Abstract :
The paper indicates how a whitened cube can be made to give accurate results when used in the determination of the average candle-powers of lamps having different polar distributions. As a preliminary, the relative effects of different portions of the interior of the cube in producing illumination on the window are determined, and the results used to calculate the illuminations produced by different types of light distribution. The ratios of these illuminations to that furnished by a uniform point source of unit intensity give apparent average candle-powers in the cube. It is then shown by experiment that candle-powers obtained by the usual methods of integrating photometry agree with these calculated values. Thus it is possible to determine by how much these apparent candle-powers differ from true candlepowers deduced by point-to-point methods, and hence to obtain true candle-powers from experimentally determined apparent values. In the special case when a comparison is made between lamps having identical polar distributions, it is shown that the usual method gives accurate results.
