338 



TRANSACTIONS OF SECTION A. 



In order to show to what extent temperature affects or controls curves E 

 and E, the monthly mean temperature has been plotted in conjunction with 

 these. The result demonstrates that from 1905 to 1910 one curve greatly 

 resembles the other, while from 1899 to 1904 there are great differences between 

 the two, although ' notches ' are usually noticeable in curve E upon the months 

 of maximum and minimum temperature. 



Acting upon the advice of Sir George Darwin, who kindly furnished the 

 necessary formulae, I have separated the annual inequality from the ' march 



Mean Temperature 



Number of Quakes 

 Recorded 



Mean Pos/tion 

 ofE-W.HPenc/u/um 



Fig. 2. 



of zero,' and find the result bears out the seasonal and annual movements pre- 

 viously referred to in curve E, fig. 1. 



As certain not fully understood rhythmic changes prevail in magnetic curves, 

 I have plotted the Toronto horizontal force for the entire twelve years in 

 conjunction with curve E, and several years' magnetic intensity for Chelten- 

 ham, Md., and Honolulu, and find many interesting comparisons which when 

 further developed may prove of value in helping to solve the origin of curve E, 

 and explain certain magnetic phenomena. 



The latitude variation curve for several years has also been studied in con- 

 junction with E, and it is hoped when this comparison is more thoroughly 

 pursued interesting results may be obtained. 



Fig. 2 is given to show the annual position of curve E both above and 



