This was one of the morning measurements. A brief review of literature pertaining to pulse records of this sort and a description of technique previously used has been given in another publication by one of us under the convenient name of "tetanus pulse."1 Formerly the changes in pulse-rate were produced by the subject's clenching his fists and making the muscles of arms, legs, and trunk rigid for a given period, after which he relaxed in a steamer chair as he had been previous to the exertion. From the standpoint of quick transitions from rest to exertion, and from exertion to rest, no routine can be better, but, as pointed out in the previous publication, much depends upon the subject for the amount of exertion actually put forth and in sustaining it with some uniformity during the period.

1 Miles, Carnegie Inst. Wash. Pub. No. 266, 1918, p. 92.

The method in the present research and in some previous unpublished studies is clear from an inspection of figures 22 and 23. The subject, connected to the string galvanometer, reclined comfortably in a steamer chair. Above him and within easy reach was suspended a bar of suitable dimensions, which was easily capable of bearing the subject's weight. At a given signal he grasped the bar and raised himself to the position in figure 23, with his weight entirely free from other support. He remained in this position for a period of 5 seconds and then quickly settled back into the chair and into as complete relaxation as possible. The exertion produced in this way is more uniform than when the subject stiffens his muscles but remains seated. The bar is placed at such a height that, in order to keep the body free from other support, it is necessary to thrust the feet forward and hold them up. This causes muscular tension in abdomen and legs. Thus the whole body is exercised and it is rather unlikely that a subject would find it possible to learn to isolate the muscles most used and to keep the others well relaxed.

The photographic pulse tracing (electrocardiogram) was continuous for about 8 seconds of rest preceding exertion, during the period of exertion, and for 20 to 30 seconds following the exertion. Sample records are reproduced in figure 24. It should be noted that these records are made primarily to record heart rates. They are not standard electrocardiograms for three reasons: (1) The deflection of the string of the galvanometer is not standardized, for the string is made taut in order to reduce the large deflection caused by the swift and general body movements. (2) Electrical condensers are placed in series with the string and the subject; in this case 12 microfarads were used. These condensers serve to minimize deflections in the base-line, that is, to cut out or reduce all slow changes in potential, such as those occasioned by movement of the arms and trunk of the subject. This causes, of course, a reduction in the size of the P and T waves of the common electrocardiographic complex. (3) The points of leading off from the subject's body, that is, directly below the arms, while corresponding fairly well to Einthoven's lead No. 1 (Ll), are not exactly the same as this or any other commonly used lead.

Figure 24 shows records for three subjects. In the illustration it has been necessary to cut each record in two. Thus, A' must be considered as belonging at the right hand end of A, and together they form one continuous record. The same applies to B and B' and C and C'. The three sections of a record are clearly marked: At the left is a short section covering 6 or more pulse cycles representing a period of rest before the exertion (the paper may be regarded as having moved from right to left when the record was being taken). Near the center is the section of the pulse record taken during the moments when the subject was supporting his weight on the bar. The string is seen to have been disturbed by many action currents coming from the body muscles. Suddenly these action currents cease, indicating that the subject has settled back into the chair; the pulse remains rapid or perhaps becomes more rapid than during the actual intervals of exertion, and then, gradually, the heart cycles lengthen, while the subject is quietly resting in the chair. Only the large R wave of the electrocardiogram complex can be seen during the period of exertion. Pulse-cycle length was measured from R to R.1

Fig. 22. - One corner of room C, the main psychological laboratory.

The figure shows the arrangement of a portion of the apparatus used in taking electrocardiograms under conditions of short periods of muscular exertion. The string galvanometer and arc lamp are at the left and not shown in the picture. The subject is connected to the galvanometer by electrode leads, E. A standardizing resistance arrangement H and a protection resistance R' make possible control of the string's deflections. Electrical condensers C in series with the subject and the galvanometer string minimize its disturbance at the moments of physical activity. The photographic camera P' is driven by a motor on the other side of the wall at M. The actum of the camera is very quiet, its speed is controlled at S, and the paper or film after exposure is deposited in the large dark box D. The bar H, upon which the subject exercises, is directly above him and within easy reach.

Fig. 23. - The subject in position during the moments of physical exertion.

Previous to exertion the subject reclines comfortably in a steamer chair. At a given signal he reaches, grasps the bar above him, and raises himself, thrusting his feet forward to keep all his weight free from other support, and maintains the position shown in the figure until the signal "down,"' at which he relaxes into tne chair. The electrocardiographic record is continuous through rest, exertion, and rest. The experimenter occupies a position at the left.

Fig. 24. - Sample pulse records showing changes in the pulse-rate occasioned by short periods of exertion.

A and A' are parts of the same continuous record, which has been cut for purposes of illustration. From the left-hand end of A to the point U, the subject was resting quietly in the steamer chair; at U (up) the weight was lifted and this muscle tension was continued to D (down, shown in A'), when the subject let himself down into the chair and was relaxed until the end of A'. The same description would apply to B and B' or to C and C". The pulse-cycle length is measured from R to R, and the time is in one-fifth second. The small irregular vibrations in the pulse line show the period of activity, which obviously began before U and continued slightly beyond D . In the post-activity section of B', the pulse-cycle length returns to normal rather slowly as contrasted with the changes shown in the similar section of C.