AT CO-OPERATING STATIONS. 155 



a sudden commencement at about 11.59 a.m., G.M.T., on May 8, 1902. At Kew, Falmouth, Colaba, 

 Mauritius (curve not shown), and Christchurch the initial change in II was an increase, which continued 

 for about 10 minutes. The Falmouth curve in Plate XIV shows a short stoppage of the upward motion 

 at about 12.4. This stoppage actually existed, being 'as apparent in the Kew as in the Falmouth curve, 

 but it has been somewhat exaggerated in the tracing. The original Colaba curve gives just a suggestion 

 of a corresponding stoppage. It is not shown in the Mauritius curve (a tracing), nor on the Christchurch 

 curve. The latter, however, is somewhat fuzzy, and a very short stoppage might not leave a visible effect. 

 To the nearest ly the total increase in H from 11.59 a.m. to 12.9 was 18y at Kew, 19y at Colaba, 9y at 

 Mauritius, and 9y at Christchurch. The corresponding changes in Declination (W. to west, E. to east) 

 were at Kew l'-3 W., Colaba 0'-7 E., Mauritius 0''7 W., and Christchurch 0'-2 E. The disturbance took 

 place at an hour when the regular diurnal change at Kew and Falmouth was rapid and the elimination 

 which was made of the regular diurnal change was, doubtless, only approximately correct. 



The changes in Vertical Force were at Colaba - 7y and at Mauritius + 5y. No Falmouth V curve 

 was available. At Kew no change was detected, but a small change might have been hidden by electric- 

 tram currents. At Christchurch the Vertical Force was, if anything, diminished, but if so, only to the 

 extent of about 5y. At Colaba the Vertical-Force scale was 16 2y to 1 mm., so that an error of ' 2 mm. 

 in the measurement of the ordinate means an error of 3y in force. 



The initial change of force at 11.59 a.m. is not the only one common to the different stations. A very 

 similar movement, occurring from about 7.4 to 7.11 p.m., G.M.T., is prominent in the Kew, Falmouth, 

 Colaba, and Mauritius curves, and seems recognisable also at Christchurch. This was followed by a 

 decrease in H, which is very prominent at Kew, Falmouth, and Colaba, but which at Christchurch is 

 interrupted by oscillations. We have here an excellent example of the difficulty of getting comparative 

 results from different stations. To do this satisfactorily, one requires a movement practically continuous 

 in one direction, occurring simultaneously in all the elements at all the stations. One can obtain a 

 variety of such corresponding points peaks and hollows as Dr. BALFOUR STEWART called them on the 

 Kew and Colaba curves, but when it comes to bringing in Mauritius and Christchurch it is comparatively 

 seldom that turning-points common to all can be found. In the present instance only three movements 

 could be found at all suitable for comparison. These answered to the intervals 11.59 a.m. to 12.9, 

 2.10 to 2.28 p.m., and 7.4 to 7.11 p.m., and even in their case one could not feel absolutely certain that 

 the movements in all the elements at all the stations were really synchronous. 



The magnetic storm illustrated in Plate XIV has received considerable prior attention owing to the 

 fact that its commencement began within 4 or 5 minutes, possibly less, of the beginning of the eruption of 

 Mont Pele"e, which devastated St. Pierre in Martinique. This fact was pointed out by Dr. BAUER,* 

 dealing with magnetic curves from American observatories. 



The shortness of the interval between the commencement of the magnetic storm and the eruption not 

 to speak of the character of the movements showed that it was not a case of mechanical shock due to 

 seismic waves, and Dr. BAUER'S remarks suggest the possibility of a true magnetic action at a distance 

 due directly to the eruption. This was, I think, only a suggestion. Dr. BAUER'S final opinion on the 

 subject I have not seen. My own impression is that the coincidence was merely accidental. The reasons 

 for this view will appear later (100). 



58. Plate XV shows Horizontal-Force movements referring to three different disturbances, those of 

 August 20, November 6, and November 24, 1902. 



The storm of August 20 was ushered in, like that of May 8, by a sudden movement. This commenced 

 at 9.6 p.m., G.M.T., and terminated practically in 4 minutes. During this time H increased by 14y at Kew, 

 lOy at Colaba, 6y at Mauritius, and 4y at Christchurch. The changes in the other elements were mostly 

 almost too small to measure, being in D 0' 6 W. at Kew, 0' 2 E. at Colaba, '0 at Mauritius, and 0' 6 AV. at 

 Christchurch. In V there was an increase at Kew of about ly, and a fall at Colaba of under 2y; no change 

 could be detected at Mauritius or Christchurch. The movements between 9.6 and 9.10 p.m. at the 

 several observatories appeared synchronous, and there was no visible arrest of the movement corresponding 

 to that seen in the Kew and Falmouth curves of May 8. 



* ' Terrestrial Magnetism,' vol. 7, L902, p. 67. 



x 2 



