indeed, in India, tend to become hereditary, and, when their nature permits it, to vest in the eldest member of the oldest stock. Comparing these Indian successions with some of the ruder social organisations which have survived in Europe almost to our own day, the conclusion suggests itself that, when Patriarchal power is not only domestic but political, it is not distributed among all the issue at the parent's death, but is the birthright of the eldest son. The chieftainship of a Highland clan, for example, followed the order of Primogeniture. There seems, in truth, to be a form of family-dependency still more archaic than any of those which we know from the primitive records of organised civil societies. The Agnatic Union of the kindred in ancient Roman law, and a multitude of similar indications, point to a period at which all the ramifying branches of the family tree held together in one organic whole; and it is no presumptuous conjecture, that, when the corporation thus formed by the kindred was in itself an independent society, it was governed by the eldest male of the oldest line. It is true that we have no actual knowledge of any such society. Even in the most elementary communities, family-organisations, as we know them, are at most imperia in imperio. But the position of some of them, of the Celtic clans in particular, was sufficiently near independence within historical times to force on us the conviction that they
