REVERSIBLE AND IRREVERSIBLE SYSTEMS UNDER INFLUENCE OF LIGHT. 357 
current in the solution, the red and blue thick leads from the quartz vessel were 
always connected in the same way to the galvanometer (Nalder Nl), the red lead 
with the plate exposed to light. On replacing the light cell by a Daniell, it was 
found that the plate in light behaves to the plate in the dark as Zn to Cu in the 
Daniell, i.e ., under the action of light the current passes in the solution from the Ag 
plate in light to the Ag plate in the dark. 
x4.fter an hour the E.M.F. in the dark was =2 ‘7 centims. to the right, and next 
day (June 4) = 1*5 centims. to the right = 17'6 x 10~ y ampere = 15‘0 x ICU 6 volt. 
The arc (Dubose) was at 65 volts, 6 amperes. The intensity of light (measured by 
the Rubens thermopile on Nalder N2)=1D6 centims. (right and left), while the 
calibration with the Clark and 100,000 ohms resistance and 10 ohms shunt of 
galvanometer gave a deflection (right and left) 19'8 centims. The quartz vessel was 
removed from the quartz window 8 centims. There are 16 curves on this plate, 
numbered in the order they were taken. After the curves were taken, the sensitive¬ 
ness of the galvanometer (Nalder Nl) was measured in standard units (Clark, 
83,600 ohms in circuit; 5 ohms in shunt of galvanometer); the same was (right and 
left) = 17‘4 centims. The resistance of the solution in the quartz vessel between the 
plates 13'4 ohms. 
Curve Nl shows that, after the first exposure to light, the system passes on 
removal to the light, in the dark, through a deduction period, back to its former 
state in the dark before illumination, approaching it however very slowly and 
asymptotically. 
Curves N2, 3, 4, 5, 6 (distance of the arc from the quartz vessel 37 centims.) show 
that after the first exposure we get the same curves on repetition of the experiment, 
at the same distance of the arc from the quartz vessel. The return to the state in the 
dark on removal of the light is complete in (2), (3), (4), (5), (6), after curve (l), when 
the small residue of energy created by the heating effect of light, which leaves the 
system only very slowly, is allowed to remain in it (2 millims. deflectiou). The 
deflections are in (1) 20 millims., in (2) 19 millims. (corr. 21 millims.), in (3) 17 millims. 
(corr. 19 millims.), in (4) 18 millims. (corr. 20 millims.), in (5) 18 millims. (corr. 
20 millims.), in (6) 16 millims. (corr. 18 millims.), i.e., if not corrected, 20, 19, 17, 18, 
18, 16, average = 18 millims. = 21*1 x 10“ 9 ampere = 18'OxlO -6 volt; if corrected, 
20, 21, 19, 20, 20, 18 millims. = average ] 9*7 millims. = 23'1 x 10~ 9 ampere 
= 19 6 x 10“ 6 volt. 
This shows beyond dispute that the system is in the dark in one constant state, and 
in light (of a given intensity) again in another. 
The curves N7, 8, 9, 10, 11, 12 test the same point at a distance of 33 centims. 
After N7, 8, the curves were taken uninterruptedly one after another to see whether 
the system quickly returns to the same two states in light and in the dark. We find 
that on every occasion the curves reached, or almost reached, in the dark the same 
line o-6, and again every time it returned in light to the same line c-d. The author 
