No. 1239. The end of the work is wanting in this copy. The second part has forty-nine chapters, as the introduction to Part II. is counted as chapter i.; Part III. has fifty-six chapters, the introduction being counted as chapter i., and chapter xxiv. being divided into two chapters. The index of passages from the Pentatcuch follows the ordinary mode of counting the chapters of the Guide.
No. 1240. Arabic text transcribed in Arabic characters by Saadiah b. Levi Azankot for Prof. Golius in 1645.
No. 1245. First part of the Datalas al-hairis, written by Saadiah b. Mordecai b. Mosheh in the year 1431.
No. 1242 contains the same Part, but incomplete. Nos. 1243, 1244, 1245, and 1246 contain Part II. of the Arabic text, incomplete in No,. 1245 and 1246.
Nos. 1247, 1248, and 1249 have Part III.; it is incomplete in Nos. 1248 and 1249. No. 1249 was written 1291, and begins with III, viii. A fragment of the Arabic text, the end of Part III., is contained in No. 407, 2.
No. 2508 includes s fragment of the original (I. ii.-xxxii.), with a Hebrew interlinrary translation of some words and a few marginal notes. It is written in Yemen square characters, and is marked as "holy property of the Synagogue of Alsiani."
A fragment (I. i.) of a different recension from the printed is contained in 2422, 16. On the margin the Commentaries of Shein-tob and Ephodi are added in Arabic.
A copy of the Datalat is also contained in the Berlin Royal Library MS. Or. Qu., 579 (so; Cat. Steinschneider); it is defective in the beginning and at the end.
The Cairo Genizah at Cambridge contains two fragments (a) I.lxiv. and beginning of lxv; (b) II. end of xxxii. and xxxiii. According to Dr. H. Hirschfeld, Jewish Quarterly Review (vol. xv. p. 677), they are in the handwriting of Maimonides.
The valuable collection of MSS. in the possession of Dr. M. Gaster includes a fragment of the Dalalat-al-hairin (Codcx 6o5). II. xiii--xv., beginning and end defective.
II. Translations, a. Hebrew.--As soon as European Jews heard of the existence of this work, they procured its translation into Hebrew. Two scholars, independently of each other, undertook the task: Samuel Ibn Tibbon and Jehudah al-Harizi. There is, besides, in the Moreh ha-moreh of Shemtob Paiquera an original translation of some portions of the Moreh. In the Sifte yeshenim (No. 112) a rhymed translation of the Dalalat by Rabbi Mattityahu Kartin is mentioned. Ibn Tibbon s version is very accurate; he sacrificed elegance of style to the desire of conscientiously reproducing the author's work, and did not even neglect a particle, however unimportant it may appear. Ibn Tibbon went in his anxiety to retain peculiarities of the original so far as to imitate its ambiguities, e.g., meziut (I. lviii.) is treated as a masculine noun, only in order to leave it doubtful whether a pronoun which follows agrees with meziut, "existence," or with nimza, "existing being," both occurring in the same sentence (Br. Mus. MS. Harl. 7586, marg. note by Ibn Tibbon). When he met with passages that offered any difficulty he consulted Maimonides. Harizi, on the other hand, was less conscientious aboot words and particles, but wrote in a superior style. Vox populi, however, decided in favour of the version of Ibn Tibbon, the rival of which became almost forgotten. Also Abraham, the son of Moses Maimooides, in Milbamoth ha-shem, describes Harizi's version as being inaccurate. Most of the modern translations were made from Ibn Tibbon's version. There are, therefore, MSS. of this version almost in every library containing collections of Hebrew books and MSS. It has the title Moreh-nebuchim. The British Museum has the following eight copies of Ibn Tibbon's version.:--
Harl. 7586 A. This codex was written in tlse year 1284, for Rabbi Shabbatai ben Rabbi Mattityshu. In the year 1340 it came into the possession of Jacob b. Shelomoh; his son Menabem sold it in the year 1378 to R. Mattityaho, son of R. Shabbatai, for fifty
