476 M. 11. Kohlrausch's Theory of the 



tions as applied to any particular problem (as has been done with 

 reference to the problem last discussed), we should find them to 

 be such as do not strike us as particularly applicable, as was the 

 case with that problem. If, being in ignorance what system of 

 assumptions ought to be made to render the problem determi- 

 nate, we were to wish to give a definite answer to the problem, 

 it might be in the following form : ascertain the chance of the 

 requh-ed event happening on any one system of assumptions, 

 and the chance of that system representing the true connexion 

 among the simple events, and multiply the values of these 

 chances together ; the sum of a series of these products compri- 

 sing every possible system of assumptions would be the true 

 chance of the event. But Professor Boole's method evidently 

 does not attempt to solve any question of this nature. It would 

 seem that though Professor Boole gives a uniform and emi- 

 nently elegant method of solving a class of cases of such in- 

 determinate problems, that class is not one of much practical 

 application. 



LXXV. Theory of the Electric Residue in the Leyden Jar. 



By R. KoHLRAUSCH. 

 [Concluded from p. 426.] 



§11- 



WE will now endeavour to obtain an equation for the residue- 

 curve by help of the principles stated in the foregoing 

 paragraph. 



The charge Qq being suddenly imparted to the jar, generates 

 an electric moment m, which increases with the time, and, in order 

 to re-establish an equilibrium between the action of the charge 

 Qq and a contrary action which has been elicited in the glass, 

 approaches a certain limit M, proportional to Q^. In order to 

 bring this electric moment into calculation, we must select some 

 unit by which to measure it. Let the unit of moment be that 

 which can detain a residue equal to unity ; the latter unit being 

 a certain quantity of electricity, indeed the same quantity ac- 

 cording to which Qo is measured. As the moment, however, 

 may be assumed proportional to the residue, instead of the former 

 we shall substitute the latter, which is its effect, and say the 

 primitive charge Qq has produced the residue r^ in the time t, 

 which, in order to restore the equilibrium, must increase to R, 

 so that then 



R=;>Qo, 

 where ^) is a constant. 



If, however, the primitive charge Qq continually suffers a loss 



