The rate of flow of water through the apparatus is determined before the experiment begins. This is done by deflecting the water for a certain number of seconds into a graduate or by deflecting it into the small can and weighing the water thus collected. The water is then directed into the drain during the preliminary period. Meanwhile the main valve at the bottom of the water-meter is opened, such water as has accumulated from tests in preceding experiments is allowed to run out, and the valve is closed after the can is empty. The meter is then carefully balanced on the scales and the weight is recorded. At the beginning of the experiment the water is deflected from the drain into the meter. At the end of the period, while the water is running into the small can, the water-meter is again carefully weighed and the weight recorded. Having recorded the weight, the water is again deflected into the large meter and what has accumulated in the small can is carefully poured into the large meter through a funnel. If the meter is nearly full, so that during the next period water will accumulate and overflow the meter, it is emptied immediately after weighing and while the small can is filling up. About 4 minutes is required to empty the can completely. After it is emptied, it is again weighed, the water-current deflected from the small can to the meter, and the water which has accumulated in the small can carefully poured into the meter. All weights on the water-meter, both of the empty can and the can at the end of each period, are checked by two observers. ABSORBER TABLE. Shortly after the subject has entered the chamber and in many instances before the sealing-in process has begun, the ventilating air-current is started by starting the blower. The air passes through one set of purifiers during this preliminary period, and as no measurements are made for this period it is not necessary that the weights of the absorbers be previously known.