centre of each community would be the perpetual fire on the king's hearth tended by Vestal Virgins of the royal clan. The king would be a man of another clan, perhaps of another town or even of another race, who had married a daughter of his predecessor and received the kingdom with her. The children whom he had by her would inherit their mother's name, not his; the daughters would remain at home; the sons, when they grew up, would go away into the world, marry, and settle in their wives' country, whether as kings or commoners. Of the daughters who stayed at home, some or all would be dedicated as Vestal Virgins for a longer or shorter time to the service of the fire on the hearth, and one of them would in time become the consort of her father's successor. This hypothesis has the advantage of explaining in a simple and natural way some obscure features in the traditional history of the Latin kingship. Thus the legends which tell how Latin kings were born of virgin mothers and divine fathers become at least more intelligible. For, stripped of their fabulous element, tales of this sort mean no more than that a woman has been gotten with child by a man unknown; and this uncertainty as to fatherhood is more easily compatible with a system of kinship which ignores paternity than with one which makes it all-important. If at the birth of the Latin kings their fathers were really unknown, the fact points either to a general looseness of life in the royal family or to a special relaxation of moral rules on certain occasions, when men and women reverted for a season to the licence of an earlier age. Such Saturnalias are not uncommon at some stages of social evolution. In our own country traces of them long survived in the practices of May Day and Whitsuntide, if not of Christmas. Children born of more or less promiscuous intercourse which characterises festivals of this kind would naturally be fathered on the god to whom the particular festival was dedicated.