9O A RESPIRATION CALORIMETER. 



all practical purposes. It is seen that this method determines oxygen 

 by difference, while usually the factor in air analyses that is determined 

 by difference is the nitrogen. 



The calculation may be expressed algebraically in the following way : 



z> = volume of air sample. 



Vj = volume of air containing water = (I + II). 



V 2 = volume of air containing CO 2 = (I + II + III) . 



V 3 volume of air containing O + N = (I + II + III + IV). 



a = total volume of water vapor. 



b = total volume of carbon dioxide. 



c = total volume of nitrogen. 



d total volume of oxygen. 



W= total weight of water vapor in system; w weight in air 



sample. 



W'= total weight of CO 2 in system ; w l = weight in air sample. 

 W = 'XV 1 Tpi^XV. 



^0 



i. 2434 a> XV! , .5091 w 1 XV, 



- 



It was formerly assumed that at the beginning of the experiment 

 ^=.2091 (V g a ); ^=.7909 (V 3 a ). 



These values for the amounts of oxygen and nitrogen were deter- 

 mined by assuming the composition of the air free from carbon dioxide 

 and water vapor as 20.91 per cent oxygen and 79.09 per cent nitrogen. 

 We now secure greater accuracy, however, by using the actual analysis 

 of the carbon-dioxide and water free air as made at the beginning of an 

 experiment, i. e., at 7 a. m. 



This consequently changes the factors used in the last two equations 

 from o. 209 1 and o. 7909 to those found by analysis. In calculating the 

 composition of the air at the end of the first experimental period, c is 

 determined from the record of the amount of nitrogen entering with 

 the oxygen, lost or gained through interchange through the food aper- 

 ture, rejected with the absorbers, and lost if a sample of air has been 

 rejected. All of these corrections are applied to the original initial 

 volume of nitrogen found by analysis. 



Under these conditions, then, we have 



<t=V s (.a+b + c). 



