dependent on the Motion of the Conductor. 95 



between the two outer disks, and the alteration of polarization 

 thereon depending, had no effect on the direction of the deflec- 

 tion. If, as in the first experiment, the liquid moved from a 

 to b, and the negative instead of the positive pole of the elec- 

 tromotor was connected with the disk in the middle side-tube, 

 the deflection obtained indicated that the part of the current 

 which passed through the liquid column c d (and therefore had 

 the same direction as the flow of the liquid) possessed a greater 

 intensity than the other. In the first two experiments, there- 

 fore, the more powerful part of the galvanic current went be- 

 tween the middle polar disk and that one of the two outer disks 

 which was exposed to the less pressure from the flowing liquid ; 

 but in the last experiment the more powerful current passed 

 between the middle disk and that one of the two outer disks 

 which received the greater pressure. The difference in the 

 pressures to which the outer disks were exposed during the flow 

 of the liquid cannot, then, have exerted any influence on the 

 direction of the deflection. On the other hand, it appeared in 

 all the experiments that that part of the current which had the 

 same direction as the flow of the liquid was always the intenser ; 

 and for this result no other reason can be given than that the 

 galvanic resistance becomes less when the current moves in the 

 same direction as the liquid, and greater when the directions 

 are opposite. 



In order that the deflections may turn out regular, the liquid 

 employed must exert a great resistance. The importance, 

 in this respect, of the resistance will be seen from the follow- 

 ing calculation. 



We will denote by r the sum of the principal resistances in 

 the liquid column c d, in the conducting-wire c/, and in one 

 position of the wire of the magnetometer, and, in a correspond- 

 ing manner, by r + p the sum of the resistances in the liquid 

 column d e, in the conducting-wire e n, and in the other posi- 

 tion of the magnetometer-wire. Because these sums have 

 been made as nearly as possible equal to one another, p must 

 be a very small quantity. If we now assume that the negative 

 pole of the electromotor is connected with the gold disk in the 

 middle side-tube, and that the liquid employed is water, this 

 disk covers itself with hydrogen, the electromotive force of 

 whose polarization may be called w. The two outer disks, on 

 the contrary, become coated with oxygen ; and the electromo- 

 tive force of its polarization in the side-tube c we will denote 

 by s, and in the side-tube e by s + <r. As the polarization is 

 all but perfectly equal in both these disks, the value of cr must 

 be inconsiderable. Let i denote that part of the current which 

 flows through the liquid column cd, and i\ that part which 



