172 A. TANAKADATE. 



Bauer gives the maximum disturbance of the vertical field 

 near Shetland Islands to be .1061 C.G.S. units upward at the 

 point, Longitude 0°, Latitude 60° N. Either of the values now 

 found come much short of that amount. 



Were the data and the assumption sufficiently reliable, we 

 might have proceeded to the second and further approximations 

 by taking account of the disturbing forces in the observed values 

 of X,Y,Z. For the present, we content ourselves with the result 

 that the variational anomalies in Japan show an excessive distribu- 

 tion of negative magnetism above the mean value in the east of 

 Asia, and those in Austria-Hungary and Great Britain that of 

 positive magnetism in north west of Europe, similar to those shown 

 by the higher harmonics in Gaussian expansion. 



Strictly speaking, the mode of distribution of the magnetism 

 must remain perfectly indeterminate as long as we adhere solely 

 to the observed elements, there being an infinite variety of ways 

 to fulfil those values ; and inasmuch as the surface integral of the 

 force over the earth vanishes, the so called seat of action may 

 be placed either inside or outside. 



It is often erroneously believed that the expansibility of the 

 earth magnetic potential in negative powers of the radius vector 

 is a proof that the source of action is inside the earth. To say 

 nothing of the possible magnetization of the surrounding medium, 

 such is no proof even from the pure theory of action at a distance. 

 To turn the subject into hydrokinetic analogy, a circulation 

 very similar to the lines of force in the earth magnetism can be 

 produced in frictionless liquid of infinite extent, by initial pressure 

 applied over any singly connected surface of finite or infinite 

 extent with a circular Hole. The imitation can be made closer by 

 taking several such surfaces instead of but one. Imagine a spherical 



