520 PROCEEDINGS OF THE AMERICAN ACADEMY. 



the residual charge will be liberated at precisely the rate which char- 

 acterized the residual forming current on its entrance into the condenser 

 during the long- continued charge. No experiments have been made 

 in the present work in which the rate of liberation of residual charge 

 was observed, but the law, if closely tested, will probably be found veri- 

 fied fairly well, and, if this is so, we may conclude that the so-called 

 " free charge " of condensers such as glass and paraffined paper contains 

 an appreciable quantity of very mobile residual charge. 



Many investigators have noticed that the capacities of most con- 

 densers vary considerably with the frequency of the alternating current, 

 when determined by one of the bridge methods, the capacities invari- 

 ably decreasing as the frequency is increased to high values. Now 

 if the results of the present research can be applied to chargings by 

 means of an alternating electromotive force, and we see no reason why 

 they should not apply, then it follows that the variation in the capacity 

 of a condenser is not primarily due to the increased frequency, or de- 

 creased period, but to the decreased charging interval, or time of con- 

 tact of the vibrating tongue with the condenser terminal. In fact, it 

 seems that the measured capacity should increase with increasing fre- 

 quency of alternation, provided the contact time of the vibrating tongue 

 is made longer at the same time. Of course this condition can be 

 realized for a certain range of frequency only. 



The fact that a considerable part of the residual charge is very mo- 

 bile is well illustrated by some observations on one of the condensers 

 made of pure paraffin sheets. As shown by the results tabulated above, 

 no satisfactory evidence was obtained of a measureable quantity of re- 

 sidual charge formed in such condensers within 0.0017 of a second after 

 the beginning of charging. When this condenser was charged for two 

 minutes, it was found to have formed 0.7 per cent of residual charge, 

 as measured by the air condenser neutralization method, in which no 

 residual charge is lost. But when the same condenser had been 

 charged for many minutes and then discharged by momentary short 

 circuit, only 0.1 per cent of residual charge was obtained, all the rest 

 having apparently disappeared along with the main discharge. Yet 

 this momentary short circuit forms an essential feature of the experi- 

 ments carried out by many investigators, who have studied, by means 

 of the quadrant electrometer, the reappearance of residual charge after 

 a momentary short circuit. 



^ As to the cause of residual charge, the results of the work cannot 

 give much information. It seems likely, however, that air bubbles in 

 the dielectric medium play a very important role in absorption of 

 charge. I hope to be able to carry on further investigations with even 



