The general routine of the treadmill experiments was as follows:

On the completion of the measurements of the standing metabolism, which were conducted in an adjoining room on the portable respiration apparatus, the blood pressure of the subject was measured and the body electrodes were applied. The subject then came to the treadmill room, where these body electrodes were plugged into one branch of the leads connecting with the string galvanometer. During the 5 to 10 minutes that the subject was sitting quietly a photographic record was taken of his pulse by the string galvanometer. The subject then stood and after 5 to 10 minutes his standing pulse was again recorded by the string galvanometer. As soon thereafter as conditions permitted he entered the chamber. On his entering, the respiration pneumograph, which had been placed around his waist, was connected to the special tambour, the body electrodes were plugged into the branch of the galvanometer line extending to the chamber and the step counter was fastened to his ankle.

The subject stood quietly on the treadmill with the cover of the chamber lifted to its full height, which brought his head just inside the base of the cover. During this time the three fans in the chamber were running and the 15-cm. blower was delivering air over the side of the skirt into the chamber at its full capacity of 12 cubic meters per minute.

The assistant at the observer's table recorded the readings of the various counters and gave a warning of 15 seconds to the other assist-ants before the starting of the treadmill. In the experiments of January 6 the first walking pulse taken with the string galvanometer was at the end of the first full minute. In the experiments of January 28 and February 3 a photographic record was taken of the pulse at the time of the transition from standing to walking, which gave the pulse for the last 15 seconds of standing and continuing through the first full minute of walking. After 2 minutes of walking the large blower was stopped and wheeled back from the edge of the chamber. At the end of 2 1/2 minutes1 of walking the cover was lowered into its seal and the connecting pipe in the circuit from the drier to the rear of the chamber (L in fig. 14) was put in place; the chamber was now completely sealed. By this time the assistant had adjusted the mill to a speed of 70 meters per minute. At the end of 3 1/2 minutes of walking one assistant read the psychrometer; at the end of 3 3/4 minutes a second assistant read the level of the spirometer bell and a third balanced the Wheatstone bridge of the resistance-thermometer circuit. At the end of exactly 4 minutes of walking1 which marked the beginning of the period proper, the temperature was taken and the air blowing through the drying circuit was stopped. Air samples were then drawn into the two Haldane gas-analysis apparatus with the necessary precautions. As soon as this was completed the air was again started in the drying circuit, and the analysis of the air proceeded forthwith.

1 On January 6 the cover was lowered directly after the subject entered the chamber.

As it was impossible for the assistant to read both step and distance counters at the same moment, it was the practice to read the step counter 10 seconds before and the distance counter 10 seconds after the 4-minute signal. As this was done at the 4,14 and 24 minute readings, the elapsed time was the same, namely, 10 minutes in each instance, for the experimental period proper. The barometer was read as soon after the 4-minute signal as possible. During the next 10 minutes the assistants were occupied in watching the psychrometer, temperature, counters, and especially the spirometer bell, which rose with the temperature of the chamber as the experiment progressed. The temperature was controlled by means of electric fans blowing over the chamber and by opening the windows of the room. On account of the large radiating surface of the chamber and the thorough stirring of the chamber air, any tendency of the air to expand unduly could be quickly checked in this way.

At the close of 6 and 12 minutes of walking photographic records of the pulse were made. As stated earlier, the photographic record thus secured includes also records of the step and respiration rates. During the interval when the pulse was not being photographically recorded the operator made a visual count of the deflections of the string for 15 to 20 seconds each minute. By this means and the aid of a stop-watch records were secured for each minute. At 9 1/2 minutes after the period began, which was 13 1/2 minutes after the walking began, a reading was taken of the psychrometer; the spirometer was read at 9 3/4 minutes after the beginning of the period; and at 10 minutes gas samples were drawn, and the temperature, barometer, and counters were read in the same way as at the beginning of the period on the fourth minute of walking. This marked the end of the first half of the period, and these data were taken as a control and possible check should anything irregular appear later in the experiment.

1 These 4 minutes of walking were looked upon as a period during which the carbon-dioxide production would have reached a uniform rate. That 4 minutes is sufficient for the purpose is indicated by the fact that both the pulse and respiration rates are fairly uniform by the fourth minute of walking; furthermore, unpublished results of experiments made in the Nutrition Laboratory show that the rate of oxygen consumption when a person walks at 30 per cent grade at a rate of 50 meters per minute becomes uniform by the end of the second minute.

The walking then continued without interruption for another 10 minutes, when the same procedure was followed, except that at this time a sample of air was likewise drawn into the Sonden gas-analysis apparatus for the determination of the oxygen present at the end of the experiment. At this time a photographic record of the pulse was again made. The walking continued until the samples of the air were drawn, when the cover of the chamber was lifted and pressure was put upon the cuff of the sphygmomanometer. At this point the treadmill was stopped and the blood pressure was determined as rapidly as possible during the next 2 minutes. The subject was then disconnected from the respiration tambour, galvanometer leads, and step counter. On leaving the chamber he was weighed and the final records of blood pressure and radial pulse were made in an adjoining room.

At the conclusion of the experiment the large ventilating fan was moved up to the chamber, the windows were opened, and both the room and chamber were given a thorough ventilation during the following 10 to 12 minutes. During this time preparations were made for the next experiment.