Nor is this all. It is hard to understand some of the folklore and tradition unless we recognise that at a time before marriage was instituted, at some of the sacred festivals the intercourse of the sexes was permitted if not encouraged. This view is strengthened by the researches of Westermarck[128] and Rhys.[129] Given such a practice, the origin of matriarchal customs and of the _couvade_ is at once explained; and it is clear that the charges against the Druids of special cruelty and impurity must be withdrawn. Their sacrifices and customs were those common to all priesthoods in the ancient world. I have shown that some circles used in the worship of the May year were in operation 2200 B.C., and that there was the introduction of a new cult about 1600 B.C., or shortly afterwards, in southern Britain, so definite that the changes in the chief orientation lines in the stone circles can be traced. To the worship of the sun in May, August, November and February was added a solstitial worship in June and December. The associated phenomena are that the May-November Balder and Beltaine cult made much of the rowan and may thorn. The June-December cult brought the worship of the mistletoe. The flowering of the rowan and thorntree in May, and their berries in early November, made them the most appropriate and striking floral accompaniments of the May and November worships, and the same ideas would point to a similar use of the mistletoe in June and December. The fact that the June-December cult succeeded and largely replaced the May-November one could hardly have been put in a cryptic and poetic statement more happily than it appears in folklore: Balder was killed by mistletoe. This change of cult may be due to the intrusion of a new tribe, but I am inclined to attribute it to a new view taken by the priests themselves due to a greater knowledge, among it being the determination, in Egypt, of the true length of the year which could be observed by the recurrence of the solstices, and of the intervals between the festivals reckoned in days. However this may have been, all the old practices and superstitions were retained, only the time of year at which they took place was changed. As the change of cult was slow, in any one locality the celebrations would be continued at _both_ times of the year, and for long both sets of holidays were retained. Since I have shewn that the solstitial worship came last, traces of this, as a rule, would be most obvious in places where it eventually prevailed over the cult of the May year. In such places the absence of traces of the May festival would be no valid argument against its former prevalence. In other places, like Scotland, where the solstitial cult was apparently introduced late and was never prevalent, we should expect strong traces of the May worship, and, as a matter of fact, it is very evident in the folk lore and customs of Scotland; even the old May year quarter days are still maintained.