throwing back, no doubt more or less unconsciously, to more ancient times the ideas of their own age, when the national thought and institutions had become deeply modified by Chinese influences. As Dr. Florenz very justly remarks, "The little which European inquiry has hitherto been able to teach us of the real condition of Japan in the most ancient times shows that the historical representation of this period in the Kojiki and Nihongi (upon which rest all the later statements of the Japanese) is most profoundly penetrated by false principles. The newer relations, partly developed from native material, partly influenced by Chinese culture, are reflected back upon the oldest without due distinction, and the result is a confused picture in which the critical inquirer can, it is true, frequently separate what is original from subsequent additions, but must often let fall his hands in despair." A conspicuous instance of this is the way in which the Imperial theory of the universal authority of the Mikados is extended backwards to a time when their sway was really restricted to the provinces round the capital and a few other places. It is also exemplified by the treatment of territorial and official designations in the older part of the history as if they were already family names, which they did not become until a later period.
Chronology.—The Kojiki wisely has no chronology. But the authors of the Nihongi, or more probably of some of the works on which it is based, thought it necessary, in imitation of their Chinese models, to provide a complete system of dates extending as far back as the middle of the 7th century B.C., and giving the exact years, months, and even days for events which are supposed to have happened in this remote period. When it is remembered that there was no official recognition of the art of writing in Japan until A.D. 405, and that the first mention of calendar-makers belongs to A.D. 553, the historical value of such chronology may be readily estimated. After the Christian epoch there may have been some blundering and unsuccessful endeavours to give the right years, but for several centuries longer the months and days must have been simply supplied from the writers' imagination. Even so late as the beginning of the 5th century the chronology can be shown to be wrong in several cases by no less an interval than 120 years. Abundant proofs of its inaccuracy are revealed by a
