diacal sign to which the imaginary hands happen to be pointing,
just as clock-time is indicated by the minutes read from the long
hand and the hours from the short. The sexagenary cycle came into use
in China in 623 B.C. The exact date of its importation into Japan is
unknown, but it was probably about the end of the fourth century A.D.
It is a sufficiently accurate manner of counting so long as the tale
of cycles is carefully kept, but any neglect in that respect exposes
the calculator to an error of sixty years or some multiple of sixty.
Keen scrutiny and collation of the histories of China, Korea, and
Japan have exposed a mistake of at least 120 years connected with the
earliest employment of the sexagenary cycle in Japan.

The fourth method corresponds to that adopted in Europe where the
number of a year is referred to the birth of Christ. In Japan, the
accession of the Emperor Jimmu--660 B.C.--is taken for a basis, and
thus the Occidental year 1910 becomes the 2570th year of the Japanese
dynasty. With such methods of reckoning some collateral evidence is
needed before accepting any of the dates given in Japanese annals.
Kaempfer and even Rein were content to endorse the chronology of the
Chronicles--the Records avoid dates altogether--but other Occidental
scholars* have with justice been more sceptical, and their doubts
have been confirmed by several eminent Japanese historians in recent
times. Where, then, is collateral evidence to be found?

*Notably Bramsen, Aston, Satow, and Chamberlain.

In the pages of Chinese and Korean history. There is, of course, no
inherent reason for attributing to Korean history accuracy superior
to that of Japanese history. But in China the habit of continuously
compiling written annals had been practised for many centuries before
Japanese events began even to furnish materials for romantic
recitations, and no serious errors have been proved against Chinese
historiographers during the periods when comparison with Japanese
annals is feasible. In Korea's case, too, verification is partially
possible. Thus, during the first five centuries of the Christian era,
Chinese annals contain sixteen notices of events in Korea. If Korean
history be examined as to these events, it is found to agree in ten
instances, to disagree in two, and to be silent in four.* This record
tends strongly to confirm the accuracy of the Korean annals, and it
is further to be remembered that the Korean peninsula wa