'238 Mr. G. W. Walker on the 



used. This is what one would expect from our knowledge 

 of other types of radiation. 



The experiments with temperature show that the ionization 

 is independent of the temperature, or that the radiation from 

 the w^alls of the vessel is not affected by temperature up to 

 450° C. at least. 



Since these experiments were performed Strutf^ and 

 McLennan and Burton f have confirmed the conclusion 

 that the ionization is due to the w^alls of the vessel being 

 radioactive by showing that the radiation varies with the 

 material of the vessel, and the latter have further shown that 

 there are very penetrating rays that can pass through the 

 w^alls of the vessel. The radiation is thus evidently complex 

 in character, and the greater part of the ionization is produced 

 by the easily absorbed rays. It thus resembles the radiations 

 from thorium and radiam. 



The average of a large number of observations gave 61 

 as the number of ions produced per c. c. per sec. in the iron 

 cylinder at atmospheric pressure, 3*4 x lO""^*^ E.S. units J as 

 the charge on an ion, and the volume of the cylinder as 

 12,800 c. c. Using this new value for e, and reducing the 

 values of other experimenters, we find that Wilson (loc. cit.) 

 by using a glass flask with a slight coat of silver obtained 

 36 for n, Butherford {loc. cit.) with zinc cylinders 27, while 

 Harm § wath a large glass flask obtained values between 53 

 and 63. The difference in these results is most probably due 

 to the different materials used in the several experiments. 



In conclusion I desire to express my sincere thanks to 

 Prof. Thomson for his kindly interest and suggestions during 

 the course of the investigation. 

 Cavendish Laboratory, Cambridge, 



XXYIII. On the Theory of the Quadrant Electrometer. By 

 George W. Walkee, M.A., A.R.C.Sc, Fellow of Trinity 

 College, Cambridge ||. 



FOR the purpose of some experiments which I am taking 

 up, I have found it necessary to examine the theory of 

 a symmetrical quadrant electrometer more carefully than I 

 have formerly had occasion to do. The results seem to me to 

 be of considerable importance. 



* ^Nature; vol. Ixvii. p. 369 (1903). 



t Ibid. p. 391 (1903). 



X Phil. Mag. vol. V. p. 346 (1903). 



§ Physikalische Zeitschrift., iv. Jahrgaug, Xo. 1, 1902. 



Jl Communicated by the Physical Society : read Jan. 23, 1903. 



