Unit of Candle-power in White Light. 63- 



4. The Colour of Radiation from the Violle Standard. 



At the beginning o£ this paper the importance was ex- 

 plained of giving proper consideration to the colour of the 

 radiation from any proposed standard of light. The in- 

 creasing efficiencies of modern sources of light owe these 

 increases mainly to higher running temperatures, and hence 

 the light emitted tends to consist of a relatively larger pro- 

 portion of shorter waves. The light from the Hefner lamp 

 has been found by the authors to correspond in colour with 

 that from a black bodv at 1540° 0. That from the Pentane 

 lamp to 1610° C. 



The Violle standard, consisting as it does of platinum at 

 the melting-point (1750° C) has a radiation which corre- 

 sponds with the colour of a carbon filament glow-lamp running 

 at an efficiency of 2'6 5 lumens per watt, or 4*7 5 watts per 

 mean spherical candle (see fig. 3 ). Assuming a reduction 

 factor of 0'85 this is equivalent to an ordiuary carbon filament 

 lamp with a specific consumption of 3*8 watts per mean 

 horizontal candle. 



The authors wish to express their acknowledgement in 

 connexion with this work to Dr. R. T. Glazebrook, C.B. y 

 F.R.S., Director of the National Physical Laboratory, and 

 also to Dr. J. A. Harker, F.R.S., for facilities afforded in 

 the use of the black body furnace referred to at the beginning 

 of the paper. 



IV. The Unit of Candle-power in White Light *. By 

 Clifford C. Paterson and B. P. Dudding, A.R.C.Sc. 

 (From the National Physical Laboratory) f . 



Synopsis. 



1. npHE differences in the colour of the light radiated 

 _L from different sources of white light causes uncer- 

 tainties in photometric determinations. The paper describes 

 the methods used at the National Physical Laboratory to 

 minimize this well-known difficulty by the use of the cascade 

 principle. 



* The work dealt with in the various sections of this paper has 

 entailed so many thousands of measurements and candle-power deter- 

 minations of individual lamps, that to give them in detail would greatly 

 overweight the paper with tabular matter. The authors have, therefore, 

 felt obliged to limit themselves to summaries of results and sometimes 

 to bare statements of the results obtained, computed from their manu- 

 script tables. 



t Communicated by the Authors. From the Proc. Phys. Soc. London, 

 April 15, 1915, p. 26:'}. 



