Inductive Capacity of Electrolytes. 195 



In numbers 1 and 2 the distance and potential are alike, 

 but the force is eighteen times greater without the plate than 

 with it. This cannot be explained by the simple fact that the 

 specific inductive capacity of mica is perhaps a dozen times 

 less than that of water, for that renders the equivalent thick- 

 ness of water but a few millimetres greater, and so reduces 

 the force by less than one half. On account of its compara- 

 tively large conductivity, however, the water on opposite sides 

 of the mica quickly becomes charged as shown * (fig. 3) , and 

 the entire fall of potential just between the electrodes occurs 

 in the mica. There is then no force on the needle in the 



dY 

 direction of x. —r- being zero. Around the edges of the 



3 dx rt ° 



mica plate, however, there is a fall of potential, and therefore 



dY 

 a force, -j- 9 acting in different directions at different points on 



the needle. In (2) Table VII. the resultant of these separate 

 forces is expressed by a deflexion of 0*84 centim. toward the 

 plate. In (3) the resultant toward the plate is zero, while in 

 (4), the plate being much nearer the needle, and therefore the 

 lines of force curved backward to a greater degree, the force 

 is apparently repulsive ; that is, the resultant is in the direc- 

 tion of — x. 



B. — If now the fixed electrode be completely insulated 

 instead of merely shielded by the mica, there is no appreciable 

 force on the ueedle for the same difference of potential and 

 rate of alternation used before. The 26-part wheel of the 

 rotating commutator was substituted and the speed increased 

 to 3000 per minute ; this gave 78,000 reversals per minute. 

 The potential was increased to 75 and 90 volts, and a new 

 apparatus arranged with two fixed electrodes and a larger 

 double needle, as shown in section in fig. 4. The sensibility 

 of this instrument was many times greater than the other, and 

 enabled me to observe the very small force exerted on the 

 needle, which it had been seen must exist when using in- 

 sulated fixed electrodes and alternating potentials. For, just 

 after each reversal there is a fall of potential in the water 

 between the needle and the insulated plate, and therefore a 



dY 

 force —r- acting on each unit of electricity of the charged 



. . dY . 



needle. This continues until -7- is reduced to zero by con- 



(XX 



duction, which is accomplished in much less than a millionth 



* The current is of course alternating-, and so these charges as indi- 

 cated in the figure are reversed perhaps fifty times per second. 



