AT WINTER QUARTERS. 163 



many points of similarity to that shown at the commencement of the storm of November 7, 1902. 

 It is, however, a much larger movement, and is exactly opposite in direction, the first movement on 

 April 6, 1903, representing decrease in all the elements, whilst that on November 7, 1902, represented 

 increase at least in H and V. The range shown in the second portion of Plate XXVI is 3 45' in D, 

 lOi'y in H, and 105y in V, but as the D trace got off the sheet towards the end of the time, and the H trace 

 was off for 2| consecutive hours, the range in these two elements may have considerably exceeded that 

 shown. The end of the traces in Plate XXVI answers to the end of the photographic sheet of April 5-6. 

 The commencing portion of the next sheet (not reproduced) shows a similarly disturbed set of conditions, 

 lasting until nearly 2 a.m., L.T., on the 7th. 



69. Plates XXVII to XXXI afford illustrations of what I have called the " special typo of disturbance," 

 which is discussed in detail in Chapter X. Plate XXVII is a good example of moderate size occurring 

 on June 19, 1903. The disturbance referred to is that shown from about 6.40 to 7.20 p.m., L.T. During 

 what I have termed the "first phase"- extending from 6.40 to 7.1 p.m. the D and H traces go up, 

 the V trace down the sheet, i.e. all three elements diminish. During the second phase the movements 

 are in the reverse direction, i.e. the three elements increase. In this individual case the return D and 

 H movements are less than the original movements, something occurring to check them. The return 

 movement in the V curve exceeds the first movement, as nearly always happened. The curves show 

 the continual minor oscillations characteristic of practically all the Antarctic D and H records. Thus 

 it is difficult to assign a very definite time for either the beginning of the first phase or the conclusion of 

 the second. Most likely the duration of the phases is underestimated. The H trace went beyond the 

 limits of registration apparently just exceeding them about 7 p.m., L.T. The summit in the H curve 

 must have occurred a few minutes after the hour 7.1 p.m. accepted above for the end of the first phase. 

 The curve shows, however, a peak somewhat earlier, about the time of the turning-points in the D and 

 V curves. The behaviour of the II trace during the special type of disturbance was less uniform than 

 that of the other two elements the first change in H being sometimes an increase and in fixing the end 

 of the first phase more attention was given to the two other elements, especially to V. 



The changes in H during the special disturbance of June 19, though looking much larger to the eye than 

 those in D, represented in reality considerably less change of force. Thus, during the first phase, the 

 change shown in II represents a decrease of 91 7, but the corresponding fall in D was 72', answering to a 

 force of about 138y. The second phase movement in V, relatively insignificant as it looks, represents an 

 increase of 53y. After reaching the maximum shown at about 7.20 p.m., the V curve shows only a very 

 gradual diminution of ordinate, as if the cause of the special disturbance, whatever that may be, had an 

 effect on the Vertical Force which required time to disappear. 



Synchronously with the special disturbance in the Antarctic on June 19, 1903, there was a well-marked, 

 though much smaller movement at Christchurch, shown in the upper curves of Plate XVII. The 

 prominent crest in the Christchurch D curve, appearing a little before 8 a.m. (G.M.T.), seems to answer 

 to the end of the first phase in the Antarctic movement. 



70. Plates XXVIII and XXIX show movements of a very similar character to that shown by Plate 

 XXVII, occurring on successive days, June 28 and June 29, 1903. The estimated times of ending for the 

 first phase were : for June 28, 7.54 p.m., L.T. ; for June 29, 8.33 p.m., L.T. The occurrence of closely 

 similar movements in Declination on two successive days has been remarked on before. In the 9th edition 

 of the ' Encyclopaedia Britannica '* Dr. BALFOUR STEWART refers to a discussion by Senor CAPELLO of 

 a series of cases in which such changes appeared in the Lisbon magnetic curves, and mentions that he had 

 observed correspondences of similar closeness in the Kew curves. It is not clear whether Dr. STEWART 

 himself considered that the number of the instances of apparent repetitions and the closeness of the 

 resemblance were such as to demonstrate that more than chance was involved. But provisionally, 

 supposing some definite physical cause to be involved, he suggests that something might lead to the 

 neighbourhood of some particular meridian of the Earth's surface becoming sensitive to radiations received 

 from the Sun, and that the consequence might be the repetition of a special type of disturbance on 



* Terrestrial Magnetism (under Meteorology), Art. 87. 

 Y 2 



