262 PSYCHOLOGICAL EFFECTS OF ALCOHOL. 



ence of alcohol as we might reasonably expect from other tests. That 

 in spite of this repetition effect dose B shows a unanimous depressing 

 effect of alcohol as measured by the differences, giving an average of 

 14.5 per cent, is all the more striking. So, as we have already pointed 

 out, the 5.36 per cent shortening of the reaction time after dose A must 

 not be regarded uncritically as a reversal of the sign of the effect of a 

 small dose. It is due in part at least to the position of the experiment 

 with dose A in the group of repetitions. From all the data taken 

 together it is clear, however, that the effect of 30 c.c. of alcohol on the 

 eye-reaction time must be exceedingly slight. 



The relatively large average difference between the two normal days 

 in the threshold for Faradic stimulation (Z values) is less easily ex- 

 plained. It is probably not so much an average effect of repetition 

 as it is a result of gross change in a single individual (Subject IX). 

 Into the conditions that govern the changes in the threshold to Faradic 

 stimulation we have too fragmentary an insight to venture a hypothesis 

 in this case. We have already seen that changes in the degree of 

 assurance which is demanded by the subject may seriously influence the 

 apparent results. The influence of other mental or physiological 

 factors are evident in the daily rhythm. Effects of fatigue or the dis- 

 turbance of accidentally distracting noises are difficult to eliminate or 

 to equalize. 



As compared with the measurements on which most of the classical 

 work on the effect of drugs has been done, all of our measurements 

 show conspicuously little effect of practice. Least affected are the 

 latency of the lid-reflex and the angle velocit}' of the eye-movements. 



CORRELATION OF THE VARIOUS MEASUREMENTS WITH 



THE AVERAGE. 



Of practical as well as of theoretical interest is the question which of 

 the various measurements shows the closest correspondence with the 

 average performance of the various subjects. This is of importance, 

 in the first place, in the effort to estimate the relative value of the 

 different measurements as an indicator of individual differences. It is 

 of importance, in the second place, as an indicator of the central ten- 

 dency of the effects of alcohol, if there is any such thing. In table 50 

 we give the variations of the several normal subjects from the average 

 of the group in the several kinds of measurements. 



A plus sign ( + ) before a value in table 50 shows that the individual's 

 performance in that test after alcohol varied in the same direction as 

 the average of the group, but more in amount. Conversely, a minus 

 sign ( ) shows that the individual's performance after alcohol showed 

 less change than the average of the group or that it was in the opposite 

 direction. Thus the signs of the values entered opposite Subject II 

 show that in all but two tests the effect of alcohol was greater on this 



