REVERSIBLE AND IRREVERSIBLE SYSTEMS UNDER INFLUENCE OF LIGHT. 351 
were polished with the greatest possible care, it is found that their E.M.F. in the 
dark on immersing them in a liquid is, in comparison with the above values, enormous. 
Moreover, the E.M.F. of the plates is unceasingly varying, the mirror of the galvano¬ 
meter is continuously in movement. It is only on account of the observations of 
Becquerel that the E.M.F. of AgCl, AgBr, Agl plates in the dark, in a closed circuit, 
gradually diminishes and becomes stationary, that research in this region becomes at all 
possible. But it is only with exceedingly well polished plates that the constant very 
small values of the E.M.F. finally obtained in the dark can be of real use. This is one 
of the very first conditions for getting good, reliable results in this region of research. 
After the plates were polished, they were annealed by heating them gradually to 
about 120° to 140° C., and gradually cooling for several hours, repeating the 
operation. The plates were then polished again, to remove the thin film of oxide, 
and uniformly covered on one side with shellac and over this with paraffin, or with 
paraffin only, so that the current should pass between the two surfaces of the two 
plates only. The plates so prepared are then placed in the liquid of the quartz 
vessel for the investigation, the circuit is left closed until the E.M.F. in the dark not 
only ceases to vary, but gradually diminishes till it reaches the smallest possible value 
(in case of many plates it can be brought to about 1 centim. or less on my galvano¬ 
meter, i.e., to 0-00001 volt and less), which remained quite stationary for longer 
periods, so as no more to interfere with the research. This takes hours, sometimes 
24 hours, 48 hours, sometimes much longer. Once the plates have been further 
finally polished by the current (this is what actually happens here) they give no 
more trouble. They require very little time for reaching their zero deflection on 
exposure to and on removal of the light, or on opening and closing of the circuit. 
Photographing the Effect of Light upon the Plates. 
The reactions which take place under the action of light are of so great speed, and 
the curves themselves are of such a complicated course, that there can be no correct 
and complete study of the phenomena, unless they are put on record by photography. 
The nature of the phenomena with the plates is such that it became necessary to 
take more than one curve on the same sheet. Special arrangements were made so as 
to know with absolute certainty the meaning of each of the curves from their position 
on the plate obtained, and not to mix them up with one another, and also not to fog 
the very sensitive bromide paper while there was light in the room and the different 
measurements made. When inconstant cells, showing polarisation, will be dealt with, 
proof will be given in several ways that the special course of the parts of curves 
giving the induction and deduction periods is not due to the inertia of the moving 
galvanometer mirror, but represents a true phenomenon. The curves (Plates 1-8) give 
the originals reduced to half their size. All the details which were written down on 
each plate at the time of the experiment can only now be read with the help of a lens. 
All the data given on the plates are contained in the accounts of the experiments. 
