Watt’s Cottage

Conception of an Idea


James Watt had worked as an instrument maker in Glasgow and came into contact with many of the leading scientists of the era, including Professors John Robison and Joseph Black.  He repaired a working model of a Newcomen steam engine for Glasgow University and realised that it was very inefficient.  Hot steam was put into the cylinder and then condensed by introducing water.  This created low pressure below a piston and atmospheric pressure forced the piston down and so drove a connecting rod.  Most of the energy was lost in cooling the cylinder.  Watt’s critical insight came in May 1765 when he was 31 years old and conceived the notion of using a separate chamber to condense the steam and the commensurate apparatus necessary to do this.  This meant that very little heat was absorbed into the cylinder itself on each cycle, and thus far more heat from the steam was made available to perform useful work.  Watt made a working model later that same year.

Amongst the learned men of industry to whom Watt had been introduced was Dr John Roebuck.  Roebuck had very successfully upscaled the production of sulphuric acid and is often seen as the father of process chemistry.  In 1759 he had been one of the founders of the Carron Iron Company and provided much of the initial technical advice.  On the back of this he had leased the coalfield and saltpans at Kinneil near Bo’ness from the Duke of Hamilton and moved into Kinneil House in 1760.  He built a new saltpan whose scale was far in excess of anything previously seen.  The coalfield was working deep seams and he replaced the horse gins with Newcomen engines to pump water out.  Not surprisingly therefore Watt wrote to Roebuck in August 1765:

I have tried my small model of my perfect engine, which hitherto answers expectation, and gives great, I may say greatest, hopes of success… in short, I expect almost totally to prevent waste of steam, and consequently to bring the machine to its ultimatum.” 

It was very timely for Roebuck’s coal mining ventures suffered from disastrous flooding that year.  However, it was to take ten years to achieve the ultimatum.

Illus 2: Kinneil House c1920.

The first point of contact between Watt and Roebuck is not known, but it is clear from the letter just referred to that they had been corresponding for some time.  In that letter Watt also mentions a coal wagon design that he had executed for one of Roebuck’s waggonways.  That autumn the two collaborated in the design of some boilers.  One described as a ‘testudo’ had an arched top.  It evaporated 25 lbs of water with 4 lbs of coal in about 44 minutes once boiling had commenced.  The grate was formed from tubes filled with water while other tubes carried the heat through the water space.  Roebuck’s boiler was 9ft in diameter and deep at the crown to contain the steam.  At the same time Watt was designing a much larger version for his own experiments with the fire in a central vertical tube.  He also experimented with methods of firing to produce as little smoke as possible by stoking with only a few coals at once and keeping the fire “thoroughly inflame.”  Roebuck probably paid for this work and this may be how he first became involved financially. 

The two men must have discussed the new concept for improving the steam engine and later in August Watt wrote to Roebuck  saying:

I have tried my new engine with good success; for though I have not been able to get it perfectly air-tight from its bad materials, yet, immediately on turning the exhausting cock, the piston, when not loaded, ascended as quick as the blow of a hammer, and as quick when loaded with 18 lbs (being 7 lbs on the inch) as it would have done if it had an injection as usual.” 

The valves were probably being worked by hand.  A traditional story tells of Watt demonstrating the model in the dining room at Kinneil House, presumably using the fire in the hearth to heat the water, but this may be apocryphal.  However, Roebuck later told a colleague that he had been convinced of the “justness of the Principles” of Watt’s separate condenser through the trial of a small one at Kinneil and so supported him.  It was claimed that this had a 9ins cylinder.  The performance was sufficiently impressive to stimulate Roebuck’s interest and thereafter he became more and more involved.  For Watt, having an influential and renowned industrialist on board to develop and promote his invention must have seemed ideal.

It appears that, at first, Roebuck took some convincing about the doctrine of latent heat and the other discoveries of Black and Watt.  Black said that Roebuck took a lot of persuading to understand fully

the nature of steam, of the great quantity of heat, and consequently of fuel, necessary to produce it, and of the importance, therefore, of preventing the waste of it.”  “Mr Watt was obliged to yield for some time to the Doctor’s confidence in his own great experience… But Dr Roebuck had too much judgement not to see the conclusiveness of the experiments by which the doctrine of latent heat is established not to yield to their force.” 

On 13 September 1765 Roebuck wrote to Watt saying that he hoped to see him “at Kinneil before the end of this month, and hope Dr. Black will accompany you.”  It would have been on that occasion that Watt carried out an experiment with a piece of apparatus which he had devised and demonstrated the latent heat of water to Roebuck.  It may not have been the concept of latent heat itself which Roebuck had difficulty in accepting but Watt’s extension of this doctrine into what later became Watt’s Law.

Illus 3: The Loading Yard at Carron Works with hoops and sugar pans lying about.

Despite this apparent reticence, Roebuck’s letter of 13 September also encouraged Watt to proceed with his work on improving the steam engine: “It is certainly best to press it forward with all speed, whether you pursue it as a philosopher or as a man of business.”  This is typical of Roebuck’s character, wanting to see it running as soon as possible to ensure that no one else took up the idea and that a return could be made on the investment.  Roebuck then helped Watt with his steam engine experiments in many ways.  Not only did he procure materials for models but he carried out experiments with Watt into the properties of steam. 

