74 



it, to name an example of it, to name a whole to which it 

 belonged, to name a part, to name another of the same class 

 or another part of the same whole." 



Over three thousand experiments were made in all. "In 

 almost every case the subject is able to accomplish his task 

 correctly. His description of his experiences shows that there 

 are in the main three kinds of complexes of experiences. Most 

 frequently the subject follows one line right through the ex- 

 periment, which then leads to the spoken word. In the other 

 cases, he may seek a word which he does not find, and which 

 he even afterwards cannot name, or he may have intended 

 to say a certain word, but for some reason or other, wittingly 

 or unwittingly, have said another. In general the first class, 

 the simple reproductions, takes place in a good deal less time 

 than the other two classes, the complex reproductions, of 

 which two the second named usually and naturally last longer. " 



"How does any one particular reaction come about and 

 not another?" Watt asks. "The first influence at work on 

 the subject is the given task. This he hears spoken by the 

 experimenter, and generally repeats to himself in words, 

 e.g. 'find a part!' 'name an example!' or he may exemplify 

 the experiment to himself, e.g. 'animal dog', and so on. 

 The scanty description of the preparation for the experiment 

 given in the subject's account of it does not help us to form a 

 very clear idea of what the process itself is. It was found , how- 

 ever, as a series of detailed curves show, that of all the simple 

 reproductions the percentage of occurrence of each of the three 

 above-named classes changes regularly and similarly with 

 each subject from one task to another. This leads to the 

 assertion that the task has a regular influence on the nature 

 of the experiences of each subject, which becomes particularly 

 evident between the two larger groups of simple reproductions, 

 those containing visual representations and those containing 

 nothing at all." In both simple ajid complex reproductions, 

 it is found that the duration of the reaction is on the average 

 dependent on the nature of the task. 



Further, Watt's results are confirmatory of experiments 

 conducted by Kiilpe in Wiirzburg in the same year. 1 We 

 may here, in a few words, sum up the results of these latter 

 experiments. In the first place, answers were given in ac- 

 cordance with the task set. Further, it was found to be more 

 agreeable for the observer to work with a task than without 

 one. To work without any task at all proved to be very dif- 



^'Versuche iiber Abstraktion", Bericht des I. Kongr. f. exp. Psych. 

 Gieben, 1904. 



