Profs. John Perry and W. E. Ayrton. [May 27, 



make tlie present paper too long to describe in full the slightly wavy 

 nature of the actual curves connecting v and t obtained in our experi- 

 ments. That the variations are mainly due to slight fluctuations in 

 the temperature of the room is seen from our finding a diurnal period 

 in the waves. Indeed, we have found that small changes of tempe- 

 rature very materially affect all the phenomena with which we deal 

 in this paper. But there is something more in these small waves than 

 can be explained by changes of temperature, and possibly the solu- 

 tion (2) of the differential equation may have some coefficients which 

 are sine and cosine functions of the time. A study of the constants 

 on Professor Maxwell's theory ought to tell whether 



x n + ax ll - l + . . . . +mx=0 



has any unreal roots. 



In the same way as that above described, we have obtained the ex- 

 ponentials of 



«=A + Be-«+C«r rf + (5), 



the solution of equation (1) when v is kept constant, from actual 

 observations of the current flowing into a condenser from a battery 

 with a constant electromotive force. The curve for u, just like the 

 curve described above, possesses a slight waviness, and when this is- 

 neglected its simple logarithmic nature subsequently to a certain 

 period of time is very striking. 



In former papers read before this and other Societies we have given 

 our reasons for believing that the phenomena of polarisation in volta- 

 meters, under the action of an electromotive force insufficient to pro- 

 duce disruptive discharge in the liquid, that is rapid and visible 

 decomposition, are of exactly the same nature as the phenomena 

 exhibited by condensers having glass or other solid dielectrics. And 

 we think that if our analogies fail for greater electromotive forces it is 

 because rupture of a solid dielectric introduces instantaneously entirely 

 different conditions. An examination of the following reductions, 

 which we give as a sample of the observations of current flowing into 

 and out of a voltameter containing ordinary water and charged with 

 one Minotto's cell (which has an electromotive force less than will 

 produce visible decomposition), will show that there is a striking 

 adherence to the logarithmic law, and that we here add another 

 analogy to those we have already given between condensers with 

 liquid and with solid dielectrics. 



In the following table t is the time, in minutes, after applying one 

 Minotto's cell, u is the current flowing through the voltameter, the 

 unit of u being 0*371 microfarad per second ; u' is the current calcu- 

 lated from the formula 



w'=180-l + 426-6e-°"°i773/ 



(6), 



