Page:Historical eclipses.djvu/25

 covered by these annals, and there are numerous similar observations recorded down to modern times. It is found that thirty-two of the thirty-six eclipses of the Annals of Lu can be confirmed by astronomical computation. The other four never happened. But the whole style of the annals is such that none but a contemporary could or would have composed them, and it seems to me the easiest explanation of the four false eclipses that a watch was kept for eclipses at every new moon and that observers sometimes imagined that a part of 'the Sun was eaten', as the phrase went, when the phenomenon was not really astronomical at all. If an astronomer was liable to be executed for failing to observe an eclipse, he would probably take no risks, and would record anything that the civil authorities might be likely to regard as an eclipse.

Eclipse of Archilochus. Selecting only the eclipses which have excited some unusual interest, I come next to the eclipse of Archilochus, who wrote:

'Nothing there is beyond hope, nothing that can be sworn impossible, nothing wonderful, since Zeus father of the Olympians made night from midday, hiding the light of the shining sun, and sore fear came upon men.'

This eclipse must have been that of B.C. 648 April 6, the earliest fixed date in Greek history. Archilochus is known to have divided his life between Paros and Thasos. According to my solution, the eclipse was