nearly half a century earlier, if the depth of the deposit (9 ft.) containing amphora sherds under the warehouse be taken as an index.5 Moreover, some of the Arretine ware dates probably from the first quarter of the 1st century A.D. For the subsequent duration of the town we have at present insufficient evidence. In the industrial area where the tanks or dye-vats are situated there were two or more phases of reconstruction, but there was a general continuity in the main units of the plan, and sherds of Mediterranean amphora occurred in all strata. In other words, the function and contacts of the site remained unchanged. To interpret these factors in terms of time is guesswork; a minimum of a century might appear to have been required by the renewals, but there seems to be no compelling reason to allow more than two centuries, and a terminal date soon after A.D. 200 is suggested. The quantity of Western pottery in the relatively small areas uncovered is impressive. At least fifty sherds of Arretine, including four potters' stamps (VIBIE, ITTA, CAMVRI, C.VIBI OF) are recorded, and others have been found, mostly in the northern (warehouse) sector where, it may be supposed, the shippers and merchants were congregated. Of amphora sherds something like 150 are known to have been unearthed over a wider area; and incidentally an internal incrustation on some of the sherds has been shown by analysis to contain resin, a traditional component of certain Greek wines though here (as elsewhere) perhaps applied deliberately by the potters to render the vessels impermeable. Other wares which may have been imported from the West include above all an extensive group of flat-bottomed dishes of a hard, metallic pottery, whitish in section but with a polished slip blackened internally by inverted firing and with concentric rings of rouletting on the upper surface of the base. Local imitations of this ware are distinguished by softer fabric and coarser rouletting. This 'rouletted dish' is widely distributed in southern and central India and has become what geologists would call a 'type fossil' in the dating of associated Indian cultures. It will be considered again at a later stage. Roman glass, including a 'pillared' bowl of 1st-century type, reached the site in small quantities, and fragments of at least two characteristic Roman lamps of early date have been noted. On the other hand, coins of the period, whether Roman or Indian, are completely absent. This is in accordance with the postulate, already discussed, that the imported currency was not as a rule circulated in India in the normal processes of monetary exchange, the country ------------------------------------------------------------------------------------------------------------------ 5 This deposit was largely alluvial mud and probably accumulated fairly rapidly. |