PULSE DURING MENTAL AND PHYSICAL WORK. 193 



line was run on the kymograph, in which there were neither sphygmo- 

 graphic records nor reaction records. Each group of 50 reaction 

 experiments was interrupted in the middle for readjusting the drum. 

 The full length of the drum just sufficed to give 50 reaction lines, plus 

 the blank lines and a few accidental blanks that occurred in the course 

 of the series. 



The reading of these records was a painstaking and a time-consuming- 

 process, but presented no special difficulties. A small probable error 

 is involved in each pulse-cycle record. It depends on the fact that the 

 pointer will leave the mercury at some point of the systolic rise. It 

 may be near the beginning or it may be near the end. However, since 

 the duration of the systolic rise is relatively short (about 0.08"), and 

 since extreme positions give no break, the probable error of any one 

 record will not be over 0.02." In an evenly running series it is much 

 less. On account of this error it seemed inexpedient to read the curves 

 closer than 0.01 ." For most purposes of correlation this is close enough. 



The records are often complicated by movements of the recorder 

 incident to the dicrotic and the post-dicrotic waves, which may give 

 two or three breaks for each pulse-wave. These breaks, however, are 

 usually of regularly decreasing lengths, and seldom interfere with the 

 reading of the record. Gross body-movements, however, produce 

 serious disturbances in the curve, which render a record illegible as long 

 as they persist. Such irregularities are not without their advantage, 

 since they indicate the presence of body-movements and prevent a 

 misinterpretation of physically conditioned pulse-changes. 



ELECTRO-CARDIOGRAMS FROM BODY LEADS THROUGH CONDENSERS. 



The arrangement is as follows : Instead of taking electro-cardiograms 

 from any of the well-known Einthoven or Waller leads, which require a 

 relatively complete relaxation of the subject and prevent any other use 

 of the limbs, the method we employed attaches the electrodes to the 

 body, on either side, directly under the armpit. For registration of the 

 duration of the pulse, this device has certain advantages over all other 

 forms of recorders: (1) In general, the electro-cardiogram has ideal 

 configuration for accurate measurements. The shape of the curves is 

 relatively constant, and the sharp first systolic spike (Einthoven's R 

 spike) is peculiarly clear-cut. (2) Use of the body leads gives electro- 

 cardiograms which can not be used diagnostically because of the 

 uncertainty of the position and contact of the electrodes. On the 

 other hand, they leave the limbs free for any sort of concurrent activity. 

 (3) Situated directly over relatively small trunk-muscles, even violent 

 activity need not interfere with the records. This is a unique advan- 

 tage of the body leads. As elaborated in the Nutrition Laboratory by 

 the assistance of Professor H. M. Smith and Mr. K. H. Brown, the 

 device gave excellent records of the pulse of a man walking on a tread- 

 mill for hours at a time. 



