ON CDRVKS Of DKCLINATION MAGNETOGUAPHS, ETC. 207 



the changes in the horizontal force and the differences in the declination 

 curves at St. Petersburg and at Kew. 



At 11.45 a.m. on March 18 there is a sudden kick to the westward, 

 lasting for about t-\vo minutes and measured by a length of 2 millimetres 

 on the Kew curve, i.e. giving a deflection of about 2'. This kick takes 

 place simultaneously at St. Petersburg and at Vienna, and is nearly equal 

 at all the stations. It is also felt at the same instant at Coimbra and at 

 Lisbon. 



A similar kick, but less marked at St. Petersburg, occurs next day at 

 11.30 a.m. (Greenwicli time) at all the stations. 



After an entire agreement between the curves through the day, at 10 

 p.m. a disturbance occurs wliicb deflects the needle eastward at Kew and 

 westward at St. Petersburg, but by midnight the cui-ves coincide again, 

 and remain coincident with the same very small variations through the 

 night. 



Between 3 and 4 p.m. on March 20 we get disturbances opposite ways, 

 first westward at Kew and eastward at St. Petersburg simultaneously, 

 again followed by coincidences through the day. 



Another disturbance commenced by a tremulous motion of the mao-net 

 about 7 a.m. on the 23rd, and lasted until 11 p.m. ° 



From the beginning of this storm until 1.45 p.m. the several east and 

 west disturbances or oscillations of the needle are simultaneous and of the 

 same character, and are verj nearly equal in amount at Kew, Stonyhurst, 

 and at St. Petersburg. From 1.45 to 2.30 p.m. the deflections to the 

 eastward were far greater at St. Petersburg than at the other stations, 

 but were still simultaneous at all the stations. The record at Stonyhurst 

 shows that the vertical force increased in value about 2 p.m., so that here 

 an increase in the vertical force is accompanied by greater eastward deflec- 

 tions at St. Petersburg. 



The St. Petersburg curve remains below the Kew and Stonyhurst 

 curves, with the same smaller disturbances, until 7.12 p.m., just 

 after one but before another violent disturbance, each of which 'lasted 

 half an hour. The first of these two violent disturbances was first east- 

 ward and then westward at all stations, but greater at St. Petersburg 

 than at Kew, and was accompanied by a con-esponding decrease, and 

 then an increase of the horizontal force. At 7.25 p.m., accordino- to 

 the Stonyhm-st record, the V.F. had diminished to its mean value, 'and 

 simultaneously with this diminution the horizontal force had been in- 

 creasing. The second violent disturbance was westward at St. Peters- 

 burg, and eastward at Kew and Stonyhurst. This second disturbance 

 was also westward at Vienna, but less violent in character. The maxi- 

 mum was reached at 7.30 p.m. 



The simultaneous disturbances become alike again in character and 

 direction at 7.50 p.m., but from 8.15 p.m. until 11 p.m. (the end of the 

 storm) the disturbances at Kew and at St. Petersburg do not correspond, 

 but are at times in opposite directions. From 11 p.m. the curves are 

 again agreeing with one another. 



The time scales for different stations are nearly but not quite the 

 same ; the St. Petersburg is slightly shorter than the Kew scale, and the 

 Kew is slightly sliorter than the Vienna scale. They are so nearly equal 

 that for short lengths the difference is not perceptible. In Plate VIII., 

 fig. 1, where tlie disturbances daring seventeen hours on March 23-24 are 

 represented ia one diagram, an attempt has been made to guide the eye by 



