FREE ASSOCIATION. 



121 



number of persons who had given responses to them of less than the 

 above standard of 9.0. These words were then paired, and one of each 

 pair assigned to either series, which are called series D and D' respec- 

 tively. The order of the words was kept the same as in the original 

 Kent-Rosanoff series, so the two lists were finally constituted as follows : 



KENT-ROSANOFF WORDS, SERIES D. 



Head. 



Religion. 



Bitter. 



Hammer. 



Thirsty. 



Square. 



Loud. 



Thief. 



Bed. 



Heavy. 



Baby. 



Scissors. 



Quiet. 



Street. 



King. 



Blossom. 



KENT-ROSANOFF WORDS, SERIES D'. 



Dark. 



Music. 



Soft. 



Eating. 



Mountain. 



House. 



Comfort. 



Short. 



Fruit. 



Butterfly. 



Common. 



Woman. 



Wish. 



Beautiful. 



Window. 



Rough. 



Needle. 



Red. 



Sleep. 



Anger. 



Girl. 



High. 



Earth. 



Cabbage. 



Eagle. 



Stomach. 



Lamp. 



Boy. 



Health. 



Bible. 



Sheep. 



Bath. 



Cottage. 



Swift. 



Blue. 



Stove. 



Long. 



Whisky. 



Child. 



City. 



Butter. 



Doctor. 



Lion. 



Joy. 



Tobacco. 



Moon. 



Green. 



Salt. 



Cheese. 



Afraid. 



One or the other of these constituted the fourth series of each experi- 

 mental day. The figures given in table 16, as c, represent the median 

 of the different figures for the usualness of each response, as calculated 

 from the Kent-Rosanoff tables. The higher this figure, the more usual 

 are the subject's responses. The results of these calculations for the 

 4 days are shown in table 16. 



TABLE 16. Index of community "c" of 50 Kent-Rosanoff words normally and under alcohol. 



Other experiments had made it apparent that practice increases the 

 individuality of the response, and it is borne out in these figures. It 

 seems certain, also, that the object of splitting the series, namely, to get 

 two series of equal tendency in respect to the frequency of response, 

 was not achieved, at least for this group of subjects. There is an 

 obvious tendency for Series D' to show more unusual responses than 

 Series D that is beyond any reasonable expectation from practice. In 

 further experiments it would be advisable to repeat the whole Kent- 

 Rosanoff series. This unfortunate difference in the series materially 

 interferes with interpreting the results in reference to alcohol. Marked 

 alcohol effects are clearly not present. The figures for the alcohol days 

 are generally between or on either side of those for the normal days. 

 Diagrammatically the relationship is shown in figure 25, the general 



