1880.] Strained Material, Lay den Jars, and Voltameters. 411 



phone, and attributed the effect to its true cause, viz., the expansion 

 of bodies under the influence of heat, which, in fact, is the explanation 

 of all microphone receivers. 



Ader reproduced speech by the vibrations of a wire conveying 

 currents of electricity, but he found that only magnetic metals were 

 effective, and therefore, like De la Rive, he attributed the result to 

 magnetic agencies. {Vide Count du Moncel, " Telegraphic Journal," 

 March 1, 1879.) 



These and many other sonorous effects of currents on wires may be 

 really due to such heat effects as I have described. 



IL " A Preliminary Account of the Reduction of Observations 

 on Strained Material, Ley den Jars, and Voltameters." By 

 John Perry and W. E. Ayrton. Communicated by Pro- 

 fessor G. G. Stokes, Sec. R.S. Received April 17, 1880. 



It has been shown by Dr. Hopkinson that, if two Leyden jars be 

 made of the same glass, but of different thicknesses : — 



1st. If they be charged with the same difference of potential for 

 «qual times, discharged for equal times, and then insulated, that the 

 residual charge will, after equal times, have in both cases the same 

 difference of potential. 



2nd. That residual charge is proportional to exciting charge. 

 These propositions may be included in one law — the superposition of 

 simultaneous forces is applicable to the phenomena of residual charge. 

 All the investigations in Dr. Hopkinson's paper in the "Transactions 

 ■of the Royal Society," vol. 167, part 2, serve to|prove this law, and, so 

 far, they support the theory of residual charge, which we owe to the late 

 Professor Clerk Maxwell. We should, therefore, be inclined to think 

 that one of the best methods of investigating the relation between the 

 relative powers of different glasses to possess residual charge would 

 be simply to charge jars made of these glasses for the same great 

 length of time, discharge for the same short intervals, and insulate, 

 measuring in each case the time increase of soaking out of the residual 

 charge. Whatever the thickness of the glass or the amount of the 

 original charge, we know that the same glass will always give the 

 same proportion of residual charge at the same times from insulation. 

 Any change of the state of the glass caused by heat would be shown 

 as a change on the curve of increasing residual charge. There 

 seems to be no doubt that this method would give what may be called 

 a measure of the specific power of producing residual charge pheno- 

 mena of the glass experimented on. 



In this way, since 1875, we have obtained a considerable number of 



