566 VITALITY AND EFFICIENCY WITH RESTRICTED DIET. 



Har and How of 77 and 73 per cent (see table 150), which are consider- 

 ably lower than usual for these two subjects. From the fifth session to 

 the close of the experiment the two squads show uniform results. 



With 3 vd., improvement is evident between the first and second 

 experiments, Squad B indicating about the same improvement as with 

 5 vd. Squad A shows considerable improvement, but this is due to 

 the unexplainable fact that the threshold for 3 vd. on September 29 

 was abnormally low. The record for several of the subjects, i. e., 

 Bro, Moy, Gar, Tom and Vea, is only 50 to 60 per cent correct judg- 

 ment (see table 149). Some improvement is evident in the next two 

 sessions (experiments 3 and 4). The decHne in experiment 5 of itself 

 is perhaps not significant, but it is the beginning of the depression in 

 the curve during the next four sessions, that is, December 8 to January 



Squad A 

 Squad B 



100 



Fig. 103. — Percentage of correct judgments in discriminating the 

 pitch of tones. 



26, which is fairly well marked. There is recovery to the highest level 

 on February 2. Squad B shows depression for 3 vd. on December 15 

 and January 5, that is, just before and just following the Christmas 

 vacation, with a drop on January 27. 



For 2 vd. the curves start at about the same level; there is almost no 

 improvement between the first and second experiments ; there is, how- 

 ever, considerable improvement between the second and third experi- 

 ments. Specifically, Squad A continued to improve after this date and 

 reached their highest value on November 24. From December 8 to the 

 end of the experiment there was a decrement. Squad B showed a 

 decrease on January 5, to a score of 71 per cent. January 13, the first 



