Eye Reaction. 61 



expected to produce a noticeable decrease. Particularly would this 

 be the case when so relatively few reactions were taken, that is, only 

 5 every period. The general decrease in reaction time, attributed to 

 practice for the Dodge and Benedict normal group, unfortunately rests 

 on fragmentary normal data for Subject VI. There was but one normal 

 day and it so happened that the records for the first period of this day 

 were illegible. There was also only one alcohol day with dose A and 

 the values obtained on this day were peculiar. Subject VI showed a 

 marked variation in his average eye-reaction time between the normal 

 day (October 22, 1913) with an average reaction time of 209 a and a 

 mean variation of 22 a, and that recorded on the alcohol day with dose 

 A one week later (October 29), with an average reaction time of 159 cr 

 and a mean variation of 24 cr, there being a difference of 50 cr between 

 the average reaction time on the two days. 1 The shortness of the reac- 

 tion time on the second day is astonishing, as averages such as those for 

 periods 2 and 3 (145 a and 147 cr, respectively) are extremely short and 

 unusual in data thus far collected. 2 No similar averages are found 

 among those obtained for the other four days of the Dodge and Bene- 

 dict series or in those obtained for the six days of the repetition series 

 on this same subject. Considering these peculiarities of the eye- 

 reaction data for the first series, it is not surprising that the later exper- 

 iments fail to agree with them. 



Careful investigation and criticism have failed to reveal any factors 

 which would seem to modify the effect of alcohol on the eye reaction as 

 shown in table 9. The average reaction time in all of the periods after 

 the alcohol dose, with the exception of period 6, is larger than the com- 

 parison values for the normal days, the differences between them rang- 

 ing from 9 to 23 a. The total average reaction times for the two groups 

 of data (see next to last column) are 208 cr for normal days and 227 a for 

 alcohol days. The effect of alcohol in terms of differences (see section 

 in) was to lengthen the reaction time 25 cr in period 2, 27 a in period 3, 

 20 o- in period 4, and 13 a in period 5. In period 6 the sign is reversed, 

 with a decrease in the reaction time amounting to 4 cr. The general 

 average shows an increase in reaction time of 22 cr as the effect of the 

 alcohol. The differences just referred to for periods 2, 3, and 4 and for 

 the average of all periods have, as shown in section in of table 9, in 

 each case a probable correctness above 0.900. The differences for 

 periods 5 and 6 can not be regarded as especially significant, since the 

 probable correctness figures are smaller. 3 In terms of per cent, periods 

 2 to 5 show increases of 12, 13, 9.5, and 6.2 per cent, respectively, while 



1 See Dodge and Benedict's report, p. 85, table 7, and p. 88, table 8. 



2 See Dodge and Benedict's report, p. 84, table 7, Subject III, January 26, 1914. The average 

 latency of eye reactions is given as 103 <r. This is a misprint; the figure should be 183 a. The 

 total average and the differences are figured on the basis of the latter value. 



3 This is due partly, at least, to the smaller differences in the last two periods and also to the fact 

 that here fewer reactions enter into the computations of the probable errors. 



