It was desirable, then, to maintain a rate of walking within this optimum limit and a rate of 70 meters per minute was selected. This is a fair rate of walking and does not in 20 to 25 minutes introduce the element of fatigue. This rate was also convenient because the amount of carbon dioxide eliminated during the time available for the experiment, judging from other data, would in all probability not exceed 1 per cent of the chamber volume. The capacity of the Haldane gas-analysis apparatus used for the carbon-dioxide determinations was 1 per cent, as previously stated, and it was desired to use this small apparatus rather than other larger and more complicated forms. Furthermore it was desirable not to have so great a concentration of carbon dioxide within the chamber that a marked stimulus to the pulse or respiration would be produced.

1 Dune. Denkachrift. d. math.-natur. Kl. d. kaiserl. Akad. d. Wissenschaften, 1909, 86; also Benedict and Murschhauser, Carnegie Inst. Wash. Pub. No. 231, 1915, p. 85. 2 Brexina and Kolmer, Biochem. Zeitschr., 1912, 38, p. 142.

Fig. 13. - Photographic record of the pulse, respiration, and steps of Bro while he was walking in the treadmill chamber.

A and B, record for the standing-walking period. A and B together composed one continuous record which has been cut in two approximately equal lengths for illustration. X, point at which walking began. C, record at end of sixth minute. D, record at end of twenty-fourth minute. E, record for the walking-standing period. Y, point at which walking ceased.

The Step Recorder

A record of the steps taken during walking was obtained by a small electrical contact device at the rear of the treadmill to which the subject's left foot was attached by a long coil spring. The spring had sufficient tension to operate the contact when the foot was thrown forward, but at the same time was so weak that it was not noticeable to the subject as he walked. The electric contact operated a message register, also a signal magnet placed before the camera which was used in connection with the string galvanometer for the pulse records. By the latter, photographic records showing the steps were secured on the same paper with the electrocardiograms which were taken at the close of the first, sixth, twelfth, and twenty-fourth minutes of walking. A specimen record is shown in figure 16 with the pulse records.

There was an occasional chatter of the signal magnet due to a rebound by the coil spring. This chatter is in most cases evident in the spacing on the records and such irregular movements of the signal magnet are neglected in the counting. From these records is found the number of steps per minute at the time of the observations and the average of these is used as the steps per minute for the period.

Pulse-Rate

In order to secure pulse records while the subjects were walking, leads from the string galvanometer in the psychological laboratory were carried to the room where the treadmill was situated and entered the chamber through the hard-rubber plug in the front of the cover (see P, fig. 14). The leads terminated inside the chamber in a receptacle into which the subject plugged the terminals of the body electrodes after he entered the chamber. These electrodes were the same as those used in securing electrocardiograms during short periods of muscular exertion. They are shown in figure 26 and described on page 153.

Since there was always a certain amount of electrical leakage from the treadmill motor, as well as static electricity of the leather belt, it was necessary to protect the string of the galvanometer from these stray charges which interfered with the electrocardiograms as well as endangered the safety of the string itself. This was successfully done by grounding the subject by means of the third electrode, G in figure 26, worn on the lower chest and connected to an iron water-pipe outside the chamber. This arrangement had previously been found more satisfactory than grounding the treadmill or the string galvanometer and made it possible for the men to walk in ordinary shoes without further insulation.

A short branch line connected by a double-pole double-throw switch, easily accessible to the operator, allowed either the subject walking on the treadmill in the chamber or a second subject resting in an adjoining room to be in circuit with the string galvanometer. In this way electrocardiograms could be taken alternately from two subjects with but little loss of time.

As a rule the pulse was recorded photographically for a period of 15 to 20 seconds, first while the subject was sitting in an adjoining room, then while standing either in the adjoining room or on the treadmill before the walking started, and again at the end of the first, sixth, twelfth, and twenty-fourth minutes of walking. During the walking periods of January 28 and February 3 a visual pulse count was also made each minute by observing the deflections of the string for 15 to 20 seconds. A record of the pulse-rate was thus secured on these two dates for each minute of walking, either by the photographic method or by visual count with the aid of a stop watch.

