Opal is hydrated amorphous silica, containing between three and twenty percent water. It has no crystal structure, which makes it a mineraloid rather than a mineral, and the water content makes it unstable: opal can craze, developing a network of fine cracks, if it dries. Ethiopian hydrophane opal absorbs water and changes appearance temporarily when wet, which is disconcerting and reversible.
The play of colour was not explained until the 1960s, when electron microscopy showed that precious opal consists of silica spheres packed in a regular three-dimensional array. Light diffracts from this array exactly as it does from a crystal lattice, and the colours seen depend on the sphere diameter and the viewing angle. Larger spheres give reds, smaller give blues; a stone showing red generally shows the whole range, which is why red-dominant opal is the most valuable. Common opal, with irregularly sized spheres, shows no play of colour at all.
This is one of the more satisfying explanations in gemmology: an optical phenomenon that looks like an internal fire turns out to be diffraction grating physics, and the structure responsible is invisible at any magnification available before the twentieth century.
Australia produces the great majority of the world's precious opal, principally from Lightning Ridge, Coober Pedy and Andamooka, and the Australian fields were only developed from the 1880s onward. Before that the source was Hungary, in what is now Slovakia, which supplied the opals known to the classical and medieval world. Ethiopian material entered the market in quantity from the 2000s.