Thirdly, coins of Augustus and Tiberius predominate. After Nero they dwindle markedly.

Fourthly, the gold coins are liable to be either pierced for suspension or mutilated by a cut across the obverse. Only one of the very numerous Roman silver coins is known to have been similarly mutilated.

Fifthly, there is a notable grouping of early coin-finds across southern India from the western to the eastern coasts.

Of these five points, the first is readily explained in relation to the second. The significance of hoards in general has often been discussed, and they are commonly regarded as savings lost, or rather not recovered, during a period of disturbance. But it would be easy to overstress this aspect of the matter, and, whatever the accidents of their final deposition, it may be suggested that the preponderance of early Roman hoards in South India first and foremost represents an essentially economic factor. The great quantity of gold and silver coins imported by 1st-century Roman trade into India is noted or implied by more than one classical writer. The Periplus mentions it in relation to Barygaza and as first amongst the imports carried by Western merchants to the Malabar ports; Pliny (VI, 101) remarks that 'in no year does India absorb less than fifty million sesterces', which presumably represented actual cash rather than the estimated value of wares; Tiberius may be thought to have had this extravagant efflux specially in mind when he complained to the Roman senate of the reckless exportation of money 'to foreign nations and even to the enemies of Rome' in exchange for gew-gaws (Tacitus, Ann. II, 53). But how was all this currency fitted into the alien economies of recipients far beyond the boundaries of the Roman Empire?

The clear answer is that, for the most part, it was employed not as currency but as bullion. In the whole of peninsular India there was no native currency of gold or silver to which the Roman coinage could be approximated. Pausanias (III, 12, 24) in the 2nd century observed that the Indians exchanged their wares with those of the Greeks without understanding the use of money. The potin or lead coinage struck by the Andhra empire of central India in the first two centuries A.D. is a partial exception to this rule; and half a dozen scattered denarii of Augustus and Tiberius found over a period of years in a restricted area of an Andhra town at Chandravalli, near Chitaldrug in northern Mysore, if they do not represent a hoard broken anciently, may here have circulated in some ratio to the native base metal issues. Outside the Andhra zone, however, this possibility does not arise. For the most part, the imported coins can only have been used as bullion, to be weighed out in exchange for