REVERSIBLE AND IRREVERSIBLE SYSTEMS UNDER INFLUENCE OF LIGHT. 337 
“ such has actually been found to be the case in many experiments” (p. 222); on the 
whole, just as many mysterious results were obtained as there were experiments 
made. 
The E.M.F.’s obtained changed also in their value. The “tin plates, e.g ., placed 
in alcohol, gave large deflections, but almost none, if any, chloride was introduced 
into the alcohol, &c.,” though this ought to increase the conductivity (p. 216), &c. 
Now what is more characteristic of a galvanic combination than the direction of 
the current in it ? No galvanic combination we know of gives a current sometimes in 
one direction and at others in the opposite, the anode and cathode exchanging their 
roles. At the hands of two such excellent investigators as Becquerel and Minchin 
only one conclusion would seem possible, a conclusion which Minchin, and apparently 
also Becquerel, draw all along, that the action of light upon AgCl, AgBr, AgT, and 
especially metal plates, leads to some special kind of surface phenomena, depending in 
some unknown way upon the thickness of the layer or films, and which for still 
unknown reasons are of a varied and complicated character. 
From the first it was clear to the author that, to make substantial progress, the 
research must be carried out on cjuite different lines. It is not sufficient to notice 
the direction of the current; the total curve in all its complexity from beginning to 
end, including the deduction and induction periods, must be studied and photo¬ 
graphed so as to get a complete, objective, and not only a partial,' knowledge of* the 
total phenomenon. The investigation must be carried out with much greater detail 
than hitherto, and be of quantitative, not of qualitative character, and the results so 
obtained be studied in connection with the chemical composition of the whole hetero¬ 
geneous system and with those reactions which must take place in it under the 
action of light. These problems, which Becquerel and Minchin did not attempt to 
study, confining all their attention to the plates only, alone can lead to a satisfactory 
knowledge of this region. 
After a long and detailed study of the method adopted, necessarily spent in obviating 
thermo-effects in the metallic circuit and 'in the cells and other sources of error, 
the phenomena under consideration were isolated, all complications caused by the 
variation of the intensity and composition of the light and other interfering phenomena 
being removed from the curves. The numerous varieties of curves were thus finally 
reduced to two special kinds only. It was only from this point that serious attention 
could be given to the meaning of the special course of the curves, and real advance 
made. 
The curves of a great many of the systems studied (Ag plates, Sn plates, Pt plates, 
Au plates in NaCl solution) had a peculiar course. On exposing to light, a deflection, 
say to the right, is first obtained (the E.M.F. is positive), which gradually decreases 
till it becomes stationary in light, the E.M.F. still remaining positive ; or it con¬ 
tinues to decrease until it becomes “ zero ” and then becomes negative, and this 
deflection in the opposite direction goes on for some time till it becomes stationary in 
VOL. ocvi.— a. 2 x 
