Clement, Joseph

JOSEPH CLEMENT & BABBAGE'S DIFFERENCE ENGINE

Difference Engine - so called because it was designed to compute tables of numbers (primarily, Babbage envisaged, tables for navigation) according to the method of finite differences (mechanical analogue of De Prony's system), and then automatically to print the tables as they were computed. First Difference Engine was Babbage's great practical engineering project, vastly more complex than any previously conceived mechanical calculator - the advances in machine tools and machining techniques developed in making the first Difference Engine (1823-33) were to have a far-reaching effect on subsequent development of precision engineering.  Difference Engine No. 1 was 8 ft high, 7 ft long, and 3 ft deep, weighing an estimated 15 tons. Design called for some 25,000 separate parts, equally split between the calculating section and the printer.

In June 1823, Charles Babbage received £1,500 from the Chancellor of the Exchequer, Mr Robinson, for construction of his Difference Engine. Project was now launched, and two rooms in CB's house in Devonshire Street were turned into workshops, while a third was converted into a forge. On Marc Brunel's recommendation, CB employed an engineer named Joseph Clement, a highly skilled toolmaker as well as a first-rate draughtsman: this a rare combination at time, and Clement rarer still, as excelled at both of these while also able to produce detailed drawings of elaborate parts and meticulous illustrations of machines. Time and again in the past this unusual combination had secured JC advantage: in 1813, at age of 34 and with £100 savings, he had moved from Glasgow to London to advance himself, and almost immediately after getting off coach at Holborn he offered his services to Alexander Galloway, proprietor of a local machine shop, and was taken on at a guinea a week. Soon proved himself better than anyone there, and so approached Joseph Bramah, one of great pioneering engineers of day, taking with him examples of his drawing - was hired on the spot, and was soon earning 3 guineas per week as chief draughtsman and superintendent of Bramah's Pimlico works. He then moved to become chief draughtsman for Henry Maudslay, and then in 1817 - with savings of £500 - he set up on his own, taking premises at Prospect Place, Lambeth, and soon established a reputation for precision work and fine finishing. He was forever making new machines, improving existing ones, and innovating techniques of manufacture. Was, however, notoriously difficult man to get on with: was blunt and gruff to point of belligerence, and is described as a "heavy-browed man, without any polish or manner of speech". However, despite only rudimentary education and near-illiteracy, he knew his worth, and was quite unintimidated by rich or illustrious customers. Was many times taken to arbitration when people thought he had charged too much - eg. on one occasion I.K. Brunel, engineer to the Great Western Railway, asked if JC could supply a louder steam whistle for his locomotives to replace old ones, which were reedy and feeble: JC inspected whistles, which he scorned as "mere tallow chandler's work", and devised the tools he needed for a new whistle, which worked so well that could be heard miles away. Brunel ordered a 100 of them, but was appalled when got the bill, saying was 6 x more than what they usually paid. JC replied that his whistles were more than 6 x better than previous whistles, and arbitrator ruled in JC's favour.

Thus, if it was precision engineering that was needed, new machine tools, and exquisite engineering draughtsmanship, then JC was the man - as long as you had the money. This was precisely Babbage's situation in the mid-1820s: had a mandate from govt to build his engine, money in the bank, and a colossal task that would make unprecedented demands on the state of the mechanical arts in design, precision and intricacy. With JC hired, CB began in earnest the design of the first fully engineered difference engine, which he called Difference Engine No. 1. He designed and then sketched the mechanisms, and JC then drafted them; CB would invariably consult JC on how he should go about making the parts, and was the inspirational source of the concept and function of the elaborate mechanisms - while JC was the source of practical skill and know-how. Collaboration was v. close, and is now impossible to establish how much of the technical detail of the engine was JC's and how much CB's.

Precision was incessant anxiety, and where existing practice was inadequate, CB and JC devised new tools and modified machines that were to hand, but this pushed costs up. Meanwhile, profounder challenge was need to produce large numbers of similar parts: CB's calculating engines were like no other machines of the time - their massive cast-iron frames identify them as part of the great age of machinery, but they do not resemble clocks, steam engines or textile machines of early-mid C19th, as what is entirely distinctive about CB's designs is that they call for more repeated parts than any device designed up til that time - hundreds and sometimes thousands of near-identical parts. While the manually-operated desktop calculators of the time had 6 or perhaps 8 digits, CB's engines were designed to handle numbers with 20, 30, and even 50 digits - this was then v. ambitious numerical range, and way beyond anything previously proposed. Even a modest-sized Babbage engine has several hundred similar number wheels, and one of the most dramatic visual features of all CB's engines is the overall impression of a regular array of indistinguishable parts.

