138 PSYCHOLOGICAL EFFECTS OF ALCOHOL. 



measuring the skin-resistance were carried to the galvanometer table, 

 whence they might be switched to the string galvanometer for optical 

 measurements, or to a suitable high-resistance watch-case telephone 

 receiver for the acoustic method. The electrical connections for the 

 system are diagrammatically represented by the fine lines with appro- 

 priate legends. 



Position of the subject. Two positions were occupied by the subject 

 for sensory-threshold measurements by the Martin method. In the 

 earlier measurements and in the 12-hour experiments, the subject 

 reclined in a comfortable chair in position I, figure 1. In the measure- 

 ments which form the bulk of the experiments here reported, that is, 

 in all measurements which were made on Subjects II to X subsequent 

 to January 3, 1914, the subject reclined in a steamer-chair on the 

 balcony of the laboratory. In the former case, threshold measure- 

 ments were a part of Group I of the experimental series. In the latter 

 case, threshold measurements alternated with association experiments. 

 In every respect the balcony position corresponded more closely with 

 the conditions that are recommended by Martin. In this position the 

 subject faced a blank wall and responded to the stimulation by signaling 

 with a telegraph key. When position I was used and the subject sat 

 in the same room with the apparatus, the inductorium and its connec- 

 tions were hidden from view by other apparatus. In this case the 

 subject indicated a perceptible stimulation by saying "now" or "yes." 

 It is doubtless always more or less unsatisfactory to have the subject 

 in the same room with the apparatus, even when the utmost precau- 

 tions are taken to prevent his hearing the key or seeing the movements 

 of the secondary coil. If the threshold work alone was under consid- 

 eration the ideal condition of isolating the subject could be rigidly 

 enforced. When a series of measurements was undertaken, such as 

 ours was, such isolation becomes more difficult. Periodic movement 

 of the subject from one room to another would have been indefensible 

 in our case. Nevertheless, it seems probable that with increasing 

 definiteness of the various controls in this type of experimentation 

 isolation of the subject from the apparatus, both for the threshold 

 measurements and for other psychological experiments, must not be 

 neglected. 



Electrodes. In all our threshold experiments zinc sulphate non- 

 polarizable electrodes were used. Amalgamated zinc electrodes were 

 immersed in concentrated sulphate of zinc. The fingers were placed 

 in a porous porcelain inner vessel in which there was a physiological 

 salt solution. Martin reported that the value of /3 was not changed by 

 changes in the amount of the finger immersion. Assuming on this 

 ground that it made no difference, we found it more convenient to have 

 only the first joint of the finger immersed. 



Primary current. For sensory threshold experiments we universally 

 used a primary current of 0.5 ampere taken from two accumulators of 



