Friday, 27 January 2012

Georg and Edvard Scheutz and the Construction of Difference Engines

So after two inactive years, I've decided to reactivate my blog. This post is something I wrote about four years ago and never published, but enquiries made today on Twitter by @Dr_Black about surviving difference engines made me drag it out.

Georg and Edvard Scheutz and the Construction of Difference Engines

Johann H. Mueller, (1746-1830) began making calculating devices in the 1780s, and it was he who conceived of the idea of calculating and then printing mathematical tables. Evidence now suggests that Charles Babbage, who started thinking about calculating machines in the 1820s, was influenced by Mueller. However, neither Mueller nor Babbage were able to construct a successful machine.

Babbage described his Difference Engine No. 1 in his publication A Letter to Sir Humphry Davy... on the Application of Machinery to the Purpose of Calculating and Printing Mathematical Tables. London, 18221. However, an attempt at construction ground to a halt after some 25,000 parts had been assembled, and it was never completed. By 1834 Babbage had conceived of a far more complex machine, Difference Engine No. 2. This too was never completed in Babbage’s lifetime2, and once again, Babbage intellectually moved on, conceiving of an even grander project, his Analytical Engine.

While the Difference Engine was intended to calculate mathematical tables and print the results, the far more ambitious Analytical Engine was designed to execute any mathematical operation by following a program on punched cards3.

Meanwhile in Sweden, Georg Scheutz (1785-1873), publisher, translator and inventor, read an article on Babbage’s Difference Engine in the Edinburgh Review4. Stimulated by this, he conceived of his own version of the Difference Engine, but decided to concentrate on the output, i.e. the printing of the tables. In this he was radically different from Babbage. Scheutz’s original design was realised by his son Edvard, who in 1837, at the age of 16, offered to build it. Working in a workshop at home with a lathe and a few other simple tools, the engineering student managed to complete a functioning machine by 1843. This was the world’s first fully functioning mechanical calculator, and it was complete with a printing unit for output5.

This machine was tested by the Swedish Royal Academy of Sciences under the watchful gaze of the famous chemist J.J. Berzelius, who testified favourably on its construction and function. With this endorsement, the Scheutzes attempted to obtain orders from several European countries, but in vain. The British Treasury declined on the grounds that they had already expended a huge amount of government money on Babbage’s designs6 and could not invest in a foreign invention. In the end, the Swedish government gave the Scheutzes a grant for further development. J.W. Bergström’s engineering workshop in Stockholm constructed the Scheutzes’ second difference engine, completing it after only a year’s work, in 1853.

This was a far more complex machine. Whereas the first difference engine used five decimal places and three differences, this new model used fifteen places and four differences. The Scheutzes applied successfully for a patent for this second machine at the London Patent Office in 1854. This calculator won a gold medal at the Universal Exposition in Paris in 1855, and was later sold for use in the observatory at Albany, New York. It is now on display at the National Museum for American History.

A third machine, built by Bryan Donkin’s mechanical workshop in London in 1859 was used for many years at the Registrar’s office for calculating and printing statistical tables. This is now on display at the Science Museum, London.

1 The last four copies of this work to have sold at auction are (all prices hammer):

Christie's New York, Feb 23, 2005, lot 16, $32,000; Sotheby's, Mar 20, 2003, lot 385, £7,000; Swann, Apr 19, 2001, lot 165, $8,000; Bloomsbury, Oct 28, 1999, lot 15, £5,500.

2 The Science Museum in London completed the build in 1989-91 using Babbage’s plans and engineering at 19th century tolerances.

3 This was first reported by Luigi Federico Menabrea, which was in turn translated by Ada Lovelace and published in England in 1843.

5 This machine was rediscovered in 1979 after it had been lost for 100 years. Michael Lindgren, while completing a study of the Scheutzes noted that in the inventory of the estate of Edvard, it had been deposited in the Nordiska Museet. Lindgren found it in the storehouse of that same museum still in its original mahogany box.

6 Babbage received some £17,000 for the development of his designs; a steam locomotive in the 1840s cost roughly £800 to build.