Simon Winchester, Exactly (William Collins 2018)
https://www.goodreads.com/book/show/40037112-exactly?ac=1&from_search=true&qid=gFAM49o7Yk&rank=1
This is the exciting true-life history of precision engineering, beginning with establishing the difference between the concept of ‘accuracy’ and the concept of ‘precision’.
The book features a fascinating structure, being enchaptered according to ascending measures of precision. A 1/2 inch ‘tolerance’ in shoes, e.g., is acceptable, but for the Hubble telescope, a measurement error on the mirror (later repaired) 1/50 the thickness of a hair rendered its images from outer space useless.
The remarkable Antikythera Mechanism was supposed to predict eclipses, but when tested in modern times proved to be very inaccurate at it. Interestingly, it never occurred to ancient Greeks to use it as a clock.
The centuries long struggle to measure longitude, the time difference between a ship and homeport, was finally solved by the British success of Harrison, driving the success of the British Empire. But the story shows that a mechanism has to be both duplicable and manufacturable at a reasonable cost.
John ‘Iron-mad’ Wilkinson’s patent for boring cannon barrels and James Watts steam engine fuelled the industrial revolution.
Henry Maudsley’s micrometre measured the actual dimensions of an object. ‘The ideal of precision’ we owe to Maudsley.
In early America muskets were not made with interchangeable parts but had to be repaired individually by a blacksmith. This led to defeat at the Battle of Bladensburg in 1814. Humans will inevitably err; precise machines will not.
Galileo had first noted the relationship between length and time, as a pendulum swing rate depended on the length of the pendulum.
Eventually, mankind developed the concept of ‘traceability’ – the ability to connect time to an official US atomic clock. James Clerk Maxwell discovered that constants in nature were to be found at the atomic level i.e., the wavelength of light.
There is an interesting appendix on the history of measurement. A six-year survey on the length of the meridian from Dunkirk to Barcelona divided by 10 became the standard post-revolutionary French meter, then cast in a platinum étalon.
However, this was before continental shift was understood, and the measure of the meridian was eventually found to be off by 2/10 of a millimetre. It took 7 decades for the international community to agree on new standards.
I borrowed this book at my scientist sister’s house and only read up to page 124, but I can’t wait to read the rest of it someday.









