234 KANSAS UNIVERSITY SCIENCE BULLETIN. 



greater effect. It is unfortunate that we have two sets of con- 

 ditions here, as it makes it uncertain whether the effect is due 

 entirely to the mode of producing physical fatigue or to the pos- 

 sibilities of learning in the tests. However, it is certain that 

 both modes of physical fatigue have decreased mental efficiency 

 in a marked degree. All the subjects have shown some de- 

 crease in mental ability after physical work as tested with 

 the sounders. This decrease has been greatest for Be. and 

 S., and moderately large for Bt. The other subjects, except 

 Bu., have been affected definitely. Bu.'s results are uncertain. 



ASSOCIATION TEST. 



As in the sounder test, any variation in speed that can 

 affect in any way the accuracy of the association processes has 

 been eliminated. The variations in the time of recall on the 

 following day represent the degree to which the syllables had 

 been learned, and do not indicate any possible changes due to 

 the influence of fatigue upon the process of recall. The 

 syllables were exposed at a regular rate after rest and fatigue 

 periods and the only possible effect of the physical fatigue 

 would be upon the learning process. The effect of fatigue 

 upon the process of recall constitutes a distinct problem that 

 should be studied, but the present series deals only with the 

 influence upon the learning process as evidenced by recall 

 after a period that is long enough that fatigue can not affect 

 recall directly. 



The per cent correct and the average association time for 

 each day's test are given in Table IVa. As the fatigue days 

 alternated with the rest days, the results for corresponding 

 days are placed opposite each other in the first four columns. 

 The per cents of increase or decrease for the fatigue days as 

 compared with the results for the rest days are given in the 

 fifth and sixth columns. The total per cent correct, and the 

 average time, taken from the total scores, without reference 

 to the individual days, are given in Table IVb, which also shows 

 the number of cases used in determining the average time, and 

 the number of cases in which the time was more than 10,000 a 

 and the number of cases in which it was less than 3000 <* . It 

 should be remembered that the total number of cases given 

 represents those responses that were correct or partially cor- 

 rect. The incorrect responses were not used in determining 

 the average association times. 



