3 905 - 6 .] Electrolysis through Precipitation Films. 
459 
Discussion of Results. 
1. A glance at the main tables of results reveals the fact that in 
many cases the final conductivities in any set for a given cell are 
generally much more concordant than are the initial values. It 
might have been better, perhaps, to take as initial value the 
•conductivity of the solutions in the cells without the porous 
diaphragm. 
2. The series of measurements show clearly that the diminution 
of conductivity of a solution across a precipitation film is dependent 
•on the nature of the ions. Thus, taking the series NH 4 C1, 
k T H 4 Br, (HH 4 ) 2 S0 4 , a film of aluminium hydroxide reduces 
the conductivity of the chloride by only 3 per cent., of the bromide 
by 7 per cent., and of the sulphate by over 20 per cent. The 
■diminution measured must be the sum of the effects on the cation 
and the anion, but in these cases, with a common cation, the above 
values may be taken as an approximate measure of the differences 
between the anions Cl', Br', S0 4 ". 
A similar but distinctly smaller effect is found for the cations in 
the series (NH 4 ) 2 S0 4 , K 2 S0 4 , Na 2 S0 4 , with a chromic hydroxide 
film ; it is noticeable that K’ and lSTH 4 ‘ are affected to about the 
same extent, whilst Na* stands apart from them, in accordance with 
other physical and chemical characteristics. 
3. A comparison of the results obtained with films of aluminium 
hydroxide and chromic hydroxide shows that if the ions are 
arranged in order of increased effect of the film, the order is the 
same for the two films, but that the amount of diminution is quite 
different, nor are the ratios the same in the two series. Whether 
this is due to increased influence of the chromic hydroxide film on 
the anion or on the cation, there is, at present, no evidence to 
indicate. 
It may be of interest to mention that several attempts were 
made to form a ferric hydroxide film, but without success. There 
were indications, in the increased resistance, that a film was being 
formed, but it did not remain for long, and a gritty precipitate 
soon made its appearance on one side or the other of the porous 
plate. Apparently the precipitated ferric hydroxide rapidly 
disintegrated, probably losing its colloid character. 
