A candle, a mouse, and a jar
How We Found Out What Air Is
Air was thought to be a single simple thing. A plant on a windowsill proved otherwise.
-
Step 1 of 5
Air was assumed to be one simple substance.
The fact was known for centuries. The explanation was wrong just as long. For most of history air was an element: one thing, everywhere, unmixed. The obvious problem was well known and unexplained — a candle in a sealed jar goes out long before the jar is anywhere near empty.
-
Step 2 of 5
A mouse in the same jar does the same thing.
Two very different processes, using up exactly the same thing. Joseph Priestley put a mouse under a jar and it collapsed. He put a burning candle in and it went out. Crucially, air that had already killed a mouse would not support a flame either, and air that had put out a flame would not keep a mouse alive.
-
Step 3 of 5
Then he put a plant in.
A sprig of mint, sealed in used-up air and left in the light for several days, restored it. A candle would burn in it again, and a mouse could live in it. Something was being consumed by burning and breathing, and made again by the plant.
-
Step 4 of 5
So air is a mixture, and only one part of it matters here.
Priestley isolated that part by heating a red powder and collecting the gas. A candle in it burned fiercely bright. A mouse in it stayed lively for far longer than the same volume of ordinary air would allow.
-
Step 5 of 5
It took a third person to say what it meant.
Priestley had the gas but explained it with the old theory, believing that burning released something rather than consumed it. Antoine Lavoisier weighed everything before and after — and metals got heavier when burnt, which only works if burning takes something from the air.
Weighing the results is what settled it. The measurement was the argument, not the experiment.
The short version
Air turned out to be a mixture: burning and breathing use up the same one fifth of it, and plants in light put it back.
Try it yourself
Stand a candle in a dish of water and cover it with a jar. It goes out, and the water rises a little way up the jar — that gap is the part that was used.