14 
The table shows the duration of the tests, the amount of ethyl 
alcohol burned in the lamp, and the theoretical amount of carbon 
dioxid, water, and heat which would have been produced by the com- 
bustion of the alcohol as compared with the amounts actually found. 
As previously noted, the agreement between theoretical values and 
those actually obtained is, as a rule, very close. 
SAMPLING OF THE AIR CURRENT FOR ANALYSIS. 
Two methods have been employed for sampling the outgoing current 
of air for analysis. After this current of air has left the cooling appa- 
ratus where the major portion of the water has been condensed, small, 
samples of about 150 liters each during every six-hour period are drawn 
by the aspirators. The details of this method and the formulas used 
for calculating and correcting the results are given in a previous 
publication.* 
The greatest objections to the use of the aspirators as a method of 
sampling the air are: (1) The number of corrections for temperature, 
barometric pressure, and tension of aqueous vapor which have to be 
made in finding the actual volume of air drawn into the aspirators; 
(2) the variations in the speed of the motor driving the meter pump 
and, consequently, in the volume of the ventilating air current, these 
not being accompanied by corresponding variations in withdrawal of 
the sample, and (3) the irregularity with which the sample is drawn. 
By means of an automatic device for regulating the speed at which 
the water runs out from the aspirators it has been possible of late 
to make this rate very uniform. At the same time relatively small 
fluctuations might introduce errors. Thus, if the aspirators emptied 
more slowly than the normal at a period when the man was giving off 
more than the average amount of carbon dioxid, as, for instance, when 
he was weighing the " absorbers" or doing other special work, the 
measured amount of carbon dioxid. as determined by the aspirators, 
would be smaller than the amount actually eliminated, and vice versa. 
The meter-pump method, briefly described in a previous bulletin, 6 
affords, theoretically, an almost perfect means of sampling the venti- 
lating air current. The arrangements for carrying out this method are 
as follows: The air from each fiftieth stroke of the pump is collected 
alternately in two tin receptacles, from which it is withdrawn by suc- 
tion and passed through tubes filled with pumice stone and sulphuric 
acid, and with soda lime for the absorption of the water and carbon 
dioxid, respectively. In the earlier experiments in which the meter 
pump was used for moving the ventilating air current this method of 
sampling had not been perfected. In test experiments Nos. 9-12 the 
a U. S. Dept. Agr., Office of Experiment Stations Bui. 63, pp. 34, 35. 
b U. S. Dept. Agr., Office of Experiment Stations Bui. 63, p. 31. 
