Polishing a gemstone, however, with powder ground down from a gemstone harder than itself was a different matter. By doing this the jeweller could achieve a smooth surface, smooth edges and a uniform shape; whereas if he tried to cut it, he might shatter it into splinters. Polishing and engraving-another sure and tried technique which was unlikely to damage the stone-were much safer than cutting.
Thus the gemstones set in jewellery of the Carolingian period are polished or engraved stones, mined in Europe, the Middle East or India, believed to have magical or medicinal power, and with a depth of rich and glowing colour. There were rubies from India; garnets from Bohemi; lapis lazuli and sapphires (the latter’s name being applied to the former); emeralds, amethysts, beryl and aquamarine; amber, the fossilized resin of the ancient forests, which had been known and valued since the beginning of recorded time: turquoises and cornelians, both of which the Egyptians too had gathered; pearls and jet, for which the Romans had had a great liking: all of these were gems which were relatively easy to find, or which could be mined in fairly shallow workings.
In the main there were two styles of setting for the gemstones used in early jewellery: the box setting and the collet. In the former, the jeweller made a small metal box without a lid, placed the gemstone inside, and hammered the metal edges carefully down to hold it in place. A collet set is very similar, but the sides of the box were cut down so that more of the stone might be seem, and claws were sometimes incorporated for the sake of safety.