90 Lt. Hl. Bigelow—The Earth a Magnetie Shell. 
From an inspection of the diagram of the observed vectors 
it is clear that we have to deal with each system, Inflected and 
Exflected, in turn, but the following fact confronts us at the 
outset. If w be computed for the polar stations from the 
formula aS the value of »=0°70. Now this would imply a 
physical condition of the material of the earth which is highly 
improbable, if » is taken as the magnetization constant of the 
substance of the interior of the earth. For in all known sub- 
stances except bismuth, w is greater than 1, and this differs 
very slightly from 1. Hence we must interpret the polar 
vectors as stream lines flowing around an obstacle in the 
interior of the earth. In other words, the outer stratum of the 
earth is permeable to the external magnetic forces, while the 
nucleus is not, that is to say, the earth is a magnetic shell. 
The diameter of the central nucleus can be approximately 
determined as follows: If uw is to be taken greater than 1, its 
value from the tropical stationsis 2. The value of » for nearly 
all dielectrics is 2, and is adopted for this computation. To 
obtain H, I have taken the sum of the vectors from the pole of 
the equator, as given by the boundary drawn through the ends 
of the vectors, and distributed it uniformly, as if undisturbed 
by the earth. The result is 0°:00035 C.G.S. Substituting 
these values in the formula for exflected forces 
N will have values from 0 to 8°8 and the radius of the internal 
nucleus is 0°794 R. 
For the tropical vectors yu is 2, and the formula is 
O50Y) says 
a ta AS f ae 
35. 7 
N ranges to 26°83 in this case, and the value of the nucleus is 
0-794 R. The true value of the radius of the nucleus of the 
earth is about 3170 miles, of the shell 790 miles. 
The difficulty with the problem is this: the formule are 
deduced on the supposition that the lines of force inside the 
surface of the sphere pass parallel to the axis of w In the 
case of a shell we do not know exactly the internal paths, and 
hence are at the disadvantage of having either very complex 
formule for practical solutions, or else of resorting to less 
rigorous methods. It will be proper, whenever the magnetic 
observations are sufficiently numerous and accurate to justify 
