in ilu> Charging of a Condenser* 



391 



same point and allowing it to swing freely through a definite 

 arc. The diameter of the smaller sphere used in the above 

 determination was 2*1G centim. The data for calculating the 



Diagram 1. (See Table I.) 



10X 

 Time of Charging. 



20X 



actual time of contact not being satisfactory it has seemed 

 best to express the time during which the condenser was 

 receiving a charge in terms of the time of a single contact. 



An investigation of the cause of this difference in the 

 results obtained with the two different cells showed that it 

 was at least largely due to the difference in their internal 

 resistances. Upon investigation it was found that the internal 

 resistance of cell No. 489 was 20 ohms, and that of cell 491 

 396 ohms. 



As a result of these observations, in all later experiments 

 the battery used in charging the condensers consisted of five 

 accumulators joined in multiple arc. The internal resistance 

 of each accumulator was about five thousandths of an ohm, 

 and consequently the internal resistance of the battery about 

 one thousandth of an ohm. The connecting wires were made 

 as short as possible, and of No. 11 B.W.G. copper wire, except 

 the suspension of the smaller sphere, which, when not other- 

 wise stated, consisted of a piece of soft silver wire 0*05 millim. 

 in diameter and 6 centim. in length. 



As soon as these changes were made in the battery used 

 and in the connexions it w^as found, as will be seen from the 

 results given further on (see Tables II.-X.), that the charge 

 given to the condenser during the time the spheres were in 

 contact was no longer constant. On the contrary, it varied 

 between wide limits, and frequently was much greater than 

 what might be called the normal charge, that is the constant 



2 E 2 