He allowed Watt to run and test a Newcomen engine in one of his coal mines, probably not that at the Schoolyard Pit.  The cylinder diameter of that engine was 5ft 3ins (and it is now in Kinneil Park), but that tested was smaller with a cylinder of 32ins and stroke of 5ft 2ins.  Watt calculated the evaporation rate of the boiler, the volume of the cylinder and the surface area of the piston.  He determined how much steam was condensed in warming the cylinder before each stroke and found it was 97.2 ‘solid feet’ while the cylinder only contained 28.6 ‘solid feet.’   This confirmed just how much steam was wasted reheating the cylinder.  The pressure on the piston was 8.5psi.  This was the first of many trials on Newcomen engines which Watt would conduct and it may have helped him to convince Roebuck of the correctness of his theories because “it served to show the great waste of fuel and steam.

Watt developed a great understanding of the Newcomen engines at Bo’ness and suggested improvements.  The expertise was not wasted and at some stage in 1765 he went into partnership with Robert Mackell, a millwright and engineer in Falkirk, to build Newcomen engines for other customers.  He needed the money to look after his family and to fund further work on his new design of engine.  The first of the engines produced by the partnership was a small engine to pump water to a turpentine still-tub at Carron Wharf near Carronshore, which was completed by the end of the year.  The parts must have been made by Carron Company and Watt was to become familiar with the works there.


By October 1765 Watt’s model of his improved engine was working well and he provided it with a steam jacket for the cylinder in order to retain the heat.  There must have been long and intricate discussions in the evenings in front of the fire at Kinneil House where Watt was staying as Roebuck’s guest.  Roebuck was eighteen years Watt’s senior.  He was a knowledgeable scientist and recognised this as a nascent area of study – the dawn of a new era in thermodynamics.  Professor Black noted that Roebuck, in addition to his knowledge of medicine and chemistry, “had no small degree of mechanical knowledge and ingenuity.”  He was also a great moral support, a point later emphasised by Professor Robison.  Roebuck, he said, was “well accustomed to great enterprises, of an undaunted spirit, not scared by difficulties, nor a niggard of expense.  Such a man was indispensably necessary to one of Mr Watt’s character, – modest, timid, easily frightened by rules and misgivings, and too apt to despond.”  He too had a wide circle of scientific and industrial contacts amongst whom was Matthew Bouton, about whom we shall be hearing more.  Roebuck was also a pragmatic industrialist accustomed to resolving the problems associated with scaling up ideas.  He would have advised Watt to construct a large engine to his new design.

Illus 4: Scale Drawing for the 2½ft diameter cylinder of the first Kinneil Engine, November 1765.

Illus 5: Muirhead’s Drawing of the 1766 Engine at Kinneil.

In November 1765 Watt sent Roebuck a drawing of an inverted cylinder with a separate condenser.  The cylinder was to be cast and bored at Carron for an engine to be erected at Kinneil.  As well as details of the cylinder design, there are sketches for covers and for the piston.  The bore at 24ins was “to be made as truely round & equally wide as possible.”  The fact that this even had to be stated illustrates how difficult it was to achieve in practice.  The total length was 7ft but, allowing for packing flanges and the like, the stroke must have been nearer six.  The diameter of the piston rod was 2¾ins.  Watt appears to have expected Roebuck to order this to be cast and machined at Carron but there are no details about payment.  Roebuck obviously agreed to help finance Watt’s first attempt to erect a large engine at Kinneil.

As Watt was intending to use steam to force the piston up and down he could no longer cover the piston with water to provide a seal, as was the case with the Newcomen engine.  He experimented with cloth instead, and then with pasteboard made from old rope.  Watt was at Kinneil in December directing the construction work.  The boiler was to be of the “common kind” made with the thinnest plates.  Watt’s greatest problem was the design of the condenser. 

He had decided to use a surface condenser rather that a direct injection of water and proposed a plate type with sixteen double plates of a foot square set half an inch apart. There had been some argument between Roebuck and Watt over the area for condensers, Roebuck advocating a large area for speed of condensation, while Watt was concerned about the large quantity of water required.  Both were right.  Watt also had some new ideas for the air pump.

Illus 6: Segment from a Boiler cast at Carron Works in 1766 and later built into its offices.


The next we hear about the first engine for Kinneil is in February 1766 when Watt wrote to Roebuck:

The piston-rod for the large machine is near finished.  I am somewhat at a loss about sending it, as if sent on a common cart Mr Singleton and the rest of your people will unavoidably see it, and it being hollow will occasion speculation.  I believe it will be best to send it in a box.” 

The only reason for the problems and expense of making a hollow piston rod was for it to act as a drain for the condensate out of the cylinder in an inverted engine.

Watt and Roebuck must have been very confident of success to have ordered a cylinder for so large an engine so early in the experiments.  The illustration of the engine published by Muirhead shows a heavy weight hanging on the end of the piston rod.  This enabled the engine to perform work and provided a measure of its capacity.  There are clear omissions in the design of the condenser and there are no valve or pump operating mechanisms shown.  The dimensions correspond with the drawing of the cylinder which Watt sent to Roebuck. 

