90 Effect of Alcohol on Psycho-Physiological Functions. 



trast, the average for the alcohol curves shows period 2 as clearly lower 

 than period 1, while 3 and 4 tend to a progressive rise, but not quite to 

 the level of period 1. The curve drops somewhat at period 5, particu- 

 larly in the third and fourth 2-second intervals, but period 6 shows 

 decidedly the best average performance for the alcohol days, and 

 reaches a level which is not equaled by any period in the average for 

 the non-alcohol days. All of the alcohol curves show a drop between 

 periods 1 and 2, but this is much more prominent in the first 2-second 

 interval, indicating that the alcohol had a particularly depressant 

 effect upon the beginning of action. 



The differences between the periods before and after the control mix- 

 ture or the alcohol are given in section n of table 15. Computation 

 has been made for records of 2, 4, 6, and 8 seconds, as in the presenta- 

 tion of data by Dodge and Benedict. An examination of these values 

 shows that following the control dose the differences are quite regularly 

 positive, while after the alcohol dose they are more generally negative. 

 The course of the normal differences is shown by figure 10 1 (solid dots). 

 As would be expected from A in figure 9, the differences are almost zero 

 in period 2, as the performance in this period was almost identical with 

 that in period 1 in the normal curves. In period 3 they are markedly 

 plus, becoming continuously smaller but without change of sign in 

 periods 4 and 5 and with a return to practically zero in period 6. The 

 course of the alcohol differences (fig. 10, open dots) are in sharp contrast 

 with the normal series. Period 2 is far below the base line; periods 3 

 and 4 show a relatively large rise over period 2, but still remain minus 

 and are decidedly different from periods 3 and 4 on the non-alcohol 

 days; period 5 is distinctly minus for the 6 and 8 second curves; period 

 6 rose far above the line. 



While the effect of alcohol is evident in these curves to give it a con- 

 crete numerical statement requires an expression for the differences 

 between the course of the two groups of curves, i. e., a statement of the 

 true difference between the results obtained on the non-alcohol and 

 alcohol days. This has been computed for the full 8-second performance 

 and is presented in table 16. The differences between the values before 

 and after the control dose or the alcohol dose are obtained in the usual 

 manner from section n of table 15, these differences for the respective 

 periods being —1.46, —1.50, -1.15, —1.30, and +2.65 complete 

 oscillations. The percentile differences are obtained by dividing these 

 differences by the average of the relevant normals, i. e., the average of 

 the two average values for period 1 in the normal series (37.28) and in 

 the alcohol group (37.70). Using the average (37.5), we find that the 

 alcohol depression amounts to about 4 per cent in periods 2 and 3 and 



1 Figure 10 has been drawn for 2, 4, 6, and 8 second differences, i. e., the 8-second differences in 

 a way include those for 6, 4, and 2 seconds, and the 6 seconds include the 4 seconds and 2 seconds, 

 etc. In A and B of figure 9, on the contrary, each curve is for the performance of a separate 2- 

 second interval. 



