onsequently the dates of the feasts), which the Rabbanites
fixed by calculation of the conjunctions, while the Karaites depended on
observation of the visible new moon. The ancients, he says, required
evidence of the appearance of the new moon.(95) Saadia, who mistakenly
assumed that the beginning of the month had been determined astronomically
from remote antiquity--the calendar was, in fact, of Sinaitic
origin(96)--asserted that the taking of testimony about the appearance of
the moon was an innovation occasioned by the contention of Zadok and
Boethus that the law required the beginning of the month to be determined
by actual observation; witnesses were heard only to prove that observation
confirmed the calculation. To this the author replies: "The book of the
Zadokites (Sadducees) is well known, and there is no such thing in it as
that man (Saadia) avers. In the book of Zadok are various things in which
he dissents from the Rabbanites of the second temple with regard to
sacrifices and other matters, but there is not a syllable of what the
Fayyumite (Saadia) says."(97) Saadia himself appears not to have
questioned the authenticity of the writings that went under the name of
Zadok, with which he seems to have been acquainted, directly or
indirectly, for in a passage quoted by Yefet ben 'Ali he says that Zadok
had proved from the one hundred and fifty days in the story of the flood
just the opposite of what the Karaites try to prove from them.(98)

Zadokite books thus meant, for all those from whom our information comes,
Sadducean books; and so, in the sense that, whatever their age and origin,
they contained substantially Sadducean teachings, most modern scholars,
also, have understood the name.

The possibility that Sadducean writings from the beginning of the
Christian era had survived to the Middle Ages cannot well be denied,
especially in view of the preservation of the book of the unknown sect
that forms the subject of our present study in copies as late as the tenth
or eleventh century; and even if the book which the Karaites took for
Sadducean was erroneously attributed to that sect, there is no sufficient
ground for identifying it with the texts in our hands or for ascribing it
to our sect. A thirty-day month, and the prohibition of divorce and of
marriage with a niece, are much too slender a foundation to support so
large an inference, and it is hardly legitimate to argue that if we had
the entire book, of whi