dependent on the Motion of the Conductor'. 105 



ceptible influence on the accuracy of the measurements of the 

 resistance. 



We will imagine that the polar disks in the outer side-tubes 

 c and e (fig. 1) are connected with the windings of the magne- 

 tometer in such wise that a galvanic current passes through 

 all the windings in the same direction. The middle side-tube 

 is closed, and the disk in it does no service in these experi- 

 ments. If now an electromotor which emits alternating cur- 

 rents of equal intensity and short duration be inserted between 

 c and/, it is evident that the magnetometer-needle must con- 

 tinue in its position of equilibrium if the water in the glass 

 tube is at rest, because the actions of the currents neutralize 

 one another ; but when the water is set in motion, then, because 

 this motion alters the resistances, that current which has the 

 same direction as the water is augmented, and the other dimi- 

 nished, and on this account the needle must make a deflection. 

 In the experiments which were made on the present occa- 

 sion, induction-currents were not employed, as in Kohlrausch 

 and Nippoldt's investigation, but an arrangement of the follow- 

 ing nature. A circular disk of boxwood 75 millims. in dia- 

 meter, and 20 millims. in thickness is covered on each side 

 with a brass disk. By aid of a winch and cog-wheel the disk 

 can be turned on its axis, making five turns during one revo- 

 lution of the winch-handle. Fig. 2 represents a portion of the 

 circumference of the disk. From the two brass 

 disks A and B little brass plates a, b, c, d, and e £' * 



stand out, all perfectly equal in width, and situ- 

 ated at equal distances one from the other. They 

 are let into the wood so that their surface coincides 

 with the periphery of the disk. Every other of 

 these brass plates is in connexion with the brass 

 disk A, and every other in connexion with B. The 

 number of these plates is 24, of which, therefore, 

 12 are in conducting-union with A, and the other 

 12 with B. The distance of the plates from one 

 another is equal to their width. By a suitable arrangement 

 the brass disk A was connected with one pole of an electro- 

 motor, and B with the other pole. On the periphery of the 

 boxwood disk, two brass springs, tipped with steel, insulated 

 from each other, slide in such a way that they are simulta- 

 neously in contact with two neighbouring brass plates. At 

 the same time, therefore, the one is in contact with the plate a, 

 and the other with plate b. One of these brass springs was 

 connected with the polar disk in the upper side-tube (fig. 1) ; 

 from the disk in the lower side-tube e a conducting-wire went 

 to one end of the windings of the magnetometer ; from the 

 Phil Mag. S. 5. Vol. 1. No. 2. Feb. 1876. I 