The cylinder had not been cast at Carron by the middle of February.  Watt may have referred to it in April when he wrote “Your cylinder is finished and sent to Carron wharf, with the ends on.”  From Carron Wharf it would have been shipped along the coast to Kinneil.  The wording suggests that Roebuck was taking the lead.  Early in 1766 Watt, still in Glasgow, congratulated Roebuck on hearing that the “Quality of your Coal, when found was equal to your most sanguine hopes & the richness little inferior to expected.”  This was good news indeed as coalmining operations were to fund the experimental steam engine.  Meanwhile Roebuck’s sulphuric acid works had helped to subsidise the improvements to the colliery but the market price of the acid was falling as other producers came on stream.  However, when the cylinder arrived it was found to be too badly bored to be of any use, “though the best Carron could make.”  The result was that this engine was abandoned.  An undated list of expenses on “Doctor Roebucks new engine” must refer to this first attempt at Kinneil because this would have been the only one needing a hollow piston rod as well as a partitioned condenser which cost £4.4s.  While there are costs for other small items and a steam chest at £1.6s, there is no price for the cylinder or framing.  Gardner was paid £3 7s 6d for 45 days work and Osborne 9s 4d for 8 days.  These were costs which Watt may have met himself – John Gardner and Murray Osborne were apprentices in his Glasgow instrument-making shop.  At this point, Watt ceased further experiments on any of his engines for well over a year.

Carron may not have charged for the defective cylinder.  It had been grappling with the problem of precision boring for some time and realised that cylinders could provide a lucrative market.  However, when its cannon were rejected by the Admiralty it had to take the problem more seriously.  It was evident that guns cast solid and then bored were far more reliable than those cast around a core.  Accordingly John Smeaton was commissioned to design boring mills for ordnance and cylinders but it was 1771 before they were up and running.

Watt needed practical experience on existing steam engines and continued his partnership with Robert Mackell.  In this period they erected:

  • 1765 – a small engine to pump water to a turpentine still-tub at Carron Wharf near Carronshore
  • 1766 – an engine with a 24ins cylinder for Bruce at Kennet.
  • 1767 – 45ins cylinder at Newton of Ayr.
  • 1768 – Engines with 18ins & 20ins cylinders for Colville at Torryburn.

This work took Watt away from Kinneil and so Roebuck engaged Joseph Hately to maintain his pumping engines.  Carron Company recommended Hately as “a Gentleman of Merit & Ingenuity” to those wishing to purchase engine parts from them.  By October 1766 he is mentioned in connection with “Doctor Roebuck’s new Engine.”  That was probably the Newcomen engine installed at the Schoolyard Pit which, with a diameter of around 5ft, was massive by the standards of the day.  Relations between Watt, Roebuck and Hately seem to have been extensive and at first friendly. 


Pressure of other work prevented Watt from carrying out any more trials until the early part of 1768.  He had his instrument business and retail shop to supervise; as well as experiments to obtain alkali from salt.  The shop had been losing money and was the chief cause of Watt’s debts.  He also helped Mackell to survey the line of the Forth and Clyde Canal.  The two men went to London to promote the Parliamentary Bill for it.  On the way back Watt called in at the Soho Works in Birmingham and was impressed by what he saw.  Its owner, Matthew Bouton, was not there but Watt met other members of the Lunar Society including Dr William Small who became an important supporter of Watt’s work.  Watt went on to do more survey work on canals and river improvements.   It may have been at this time that he agreed with Roebuck for a two thirds share in any patent for his perfect engine in return for the payment of his existing debts, which came to £757.4s.7d, and the cost of obtaining the patent.  At all events, in January 1768 his thoughts again turned to steam engines.  At the end of January 1768 Watt wrote to Roebuck concerning the rotary engine.  He then returned to experiments on small models of around 8ins diameter.  In the middle of April 1768 Watt informed Roebuck that he had been testing his engine, which seems to have had a conventional layout with a steam jacket.  Most changes probably focused on the design of the condenser which was found to heat up during long running.  Watt also introduced self-acting valves and twin air pumps.  On 29 April he reported to Roebuck that the engine “doing twice as well as the common, is, I think, absolutely certain from what Mr Jardine saw; and there is little reason to fear but what it will do all we expected.”  George Jardine was Roebuck’s secretary.  At the end of May 1768 Watt was optimistic and wrote to Roebuck wishing him “joy of this successful result, and hope it will make you some return for the obligations I ever will remain under to you.”  

The partnership with Mackell ended in March 1768 when Mackell became sub-engineer for the approved Forth & Clyde Canal under John Smeaton.  Watt had learned a lot from this practical experience and made modifications to the designs of the Newcomen engines.  Two of his designs, for Leadhills and Carron Iron Works, which were not executed, replaced the traditional beam with a wheel.  The Bo’ness mines returned to profitability in 1768 and Roebuck told Boulton that there as a “good prospect of its repaying us all our expences.”  These expenses had not just included the new engines but also waggonways and harbour facilities.  Roebuck had borrowed money from Boulton.

