1910] on the Telegraphjj of Photographs, Wireless and hij Wire. 848 



curves now shown on the screen ; that above the axis along which 

 exposure is measured is the sensitive cell, that below this axis the 

 cell of low sensitiveness. Clearly the current passed through the 

 galvanometer is that obtained by joining the sums of the ordinates. 

 This gives the small curve shown as the shaded portion. When the 

 illumination is thrown on the cell the current rises very rapidly 

 instead of gradually, whilst when it is suddenly shut off (at P in the 

 upper curve) it drops to zero almost instantly instead of falling- 

 gradual ly. 



I shall now show, by means of a meter, an image of the pointer 

 of which will be projected on to the screen, how the inertia of sele- 

 nium is overcome. You will first see that if I take away the screen 

 so as to allow light to fall on the selenium cell, current passes into the 

 galvanometer, and the needle slowly deflects several degrees. Now, I 

 quickly shut off the light by intercepting it with the screen, and the 

 needle comes slowly backwards. Such sluggish movement would be 

 impossible for the purposes of photo-telegraphy, where at least half a 

 dozen changes per second are required to be recorded abruptly even 

 in transmitting the simple portraits to which the selenium process is 

 limited. 



Now, using two cells of different characteristics and a Wheatstone 

 bridge arrangement, I will once more allow light to fall suddenly on 

 the two cells simultaneously, and you will see that the galvanometer 

 needle records the change in resistance of the combination quite 

 quickly ; the combination is even more noticeable when the light is 

 suddenly shut off again, the needle returning to zero with great 

 rapidity. This compensated arrangement of selenium cells at once 

 renders their use of practical value for various physical and optical 

 measurements. Professor Korn has found that for an increase 

 in the illumination 8 1, the current obtained is given by the equation 



y = a . hi . e , where ij is the current, a the sensitiveness of 



the cell, ^ and m its inertia constants, and e the base of Naperian 

 logarithms. For two cells to be combined to the greatest advantage 

 we must have them such that if their equations are respectively 



and 

 then 



This makes the condition for good compensation that 



