THE PRESIDENTIAL ADDRESS 15 



artificial counterpart using the photoelectric surface has the valuable 

 property that the electric current which indicates that light is falling 

 upon it can be precisely measured, so as to determine the intensity 

 of the light. In contrast with photographic action, the energy 

 available to produce the record comes not from the original source 

 of light, which only, as it were, pulls the trigger, but from the battery 

 in the local circuit, and it may be amplified so as to actuate robust 

 mechanisms. It has been applied with success to guiding a large 

 telescope or, in a humbler sphere, to open doors, or even to catch 

 thieves. 



However, the scientific interest lies more in the possibility of 

 accurate measurement. As an interesting example we might take 

 the problem of measuring the apparent diameter of the great nebula 

 in Andromeda. As is known, modern research tends to indicate 

 that the Andromeda nebula and other like systems. are the counter- 

 parts of the galaxy, being in fact island universes. But until lately 

 there was such a serious difficulty in that all such systems appeared 

 to be considerably smaller than the galaxy. Stebbins and Whitford, 

 by traversing a telescope armed with a photoelectric cell across the 

 nebula, have found that its linear dimensions were twice as great 

 as had been supposed, reducing the discrepancy of size to com- 

 paratively little. 



But, it may be suggested, could we not go further and make a 

 photoelectric equivalent, not only for the rudimentary kind of eye 

 which has only a single sensitive element, but for the developed 

 mammalian eye which has an enormous number ? Could we not 

 build up on separated photoelectric elements a complete and detailed 

 picture .? In point of fact this has been done in the development of 

 television ; and since this new art which interests us all can properly 

 be considered as an extension of the powers of nornial vision, no 

 excuse is needed for devoting some consideration to it. We must 

 divide the photoelectric surface into minute patches which are 

 electrically insulated from one another. This is not too difficult ; 

 but if it were proposed directly to imitate nature, and attach a wire, 

 representing a nerve fibre, to each of these patches, so as to connect 

 it to the auxiliary apparatus, we might well despair of the task ; for 

 there are probably half a million such connections between the human 

 retina and the brain. In the artificial apparatus for television, one 

 single connection is made to serve, but it is in effect attached to each 

 of the patches in rapid succession by the process of ' scanning ' the 

 image. The photoelectric mosaic is on one side of a thin mica sheet, 

 and a continuous metal coating on the other side gives the connec- 

 tion, which is by electrostatic induction. Each element of the surface 

 forms a separate tiny condenser with the opposing part of the back 

 plate. Scanning is achieved by rapidly traversing a beam of electrons 



