Given a Medieval or Middle Ages level of technology - leather bellows, etc how close could craftsmen come to achieving a near total vacuum in an enclosed space.

In context, I'm trying to create a Medieval society capable of producing chemicals through vacuum distillation, ceramic foams through vacuum kilns and the other benefits vacuums bring to high temperature applications.

If possible - any maths on how close they could get to hard vacuum - or even an approximation of the vacuum level within Magdeburg hemispheres - would also be appreciated.

  • 1
    $\begingroup$ The Magdeburg hemispheres experiment does not demonstrate any kind of hard vacuum. If there was 1 psi inside, the force required to separate the hemispheres would still be 13.7/14.7 of the force needed for a perfect vacuum. $\endgroup$ Nov 12, 2014 at 3:48

5 Answers 5


You can easily generate quite a high vacuum (0.1 bar or less) by driving the air out of a vessel with steam from a kettle, sealing the vessel, and then letting it cool.

No pumps necessary.

Hero of Alexandria experimented with steam but may not have discovered this effect.

Now consider what might happen if an able researcher in the Alchemical Arts, reproducing this ancient knowledge, tried to store Hero's magical stuff in a bottle...

Edit : Interesting! the Wiki page for "vacuum" (historical interpretation)" asserts that Hero tried to create a vacuum (but presumably not via this technique!), Islamic scientists were experimenting before AD 1000, and in the 1300s, answering the bellows suggestion, "ten horses could not pull open bellows when the port was sealed"...

Looking up "steam tables" I found the vapour pressure of water is 2.5% of atmospheric pressure at 20C, and 1.2% at 10C, so there's potentially 0.012 atmospheres in the bottle on a cool day...

  • $\begingroup$ For a demonstration of what condensation-made vacuum can achieve, see youtu.be/Zz95_VvTxZM or youtu.be/JsoE4F2Pb20 . $\endgroup$
    – Peteris
    Nov 12, 2014 at 20:18
  • $\begingroup$ The earliest steam engines that did "real work" fed steam into a cylinder at atmospheric pressure, closed the air outlet, and then exploited the partial-vacuum formed as the steam condensed. $\endgroup$
    – supercat
    Feb 21, 2015 at 20:51
  • $\begingroup$ Indeed, and the low pressure stage of any triple-expansion engine or turbine generating plant still works that way (2psi above atmospheric on the last one I saw). Unlike the older "condensing engines" this is economical because you get the atmospheric pressure steam "for free" from the higher pressure stages. $\endgroup$ Feb 21, 2015 at 23:02

It has been done in 1654 with Magdeburg hemispheres. They're essentially two halves of a metal ball, attached to each other. Then the air is pumped out. They used horses to show that the two halves couldn't be pulled apart afterwards.

You won't get a vacuum, but very low pressures, sure.

What you'll need is an air pump (bellows) and the two halves of metal.

I think it could have been done earlier than 1654, if people realized the ideas earlier. In 1643, Torricelli invented the barometer. This is what lead to the idea of atmospheric pressure (or so I think, you'll want to ask on History.SE to be sure, I'm just researching wikipedia here).

With an air valve, I imagine one could easily pump in gases that could react after sucking the normal air out.

  • 5
    $\begingroup$ History of Science and Mathematics might actually be even better than History for this question. $\endgroup$
    – David Z
    Nov 11, 2014 at 12:04
  • $\begingroup$ do you have anything earlier ? It's nnot bad but the middle age was a thing of the past in 1654 $\endgroup$
    – Vincent
    Nov 12, 2014 at 18:43

Depending on exactly what you're trying to do another approach comes to mind:

Fill your apparatus with mercury. The drain pipe extends down more than 1 meter. Chill the whole apparatus in ice. Open the drain valve, you'll have less than 1 pascal of mercury vapor in there. Close the drain valve, now you can warm it up to the desired temperature.

You could also construct a pump based on this: Open a valve between the apparatus and a chamber. Close the valve. Open a valve to the atmosphere, flood with mercury. Close the atmosphere valve, open the drain valve. Close the drain valve, repeat. Your pump cycle is very slow but you'll get a pretty good vacuum eventually. Note that this doesn't require the apparatus to be chilled, only the mercury side of things needs chilling.

I rather suspect this will produce a better vacuum than they can actually maintain anyway.


I remember to have read that the Magdeburg hemispheres were heated and then pressed together. The pressure inside went down while the hemispheres cooled down.

From my chemistry practical works, I remember that a vacuum of 20% (remaining pressure is 0,2 bar) can be obtained by the Bernouilli effect: we opened a small tap, the water flowing through the hose had a lower pressure. (the container was connected by a T-shape piece).

  • $\begingroup$ According to Wikipedia, the point of the Magdeburg hemispheres was to demonstrate the piston-based vacuum pump that Otto von Guericke had invented. The vacuum was from pumping, not heating and cooling. $\endgroup$ Nov 11, 2014 at 23:54

"... but nobody withstands The Machine."

We need the pump before we can play with the spheres and for that we need to be just this side of Enlightenment and know how to machine. I don't think a bellows will work (at all?) on an industrial scale, even his toy spheres needed a legitimate vacuum pump.

Machine tool, History -Wiki

Forerunners of machine tools included bow drills and potter's wheels, which had existed in ancient Egypt prior to 2500 BC, and lathes, known to have existed in multiple regions of Europe since at least 1000 to 500 BC.[6] But it was not until the later Middle Ages and the Age of Enlightenment that the modern concept of a machine tool—a class of machines used as tools in the making of metal parts, and incorporating machine-guided toolpath—began to evolve.

Clockmakers of the Middle Ages and renaissance men such as Leonardo da Vinci helped expand humans' technological milieu toward the preconditions for industrial machine tools. During the 18th and 19th centuries, and even in many cases in the 20th, the builders of machine tools tended to be the same people who would then use them to produce the end products (manufactured goods). However, from these roots also evolved an industry of machine tool builders as we define them today, meaning people who specialize in building machine tools for sale to others.

Historians of machine tools often focus on a handful of major industries that most spurred machine tool development. In order of historical emergence, they have been firearms (small arms and artillery); clocks; textile machinery; steam engines (stationary, marine, rail, and otherwise)...

So, we need to know how to machine and have a reason why we do. You have vacuum distilleries?... NP.

enter image description here

He probably owes you money, eh?... Well I'll ask 'em...

Failing this criteria, then I'd guess we're looking for the maximum bar that a bellows can pull and whether or not that's acceptable for the process in question.

I'm having a hard time finding data, or use of bellows ->backwards. I'm guessing no one came forward and said "That's great Mr. Magdeburg Otto, but all you've done is make a metal contraption [vacuum pump] just like my (imaginary) leather bellows thingy".

  • 1
    $\begingroup$ "Mr Magdeburg" was actually Mr von Guericke (Otto, to his friends). $\endgroup$ Nov 12, 2014 at 17:18

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .