Until recent centuries, historians had to reckon with only limited social movement. I use the phrase in a very wide sense, to cover change of habitation, profession, or class; but even so defined, in a world dominated numerically by small farmers, people rarely moved around, or up or down, socially. This is as true of the thirteenth as of the third century.But it is usual to say that, granted these limitations, social movement was still much more restricted after 250 than before. This has recently been challenged by A. H. M. Jones. ‘The late Roman empire is often conceived of as a rigid hierarchical society, in which every man was tied to the station in life to which he was born.’ But the laws directing this confess, in their repetitions and relaxations, that they could not really be applied. Society, he says, was actually less static after 250 than before. Despite what has become the almost canonical view of the question, Professor Jones is certainly in the right. Under three major headings, all perfectly well known, many people did change jobs and homes. Some fled from invasion, in numbers powerfully suggested by the thousands of coin hoards from Britain to the Black Sea—we must assume that for every treasure we find, a hundred are still hidden, and for every man who buried his money, a hundred took it with them—; another group, from one to two hundred thousand, joined an expanded army; an equal number in Egypt alone turned monk. If we say, then, that over the century 250–350, half a million people were taken from one life and newly rooted in another, we have a figure by no means fantastic, yet without parallel in any earlier era of the Empire.