272 Dr. Wood and Mr. Campbell on the Diurnal 



values of the ionization were then corrected for the rise, and 

 the excess or defect of the value of the ionization for each 

 hour of the day from the mean value for the day was plotted. 

 In this way all the later curves were obtained. 



Fig J 











::::::::z\;: 





tl\ 





f ^ 





___ 1 



■A 



i 



/ \ 



r 



^«™ ""~v L ^ 



r_ 



r v" 



L_Z 



i 



1/ 



i 



\r 



b 





§ 





•a auu 





















6 8 10 12 I* 16 18 



ftoUftS O^ T//S D4Y- - 



Mean of Curves in Fig-. 4. 



20 22 24 



The periodic character of the variation was at first attri- 

 buted to temperature as the most probable cause, and one of 

 Callendar's patent recording; thermometers, kindly put at 

 our disposal by the Cambridge Scientific Instrument Com- 

 pany, was installed alongside of the apparatus. It was found 



