HABIT FORMATION IX THE DOG 



59 



variation ; the ordinary cheap laboratory stop-watch may 

 show a difference of 0.2 second in the time necessary to start 

 and to stop it respectively; the instrument gets out of order 

 easily, and the experimenter does not react to a movement of 

 the animal which is absolutely constant for any two reactions. 

 Further the writer does not believe a difference, of 0.2 seconds 

 to be significant in an animal's making so complicated a reac- 

 tion as that of o^Dening the problem-box. 



In the experiments on Dogs i, 2 and 5, twenty trials a day 

 were allowed. Dogs 6, 7 and 8 were allowed only ten 'daily 

 trials. It will be seen from the records that the dogs allowed 

 onh^ ten trials a day required fewer trials and in some cases 

 even fewer days than did the animals gi\'en twenty trials under 

 similar conditions. This ma}^ be explainable by the fact that 

 after a certain amount of work has been done an animal may 

 tend to become careless, through fatigue perhaps, and will 

 relapse into errors, which, becoming habitual, persist for a consider- 

 able time. The tables show the distribution of errors made in 

 this work by each dog, according to the place in the animal's 

 daily series. It would be interesting to know just what is the 

 optimal number of daily trials which should be allowed to a 

 dog as compared with other animals in order to economize the 

 learning process to the greatest degree. 



TABLE 1 

 Learxixg Records ox Box 1 

 Dog 1 (Blixd $ ) 

 Problem 1 



Day, 



Trials 

 1 



1 2 



8 9 10 n 12 13 14 15 16 17 IS 19 20 21 



1 



Total 

 ■ errors 

 11 

 12 

 11 

 10 

 12 

 11 

 10 

 11 

 10 

 12 

 11 

 11 

 12 

 11 

 10 

 10 

 10 

 10 

 10 



11 



Total errors 20 20 20 17 16 17 8 12 5 12 10 15 17 11 







216 



% Error... 100 100 100 85 80 85 40 60 25 60 50 75 85 55 35 30 10 5 

 % .A.ecuracv 15 20 15 60 40 75 40 50 25 15 45 65 70 90 95 100 100 100 

 Av.time 5.010.4 2 7 2 9 2.1 1.5 1.5 1.9 19 15 1.3 12 1.1 11 10 1.0 1.0 



