show that this difficulty is exaggerated, and that the objection founded upon it is one on which all attempts at definition in social science would suffer shipwreck. No doubt, if we were simply to put forward "the city" as a social classification, we should expose ourselves to the very serious objection that so vague a term confuses clan-cities like the Hebrew, in which the inhabitants are regarded as "sons" and "daughters" of the place—a curious combination of kinship and local contiguity as social ties—with municipal life, like that of Athens and Rome, in which kinship of communal nature is gradually forgotten, and with royally or imperially chartered towns, like those of England, France, Spain, Germany, in which kinship ties are altogether lost and the connection between local and central government alone regarded. The term "city commonwealth" has been used to prevent any such ambiguity. It is intended to call attention to the fact that the "city" which occupies so conspicuous a place in social history is neither a village commune nor a chartered town, but a self-dependent unit, rising indeed out of clans and villages, and expanding, it may be, into an empire, but clearly distinguishable alike from communal and imperial systems. But the objection is really based on a mistaken view of social science which would destroy all its definitions. According to this view, our social classes must possess a clear-cut regularity of outline such as the insensible gradation of forms of social life renders impossible. We have previously referred to this irrational requirement, and can only repeat that the fallacy proceeds from the assumption that a class can ever possess the defined unity of individual being.
So far as the social classification under discussion is concerned, it may be added that the range of the city commonwealth, like that of the clan itself, has been