However in June 1768 Roebuck and Watt were stunned by news of a patent obtained by Joseph Hately for a steam engine with a double cylinder with cold water between the two with the object of keeping the interior as cool as possible.  This would have improved the vacuum through greater condensation against the cylinder walls but would also have substantially increased steam consumption – an example of the right principle wrongly applied.  It also begs the question as to whether or not Hately had seen the steam jacket on the cylinder of Watt’s 1766 engine at Kinneil and misunderstood the use to which it was intended.  It may have been for this reason that relations between Roebuck and Watt on the one hand and Hately on the other deteriorated.

June 1768 also saw Joseph Hately employed by the Carron Company to improve its returning engine, which had been erected by John Smeaton to raise spent water from the waterwheels back into the Forge Dam.  After Mackell’s departure Watt continued to erect and maintain Newcomen engines, presumably because he had already contracted to do so. 

In the light of Hately’s patent Roebuck and Watt decided to take out a patent of their own.  Watt had to prepare a petition outlining the nature of his invention and then to make his patent application in person.  So at the beginning of August 1768 he set off for London and made the necessary oath on 9 August.  Thomas Handley was appointed as his patent agent in London and the application began its slow passage through the offices of the Attorney General, the Secretary of State, the Privy Seal, the Signet Office and the Lord Chancellor.  At the end of August 1768 Watt met Matthew Boulton in Birmingham.  Ever since his previous visit Watt had been corresponding with academics in the area and felt at home there.  Boulton had been about to erect a steam engine at his Soho works using some of Watt’s principles but held back on hearing of the patent application.

In October 1768 Watt was at Torryburn erecting the Newcomen engine.  Roebuck was getting impatient with the delays on the new design and told Watt,

I want much effectually to try the machine at large.  You are letting the most active part of your life insensibly glide away.  A day, a moment, ought not to be lost.  And you should not suffer your thoughts to be diverted by any other object, or even improvement of this, but only the speediest and most effectual manner of executing one of a proper size, according to your present ideas

(letter dated 30 October 1768). 

These ideas had advanced considerably since his failed 1766 trial at Kinneil.  Watt evidently wanted to continue with his experiments on models in order to make yet further improvements and Roebuck on 9 November stressed “I am as solicitous as yourself that every trial should be made of the small engine, which can be speedily executed, particularly the alteration of the condenser.  But I would not lose a single moment in completing what was resolved on.”

Clearly Roebuck was frustrated by Watt constantly seeking further improvement before he had put those already deigned into practice.  At around this time George Jardine wrote to Watt from Bo’ness:

“The doctor was been with you at Glasgow… I waited to find… if he had in any respect consented to the proposal from the South [a partnership with Boulton]; but understand, that the more he is convinced of the practicality of the scheme, the keener he is of carrying it to practice yourselves for your mutual advantage; and I am perfectly convinced that though he may mistake particulars, it is one of the strongest motives that you may not only have the honour but the advantage of your own ingenuity… My opinion is, James, that you will find it necessary, on account of your intimate connection, to fall in with his sentiments, and contribute as speedily, and as vigorously as you can, to bring it to some conclusion.  He has proposed to erect one immediately at Bo’ness.  I know very well the objections you have to this, and indeed to everything which will give you further trouble; but I apprehend when you consider deliberately the following particulars, you will endeavour to pluck up spirits for a few months longer.  You are yourself perfectly satisfied as to the success you expect in practice, and I really apprehend that there is now so little to do for that purpose, that it is unnecessary to make any further division of your profit to a third person; for the very nature of your improvements is such that it is impossible it can fail to succeed much to you interest, even though it should not be carried to such perfection as might be expected from the gentleman in the south’s assistance.  I know you will think the erection of one at Bo’ness a further trouble, but I am really of opinion that even that trouble will be productive of considerable advantages.  It would be hard to be disappointed in any respect in the first attempt that you make, and it would surely be better that you should have not only a model, but a finished engine, either to proceed upon as a pattern or to dispose of, as may be convenient.  Nor do I apprehend that your trouble will be so great as you are ready to imagine.  You judge of it too much from what is past.  You are now in a very different situation, and can begin at principles which you had formerly to labour at by experiment and calculation.  The person who is your operator has surely now learned to take some of the trouble off your hand; and I flatter myself that it would now be nothing more than mere amusement to attend to the proportions and accuracy of the execution.  The Dr proposes, not only in the present but in future undertakings, to have everything that requires a great nicety and exactness, to be made by the best workmen, and, as far as possible, to follow the method of watchmakers, who have little more to do than to adjust and combine materials.  This may indeed be not such a perfect scheme as to have the whole conducted in the same manner, but under particular inspection; but it may nevertheless do well, and at least ease you of a part of your trouble; and all improvements in this respect will occur to you much better afterwards.  I wish earnestly, therefore, that you would endeavour to keep away those plague headaches for a little longer.  You have many motives, man, to have good spirits.  You are surely very near to something that will be much to your advantage: the happiness, or at least the interest, of your family, – your own ease and amusement, – that life of ingenious indolence which you have often figured out to yourself, are all within prospect; – not to mention the honour of a discovery of so much importance, a circumstance which few would think so moderately of as yourself.  I would, therefore, have you resolve immediately to spend the great part of the summer here…”

1769.  Watt told Roebuck in November 1768 that he had “written the specification two or three times over, but am not yet satisfied with it, though it is better than it was.”  The patent was sealed on 5 January 1769 which gave Watt till the beginning of May to submit this specification.  As the trials were not completed he took Small’s advice and concentrated upon principles rather than the actual machine – “a Method of lessening the Consumption of Steam and Fuel in Fire Engines.”  It was all-embracing and provided security for the scope of later improvements.  Watt swore to the specification at the end of April 1769 at Berwick on Tweed.  The fees eventually came to £88.8d.9d.  With travel and other expenses it had cost the not inconsiderable sum of £170. 

