398 



SCIENCE 



[N. 8. Vol. XLV. No. 1165 



way. My preliminary experiments show 

 that where persistent images are found, as 

 in the various forms of hallucinations of 

 normal persons, or persistent ideas as in the 

 incipient forms of delusions, one can break 

 them up and ultimately destroy them in 

 some cases and in others substitute for them 

 images and ideas having a more agreeable 

 nature. The' breaking down and destruc- 

 tive process was actually accomplished in 

 cases of apparitions, accounts of which 

 I have already published. I am now work- 

 ing upon experiments having to do not 

 so much with the possibility of supplanting 

 spontaneous images having a disturbing 

 character by those that are more agreeable, 

 for that I now know is possible, but more 

 particularly with the various modes of 

 bringing it about. I can already see that 

 the photograph, stereoptieon views, the 

 moving picture and its accessories appeal- 

 ing to the other senses and the phonograph 

 will be very useful in the building up and 

 supplanting process just mentioned. 



Before suggesting a new method of 

 studying personality one naturally em- 

 ploys other methods to test and control his 

 own. Biographical, autobiographical, his- 

 torical, observational, introspective, etc., 

 methods have actually been emplo3^ed in 

 the course of these experiments, in that the 

 opinion that images are betrayers of per- 

 sonality is based, as has been said, upon 

 the general agreement between what the 

 images reveal regarding the observer and 

 what others, including myself, know of him 

 and of what he himself said regarding their 

 revealing power. 



It is everywhere evident that the more 

 impersonal, systematic, and exact observa- 

 tion and introspection employed in the 

 image method, give more reliable informa- 

 tion regarding an individual than the 

 chance and uncontrolled observation and 

 introspection used in the methods just men- 



tioned. Again, the observer's powers of 

 discrimination in employing the image 

 method are not called upon to the extent 

 they would be if he was asked to make a 

 series of judgments upon his own charac- 

 teristics, since he was not informed until 

 the experiments were completed as to the 

 object of the investigation. What was 

 learned by the image method was much 

 more intimate and detailed than by the 

 others. The experiments yielded not alone 

 information agreeing with what was previ- 

 ously known of the observer by others and 

 by myself, but what was not even known to 

 the observers themselves until after the in- 

 vestigation had been made, that is, the 

 images actually brought characteristic re- 

 actions to the observers' attention which 

 they had not noticed before. 



As it is in the association methods that 

 one would expect to find the image method 

 competitors, I determined to make some di- 

 rect tests and asked each of the observers 

 who took part in these experiments to write 

 one hundred words in succession. The 

 words written show that very much more 

 can be learned through using the image 

 method. The same thing was shown in 

 applying the Kent-Rosanoff associative 

 method. 



It may be said in a general way regard- 

 ing all the association methods, and among 

 these I include the methods of psycho- 

 analysis, that the image method gives re- 

 sults as regards personality that are far 

 more definite and clear cut. 



The experiments of Ach and Barrett 

 along the lines of temperament and char- 

 acter mark an important step in experi- 

 mental psychology, and it would be very 

 interesting to compare results obtained by 

 applying their methods and material with 

 those obtained by the image method. How- 

 ever, the confining of character study 

 largely to data having to do with choice, 



