INTRODUCTION

To study mint-output year by year, we need big hoards and coinage which is dated by year. This evidence can reveal big fluctuations. Thus annual differences in the Egyptian hoards illustrated in fig. 8.1 show a coefficient of variation of 245% (hoard 3) and 230% (hoard 6). This is primarily due to the enormous production surge in the mid-60s AD. In denarii of Severus Alexander (fig. 8.3), the coefficient is still high, 80% overall at Reka Devnia and 81% at Elveden. At Reka Devnia between 148 and 161, the variation in year-dated coin of Antoninus Pius struck for the Emperor is 55% (n = 4,081). Emperor coin under Commodus at Reka Devnia shows variation of 42% (n = 2,364).

These results undermine any assumption that annual output was constant. Usually there seems to be movement within the year-to-year figures, a movement which at times becomes regular and even systematic. Nevertheless, precious metal output from the Rome mint appears to have been broadly continuous in the main period studied. This contrasted with mints for provincial coinage, which were sometimes closed for years on end. But usually lack of year-dating means that the Rome evidence has to be conflated into blocs of several years. The longer blocs still show some fluctuation. Gold and silver indices run in parallel reign by reign for part of the period (Table 8.8). But over shorter periods they do not do so consistently.