REVERSIBLE AND IRREVERSIBLE SYSTEMS UNDER INFLUENCE OF LIGHT. 381 
S0 4 Hg 2 was mixed up with the mercury. For this reason no comparative experi¬ 
ments with different systems can be made. 
In all cases however we get constant cells, showing no polarisation, and the current 
always jmssed from the electrode in the dark to the electrode in light, as in all cells 
reversible in respect of the anion. Plate N23 gives an illustration of this for the 
system Hg, solid S0 4 Hg 2 in Off normal S0 4 (NH 4 ) 2 . 
XIY. On the Conditions under which a Constant Reversible Galvanic Cell can be 
realised under the Action of Light. 
It was explained before that according to the composition of the system and the 
reactions going on in the same, under the action of the current created by light, we 
get constant cells reversible in respect of the cation or the anion, or inconstant cells. 
We must however realise the following: the composition of the system and the 
reactions going on in the same under the action of the current give here, as in 
ordinary galvanic cells, the conditions of constancy and reversibility only in general , 
but in no galvanic combination we know of does the principal reversible reaction, 
upon which the constancy of the cell depends, take place alone to the entire exclusion 
of all other phenomena and reactions simultaneously taking place in the galvanic cell. 
A constant reversible galvanic cell is in all cases only “ practically ” reversibly 
constant. To start with, there are no two electrodes (even when highly polished) 
made of the same metal, which do not give an E.M.F. in the dark when connected to 
a circuit. Therefore no two galvanic cells of the same combinations can give the 
same E.M.F. In the next instance, it is well known that, to keep the E.M.F. of a 
constant reversible for any length of time practically constant, only small currents 
must be used. The reason of this is that, besides the principal reaction under the 
action of the current, subsidiary reactions take place which become more prominent 
the greater the current density; these reactions also lead to the formation of gas 
batteries, or attack the electrodes, &c. In the third instance, there are very few 
metals which can be got chemically pure (absolutely pure perhaps none), and this 
leads to local currents, local E.M.F., &c. In the fourth instance, there is a variation 
of concentration at the electrodes. In ordinary constant reversible galvanic cells the 
principal E.M.F.’s of the principal reversible reaction, due to the composition of the 
system, happen to be, in comparison with the interfering E.M.F.’s, very great, since 
the difference in the electrical potentials of the electrodes, consisting of two entirely 
different metals or combinations, to the solution are great—hence ordinary constant 
cells caii practically be easily realised. When, however, we pass to galvanic cells 
created by light, where the E.M.F.’s created by light, in connection with the com¬ 
position of the same, are very small, it is evident that the difficulties for a practical 
realisation of constant cells must become enormous. 
The above fixes the problem and also shows the limit of possible success in this 
