422 VITALITY AND EFFICIENCY WITH RESTRICTED DIET. 



gether with a similar average obtained for all of the records of the 

 63 men of the series of 1917, are plotted in curves and embodied in 

 figure 93. A similar series of curves for Squad B, together with the 

 average for the series of 1917, are presented in figure 94. As explained 

 on page 416, two records, such as are illustrated in figure 24, were taken 

 for each subject and date; therefore, as 10 subjects are averaged, each 

 plotting-point is usually the average length of 20 pulse-cycles. The 6 

 pulse-cycles immediately preceding the beginning of activity were used 



654321 1234- 



12 3 4 5 



Fig. 93. — Composite curves for changes in pulse-cycle duration with exertion, Squad A — 

 Records for 3 days with Squad A compared with the results for normal series of 1917. 



to indicate the level of the quiet pulse. The first 20 cycles following 

 the end of activity were taken as indicative of the changes in the period 

 of rest after exertion. Since in our procedure the duration of activity 

 was made as constant as practicable, the number of pulse-cycles which 

 came within this limit was naturally a function of the pulse-rate and 

 varied with different subjects. This, unfortunately, when one aver- 

 ages the records, causes an irregularity at the point of transition from 

 activity to rest. In figure 93 the curves for Squad A show an irregular 

 rise at this point due to the predominating influence of these faster 

 pulse-rates. (See Gul and Tom, table 94.) 



In all of the curves for Squads A and B, the major portion of the 

 period of activity and also of the period of subsequent rest have a very 