In the experiments of January 28 and February 3, electrocardiograms were also secured of the pulse at the time of transition from standing to walking at the start, and again from walking to standing at the close of the period. As a rule these transition records consisted of 15 seconds of the first stage and 60 and 30 seconds of the final stage of the transition. A reproduction of a typical group of records for one subject is shown in figure 16. Finally the radial pulse was counted after the walking had ceased and the subject had been sitting quietly for 4 and 8 minutes.

For the records taken during sitting, standing, walking, and sitting after walking, the counts were in terms of pulse-rate per minute. In the transitional pulse records the individual pulse cycles were measured, as was done with the electrocardiograms during short periods of exertion described on page 151.

Respiration-Rate

A pneumograph worn by the subject around the lower chest was connected to a tambour, shown at A in figure 17, by means of a rubber tubing, B, which left the chamber by a brass tube in the rear of the skirt-wall. On one radius of the diaphragm of the tambour was cemented a light aluminum lever, C, the base of which pivoted on the edge of the tambour which was segmented at this point.1 Supported by this lever a fine copper wire, D, dipped in a cup of mercury, E. Each respiration, acting through the pneumograph and tambour, made a contact between the wire and the mercury and operated a signal magnet placed in front of the camera used for the electrocardiograms. Adjustment of the contact in the mercury cup was made by means of the reservoir, F, while a screw pinchcock, G, on the tube from the pneumograph damped the movements of the lever. The apparatus was so designed that if desired a second subject might be alternately connected to the tambour by means of a second tube, H, and the 3-way cock, I.

1A similar form of tambour making use of the principle of Frank's segmented capsule is illustrated by Wiggen. Circulation in Health and Disease, Philadelphia, 1915. See figure 11, page 57.

When the subject entered the treadmill chamber the pneumograph was connected to the tube in the rear of the skirt-wall and the proper adjustment of the reservoir made so that the hair wire opened and closed the circuit with each respiration. This actuated the signal magnet in front of the camera of the string galvanometer so that each time the pulse was photographed a record of the respiration was also secured on the same photographic film. A typical section of such a record is given in figure 16, in which the time in two seconds and fifths of a second, the pulse, respirations, and steps are shown.

Fig. 17. - Respiration recorder.

The subject on the treadmill wore a pneumograph which was connected to a segmented tambour A by means of the rubber tubing B; with each respiration the tambour caused the aluminum lever C to complete an electric circuit through the fine wire D and the cup of mercury E. This circuit operated a signal magnet not shown in the figure. F, mercury reservoir for adjusting the level of the mercury in the cup E; G, pinchcock for damping the movement of the tambour; H, a duplicate of B for use with a second subject; I, 3-way stopcock for throwing either B or H into connection with the tambour.

Blood Pressure Following Walking

Determination of the blood pressure by any of the various sphygmomanometers is exceedingly difficult during walking, since the sway of the body, jar of the step, etc., all tend to make the readings uncertain. Moreover, in these walking experiments the fact that the subject was in a closed chamber precluded any thought of such an attempt during the actual period. The only determinations feasible under these conditions were records of the blood pressure shortly before and immediately after walking, the latter continuing several minutes to note any changes in pressure as the stimulus from walking decreased.

The pressure was therefore taken with a Tycos sphygmomanometer immediately at the close of the standing metabolism experiment. The subject wore the cuff in place on his arm during the time he was walking in the chamber. At the close of the walking period the chamber cover was lifted, and while the subject was still walking the cuff was inflated. The treadmill was then stopped. The systolic and diastolic pressures were taken in rapid succession during the next two minutes which usually resulted in readings being secured at 15 seconds, 1, 1 1/2, and 2 minutes after walking ceased. The subject then stepped off the treadmill, was weighed, and after sitting quietly in an adjoining room the pressures were again read on the fifth and ninth minutes after walking.