Good thinking! That experiment always irritated me. It gives very inconsistent results and makes no sense anyway. The volume of CO2 produced is notionally the same as the oxygen consumed! Mind you, the burning material also produces water that condenses, also reducing the volume on cooling.

The reason that the air shrinks is that the flame had heated it. When it cools it shrinks. Some CO2 might also dissolve, reducing the volume further.

You will note that the air doesn't begin shrinking until the flame goes out!

I think carbon dioxide should dissolve in water much better than oxygen, as it consists of two different kinds of atoms so that the charge is distributed less evenly (though not as unevenly as in water where the hydrogen atoms are at an angle, not just at each end of the O atom). Carbon dioxide forms carbic acid with water.

Hello,

the candle is paraffin wax, the "formula" is (neglecting end groups) CH² .

So, total combustion would result in one molecule of carbon dioxide

per  one  molecule of water. Given enough time, the cabon dioxide

would dissolve in the water entirely.

But

problem is, that the combustion will not be total!

When oxygen becomes rare, some carbon monoxide will be

formed, along with a lot of cracked products from the wax.

The carbon monoxide will not easily dissolve in water.

This is reason for the "inconsistent" results.

For that reasons this experiment is better done with

iron wool or white phospor.

Georg