486 EVENING DISCOURSES 



of course, makes the luminous cord intrinsically much brighter. We can 

 in this way force electricity to make its track through denser and denser 

 masses of molecules, with the result that we achieve brighter and brighter 

 luminous cords. The temptation to push this process to the limit is thus 

 very great, for although we do not necessarily obtain much more light for 

 the electricity we use, we do obtain sources of light which have a most 

 remarkable brilliancy — a quality which has important practical uses. 



The tube containing the discharge naturally gets hotter during this 

 process of forcing electricity through it, and we soon find that the keeping 

 of this tube cool is the main problem before us if we would push the process 

 to extremes. 



Let us examine the phenomenon in stages. We are looking at the lamp 

 in which the mercury vapour pressure is one atmosphere. The envelope 

 is made from a special hard glass to withstand a temperature of about 

 6oo° C. It takes 10 volts per centimetre of arc column to operate the 

 discharge, and the brightness is 150 candles per square centimetre. 



Here is a small edition of the same lamp, using about the same amount 

 of electricity but operating at a mercury vapour pressure of 10 atmospheres. 

 The transparent envelope is of pure quartz, since the temperature now rises 

 to about i,ooo° C. It takes 50 volts per cm. to operate the discharge, and 

 the brightness of the column is 700 to 800 candles per sq. cm., which is 

 about five times as bright as the previous atmospheric pressure type. The 

 bare lamp simply seems to have the same intense light as the previous 

 atmospheric type ; this, however, is an optical illusion. We can perhaps 

 appreciate better the difference in the brightness of these two sources if 

 we project their images side by side on the screen. On these images, the 

 smaller one is about five times brighter than the other. 



However, even quartz will not stand up to the very high temperatures 

 it is possible to attain, and unless we make provision for adequately cooling 

 the quartz envelope, the life of the lamp would be very short. 



In the next example we help matters further by keeping the quartz cool 

 with a stream of water. In this way we are able to dissipate still more 

 power in the tube. The volts per centimetre of arc column are now 300 to 

 500, and the brightness reaches 30,000 candles per sq. cm., which is 

 comparable with that of the high current density arc used for cinema 

 projection and searchlights. We will look at the projection of its image 

 first, it is above the other two. In comparison with the previous two lamps 

 the greater brilliancy is obvious. We will now switch on a naked lamp. 

 It does not do to look at it direct, or after image on our eyes may bother us 

 for five or ten minutes. 



When we tax the qualities of quartz to the limit we can reach 150,000 

 candles per sq. cm., i.e. 1,000 times brighter than our first lamp. This is 

 comparable with the brilliance of the sun itself. But under these conditions 

 the quartz lasts only for a few minutes or seconds, even when cooled with 

 water. 



Lamps of this type, when they become practicable, have one main 

 purpose. Used in conjunction with mirrors or lenses they may take the 

 place of carbon arcs and tungsten lamps for projectors — whether in aero- 

 dromes, cinemas or searchlights — where they may possibly have advantages 

 over the lamps in use for these purposes at present. 



Through the medium of these experiments I have been trying to picture 

 to you the outlook at the moment of the scientific worker in the field of 

 electric lighting. That outlook changes rapidly. We have seen that only 

 during the past four years an increase of three times has been achieved in 

 the amount of light obtained from a given amount of electricity. Secondly, 



