Most gems of the Greco-Roman period were made of hard stones of which quartz (silicon dioxide, SiO2; hardness of 7 on the Mohs scale) is by far the commonest type. There exists a wide range of quartzes of varying colours and appearance to which different names are applied. Colour could be invested with a quasi-magical function: red was the colour of blood and flesh, and green that of vegetation; amthyst was the colour of wine and as its name implies was a prophylactic against drunkenness.

The list given below is not intended to be comprehensive and only includes the minerals most frequently found. The task of identifying these gems is not a difficult one, since only about a dozen minerals were in common use. The more exotic stones represent just a fraction of ancient engraved gems. However, it is not always possible to match the ancient Greek and Latin terminology with known gemstones. It is a remarkable fact that Pliny gives some three hundred names to describe stones known and used in his time.

Chalcedonies or microcrystalline quartzes

Cornelian is a translucent red form of chalcedony, shading from dark red to golden yellow. The name derives from the Latin cornum, the red berry of the cornel-tree. The alternative name carnelian is generally considered incorrect, but the false etymology from carnis, 'flesh', has popularised its spelling. The whitish appearance of some ancient specimens is due to exposure to great heat.

Sard is a translucent brown variety of chalcedony, shading from light yellowish brown to opaque dark brown. Dark inclusions can sometimes be observed. Sard is often difficult to distinguish from cornelian. The cornelian and sard are the most widely used stones in Greek and Roman glyptic art. The name is derived from Sardis in Lydia, the place where it was chiefly found.

Sardonyx is used to describe chalcedony with straight bands of alternating brown or blue bands. It was the preferred stone for engraving cameos. The carver would take advantage of the colours in the layers to show, for instance, cream-coloured figures on a dark background, or to depict details of a drapery or a wreath. The term nicolo is used to describe a Roman banded intaglio with a blue or brown top layer and a dark brown bottom layer. Onyx, which derives from the Greek word for fingernail, a reference to the colour of the pale bands, is the name usually given to a black and white two-layered banded chalcedony.


Jasper is an opaque form of chalcedony and the most popular colours were red, orange and yellow. There is a green variety sprinkled with red spots popularly called bloodstone or heliotrope. Yellow and especially red jaspers became very fashionable for Roman gems in the second and third centuries A.D. Mottled jasper with small patches of white, brown, yellow and black was also occasionally used in the same period.

Chalcedony is a microcrystalline form of quartz and its different colours are due to the impurities that it has absorbed. The name is also generic, but is convetionally applied to colourless, grey and blue varieties. It derives from the name of the city Chalcedon (modern Kadıköy in Türkiye).

Plasma is a green variety of chalcedony and often contains dark inclusions. The green colour is generally due to the presence of chrome. It is not a very accurate term, as it is applied to several different green stones, including aventurine, prase and chrysoprase.

Macrocrystalline quartzes

Rock crystal is transparent and colourless. It was more popular in the Greek period than in the Roman, when it appears only in the first century B.C.. Ancients believed that rock crystal was a form of petrified ice, the result of water being frozen at a very low temperature, the word 'crystal' being derived from the Greek word for ice. In Asia Minor, Pliny mentions that a poor variety occurred around Alabanda and Orthosia in Caria.

Amethyst is transparent and ranges in hue from dark purple to pale mouve. The colour is usually not distributed evenly in the stone, some parts being lighter, others darker. The name derives from the Greek word meaning 'not drunken', which originated from the belief that the wearing of the stone gave immunity against the after-effects of indulgence.

Other minerals

Lapis lazuli is deep blue and sometimes contains brassy specks of pyrite. It was highly prized since its only quarries were in Afghanistan, although Persia was also a possible source. In Roman times lapis lazuli was seldom used as a gemstone and most examples date to the second and third centuries A.D.

Garnet is a crystallised silicate. It is transparent and ranges in colour from dark red to orange, and sometimes purple, variants to which different names were applied in antiquity. It was not used until the Hellenistic period, when it became particularly fashionable. Its hardness was superior to that of quartz and it was thus more difficult to carve. In Asia Minor, Pliny mentions that it was extracted around the cities of Alabanda and Orthosia in Caria; the modern name almandine is a corrupt form of alabandina. A fiery red variety called lychnis is said by Pliny to have been found around Orthosia and throughout Caria. Garnets were very often carved with a curved surface and, to lighten the colour, the underside was holowed out.

Haematite is an iron oxide of dark metallic grey appearance. According to Theophrastos it was given that name because it looked like congealed blood, the word 'haematite' being derived from the Greek word for blood. Another explanation for the name is the red colour that haematite takes when it is powdered. It was infrequently used in the Greek period and most examples date from the Roman imperial period, when it was often made into magical intaglios.

In addition, glass was used throughout antiquity as a substitute for expensive gemstones. Some glass gems were directly engraved as in the same way as stone gems, other were cast from terracotta moulds made from the actual engraved stone gems, both intaglios and cameos. When the result was not entirely satisfactory, the glass impression was retouched to give it a sharper edge. Glass gems could also consist of several layers of varying colours, to reproduce sardonyx, nicolo, or banded agate. The glass could be opaque or translucent, with air bubbles often being visible on the latter variety. According to Pliny and other ancient authors, glass gems were often sold as stones by fraudulent dealers. One story tells us that the wife of Gallus, in the mid-first century B.C., bought a necklace of expencive beads only to find out that they were cheap glass. The dishonest seller was caught and duly hauled off to the arena where after a fearful wait he was confronted by a capon, not a lion, to the amazement of spectators; this was said to be a punishment that fitted the crime. It is clear, however, that most glass gems were sold as such, being more affordable and available in a whole range of fancy colours. Demand for them increased sharply from the first century B.C. onwards.