(2) Latin Versions. — (a) Vetus Itala or Old Latin, which Ilgen, in 1800, coiTectly surmised was based on a then unknown Gr. text, which has proved to be that of K. Tliougli all codices of this Version agree substantially with N, there are clearly three recen- sions, (a) It. I., the text edited by Sabatier (Bihli- orum sacrorum Latinw versioties antiqiKr, Paris, 1751) and by Neubauer (in his excellent little work, The Book of Tobit ). It is based on a Parisian Codex, Regius 3654, and on Cod. 4 in the library of S. Germain. (^) It. II., a text found in Cod. Vat. 7 which contains only 1-6'-, and once belonged to queen Christina of Sweden. It was collated by Sabatier in the above work, and was edited by Bianchini, Rome. 1740. (7) Fragments of a third recension (It. III.) are given in the Spccuhcm of Augustine, edited by Mai {Spicilegium, ix.).— (6) The Vulgate. Jerome afbrms that lie translated Tobit in one day from the Syro-Chaldee. As he was not familiar mth this language, a Jew, who knew both languages, translated it for him into Hebrew, from wliich he made his Latin transla- tion. There are many readings in Vulg. that were not found in any otlier text, until Gaster, 18<J6, discovered a Heb. MS, which in tlie narrative, as distinct from tlie exhortations and prayers, agrees in the main with Vulg. (see below, HL).
(3) Hi/riac Version. — This has been edited by Walton in his Polyglot ; and by Lagarde in Libri apokr. Syriace. As far as 7" it is a close transla- tion of B. After that, it agrees with K or the Gr. cursives. It lacks 13""'^
(4) Chaldee or Aramnic Ver.tion (Aram.). — This was first edited by Neubauer from a collection of Midrashim, copied in the 15th cent, in Greek- rabbinical characters. The Book of Tobit is an extract from the Midrash rabbah-de-rabbnh on Genesis, and forms a liaggada on Jacob's promise to give a tenth of his proceeds to God (Gn ii-'). Neubauer tliinks that the Chaldee text of Jerome was Aram, in a fuller form ; but in the view of the present writer there are facts which seem to imply that the Aram, is a translation from the Greek. The facts that the dat. 'Pd7ois (4' 5°) is found in Aram, as [?•:■!, and 'EK/Jardvois (3' 6') as D':n3;i<, and the ace. liypi.t' (6') as I'li'B ; and that the Gr. words 6.piaTov (2') and a-q^Luov (5-) are transliterated in Aram., atl'ord strong proof that Aram, is based on a Greek text : not on X (as Schiirer), for Aram, agrees more often with B than with N ; but on a briefer text tlian either, and more free from Christian influences.
(5) Hebrew Versions. — (a) Heb. Munsteri (HM), BO called because it was published, with a Lat. tr., by Seb. MUnster, at Basle, in 1542. The first edition, however, was printed at Constantinople in 1516. It is included in Walton's Polyglot, and also in Neubauer's Tobit. Neubauer gives, in the foot- notes, various readings from No. 1251 of the Heb.
MSS in the National Library at Paris: from a Persian tr. from the Heb. which is No. 130 in the same Library ; and No. 104 of de Rossi's catalogue, at Parma. It is noteworthy that HM usually agrees with Aram, when tlie latter dissents from the Greek. In chs. 12. 13, where Aram, is lacking, HM presents an eclectic text, agreeing in the main with SjT., but for 13^-'» it has an original and very brief doxology, and omits ch. 14 altogether. Gins- burg assigns it to the 5th century.— (6) Heb. Fagii (HF). This is a free, independent translation, made perhaps in the 12th century. The translator was a learned Jewish scholar, fond of precise, technical terms ; very familiar with the Heb. Bible, and fond of introducing suitable Bible texts, and of reducing the text of Tobit to biblical
phraseology. Tiiis is also given in Walton's Poly- glot.—(c) Heb. Londinii (HL) is a text found by Gaster in the British Museum, Add. 11,639. A description and translation of the MS, which lielon"S to the 13th cent., is given by Gaster in PSBA, vol. xviii. 208 ff., 259 ff., and vol. xx. 27 ff. So far as the exhortations, prayers, and doxologies are concerned, they are certainly late. They develop, in a remarkable decree, the tendency observable in HF to reduce the text to biblical phraseology. In the exhortations, etc., HL gives us a cento of Scripture texts, skilfully selected as being most cognate to the Gr. text. As to the narrative, it is intensely interesting to note how closely HL agrees with Vulg., and Gaster claims for the MS as a whole a close relationship to the ' Syro-Chaldee ' used by Jerome. As to the narra- tive portions, tlie author of HL certainly nuit/ have used an Aramaic or Heb. text closely related to Jerome's 'Syro-Chaldee,' though, if the doxologies, etc., are of late composition, one cannot escape the unpleasant surmise that HL may be drawn from the Vulg. itself.— (rf) Heb. Gasteri (HG). Thiswas copied some years ago by Gaster from a Midrash on the Pentateuch, which he fears has now perislied. It is a condensation in Heb. of the narrative por- tions of Aram., witli the exhortations, prayers, and doxologies rigorously excluded, ;ind all approach to verbosity in tlie narrative sternly checked. It is possible that tlie author of HL may have possessed a similar History, exhibiting those peculiarities of the Vulg. which, until tlie publication by Gaster of the translation of HL, were considered unique in tlie Vulgate. The tr. of HG is given in PSBA vol. xix. 33 f. Its agreements with Aram, are very siguiticant.