While Watt was struggling with the text of his patent specification he was continuing to experiment with a model at Glasgow and preparing plans for a large engine at Kinneil.  He had trouble in obtaining sufficient cooling water at Glasgow, at least at Kinneil there was a small stream and the water there could be used to drive a waterwheel which in February 1769 Watt proposed should be used to work his condenser. 

In 1769 he returned to Bo’ness, erecting two Newcomen engines for Dr Roebuck at Bo’ness of the same sort as those at Torryburn.  One of these was at the Taylor Pit.  It pumped water out of the coal workings there, which were higher than those of the large Schoolyard Pit, and so relieved that engine.  Of the engine at Taylor’s pit the workmen could only say that it was the fastest one they ever saw.  From its size, and owing to its being placed in a small timber-house, the colliers called it the “box bed”.  The second one must have been at the Burn Pit near Kinglas (afterwards transferred to the Temple Pit where it continued in operation until 1825) and for many years was known locally as “Watt’s spinning wheel.”  It was made at Carron Iron Works and was the first engine to be used for winding coal. 

Roebuck terminated Hately’s employment on 19 May 1769 and wrote to Watt:

“Mr Hately left this place today.  I have closed all accounts with him.  He is to call on your house and bring the contract with Ld Crawford which I ?signed? and you are desired to see my name scratched out.  I am also to pay him eighty pounds at your house in three weeks which is in ?—? Of the Two Hundred that I received when I met him at Glasgow – I hear that he has already sent for a copy of your specification and that he charges me with revealing his secrets to you.  I am glad however that he cannot charge me with revealing your secrets to himself.”

Roebuck junior consulted Watt about problems they were having with the pumps at the West or Wester engine in November 1769 and the following April Watt sent his “lads” to effect some repair there.  The engine was located at Capies Point on the low road into Bo’ness from the west.  During the summer of 1770 Watt designed a new type of beam for this engine made out of four baulks of timber, braced together in an open framing, possibly based on something he had tried on his own engine at Kinneil.  Roebuck junior also wanted Watt to look at the Dunmore engine in order to keep it running for a further couple of years while they extracted the coal reserves.  The coal and saltpans at Dunmore were also leased by the Roebuck family – there had been a steam engine there since 1720, amongst the first in Scotland.  The design of the Newcomen engine had hardly changed in those 50 years.  During September Watt carried out tests on this engine. He assisted with the maintenance of Roebuck’s other colliery engines.  Ironically, this may have taken him away from getting his own engine running properly. 


Illus 7: Kinneil House with the Medieval Church on the top left and Watt’s Workshop below it, c1947.

During the autumn of 1768 Watt had begun designing his much larger engine to be erected at Kinneil and wrote to Roebuck on 9 November “On considering the engine to be erected with you, I think the best place will be to erect a small house in the glen behind Kinneil.”  This workshop, now known as “Watt’s Cottage,” was duly constructed 20m to the south of Kinneil House. Watt was preparing drawings at the end of April and into May 1769.  On 8 April Roebuck told Watt to bring his wife and family to stay at Kinneil whilst he conducted the experiments.

Illus 8: Drawing of Watt’s Cottage from Samuel Smiles “Lives of Boulton & Watt,” 1865.  The roof was already covered with ivy.  It is interesting to note the amount of water in the stream and the waterfall downstream of the footbridge.



On 28 April 1769 Watt told Small :

“I have begun to erect an engine of an 18 inch cylinder and 5 foot stroke, at Kinneil, the doctor’s place. However, it is to be done as privately as we can, as we do not propose to puff.”

Despite this statement he did not send the plan and elevation to Roebuck until the first of May with the request that “I wish the wood of the frame to be got ready, but not to be put up till I come.” The engine was to have the normal layout of a reciprocating beam engine. As the smith work might cause delays. Watt suggested that a start might be made on the boiler as well as the chains which were to have three rows of links made from the best tough iron.  He told Small at the end of May that he hoped to have the engine going within a month.  In July the great beam was hung, the boiler 5½ft (presumably the diameter at the base) was set and the condenser finished.  The condenser, which was partly of lead with a wooden frame, had block tin air pumps, 9ins in diameter.  The upright timbers supporting the main beam were 14ins square.  He originally proposed mounting the main beam, which was about 10ft long, with unequal centres at 6ft and 4ft, with the longer length on the stream side.  The first pump would have been 18ins square and 15ft in depth but a later one had a diameter of 18½ins with a depth of 25ft suggesting that he intended to lift water out of the adjacent Gil Burn.

