The theory and fact of the Roman emperor's position were strikingly at variance in the fourth century. ‘Quod principi placuit legis habuit vigorem’ may hav been the established doctrine, but when he spoke to his proconsuls, the emperor said rather, ‘You must order the proceedings [of civil suits] to be on luded on the third day, or at the latest on the fourth day, or at any rate on the fifth’; or, to the civil service, ‘Let the greedy hands of the officials now forbcar, let theln forbear, I say. The stutter and shrillness do not fit an autocrat, nor the repetitions; but repetitions are characteristic of the Codes. An entire decree might have to be re-enacted, in almost identical form, eight or ten time, evidently because no one had paid any attention to it. Against this merely ornamental legislation, against ‘the ruler who lays down a law and is reckoned unable to enforce it,’ Libanius might protest, yet, three times imperatively summoned to the capital, he simply remained in Antioch. A contenlporary, Valerianus v. c., flouted one decree after another, one judicial decision after another, and flouted, too, the authority of the praetorian prefect, the proconsul, the urban prefect, and the imperial secret police, one of whom he allegedly killed with his own hands. Between the emperor and his subjects, and between his theoretical omnipotence and his actual powers, the essential link was the bureaucracy. Plainly, it was not doing its job. It is the argument of this paper that part of the difficulty lay in the failure of all ranks of government to get through to each other.