SEYFFARTH — OX THE THEORY OF THE MOON S MOTIONS. 479 



21. Diodor (xvi. 9), Plutarch (Nic. 23, p. 394 R., xad^ 6v 

 yp6\^oiy Yj/xe/J.su dpa^ ix Zaxvi^&ou tzasiu i~c Jcoi^'jatov ixAc-06- 

 ar^:; r^C ff£/3yv3jc oudku dearapa^dsc^ a^^r^yd^r^^ Plutarch (Dion. 

 24, //sra rac oTOvoa^ xat rac '^^I'OfJuaixha^ xaxvjya^ i^khntv ^ 

 acArji^rj), and Quinctilian (In. or. i. 10, 48, "Dion cum ad destru- 

 endam Dionysii tyrannidem venit, not est tali casu deterritus), 

 report that in Ol. 105. 4, in the course of the archonship of Aga- 

 thocles, a total eclipse of the moon happened in Sicily, soon after 

 sunset. This eclipse must have been total, because it is paralleled 

 with that of Nicias (No. 15) and called a -'terrible" one. About 

 that time only one lunar eclipse was possible soon after sunset, 

 viz. that in — 356, Aug, 9, 6h. 45m. P. T., 15 10° E., obscuration 

 2i inches (Pingre), or 4 inches (Calvisius). Since, however, the 

 longitude of the y was nearly 5° 14' (p. 429) shorter, this eclipse 

 was total indeed. Diodor refers the eclipse to the same year, 

 because Ol. 105,4, commenced, according to his Olympiads, with 

 July 2d in — 356. Even Calvisius recurred to the same eclipse, 

 because all the following eclipses disagreed with the ancient re- 

 ports. Agathocles ruled two years later ; but in — 354 no simi- 

 lar eclipse occurred, as Pingre demonstrates. 



22 & 23. The ancient authors erroneously refer two different 

 eclipses of the moon to the same battle near Arbela, for they refer 

 those eclipses to different Greek months and hours. Cicero and 

 Arian, who are the most reliable authorities, place the battle in 

 — 328, Sept. 10, and the eclipse preceding the latter in Aug. 29, 

 I2h. P. T. ; for Cicero (Div, i.53) says, "si lunapaullo ante solis 

 ortum defecisset et in signo Leonis, fore ut armis Darius et Persae, 

 praelio vmcerentur." Arrian (Exp. Al. iii. 7, 6, & 15, 7) reports 

 that the battle, 1 1 days after the eclipse, was fought both in the 

 month of Pyanepsion, i.e. in September (p. 408), and during the 

 archonship of Aristophanes; moreover, that the eclipse was a 

 partial one (r:^c 'xs/jjr/iyc ro TzoXb ixhizsQ iyevsTo). About that 

 time onlv one lunar eclipse coincided with sunrise near Arbela 

 (41° 40' E.), viz. that in — 328, Aug. 29, I2h., i.e. 3h.44m. after 

 midnight, Arbela time, and, according to our Table (p. 429-30), 

 about three hours later, 6h. 50m. local time. In consequence of 

 the parallax, the obscuration became visible at Arbela nearly two 

 hours earlier. Since Plutarch (Al. 31), however, reports that the 

 Persian army. 1 1 days after the eclipse, came in sight of Alexan- 



