198 PSYCHOLOGICAL EFFECTS OF ALCOHOL. 



at approximately the same point in all records, and consequently 

 permits comparison of approximately the same number of pre-stimu- 

 lation pulses. The main purpose of the present statistical arrangement 

 of the data is to isolate the post-stimulation pulse-change. On this 

 account, the arbitrary break between the post-reaction pulse of one 

 experiment and the pre-stimulation pulse of the next is without signifi- 

 cance. Since all the pulse- waves were read, the data are capable of any 

 other arrangement that statistical interests might demand. 



The theory of the statistical elaboration of the pulse data by which we 

 hoped to realize the correlation between the different phases of the 

 experimental process and the pulse-rate probably needs some expla- 

 nation. An examination of the duration of successive pulse-cycles, as 

 given in table 34, will show that, except by accident, no two successive 

 pulse-cycles take equal times. This corresponds with the well-known 

 physiological laws of the extreme susceptibility of the pulse to waves 

 of nervous excitement. The pulse of relaxed subjects is accelerated by 

 the slightest physical or mental activity. Even without conscious 

 activity, as, for instance, in sleep, it is yet complicated by a considerable 

 group of rhythmic and arrhythmic physiological processes. In normal 

 life there are short rhythms, corresponding to respiration and the 

 Traube-Hering waves. There are longer rhythms corresponding to the 

 ingestion and digestion of food, to work and relaxation, to the sequence 

 of day and night, etc. A constant base-line with clear-cut experimental 

 deviations does not exist in practice. Experimental deviations from 

 the normal, if they occur at all, will be superposed on a complex of the 

 rhythmic and the arrhythmic changes to which the pulse is normally 

 liable. The obvious problem in any statistical treatment of the pulse 

 data for experimental purposes is to disentangle the significant experi- 

 mental changes from the various rhythms. 



Even the most common use of pulse data is not free from the neces- 

 sity for similar statistical treatment. In the so-called pulse-rate one 

 may not regard as significant the measure of any individual pulse-cycle. 

 However accurate such a measurement might be, it would be meaning- 

 less unless it were known in what phase of the various rhythms it 

 occurred. At once a simple and more accurate measure of the pulse- 

 rate is to average such a large number of pulsations that it is safe to 

 assume that the lesser rhythms have run their course a number of 

 times. Under such precautions, the " pulse-rate" of relaxed subjects 

 will be relatively constant, not because the pulse-cycles do not vary, 

 but because their variations in successive periods tend to counter- 

 balance each other. Just how long a period is necessary for such 

 measurements is not a matter of universal agreement. Common prac- 

 tice finds 60" a convenient and satisfactory unit. In using such a 

 unit, one assumes that in 60 successive pulse-cycles (if 60 happens to be 

 the pulse-rate) the various lesser rhythms have become statistically 

 eliminated by counterbalancing one another. 



