84 Effect of Alcohol on Psycho-Physiological Functions. 



FINGER MOVEMENTS. 



Although it can not be claimed that the motor activity involved in 

 this test is a customary or practice act for every one, yet it seems 

 a serviceable measurement. Finger movements, as in the ordinary 

 tapping test, are used widely as a measure of motor ability. The test 

 employed in these experiments was very simple to perform; including 

 all adjustments it occupied scarcely 5 minutes of the experimental 

 cycle, and 9 seconds of movement was not sufficiently long to make it 

 at all objectionable to the subject. Neither the connection with the 

 string galvanometer for records of the pulse nor the simple arrangement 

 for obtaining respiration curves distracted or annoyed the subject in 

 any way, while at the same time the simultaneous recording of these 

 processes supplied valuable contributory data. The homogeneity and 

 regularity of the records later presented must be the real proof of the 

 serviceability of this measurement. 



The photographic record, which included the finger movements, 

 pulse, and respiration curves, were without exception legible. For 

 typical records of the finger oscillation, see Dodge and Benedict's report, 

 figure 28, opposite page 171, and this monograph, figure 11, E and F, 

 page 96. In elaborating, the record was divided from the beginning 

 of finger-movements into blocks of 2" each. The number of complete 

 oscillations in the first 2" was recorded, then the number of com- 

 plete oscillations in the first 4", likewise for 6", and finally for 8". 

 There were two records in each period which were, as outlined in our 

 experimental cycle, separated by a rest interval of 1 minute; these two 

 records were considered individually and the count for each record is 

 entered separately in table 15, which gives the results of the measure- 

 ment. While this method of presentation has, of course, doubled the 

 size of the table and perhaps makes it somewhat more difficult for the 

 reader to comprehend the results, the detailed presentation proves that 

 the first record in each period shows the better performance. 



The table is arranged by periods and in two sections, section i giving 

 the experimental data and section n the differences. As an example of 

 the way in which the data for this particular observation are entered in 

 the table, we will consider period 1 of the first normal day, June 29, 

 1914. From the first figure in the left-hand column of section i of 

 table 15 it is seen that during the first 2 seconds 10.2 complete oscilla- 

 tions were performed. Lower in that column we find 20.0 oscillations 

 for 4 seconds, these including, of course, the number of oscillations 

 made in the first 2 seconds. Farther down in the column we read, 



where movements both to the right and the left were legible or included, hence the values are not 

 the sum of right and left averages. This is statistically correct if canied out consistently, but it 

 perhaps exposes tables 26 and 27 to the unfounded criticism of containing several large errors. 

 Apparently, it would not have been unfair to have taken the sum of the two columns referred to, 

 for with the normal group there are ten omissions of both the duration right and the duration 

 left values, because there were no corresponding values to include. 