Reason for so many near-identical parts lies in the way the engines store and manipulate numbers - each digit stored on a gear wheel that he called a 'figure wheel', inscribed with the numbers '0' to '9', and each individual digit in a number has its own separate identical figure wheel. So number '9,542' will be held in the machine by 4 separate figure wheels in a vertical column with the units at the bottom, tens next up, and so on, with the units wheel at the bottom of the stack turned two teeth to represent '2', the tens figure, turned to represent '4', etc. In this way, teh value of a number is represented by a vertical stack of inscribed figure wheels, each one rotated according to the value of the digit it represents. The early designs for Difference Engine No. 1 had 6 columns, each with at least 12 figure wheels; by 1830 this had grown to 7 columns of figure wheels and 16 digits; and by 1840s, CB's Difference Engine No. 2 had 8 columns each with 31 figure wheels. So even the early designs called for near-identical parts in their tens or hundreds, rather than the one-off or small quantities typical of production at the time.

Thus, CB had misfortune to conceive of his engine when manufacturing technology was in transition between craft and mass-production, before there were established methods for the production of similar parts in quantity. Components were made one at a time, and each was slightly different - could only be 'made to fit' by careful comparision of dimension parts, then by finishing with hand tools or by incremental machining, and finally by careful filing, tweaking and reaming by hand. But process was highly labour intensive, time consuming and therefore costly - plus CB also hampered by fact that there was practically no standardization in manufacturing, which meant that he couldn't farm work out to several workshops: each mechanical engineer had his own taps and dies used for cutting screws, and no two were the same, plus each lathe had a different master screw used as the pattern for cutting screw threads, so a screw cut on one lathe would be different from one cut on its neighbour, even in the same workshop. Without standardization, parts produced on different machines were simply not interchangeable - thus CB forced to used only Clement, and had to live with the frustrations of grindingly slow and expensive progress.

N.B. The demand for repeat parts for CB's engine had a far-reaching effect on production engineering in later yrs. If a machine made elsewhere needed a new bolt, the thread cut into the machine, even if undamaged, had to be drilled out and retapped to match the new component. JC saw the waste and trouble of this, and in about 1828 settled on a fixed number of threads in a given length of screw. The scheme was not an immediate success, but a journeyman working for JC on CB's engine, Joseph Whitworth, played a major role in the adoption of standards 20 yrs later through the introduction of the standard Whitworth screw thread, which dominated engineering practice for over a century.

Back with the Difference Engine, the relationship between CB and JC started off well, but by 1828-29 things getting a little fraught, mainly over money. CB's wife had died in 1827, leaving him desolate and as a result he took a Continental tour during 1827-28 meaning he was absent for long periods. Clement was a highly skilled mechanic but had not previously been responsible for so large a project, and even before leaving England, CB had become worried that Clement was less than straightforward - suspected him of developing and constructing at CB's expense special lathes and other tools valuable in constructing the Difference Engine but intended to remain in Clement's workshop. Meanwhile, the £1500 advanced by the Govt had been spent and CB was financing construction from his own resources. However, he successfully petitioned the Govt - with the help of the RS - in 1829, and received a further £7,500 by Feb 1830.

While the finances were being sorted out, CB then took long overdue steps to organize the work more satisfactorily, and refused to pay any more of Clement's bills until a proper system was established for verifying that they accurately represented the work carried out. Two engineers, one chosen by each party, were jointly to inspect the work and accounts, and if they could not agree a third engineer would be appointed as an arbitrator. (This procedure was standard practice, providing an alternative to law courts). Clement chose Henry Maudslay, and CB chose Bryan Donkin; accounts were duly inspected and approved, but in reality so novel was the work  upon which he was engaged that Clement was fairly free to charge what he wanted and inspection subjected him only to the loosest control. Clement doing extremely well - was making a great deal of money out of the Difference Engine, and was determined to keep doing so while he could. CB meanwhile remained v. worried about him, although he had no doubt that he was an excellent craftsman, but big problem was that all the novel and ingenious high-precision machine tools that had been developed and constructed at CB's instigation and funding, now belonged to Clement. Things eventually sorted out, and JC's account settled in May 1830, and work on engine resumed - but only after had lain dormant for over a year.

