THE LUMINOUS EFFICIENCY OF ILLUMINANTS 541 



of the mantle ; but it appears to be well recognised that its 

 high luminous efficiency is partly due to selective radiation. 



At the time of the introduction of metallic filament, a lively 

 controversy took place regarding the explanation of the higher 

 efficiency of the new lamps. By some authorities, the im- 

 provement was ascribed simply to a higher temperature of 

 incandescence; others contended that the temperature was 

 actually lower than that of the carbon filament and that the 

 gain in efficiency must therefore be due to selective radiation. 

 At the present time, the temperatures attained under practical 

 conditions by the various metallic filaments appear to have been 

 computed fairly definitely and it seems to be agreed that this 

 is considerably above that of the carbon filament ; but the 

 recent work of Hyde, Coblentz and other workers in the United 

 States makes it probable that selective radiation is also partly 

 responsible for the improvement in efficiency. 



Coblentz has recently carried out a most interesting series 

 of researches on the radiation from polished specimens of 

 various metals. His conclusion appears to be that metals 

 which yield highly efficient lamp filaments also exhibit high 

 reflective power and therefore marked suppression of radiation 

 in the infra-red. Thus the reflective power of tungsten was 

 found to be 50 per cent, in the yellow of the visible spectrum, 

 89 per cent, at 2*5^ and as much as 95 per cent, at 10^. This 

 means that there is a diminution in the amount of useless 

 infra-red radiation and that the luminous efficiency is cor- 

 respondingly improved. We do not know yet how far it is 

 possible in practice to improve the luminous efficiency of solids 

 by this means. But in view of the enormous extent of the 

 infra-red range of radiation from most incandescent materials, 

 the results of even a small suppression in this region might 

 be very beneficial. 



Another curious fact mentioned by Coblentz is that the 

 quality of low emission in the infra-red is associated with 

 good electrical conductivity. Metals having a high conductivity 

 possess this quality in a high degree and therefore make 

 efficient sources of light when their other physical qualities 

 are such that they can be used. Most oxides, on the other 

 hand, which are poor conductors, also possess a uniformly high 

 reflecting power in the infra-red. It would almost appear as 

 though the high electrical conductivity which proves so em- 



