65 6 Mr. McClung on the Relation between Rate of 



well-defined circular opening, A, in the lead. The cylinder 

 in which the air was ionized is shown more in detail in 



Fijr. 1. 



WWWW\AAA/V 



^J. 



-A- 



£arth 



SA/fT^ 



L' 





Q' 



\ 1 

 t 



fig, 2. XY was a circular brass cylinder about 34 cms. 

 long and 14 cms. diameter. The end Y was closed by a 

 brass cap which screwed on to the end o£ the cylinder, 

 while at the other end was a thin aluminium window, B, 

 through which the rays were allowed to pass into the 

 cylinder. 



The arrangement o£ electrodes inside the cylinder was 

 similar to that used in the previous experiments, for the 

 same reasons as were explained in that instance, but the 

 method of supports and insulation had to be modified some- 

 what on account of the higher temperatures which were to be' 

 used. The electrodes consisted of extremely thin aluminium 

 foil w^hich was stretched on brass annular rings for support. 

 These rings were 9*1 cms. inside and 10 cms. outside 

 diameter, and were grooved round the outside edge, and the 

 aluminium foil was folded over this edge and then bound 

 tightly round with thin copper wire. These rings were 

 eleven in number, and were placed about 2 cms. apart, so 

 that there was a length of about 20 cms. of air on which the 

 measurements were made. Six of these alternate rings were 



