

Periodicity of the Spontaneous Ionization of Air, etc. 273 



to be impossible to connect the temperature curve with the 

 ionization curve in any way. The temperature curve had 

 only one maximum and one minimum, so that i£ there be 

 any connexion between the two curves the periodicity o£ the 

 ionization must depend on some differential coefficient o£ 

 the temperature. Then, again, the nature o£ the temperature 

 curve depended entirely on the hot-water pipes, and was 

 quite different on Saturdays and Sundays from the rest of 

 the week, whereas we were unable to detect any change 

 whatever in the character o£ our curves on these days. 



The next quantity, variations o£ which were measured with 

 a view to connecting them with the variations of the ioniza- 

 tion, was atmospheric potential. It was at once evident that 

 there was a close similarity between the curves representing 

 the simultaneous variation of the two quantities. For some 

 time, the atmospheric potential recorder at the Observatory 

 here was run with the kind assistance o£ Mr. A. R. Hinks, 

 but in the end it was found to be unnecessary, as the very 

 full data published of the observations at Kew, when used 

 instead, brought out the same similarity and at the same 

 time showed that it could not be due to local causes. 



In figs. 6 to 10 the ionization curves and the atmospheric 



Fio-. 6. 



X /on/jr/rrioH Ct/m-e f£d. { /306. 



potential curves for the same month are plotted together. 

 There is no connexion between the scales representing the 



