176 ANALYSIS OF DISTURBANCES. 



The disturbances of June 19, 28, 29 and July 26, 1903, were simultaneous with the two phases of the 

 special type of disturbance recorded in the Antarctic. In this case also an imaginary mean disturbance 

 has its components calculated, and the resultant and its direction angles thence deduced. 



Owing to electric tram disturbances, changes of the order 0'5y in V could not be recognised with any 

 approach to certainty. This is no doubt partly responsible for the fact that the value is so frequently 

 assigned to AV. When the vector is small there is inevitably a good deal of uncertainty in the 

 determination of both and <f>, and the values given must not be regarded as usually more than somewhat 

 rough approximations to the truth. 



On examining the results it will be seen that practically all the disturbances may be regarded as included 

 under one or other of two types, distinguished by the letters A and B. The disturbances of the type A 

 have a value of 6 in the neighbourhood of 60 and a value of <j> in the neighbourhood of 220. They include 

 those sudden commencements in which the increase of H is the most noticeable feature, and the movements 

 corresponding in time with the second phase of the Antarctic disturbances of the special type. An exception 

 is provided by the disturbance of June 19, 1903, in which the value of < departs largely from 220. On 

 this occasion, however, the disturbance at Kew was so small that comparatively little weight attaches to 

 the results deduced. The disturbance on May 8, 1902, from 7.4 to 7.11 p.m., is also included amongst the 

 A's. The disturbances to which the letter B is attached have a value of which is in the neighbourhood of 

 120. The values of < are more variable, but are all included between +63 and - 63. The movements 

 corresponding in time to the first phase of the special type of Antarctic disturbances all come fairly under 

 this category ; so does the first phase of the sudden commencing movement of August 25, 1903. 



93. The disturbances at Colaba, in Table LXIII, p. 182, are also mainly of two types, distinguished as 

 before by the letters A and B. The characteristic feature at Colaba is the large size of AZ, the component 

 parallel to the Earth's axis, and the small size of AX and AY. Owing to this latter fact the evaluation of 

 <f> is particularly uncertain, and comparatively little significance can be assigned it. We may thus 

 regard the essential feature of the type A disturbances as the possession of a small value for 6, and the 

 essential feature of the type B disturbances as the possession of a value for approaching 180. 



The sudden commencements in which H increases, and the disturbance synchronising with the second 

 phase of the Antarctic disturbance of July 26, 1 903, are included amongst the A's, and so is the movement 

 between 7.4 and 7.11 p.m. on May 8, 1902. Thus the A's at Colaba and the A's at Kew correspond to one 

 another. 



Comparing the values of AK at Kew and Colaba, it will be seen that the former are very decidedly the 

 larger, except on May 8, 1902. 



94. For the Mauritius Table LXIV, p. 182, our information is less extensive, as the copies of disturbances 

 in D and V were mostly confined to 1902. 



As at Kew and Colaba, the sudden commencements of May 8 and August 20, 1902, and the movement 

 from 7.4 to 7.11 p.m. on the former date, are of similar type and are classed as A's. The values obtained 

 from AR are all smaller than the corresponding values at Kew or Colaba. 



95. The data for Christchurch, Table LXV, p. 182, include all the disturbances already considered for 

 Kew. The results are more difficult of classification than those at the other co-operating stations. The 

 commencing disturbances of May 8 and August 20, 1902, and of April 5, 1903, as well as the movements 

 answering in time to the second phase of the Antarctic disturbances of June 19, June 28, and July 26, 

 have been classed as A's. They present, however, a considerable range of values in 6 as well as in <f>. 

 This group also includes the first of the to-and-fro movements experienced on August 25, 1903. This last 

 movement was, however, so minute that too much weight should not be ascribed to the fact. 



As elsewhere, the movement from 2.10 to 2.28 p.m. on May 8, 1902, and that answering to the first phase 

 on July 26, 1903, appear of similar type and have been classed as B's. But the first movement on 

 July 26 is clearly of the same type as the second on June 29, so this also must be classed as a B. The 

 first-phase movements on June 19, June 28, and June 29, 1903, are very similar in type, but differ from 

 either the A or the B class, and have accordingly been classified as C's. This class seems also to include 

 the second and larger movement on August 25, 1903. 



If we consider the sudden commencements, we find that the values of AR at Christchurch are decidedly 



