Science

Magenta and the missing wavelength

The Rainbow, Achères la Forêt by Jean-Charles Cazin, a rainbow over a French village landscape

Graph of the three human cone types and how strongly each responds across the visible spectrum

GOSSIP - hot pink, high grade professional acrylic paint, by Stuart Semple 100ml Stuart Semple

PINKEST PINK POTION - high grade professional acrylic paint, by Stuart Semple 100ml Stuart Semple

There is no single wavelength of light that looks magenta. This is how the eye and brain make the colour anyway, and why it got its name from a battle in Lombardy.

A colour with no place in the spectrum

Split sunlight with a prism and you get a band of colour running from violet at one end, through blue, green, yellow and orange, to red at the other. Each part of that band is light of a particular wavelength. Violet sits at the short end, at around 400 nanometres, and red at the long end, at around 700. A rainbow shows the same order, because raindrops bend shorter wavelengths more strongly than longer ones and so spread sunlight out by wavelength, with red on the outside of the bow.¹

Isaac Newton named seven colours in that band, including indigo between blue and violet. Magenta is not on anyone's list. There is no single wavelength of light that people see as magenta, which is why colour scientists call it a non-spectral, or extra-spectral, colour.²³

How the eye builds it

Colour vision in daylight relies on three kinds of cone cell in the retina. They are often called red, green and blue cones, but their sensitivities are broad and overlap heavily, so scientists prefer short, medium and long wavelength cones (S, M and L). Measurements of single human cones by James Bowmaker and Herbert Dartnall, published in 1980, put their peak sensitivities at about 420, 534 and 564 nanometres.⁴ No cone reports a colour on its own. What we see depends on how strongly the three types respond relative to one another.²⁵

Light from a single wavelength always produces one of a limited set of cone patterns. Yellowish-green light, for example, excites the M and L cones almost equally. Magenta comes from a pattern that no single wavelength can produce: a strong response from the S cones and the L cones at once, with little from the M cones in between. That happens when light from both ends of the spectrum, reddish and bluish, reaches the eye together.³² The visual system reads that mixture as a hue of its own, somewhere between red and violet, and depending on the balance of the two ends it looks more like magenta or more like purple.³

Colour scientists draw this on the chromaticity diagram, a horseshoe-shaped map of every colour the average eye can see. The pure spectral colours run around the curved edge. The straight edge that closes the horseshoe, joining spectral red to spectral violet, is known as the line of purples. Every colour along it is a mixture of light from the two ends of the spectrum, and only its two end points are spectral.⁶²

Is it real?

Popular science often says magenta "doesn't exist" or that the brain invents it.⁵ The Colour Literacy Project, a group of colour educators, calls this a misconception. Light of 550 nanometres is not green in itself either. Green is also what the eye and brain make of a particular balance of cone signals. In the project's words, magenta is made up by our brains, "but so are all the hues that seem to exist in the spectrum".² The difference with magenta is only that no single wavelength can trigger it. It needs two parts of the spectrum at the same time.

This is why magenta is easy to make and hard to find in a rainbow. A phone or computer screen makes it by lighting its red and blue sub-pixels and leaving the green off.³⁵ Printers go the other way. Magenta ink is one of the three standard process colours, alongside cyan and yellow, and it works by absorbing green light from the white page and letting the red and blue through.³⁷

Named after a battle

The word itself comes from chemistry and war. In 1858 and 1859 the French chemist François-Emmanuel Verguin, working near Lyon, made a rose-red dye from aniline, a coal tar product, and named it fuchsine after the fuchsia flower. He patented it on 8 April 1859.⁸⁷ On 4 June that year French and Sardinian forces beat the Austrian army at Magenta, a town near Milan. The firm that had taken on Verguin's patent sold the dye under the name magenta, and the name stuck.⁷³ In Britain, Simpson, Maule and Nicholson sold the same dye at first as "roseine" before adopting the new name.⁷³ One widely used method oxidised aniline with arsenic pentoxide, and dye made that way could contain as much as 6% arsenic.⁷