The First Artificial Cold

Wrap two thermometer bulbs in cotton. Wet one with water, leave the other dry, and fan both equally.

Log both every 30 seconds for five minutes.

Expect the wet one to settle several degrees below room temperature and stay there while it is still wet.

Nothing cold touched it. The drop is entirely the cost of turning liquid water into vapour, paid out of the thermometer's own heat.

Repeat step 1 with surgical spirit instead of water.

Expect a faster and deeper drop.

Alcohol evaporates more readily at room temperature than water does — it has a higher vapour pressure — so it takes its latent heat quicker.

Record both curves on the same axes. You have just done the experiment that every refrigerant selection since 1834 is a more careful version of: the fluid is not incidental, it is the design decision.

Put warm water in a syringe, seal the tip, and pull the plunger back hard.

Bubbles appear. The water is boiling — at a temperature you could put your finger in.

This is Cullen's move, and it is worth stating exactly what it overturns: boiling is not about being hot. It is about the liquid's vapour pressure matching the pressure above it. Take the pressure away and the boiling point comes down to meet the liquid.

Now put the two halves together. Wrap a small water-filled vessel in wet cloth, place it in a jar you can evacuate, and pump the pressure down while logging its temperature.

Expect a much steeper fall than fanning ever gave you — the low pressure makes the water evaporate far faster, and faster evaporation means heat leaving faster.

With a laboratory pump and ether, this reaches freezing. That is precisely Cullen's 1755 apparatus, and the ice he made formed on the outside of the vessel, from room air.

Keep pumping and watch what happens over ten minutes.

The cooling slows and stops. Your liquid is running out, and the vapour you have pumped away is gone.

That is the gap between a demonstration and a machine. Cullen's apparatus cools until the ether is spent, and then you must refill it.

To cool continuously you must catch that vapour and turn it back into liquid — which needs a compressor to raise its pressure and a condenser to take the heat away. Eighty years later, that is exactly what Perkins added, and nothing else.

The demonstration. William Cullen, professor at Glasgow, showed artificial refrigeration in 1755 by evaporating ether under reduced pressure and forming a small amount of ice. It is generally taken as the first documented instance of cold produced deliberately rather than collected. He published; he did not patent, did not build a working machine, and did not pursue it commercially.

Why nothing happened for eighty years. Cullen had the physics complete and the economics against him. Natural ice was abundant and, once Wyeth's ice plough arrived in 1825, extremely cheap. There was no market to disturb. The idea sat in the literature until city populations, brewing and meat shipping created a demand that lakes could not meet — at which point Perkins' cycle and Linde's ammonia machine had something to be better than.

Cullen was a physician, and it shows. He came to the problem through medicine and chemistry rather than engineering, which is probably why he framed it as a phenomenon to be understood rather than a device to be sold. He is better remembered in medicine than in refrigeration, and his students included Joseph Black, whose work on latent heat a few years later gave the effect its proper name and measurement.

The same physics, everywhere. Sweating is Cullen's experiment run by your own body. The unglazed zeer pot and the Spanish botijo keep water cool in hot dry air by exactly this route, with no moving parts at all. A modern vacuum freeze-dryer is Cullen's apparatus with better pumps. And every refrigerant in this batch is a search for a liquid whose boiling behaviour is convenient — which is the question Cullen's experiment first made askable.

What it does not do. Evaporative cooling fails in humid air, because the air is already carrying as much vapour as it can and the liquid will not leave. That is why swamp coolers work in Arizona and not in Singapore, and why the whole approach eventually had to be sealed into a closed loop where the machine, not the weather, decides the conditions.