One of the oldest exhibits at Summerlee Museum isn’t actually part of our collection. The Farme Colliery engine belongs to Glasgow Museums but has been at Summerlee since 1987 where it is a star attraction in the coal mine exhibit. Now partially-rebuilt in a replica engine house, the engine is one of the rarest steam engines in the world.
The winding engine which we illustrate is in many respects the most remarkable old steam engine in existence.
‘The Engineer’ magazine, 1879
The Farme Colliery Engine, built in 1809 or 1810 is a so-called Newcomen, or Atmospheric Engine of which only 8 or 9 survive in varying degrees of completeness.
What makes this particular engine very unusual, though is that it is a rotative Newcomen engine. What does that mean? Well, quite simply the engine turns a wheel rather than just moving a pump rod up and down as most Newcomen engines did. The Farme Colliery Engine is one of only three rotative Newcomen engines in the world and the only one that is still in Europe.
What’s a Newcomen Engine?
Named after its inventor, the engineer Thomas Newcomen this type of machine had the distinction of being the first workable type of steam engine. We often think tend to think that James Watt invented the steam engine but Newcomen’s device goes back half a century before Watt’s celebrated invention of the separate condenser. Believe it or not, Newcomen’s first successful engine was built in 1712. This was during the reign of Queen Anne and only five years after the Act of Union.
Unlike later steam engines, Newcomen’s didn’t use the expansive power of steam to drive its piston. Instead, Newcomen filled his cylinder with steam and then used a jet of water to condense it, creating a partial vacuum.
The piston was then driven down into the open-topped cylinder by the pressure of the surrounding atmosphere. The rim of the piston had hemp rope wound round it and water was dribbled onto the top of the piston to keep the rope wet and maintain a seal. You can see rope lying next to the top of the cylinder in the pictures above.
Newcomen engines were very inefficient but that didn’t matter at a colliery where the engine could be fed the plentiful but unsaleable ‘dross’ coal that was brought up.
Situated on the Farme estate on the Cunningair Loop to the east of Rutherglen, Farme Colliery was established in the early years of the 1800s. Our engine was the colliery’s first and is said to have been built in 1810. It was a winding engine and is also said to have been used to pump water from the mine (we will come back to this).
Although an engine house was built around the 1810 engine it was actually free-standing, supported by a four-legged iron frame. This would have enabled the engine to be moved if needed. After all, coal mining was a highly speculative activity in the early 1800s, with no guarantee of success. At the same time, this being before the era of mass-production, steam engines were expensive pieces of equipment. An engine might have to be moved several times during its life.
The Engines at Farme
We are lucky that even in the late 1800s the Farme Colliery Engine was considered to be an important relic. This, no doubt helped it survive.
In fact there were no less than three Newcomen engines at the colliery earlier in the century. All were still working away in 1879 when an article about them appeared in Engineering magazine. However, this article contains some confusion which we are still trying to untangle.
The first engine was installed in 1809 or 1810, depending on who you believe. As we have seen, this particular engine was free-standing, presumably in case the new mine wasn’t successful.
In 1820 a second Newcomen winding engine was installed. This was a much bigger engine with a 42 inch diameter cylinder. Instead of being free-standing the engine was built into a substantial engine house. Later that year another Newcomen engine was built, this time a dedicated pumping engine with a whopping 60 inch cylinder.
The article in the Engineer featured engravings of the 1820 winding engine and the 1821 pumping engine but not the 1810 engine, although it did have photos of the earlier engine.
Mystery Number 1: who made the engine?
A 1903 paper by Henry Davey repeated much of the information from the 1879 article in the Engineer. Davy tells us that the 1810 engine was built by a John McIntyre, who made his own patterns for the engine’s components. These were, says Davey, cast at Camlachie Foundry in the east end of Glasgow.
Here things get a little confusing because we know that Camlachie Foundry didn’t exist in 1810, being built in 1814. However, there was a foundry in the Camlachie area of Glasgow at the time, called Greenbank Foundry. Perhaps the foundry Davey referred to was Greenbank Foundry or maybe he was confused and it was actually the 1820 winding engine that was cast at the Camlachie Foundry. If we can find out more about John McIntyre that may help to answer the question.
The confusion deepens when we look at the drawings and photographs in both articles. A photograph of the engine now at Summerlee is shown next to an engraving of what must surely be the 1820 winding engine, both presented as being the 1810 engine!
Mystery Number 2: was it a pumping engine?
The 1810 Farme Colliery engine was a rotative engine, that is it used a crank to convert the up-and-down motion of its rocking beam to create circular motion. The circular motion turned a winding-drum, or more accurately, two winding-drums on a single axle.
Pumping engines were usually larger and had at the end of the beam a pump rod that descended deep into the mine to pump water out. This was what Thomas Newcomen’s original engines did and an example from Kilmarnock is preserved in the National Museum of Scotland.
The 1879 article in The Engineer tells us of the 1810 engine:
It… until very recently pumped water from a depth of twenty –three fathoms with an 8in. bucket pump. It drew from 150 to 200 tons per day, and pumped for four or five hours in the twenty-four on a consumption of 34 cwt.
Contemporary photographs do show two vertical rods leading from the beam of the 1810 engine and these likely powered pumps. However, these pump rods are about half-way along the beam from the pivot and are quite thin so it seems unlikely that they pumped large volumes of water. One of the pumps would have been to pump water from the ‘hot well’ where condensed water from the cylinder was collected, back up to the tank that held the condensing water. The other pump may have been a boiler feed pump, or maybe it did descend into the mine?
Another possibility is that there was a pump connected to the other end of the winding drum axle. Since the winding drum wasn’t preserved or recorded and none of the photographs give us a clear view of that side of the winding drum we can’t be sure.
However the pump was placed there must have some kind of clutch that allowed the engineman to disconnect the power to the winding drum while the engine was on pumping duty.
The final possibility is that this is another case of the confusion between the 1810 and 1820 winding engines. After all, the engine was described in 1879 as having pumped water “until recently” so the author didn’t see it actually working as a pump.
Mystery Number 3: Why was the engine manually-operated?
The valve gear is of the simplest possible kind, and is worked by hand, just as was the case with the original Newcomen engine before Humphrey Potter invented the scoggan or Beighton the plug frame. We cannot call to mind any other instance in which an engine is employed either for winding or any other purpose, the valve gear of which does not operate automatically.
‘The Engineer’ magazine, 1879
The very earliest workable engines built by Thomas Newcomen had automatic valve gear that was controlled by a rod connected to the engine’s rocking beam. So why did a late example like the Farme Colliery Engine work purely manually?
We shouldn’t underestimate the amount of work involved in driving the engine by hand. Although there were only two controls, a handle that controlled the valves and another for the throttle, the valve lever needed to be moved for every stroke of the engine. This wouldn’t be so bad for winding as the engine would only run for a few minutes each time, giving the driver a break. However, pumping would take place for hours at a time. Perhaps then this is a clue that the 1810 engine wasn’t a pumping engine after all.
As for the question of why the engine worked manually, well the photographs show that the plug rod (which transmitted movement from the rocking beam to move the valve gear) was split. The reason for a plug rod to have a split section was to engage with the valve gear for automatic operation. This suggests that the engine was once automatic, unless the plug rod came from another engine.
Why the engine was turned to manual operation we can only guess. A single-acting engine like this wasn’t ideal for creating rotative motion so maybe it couldn’t be made to work reliably.
In 1913 the 1810 engine, by then Scotland’s last working Newcomen engine was replaced by a more modern winding engine. This was because the mine was being dug deeper and the old engine simply wasn’t up to the job any more. The old engine was given to the City of Glasgow Corporation for preservation but first every component was carefully measured and drawn.
Only two other rotative Newcomen engines, both English, survive. They were collected by the enormously wealthy American motor car manufacturer Henry Ford, who visited England looking for exhibits for his museum in 1928. Ford’s agent heard about the Farme Colliery Engine and looked into whether he could buy it. However, the idea was abandoned on learning that Glasgow Corporation had beaten him to it! You can read David Perrett’s account of Henry Ford’s extraordinary engine-buying expedition in The International Journal for the History of Technology Vol.88, No.1 (2018).
The engine wasn’t re-built in Glasgow but stayed in storage, probably because the Victorian industrial museum at Kelvingrove House no longer existed. Unfortunately, many of the smaller components (including that puzzling valve control gear) were lost during moves between buildings. Then in the 1980s the Summerlee Heritage Trust who were building a new industrial museum in Coatbridge approached Glasgow Museums. In 1987 the Farme Colliery Engine came to Summerlee Museum and the work of reassembling it began. In 1999 the engineers at Summerlee won an award for their reconstruction of the winding drum (which hadn’t been preserved or measured) based purely on photographs.
I am hugely grateful to Iain Drysdale who was the first person in recent times to research the Farme Colliery Engine, writing an unpublished paper that raised the questions discussed here in 2000. Iain has generously shared his research with me and I hope that in time we will reach some firmer conclusions over the questions I raise here.
If you are interested in the early history of steam power I recommend you visit the website of the International Early Engines Conference which includes lots of fascinating information, including a database of engines built before 1800.
About the Author
Justin Parkes has been Industrial History Curator for CultureNL since 2008.