410 Messrs. J. J. Thomson and H. F. Newall. [June 16, 



wide to give a good average. I donbt if such variations can be 

 attributed wholly to experimental error ; but on the other hand, it is 

 difficult to imagine that potassium should have more than one 

 dispersion equivalent, while in the same series of dissolved salts it 

 has apparently one and the same refraction equivalent. I am more 

 disposed to believe, that the uncertainty lies in the value of the 

 radicles to which the metal is joined; but this will require a more 

 extended research. 



It is also an important enquiry : — To what extent does the modifica- 

 tion of the dispersion equivalent affect the refraction equivalent for 

 the line A ? On this question, and others of a similar nature, I hope 

 shortly to submit a further communication. I think it will be already 

 sufficiently obvious that the specific dispersive energy of a compound 

 body is a physical property analogous to, but distinct from, its specific 

 refractive energy, and that it is capable in like manner of throwing 

 light upon chemical structure. 



XXL " On the Rate at which Electricity leaks through Liquids 

 which are Bad Conductors of Electricity." By J. J. 

 Thomson, M.A., F.E.S., Fellow of Trinity College, and 

 Cavendish Professor of Experimental Physics in the 

 University of Cambridge, and H. F. Newall, M.A., 

 Assistant Demonstrator in Physics, Cambridge. Received 

 May 26, 1887. 



The experiments here described were undertaken to test whether 

 the rate at which electricity leaks through a liquid which conducts 

 electricity badly, does or does not follow Ohm's law. 



The method used is described later on ; it consists in establishing 

 by a battery a difference of potential of about 100 volts between the 

 plates" of a condenser, in which the dielectric is the faulty insulator to 

 "be experimented on, then disconnecting the battery, and measuring 

 with an electrometer the rate at which the difference of potential 

 dies away. 



Let v 1 and v 2 be the differences of potential at the beginning and end 

 of an interval T, and let 



If c be the capacity of the condenser, q the quantity of electricity 

 which has leaked away in the time T, then 



q 

 c 



