The history of pigment chemistry is in many respects the history of art itself, because the colours available to a painter at any given moment determined what could and could not be represented. A Byzantine mosaic could deploy the gold tesserae that were beyond the financial reach of most later panel painters; a Venetian painter of the sixteenth century had access to a deep blue made from lapis lazuli imported from Afghanistan, which no northern European workshop could afford in significant quantity; an Impressionist working in Paris in the 1870s could buy tubes of chrome yellow and cobalt blue that had been invented only within the preceding few decades. Each of these circumstances shaped what was painted, how it was painted, and how it has survived.

The oldest pigments in consistent use are the earth colours, obtained by grinding naturally occurring mineral deposits. Iron oxides produce red and yellow ochres; manganese oxide produces a dark brown; carbon black is made by charring wood or bone. These pigments are chemically stable and have survived essentially unchanged on cave paintings tens of thousands of years old. A disadvantage is that they offer a relatively narrow range of colours and produce muted tones, rather than the intense saturated hues that later painters often sought.

The first major expansion of the palette came with pigments derived from processed minerals. Azurite, a blue copper carbonate, and malachite, a green copper carbonate, were ground and washed to produce pigments used extensively in medieval European and East Asian painting. Vermilion, a brilliant red made from the mineral cinnabar or synthesised by heating mercury and sulphur, was known in both China and the Mediterranean by the classical period and remained among the most expensive pigments of European painting until the nineteenth century. The Italian pigment historian Giuseppe Castelli has documented the trade in mineral pigments and has shown that some works of major painters can be approximately dated by the sources of their pigments, as the trade routes shifted over the centuries.

The deepest and most prized blue of pre-modern European painting was ultramarine, made from the semi-precious stone lapis lazuli, mined almost exclusively in the mountains of northern Afghanistan. Extracting usable pigment from lapis was laborious: the stone was ground, mixed with wax, kneaded in alkaline solution to separate the deep blue from duller minerals, and the resulting paint was more expensive, weight for weight, than gold. For much of the Renaissance, contracts for major paintings specified the weight of ultramarine to be used in particular areas, with premium prices for its deployment in the robes of the Virgin. The combination of rarity and symbolism gave the colour a cultural significance that modern synthetic blues cannot reproduce exactly.

Organic pigments, derived from plants and animals rather than minerals, added important colours but introduced significant stability problems. Indigo from the Indigofera plant gave a strong blue; madder from the Rubia plant produced red lakes (pigments made by precipitating a dye onto a colourless base); cochineal insects from Mexico, after the Spanish conquest, supplied a particularly brilliant red lake. Many of these pigments fade rapidly in strong light, and paintings in which they predominate often appear very different today from how they looked when new. Conservators working on Flemish paintings have noted that the green glazes over gold leaf, made from yellow lakes over copper pigments, have sometimes lost their yellow component entirely, leaving the green areas looking grey.

The industrial revolution transformed the pigment market. Prussian blue, synthesised accidentally in Berlin in 1706, was the first of a long series of cheap, stable synthetic pigments that gradually displaced the natural products. Chrome yellow, lead-tin yellow, cobalt blue, cadmium red, and many others followed during the nineteenth century. By 1900, the painter's palette contained dozens of new colours that would have been inaccessible a century earlier, and at prices that allowed the mass production of artists' tube paints for the first time. The new availability coincided with, and to some extent enabled, the formal experiments of Impressionism and the subsequent modernist movements.

Not every new pigment proved stable. Lead-based pigments, toxic in manufacture and use, caused both health problems and colour changes when exposed to atmospheric sulphur. Chrome yellow darkens under strong light because of a slow chemical change. Cadmium red and yellow are toxic in their soluble forms and have been increasingly restricted in professional use. Modern conservation science has developed the analytical tools - X-ray fluorescence, Raman spectroscopy, and micro-sampling - required to identify the pigments of a particular painting and to anticipate the changes they will undergo. Castelli has argued that every major collection now needs such analytical capacity, because decisions about display lighting, humidity, and exhibition loans depend critically on which pigments are present.

A more recent development is the study of what is no longer visible. In some paintings, a pigment has faded so completely that the original appearance of the work has to be reconstructed from traces on the canvas and from comparison with better-preserved paintings. The conservators of the Rijksmuseum produced a reconstruction of the original colours of Rembrandt's Night Watch by combining pigment identification from micro-samples with computational modelling of the fading that had occurred since the seventeenth century. The resulting reconstruction, projected onto the painting in carefully controlled conditions, allows visitors to see what the work might have looked like when new - a view that differs substantially from the darkened image familiar from reproductions.

The study of pigments therefore connects several disciplines that once operated separately: art history, chemistry, materials science, and conservation. Castelli has observed that the boundary between these fields has become steadily less useful, and that the most interesting recent work has been done by teams drawn from several of them. Whether the pigment is a blue from an Afghan mountain, a red from a Mexican insect, or a modern cadmium formulation from a chemical plant in Germany, its story is simultaneously a story of materials and a story of the images they have made possible.