SUMMARIES AND CORRELATIONS. 261 



The most consistent effect of repetition, though not the least, appears 

 in the case of the eye-movements. Its direction, however, is reversed. 

 That is, instead of being shortened by repetition, as one might expect, 

 the average duration of the eye-movements is increased 7 per cent in 

 both normal and psychopathic subjects from the first to the last normal 

 day. There are no published data which would have led us to expect 

 this apparent reversal of practice. Its probable explanation is to be 

 found in the increased accuracy of the eye-movements of 40 after 

 practice in looking from one fixation-point to the other. Since the 

 errors of fixation are due in general to an underestimation of the 

 distance, and are commonly corrected by positive corrective move- 

 ments in the same direction as the main eye-movement, the practice 

 that results in increased accuracy of movement would produce longer 

 arcs of movements. But longer arcs of eye-movement regularly take 

 longer times. That something of this sort actually did take place is 

 indicated by the record of decreased length in the corrective movements 

 which is found in the full tables. Making allowance for these changes 

 in the arc of movement it appears that the actual angle velocity of the 

 eye-movements is practically free from practice effect during our 

 experiments. 



The effect of repetition on the eye-reactions is relatively large. This 

 appears not only in a comparison of the first and last normal days, but 

 also in the relation between normal and alcohol days. If one notes the 

 following summary of eye-reaction averages, it will be seen that the 

 gradual decrease of latency appears quite independent of the alcohol 

 dose. 



Normal I. Dose A. Dose B. Normal II. 

 Av 0.216" 0.206" 0.201" 0.193" 



These results were unexpected. Previous experiments with the ocu- 

 lar reactions led us to expect no effect of repetition. Theoretically it 

 seemed to us that the long training of practical life would make the 

 relatively small number of laboratory cases entirely insignificant. The 

 cause of the discrepancy between our expectation and the results in 

 this case is probably to be found, as we have already noted, in the small 

 number of positions for the eccentric visual stimuli to which the eyes 

 moved. The effect of repetition in this case constitutes probable cause 

 for an effort to improve the technique. If it is finally found to be 

 expedient, it should not be difficult to arrange apparatus so that each 

 peripheral stimulus should occupy a different position in the field of 

 vision. The reduction of the number of positions to 6 in the present 

 experiments was made to gain uniformity of succeeding series. It was 

 not necessary and possibly it was not expedient. Given the effect of 

 repetition, it is not surprising that the first alcohol day should show a 

 marked improvement in the reaction-time after the first series. Such 

 an improvement would follow the general rules of habit formation. 

 In our case it would work directly opposed to such a depressing influ- 



