100 PSYCHOLOGICAL EFFECTS OF ALCOHOL. 



subjects were shown each word separately before the first day's experi- 

 ments. The psychopathic subjects were shown each word separately 

 at the beginning of each day's experiments. 



Exposure. The words were exposed by the exposure apparatus in 

 chance order at intervals of 10 seconds. They were changed by the 

 operator by hand. 



Operation. With the subject in position, the time and reaction 

 markers properly adjusted to the drum, and, the Blix-Sandstrom kymo- 

 graph running at the rate of 100 mm. per second, one of the stimulus 

 cards was selected at random and inserted by the operator in the 

 exposure apparatus. From 2 to 2.5 seconds before the exposure the 

 operator touched the kymograph lever to change the circular movement 

 of the drum to a spiral, and withdrew his hand from the apparatus 

 as a signal for attention. As the drum continued to revolve, an off- 

 set on the kymograph shaft engaged a circuit breaker, with which 

 the voice key, the electrical marker, and the electric magnets of the 

 exposure apparatus were in series. The consequent movement of the 

 marker indicates the beginning of the movement of the exposure appa- 

 ratus which eventuates in the exposure of the word stimulus. As was 

 previously explained, this movement of the marker is not coincident 

 with the exposure; the latter followed after 37 a. While this discrepancy 

 between the registered and the actual beginning of exposure is theo- 

 retically inexpedient, it can not affect comparative values as we have 

 shown. Absolute values for the reaction time can be obtained by 

 deducting the latencies of the exposure apparatus (37 cr). 



The variation of this latency as is indicated above is considerably less 

 than half the unit of measurement. As the drum moves on, the circuit- 



FIGS. 17 to 20.-Records of the latency of the voice key. 



Figure 17 is a record of the sound of "sh," recorded by three methods. Two records were 

 produced respectively by a Harvard Apparatus Company marker and a Deprez signal. Both 

 were in series with each other and with the sound key. The middle record was made by a Cam- 

 bridge string galvanometer (sensitivity, 3 cm. per 0.001 volt) in series with a telephone receiver 

 which is pressed against the throat over the thyroid cartilage by an elastic band. This record 

 shows an exceedingly small difference between the various forms of recording devices. None of 

 the records shows a relative delay of more than 0.002". Of the three the string-galvanometer 

 curve naturally shows the most details. 



Figure 18 is a record of the word "cake," recorded similarly as the above sound "ah." The 

 various sounds of the word appear plainly in the record of the string galvanometer movement. 

 The vowel is especially conspicuous. Almost identical time relations exist between the various 

 lines in this record as in figure 17, i. e., the initial C is recorded by our voice key with as little error 

 as the open vowels. 



Figure 1 9 is a similar record of the word ' ' yolk. ' ' The character of the vowel is notably changed. 

 The initial "y" and the final "k" are obvious in the galvanometer record. The relative latencies 

 do not change. 



Another record of the word "cake," using the string galvanometer as before, is reproduced in 

 figure 20. But instead of actuating the galvanometer from the throat, in this record the telephone 

 receiver was placed at the side of our voice key. The latency appears not to be materially modi- 

 fied by this process, i. e., the difference between the throat-movement and actual vocalization in 

 the sound C is negligible, but the record contains some details which are not found in figure 18, 

 namely, at the beginning of the record, the initial C of "cake" appears in the string galvanometer 

 record of figure 20 as a high-pitched tone. This corresponds with the fact that the pitch of C 

 is not determined at the vocal cords, but at the front of the mouth. 



