35 The <I>Megillat Taanit</I> is a collection of ephemerides or calendars, indicating the days on which happy events occurred, and on which it is forbidden to fast. The little work, written in Aramaic, but enlarged by Hebrew glosses, is attributed by the Talmud to Hananiah ben Hezekiah ben Garon, or Gorion (first century); the nucleus about which the book was built up seems to go back as far as Maccabean times. 36 See Note 94. 37 Collection of texts not incorporated in the Mishnah, the order of which is followed, now to explain it, now to complement it, and sometimes to contradict it. The redaction of the Tosefta is attributed to R. Hiyyah bar Abba (third century). 38 When the aim of the Midrash is to interpret the legal and ritual portions of the Pentateuch, it is called Halakic; it is Haggadic when its aim is to interpret the narrative and moral portions (see chap. VI, p.107) - The Halakic Midrashim nevertheless contain much Haggadah. The redaction of the Mekilta, the commentary on Exodus, is attributed to R. Ishmael; that of the Sifra, or Torat Kohanim, the commentary on Leviticus, to R. Judah ben Ilai; that of the Sifre, the commentary on Numbers and Deuteronomy, to R. Simon ben Yohai and to the school of Rab, all scholars of the second and third centuries. The Sifra that Rashi employed was more complete than the one now available, and he cites a second Sifre, at present unknown. 39 The Midrash Rabba, or Rabbot, consists of Haggadic compilations on the Pentateuch and the Five Rolls; the elements of this Midrash are comparatively ancient, but its definite redaction without doubt does not go farther back than the eighth century. Rashi did not know those portions of the Midrash Rabba which explain the Books of Exodus and Numbers. 40 By this name are designated Haggadic collections for various distinguished times and seasons of the year. There are two Pesiktas, the Pesikta attributed to R. Kahana, a Babylonian Talmudist, though its redaction falls in the seventh century, and the Pesikta Rabbati, or Great Pesikta, doubtless compiled in Southern Italy in the ninth century. Rashi knew the first of these collections; and his citations aided Zunz in the reconstruction he made of this Midrash before the discovery of a manuscript by Buber confirmed his clear-sighted suppositions. 41 Name of a Midrash on the Pentateuch, redacted by the pupils of R. Tanhuma. Quite recently the endeavor was made to prove that Rashi did not know the Tanhuma either in the current text or in the more extended text published by Buber in 1885, and that he called Tanhuma the Midrash Yelamdenu, which is lost, and which is said to be the prototype of the two versions of the Tanhuma. See Grunhut, in <I>Festschrift Berliner,</I> pp.156-63.