THE DISTURBANCE OF ISOMAGNETICS. 



171 



The mean values of these elements are the arithmetical mean of 

 all the observed values, except those for the declination, which are 

 obtained in a different way. The curves given in PI. XA'Iand XVII 

 were traced upon tracing cloth, cut out and weighed. From the 

 ratio of their weights to that of a rectangular piece of the same cloth 

 covering 24: hours the mean value was easily deduced. This pro- 

 cedure w^e found more practical than that with Amsler's planimeter. 



Through the kindness of Mr. K. Xakamura of the Central 

 Meteorological Observatory, the pliotographic records of the magnetic 

 elements at Tokyo during the present survey were placed at our dis- 

 posal. The inspection of these records showed us that the magnetic 

 weather during the survey wan generally fair, except on three days 

 on which we were taking observations at Obama, Shioya, and 

 Isshiki. The disturbances do not appear to be very serious, and 

 as our curves of declination obtained in tliose places are not far 

 from being regular, wc take the mean value without applying any 

 c jrrection. 



During the course of the survey, we felt two earthquakes, one at 

 Nagoya and the other at Shioya. The former was not very severe 

 but was accompanied by the usual rumbling, the latter though not ac- 

 companied by sound was a much stronger one. lîoth of them hap- 

 pened while we were observing declination, and we took observations 

 of declination immediately the shock wan over. PI. XYI shows that 

 the declination suffered a distinct fall at Nagoya but resumed its 

 regular course after an interval of some minutes. In Shioya the fall 

 was not so marked as in the case of Xagoya, especially as the shock 

 happened in that part of the day in which the declination was rapidly 

 falling. AVhether this is due to chanize in the direction of the earth's 

 field, or to distortion of azimuth on the part of the instrument, or in 

 the ground on wliich it stood, we can not tell. In the case of tlie 



