CALCULATION OF RESULTS. 83 



INFLUENCE OF FLUCTUATIONS IN TEMPERATURE AND PRESSURE ON THE 

 APPARENT VOLUME OF ATR IN THE SYSTEM. 



The air, being confined in a space with semi-rigid walls, is subjected 

 naturally to variations in true volume, depending upon the temperature 

 and barometric pressure. If the air inside of the chamber becomes con- 

 siderably warmer there is naturally an expansion, and were it not for the 

 tension-equalizer there would be pressure in the system. Also, if the ba- 

 rometer falls, there is an expansion of air which, again, in the absence of 

 the tension-equalizer, would produce pressure in the system. It is neces- 

 sary, therefore, in calculating the true volume of air, to take into account 

 not only the apparent volume, which, as is shown above, is always a con- 

 stant amount at the end of each period, but the changes in temperature and 

 barometric pressure must also be noted. Since there is a volume of about 

 1,400 liters, a simple calculation will show that for each degree centigrade 

 change in temperature there will be a change in volume of approximately 

 4.8 liters. In actual practice, however, this rarely occurs, as the tempera- 

 ture control is usually inside of 0.1° C. and for the most part within a few 

 hundredths. A variation in barometric pressure of 1 millimeter will affect 

 1,400 liters by 1.8 liters. 



In actual practice, therefore, it is seen that if the barometer falls there 

 will be an expansion of air in the system. This will tend to increase the 

 volume by raising the rubber diaphragm on the tension-equalizer, the 

 ultimate result of which is that at the final filling with oxygen at the end 

 of the period less is used than would be the case had there been no change 

 in the barometer. In other words, for each liter expansion of air inside 

 of the system, there is 1 liter less oxygen required to bring the apparent 

 volume the same at the end of the period. Similarly, if there is an increase 

 in temperature of the air, there is expansion, and a smaller amount of 

 oxygen is required than would be the case had there been no change; and 

 conversely, if the barometer rises or the temperature falls, more oxygen 

 would be supplied than is needed for consumption. It is thus seen that the 

 temperature and barometer changes affect the quantity of oxygen admitted 

 to the chamber. 



INFLUENCE OF FLUCTUATIONS IN THE AMOUNTS OF CARBON DIOXIDE AND 

 WATER-VAPOR UPON RESIDUAL OXYGEN. 



Any variations in the residual amount of carbon dioxide or water-vapor 

 likewise affect the oxygen. Thus, if there is an increase of 1 gram in the 

 amount of residual carbon dioxide, this corresponds to 0.51 liter, and con- 

 sequently an equal volume of oxygen is not admitted to the chamber during 

 the period, since its place has been taken by the increased volume of carbon 

 dioxide. A similar reasoning will show that increase in the water-vapor 

 content will have a similar effect, for each gram of water-vapor corresponds 



