74 Scientific Proceedings^ Royal Duhlin Societij. 



It soon became evident that the activity was enoiiuously increased, so 

 that the natural leak of the testing vessel was negligible. In making tests 

 with chlorophyll the plate system was isolated, and a reading taken with 

 the arc burning, but shuttered off. The shutter was then raised for about 

 ten seconds, as timed with a stop-watch, and then replaced. The electro- 

 meter spot of light soon became almost stationary again, and the difference 

 between the two readings measured the effect for the time that the light was 

 acting. From this the displacement per minute, and so the current, was 

 found. This varied considerably in different tests, the variations being 

 probably chiefly due to changes in the arc. When a clean zinc plate was 

 used the shortest flash of light that could be given sent the electrometer off 

 the scale. A condenser of capacity ^ microfarad was put in parallel with the 

 electrometer, thus reducing the sensitiveness to about 7/0?''^ P^-i't of its 

 previous value. An exposure of about half a minute then produced a con- 

 venient deflexion. The results obtained are shown below : — 



Substance on Test Plate. Cunent in E. S. U. 



Zinc plate coated with chlorophyll, . . . 12 x 10"' 



Layer of leaf-powder about I'o mm. thick, . . 16 ,, 



Clean zinc plate, ..... 6650 ,, 



It will be seen that tlie chlorophyll is about 2000 times, and the zinc 

 over 10,000 times, as active with this source compared with the previous 

 one. The interposition of the red screen reduced the activity almost, if not 

 quite, to zero. The measurement of very small currents could not be made 

 quite so accurately as before, as the motion of the electrometer needle was 

 not quite so regular. The variations were probably due to draughts of 

 slightly ionized air from the arc reaching the connecting wires. 



Discussion of Kesults. 



All the evidence points to the fact that the extremely small effect obtained 

 with the " half-watt " lamp was due to the small amount of ultra-violet 

 light present. The last test shows that the latter produces a pronounced 

 effect. The light which passes through the red screen is apparently inactive. 

 Now, this red light is very effective in causing photosynthesis, so apparently 

 photosynthesis is caused by light whose frequency is too low to effect the 

 expulsion of an electron from the chlorophyll molecule. 



This conclusion is greatly strengthened when we consider quantitatively 

 the results obtained. The largest photo-electric current obtained with leaf- 

 powder illuminated by the "half-watt" lamp was about 9 x lO"'^ E.s.u. Let us 

 assume that the current was magnified twenty times by collision — an estimate 

 which, in view both of the results obtained with the zinc plate and of the value 



