MAGNETIC SLKVEY UF JAi'AN. '20S 



From this we obtain the tbllovviDg values for the inclination of the 

 Line of Eqnnl Horizontal Force to the direction of geographical north, 

 and for the greatest rate of change «^f Horiz<jntal Ftnve ])er kilometre : 



u = 103° 14'.9 



r = .000034 



III. — The Total Force. 



Tile Total Forces were calculated from liic Horizontal Forces and 

 Dips ; and the values at the fifty selected stations, when combined 

 by the method of least squares, gave the following expres.-iion for the 

 Total Force (F) in terms of the co-ordinates : — 



F = .46407 + .000094 9 - .000045 X 



F is measured in absolute C. G. S. electromagnetic units : and 

 9 and X, the usual geographical co-ordinates referred to tlie Mean 

 Station (36° 30' X. Fat., 137° 0' E. Long.), are measured in mi- 

 nutes of arc. 



Table IV. is constructed after the same fashion a>. Tables II. 

 and III. 



If we take all the 81 Stations, the mean probable error of a sin- 

 gle observation comes out ± .00128. If we neglect Xos. 11, 18, and 

 38, which are conspicuous l)y tlie hugeness of their differences the 

 mean probable error is at once reduced to ± .00099. 



If we take into account only the fifty selected Stations, the mean 

 probable error is ± .00089. 



It will be noticed that, out of the six Station.s ( Xos. 5, 35, 43. 

 68, 1'l. 73) which were conspicuous in Tables IL and III. by the mag- 

 nitudes of the Differences, only one is so distinguished in Table W . 

 With the exce})tion of Kashiwazaki (Xo. 35), these Stations just enum- 

 erated are rather to be distino-uished bv the smallness of the differ- 



