The Patria Potestas, in its normal shape, has not been, and, as it seems to me, could not have been, a generally durable institution. The proof of its former universality is therefore incomplete so long as we consider it by itself; but the demonstration may be carried much further by examining other departments of ancient law which depend on it ultimately, but not by a thread of connection visible in all its parts or to all eyes. Let us turn for example to Kinship, or in other words, to the scale on which the proximity of relatives to each other is calculated in archaic jurisprudence. Here again it will be convenient to employ the Roman terms, Agnatic and Cognatic relationship. _Cognatic_ relationship is simply the conception of kinship familiar to modern ideas; it is the relationship arising through common descent from the same pair of married persons, whether the descent be traced through males or females. _Agnatic_ relationship is something very different: it excludes a number of persons whom we in our day should certainly consider of kin to ourselves, and it includes many more whom we should never reckon among our kindred. It is in truth the connection existing between the members of the Family, conceived as it was in the most ancient times. The limits of this connection are far from conterminous with those of modern relationship. Cognates then are all those persons who can trace their blood to a single ancestor and ancestress; or, if we take the strict technical meaning of the word in Roman law, they are all who trace their blood to the legitimate marriage of a common pair. "Cognation" is therefore a relative term, and the degree of connection in blood which it indicates depends on the particular marriage which is selected as the commencement of the calculation. If we begin with the marriage of father and mother, Cognation will only express the relationship of brothers and sisters; if we take that of the grandfather and grandmother, then uncles, aunts, and their descendants will also be included in the notion of Cognation, and following the same process a larger number of Cognates may be continually obtained by choosing the starting point higher and higher up in the line of ascent. All this is easily understood by a modern; but who are the Agnates? In the first place, they are all the Cognates who trace their connection exclusively through males. A table of Cognates is, of course, formed by taking each lineal ancestor in turn and including all his descendants of both sexes in the tabular view; if then, in tracing the various branches of such a genealogical table or tree, we stop whenever we come to the name of a female and pursue that particular branch or ramification no further, all who remain after the descendants of women have been excluded are Agnates, and their connection together is Agnatic Relationship. I dwell a little on the process which is practically followed in separating them from the Cognates, because it explains a