214 



0. C. Glaser 



the violent progression next, whereas the strippings and the wav- 

 ings are the commonest of all. 



These observations open two ways in which the problem of 

 resolution may be attacked; by studying the time taken to solve 

 the problem and by noting the relative frequency of the most effec- 

 tive movements. The time and the frequency might both remain 

 constant, or might change, or only one might change. As a reduc- 

 tion in the amount of time taken to solve the problem need not 

 necessarily be due to an increase in the relative frequency of the 

 most effective strokes, these two must be considered separately, 

 although an increased frequency of strokes best fitted to solve the 

 problem would involve a reduction in the amount of time. If a 

 reduction in the amount of time required does occur, it means that 

 the physiological state produced by the rubber tube has been 

 resolved into the normal state more rapidly than it was resolved the 

 first time. In other words, the animal has learned by experience. 



The following Table IV contains my measurements of problem 

 solving time. In every case the animal was given the same prob- 

 lem consecutively, viz: the same rubber tube was placed on the 

 same arm, under the same conditions. As little time as possible 

 was lost between trials. 



TABLE rv* 



Trials 



*These measurements include the latent periods. 



The number of trials recorded in Table IV is small. I was 

 prevented ftom collecting more data by the sudden changes of 

 behavior before alluded to. Other animals were tried but failed 

 to react regularly even five times. The results as they stand, how- 

 ever, are worthy of confidence; they are representative of the whole 

 behavior which is varied and uneven; like the measurements of 



