This is the second of theree essays on colour commissioned for the catalogue of a now-cancelled exhibition on colour at the Musée d’Orsay in Paris.
In 1954, French artist Yves Klein said that “I believe that in future, people will start painting pictures in one single colour, and nothing else but colour.”
Klein did not wait for the future – he began painting monochromes in the 1940s. But it was only in the late 1950s that he truly found what he was looking for: a blue so glorious, so lustrous and deep, that it spoke for itself, and said all that Klein wanted to express. “Blue is the invisible becoming visible”, he said.
Yves Klein, IKB 79’ (1959).
It is not always easy to interpret Klein’s remarks, but I believe this one is not so hard to fathom. Blue has always spoken to something beyond ourselves: it is a colour that draws us into the void, the infinite sky. “Blue is the typical heavenly colour”, said Wassily Kandinsky – and who would doubt it after seeing the ceiling of the Arena Chapel in Padua, painted by Giotto around 1305, a vault coloured like the last moments of a clear Italian twilight? Some cultures don’t even recognize the sky as having a hue at all, as if to acknowledge that no earthly spectrum can contain it. In the ancient Greek theory of colour, blue was a kind of darkness with just a little light added.
There’s a strong case to be made, then, that shades of midnight have always been the most treasured of artists’ colours. One of the earliest of the complex blue pigments made by chemistry was virtually an ancient industry in itself. The blue-glazed soapstone carvings known now as faience produced in the Middle East were traded throughout Europe by the second millennium BCE. Faience is typically now associated with ancient Egypt, but it was produced in Mesopotamia as long ago as 4500 BC, well before the time of the Pharaohs. It is a kind of glassy blue glaze, made by heating crushed quartz or sand with copper minerals and a small amount of lime or chalk and plant ash. The blue tint comes from copper – it is of the same family as the rich blue copper sulphate crystals of the school chemistry lab, although faience could range from turquoise-green to a deep dusk-blue. These minerals were typically those today called azurite and malachite, both of them forms of the compound copper carbonate. It’s not at all unlikely, although probably impossible to prove, that the manufacture of glass itself from sand and alkaline ash or mineral soda began in experiments with firing faience in a kiln somewhere in Mesopotamia.
Egyptian faience: a statue of Isis and Horus, c. 332-30 BC.
Similar experimentation might have given rise to the discovery of the trademark blue pigment of the Egyptians, simply known as Egyptian blue or frit. The recipe, at any rate, is almost the same: sand, copper ore, and chalk or limestone. But unlike faience glaze, this material is not glassy but crystalline, meaning that the atoms comprising it form orderly arrays rather than a jumble. Producing the pigment requires some artisanal skill: both the composition and the kiln temperature must be just so, attesting to the fact that Egyptian chemists (as we’d call them today) knew their craft – and that the production of colours was seen as an important social task. After all, painting was far from frivolous: mostly it had a religious significance, and the artists were priests.
Azurite and malachite make good pigments in their own right – the first more bluish, the second with a green tint. They just need to be ground and mixed with a liquid binder. In the Middle Ages that was generally egg yolk for painting on wooden panels, and egg white (called glair) for manuscript illumination. Good-quality azurite wasn’t cheap, but there were deposits of the mineral throughout Europe. To the English (who had no local sources) it was German blue; the Germans knew it as mountain blue (Bergblau).
The mineral form of the blue pigment azurite (hydrated copper carbonate).
A cheaper blue was the plant extract indigo, used as a dye since ancient times. Unlike most organic dyes – those extracted from plants and animals – it doesn’t dissolve in water, but can be dried and ground into a powder like a mineral pigment, and then mixed with standard binding agents (such as oils) to make a paint. It give a dark, sometimes purplish blue, sometimes lightened with lead white. The Italian artist Cennino Cennini, writing in the late fourteenth century, described a “sort of sky blue resembling azurite” made this way from “Baghdad indigo”. As the name suggests – the Latin indicum shares the same root as “India” – the main sources for a medieval artist were in the east, although a form of indigo could also be extracted from the woad plant, grown in Europe.
But the artist who could find a patron with deep pockets would be inclined towards a finer blue than any of these. When the Italian traveller Marco Polo reached what is today Afghanistan around 1271, he visited a quarry on the remote headwaters of the Oxus River. “Here there is a high mountain”, he wrote, “out of which the best and finest blue is mined.” The region is now called Badakshan, and the blue stone is lapis lazuli, the source of the pigment ultramarine.
Cennino shows us how deeply ultramarine blue was revered in the Middle Ages, writing that it “is a colour illustrious, beautiful, and most perfect, beyond all other colours; one could not say anything about it, or do anything with it, that its quality would not still surpass” [Cennino Cennini, Il Libro dell’ Arte (The Craftsman’s Handbook) (c.1390), p. 37-8. (Dover, New York, 1960)]. As the name implies, it came from “over the seas” – imported, since around the thirteenth century, at great expense from the Badakshan mines.
