SUMMARIES AND CORRELATIONS. 265 



alcohol. That finger-movements would be serviceable in considerably 

 less degree for a general test, when for any reason the eye-movements 

 were not available, is obvious if one remembers the gross differences in 

 the pre-experimental practice of the finger-movements of different 

 individuals, and the relative ease with which they can be arbitrarily 

 modified. In every respect we believe that the eye-movements are the 

 most reliable and the most important measurements of the group. 

 They are least open to arbitrary modification, vary directly with the 

 dose of alcohol, come closest to the total average of all the tests, cover 

 the most general characteristics, and come nearest to being a true test 

 of the individual's susceptibility to the effects of alcohol. 



Aside from the practical value of this correspondence between the 

 effects of alcohol on the coordination processes and the average effects, 

 it has a rather far-reaching theoretical implication. If, in all the diverse 

 processes which we have measured, the coordination processes represent 

 a central numerical tendency, it must be that they correspond in some 

 closer way than the rest to a real central tendency of the alcohol effect. 

 It would seem to indicate that the alcohol change in the average per- 

 formance of our subjects is a function of central coordination. If this 

 indication is substantiated by later investigations it should prove to be 

 not only of the utmost importance for an understanding of the various 

 manifestations of the effect of alcohol in individual cases and for the 

 general phenomena that accompany its excessive use, but it would 

 throw a flood of light on the complex organization of normal psycho- 

 physical processes, as well as on the effects of fatigue and other de- 

 pressing agents. 



NUTRITION LABORATORY OF THE 



CARNEGIE INSTITUTION OF WASHINGTON, 



Boston, Massachusetts, May, 28, 1915. 



