662 Dr. S. R. Milner on the 



V-f- CRab by assuming R AB = 59 co, which makes the different 

 portions of the curve join on well to each other *. The second 

 and fifth portions of the curve are very interesting. The 

 sudden reduction in the current which has taken place causes 

 the polarization to tend to diminish, but at a continually 

 decreasing rate. Meanwhile the current still passing is all 

 the time tending to increase the polarization, and the com- 

 bination of the two effects gives rise to a minimum value of 

 the polarization which is weil marked, both theoretically and 

 experimentally. It is striking that the extremely complicated 

 state of distribution of the silver nitrate near the surface of 

 the anode produced by so complex a cycle of changes should 

 reduce at the surface itself to a state expressible in such simple 

 terms as those of (25). 



Mercurous-Nitrate Cells and Continuously Varying Current. 



When the polarizing current undergoes a continuous 

 variation instead of a series of sudden ones, formula (23) 

 becomes applicable instead of (22). The simplest case to 

 observe experimentally is that in which the polarizing current 

 is produced by a small applied E.M.F. instead of by a high one 

 acting through a high resistance. The resulting polarization of 

 the cell will then appreciably reduce the effective E.M.F., and 

 cause a gradual decrease in the polarizing current. This type 

 of variation in the current was investigated in some experiments 

 made on mercurous-nitrate cells, the results of two of which may 

 be quoted. When the cathode was made unpolarizable, so that 

 the anodic polarization was the whole polarization developed, 

 1 found in every case that over the greatest part of its course 

 the current fell off at a rate very approximately proportional 

 to the logarithm of the time. To obtain an analytical 

 expression for the polarization we might thus represent the 

 current by the formula 



C = B -Mog* (26) 



where B and b are constants, but it would be unsafe to intro- 

 duce this empirical expression into the integral of (23) without 



* The slight want of agreement in the latter half of the curve is pos- 

 sibly due to the fact that some lamps in another part of the building 

 worked by the secondary cells used to supply the polarizing current may 

 have been switched off temporarily, increasing the E.M.F. of the secon- 

 dary cells and thus the polarizing current. Readings of the current were 

 taken at 31 - 5 and 37*5 minutes, so that, if the lamps were turned on again 

 before the latter reading, the effect would be missed in the current 

 measurements, although a permanent increase in the polarization would 

 be produced. This had happened in previous experiments which were 

 rejected in consequence, but I have been unable to verify it in this 

 instance. 



