ACTION OF LIGHT ON SELENIUM. 
339 
With the magnesium-light in the place of the gas-flame we obtained the following 
results, the 10th shunt being in the galvanometer: — 
Deflection due to current of 1 Leelanche’s cell 180 to right. 
ditto ditto and magnesium-light 280 „ 
Current cut off and only magnesium-light on 11 to left. 
In order to examine into this point more fully we made use of the lime-light. The 
current from one Leclanche’s cell was sent through the selenium, from the marked end, 
and produced a deflection (with the 10th shunt) of 137. The lime-light was then 
focused on the end furthest from the marked end, and this sent the deflection up to 155. 
On screening otf the light, the deflection was at once reduced to 150. Hence we see 
that though at this end (see p. 336) light produces by itself no sensible current, yet, 
when the battery-current is passing, it facilitates its passage. 
The lime-light was then focused on the marked end, and the following results were 
obtained : — 
Deflection due to the current of 1 Leclanche’s cell 160 with 10th shunt. 
ditto ditto and lime-light. .. . 185 
When the light was focused on the central part, we obtained the following results : — 
Deflection due to the current of 1 Leclanche’s cell 171 with 10th shunt. 
ditto ditto and lime-light 176 „ „ 
When the light falls on the central part, and there is no battery-current passing, 
there is no deflection. 
Hence it appears that the action of the lime-light on any portion of this selenium 
plate tends to facilitate the passage of a battery-current through the plate, whatever be 
its direction, but that, when no battery-current is passing, the lime-light falling upon 
certain portions produces a small current of its own in one particular direction, viz. from 
the marked to the unmarked end of the selenium. 
In order that the action of the photo-electric current might be comparable with the 
original current, it was evident that the original (or battery) current should be one of 
weak intensity, and also that the resistance of the circuit should be as small as possible. 
The tubes numbered 23, 24, and 25 were accordingly used, and a small thermopile was 
made use of to send a current through them. 
In the first series of these experiments tube No. 25 was used. 
The positive current entering the selenium at the marked end, gave a deflection on 
the unshunted galvanometer of 65 to the left. While this current was passing, the 
lime-light was focused on the marked end, and the deflection was at once reduced to 
38 to the left of zero. On screening off the light the needle took a swing up to 72, and 
then remained steadily at 70. 
The position of the battery-poles was then reversed in the mercury-cups so that the 
positive current should enter the selenium at the unmarked end. Before exposure, the 
3 b 2 
