Skip to content

Percy Ludgate

Abstract

Percy Edwin Ludgate (1883–1922), a clerk and accountant in Dublin, published the design of a program-controlled mechanical computer, his “Analytical Machine,” in 1909. It was the second such design after Charles Babbage’s Analytical Engine, conceived independently, smaller, and built on unusual ideas: numbers stored as sliding rods, and multiplication by adding integer “logarithms.” It was never built, and Ludgate was forgotten until the computer scientist Brian Randell rediscovered him around 1971. In between, his paper turned up in 1960 as evidence against Konrad Zuse’s computer patent.

A Clerk in Dublin

Ludgate was born on 2 August 1883 in Skibbereen, in the south-west of Ireland, the son of a former soldier. The family moved to Dublin, where he went to school. Poor health kept him out of the civil service, and he worked for years as a clerk for a corn merchant; during the First World War he was involved in organising fodder supplies for the cavalry. He later qualified as an accountant. He died of pneumonia in Dublin on 16 October 1922, aged 39. No papers survive, and only one photograph of him is known.

The Analytical Machine

For six years Ludgate spent his spare time designing a calculating machine, working out the idea himself and learning of Babbage only along the way (see Charles Babbage). He published it in April 1909 in the Scientific Proceedings of the Royal Dublin Society as “On a Proposed Analytical Machine,” fifteen pages long. Of his drawings, all but one are lost.

The machine was to be about 65 centimetres long, 60 wide and 50 high, far smaller than Babbage’s room-sized engine. A program, coded numbers and instructions on a perforated paper tape he called the “formula paper,” controlled it; a keyboard punched the tape, one row of keys for numbers and one for operations. Results were printed or punched onto tape. The store consisted of two sets of 192 compartments arranged in rings that rotated about a common axis; each compartment held twenty rods that a mechanism slid into one of ten positions, representing a decimal number of up to twenty digits, plus a two-position rod for the sign.

Instead of multiplying digit by digit, the machine looked up integer index numbers for the two digits in a fixed table, a kind of discrete logarithm, added them, and converted the sum back into the product. Division used a converging series, and to compute it Ludgate described what would now be called a subroutine. An electric motor was to turn the main shaft three times a second; Ludgate estimated three seconds for an addition, about ten for a multiplication and up to ninety for a division.

Nature, Engineering and the English Mechanic reported on the design in 1909. In 1914 Ludgate wrote a short essay for the handbook of an Edinburgh exhibition marking the tercentenary of John Napier’s logarithms, describing Babbage’s Difference and Analytical Engines and mentioning, without details, that he had designed a difference engine of his own. Leonardo Torres Quevedo, who published his “Essays on Automatics” in 1914, was the next to describe a program-controlled calculator (see Leonardo Torres Quevedo and El Ajedrecista). In Germany the civil servant Emil Schilling filed a patent in 1926 for a punched-tape program control that drove an ordinary calculating machine (see Germany’s Computing Pioneers). Ludgate’s index-number multiplication reappeared, presumably without any knowledge of his paper, in a Faber-Castell slide rule, the 366 “System Schumacher.”

Ludgate against Zuse

In November 1951 Konrad Zuse applied for a German patent on a computer derived from his Z3. The Triumph Werke of Nuremberg, which made motorcycles as well as typewriters and accounting machines, opposed it. In 1959 the examiners at the patent office in Munich were inclined to grant the patent, but on 12 January 1960 a new objection arrived from Triumph’s patent attorney, G. B. Hagen, who described Ludgate’s 1909 paper over nine pages with a precise citation and also invoked Babbage, Claude Shannon and Louis Couffignal. Zuse’s attorney warned him on 28 January that it looked like a serious piece of prior art. Zuse replied that Ludgate’s paper tape had no counterpart in his application and that the coding system Triumph claimed to find there was “a product of the opponent’s imagination.”

In 1961 Hagen added Couffignal’s 1937 doctoral thesis. The patent office refused the patent in 1962, and on 14 July 1967 the Federal Patent Court dismissed Zuse’s appeal: the invention was new and an advance, but lacked inventive height (see Konrad Zuse and the Z3). How Triumph came across an obscure Irish paper is unknown. Zuse believed IBM had helped, and IBM had had an office in Dublin since 1956; but a copy of the paper had also reached the library of the Deutsche Seewarte in Hamburg.

Rediscovery

Brian Randell, a British computer scientist and historian, came across the 1909 paper and published it again in 1971, which made Ludgate known among historians. From 2016 Brian Coghlan, a former lecturer at Trinity College Dublin, led renewed research into Ludgate’s life and produced a detailed reconstruction of the machine, published in the IEEE Annals of the History of Computing in 2021. On 15 October 2022 a plaque was unveiled at 30 Dargle Road in Dublin, the house where Ludgate worked on his design.

📚 Sources