110 PSYCHOLOGICAL EFFECTS OF ALCOHOL. 



as were regularly used by us for measuring the sensory threshold to 

 Faradic current. Two evaporating dishes about 6 cm. in diameter were 

 one-quarter filled with a saturated solution of zinc sulphate. Each 

 dish held an amalgamated zinc rod, through which the electrode was 

 connected with the wiring from the bridge, and a porous porcelain cup, 

 which was half filled with physiological salt solution, in which the 

 respective fingers were immersed. The Wheatstone bridge was the 

 same as that used in determining the skin-resistance for the Martin 

 measurements of Faradic threshold; but in the present case it was 

 operated by a constant current of 3 volts, instead of the alternating 

 current which must be used for skin-resistance measurements. In 

 place of the usual telephone receiver we connected the string gal- 

 vanometer. (See fig. 1.) 



The recording beam of light from the string galvanometer was 

 reflected at the eyepiece of the projection microscope at an angle of 90 

 to a millimeter scale which was attached to the side of the eye-reaction 

 camera. The string was loosened to a sensitivity of about 20 cm. per 

 0.001 volt. Its position on the scale was kept approximately constant 

 by balancing the Wheatstone bridge between the experiments. The 

 experimental movement of the string shadow resulted from a lack of 

 balance in the arms of the bridge, and showed at once the direction of 

 change and its amount. 100 mm. of scale was measured in terms of 

 millimeters of balanced bridge at the beginning and at the end of each 

 experimental period, so that the experimental changes could be reduced 

 to terms of resistance changes. 



Two circumstances greatly reduced the value of the resulting readings: 

 (1) Long immersion of the fingers in the fluid electrodes was found 

 almost to annihilate the phenomenon. It was consequently measured 

 only in the D-D' series (Kent-Rosanoff series). (2) In the predeter- 

 mined sequences of reactions, 6 per minute, it appears that there is not 

 sufficient time between experiments for a return of the psycho-galvanic 

 equilibrium. At any rate, in our experiments the resistance changes 

 seemed cumulative. For some cause the apparent resistance at the 

 end of a series was regularly different from that at the beginning. 

 These circumstances make it doubtful if our measurements of the 

 psycho-galvanic reflex are of any real significance. 



APPARATUS FOR RECORDING THE ASSOCIATION TIME. 



The arrangements for recording the latent time of the responses and 

 the synchronous pulse- waves were somewhat complex. It will be 

 remembered that both subject and operator occupied the balcony of 

 the research room. There was no apparatus on the balcony except the 

 tambour and the mercury-cup devices to transform the mechanical 

 pulse and respiration waves into electric impulses. All graphic records 

 were taken on the Blix-Sandstrom kymograph on the floor below. It 



