How an astronomer's iron-salt experiment from 1842 became the standard way to copy engineering drawings, why the lines were white on blue, and why the word outlived the object.
Ask someone to picture a blueprint and they will describe white lines on a deep blue sheet, probably with a coffee ring on it. The process was already fading by the 1950s, and yet the word is everywhere: policy papers, genetics, software frameworks. It is worth asking how a chemical trick from the 1840s became the general word for a plan.
In 1842 the English scientist John Herschel, better known for astronomy, was experimenting with the way light acts on iron salts. He found that paper coated with ferric ammonium citrate and potassium ferricyanide, exposed to sunlight and then rinsed, turns a deep and permanent blue wherever the light has reached it. He called the result a cyanotype.1
The chemistry is short. Ultraviolet light reduces the iron in the citrate from iron(III) to iron(II). The iron(II) then reacts with the ferricyanide to form Prussian blue, an insoluble pigment with the idealised formula . The unexposed chemicals stay soluble, so fixing the image needs nothing but a rinse in water: no silver, no darkroom, no fixer.
That simplicity is the whole story. Silver photography was delicate and expensive; cyanotype was cheap, tolerant and slow, and could only ever be blue, which made it a poor way to take portraits and a very good way to copy things.
The first person to see that clearly was the botanist Anna Atkins. She wanted accurate records of algae, which are tedious to draw, so she laid dried specimens directly on sensitised paper and put them in the sun. Where the plant blocked the light, the paper stayed white. The result is a photogram: a life-size white silhouette on blue.
From 1843 she issued these prints in parts under the title Photographs of British Algae: Cyanotype Impressions, generally regarded as the first book illustrated with photographs. More to the point here, it showed what cyanotype was good for: no camera, no lens, and a same-size copy of any flat thing with no draughtsman in between.
Engineering offices had exactly that problem. A ship or a bridge needs many copies of each drawing, and for most of the nineteenth century each copy was traced by hand, slowly, with a fresh chance of error every time. Once cyanotype paper was sold commercially, from the 1870s, the routine changed. The draughtsman inked the master drawing on translucent tracing paper or cloth. The tracing was clamped in a glass frame over a sensitised sheet, put out in the sun for some minutes, then washed and hung up to dry.
The colours follow directly from the chemistry. Light passes through the blank parts of the tracing and turns the sheet blue. The ink lines cast shadows, the paper beneath them never reacts, and the unexposed salts wash away to leave bare white paper. A blueprint is a negative, a contact print of a drawing. Nobody chose white on blue for its looks. It was what the cheapest copying process produced.
It had drawbacks. Washing and drying made the sheets cockle and shrink a little, so dimensions were read from the figures and never scaled off the sheet. Pencil notes do not show on dark blue. And Prussian blue breaks down in alkali, which is why conservators keep cyanotypes away from the buffered papers used for most other documents.
The replacement came out of the dye industry. Diazo papers, sold from the 1920s, are coated with a light-sensitive diazonium salt and a colour coupler. Light destroys the salt wherever it gets through the tracing. Under the lines the salt survives, and when the sheet meets ammonia vapour it couples into a dark dye. The copy is a positive, dark lines on a white ground, nearly dry out of the machine, and you can write on it. By mid-century the whiteprint had pushed the true blueprint out of most offices. Anyone who worked near one of those machines remembers the ammonia.
Then the copying problem dissolved. Large-format photocopiers took over from diazo machines, and computer-aided design, which grew from research systems in the 1960s into ordinary desktop software in the 1980s, removed the hand-drawn master itself. A drawing became a file. Copies were plotted on demand, and later not printed at all.
So why do we still say blueprint? Partly habit. The trade kept calling its drawings blueprints through the whiteprint era, when they were no longer blue, and the plotter era, when they were no longer prints. But the word also names an idea nothing else names quite as well. A blueprint is complete, it comes before the work, it is meant to be followed exactly, and everyone on the job holds an identical copy. "Plan" is vaguer. "Drawing" says nothing about authority.
There is an irony in that. The blueprint was never the design. The design was the inked tracing, kept safe in a drawer. The blueprint was the cheap, disposable copy sent out to get dirty on site. The word we now use for the master plan originally meant the thing you were allowed to spill coffee on.
Cyanotype is still practised, by artists and in classrooms, because it remains one of the simplest photographic processes there is. The blue that made Atkins' seaweed and a century of bridges is the same blue.
From the Greek word for dark blue. Herschel is also usually credited with bringing the words "negative" and "positive" into photography, which makes the blueprint, a negative made without a camera, doubly his. ↩