﻿[ 618 ] 



LXIII. The Photoelectric Effed— III. By 0. W. Richard- 

 son, F R.S., Wheatstone Professor of Physics, University 

 of London, King's College, and F. J. Rogers, Associate 

 Professor of Physics, Stanford University, California*. 



IN the Philosophical Magazine, vol. xxvi. p. 549 (1913), 

 Dr. K. T. Oompton and one of us published data for the 

 metals platinum, aluminium, sodium, and caesium, showing 

 the relative photoelectric efficiency when a given amount of 

 energy fell upon them in the form of light of different wave- 

 lengths. It was intended to reduce the measurements to 

 absolute values, but this was not possible at the time owing 

 to the absence of a suitable radiation standard. This de- 

 ficiency has since been remedied and the results of the 

 measurements are given below. They are to be taken in 

 conjunction with the numerical data and curves published 

 in the paper already referred to. 



To reduce the current measurements to absolute values, 

 the only new data required were the capacities of the electro- 

 meter and the variable condenser which was added to the 

 quadrants in some of the experiments. The capacity of the 

 electrometer and its connexions was found to depend to a 

 considerable extent upon the voltage on the needle. This 

 was therefore determined separately for each of the different 

 voltages which had been used. Both the method of mixtures 

 and a leakage method were employed, and the values obtained 

 when plotted against the potential of the needle were found 

 to increase regularly with the applied potential. The new 

 values also agreed with the values in terms of the variable 

 capacity which had been used in reducing the measurements 

 described in the former paper, showing that the capacity of 

 the electrometer with a given potential on the needle had not 

 changed in the interval. 



In the energy calibration in the former paper, the mono- 

 chromatic light from the illuminator was allowed to fall on a 

 linear thermopile provided with a slit adjusted to the same effec- 

 tive width as the strip used in the photoelectric measurements. 

 The steady therm o-electromotive force developed under the 

 influence of this radiation was balanced against the drop of 

 potential produced by the flow of a known small current 

 through a known small resistance. In order to reduce the 

 former observations to absolute values, it w T as therefore 

 necessary only to determine the electromotive force gene- 

 rated in the thermopile when a beam of radiation of known 

 energy density was allowed to fall on it through the same 

 slit. The source of radiation of known energy density was 

 a standard incandescent-lamp obtained from the Bureau of 



* Communicated by the Authors. 



