HISTORY OF PHYSICAL ASTRONOMY.

868

calculation that the utmost possible duration of an annular eclipse underthe equator, is 12 m 24 s *, and under the latitude of Paris is 9 m 56 s f.

The writings of ancient authors contain some interesting allusions toeclipses, both of the sun and moon. The most remarkable is a totaleclipse of the sun, which Herodotus relates to have happened while theMedes and Lydians were actually engaged in battle. According to thestatement of the Greek historian, “the war between the two nations hadcontinued during five years with alternate advantages to either party. Inthe sixth there was a nocturnal combat; for after an equal fortune on bothsides, and while the two armies were engaged, the day suddenly becamenight.” | He mentions that Thales the Milesian had predicted this pheno-menon to the loniaus, and had ascertained the time of the year in whichit would happen. He then adds that the Lydians and the Medes, seeingthat the night had thus taken the place of day, desisted from the combat,and became desirous, on both sides, of making peace.

This remarkable eclipse is alluded to by several ancient writers whoflourished subsequently to Herodotus . As the last-mentioned author hasnot assigned the time of its occurrence, the ascertainment of this pointhas given rise to a good deal of discussion. Cicero and Pliny both con-cur in asserting that it happened in the fourth year of the forty-eighthOlympiad. This would place it in the year 589 a.c. This date has beenadopted by Pdccioli, Newton, and various other authorities of modern times.On the other hand, Scaliger found by actual calculation that it happenedon the 1st of October, 583 a.c.; Usher placed it in the year 601 a.c.;Bayer, Costard, and several others, have maintained that a total eclipseof the sun, which appears from theory to have happened on the 18th ofMay, 603 a.c.., was the one alluded to by Herodotus .

The question with respect to the date of the eclipse, continued in thisstate of uncertainty until the year 1811, when Baily finally communicateda paper to the Boyal Society containing its true solution. By a skilfulcriticism of the passage in which the Greek historian alludes to the event,he has proved beyond all doubt that it could not have happened earlierthan 629 a.c., or later than 595 a.c. Out of seventy eclipses which hap-pened within that period, he found only one which was total in thepeninsula of Asia Minor . That eclipse happened on the 30th of Septem-ber, 610 a.c. It was central and total to part of Asia Minor, Armenia ,and Media, and the moon’s shadow passed over the very locality wherethe two armies most probably were engaged §. It seems impossible towithhold the conclusion that this is the eclipse mentioned by the Greekhistorian as having produced so memorable an impression on those whowere spectators of it.

Herodotus also mentions a total eclipse of the sun, which happenedwhen Xerxes was advancing with his army from Sardis to Abydos . Hestates that the sun became invisible, although the heavens everywherepresented a serene and cloudless aspect; and, that in consequence, nighttook the place of day [|. The date of this eclipse has been referred to theyear 480 a.c. The question of its actual occurrence is liable, however, tosome doubt, arising from a difficulty experienced in reconciling the state-ment of the historian with calculations founded on the solar and lunartables.

Thucydides the historian alludes to an eclipse of the sun which hap-

* Mem. Acad, des Sciences, 1777, p. 317. f Ibid., 1777, p. 316.

4 Herod., lib. i., sec. 74. § Phil. Trans., 1811, p. 220, et seq. |] Herod. , lib. vii.