by marriage with a royal princess; for two of the most ancient kings of Athens, namely Cecrops and Amphictyon, are said to have married the daughters of their predecessors. This tradition is to a certain extent confirmed by evidence, pointing to the conclusion that at Athens male kinship was preceded by female kinship.
Further, if I am right in supposing that in ancient Latium the royal families kept their daughters at home and sent forth their sons to marry princesses and reign among their wives' people, it will follow that the male descendants would reign in successive generations over different kingdoms. Now this seems to have happened both in ancient Greece and in ancient Sweden; from which we may legitimately infer that it was a custom practised by more than one branch of the Aryan stock in Europe. Many Greek traditions relate how a prince left his native land, and going to a far country married the king's daughter and succeeded to the kingdom. Various reasons are assigned by ancient Greek writers for these migrations of the princes. A common one is that the king's son had been banished for murder. This would explain very well why he fled his own land, but it is no reason at all why he should become king of another. We may suspect that such reasons are afterthoughts devised by writers, who, accustomed to the rule that a son should succeed to his father's property and kingdom, were hard put to it to account for so many traditions of kings' sons who quitted the land of their birth to reign over a foreign kingdom. In Scandinavian tradition we meet with traces of similar customs. For we read of daughters' husbands who received a share of the kingdoms of their royal fathers-in-law, even when these fathers-in-law had sons of their own; in particular, during the five generations which preceded Harold the Fair-haired, male members of the Ynglingar family, which is said to have come from Sweden, are reported in the _Heimskringla_ or _Sagas of the Norwegian Kings_ to have obtained at least six provinces in Norway by marriage with the daughters of the local kings.
Thus it would seem that among some Aryan peoples, at a certain stage of their social evolution, it has been customary to regard women and not men as the channels in which royal blood flows, and to bestow the kingdom in each successive generation on a man of another family, and often of another country, who marries one of the princesses and reigns over his wife's people. A common type of popular tale, which relates how an adventurer, coming to a strange land, wins the hand of the king's daughter and with her the half or the whole of the kingdom, may well be a reminiscence of a real custom.
Where usages and ideas of this sort prevail, it is obvious that the kingship is merely an appanage of marriage with a woman of the blood royal. The old Danish historian Saxo Grammaticus puts this view of the kingship very clearly in the mouth of Hermutrude, a legendary queen of Scotland. "Indeed she was a queen," says Hermutrude, "and but that her sex gainsaid it, might be deemed a king; nay (and this is yet truer), whomsoever she thought worthy of her bed was at once a king, and she yielded her kingdom with herself. Thus her
