196 Mr. E. B. Rosa on the Specific 



of a second. During the remainder of the interval before 

 the next reversal there is no force on the needle. 



With 90 volts the deflexion was only a few millimetres, 

 but the needle was perfectly steady, and readings were made 

 to a tenth of a millimetre. For steady or slowly reversed 

 potentials there was no perceptible deflexion, but above 

 10,000 reversals per minute the deflexion was perceptible, 

 and increased to a maximum of 4*5 millim. at 78,000 per 

 minute, the deflexions being sensibly proportional to the speed, 

 as would be expected. This is therefore a very different case 

 from that of uninsulated electrodes, where the force is many 

 thousand times greater and practically independent of the 

 speed. 



C. — The secondary of a Ruhmkorff coil was substituted for 

 the battery and commutator, and an alternating current sent 

 through the primary. Comparatively large deflexions were 

 observed with the same arrangement of insulated fixed elec- 

 trodes. This was due to the much higher potentials, only an 

 approximate estimate of whose value could be obtained. 

 Owing to the nearly instantaneous character of the current 

 from the Rubmkorff coil, a far larger proportion of the effect 

 observed with uninsulated electrodes would be observed with 

 insulated. 



D. — Since the small force on the needle observed with an 

 insulated fixed electrode is due to the conductivity of the 

 liquid, the same effect must occur, but in less degree, with 

 liquids of less conductivity. To show this, two fixed elec- 

 trodes, T and T', were fastened, one to the inside and the other 

 to the outside surface of the glass dish (fig. 5), and the latter 

 filled with turpentine. The fixed electrodes are thus at nearly 

 equal distances from the needle, and with rapidly alternating 

 potentials the attraction was practically the same whichever 

 was used. But for very slow alternations the force diminishes 

 slightly, and for steady potentials falls to zero, when using 

 the outer electrode, remaining practically constant for the 

 inner electrode. In other words, when using the outer elec- 

 trode and steady potential, the turpentine comes to a uniform 

 potential by conduction, and the 95 volts fall of potential 

 occurs entirely in the glass. 



5. Variation in Temperature. — In order to determine the 

 variation of the force with temperature, the apparatus (fig. 1) 

 was set into a large glass jar, which was then filled with water 

 up to the ebonite cover, e, or about a centimetre above the 

 level of the liquid in the glass dish. The mirror remained in 

 view above the top of the jar. Using hot water and ice as 

 needed, the temperature of the water-bath was varied between 



