It is a midrash comprising a collection of ancient rabbinical homiletic interpretations of the Book of Numbers (Bamidbar in Hebrew).
It is the latest component of Midrash Rabbah on the Torah, and as such was unknown to Nathan ben Jehiel (c. 1035–1106), Rashi (1040–1105), and Yalkut Shimoni.
M. Beneviste drew attention as early as 1565 to the fact that Tanchuma and Numbers Rabbah are almost identical from the section Behaalotecha onward.
This is true also, with the exception of the interpretation of the numerical value of the Hebrew word for fringes, of the other passages pointed out by Leopold Zunz as originating with later, and notably French, rabbis.
This numerical interpretation of “fringes” forms a part of a passage, also otherwise remarkable, at the end of the section on Korach (18:21), which, taken from Numbers Rabbah, was interpolated in the first printed edition of the Tanchuma as early as 1522, but is absent from all the manuscripts.
Another long passage (18:22) which belongs to the beginning of Chukat, as in Tanchuma, is erroneously appended in the editions to the section on Korach.
According to Epstein, some unknown author wrote the Midrash upon the Torah portion Bamidbar to complete the Sifre, which commences with Numbers 1:1, another then continued it with the commentary on Naso, and in order to complete the work for the remainder of Numbers, the commentary for the remaining Torah portions was drawn from Tanchuma.
Even the first part contains much that is taken from the Tanchuma, but, as Zunz wrote, "a copious stream of new Haggadah swallows the Midrash drawn from this source and entirely obscures the arrangement of the Yelamdenu."