(M40) But the exact date at which the Proerosia or Festival before Ploughing took place is somewhat uncertain, and enquirers are divided in opinion as to whether it fell before or after the Great Mysteries, which began on the fifteenth or sixteenth of Boedromion, a month corresponding roughly to our September. Another name for the festival was Proarcturia, that is, “Before Arcturus,”(176) which points to a date either before the middle of September, when Arcturus is a morning star, or before the end of October, when Arcturus is an evening star.(177) In favour of the earlier date it may be said, first, that the morning phase of Arcturus was well known and much observed, because it marked the middle of autumn, whereas little use was made of the evening phase of Arcturus for the purpose of dating;(178) and, second, that in an official Athenian inscription the Festival before Ploughing (_Proerosia_) is mentioned immediately before the Great Mysteries.(179) On the other hand, in favour of the later date, it may be said that as the autumnal rains in Greece set in about the middle of October, the latter part of that month would be a more suitable time for a ceremony at the opening of ploughing than the middle of September, when the soil is still parched with the summer drought; and, second, that this date is confirmed by a Greek inscription of the fourth or third century B.C., found at Eleusis, in which the Festival before Ploughing is apparently mentioned in the month of Pyanepsion immediately before the festival of the Pyanepsia, which was held on the seventh day of that month.(180) It is difficult to decide between these conflicting arguments, but on the whole I incline, not without hesitation, to agree with some eminent modern authorities in placing the Festival before Ploughing in Pyanepsion (October) after the Mysteries, rather than in Boedromion (September) before the Mysteries.(181) However, we must bear in mind that as the Attic months, like the Greek months generally, were lunar,(182) their position in the solar year necessarily varied from year to year, and though these variations were periodically corrected by intercalation, nevertheless the beginning of each Attic month sometimes diverged by several weeks from the beginning of the corresponding month to which we equate it.(183) From this it follows that the Great Mysteries, which were always dated by the calendar month, must have annually shifted their place somewhat in the solar year; whereas the Festival before Ploughing, if it was indeed dated either by the morning or by the evening phase of Arcturus, must have occupied a fixed place in the solar year. Hence it appears to be not impossible that the Great Mysteries, oscillating to and fro with the inconstant moon, may sometimes have fallen before and sometimes after the Festival before Ploughing, which apparently always remained true to the constant star. At least this possibility,