45 2 MEASUREMENT OF CAPACITY. DIELECTRICS. 



depends not merely on the duration and strength of the primitive 

 charge, but also on the previous charges it has received. 



If a Leyden jar, for instance, has had first a positive charge for 

 several weeks, then a negative charge for twenty-four hours, and a 

 new positive charge for five minutes, the residue may give oscil- 

 lations of potential which are alternately positive and negative.* 

 Everything happens as if electricity gradually penetrated into the 

 dielectric, to become dissipated by a displacement in the contrary 

 direction, when the armatures are brought to the same potential. 



These residual charges have been observed since the invention 

 of the Leyden jar. The observations of M. Gaugain,f in particular, 

 have shown that 



i st. The charge of a condenser for a given difference of potential 

 increases with the duration of the connection with the source. This 

 charge is only definite for a pretty short contact ; it is then said to be 

 instantaneous. 



2nd. The instantaneous discharge that is to say, that which 

 corresponds to a connection of the armatures for less than two 

 seconds is virtually equal to the instantaneous charge. 



Apart from losses by conduction, the residual charge is virtually 

 equal to the excess of the total charge over thie instantaneous dis- 

 charge. 



ClausiusJ and Maxwell have endeavoured to ascribe the phe- 

 nomena of electric absorption to the heterogeneity of the dielectric. 

 Their theories agree in showing that there would be no absorption in 

 a perfectly homogeneous medium, and the fact has been verified by 

 Professor Rowland for Iceland spar, the natural substance which 

 suggests itself as purest and most homogeneous. Other natural 

 crystals gave greater or less residues ; the absorption of quartz was 

 found to be the ninth of that of glass. || 



When quantities of electricity are to be exactly measured, air- 

 condensers are the only ones which offer complete certainty, provided 

 that no dust gets between the two plates. Unfortunately, they have 

 only a small capacity. 



1043. STANDARDS OF CAPACITY. In the electrostatic system 

 (607) the capacity of a conductor is a length, and it might be 



* SIR W. THOMPSON. Congres Intern, des Electriciens. Paris, 1881, p. 217. 



t GAUGAIN. Ann. de Chim. et de Phys. [4], Vol. II., p. 264. 1864. 



CLAUSIUS. Pogg. Ann., Vol. LXXXVI., p. 337. 1882. 



MAXWELL. Electricity and Magnetism, Vol. I., p. 376. 



|| ROWLAND and NICHOLS. Phil. Mag. [5], Vol. XL, p. 414. 1881. 



