THE ECLIPSE 

A second line of inquiry has centred on the path and time of the eclipse itself. That Herodotus’ account concerned the eclipse 585 was understood by Pliny (NatHist 2.53) and rediscovered by Airy, the Astronomer Royal (Airy 1853). Roller, in a somewhat negative paper, (Roller 1983), suggested that Airy was perhaps influenced by Pliny, but Airy seems to have been more concerned with astronomy than with ancient history and made no reference to Pliny. Modern calculations of the exact time and precise path are not as accurate as might be hoped because of uncertainties caused by minor changes in the speed of the earth’s rotation. It is nevertheless certain that a total eclipse of the sun would have been visible from north central Anatolia in the late afternoon of May 28, 585 (most recently Stephenson and Fatoohi 1997). There is little to be gained by suggesting the eclipse was too close to sundown to have made an impact (Mosshammer 1981, 148), since the eclipse had obviously made a considerable impression, whether or not Herodotus was correct to associate it with the end of the battle. Given that the sky was clear a total eclipse, even shortly before sunset, would have had a profound impact on the observer.

In the course of our own observations of the August 11 eclipse in 1999 we were stunned by the suddenness of the totality and the fact there was no indication that a total eclipse was about to take place until a slight darkening and cool breeze shortly before totality itself. On the Cappadocian Plain even an eclipse late in the day would have bee seen as a sudden and awe-inspiring event. The maximum duration of totality during the the 585 eclipse was a fraction over six minutes (Espenak 1999), almost two and a half times longer than our view of the A.D. 1999 event.

Worthen has suggested that the whole passage is misunderstood. He would combine the battle that took place at night with the eclipse and retranslate Herodotus in such a way as to make the eclipse lunar (Worthen 1997). This approach requires much rewriting of Herodotus, denies any ancient tradition of Thales prediction of a solar eclipse and is strongly based on calculations concerning the path of totality for the 585 solar eclipse. Further, Worthen postulated a total lunar eclipse at sunset because lunar eclipses are not sudden events but take a considerable time to gradually reach totality. Worthen also rejects the 585 eclipse because his calculations suggest that it would not have been total anywhere near the battle and would not, therefore, have had the impact ascribed to it by Herodotus. But Fatoohi and Stephenson, as well as Espenak, produce maps showing the band of totality passing right across Central Anatolia (Stephenson and Fatoohi 1997, Espenak 1999).