Ultramarine was precious not just because it was a rare import, but because it was extremely laborious to make. Lapis lazuli is veined with the most gorgeous deep blue, but grinding it is typically disappointing: it turns greyish because of the impurities in the mineral. Those have to be separated from the blue material, which is done by kneading the powdered mineral with wax and washing the wax in water – the blue pigment flushes out into the water. This has to be done again and again to purify the pigment fully. The finest grades of ultramarine come out first, and the final flushes give only a low-quality, cheaper product, called ultramarine ash. The best ultramarine cost more than its weight in gold in the Middle Ages, and so it was usually used sparingly. To paint an entire ceiling with the colour, as Giotto did in the Arena Chapel, was lavish in the extreme.
Lapis lazuli, the source of ultramarine.
More often the medieval painter would use ultramarine only for the most precious components of a painting. That seems to be the real reason why most altarpieces of this period that depict the Virgin Mary show her with blue robes. For all that art theorists have attempted to explain the symbolic significance of the colour – the hue of humility or virtue, say – it was largely a question of economics. Or, you might say, of making precious materials a devotional offering to God.
Duccio’s Maestà altarpiece (1308-11). The Virgin’s robes are painted in ultramarine.
But materials make their own demands. When, during the Renaissance in the fifteenth century, artists began to use oils rather than egg yolk to bind their pigments, they found that the switch both enalbed and demanded new techniques. The colours dried more slowly and so could be blended into subtle shadows and highlights, enhancing the realism of the work in line with the emerging humanist philosophy. But the artists also had to cope with the fact that some pigments look different when bound in oils; ultramarine was one of them.
This is largely a matter of physics: light is bent and scattered to different degrees in the two binding agents, and the result is that ultramarine looked more transparent. Painters were then forced to compromise the purity of precious ultramarine by mixing it with a little lead white to keep it strong and opaque. Its mystique waned accordingly, and artists began to feel more free to use a whole range of lighter blues in their works. The art historian Paul Hills says that “Blue by the fifteenth century was moving away from its association with starry night, the vault of the heavens, to the changeful sky of day.”
You can compare azurite and ultramarine side by side in Titian’s explosion of Renaissance colour, Bacchus and Ariadne (1523). Here is that starry vault, indeed turning to day before our eyes, and it is painted in ultramarine. So too is Ariadne’s robe, which dominates the scene. But the sea itself, on which we see Theseus’s boat receding from his abandoned lover, is azurite, with its greenish tint.
Titian’s Bacchus and Ariadne (1523). Ariadne’s robe, and the sky, are painted in ultramarine.
Over the centuries, artists accumulated a few other blues too. Around 1704 a colour-maker named Johann Jacob Diesbach, working in the Berlin laboratory of alchemist Johann Conrad Dippel, was attemtping to make a red lake pigment when he found that he had produced something quite different: a deep blue material. He had used a batch of the alkali potash in his recipe, supplied by Dippel – but which was contaminated with animal oil allegedly prepared from blood. The iron used by Diesbach reacted with the material in the oil to make a compound that – unusually for iron – is blue in colour. By 1710 it was being made as an artist’s material, generally known as Prussian blue.
It wasn’t entirely clear what had gone into this mixture, and so for some years the recipe for making Prussian blue was surrounded by confusion and secrecy. In 1762 one French chemist declared that “Nothing is perhaps more peculiar than the process by which one obtains Prussian blue, and in must be owned that, if chance had not taken a hand, a profound theory would be necessary to invent it.” But chance was a constant companion in the history of making colours. At any rate, Prussian blue was both attractive and cheap – a tenth of the cost of ultramatine – and it was popular with artists including Thomas Gainsborough and Antoine Watteau. It comprises some of the rich blue Venetian skies of Canaletto.
Another blue from the Renaissance and Baroque periods went by the name of smalt, which is not so very different from the cobalt-blue glass of Gothic cathedrals such as Chartres, ground to a powder. Its origins are obscure, but may well come out of glass-making technology; one source attributes the invention to a Bohemian glassmaker of the mid-sixteenth century, although in fact smalt appears in earlier paintings. Cobalt minerals were found in silver mines, where their alleged toxicity (actually cobalt is only poisonous in high doses, and trace amounts are essential for human health) saw them named after “kobolds”, goblin-like creatures said to haunt these subterranean realms and torment miners. Natural cobalt ores such as smaltite were used since antiquity to give glass a rich blue colour, and smalt was produced simply by grinding it up – not too finely, because then the blue becomes too pale as more light is scattered by the particles. As a result of its coarse grains, smalt was a gritty material and not easy to use.
