CALCULATION OF RESULTS. 81 



The quantity of air passing through the meter is so adjusted that ex- 

 actly 10 liters as measured on the dial pass through it for one analysis. 

 The air as measured in the meter is, however, under markedly different 

 conditions from the air inside the respiration chamber. While there is the 

 same temperature, there is a material difference in the water-vapor present, 

 and hence the moisture content as expressed in terms of tension of aqueous 

 vapor must be considered. This obviously tends to diminish the true 

 volume of air in the meter. 



Formerly we made accurate correction for the tension of aqueous vapor 

 based upon the barometer and the temperature of the meter at the end of 

 the period, but it has now been found that the reduction of the meter 

 readings to conditions inside of the chamber can be made with a sufficient 

 degree of accuracy by multiplying the volume of air passing through the 



meter by a fraction, A* , in which h represents the barometer and t the 



h 



tension of aqueous vapor at the temperature of the laboratory, 20 C. 

 Since the tension of aqueous vapor at the laboratory temperature is not far 

 from 15 mm., a simple calculation will show that there may be consider- 

 able variations in the value of h without affecting the fraction materially, 

 and we have accordingly assumed a value of h as normally 760 mm., and 



the correction thus obtained is -ri_ = 0.98, and all readings on the 



760 



meter should be multiplied by this fraction. 



On the one hand, then, there is the correction on the meter itself, which 

 correction is +1.4 per cent (see page 75) ; and on the other hand the cor- 

 rection on the sample for the tension of aqueous vapor, which is 2.0 per 

 cent, and consequently the resultant correction is -0.6 per cent. From 

 the conditions under which the experiments are made, however, it is rarely 

 possible to read the meter closer than 0.05 liter, as the graduations on 

 the meter correspond to 50 cubic centimeters. It will be seen, then, that 

 this final correction is really inside the limit of error of the instrument, 

 and consequently with this particular meter now in use no correction what- 

 ever is necessary for the reduction of the volume. The matter of tem- 

 perature corrections has been taken up in great detail in an earlier publi- 

 cation, and where there are noticeable differences in temperature between 

 the meter and the calorimeter chamber the calculation is very much more 

 complicated. 



For practical purposes, therefore, we may assume that the quantity of 

 air passed through the meter, as now in use, represents exactly 10 liters 

 measured under the conditions obtaining inside of the respiration chamber, 

 and in order to find the total amount of water-vapor present in the chamber 

 it is necessary only to multiply the weight of water found in the 10-liter 

 sample by one-tenth of the total volume of air containing water-vapor. 



