REVERSIBLE AND IRREVERSIBLE SYSTEMS UNDER INFLUENCE OF LIGHT. 361 
June 3 and 4 . 37 2 x 16-4 = 22 5 x 10 :! and 33 2 x 21'8 = 23*7 x 10 3 (not corrected). 
37" x 17*9 = 24'5 x 10 3 and 33 2 x 23'0 = 25'0x 10 3 (corrected). 
June 5 . . . 36 2 x 16'3 = 211 x 10 3 and 26 3 x 3U4 = 2P2 x 10 3 (not corrected). 
37 2 x 16-3 = 22’3 x 10 3 and 26 2 x 33'2 - 22*5 x 10 3 (corrected). 
June 6 . . . 36 2 x 25 - 5 = 33*0 x 10 3 and 26 2 x 49‘6 = 33'5 x ] 0 3 , 
i.e., since the thermo E.M.F. is directly proportional to the intensity of light, we 
find, as before, that the E.M.F. produced by light at a constant temperature is also 
(as the toted E.M.F.) directly proportional to the intensity of light. 
Experiments of June 18, 1903 (2 Curves ), Table I. 
The Ag plates of June 3, 1903, were polished again and covered on one side with 
paraffin, and placed into a 0T041 normal AgN0 3 solution. Resistance of liquid in 
quartz vessel = 840 ohms. After 1|- hours the E.M.F. in the dark was still 
17 centims. to the right, but further diminution became slow. At a distance of the 
arc of 27 centims. a deflection = 20 millims. was obtained. (Experimental data, 
Table I., pp. 390-391.) 
A comparison of this result with those subsequently obtained on June 9 shows 
that the formation of an oxide layer, if any, can take place with pure silver plates 
only very slowly, and that it is the Ag plate which gives the E.M.F. under the action 
of light, not a layer of oxide. 
Comparing, on the other hand, the above result with those of June 6, and calculat¬ 
ing the deflection for the same resistance of the liquid (847'8 instead of 1674), the same 
— = 0‘64lJ, the same distances (37 centims. instead of 
27 centims.), we get 20'7 millims. deflection (with the newly polished plates) instead 
of 28 millims. on June 6 (with plates which were several days in solution). This 
seems also to indicate either that the problem of getting a light of constant com¬ 
position with the arc is still awaiting its solution, or that a small diminution in the 
brightness of the Ag plates still takes place in time, though very slowly. 
Plates (a), (b), ( c) of June 19, 1903 ( Table /.). The Influence of the Concentration 
of the Solution upon the E.M.F. of the System. 
Three different concentrations of AgN0 3 were rapidly investigated one after 
another in the course of about 3 hours, between 11.30 a.m. and 2.20 p.m., passing 
from the dilute solution to the more concentrated one. The solutions used were 
0'0041 normal (Table a ), 0’025 normal (Plate b), and 0’25 normal (Table c). 
Experimental data, see Table I. 
Plate (a) contains 5 Curves. —The same solution as used on June 18, but the E.M.F. 
3 A 
intensity of the light 
VOL. CCVI.—A. 
