The principal natural resources of the earth were regarded by the Roman state as amongst the main spoils of victory (pretium victoriae). The Roman world was a huge consumer of metals of almost every type, with the coinage alone dependent on gold, silver, copper, and tin (K. Greene 2000a: 747-52). Because of an outflow of specie beyond the frontiers of the empire and the recurrent removal of material from circulation into savings hoards, the recycling of the existing coin supply was not sufficient to satisfy demand and mining activity was a vital element of the economy under state regulation. Sometimes exploitation was carried out under direct state control, though more commonly the state operated in partnership with private entrepreneurs who bid for contracts. The Roman procurators in charge of mining districts had a wide range of concessions to lease out - everything from mine workings (run on a share of production basis) to shoe making or running the bath houses in the mining villages (0rsted 1985: 203).
There are indications in some provinces, notably Spain, that mining activity reached an unprecedented scale for a pre-industrial society (D. G. Bird 1972; Domergue 1990; Woods 1987). The most extraordinary site is known today as Las Medulas after the modern village that sits in the center of a vast opencast area (over 2 km in diameter and several hundred meters deep). This remarkable manmade crater was produced through the use ofsophisticated hydraulic mining techniques, using water power to undermine the edges of the growing opencast and to wash c.17 million cubic meters of debris further down the mountain. The total opencast area of the mine occupies c.5.4 square kilometers and the outwash from it deeply buried a further 5.7 square kilometers of land (Sanchez-Palencia 2000: 225). Las Medulas is the largest of c.230 gold mines in northwestern Spain, with a combined output peaking at 20,000 pounds of gold a year in the first century CE (Plin. Nat. 33.4.78). The figure seems less implausible once you have peered over the edge of the opencast at Las Medulas!
Another way to approach the question of the scale of Roman mining activity is to look at the emerging picture of global pollution registered in the Greenland ice-cores. Analysis of the ice-cores has revealed that fresh layers of ice are formed each year and that by counting back, the ice can be accurately dated, rather like tree rings. The chemical analysis of ice across time has now demonstrated that the major pre-Industrial Revolution peak in hemispheric pollution occurred in the Roman period, with notable peaks of both copper and lead (Hong et al. 1994, 1996a, b; Rosman et al. 1997). What this translates to at a local level is best illustrated by recent work in Wadi Faynan, Jordan, where a major Roman copper mine and smelting operation were evidently responsible for a hugely toxic environment, with dangerous levels of a cocktail of poisons traceable even today in vegetation and the animals that graze on it (Barker et al. 2000: 44-6).
The use of marble in Roman construction increased in scale and extent from the reign of Augustus, and Tiberius appears to have made decorative stones an imperial monopoly alongside metals (Dodge 1988,1991; Dodge and Ward-Perkins 1992; Fant 1988, 1989, 1993). The Romans were particularly interested in the exploitation of colored stones and recent work in the Eastern Egyptian desert has revealed much about the nature of imperial operations there. Two major quarry locations have been surveyed: Mons Claudianus, producing a grey granite, and Mons Porphyrites, producing the much-prized purple porphyry (Maxfield and Peacock 2001; Peacock and Maxfield
1997) . It is clear from the wealth of epigraphic documents from these sites that the labor force was predominantly salaried and free, rather than slave-conscript. The same appears to be broadly true of ancient mining enterprises - though it is equally certain that some penal labor was involved - and the implication is that these extractive industries attracted specialist workers with above average wages. Detailed studies of the diet at these quarrying sites has revealed a surprisingly luxurious range of foodstuffs in the peak periods of operation, much of it imported from the Nile valley (van der Veen
1998) . The logistical implications are considerable (Adams 2001; Maxfield 2001).