44 A RESPIRATION CALORIMETER. 



ering of temperature. On one or two occasions, when the subject has 

 slept, contrary to the advice of the experimenters, with his head near 

 the pipe conducting the air into the chamber, slight disturbances of 

 the respiratory tract have been experienced, but when sleeping with 

 the head at the other end of the chamber no such disturbances have 

 occurred. 



APPARATUS FOR THE ANALYSIS OF THE RESIDUAL AIR. 



For purposes of calculation it is necessary to know the carbon-dioxide 

 and water content of the closed volume of air at the end of each experi- 

 mental period. While any one of the numerous methods depending on 

 the use of a solution of barium hydroxide and phenolphthalein might 

 be used for the determination of carbon dioxide, and consequently only 

 a very small volume of gas required, none of the methods of hygrometry 

 as ordinarily employed will give the water content of the air with suffi- 

 cient accuracy. It is therefore necessary to determine the water by 

 the absolute method, that is, by aspirating a large quantity of air 

 through some water absorbent and actually weighing the water vapor 

 thus removed. The arrangements for this operation are illustrated in 

 figure 1 8. 



A lo-liter sample is withdrawn from the air- pipe between the respi- 

 ration chamber and the blower through the mercury valve described on 

 page 1 8. The water vapor and carbon dioxide are removed by con- 

 ducting the sample through sulphuric acid and soda lime, respectively, 

 the volume of air withdrawn being accurately measured by a gas-meter. 

 After leaving the meter the air enters the water-pump, where the suc- 

 tion required to draw the air through the tubes and the meter is obtained. 

 A device for separating the air and water leaving this pump makes it 

 possible to return the air sample to the ventilating air current between 

 the pans and the respiration chamber. 



Theoretically the sample of air should be drawn at exactly the end 

 of the different experimental periods. In such case, however, it would 

 necessitate the complete withdrawal of 10 liters of air from the system 

 and introduce serious complications in the calculations. By means of 

 the system of U tubes described beyond, it is found that air may be drawn 

 through these tubes at the rate of 2 or even 3 liters in one minute and 

 all the carbon dioxide and water be quantitatively absorbed. At this 

 rate the air can be drawn with sufficient rapidity to furnish results that 

 agree with those from a to-liter sample drawn at one instant. By 

 returning the air to the system no loss occurs. 



The requirements for absorbents for water vapor and carbon dioxide 

 that will effect the complete removal of these substances from an air 



