THE RESPIRATION APPARATUS. 47 



The vaporization of water in the meter results in a slight lowering of 

 the temperature of the water in the meter as the air passes through it. 

 The thermometer is read in the middle of each residual sample, i. e., 

 when 5 liters of air have passed through the meter. This is assumed 

 to give the average temperature of the 10 liters. 



Calibration of Elster meter. The Elster meter was calibrated by pass- 

 ing a known weight of oxygen from a cylinder through the meter and 

 comparing the volume as calculated from this weight with the volume 

 recorded by the meter after making due corrections in the latter for 

 temperature, tension of aqueous vapor, barometer, etc. In order to 

 have the conditions under which the gas is measured correspond as 

 nearly as possible to those under which residual analyses are taken, 

 provision was made for maintaining during the test a diminished pres- 

 sure in the meter amounting to 198 mm. of water. This diminished 

 tension is that ordinarily experienced when conducting a residual 

 analysis, and is a measure of the resistance of the tubing, U tubes, and 

 meter. 



In conducting the test the apparatus was connected ready for use, 

 the oxygen cylinder weighed, and the barometer reading taken. At 

 the end of every 10 liters the temperature of the meter was recorded. 

 The barometer was read periodically, but it was found that the fluc- 

 tuations were very slight, and, in general, the average of the readings 

 at the beginning and end could be used. After about 100 liters of air 

 (apparent volume as measured by the meter) had passed through the 

 meter the cylinder was weighed and the calculation of volume from 

 weight was made. 



From a number of such tests it was found that to obtain the true 

 reading from the meter under the conditions of the experiment the 

 logarithmic factor .98895* should be added to the logarithm of the 

 apparent volume. 



TEST FOR SATURATION OF AIR PASSING THROUGH THE EI^TER METER. 



In drawing through the Elster meter the air used for the residual 

 analyses, it is assumed that, coming in contact, as it does, with a large 

 volume of water in its passage through the meter, it becomes saturated 

 with water vapor at the temperature of the meter. It was necessary to 

 verify this assumption, especially as in the calculations the volume of 

 air passing through the meter is multiplied by a large factor and so 

 must be known with great accuracy. To test this point, 20. 103 liters of 

 air, as measured by the meter, were forced through the meter at a tem- 



1 For convenience in calculations the characteristics of the logarithms of the 

 factors are neglected. 



