REVERSIBLE AND IRREVERSIBLE SYSTEMS UNDER INFLUENCE OF LIGHT. 3 77 
0‘045. The E.M.F. created by the total light passing the red screen cannot be 
measured, and that created by the total light passing the yellow-green screen is also 
very small, so as not to admit accurate measurement. It follows from this that the 
above E.M.F.’s given by the total light or the blue screen can be taken as created by 
light at a constant temperature. Plate of February 19 shows that the system repre¬ 
sents a constant reversible cell. The reactions going on in the system are the same 
as those already discussed in § 5. 
IX. Plate 15, Table VII. The Transformation of Constant Reversible Cells into 
Inconstant Showing Polarisation, and vice versa. 
The Ag-BrAg plates, which in different solutions of bromide were exposed to light 
during July, 1904 (see Table V.), began at last to show a very small polarisation. 
The Ag-BrAg plates were then kept in the dark in a 0'01 normal KBr solution. On 
November 18, 1904, a small polarisation was already very evident from the course of 
both the induction and deduction periods (see Plate Nl5, experimental data in 
Table VII.). 
We remember that similar was the result with Ag-BrAg plates in a 0T normal 
NaBr solution on July 31, 1903 (see curve N9, especially NlO of plate Nl4), and 
with Ag-ClAg plates in NaCl solutions on July 1, 1903 (Table II.) after the plates 
were exposed for sufficient time to light. The differences (the total E.M.F.—E.M.F. 
of polarisation) were found here at 76 centims. distance : 73, 60 (corr. 71, the Swan a 
little shaky), 61 (corr. 72), 60 (corr. 75 millims.), i.e., whether corrected or not, the 
E.M.F. of polarisation seems to be here for a given intensity (and composition) of 
light always of the same, or nearly the same value, as the principal E.M.F. is. 
This transformation of the constant cells, after sufficient exposure to light, into 
inconstant, follows as a necessity from the kind of reactions going on m these systems 
and discussed in § 5. Since the current passes in these systems from the plate in the 
dark to the plate in light, the upper BrAg, ClAg layer of the exposed plate transforms 
into Ag, and the systems (Ag plate in light, BrNa solution, Ag-BrAg plate in the 
dark), &c., must be formed. A consideration of the reactions going on in such 
systems under the action of light, when the Ag plate is exposed to light, and direct 
experiments instituted with the system (Ag plate in light, BrNa solution, Ag-BrAg 
plate in the dark), showed that such combinations form inconstant cells, showing 
polarisation, thus confirming the kind of reactions going on m the system reversible 
in respect of the anion. 
Moreover, a further consideration shows that we must be able to transform back 
the formed inconstant cell into the former constant cell, if we transform the formed 
under light Ag surface back into AgBr. This now is actually the case : by immersing 
the affected Ag-BrAg plate in a solution of Br 2 in water, &c., on several occasions 
I got back constant cells showing no polarisation. 
VOL. CCVI.—A. 3 c 