Illus 9: Watt’s Cottage looking north-west with the raised platform in the foreground.

The workshop had been deliberately built on the brink of the steep-sided valley to the west and this tells us a lot about the probable layout of the works.  The building was terraced into the north-facing hill slope meaning that the interior at the southern end was below the ground outside of it.  Here a platform extending the building can still be seen with three parallel N/S walls and it would seem that this was provided for the new engine.  A boiler located at the southern end of the room would have been low enough to feed steam through the wall into the cylinder and unsurprisingly this is just where twin chimney stacks were located.  

Illus 10: Watt’s Cottage c1925 looking north-west with the chimney stacks visible.

Placed here the beam of the engine could overhang the valley – simulating pumping water from a mine.  It should also be noted that the east doorway and the window to its south were later insertions and so this end of the room would neatly accommodate the boiler.

A copious supply of water was necessary for the condenser and for the proposed waterwheel.  As the sides to the narrow valley are steep and rocky this must have been conveyed from further south in either a wooden trough or a pipe.  A stone wall still crosses the valley and stream 44m to the SSE of the workshop.  It appears to be part of the formal landscaping of the grounds of Kinneil House and was probably intended to form a pond to be seen from the “Roman Bridge” on the “Duke’s Gallop.” 

Illus 11: The Wall crossing the Gil Burn looking south.

The waterfall created by it would have been visible from the house.  (There was probably a similar wall across the stream to the north of the footbridge north-west of the workshop which created a similar sequence there and which can be seen in Smile’s drawing.)

In many ways the fate of the progress of the Industrial Revolution hung in the balance for those months spent at Kinneil.  It was a very depressed Watt who finally wrote to Small on 20 September after what must have been a disastrous initial trial of his Kinneil engine and told him,

The trial has not been decisive; but I am still allowed to flatter myself with hopes…The [Clyde Survey] I would not have meddled with had I been certain of being able to bring the engine to bear; but I cannot, on an uncertainty, refuse every piece of business that offers… However, if I cannot make it answer soon, shall certainly undertake the next [Newcomen engine] that offers; for I cannot afford to trifle away my whole life, which God knows may not be long.”

Illus 12: The Drawing originally produced for the 1769 patent which must have been similar to the engine erected at Kinneil.  A – cylinder; B- condenser; C – air pump; D – lid or cylinder cover; E – stalk or piston rod; F – steam inlet pipe; G – regulating or steam valve; H – air pump piston; I – valve between cylinder and condenser; J – pipe between cylinder and condenser; K – valve at top of condenser; L – cylinder drain cock; M – steam piston.


A few invoices have survived from this period and provide the names of some of the people who supplied Watt with parts.  In May Daniel Biggs provided a steam valve, a double pump, a steam pipe, a frame for the condenser and a packing case for the steam cover.  In July John Wilson sent a variety of iron work including a “Modell of Fire Engine,” a piece of rolled plate iron, iron cut circular for valves, as well as “269½ pounds of iron work for Fire Engine.”  Stephen Maxwell submitted accounts from 1766 to 1771 for copper pipes and other copper work amongst which was a “white iron cilender” in January 1768 and the copper bottom for a cylinder in February 1769.  The most interesting item which he supplied on 10 May was a “sheet of Block Tin, £13.4s.7d”.  This sheet sounds like the one which may have been formed into the cylinder for the Kinneil engine.  It would have been heated and curved around a wooden cylinder and then the seam soldered and filed; work that may have been done in the workshop at Kinneil.

This cylinder was damaged in the course of its erection.  It was found to be slightly oval which Watt proposed to rectify where it stood.  It would be made true and hardened by “hammering it with a mallet on a piece of hard wood fitted to it” – a method that Watt had used in 1765.  It must have been the same cylinder which, in October, Watt told Small

The cylinder was a full quarter inch thick and 18 inch diameter.  At the worst place the long diameter exceeded the short about 3/8; it was made of the best tin, not hammered.  If the engine succeeds, I propose making them of copper.” 

With such inaccuracy, it is not surprising that Watt had difficulty obtaining a good seal for his piston.  He was overstretching the technology then available in Scotland to its limits.  Even after the block-tin had been repaired and appeared to be quite true Roebuck still thought that the material might be too soft and that the piston might distort it so it would have to be replaced with cast iron.  In this he was later proved to be right.

Watt admitted that fitting all the parts together and adjusting them had taken much longer than he had anticipated.  At the first trial, having pumped the air out of the condenser, the main piston descended only a couple of feet and then, on the second attempt, only a few inches.  The leather in the water pump had turned inside out and the pasteboard packing on the steam piston had torn.  After repairs, on the second trial, the engine made a few brisk strokes but soon went more and more slowly.  One fault may have been excessive friction in the air pump.  Then the steam piston failed and oil came into the condenser.  Cork packing was tried this time, then, it too failed.  There were many more tribulations but the condenser performed well enough although it was worked by a man because the waterwheel was not ready.  The boiler worked too, making sufficient steam with a small fire in spite of there being many leaks.  The conclusion he drew from this trial was that his engine

would work easily with 8 lbs on the inch, and would not consume above half the steam used by a common engine.  Even this I will not positively affirm, although I think there is reason to believe it.

