Derek de Solla Price conjectured that the Antikythera Mechanism was engraved circa 87 B.C. More recently, Tony Freeth, Haralambos Kritzas, Mike Edmunds and their collaborators, using the forms of the Greek letters in the inscriptions, proposed that the mechanism was constructed around 150–100 B.C. But scholarly opinion remains unsettled on this issue. In this paper, we offer several approaches to determining the epoch, or intended starting date, of the Antikythera mechanism. The epoch is the “year 1” from which the ancient mechanic reckoned time for the purpose of the mechanism’s display. (The epoch is not necessarily the same as the date of manufacture, though it would be surprising if they were very widely separated in time.) As is well known, the first eclipse on the Saros dial was probably located at month 2, while the most natural position would be the first month. Also it is also easy to show that this is not the first eclipse of any 8- or 7-eclipse group. Both considerations suggest that the Saros cycle was intended to be coordinated with the Metonic cycle, in such a way that both cycles have the same starting date. If this is correct, then it would be possible in principle to calculate this date if there were sufficient other evidence to restrict the possible candidates. We will show that it is possible to introduce new constraints in order to greatly restrict the possible solutions. These constraints involve a) the discovery of new solutions (in addition to those that have been published) fitting the extant eclipse glyphs and consistent with the Babylonian Saros eclipse rules; b) a fruitful new interpretation of the omitted solar eclipses on the Saros dial; and c) physical evidence in Fragment A that can be used to infer simultaneously applicable values for several angles of the lunar theory. These angles need not be chronologically meaningful in themselves, but they may offer the possibility of “winding back” the least disturbed parts of the mechanism to a configuration consistent with the starting date. When all these approaches are used in conjunction, it is possible to offer a surprisingly small number of candidates for epoch.

This presentation is based on work co-authored by Christian Carman.