to be written — e.g., Constantine ordered fifty copies through Eusebius for the new churches of his capital, and for Constans Athanasius procured " copies of the divine writings," one of which is perhaps preserved in the famous Codex Vaticanus. Other royal persons wrote them with their own hands.
Latin was probably the first language into which the Septuagint was translated. (On the Latin version, or rather versions, of the Septuagint see below, II, 1. It is a pity that so little of these labors has been preserved, and that these few remnants are so difficult of access.) After the Latin versions came
6 Versions the E§yPtian (see VI1)- Here the Made difficulty of the language makes from the these helps for restoration of the Septuag-int. Septuagint accessible to few. Similar is the case with the most neglected branch of the Semitic languages, the Ethiopic (see VIII). The Arabic versions (see B, I) are for a great part too late to have much weight for the critic of the Septuagint. The Gothic version (see X)is an outcome of the Lucianic recension, for which it would have great importance, both for age and literalness, but very little of the Old Testament is preserved in Gothic. The Lucianic recension is also the basis of a Slavonic version (seeB, XVI) and through it of the Georgian (see IX).