Method for Forcing the Condensation of Water Vapor in the Candle Experiment
Svieta, Daawar, Matt and David

A common elementary school experiment determines the oxygen content of air by removing oxygen through burning a candle and measuring the change in volume of the air. One version of the experiment is documented at One critique (found at makes the claim that the production of carbon dioxide rather than oxygen depletion smothers the flame. We experimentally tested this claim.

We looked at the chemical reaction for the complete oxidation of paraffin:

C25H52 + 38 02 --> 25 CO2 + 26 H2O
It shows that there is a net increase of gas produced in the reaction, which would lead us to expect that the water level would go down, not up. We speculated about the origin of this discrepancy.

We knew that the nitrogen and other gases in air would not be reacting. We wondered if some of the water vapor were being removed from the bubble of air under the glass by condensation on to the water surface or onto the glass walls. If enough water condensed, so there could be a net loss of gas molecules in the bubble (since 38 molecules of O2 are consumed for every 25 molecules of CO­2 that are produced). That would explain the rising water level.

We decided to force condensation of water vapor by condensing it through cooling. We could then see if the water level was where is should be if the 38/25 change had occurred.

Purpose: To test whether removing water vapor through forced condensation produces a water level increase consistent with the predicted consumption of oxygen and production of carbon dioxide.

In the original experiment, whether or not the produced water vapor is remaining as a gas and leaving some other unknown factor to cause the raise in water level or it is condensing in a significant enough way, won’t impact the raise in water level if all the water vapor is condensed.

What this test will be able to do is see if the 38 molecules of oxygen do result in 25 molecules of carbon dioxide based on the measured change in water level and the isolation of carbon dioxide through cooling. This also only works if the experiment before and after cooling are at the same temperature as the same number of particles at the same temperature take up the same amount of space. If so and there is a 38-25 ratio, then it shows that there is a high chance that the water vapor is the factor in causing the unexpected rise in water level in the original experiment. If not, then it shows that there is something else going on that remains undetermined.

To begin our experiment we first gathered together the necessary materials. First we brought out three 9’ Pyrex Pie Plates. The plates were used to hold both the candle and the water. In order to get the candle to stick to each plate, we lit a match and melted the candle, and then we stuck it in the small puddle of hot wax that gathered on its plate. We let it sit for about a minute until the wax solidified. Then, we added 100 ml of water to each plate. The water surrounded each candle, and each water level rose about halfway up the length of each candle. We then lit each candle and then covered each candle with a graduated cylinder. Then, we measured the immediate water levels for each plate. To measure these levels we looked at the scale of each graduated cylinder and thus calculated the level of milliliters. After gathering our measurements, we took all three plates and placed them in a refrigerator to cool overnight. After we let them cool over night, we took them out and let them warm to the initial temperature. Once they were at the initial temperature, we then took the measurements of the water again. We performed the same method as before when taking the new measurements. Among our major findings, we observed that the candles went out and the water levels rose.