save that the account is mutilated at beginning and end.

Boissevain concludes (by comparing the Table of Contents found with a
remark of Photius) that this particular piece of salvage was
originally Books Seventy-nine and Eighty (instead of Seventy-eight and
Seventy-nine), that Book Eighty of Dio was really what is now
commonly called Seventy-nine _and_ Eighty, and that the so-called Book
Eighty (of only five chapters) was but a kind of epilogue to the whole
work. Whatever we may decide respecting the merits of his argument,
the important fact is that here for a short distance we have Dio's
original narrative, as in Books Thirty-six to Sixty, and are no longer
obliged to depend upon epitomes.

A word of explanation about Xiphilinus must come next. This Xiphilinus
was a native of Trapezos (Trebizond) and became a monk at
Constantinople. Here, at the behest of Michael VII. Ducas (1071-1078)
he made an abridgment of Books Thirty-six to Eighty of Dio; thus it is
his version of the lost books Sixty-one to Eighty on which we are
compelled to rely. His task was accomplished with an even greater
degree of carelessness than is customary in such compositions, and it
may be said that his ability or, at least, his good will is not nearly
so great as that of Zonaras. Yet he is largely a _pis aller_ for the
would-be reader of Cassius Dio.

Whereas the original was divided arbitrarily into books, Xiphilinus
divided his condensation into "sections," each containing the life of
one emperor. Readers must further note that the present division of
Books Seventy-one to Eighty dates only from Leunclavius (1592, first
edition) and is not necessarily correct. Improvements in arrangement
by Boissevain (latest editor of Dio entire) are indicated in the