All the property thus assessed must be the object of quiritarian ownership. Originally it had been but the land and the animals associated with it (_res mancipi_),[1047] such objects as had been conveyed by mancipation, and for the evidence of the transfer of which from hand to hand the mancipation witness could be summoned. But the growing mercantile community had to take account of movables, and throughout the historical period all objects of property, corporeal or incorporeal, which constituted _pecunia_ in the later sense of the word, were subject to valuation and taxation.[1048] After the time when direct taxation ceased in Italy (167 B.C.) the valuation was no longer made for the _tributum_; but property was still for a time the determinant of the kinds of military service and voting rights, and the censors had still to scrutinise the professions of the assessed, although the scrutiny was perhaps conducted with less rigour than before. As it was the head of the family alone that could give an account of property, so it was to him that the censor put the requisite questions as to the persons dependent on his care. The respondent gave not only his own name, his father’s and his age, but made similar declarations about his son, his daughter, and his wife.[1049] Inquiries about the female members of the family were chiefly undertaken on moral grounds; they were of no importance for the work of registration, whose object was to assign voting rights and military burdens. Of the three subdivisions of the Roman state—the _curia_, the tribe, the century—the first was not considered by the censor, for the _curia_, like the _gens_, was inherited. The assignment of the tribe varied at different periods. If there was never a time in the history of the censorship when it had been confined to landholders,[1050] the possessor of an allotment was naturally registered in the _tribus_ which contained his plot of ground, the non-possessor in that wherein he dwelt. But, by the year 312 B.C., the landless citizens had already been confined to the four urban tribes; the radical censor of that year distributed them even over the country tribes, to increase the voting power of this _forensis factio_;[1051] but in 304 B.C. the landless proletariate was again confined to the _tribus urbanae_,[1052] and hence arose the permanent distinction between the more honourable country and the less distinguished city tribe. As a matter of fact, this distinction between the landed and the landless citizen could not continue when all property, personal as well as real, became of equal value at the _census_, and membership of the tribe became practically hereditary. But it was a heredity which might be broken by the censor at every period of registration. He might, as we shall see, arbitrarily transfer an individual from his paternal country tribe to one of the four urban divisions, which, partly from historical