TRIAL AND ERROR REACTIONS IN MAMMALS 39 



of strength of stimulus to rapidity of habit formation impose 

 on the student of behavior a scientific obligation to regulate 

 as accurately as possible the strength of motive-stimuli. Un- 

 fortunately, the easily regulated painful stimuli (electric shock) 

 that proved so useful in their work with the Dancing Mouse 

 cannot be applied where one seeks comparable results from 

 subjects who differ so widely in emotional responsiveness to 

 frequently recurring discomfort as do, for example, dogs and 

 monkeys. Even in the case of my naive human subjects, the 

 possibility that the size of the reward or the degree of my appro- 

 bation might be affected by stupidity or cleverness of reaction 

 had to be ruled out in order to obtain uniform results. 



By observing, for each individual, the relation between strength 

 and kind of motive on the one hand, and uniformity of reaction 

 on the other hand, I was able to secure results which I believe 

 to be quite safely comparable for the purposes of the present 

 exploratory investigation. The animal subjects were never used 

 for experiment until their hunger was partially appeased, and 

 if they chose to lie down and sleep while within the apparatus 

 they were merely urged to return to me for a bit of extra-tempt- 

 ing food. With cats and dogs an atmosphere of lazy comrad- 

 ship with their master greatly favored a steady, apparently 

 unemotional quest of the unlocked door. The horse's natural 

 habit of seeking food over prolonged periods of time rendered 

 him an excellent subject. The monkeys would doubtless have 

 given more uniform results had there been any way of over- 

 coming their natural distractibility, but the food seeking aspect 

 of the experiment proved to be most helpful in reducing the 

 effects of their tendency to shift their attention from the appa- 

 ratus to fortuitous sights and sounds. 



II. DESCRIPTION OF SUBJECTS 



The selection of subjects for experiment was made with refer- 

 ence to the desirability of covering considerable ontogenetic and 

 phylogenetic ranges without thereby impairing the value of the 

 results for comparison. Since it was more relevant to the pur- 

 pose of the investigation to explore for different kinds of adjust- 

 ment than to make an intensive study of any particular mode 

 of adjustment, no effort was made to obtain averages for large 

 numbers of individuals belonging to a given age or species. 



