Having now shown, as shortly and as intelligibly as we can, how the germs of Chinese development were sown at the dawn of true history, let us proceed to examine how far that history, as it has come down to us, contains within it testimony to its own truth. We shall revert to the description of wars and ambitions in due course; but, as so obscure a subject as early Chinese civilization is only palatable to most Western readers in small, varied, and sugared doses, we shall for the moment vary the nourishment offered, and say a few words upon eclipses.

Confucius, whose bald "Spring and Autumn" annals, as expanded by three separate commentators (one a junior contemporary of himself), is really the chief authority for the period 722-468 B.C., was born on the 20th day after the eclipse of the sun which took place in the 10th month of 552 B.C., or the 27th of the 8th moon as worked out to-day (for 1908 this means the 22nd September). Confucius himself records thirty-seven eclipses of the sun between 720 and 481, those of 709, 601, and 549 being total. Of course, as Confucius primarily recorded the eclipses as seen from his own petty vassal state of Lu in Shan Tung province (lat. 35" 40' N., long, 117" E.), any one endeavouring to identify these eclipses, and to compare them with Julian or Gregorian dates, must, in making the necessary calculations, bear this important fact in mind. It so happens that nearly one-third of Confucius' thirty-seven eclipses are recorded as having taken place between the two total eclipses of 601 and 549. This being so, I referred the list to an obliging officer attached to the Royal Observatory, who has kindly furnished me with the following comparative list:-

CONFUCIUS' DATE. OPPOLZER'S JULIAN DATE. B.C. 601, 7th moon. - -600, September 20. 
   " 599, 4th " - -598, March 5. 
   " 592, 6th " - -591, April 17. 
   " 575, 6th " - -574, May 9. 
   " 574, 12th " - -573, October 22. 
   " 559, 2nd " - -558, January 14. 
   " 558, 8th " - -557, June 29. 
   " 553, 10th " - -552, August 31. 
   " 552, 9th " 
   " 552, 10th " - -551, August 20. 
   " 550, 2nd " - -549, January 5. 
   " 549, 7th " - -548, April 19.

It will be observed that there is no Oppolzer's date to compare with the first of the two eclipses of 552; this is because I omitted to notice that there had been recorded in the "Springs and Autumns" two so close together, and therefore I did not include it in the list sent to the Observatory; but with the exception of the total eclipse of 601, all the other eclipses, so far as days of the moon and month go, are as consistent with each other as are modern Chinese dates with European (Julian) dates. As regards the year, Oppolzer's dates are the "astronomical" dates, that is, the astronomical year - x is the same as the year (x + 1) B.C.; or, in other words, the year of Christ's birth is, for certain astronomical exactitude purposes, interpolated between the years 1 B.C. and A.D. 1, as we vulgarly compute them: that is to say, the eclipses of the sun recorded 2,400 years ago by Confucius, from notes and annals preserved in his native state's archives as far back as 700 B.C., are found to be almost without exception fairly correct, with a uniform "error" of about one month, despite the fact that attempts were made by the First August Emperor to destroy all historical literature in 213 B.C. This being so in the matter of a dozen eclipses, there still remain two dozen for specialists to experiment upon, not to mention comets and other celestial phenomena. From this collateral evidence, imperfect though it be, we are reasonably entitled to assume that the three expanded versions of Confucius' history are trustworthy, or at the very least written in the best of faith.