560 VITALITY AND EFFICIENCY WITH RESTRICTED DIET. 



Berry, who used this same test in an investigation carried out at the 

 International Y. M. C. A. College in 1915.^ 



The average results for both squads, considered together, are shown 

 in figure 102. The upper horizontal line of dates at the top of the 

 figure, beginning September 29, October 13, etc., is for the sessions of 

 Squad A (see table 147). September 29 is the only normal date, the 

 others representing a reduced diet. The lower horizontal line of dates 

 beginning October 6, November 3, etc., is for the sessions of Squad B. 

 The first five of these, that is, until and including January 5, were 

 normal sessions (uncontrolled diet). The last three, January 13, 19, 

 and 27, were under conditions of reduced diet when the reduction was 

 much more pronounced than usual for Squad A. On September 29, 

 the one normal date, Squad A had an average value of 5.7. On the 

 two succeeding dates, October 13 and 27, separated by 2-week intervals, 

 the average errors were somewhat larger, 5.9 and 6.0; several of the 

 men made their poorest records on October 27. Other subjects — Can, 

 Gul, and Pea — showed very little improvement over the previous two 

 sessions. The improvement was not what would be expected nor 

 equal to that found with Squad B and shown by the curve in figure 102. 

 Squad B, it is true, showed no improvement between the first two 

 sessions, but a period of four weeks elapsed between the two experi- 

 ments of October 6 and November 3. Between the second and third 

 sessions the period was exactly two weeks; much improvement was 

 shown, the errors being reduced to 4.6. Between sessions 3 and 4 

 a period of a month again elapsed and no improvement is shown. 

 The same is to be said for sessions 4 and 5, which were separated by 

 3 weeks. When the sessions were separated by two weeks or less an 

 improvement in accurate tracing with Squad B almost always resulted. 

 While the two curves start at nearly the same point, it is evident 

 that, considering the number of practices and the intermissions. Squad 

 B made definitely the better performance. Squad A did as well as 

 Squad B at the sixth session, that is December 8 for Squad A and 

 January 13 for Squad B. This abnormal rise for Squad A is associated 

 with the more comfortable period in the early part of December, 

 when the men were receiving more food. December 19, which shows 

 a depression in the curve, was probably influenced by the fact that the 

 observations on this date immediately preceded the Christmas vaca- 

 tion. On January 12 the net energy given the subjects was small to 

 counterbalance the overeating during the vacation period. On the 

 last two dates, January 26 and February 2, the group made their best 

 records. It is unfortunate that the experiments with Squad B did 

 not come with the same regularity and were not equal in number to 



^ Fisher and Berry, Physical Efifect of Smoking, New YorkJ 1917. These authors state that a 

 fountain pen was used for drawing between the lines, that a metronome was employed to time the 

 men, and that 17 seconds were allowed for each line. Professor Berry informs us that "17" is a 

 misprint and the time should be 14 seconds, the same as that used in our investigation. 