Some art historians make no distinctions between this “cobalt blue” and those that were given the name in the nineteenth century. But the latter were much finer, richer pigments, made artificially by systematic chemistry. In the late eighteenth century the French government asked the renowned chemist Louis-Jacques Thénard to look for a synthetic substitute for expensive ultramarine. After consulting potters, who used a cobalt-tinted glassy blue glaze, in 1802 Thénard devised a strongly coloured pigment with a similar chemical constitution: technically, the compound cobalt aluminate. Cobalt yielded several other colours besides deep blue. In the 1850s a cobalt-based yellow pigment called aureolin became available in France, followed soon after by a purple pigment called cobalt violet – the first ever pure purple pigment apart from a few rather unstable plant extracts. A sky blue pigment called cerulean blue, a compound of cobalt and tin, was a favourite of some of the post-Impressionists.
The water in Renoir’s Boating on the Seine (La Yole) (c.1879) is painted in cobalt blue.
But what artists craved most of all was ultramarine itself – if only it wasn’t so expensive. Even by the mid-nineteenth century it remained costly, which is why the Pre-Raphaelite Dante Gabriel Rossetti caused much dismay (not to mention added expense) when he upset a big pot of ultramarine paint while working on a mural for Oxford University.
By Rossetti’s time, however, artists did at last have an alternative – it’s just that several of them had not yet learnt to trust it. As chemical knowledge and prowess burgeoned in the early nineteenth century, bringing new pigments such as cobalt blue onto the market, it seemed within the realms of possibility to try to make ultramarine artificially.
It was a prize well worth striving for, because pigment manufacture had become big business. The manufacture of colours and paints wasn’t supplying artists; there was now a taste for colour in the world at large, in particular for interior decoration. Factories were set up in the nineteenth century to make and grind pigments. Some sold them in pure form to the artist’s suppliers, who would then mix up paints for their customers from pigment and oil. But some pigment manufacturers, such as Reeves and Winsor & Newton in England, began to provide oil paints ready-made; from the 1840s these were sold in collapsible tin tubes, which could be sealed to prevent paints from drying out and could be conveniently carried for painting out of doors.
Mindful of the importance of the pigment market, in 1824 the French Society for Encouragement of National Industry offered a prize for the first practical synthesis of ultramarine. It is a complicated compound to make – unusually for such inorganic pigments, the blue colour comes not from a metal but from the presence of the element sulphur in the mineral crystals. This composition of ultramarine was first deduced by two French chemists in 1806, offering clues about what needed to go into a recipe for making it. In 1828, an industrial chemist named Jean-Baptiste Guimet in Toulouse described a way to make the blue material from clay, soda, charcoal, sand and sulphur, and he was awarded the prize (despite a rival claim from Germany). In England this synthetic ultramarine was subsequently widely known as French ultramarine, and Guimet was able to sell it at a tenth of the cost of the natural pigment. By the 1830s there were factories making synthetic ultramartine throughout Europe.
Artists looked upon this substitute with considerable caution, however. Ultramarine still retained some of its old mystique and majesty, and painters were reluctant to believe that it could be turned out on an industrial scale. Perhaps the synthetic variety was inferior – might it fade or discolour? Actually synthetic ultramarine is (unlike some synthetic pigments) very stable and reliable, but J. M. W. Turner was evidently still wary of it when, in the mid-century, he was about to help himself to the ultramarine on another artist’s palette during one of the “finishing days” at the Royal Academy, where artists put their final touches topaintings already hung for display on the walls. Hearing the cry that this ultramarine was “French”, Turner declined to dab into it.
But by the end of the century, synthetic ultramarine was a standard ingredient of the palette: small wonder, given that it could be a hundred or even a thousand times cheaper than the natural variety. Synthetic ultramarine is the pigment in Yves Klein’s patented International Klein Blue, which he used for a series of monochrome paintings in the 1950s. But ultramarine never looked like this before – at least, not on the canvas.
Klein noticed that pigments tend to look richer and more gorgeous as a dry powder than when mixed with a binder – another consequence of how light gets transmitted and refracted – and he sought to capture this appearance in a paint. In 1955 he found his answer in a synthetic fixative resin called Rhodopas M60A, made by the Rhone-Poulenc chemicals company, which could be thinned to act as a binder without impairing the chromatic strength of the pigment. This gave the paint surface a matt, velvety texture. Klein collaborated with Edouard Adam, a Parisian chemical manufacturer and retailer of artists’ materials, to develop a recipe for binding ultramarine in this resin, mixed with other solvents.
Yves Klein, Venus Bleue (1962).
Even in the modern era, then, some artists were still depending on chemical assistance and expertise. Despite the profusion of new pigments with complicated and recondite chemical formulations, the intimate relationship of painters to their materials has not been entirely severed.