PSYCHOLOGICAL MEASUREMENTS. 



615 



138 volts, as compared to the former average threshold of 118 volts, 

 a rise in the threshold of 17 per cent; this change is three times as large 

 as the average standard deviation. Seven of the 10 subjects show in 

 their averages this change to a higher threshold at the time of the 

 second experiment. It is of interest to compare this high electrical 

 threshold on October 13, the first experiment after food reduction began, 

 with the high visual threshold on the same date. (See figure 115.) 

 There are two much smaller depressions in the curve, that is, on No- 

 vember 24 and February 2, but in general, a very even level obtains 

 which is near that of the first and normal (September 29) experiment. 

 Squad B began high at 126 volts, for October 6, which was largely due 

 to the abnormally high threshold found for Fis of 228 volts. This was 

 a value higher than any found among the 63 normal subjects of the 



Squad A 



Squad B 

 110 



Fig. 116. — Electrical threshold averages and variability. 

 Solid line curves represent Squad A, and broken curves Squad B. 



series of 1917 and considerably higher than succeeding values with 

 this subject, although his general average is in the neighborhood of 

 190 volts, the limit of the supposedly normal range. Hence it seenis 

 that the improvement shown between the first and second experi- 

 ments is partly to be discounted as an individual pecuUarity. The 

 date of January 5 has previously been noted in other measurements 

 as showing results for Squad B somewhat below their average. The 

 first two experiments during the low-diet period of Squad B, particu- 

 larly the second one, show a markedly higher electrical threshold than 

 even that of January 5, which is itself higher than that for the three 



