of Edinburgh, Session 1885-86. 
651 
curve, and on the curve representing the results of observation for 
the same value of the time. Hence 
e~ ct 
For the curve of fig. 2 we get the equation 
j = -116 - * 1 0 1 g - * 218i + * 13l£ ” ' 53< - 
This curve is practically indistinguishable from that of fig. 2. The 
sign of d will determine whether the logarithmic curve lies above the 
actual curve or not. But all the curves I have obtained lie above the 
logarithmic curve with the exception of that in fig. 1 ; so that, especially 
since the curves 
in that figure 
are apparently 
again approach- 
ing each other, 
I believe that 
this apparent 
exception is 
really due to an 
error in draw- 
ing the curve 
through the ob- 
served points 
where t is large, 
which would 
affect the con- 
stants. 
It would seem, then, that the polarisation current can be expressed 
as a function of the time by means of a formula of the form 
j = a- be~ ct + ds ~ et ) 
where d is always positive. 
My thanks are due to Messrs Moffat and Callender, students in 
the University Physical Laboratory, who have conducted the ex- 
periments for me. 