The second engine at Kinneil was fitted with a wooden jacket around the cylinder which Watt said would protect the inner one made of block-tin.  There is no doubt that the jacket fulfilled Watt’s intention and kept the main cylinder hot.  However, Watt did not understand the true thermodynamic principles of the steam engine in which, to do work, the steam must both expand and lose heat.  Just as the cylinder should be as hot as the incoming steam, so the outgoing steam ought to have expanded and dropped in temperature, finally being as cold as the condenser in the most efficient engine.  Somehow, according to John Roebuck’s eldest son, Joseph Hately managed to gain access into the attic of Kinneil House to spy on Watt’s engine and workshop.  In fact there were no windows on this side of the house and so it may have been the flat roof of the Tower block that was used as an observation platform. 

Illus 13: Watt’s Cottage looking south-west.


Roebuck junior had an argument with him.  Unfortunately we are given no details but it is easy to imagine each other’s rage – one for the act of espionage and trespass, the other for having seen a jacket on the cylinder as which superficially looked suspiciously like that in his own patent.  It is possible that this happened in October 1769 when Hately visited Mr Addison’s brewery at Bo’ness.


Watt had to redesign his condenser unit which was being prepared that November.  Modifications were continually made until the beginning of April 1770 when Watt left to survey the line of the Strathmore Canal.  From there he sent Roebuck instructions about further improvements which were carried out by Gardner and Osborne.  Watt called at Kinneil on his way back from Perth in the middle of May and returned a month later.  He was in the course of more experiments when he was summoned to Glasgow to meet Golborne over surveys of the River Clyde and the Monkland Canal.  During the winter he prepared a survey for the Monkland Canal as well as improvements to Port Glasgow Harbour.

It was not until February 1770 that he could concentrate on the Kinneil engine again.  After this long hiatus he promised Roebuck that he would go to Kinneil and “not leave this place till the whole of the alternative engine has been effectually tried.”  Roebuck hoped that he could report better success to Boulton before the end of the month. The packing on the steam piston was still giving problems and numerous alternatives were sought and tried.  In March 1770 he reverted to cork which seemed to work.  Self-activating valve gear had been introduced but it is not clear if the waterwheel was ever put into commission.  13 June 1770 was a black day for the steam engine trials at Kinneil and Watt wrote in his journal “Made some experiments, inconclusive, on engine in the forenoon at 2 o’clock set out for Glasgow where I arrived about 10 o’ clock.”  The engine was never to run again at Kinneil.

Illus 14: James Watt in older age.


In June 1770 Roebuck was declared bankrupt and it was not until Boulton took over his share in the patent and Watt’s debts that any further progress was made.  In the middle of May 1773 Watt began to dismantle the engine at Kinneil for shipment to London and so on to Birmingham.  Roebuck recommended Watt to Lord Cochrane for the construction of a Newcomen engine at Culross.  By May 1774 it was performing economically – perhaps due to the cylinder supplied by Carron having been bored on Smeaton’s new boring machine.  Watt settled his affairs in Scotland and that May settled in Birmingham.  The engine was re-erected at the Soho Works.  Here Roebuck’s fears about the robustness of the metal were proved true and just before Christmas Day 1774 an inspection of the engine revealed that “The cylinder had been forced in by the steam.”  

Illus 15: Dr Roebuck’s Grave in Carriden Churchyard.

A new cylinder was ordered by January 1775 from John Wilkinson, the leading manufacturer of the time, but even it was not perfect.  However, it was still far better than had been achieved in ScotlandIt still took a couple of years to get it onto the market and made Watt a rich man.  Boulton too profited handsomely.  The patent was extended to June 1800 by an act of Parliament in 1775.  Roebuck struggled on, establishing a pottery at Bo’ness, and eventually died on 17 July 1794.  He was buried not at Bo’ness but in at Carriden Churchyard.  Between 1765 and 1770 he had been the person who had the greatest influence on Watt and the work at Kinneil helped to change the world for ever.

After Watt – Post 1775.

Illus 16: Watt’s Cottage c1920 looking south-west.


After the removal of Watt’s engine to Birmingham the workshop to the south of Kinneil House got a new lease of life as a washhouse.  Undoubtedly the water supply from the Gil Burn was one factor in this utilisation.  It would have been at this time that the brick floor was laid down and extended to a ramp approaching the north door.  Inside the washhouse the bricks are missing from the north-east quadrant and that this was an original feature is shown by the brick kerbing.

Illus 17: The Interior of Watt’s Cottage looking north-east.


This may have been where the wash boiler or “copper” as they were known locally, stood.  Much later this area was patched with concrete.


Other alterations probably belong to this period.  The door on the north side of the building and the window nearest to it are original and have neat slightly backset margins.  (By comparison the quoins are only roughly dressed.)  However, the door on the east side has a different dressing, the corners are slightly rounded and the stones are placed more vertically – all indicating that it is a later insertion.  The same applies to the southern window.  Iron pivots for window shutters remain.

