(> 1' Til E II O i: I /() N T A 1. Fil KC E. 



31 



The average amount by which the disturbances tend to decrease the diurnal 

 variation of the horizontal force is 1.4 scale divisions or 0.00021 in the absolute 

 Bcale. The maximum effect takes place at 8 1'. M., at exactly the same hour when 

 the declination disturbances reach their greatest effect. 



In the preceding Tables, XIII., XIV., and X V., to the hours indicated 21 1 

 minutes should be added, the observations being made so much later than the even 

 hours. 



The preceding discussion shows that for two stations, even at a comparatively 

 short distance, as for Philadelphia and Toronto, there are, generally speaking, some 

 close coincidences in the laws derived from independent observations; but there 

 are also certain differences in other results; yet it must not be forgotten that for a 

 strict comparability we require, if not simultaneous observations, at least observa- 

 tions extending over similar parts or the whole of an eleven year period. The 

 Philadelphia series includes a minimum year of that inequality, with the greater 

 extent of observations before that epoch, whereas at Toronto the series begins after 

 the minimum epoch and barely extends to a maximum year. 



For the purpose of obtaining a better view of the absolute amount of the disturb- 

 ances and their frequency of occurrence, 1 they were classified in nine groups of equal 

 differences of 20 scale divisions; the number of disturbances in each was found as 

 follows : — 



The numbers in the last column cannot be considered as entirely independent 

 of the eleven year period, and in attempting to apply the theory of probabilities in 



1 A table analogous to that given above, showing the distribution of (he disturbances in declina- 

 tion, is here added for comparison: — 



