For the moment we may overlook the cavalry and fix our attention on the bulk of the citizens who form the infantry. These are split up into five divisions, which were at a later period called _classes_. The basis of division was wealth, and the crucial question is “what kind of wealth?” It is almost certain that it could not have been wealth reckoned in money. Although Rome was a seaport and a trading state, it is doubtful whether even the old libral _as_, which was used as a medium of exchange by weight, was in current use at this time;[281] and therefore the detailed accounts given of the money valuations by which the classes were fixed must refer to a later period in the history of this organisation. The alternative that has been suggested is land.[282] There would be no difficulty in accepting this substitute, paralleled as it is by the similar organisation of Solon, were it not that the hypothesis ignores sources of wealth which the earliest Roman law seems to have classed with land, i.e. slaves and domesticated beasts. These _res mancipi_ are as much the object of quiritarian ownership as land, and they may exist without it. A man might own no land and yet be rich in cattle and sheep which he drove on the _ager publicus_, or in slaves engaged in productive handicrafts,[283] and the state was interested in all that was duly owned and was properly the subject of assessment (_res censui censendo_);[284] the Servian census must have been based on _res mancipi_, and to a certain degree it was a census based on currency, for cattle (_pecus_) were recognised as a medium of exchange (_pecunia_). On the basis of such a census five classes were distinguished; the census of each, in terms of the later assessment, which was probably expressed in _asses sextantarii_, being respectively 100,000, 75,000, 50,000, 25,000, 11,000 asses.