Old photographs show that the roof had been piended and covered with pantiles.  By 1913 when one of these images was published in Salmon’s book the roof was beginning to collapse and in 1924 Bailie McKenzie complained of the ruinous state of the little old house to the rear of Kinneil House associated with the name of James Watt.  As custodians of that little house he felt that Bo’ness Town Council should do something to preserve it.  The Burgh Surveyor indicated that £10 would meet the expense and the Council agreed to the expenditure.  Typically of local government it was then decided to send in a claim to the Ministry of Works and to notify Mr Shinwell MP of the application (Linlithgow Gazette 16 May 1924).  A fortnight later it was suggested that the cylinder that had been in the Schoolyard Pit, and was then lying in the yard of the Bo’ness Gas Light Co at the Links, should be taken to the park at Kinneil as part of the improvements.  Nothing was done.

In December 1936 Lewis I Cadell offered to contribute £50 to the cost of restoring Watt’s Cottage and the Office of Works indicated that it was willing to carry out the repairs if the Town Council met the remaining cost.  The only result was that Bo’ness Town Council recommended that the Office of Works should take over the building and its maintenance.

At the end of the Second World War the Bo’ness Gas Company decided to make the 5ft 3ins diameter cylinder from the Newcomen engine at the Schoolyard Pit more accessible to the public.  The pit itself had literally stood adjacent to the schoolyard on west side of the School Brae near the town centre and the pupils used to complain about the steam coming in through the windows.  The pit was sunk by Dr Roebuck in the early 1760s and was 400ft deep to the Easter Main Coal.  To sink it from the Wester to the Easter Main Coal a bed of whinstone 110ft thick had to be pierced – a pit of such a depth and through so much hard rock absorbed an immense sum of money.  It was the deepest shaft in the field and drained all the workings to the rise.  As the mouth of the Schoolyard Pit was some 50ft above sea level the greater part of the water had to be raised nearly 400ft.  The large Newcomen engine was erected by Dr Roebuck and must have been made at Carron.  After the pit’s closure the cylinder lay there for some time before finally being bought as scrap metal by the Bo’ness Gas Company.  It was taken to the gas works at the Links and used as a tank.  Aware of its historic value the company offered it to the Watt Club in Edinburgh to ensure that it was not melted down as part of the war effort in the 1940s.  On it the Company had placed a plaque above the steam inlet pipe which read: “CYLINDER OF ENGINE/ ERECTED BY JAMES WATT/ AT SCHOOL-YARD PIT TO-/WARDS END OF 18TH CENTURY/ WHILE AT KINNEIL AND/ AFTERWARDS ACQUIRED/ BY BO’NESS GAS LIGHT CO.” 

Illus 19: The Schoolyard Pit cylinder at the gasworks in the Links, Bo’ness, preparatory to its move in 1946.

The Club had nowhere to put it and considered that it would be best placed at Kinneil.  It therefore contacted the Carron Company and secured an offer from it to share the cost of removal of the three ton casting.  Bo’ness Town Council then agreed to the use of the site at Kinneil next to the Cottage and made a concrete platform for its reception as well as removing some of the trees in the adjacent valley.  A new plaque was added below the other: “RE-ERECTED AT KINNEIL/ 3RD AUGUST 1946/ BO-NESS TOWN COUNCIL/ DAVID LUMSDEN PROVOST.”  The area in front of the Cottage received new trees.

For the occasion the Cottage was also given a facelift by the Ministry of Works.  The wallheads had collapsed and were made up to a flat top covered with rounded concrete to shed the rainwater.  The lintels of the doors and windows had to be replaced using material salvaged from the demolition of the outbuilding immediately to the north-east of the palace wing of Kinneil House.  The inner lintel for the east door has holes in it where the security bars for a window used to sit.  The outer lintel has the initials “W S” carefully cut into it suggesting that it had been a “marriage” or “date” stone.

Illus 20: Elevation and Plan of Watts Cottage showing the amount of rebuilding required in 1946.

Fifty years later, in August 1966, 88 year old Senator WW Roebuck visited Kinneil House from Canada.  

Although it is thirty-five years since his last visit he was able to point out the spot where Watt built a dam across the Gil Burn to give him a regular supply of water for his experiments.  He recalled the incident where his famous ancestor was watching Watt working, steam was escaping from a cylinder.  The Doctor, spying the steam leak, cut the tongue from his boot and wound it round the cylinder.  This proved to be successful and was the birth of steam packing…

(Linlithgow Gazette 5 August 1966).

It is doubtful if he would really have been aware of the dam and the story of the boot is a typical family legend, though leather was used for the water pump.


Illus 21: The Early 19th century Boiler End.


What seems to be the hemispherical end of an egg-end boiler c1820-30 can be seen in the grounds of Kinneil House.  It is 1.55m in diameter and provides a fine example of riveted plate technology.  It was possibly re-used as a cattle feeder/water trough and part of the cementaceous lining remains.


Hills, R.L.2002James Watt, Volume 1: His Time in Scotland, 1736-1774.
Muirhead, J.P.1854The Origin and Progress of the Mechanical Inventions of James Watt.
Salmon, T.J.1913Borrowstounness and District.
Watters, B.2010Carron: Where Iron Runs like Water.

G.B. Bailey (2021)