ANALYSIS OF DISTURBANCES. 



181 



one hits ;i l>ar magnet in a neutral Hold one usually knocks out magnetism irrespective of where 

 uiul how one hits it. 



Suppose, for instance, that, at each station we had .1 change only in II, the horizontal component of the 

 Earth's field, then the vectors we should deduce at the several stations would have the following values for 

 and <j>, supposing II to increase at all : 



Under the circumstances supposed, these angles would ho independent of the size of the change in H. 



An examination of Tables LXII to LXVII will show that the values of and < thus found are in no 

 case very remote from the values actually calculated for these angles in the case of the disturbances 

 classed as A. 



One phenomenon that unquestionably tells against the action-at-a-distance theory is that at all the 

 stations considered AV is invariably loss than the resultant horizontal component v/AN'^ -I-~AE 3 of the 

 disturbing vector. The horizontal component is usually much the larger, even in the Antarctic, where 

 the disturbances are greatest. 



Whatever other conclusion may bo drawn from this investigation, it is, I think, clear that the application 

 of the method to a variety of disturbances drawn from different hours of the day, G.M.T., and from 

 different seasons of the year offers a promising field of research. 



TABLE LXII. Disturbance Components and Itesultants at Kew. (Unit of force 1-y.) 



