[443] Herodot. i, 74-103.
[444] Compare the analogous case of the prediction of the coming olive crop ascribed to Thalês (Aristot. Polit. i, 4, 5; Cicero, De Divinat. i, 3). Anaxagoras is asserted to have predicted the fall of an aërolithe (Aristot. Meteorol. i, 7; Pliny, H. N. ii, 58; Plutarch, Lysand. c. 5).
Thalês is said by Herodotus to have predicted that the eclipse would take place “in the year in which it actually did occur,”—a statement so vague that it strengthens the grounds of doubt.
The fondness of the Ionians for exhibiting the wisdom of their eminent philosopher Thalês, in conjunction with the history of the Lydian kings, may be seen farther in the story of Thalês and Crœsus at the river Halys (Herod. i, 75),—a story which Herodotus himself disbelieves.
[445] Consult, for the chronological views of these events, Larcher ad Herodot. i, 74; Volney, Recherches sur l’Histoire Ancienne, vol. i, pp. 330-355; Mr. Fynes Clinton, Fasti Hellenici, vol. i, p. 418 (Note ad B. C. 617, 2); Des Vignoles, Chronologie de l’Histoire Sainte, vol. ii, p. 245; Ideler, Handbuch der Chronologie, vol. i, p. 209.
No less than eight different dates have been assigned by different chronologists for this eclipse,—the most ancient 625 B. C., the most recent 583 B. C. Volney is for 625 B. C.; Larcher for 597 B. C.; Des Vignoles for 585 B. C.; Mr. Clinton for 603 B. C. Volney observes, with justice, that the eclipse on this occasion “n’est pas l’accessoire, la broderie du fait, mais le fait principal lui-même,” (p. 347:) the astronomical calculations concerning the eclipse are, therefore, by far the most important items in the chronological reckoning of this event. Now in regard to the eclipse of 625 B. C., Volney is obliged to admit that it does not suit the case; for it would be visible only at half-past five in the morning on February 3, and the sun would hardly be risen at that hour in the latitude of Media and Lydia (p. 343). He seeks to escape from this difficulty by saying that the data for the calculation, according to the astronomer Pingré, are not quite accurate for these early eclipses; but after all, if there be error, it may just as well be in one direction as in another, i. e. the true hour at which the eclipse would be visible for those latitudes is as likely to have been earlier than half-past five A. M. as to have been later, which would put this eclipse still more out of the question.
The chronology of that period presents difficulties which our means of knowledge hardly enable us to clear up. Volney remarks, and the language of Herodotus is with him, that not merely the war between Kyaxarês and Alyattês (which lasted five years, and was terminated by the eclipse), but also the conquest made by Kyaxarês of the territory up to the river Halys, took place anterior (Herodot. i, 103: compare i, 16) to the first siege of Nineveh by Kyaxarês,—that siege which he was forced to raise by the inroad of the Scythians. This constitutes a strong presumption in favor of Volney’s date for the eclipse (625 B. C.) if astronomical considerations would admit of it, which they will not. Mr. Clinton, on the other hand, puts the first siege of Nineveh in the very first year of the reign of Kyaxarês, which is not to be reconciled with the language of Herodotus. In placing the eclipse, therefore, in 603 B. C., we depart from the relative arrangement of events which Herodotus conceived, in deference to astronomical reasons: and Herodotus is our only authority in regard to the general chronology.
According to Ideler, however (and his authority upon such a point is conclusive, in my judgment), astronomical considerations decisively fix this eclipse for the 30th September 610 B. C., and exclude all those other eclipses which have been named. Recent and more trustworthy calculations made by Oltmanns, from the newest astronomical tables, have shown that the eclipse of 610 B. C. fulfils the conditions required, and that the other eclipses named do not. For a place situated in 40° N. lat. and 36° E. long. this eclipse was nearly total, only one-eightieth of the sun’s disc remaining luminous: the darkness thus occasioned would be sufficient to cause great terror. (Ideler, Handbuch, l. c.)
Since the publication of my first edition, I have been apprized that the late Mr. Francis Baily had already settled the date of this eclipse to the 30th of September 610 B. C., in his first contribution to the Transactions of the Royal Society as long ago as 1811,—much before the date of the publication of Ideler’s Handbuch der Chronologie. Sir John Herschel (in his Memoir of Mr. Francis Baily, in the Transactions of the Royal Astronomical Society, vol. xv, p. 311), after completely approving Mr. Baily’s calculations, and stating that he had been the first to solve the disputed question, expresses his surprise that various French and German astronomers, writing on the same subject afterwards, have taken no notice of “that remarkable paper.” Though a fellow-countryman of Mr. Baily, I am sorry that I have to plead guilty to a similar ignorance, until the point was specially brought to my notice by a friend. Had I been aware of the paper and the Memoir, it would have been unnecessary to cite any other authority than that of Mr. Baily and Sir John Herschel.
[446] Herodot. iv, 11-12. Hekatæus also spoke of a town Κιμμερίς (Strabo, vii, p. 294).