180 



ANALYSIS OF DISTURBANCES. 



Falmouth ; and Winter Quarters, though nearer than the first three of these stations, is comparatively 

 little nearer. The angular distances are, in short, approximately us follows : Christchurch 126", Golaba 123", 

 Mauritius 121 J, Winter Quarters 113, Kew 60f. Thus, if the disturbance had been a direct consequence 

 of the eruption, what we would have expected to find would have l>een disturbances of nearly equal 

 magnitude at Christchurch, Colaba, Mauritius, and Winter Quarters, that magnitude being much less than 

 that of the disturbance at Kew and Falmouth. Whereas we see from Table LXVIII that the disturbance 

 at Winter Quarters was of the order of ten times that at Christchurch, while the disturbance at Colaba 

 exceeded that at Kew. 



Again, if the commencement of the May 8 storm was directly due to the eruption, what are we to think 

 of the remainder of the storm which lasted 8 hours and presented a remarkable unity of appearance ; and 

 how are we to explain the fact that the largest movements occurred 6 or 7 hours after the commencement, 

 and that one of these movements, viz., that occurring between 7.4 and 7.11 p.m., G.M.T., presented at Kew, 

 Colaba and Mauritius a remarkable similarity in type to the commencing movement 1 The conclusion we 

 seem led to is that the coincidence in time with the eruption was purely accidental. 



101. The bearing of the results of Tables LXII to LXVIII on theory calls for a short comment. We 

 have found that disturbances of comparatively short period whether commencements of magnetic storms 

 or not show a general tendency to approximate to one or other of a small number of types. Except at 

 Christchurch, in the case of the special type of disturbance, there was no very clear indication of more 

 than two classes, confining ourselves, of course, to cases in which a distinct correspondence was visible 

 between synchronous disturbances at distant stations. In the limited number of cases we have considered, 

 there has also been apart from the special type of disturbance a general tendency for disturbances to be 

 larger at some stations than others. 



On what I have called the action-at-a-distance theory this would imply that the disturbing field supposed 

 has a tendency to have a more or less fixed direction relative to axes fixed in the Earth. If the direction 

 assumed by that field were largely variable as it would be if it pointed to the Sun, OF were at right angles 

 to the line joining the Earth and Sun then one would expect the maximum disturbance to be experienced 

 at widely different places on different occasions. The sudden commencements we have had to do with 

 occurred at different times, G.M.T., that of May 8, 1902, being near Greenwich noon, those of April 5 and 

 August 25, 1903, near Greenwich midnight. Again, the sudden commencements were at widely different 

 seasons of the year, yet we do not find any conspicuous difference between the phenomena experienced. 



As already remarked, Christchurch is not very far from being a diametral point to Falmouth and Kew, 

 and on the action-at-a-distance theory we should expect the disturbance vectors at diametral points to be 

 parallel in direction and equal in magnitude. So far as the parallelism in direction is considered, there is 

 unquestionably considerable support to the theory. If we take Table LXVII for instance, we have for the 

 first five (or really seven) disturbances : 



There is here unquestionably a somewhat remarkable accordance. When, however, we come to the 

 magnitude of the vectors, we find that the disturbance at Christchurch was invariably less than half that 

 at Kew. 



When considering the significance of the vector angles 6 and <f> there is one point that should be borne 

 in mind. The assumption by a vector of a nearly constant direction unquestionably suggests the action of 

 an external force having also a nearly fixed direction. It may mean, however, no more than that the action 

 of any external impulse tends to influence the Earth's magnetism in a particular way, just as when 



