Except that this longstanding experiment, used by secondary school teachers around the world for decades, is a sham - as three schoolchildren have demonstrated.
Emma, Andrew and Rebecca Fist, aged 9, 11 and 12, of Norwood in Tasmania, decided to burn the oxygen up faster, and did the same experiment with three candles. To their surprise, they found that the water rose further up the jar - even allowing for the extra volume taken up by the candles' bases. "The water level rises much more. Why?" they wrote to New Scientist magazine.
Now, the truth has emerged: the experiment does not really show the effects of the air's oxygen depletion, but how much it shrinks when the air cools after the candle, or candles, go out.
Answering the children's question, Ian Russell, of Interactive Science Ltd, in High Peak Derbyshire, who designs hands-on science experiments, commented: "Congratulations to the children, who experimentally disproved the common textbook misconception."
He said that the principal effect making the water rise is that the candle flames heat the air around them before the jar is placed over them. "There is a short pause after the candles go out, and only then does the water level rise as the remaining air cools and contracts again." As other scientists pointed out, when the candle's hydrocarbons burn, they produce carbon dioxide - a gas that dissolves a little in water - and water vapour. The volume of oxygen lost does not match the change in water level.
Yesterday, Mr Russell said that errors like these are common in school science texts. "I call them textbook viruses: they get propagated from generation to generation."