Relationship strained further when site for housing the Difference Engine was chosen as CB's house on Dorset Street - during 1831, his coach-house and stables were pulled down and a strong fire-proof building erected. New workshops had 2 floors, each 50 ft long, strongly built to support lathes and heavy machinery, with much of the roof made of glass for good lighting. Next to workshops was a fire-proof, dust-free room to house the Calculating Engine. CB also owned a small house where the carriageway from his old stables passed out into East Street: this house was allocated to Clement. Meanwhile, Clement's draughtsman C.G. Jarvis (who was later to work directly for Babbage on superb series of drawings for the Analytical Engines) was not getting on with Clement and was thinking of leaving - warned CB in a letter that "the inventor of a machine and the maker of it have two distinct ends to obtain. The object of the first is to make the machine as complete as possible. The object of the second - and we have no right to expect he will be influenced by any other feeling - is to gain as much as possible by making the machine; and it is in his interest to make it as complicated as possible."

In 1832, CB instructed Clement to assemble a portion of the Engine - this is the working Difference Engine which can still be seen in the Science Museum. Was an engineering triumph - if full Engine planned by Babbage had been completed it would have been the wonder of C19th precision engineering, and even portion which was assembled is one of the finest achievements of first half of the century. Completed portion of the Engine was moved to Dorset Street, and by May 1832 CB was pressing Clement to specify the requirements and arrangements for moving the tools, parts of the Engine, and drawings to the new premises the Govt had built in place of CB's stables and coach-house, but Clement's demands were large. In addition to exaggerated expenses for the move, he asked for £660 p.a. for the expense of keeping 2 establishments and running a divided business, but the Treasury refused to sanction JC's "extraordinary claims": "Mr Clement must recollect that £12,000 of Public Money has been expended on the Engine the great part . .  of which has passed through his hands, from which a very large amount of profit has accrued to him, and that his employment by you has constituted and still constitutes, as my Lords are informed, the principal part of his business, and that he has not been restricted in using for his general business any tools for which the public has paid, and that the further expense of completing the engine will yield him a considerable profit . . . . My Lords cannot but express their surprise that Mr Clement should have advanced so unreasonable and inadmissable a claim."

However, Clement refused to move to Dorset Street, and CB then declined to advance him any more of his own money, telling Clement to present his accounts directly to the Treasury in future. In response, Clement sacked his men and work on the Engine stopped, and this time it was not to start again. Moreover, Clement refused to deliver any of the parts or drawings until he was paid in full. Treasury finally did pay Clement in 1834, and returned drawings to CB. CB then went on trying to find a way to resume work - was loath to make a final break with Clement, because of the difficulty of replacing the specialized machine-tools, plus were not many men of Clement's ability available. However, what CB didn't then realize is that this was effectively the end of the first Difference Engine - with return of drawings he began to reconsider the whole plan of the Engine, both in detail and general principles, and from this came his Analytical Engines - first computers.

From 1834, Clement off the project; he went on to use the large planing machine, built for the Difference Engine in 1831, as his principal source of income for next few yrs. This machine was able to take large workpieces than any other of its kind, and such was demand for it that it was often kept running night and day: at 18 shillings a sq ft of planed work, the planing machine brought in £10 in a 12-hr day.

End of Difference Engine was unpleasant affair for all concerned: arbitrators and mediators had to be used in difficult three-way negotiation between a grumpy CB, an intransigent JC, and the long-suffering Treasury. CB wrote that he was "almost worn out with disgust and annoyance at the whole affair". By time of JC's last payment in August 1834, the DE project had cost the Treasury a total of £17,478 14s 10d, of which just over £2,000 was for the fireproof buildings, and the rest for manufacture and development. (This a staggering sum of money: a brand-new steam locomotive, the 'John Bull' built by Robert Stephenson for shipment to the U.S. in 1831, had cost only £784 7s). By time JC had downed tools he had made 12,000 of the requisite 25,000 parts, with most of the parts for the calculating section of the Engine complete but unassembled, but with parts for the printing apparatus barely begun. Once the loose parts were returned to CB, some were recycled for experiments and testpieces by both CB and later his son, Henry Prevost, who also assembled sets of the original Engine parts into small demonstration pieces after his father's death. Fate of the rest - JC's best and some of the finest examples of precision engineering of the day - was to be the melting pot, or sale by weight for scrap.

Verdict on JC is now difficult: is tempting to demonize him, and see him as saboteur of a great historic enterprise and CB as thwarted victim of JC's stubborness and even greed. (This the view of one of JC's workmen, Jarvis). Typical allegation was that JC was making a meal of the job to prolong the lucrative open-ended contract, and is evidence to support this: measurements made at the Science Museum in the 1980s on the portion of the DE completed by JC shown that non-critical parts, such as spacing pillars, are machined to the same high tolerances as operationally critical parts. However, no doubt that JC was highly acclaimed engineer - was awarded the Society of Arts gold medal in 1827 for lathe improvements, and the next yr the silver medal for his 'self-adjusting double-driving centre chuck' - plus made v. significant attempt to standardize screw threads.