Intelligence and Miscellaneous Articles. 495 



From Wiedemann's experiments, needles in which the magnetism 

 could be increased or reversed by striking must have been previously 

 magnetized in the opposite direction ; only one discharge of the bat- 

 tery had magnetized the needles, and therefore these discharges must 

 have taken place in opposite directions. Hence it appears to me 

 justifiable to* regard the occurrence of abnormal magnetism as a sign 

 of the alternating direction of the discharge of the Leyden jar. — 

 Monatsbericht der Akademie zu Berlin, February 1862. 



NOTE ON THE LIMIT OF THE CHARGE OF CONDENSERS. 



BY M. GATJGAIN. 



The charge of any given condenser placed in connexion with the 

 ground depends on the greater or less time during which the action 

 of the source is continued ; at least this is the case when the dielec- 

 tric is a solid body. Hence in the case of solid dielectrics, the ex- 

 pression representing the quantity of electricity accumulated on the 

 armatures of a condenser ought to be a function of the time during 

 w T hich the condenser is placed in connexion with the source. But 

 although, theoretically speaking, the state of equilibrium ought only 

 to be established after the lapse of an infinite time, in point of fact 

 the increase of the charge becomes inappreciable after a certain 

 number of hours, which is never very considerable when the dielectric 

 is not very thick ; hence it might be proposed to determine experi- 

 mentally the value of the greatest charge which a given condenser 

 could take when the action of the electrical source is infinitely pro- 

 longed. The determination of this limiting charge presents no par- 

 ticular difficulty ; it only requires considerable time, seeing that a 

 single experiment may last several days. 



I generally worked with small insulating discs with plane surfaces, 

 and with plane metallic armatures applied directly on the solid 

 dielectric. 



I first endeavoured to ascertain the relation between the limiting 

 charge and the thickness of the dielectric, and I found that this 

 thickness is quite indifferent. This result appeared to be opposed to 

 that obtained when the air is taken for a dielectric ; in this case I 

 have already shown that the charge is almost inversely as the thick- 

 ness of the layer of air, at least when this thickness is very small in 

 reference to the surface of the armatures. But it is not difficult to 

 explain this divergence, from considerations which I have presented 

 in a previous note. In fact, an air condenser is a simple condenser, 

 while a condenser with a solid dielectric is really a battery b)' cas- 

 cade. The imaginary solid dielectric (sulphur, shell-lac, or gutta 

 percha) is a true conductor. The dielectric is really the air which 

 separates the insulating disc from its armatures. Regarding matters 

 in this light, it will be understood that the thickness of the insula- 

 ting disc between the armatures ought not to influence the limiting 

 charge, but only the time necessary to obtain equilibrium ; and in 

 fact, to obtain a limiting charge, the time required is longer as the 

 insulating disc is thicker. 



Secondly, I compared fulminating panes of the same dimensions, 

 formed of different dielectrics (sulphur, shell-lac, &c), and I found 

 that the limiting charge was the same for all bodies ; so that it 



