128 Effect of Alcohol on Psycho-Physiological Functions. 



on non-alcohol days. Some indication of this reversal appeared to be 

 present in the former experiments 1 and one of the reasons for using 

 5-hour sessions in the repetition series in place of 3 hours as formerly 

 was to obtain data for a longer time following the ingestion of alcohol 

 which might illuminate this problem of possible facilitation succeeding 

 depression. For reasons which have been previously explained (see p. 

 29), it was practically impossible in the later experiments to have an 

 equal number of periods on each day. Hence it transpires that for the 

 sixth period the data are confined to one normal and two alcohol days. 

 The normal day was July 3, on which the subject tried less to carry out 

 the instructions in performing his part of the experiments. The gen- 

 eral relaxation and drowsiness of the subject are confirmed by the fact 

 that the pulse rate was decidedly the slowest on this day. This can be 

 seen in the " daily averages " at the bottom of table 20, p. 113, where the 

 general average pulse-cycle length for the day in question is 1.14 sec- 

 onds, equaling a rate of 52.6 per minute. On the other hand, the results 

 obtained on July 4, the last alcohol day, were considerably influenced 

 by a spurt of effort made by the subject. These conditions seriously 

 complicate the interpretation of the results for period 6 which might 

 otherwise appear as a post-alcohol facilitation. Obviously, however, 

 it would not be unfair, as the days were all 5 hours in length, to com- 

 pute percentile differences, using as a basis the last period of each day, 

 regardless of whether the periods were numbered 4, 5, or 6. The differ- 

 ences are already in the tables for the various measurements. It is 

 necessary to find the "last-period" average difference for normal and 

 alcohol days separately, to subtract the former from the latter, and to 

 divide the result by the average of all first-period scores. For illus- 

 tration, the percentile difference effect of alcohol in the last period of 

 eye reactions would be found as follows (see p. 60, table 9, section n) : 

 The normal differences for the last period are —23, +26, and +12, 

 average +5; the alcohol differences are +51, —16, and +32, average 

 +22; the alcohol average minus the normal average equals +17, which, 

 divided by the average of relevant first-period normals (211+208-^-2 

 = 209.5), equals 8.1 per cent; this is entered in the column headed " Last 

 period" in table 26. This percentile difference, +8.1, for the last 

 period of eye reactions shows continued depression to the end of the 

 experimental day. It would seem that the value — 1.9 in period 6 was 

 largely due to the fragmentary data for this period and not to a real 

 facilitation of the process. When the signs for all the 30 percentage 

 values of the last period are classified, it is found that half of them repre- 

 sent inferior and half superior performance, while in period 6 the pro- 



1 Dodge and Benedict's report, p. 2« r >s IT. 



2 This comparison is not strictly accurate, as there was no tetanus pulse for period 6. Since 

 there is a strong probability, judging by the other tetanus pulse data, that these three signs would 

 have been minus if present, the proportion might have stood 40, 57, and 3, which would be a little 

 less favorable to facilitation in period 6. 



