No such group could possibly form a _gens_ of its own, if it could be regarded as having been originally in dependence on a patrician clan. Although in course of time legally independent and freed from all trammels of clientship, it was yet disqualified from clan-brotherhood by this original connexion; it remained an offshoot (_stirps_), a mere dependent branch, and could never be a self-existent _gens_. This disqualification is exhibited in the definition of _gentilitas_ given by the jurist Scaevola (consul 133 B.C.), which gives as two of its conditions free birth in the second degree, and the absence of servile blood in one’s ultimate ancestry.[34] This definition excludes from membership of a _gens_ all those Plebeians who had sprung originally from emancipated slaves. No one who could be proved to have the taint of servile blood could ever be a _gentilis_. But there is every reason to believe that _servitus_ was interpreted in a further sense, that clientship was regarded as a quasi-servile position, and debarred a group of families, whose ancestor could be proved to be a client, for ever from being a clan. As a rule it would have been difficult, if not impossible, to furnish this proof; but there was one legal sign of it—the bearing by a plebeian _stirps_ of the same name as a patrician clan. The presumption of the law, in the case of the coexistence of a plebeian group of families with a patrician group of the same name, was apparently that the former had once been clients of the latter, and could never, therefore, form a _gens_ of their own.[35] But, if there were plebeian families that had no origin in clientship, there was nothing to prevent these from being _gentes_. It is true that Patricians sometimes made the claim that all the plebeian families had originated from clientship.[36] But this is, as we saw,[37] probably not true of the origin of many of the plebeian families, and there is abundant evidence that the theory was not recognised by law. We know, for instance, that gentile inheritances were shared by the plebeian Minucii, and gentile sepulchres by the plebeian Popilii.[38] The foregoing description shows that the _gens_ rests on a natural basis, that it professedly represents the widest limits of blood-relationship; hence it would seem to follow that it could not be artificially created or its members redistributed; that the numbers of the clans could not be regulated numerically, except conceivably by the addition to the existing number of a precise number of added clans—a most improbable procedure; and that, as being a natural and not an artificial creation, it was a union which was not likely to be of primary importance politically, and the rights of whose members were in all probability those of private rather than of public law. These expectations are verified, but the attempts to point out certain purely political characteristics of these associations deserve examination.[39]