Pulse and Respiration. Ill 



with the subject seated in the upright chair, holding the voice key; (5) 

 and (6) during the last third of the word-reaction measurement (taken 

 by the assistant without in any way interrupting the word reactions) ; 

 (7) following the word reactions with the subject sitting in upright 

 chair, holding the voice key; (8) and (9) at the time of recording the 

 ringer movements. The last two records were separated from each 

 other by 1 minute. 



The body electrodes and the respiration tambour were continuously 

 in place and caused the subject no discomfort. Usually connections 

 to the string galvanometer and to the respiration recorder were main- 

 tained so that no adjustment would be necessary immediately before 

 the taking of the record. The high-speed worm-gear motor used to 

 drive the paper camera was in the same room with the subject; as this 

 made a humming sound during action the subject could note approxi- 

 mately when the pulse record was being taken. It seems unlikely, 

 however, that the subject paid any attention to this sound, as he was 

 used to it in other groups of experiments in which the apparatus was 

 essentially the same. 



The usual length of a single pulse record for conditions 1 to 7, 1 of the 

 experimental hour was 15 seconds. In these records the individual 

 pulse cycles were not measured as in the case of the tetanus pulse. 

 Homologous points were chosen in the respiration curve near each end 

 of the record, and the number of pulse cycles between these two points 

 were counted. In this way any effect of respiration upon the pulse rate 

 tended to be compensated, since the same number of inspirations and 

 expirations were taken in this interval during which the pulse was 

 counted. The time distance represented in the same interval, divided 

 by the number of pulse beats, equals the average pulse-length in terms 

 of thousandths of a second for this record. Illustrative records are 

 shown in figure 11, C, D, G, and H. (See p. 96.) Here the points in 

 the respiration curves are marked so that the method of determining 

 the pulse duration can be quickly understood. The records illus- 

 trated in G and H have been reduced in length for purposes of repro- 

 duction. 



The pulse data for records 1 to 7, inclusive, are given in table 20, which 

 is really made up of 7 small tables arranged substantially in the usual 

 form employed, and is accompanied by a summary of daily averages 

 and of total averages. Since as a rule the individual pulse cycles were 

 not measured, no mean variation is available. This has made it possible 

 to condense the table somewhat by putting the differences in parallel 

 columns with the pulse duration rather than in a separate section. 

 The pulse durations and differences are given by periods. The aver- 

 ages (at the extreme right), as usual, do not include the first-period 



1 For convenience in referring to the pulse records taken under the varying experimental condi- 

 tions, the numbers 1 to 7, inclusive, as denned in the first paragraph of this section, will be employed. 



