192 Mr. H. Pender on the Magnetic 



three coaxial cylinders of soft iron, closed at the bottom and 

 top by soft-iron plates. With this shield on, the needle was 

 fairly steady during the period above mentioned. The acci- 

 dental deflexions of the spot of light with the disk uncharged 

 rarely amounted to over 5 mm. during the time required for 

 a single determination of a deflexion ; i. e. about one minute, 

 though there was at times a considerable shift of the zero. 



The expression " sensibility of the galvanometer " will be 

 frequently used in what follows. By it is meant the deflexion 

 in millimetres on the scale 1*1 metres distant from needle, 

 reduced to an infinitesimal arc, produced by a current of 

 2 x 10 -8 amperes flowing through the galvanometer. This 

 value of the current was used simply because it gave a 

 convenient deflexion. The scale remained fixed with reference 

 to the galvanometer throughout the experiment. To obtain 

 a constant testing current a Clark ceil was connected in 

 series with 101,648 ohms. Around lb'48 ohms of this was 

 a shunt circuit with a resistance of 10,000 ohms. Around 

 101 ohms of this a second shunt circuit containing the gal- 

 vanometer was formed, the total resistance of this circuit 

 being 10,000 ohms. This arrangement gives a current of 

 2 x 10 — 8 amperes through the galvanometer. A simple 

 mercury key served to throw the galvanometer in or out of 

 this circuit. 



The Reverter and Commutator.- — The reverser was similar 

 in construction to an ordinary rotary interrupter. Connected 

 to the shaft of the reverser was an eccentric 1 mm. off centre, 

 which operated a commutating device in the circuit between 

 the coil and galvanometer. To the eccentric was attached a 

 lever carrying at one end a small silver rod 1*2 cm. long. 

 When the shaft of the interrupter was turned this rod was 

 given a slight up and down motion between two flat silver 

 springs, between which also was fixed a second silver rod. 

 When the moving rod was at its highest point it made contact 

 with the upper spring, the fixed rod making contact with the 

 lower spring. When the moving rod was at its lowest point, 

 the contacts were reversed. This special form of commutator 

 was found necessary in order to avoid thermo-electric currents. 

 The two rods were connected to the terminals of the coil, the 

 two springs to the galvanometer terminals. The commu- 

 tator was entirely inclosed in a metal shield connected to 

 earth. 



The Charging Apparatus. — The disks were charged by 

 means of a large Yoss machine — diameter of revolving plate 

 50 cm. — and a battery of six gallon leyden-jars. Each 

 pole of the Voss machine was connected to the inside 



